text
stringlengths 87
880k
| pmid
stringlengths 1
8
| accession_id
stringlengths 9
10
| license
stringclasses 2
values | last_updated
stringlengths 19
19
| retracted
stringclasses 2
values | citation
stringlengths 22
94
| decoded_as
stringclasses 2
values | journal
stringlengths 3
48
| year
int32 1.95k
2.02k
| doi
stringlengths 3
61
| oa_subset
stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|
==== Front
BMC Evol BiolBMC Evolutionary Biology1471-2148BioMed Central London 1471-2148-5-331591345910.1186/1471-2148-5-33Research ArticleDo orthologous gene phylogenies really support tree-thinking? Bapteste E [email protected] E [email protected] J [email protected] D [email protected] RL [email protected] WF [email protected] GenomeAtlantic, 1721 Lower Water Street, Suite 401, Halifax, NS, B3J 1S5, Canada2 Dalhousie University, Department of Biochemistry & Molecular Biology, 5850 College St., Halifax, NS, B3H 1X5, Canada3 Dalhousie University, Department of Mathematics and Statistics, Halifax, Nova Scotia, Canada2005 24 5 2005 5 33 33 1 4 2005 24 5 2005 Copyright © 2005 Bapteste et al; licensee BioMed Central Ltd.2005Bapteste et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Since Darwin's Origin of Species, reconstructing the Tree of Life has been a goal of evolutionists, and tree-thinking has become a major concept of evolutionary biology. Practically, building the Tree of Life has proven to be tedious. Too few morphological characters are useful for conducting conclusive phylogenetic analyses at the highest taxonomic level. Consequently, molecular sequences (genes, proteins, and genomes) likely constitute the only useful characters for constructing a phylogeny of all life. For this reason, tree-makers expect a lot from gene comparisons. The simultaneous study of the largest number of molecular markers possible is sometimes considered to be one of the best solutions in reconstructing the genealogy of organisms. This conclusion is a direct consequence of tree-thinking: if gene inheritance conforms to a tree-like model of evolution, sampling more of these molecules will provide enough phylogenetic signal to build the Tree of Life. The selection of congruent markers is thus a fundamental step in simultaneous analysis of many genes.
Results
Heat map analyses were used to investigate the congruence of orthologues in four datasets (archaeal, bacterial, eukaryotic and alpha-proteobacterial). We conclude that we simply cannot determine if a large portion of the genes have a common history. In addition, none of these datasets can be considered free of lateral gene transfer.
Conclusion
Our phylogenetic analyses do not support tree-thinking. These results have important conceptual and practical implications. We argue that representations other than a tree should be investigated in this case because a non-critical concatenation of markers could be highly misleading.
==== Body
Background
Tree-thinking, the explanation of evolutionary events in the context of a tree, has inspired many philosophers and evolutionists [1]. Some tree-thinkers classically employed this pattern, labelling it the "organismal tree," and arguing that it depicts the dividing pattern of cells, the path of the envelope division of living beings through time [2] (if no cell-cell fusion occurs [3]). Other authors, even if they have retained this drawing to describe evolution, have redefined the meaning of the tree as a "prevailing trend in the evolution of genome-scale gene sets rather than as a complete picture of evolution" [4]. In any case, the reconstruction of the vertical history is decisive and relies on defining sets of congruent characters [5-7]. At the morphological level, such comparable characters are hardly identifiable. In prokaryotes, it is only with the advance of molecular phylogenetics that classification has experienced a hopeful yet limited rebound [8]. For a tree-thinker, the use of orthologue genes could unite practical and conceptual advantages. They allow us to describe the organism from the molecules, because they fit perfectly within the traditional approach of molecular phylogenetics for which the history of genes tells the history of species. They provide a vast quantity of comparable characters, and since they have been inherited from ancestor to descendants, they should likely be congruent, retracing the history of species diversification [9]. Such ideal markers are needed to reconstruct a convincing phylogenetic tree, if the tree is the right model for representing evolution.
In practice, the identification of congruent genes is mostly based on exclusion of potentially incongruent markers (i.e., paralogues, xenologues). Only broadly distributed orthologues are generally retained, if their individual phylogenies do not support apparently odd relationships [7,9-11]. The set of candidate congruent markers is sometimes further tested. Genes are concatenated to maximise the phylogenetic signal they contain, and a best tree is inferred from this large dataset. Statistical approaches are then used to test whether individual markers reject this best tree. If they fail to reject it, genes are claimed to be congruent with it. If some genes reject it, they are secondarily excluded from the core. This process is repeated until the dataset stabilizes.
Generally, these successive conditions allow the retention of a small minority of the genes present in an initial set of genomes [7]. The quantity of molecular information in these genes might thus be critical in resolving ancient phylogenetic relationships. In this context, simultaneous analyses are seen as the logical solution to produce a robust tree:supertrees [12-14] or a posteriori consensus approaches [15] can be employed. Supertree methods assemble an input set of separate phylogenetic trees with shared taxa into a larger tree [13,16] (or several trees). By fitting variously supported clades together, they allow large phylogenies based on different characters to be constructed rapidly and have been applied to a broad range of species [17]. Consensus approaches, such as sequences concatenation [5,18] or by averaging over a large number of genes [19,20] produce resolved phylogenies by overwhelming noise with signal that is presumed to be systematically congruent and historically true, though weak.
These approaches, aiming to produce a tree-like pattern whether the tree is the right model for representing evolution or not, are derived from a tree-thinking perspective. This could, however, be flawed, and deserves criticism on conceptual and statistical grounds. First, some genes have been shown not to follow a simple model of inheritance. For instance, lateral transfers of genetic material are common in nature. All living systems from viruses [21] to eukaryotes [22] can participate in the transfer of genetic material. They occur within domains of life, but also across domains, for different markers. There is now broad general agreement that lateral gene transfer (LGT) is a major force in the evolution of prokaryotes [23-27]. Additional evidence suggests that gene transfer might also be an important evolutionary mechanism in protist evolution. Andersson et al. [28] recently reported that alanyl-tRNA synthetase had been transferred from Nanoarchaeota to Diplomonads and Parabasalids. The same authors [29] showed that LGT has affected both eukaryotes and prokaryotes with respect to glutamate dehydrogenase. Recombination is also an issue for tree-reconstruction. Software such as Splitstree [30] or T-Rex [31] were developed to acknowledge this. A tree-thinker may choose to ignore conflicting signal as if it was noise, even if legitimate evolutionary events underlie it [18]. However, if this "noise" is in fact bona fide phylogenetic signal, then maybe tree-thinking is inappropriate.
The failure of individual markers to reject a concatenation-tree [7] is not a real test that genes are congruent. There are many reasons why a single gene can fail to reject a tree issued from a concatenation, many of which do not imply that these genes are effectively in favour of this single history [32]. Briefly, the best tree of a concatenation, being an average of the weak signals and noise in many genes can be a central tendency with very low discriminative power. The fact that every gene "agrees" with such a tree does not mean the concatenation tree is true, just that it reflects a part of the signal/noise of every gene. Importantly, this apparent "agreement" is also expected if individual markers contain very little phylogenetic signal. A weak marker would fail to reject most of the test trees, not only the concatenation tree.
The relative weakness of individual markers can be tested statistically when considering not only the concatenation tree but also many different trees. If several different topologies cannot be rejected by a given gene, then, unfortunately, its phylogeny does not tell us much about its actual history. For this reason, some analyses of congruence use multiple alternative test topologies [11,33]. Such analyses describe each gene by a list of likelihood or p-values associated with a set of given topologies. These lists are summarized in a large matrix of genes and topologies, which is subsequently treated by clustering methods. For instance, in principal component analyses (PCA), each gene is represented as a point in a two-dimensional projection of its position in n-dimensional space, the coordinates for each gene in that space being related to its degree of support for each of the n tree topologies tested. Thus, genes supporting and rejecting the same sets of trees should group together, constituting a cloud, while genes with atypical support/rejection patterns of topologies should be displayed elsewhere on the PCA [10]. Most of the time, PCAs produce a central cloud containing most of the markers. From this, authors generally conclude that the markers in a cloud are congruent [10,11]. However, again, a cloud of genes in a PCA may have various explanations, which differ from each gene in a cloud supporting the same history. Genes can still be lacking signal. The set of topologies might be too restricted or biased to allow discrimination between genes: if the vast majority of the topologies are very unlikely, none of them will ever be favoured, and differential clustering is unexpected. Yet, in no case would common rejection of unlikely topologies assess that the markers are congruent. Finally, even genes with different robust signal (i.e., due to recent LGT) might cluster together on a PCA, if the set of test-topologies does not allow us to identify this relationship.
Interestingly, these features can be investigated more explicitly by an alternative statistical approach: the heat map [34]. We have thus decided to re-explore the congruence in some datasets of orthologues with this method. Briefly, heat maps (HMs) generate graphs through hierarchical or partitional clustering. They allow the simultaneous display and clustering of all combinations of genes and test conditions together [35-37]. Thus genes that have the most similar responses to topologies, and topologies that are the most similar in terms of the responses they evoke from genes, can be independently identified. More precisely, when applied to phylogenetics, "responses" are p-values for each set of genes, given those topologies. Clustering of genes allows identification of one or more set of genes that might share a common evolutionary history. Clustering of topologies allows us to identify which trees are equally or nearly equally supported, and thus to assess how many distinct "best trees" there might be for a dataset of genes.
In this paper, we investigate the phylogenetic signal of four datasets in order to address a simple question: do the phylogenies of orthologs really favour tree-thinking and thus justify attempts of tree-reconstruction? Can we be reasonably confident that their history is free of LGT? We observe that no unique common history can be established for these genes. In all cases, genes fail to favour a single tree. We also observe that some of these genes support incongruent histories. Consequently, the tree-thinking on which gene concatenations rest does not proceed from phylogenetic conclusions, nor is it a priori a safe phylogenetic practice. We argue that using only the robustly resolved parts of individual phylogenies without necessarily expecting a tree as a result is likely more appropriate.
Results and discussion
Testing the phylogenetic information of datasets of orthologues
We used heat map analyses to investigate both the congruence and the absence of LGT in four selections of orthologues, two features that would be in favour of the reconstruction of an organismal tree, recently challenged by multiple analyses of comparative genomics. Indeed, it was showed that gene gains and gene losses contributed to the evolution of a substantial majority of orthologous sets of prokaryotic proteins [38-40]. Such results suggested that the simple notion of a single Tree of Life that would accurately and definitively depict the evolution of all life forms was gone forever [4]. Wolf et al. [4] concluded that the concept of a tree could only be rescued by weakening its meaning, and considering it only as "a central trend in the rich patchwork of evolutionary history, replete with gene loss and horizontal transfer events".
To test whether the reconstruction of any organismal tree was then phylogenetically justified when simultaneously using multiple orthologues, our heat maps contain two kinds of markers: artificial ones, with up to three simulated LGT events, and actual ones. A red rectangle at the left of the heat map identifies the actual markers, while a blue rectangle indicates the artificial markers (Figure 1). For each dataset, a set of plausible topologies was selected from the study of the phylogenetic signal of the markers. These plausible topologies correspond to the trees supported by a large majority of the markers. The support was estimated as the p-values from the AU test. This support is displayed in the heat map through a colour code. Lighter colours indicate a higher probability of the data given the tree (that is, stronger support) while greener colours indicate lower probabilities (stronger rejection). Heat maps were also double-clustered to group genes with similar pattern of support/rejection along columns, and to group topologies receiving similar support/rejection along rows. These hierarchical clusters are represented by a tree of genes and a tree of topologies along the heat map. Hence, to know which and how many topologies a given gene supports, one simply needs to look along its corresponding column in the heat map. If a gene is very discriminatory and favours only a few topologies, the column will be mostly green. In contrast, a gene with a weak phylogenetic signal is unable to decide between multiple topologies and its column is mostly light-coloured.
Figure 1 Figure 1A. displays the heat map for the archaeal dataset, Figure 1B. for the eukaryotic dataset. Heat maps include two kinds of markers: actual ones, indicated by a red rectangle at the left of the heat map, and artificial markers with extreme LGT (see main text), indicated in blue. They are based on a set of plausible topologies (see main text). The number of genes and topologies in the analysis are indicated on the heat map. These heat maps are double-clustered by genes and by topologies. The hierarchical clusters are represented by a tree of genes and a tree of topologies along the heat map. In the left band, the relative distribution of red and blue rectangles reflects the presence/absence of clustering of actual markers with artificial ones. Inside a heat map each dot of colour corresponds to the p-value for a given gene and a given topology. The p-values range from 0 (rejection) to 1 (support). The colour code associated with these p-values (from green for rejection to white for support) are reported above the heatmap. On the right of each heat map, the orange brackets indicate regions containing markers with a weak discriminatory power; the green brackets indicate regions containing markers with a stronger discriminatory power. Amongst the markers with a stronger phylogenetic signal, pink arrows point to some instances of conflicting signal in actual markers. They indicate different columns displaying a contrasting pattern of colour and contradictory p-values for several orthologues in a dataset.
We feel that our heat maps challenge the use of a tree-like pattern to describe molecular evolution. There was always more than one plausible topology retained (see Additional file 1 for a description of the diversity of these plausible topologies). Archaeal and eukaryotic markers favoured 60 and 92 topologies, respectively (Figures 1A and 1B), and alpha-proteobacterial markers (Figure 2A) favoured 12 different trees. Among those best topologies, none are supported by all the genes. Instead, a given topology is accepted by some markers and rejected by others, leading to multiple multicolour lines in mosaic heat maps. The absence of an entirely light line means that genes fail to agree on a single topology, even though they reject many of them and thus do contain some phylogenetic signal. This seems compatible with the redefined "weak" view of the tree. By contrast, the bacterial markers were apparently more discriminating and retained two plausible trees only (Figure 2B), one of which was supported by most of the actual markers. Furthermore, these two topologies are compatible. Could these results support an organismal tree? In fact, these trees consist of two star-phylogenies, which differ only in their ability to recover the monophyly of proteobacteria, and do not resolve any deep relationships between accepted monophyletic groups otherwise. The absence of basal resolution leaves all possibilities concerning the process of molecular evolution in bacteria open. Such a star-tree can be explained either by multiple ancient LGT events, a radiation, or the lack of ancient phylogenetic signal.
Figure 2 Figure 2A. displays the heat map for the alphaproteobacterial dataset and figure 2B for the bacterial dataset. See the legend of Figure 1 for details.
Interestingly, even though many positive controls for LGT are easily identified by the heat map by their propensity to reject the plausible topologies, the discrimination between actual markers and artificial ones is far from absolute. Some artificial genes cannot be distinguished from groups of actual markers (Figures 1B and 2A). In all the heat maps, a significant proportion of weakly discriminating genes is present. We do not know how vertical their phylogenetic history is, because not only do they agree with many different trees, but sometimes they also cluster with artificial markers. Moreover, some actual genes behave as groups of markers with transfers (see Figures 1A and 2B, for instance) and constitute common clusters of markers rejecting most of the plausible topologies. It is tempting to suggest that these genes may have undergone LGT. This would be the case for instance for rpl37ae, rpl15e in archaea (Figure 1A) or fmt in bacteria (Figure 2B).
Finally, independently of the positive controls, actual markers with a strong phylogenetic signal do not necessarily agree. We indicated by pink arrows some instances where conflicting patterns of colour are observed. Every heat map presents several such cases. These disagreements between orthologues and also sometimes the rejection of all plausible trees cannot be taken as evidence that there is a unique true tree for all these genes. We will investivage these cases further in the near future. The message of these heat map analyses is rather that we do not know if orthologues of these datasets share a unique history or not, and there is reason to suspect that they might not.
Attempted departure from tree-thinking
Let's then forget the tree-pattern and briefly consider one instance of what a different but cautious phylogenetic method could teach us. We further explored the dataset of alpha-proteobacteria, for which we had concluded that the presence of LGT could not be rejected nor could the existence of a unique history be proven. We freed ourselves from a priori tree-drawing constraints and, with the simple goal of summarizing the safe phylogenetic information of each marker, we obtained a graph that is not a tree. In this synthesis (Figure 3), 25 vertical branches are visible as well as 4 lateral branches. The comparison of the total support for the horizontal and vertical branches indicates that the vertical signal is about 15 times more important than the horizontal signal. A large majority of the genes in this dataset (30/34) does not seem to have been laterally transferred, and many parts of the inferred vertical backbone are relatively robust. However, only node 3A, the clade of Rickettsiales, is supported by all the 34 gene trees. In all the other cases, we would not be right to claim that the vertical backbone corresponds to the common history for all or even most of the markers. For instance, 21 genes do not tell anything about node 2A. Why should we then assume that these 21 markers were subjected to this pattern of vertical inheritance? Certainly, phylogeny alone does not tell us that, and the synthesis shows us clearly that we simply do not know what the history of most genes is, for most of the nodes. Finally, there are still 4 genes (rps19, rps10, rpl14 and rpoB) that have likely undergone LGT. LGT between these species occurred only once, thus no generality can be inferred from them, except that 75% of them correspond to local rearrangements of the concatenation tree. More precisely, rps19 was transmitted from Ehrlichia to Neorickettsia, rpl14 was transmitted from Mesorhizobium to Agrobacterium and rps10 of Rhodobacter apparently comes laterally from a species not studied in this dataset. The origin of the rpoB of Rhizobiales is, however, more complex and it cannot be mapped directly onto the reference tree, since multiple donors are possible.
Figure 3 Figure 3 displays the synthesis of 34 alphaproteobacterial genes (atp1, atp6, atp9, cob, cox2, cox3, nad1, nad2, nad3, nad4, nad4l, nad5, nad6, nad7, nad8, nad11, rpl2, rpl5, rpl6, rpl11, rpl14, rpl16, rpoA, rpoB, rpoC, rps7, rps10, rps12, rps13, rps14, rps19, sdh2, sdh3 and tufA). The proposed vertical-inheritance backbone representing the concatenation tree is shown in dark blue, with the line thickness of an internal branch corresponding to the frequency of its support across the whole dataset. Support was considered significant when clades received > 50% bootstrap support. Putative LGT events are in orange, connecting donors (circles) with recipients (arrowheads); where there are multiple possible donor candidates, these converge onto a double arrowhead. This happens when the clade founded by a past LGT donor may have subsequently had its species membership obfuscated by later exchanges of genetic material, yielding a non-reference assemblage of species labels in a presumed lineage. Where the apparent donor of a gene falls outside of the taxa included in the analysis, one is created as a basal group taxon, indicated in light blue. In order to avoid graphical congestion, branches in the tree may be artificially extended, as dotted segments.
Conclusion
Heat map analyses are powerful statistical methods to investigate correlations in multiple markers. Certainly, any conclusion deduced from their study depends on the set of topologies, genes and p-values investigated. However, this cannot be seen as a rebuttal in itself to reject this approach. In fact, the same issue arises for any phylogenetic and statistical analyses comparing trees. Importantly, some propositions could be extrapolated from these heat maps applied to the three domains of Life. First, we observed several cases where it was impossible to separate markers with an extremely atypical phylogenetic signal from actual markers. More than a weakness of the method (PCA does the same, data not shown), this might be explained by a relative weakness of the phylogenetic signal contained in many markers. What we report then is simple and sadly not surprising: the genes of eukaryotes and alpha-proteobacteria, for instance, cannot really discriminate between several different topologies. This absence of convergence on a single topology is obviously not evidence for LGT in itself. It is, however, a major issue, since it indicates that with phylogenetic analyses and statistical tools only, we often cannot decide if LGT is present in datasets of orthologues. The broad bacterial and archaeal datasets seem principally free from such extreme recent events. Indeed, in their heat maps, the majority of genes with transfer can be separated from actual markers. There are, however, some instances of LGT in these heatmaps too, since they present clusters of genes with LGT rejecting all the plausible topologies in which some actual markers are also found. Hence, overall, there is no strong phylogenetic evidence that any of these datasets are really comprised of congruent genes.
This could be problematic because phylogeneticists, raised as classical tree-thinkers, often desire a tree or a supertree. We feel that they might be prone to forget/accept the fundamental weaknesses of the markers they use to reconstruct the past. It could matter because if these analyses mix together markers with arguably different histories, the phylogenetic Tree of Life will be simply a phenetic tree, a "Trend of Life," averaging noise, signal, and different histories of markers to fit an a priori pattern. In other words, a phylogeneticist who would assume that he had reconstructed the organismal tree from orthologues and produced a genealogy of organisms instead of a central tendency might be a victim of an extreme version of tree-thinking. Yet, no phylogeneticist has to be an extreme tree-thinker anymore, because there is no phylogenetic evidence for that. Consequently, we see the present conclusion as a positive one. In fact, this work should encourage attempts to explore more accurately the phylogeny of organisms. We propose that a safer, more punctual use of the phylogenetic signal of orthologues could be envisioned. On one hand, "whereof one cannot speak thereof one must be silent" applies [42], while on the other hand we cautiously and resolutely report all the phylogenetic information that we can. Acknowledging that a strong version of tree-thinking has still to be proven and should not be assumed a priori, and that a weak version could be refined, we could reduce the risk of building a hazardous evolutionary history from the largely unknown phylogenetic signal of orthologues. This acknowledgement would also allow us to maximise the number of genes available for phylogenetic analysis, instead of limiting cautious simultaneous analyses to a few congruent markers. It may not produce a tree in the end, but would surely be more grounded.
Methods
Alignments and preliminary phylogenetic analyses
One eukaryotic dataset and three prokaryotic datasets (archaea, bacteria, and alpha-proteobacteria) were investigated. The eukaryotic dataset (34 genes:a-rad51, c-psma, d-rpl12e, e-ef2, rpl10a, rpl10b, rpl11b, rpl13a, rpl15e, rpl19e, rpl28e, rpl37a, rpl1, rpl10, rpl17, rpl18, rpl2, rpl26, rpl27, rpl3, rpl30, rpl9, rps11, rps15, rps16, rps19, rps20, rps23, rps4, rps8, rps15p, rps27e, sap40, srs, 17 species:S. cerevisiae, S. pombe, E. cuniculi,,G. theta nucleomorph, P. yezoensis, A. thaliana, O. sativa, C. reinhardtii, D. melanogaster, H. sapiens, C. elegans, D. discoideum, E. histolytica, P falciparum, T. gondii, T. pyriformis and P. infestans) is a subset from Bapteste et al. [5] The archaeal dataset (44 informational genes: rpl10, rpl14p, rpl15e, rpl15p, rpl18e, rpl18p, rpl19e, rpl21e, rpl22p, rpl23p, rpl24e, rpl24p, rpl2p, rpl30p, rpl31e, rpl32e, rpl37ae, rpl3p, rpl40e, rpl44e, rpl4p, rpl5p, rpl6p, rpl7ae, rps10p, rps11p, rps12p, rps13, rps14p, rps15p, rps17e, rps17p, rps19e, rps19p, rps24e, rps2p, rps3p, rps4e, rps4p, rps5p, rps6e, rps7p, rps8e, rps8p, 18 species:S. solfataricus, A. pernix, P. aerophilum, M. kandleri, P. abyssi, P. horikoshii, P. furiosus, T. acidophilum, T. volcanium, A. fulgidus, M. maripaludis, M. acetivorans, H. marismortui, M. thermoautotrophicus, S. tokodaii, F. acidarmanus, M. jannashii and Halobacterium sp.) corresponds to an update of Brochier et al. [6], including Nanoarchaea and some methanogenic species. The bacterial dataset (45 genes: efg, fmt, if1, if2, ksga, npt, rba, rf2, rpl1, rpl10, rpl11, rpl14, rpl15, rpl16, rpl17, rpl18, rpl19, rpl2, rpl20, rpl21, rpl23, rpl24, rpl27, rpl29, rpl3, rpl32, rpl34, rpl35, rpl4, rpl5, rpl6, rpl7, rpl9, rps11, rps12, rps13, rps2, rps20, rps3, rps4, rps5, rps6, rps7, rps8, rps9, trmd, 28 species:P. gingivalis, C. tepidum, P. marinus, D. radiodurans, B. anthracis, B. subtilis, C. difficile, S. pyogenes, M. leprae, M. tuberculosis, S. coelicolor, T. maritima, A. aeolicus, B. burgdorferi, T. pallidum, C. pneumoniae, C. trachomatis, C. jejuni, H. pylori, C. crescentus, R. capsulatus, R. prowazekii, B. pertussis, N. meningitidis, N. europaea, E. coli, P. aeruginosa and V. cholerae) corresponds to the core of genes identified in Brochier et al. [33]. The alpha-proteobacterial dataset (34 genes: see Figure 3 for their name, 13 species) corresponds to orthologous proteins shared by the mitochondria of Reclinomonas and all alpha-proteobacteria. All these alignments were inspected, manually refined if required, and are available upon request. For all individual markers, preliminary analyses by NJ using MUST.3.0 [43] and Maximum likelihood (ML) using PROML with the JTT amino acid substitution matrix, a rate heterogeneity model with gamma-distributed rates over four categories with the α parameter estimated using TREE-PUZZLE, global rearrangements and randomized input order of sequences (10 jumbles), were done to exclude potential non-orthologous copies, but no such copies were identified. For each dataset, all the genes were concatenated to calculate a best tree by ML (PROML + JTT model + 9 categories). Lengths of the concatenations were approximately 6300, 7300, 7800, 5900 for archaea, bacteria, alpha-proteobacteria, and eukaryotes, respectively. The best ML tree was calculated for each gene individually by the same methodology.
Constitution of matrices for statistical analyses
A set of topologies for each dataset was constructed to test the congruence and the phylogenetic signal between markers. They contain alternatives to the best concatenation tree for each dataset. The best ML tree issued from a concatenation of the markers in a dataset was used as a reference and rearranged by moving each species of the tree to any other possible location to simulate recent LGT inside a reference tree. The dataset also included (i) a star topology, (ii) topologies supporting only a single robust monophyletic group, (iii) topologies containing all the accepted monophyletic groups, but also showing one event of LGT. For instance, in case (iii), the clade of Ferroplasma/Thermosplasma includes a Methanosarcina that should have been located in another clade, under the hypothesis of an organismal reference tree. This approach generated a set of 868/1197/1142 and 443 test topologies for archaea, eukaryotes, bacteria and alpha-proteobacteria, respectively. These input trees are given in Additional file 2. They were used as user trees in TREE-PUZZLE5.1, option -wsl, with a JTT+Γ 8+I model of evolution to estimate the likelihood of each site of a given gene and global tree likelihoods for each tree. These two sets of likelihood values were used as input for CONSEL [44] to perform the Approximately Unbiased (AU) test [45] and associate a p-value to each set of generated trees.
To test if the actual markers behave differently from genes with LGT, datasets of markers presenting different degrees of LGT were generated as follows. We randomly assigned the sequence of one species to another one, as if the latter has just laterally acquired the sequence of the former. After this operation, a gene alignment presents one additional extreme and recent LGT event. We reiterated this up to three times per gene, generating up to three additional LGT events in a single alignment. These alignments are the positive controls for LGT. If the statistics of genes with LGT are identical to those of actual markers, LGT presence cannot be excluded.
Statistical analyses
Heat map analysis was implemented in R . Heat maps of p-values of the AU test were used to test that genes support similar topologies. A spot with a dark green colour indicates low p-values for a topology tested for a given gene. By contrast, a spot with a light colour indicates high p-values, i.e. good support for this topology by a given gene. These spots of colour can be clustered to highlight the presence of patterns of support/ rejection, by rearranging rows and columns separately for genes and topologies, so that they correspond to a dendrogram from hierarchical clustering. In this way, clusters of genes (topologies) showing similar patterns of support across topologies (genes) are grouped together and easily seen. Hierarchical clustering dendrograms were obtained using the Euclidean distance matrix for the vectors of p-values. The definition of the number of clusters will be discussed in a future paper.
To completely summarize patterns of support for topologies it would be necessary to include all possible topologies. This is impractical for the data sets here (even the set of a priori plausible topologies included makes visualization difficult). To utilize the information from tests over a large number of topologies while easing visualization of results, we present heat maps with a restricted set of "plausible" topologies for which the majority of genes had a p-value larger than 0.05. This set of plausible topologies is thus constructed under the hypothesis of interest that genes should share support for a single topology due to their common vertical descent. The set is also larger than required under the hypothesis that genes come from a single topology due to common vertical descent. Under this hypothesis, the p-values for that topology should be uniformly distributed across genes so that 95% of the genes are expected to have p-values larger than 0.05 for this topology. With probability larger than 0.95, at least 90 out of 100 genes should have p-values larger than 0.05 for the correct topology. Thus the set of plausible topologies could be restricted to the set with p-values greater than 0.05 for 80 to 90% of the genes. Since, for the datasets considered here, restriction to topologies with a majority of p-values larger than 0.05 eased visualization sufficiently, we did not make further restrictions. Note that the full set of topologies is being utilized, since with a larger set of initial topologies, a larger set of plausible topologies will be found. In principle, the initial set of topologies should be large enough that all topologies satisfying the criteria of plausibility are included.
Synthesis reconstruction
The synthesis of alpha-proteobacteria [32] was inferred from the analyses of the 34 ML trees for 13 species. ML trees were calculated as described above. Their bootstrap support values represent a consensus (obtained using CONSENSE) of 100 Fitch-Margoliash distance trees (obtained using PUZZLEBOOT and FITCH) from pseudo-replicates (obtained using SEQBOOT) of the original alignment. The settings of PUZZLEBOOT were the same as those used for PROML, except that global rearrangements and randomized input order of sequences are not available in this program. PROML, CONSENSE, FITCH and SEQBOOT are from the PHYLIP package version 3.6a . PUZZLEBLOOT can be obtained from the TREE-PUZZLE website . The clades supported with more than 50% bootstrap support in these 34 gene trees were compared to the concatenation tree of alpha-proteobacteria using two programs: Horizstory and Lumbermill [46]. These programs can be downloaded from . Briefly, Horizstory allows inference of the most parsimonious scenarios involving LGT and vertical descent to explain the common features and the discrepancies between the concatenation tree and each of the 34 gene trees. Lumbermill draws the synthesis by mapping the outcomes of these scenarios onto the reference tree. A strict consensus option was applied, meaning that only the relationships supported or inferred in 100% of the evolutionary scenarios resulting from the comparison between the reference and a given tree were considered in this drawing.
Authors' contributions
EB did the phylogenetic analyses, provided the ideas and wrote the paper, ES implemented the heat map analyses and provided ideas, JL created the alphaproteobacterial dataset, DM implemented Lumbermill, RLC implemented Horizstory, WFD provided ideas.
Supplementary Material
Additional File 1
The diversity of plausible topologies
Click here for file
Additional File 2
Input trees for the AU test
Click here for file
Acknowledgements
We thank O. Zhaxybayeva and D. Walsh for careful readings. Preliminary sequence data for Anaplasma phagocytophylum was obtained from The Institute for Genomic Research website at . EB was supported by a grant from CIHR (MOP4467). JL was supported by a Student Research Award from the Nova Scotia Health Research Foundation.
==== Refs
O'Hara RJ Population thinking and tree thinking in systematics Zoologica Scripta 1997 26 323 329
Philippe H Douady CJ Horizontal gene transfer and phylogenetics Curr Opin Microbiol 2003 6 498 505 14572543 10.1016/j.mib.2003.09.008
Rivera MC Lake JA The ring of life provides evidence for a genome fusion origin of eukaryotes Nature 2004 431 152 5 15356622 10.1038/nature02848
Wolf YI Rogozin IB Grishin NV Koonin EV Genome trees and the tree of life Trends Genet 2002 18 472 9 12175808 10.1016/S0168-9525(02)02744-0
Bapteste E Brinkmann H Lee JA Moore DV Sensen CW Gordon P Durufle L Gaasterland T Lopez P Muller M Philippe H The analysis of 100 genes supports the grouping of three highly divergent amoebae: Dictyostelium, Entamoeba, and Mastigamoeba Proc Natl Acad Sci U S A 2002 99 1414 9 11830664 10.1073/pnas.032662799
Brochier C Forterre P Gribaldo S Archaeal phylogeny based on proteins of the transcription and translation machineries: tackling the Methanopyrus kandleri paradox Genome Biol 2004 5 R17 15003120 10.1186/gb-2004-5-3-r17
Lerat E Daubin V Moran NA From Gene Trees to Organismal Phylogeny in Prokaryotes:The Case of the gamma-Proteobacteria PLoS Biol 2003 1 E19 12975657 10.1371/journal.pbio.0000019
Creevey CJ Fitzpatrick DA Philip GK Kinsella RJ O'Connell MJ Pentony MM Travers SA Wilkinson M McInerney JO Does a tree-like phylogeny only exist at the tips in the prokaryotes? Proc R Soc Lond B Biol Sci 2004 271 2551 8 10.1098/rspb.2004.2864
Zuckerkandl E Pauling L Molecules as documents of evolutionary history J Theor Biol 1965 8 357 66 5876245
Brochier C Bapteste E Moreira D Philippe H Eubacterial phylogeny based on translational apparatus proteins Trends Genet 2002 18 1 5 11750686 10.1016/S0168-9525(01)02522-7
Matte-Tailliez O Brochier C Forterre P Philippe H Archaeal phylogeny based on ribosomal proteins Mol Biol Evol 2002 5 631 9 11961097
Daubin V Gouy M Perriere G Bacterial molecular phylogeny using supertree approach Genome Inform Ser Workshop Genome Inform 2001 12 155 64 11791234
Eulenstein O Chen D Burleigh JG Fernandez-Baca D Sanderson MJ Performance of flip supertree construction with a heuristic algorithm Syst Biol 2004 53 299 308 15205054 10.1080/10635150490423719
Semple C Daniel P Hordijk W Page RD Steel M Supertree algorithms for ancestral divergence dates and nested taxa Bioinformatics 2004 20 2355 60 15073021 10.1093/bioinformatics/bth246
Lapointe FJ Cucumel G The average consensus procedure: combination of weighted trees containing identical or overlapping sets of taxa Syst Biol 1997 46 306 312
Steel M Dress AWM Bocker S Simple but fundamental limitations on supertree and consensus tree methods Syst Biol 2000 49 363 368 12118411 10.1080/10635159950173870
Thomas GH Wills MA Szekely T A supertree approach to shorebird phylogeny BMC Evol Biol 2004 4 28 15329156 10.1186/1471-2148-4-28
Baldauf SL Roger AJ Wenk-Siefert I Doolittle WF A kingdom-level phylogeny of eukaryotes based on combined protein data Science 2000 290 972 7 11062127 10.1126/science.290.5493.972
Clarke GD Beiko RG Ragan MA Charlebois RL Inferring genome trees by using a filter to eliminate phylogenetically discordant sequences and a distance matrix based on mean normalized BLASTP scores J Bacteriol 2002 184 2072 80 11914337 10.1128/JB.184.8.2072-2080.2002
Gophna U Doolittle WF Charlebois RL Weighted genome trees: refinements and applications J Bacteriol 2005 187 1305 16 15687194 10.1128/JB.187.4.1305-1316.2005
Filee J Forterre P Laurent J The role played by viruses in the evolution of their hosts: a view based on informational protein phylogenies Res Microbiol 2003 154 237 43 12798227 10.1016/S0923-2508(03)00066-4
Lawrence JG Ochman H Reconciling the many faces of lateral gene transfer Trends Microbiol 2002 10 1 4 11755071 10.1016/S0966-842X(01)02282-X
Doolittle WF Lateral genomics Trends Cell Biol 1999 9 M5 8 10611671 10.1016/S0962-8924(99)01664-5
Nelson KE Clayton RA Gill SR Gwinn ML Dodson RJ Haft DH Hickey EK Peterson JD Nelson WC Ketchum KA McDonald L Utterback TR Malek JA Linher KD Garrett MM Stewart AM Cotton MD Pratt MS Phillips CA Richardson D Heidelberg J Sutton GG Fleischmann RD Eisen JA White O Salzberg SL Smith HO Venter JC Fraser CM Evidence for lateral gene transfer between Archaea and bacteria from genome sequence of Thermotoga maritima Nature 1999 399 323 9 10360571 10.1038/20601
Garcia-Vallve S Romeu A Palau J Horizontal gene transfer in bacterial and archaeal complete genomes Genome Res 2000 10 1719 25 11076857 10.1101/gr.130000
Koonin EV Makarova KS Aravind L Horizontal gene transfer in prokaryotes: quantification and classification Annu Rev Microbiol 2001 55 709 42 11544372 10.1146/annurev.micro.55.1.709
Boucher Y Douady CJ Papke RT Walsh DA Boudreau ME Nesbo CL Case RJ Doolittle WF Lateral gene transfer and the origins of prokaryotic groups Annu Rev Genet 2003 37 283 328 14616063 10.1146/annurev.genet.37.050503.084247
Andersson JO Sarchfield SW Roger AJ Gene Transfers from Nanoarchaeota to an Ancestor of Diplomonads and Parabasalids Mol Biol Evol 2005 22 85 90 15356278 10.1093/molbev/msh254
Andersson JO Roger AJ Evolution of glutamate dehydrogenase genes: evidence for lateral gene transfer within and between prokaryotes and eukaryotes BMC Evol Biol 2003 3 14 12820901 10.1186/1471-2148-3-14
Huson DH SplitsTree: analyzing and visualizing evolutionary data Bioinformatics 1998 14 68 73 9520503 10.1093/bioinformatics/14.1.68
Makarenkov V T-REX: reconstructing and visualizing phylogenetic trees and reticulation networks Bioinformatics 2001 17 664 8 11448889 10.1093/bioinformatics/17.7.664
Bapteste E Boucher Y Leigh J Doolittle WF Phylogenetic reconstruction and lateral gene transfer Trends Microbiol 2004 12 406 11 15337161 10.1016/j.tim.2004.07.002
Brochier C Bapteste E Moreira D Philippe H Bacterial phylogeny based on translational apparatus proteins Trends Genet 2002 1 1 5 11750686 10.1016/S0168-9525(01)02522-7
Gordon A Classification 1999 (Chapman-Hall, London)
Alon U Barkai N Notterman DA Gish K Ybarra S Mack D Levine AJ Broad patterns of gene expression revealed by clustering analysis of tumor and normal colon tissues probed by oligonucleotide arrays Proc Natl Acad Sci U S A 1999 96 6745 6750 10359783 10.1073/pnas.96.12.6745
Getz G Levine E Domany E Coupled two-way clustering analysis of gene microarray data Proc Natl Acad Sci U S A 2000 97 12079 84 11035779 10.1073/pnas.210134797
Somogyi R Fuhrman S Anderson G Madill C Greller LD Chang B Modularity in Development and Evolution (ed Schlosser, G a G P W) 2004 202 221
Snel B Bork P Huynen MA Genomes in flux: the evolution of archaeal and proteobacterial gene content Genome Res 2002 12 17 25 11779827 10.1101/gr.176501
Kunin V Ouzounis CA The balance of driving forces during genome evolution in prokaryotes Genome Res 2003 13 1589 94 12840037 10.1101/gr.1092603
Nesbo CL Boucher Y Doolittle WF Defining the core of nontransferable prokaryotic genes: the euryarchaeal core J Mol Evol 2001 53 340 50 11675594 10.1007/s002390010224
Mirkin BG Fenner TI Galperin MY Koonin EV Algorithms for computing parsimonious evolutionary scenarios for genome evolution, the last universal common ancestor and dominance of horizontal gene transfer in the evolution of prokaryotes BMC Evol Biol 2003 3 2 12515582 10.1186/1471-2148-3-2
Wittgenstein L Tractatus Logico-Philosophicus (Routledge; New Ed edition) 1981 89
Philippe H MUST, a computer package of Management Utilities for Sequences and Trees Nucleic Acids Res 1993 21 5264 72 8255784
Shimodaira H Hasegawa M CONSEL: for assessing the confidence of phylogenetic tree selection Bioinformatics 2001 17 1246 7 11751242 10.1093/bioinformatics/17.12.1246
Shimodaira H An approximately unbiased test of phylogenetic tree selection Syst Biol 2002 51 492 508 12079646 10.1080/10635150290069913
MacLeod D Charlebois RL Doolittle WF Bapteste E Deduction of probable events of lateral gene transfer through comparison of phylogenetic trees by recursive consolidation and rearrangement BMC Evol Biol 2005 27
| 15913459 | PMC1156881 | CC BY | 2021-01-04 16:37:17 | no | BMC Evol Biol. 2005 May 24; 5:33 | utf-8 | BMC Evol Biol | 2,005 | 10.1186/1471-2148-5-33 | oa_comm |
==== Front
BMC Fam PractBMC Family Practice1471-2296BioMed Central London 1471-2296-6-201589007110.1186/1471-2296-6-20Research ArticleGPs' perspectives of type 2 diabetes patients' adherence to treatment: A qualitative analysis of barriers and solutions Wens Johan [email protected] Etienne [email protected] Royen Paul [email protected] Bernard [email protected] Joke [email protected] Department of Family Practice, University of Antwerp – Faculty of Medicine, Universiteitsplein 1, 2610 Wilrijk (Antwerpen), Belgium2 Department of Psychiatry, University of Antwerp – Faculty of Medicine, Universiteitsplein 1, 2610 Wilrijk (Antwerpen), Belgium2005 12 5 2005 6 20 20 13 1 2005 12 5 2005 Copyright © 2005 Wens et al; licensee BioMed Central Ltd.2005Wens et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The problem of poor compliance/adherence to prescribed treatments is very complex. Health professionals are rarely being asked how they handle the patient's (poor) therapy compliance/adherence. In this study, we examine explicitly the physicians' expectations of their diabetes patients' compliance/adherence. The objectives of our study were: (1) to elicit problems physicians encounter with type 2 diabetes patients' adherence to treatment recommendations; (2) to search for solutions and (3) to discover escape mechanisms in case of frustration.
Methods
In a descriptive qualitative study, we explored the thoughts and feelings of general practitioners (GPs) on patients' compliance/adherence. Forty interested GPs could be recruited for focus group participation. Five open ended questions were derived on the one hand from a similar qualitative study on compliance/adherence in patients living with type 2 diabetes and on the other hand from the results of a comprehensive review of recent literature on compliance/adherence. A well-trained diabetes nurse guided the GPs through the focus group sessions while an observer was attentive for non-verbal communication and interactions between participants. All focus groups were audio taped and transcribed for content analysis. Two researchers independently performed the initial coding. A first draft with results was sent to all participants for agreement on content and comprehensiveness.
Results
General practitioners experience problems with the patient's deficient knowledge and the fact they minimize the consequences of having and living with diabetes. It appears that great confidence in modern medical science does not stimulate many changes in life style. Doctors tend to be frustrated because their patients do not achieve the common Evidence Based Medicine (EBM) objectives, i.e. on health behavior and metabolic control. Relevant solutions, derived from qualitative studies, for better compliance/adherence seem to be communication, tailored and shared care. GPs felt that a structured consultation and follow-up in a multidisciplinary team might help to increase compliance/adherence. It was recognized that the GP's efforts do not always meet the patients' health expectations. This initiates GPs' frustration and leads to a paternalistic attitude, which may induce anxiety in the patient. GPs often assume that the best methods to increase compliance/adherence are shocking the patients, putting pressure on them and threatening to refer them to hospital.
Conclusion
GPs identified a number of problems with compliance/adherence and suggested solutions to improve it. GPs need communication skills to cope with patients' expectations and evidence based goals in a tailored approach to diabetes care.
==== Body
Background
Diabetes mellitus type 2 is an important and increasing health problem. In Belgium, the incidence is 231 new cases per 100.000 inhabitants per year [1]. It is frequently not diagnosed until complications appear, and approximately one-third of all people with diabetes may remain undiagnosed. The estimated prevalence of diabetes among adults was 7.4 % in 1995; this is expected to rise to about 9 % in 2025 [2]. To date there is strong evidence that vigorous treatment of diabetes type 2 can decrease the morbidity and mortality of the disease by decreasing its chronic complications [3-6]. However, poor patient compliance/adherence to these treatment recommendations can reduce therapeutic effects.
Earlier research on compliance/adherence showed that neither the features of a disease, nor the referral process, nor the clinical setting nor the therapeutic regimen seem to influence compliance/adherence [7]. Because of the difficulties in measuring, no estimate of compliance/adherence or non-compliance/non-adherence can be generalized. Poor compliance/adherence is to be expected in 30–50 % of all patients, irrespective of disease, prognosis or setting [8-10]. Today, more then 200 different doctor- patient- and encounter-related variables have been studied but none of them is consistently related to compliance or fully predictive. Especially in quantitative studies, little attention has been paid to patients' ideas about medicines and compliance/adherence. However, from qualitative research we know that the most salient influences on compliance/adherence are patients' own beliefs about medications and about medicine in general [7]. Their own knowledge, ideas and experiences, as well as those of family members and friends, have also been shown to correlate with compliance [11].
In order to understand and predict compliance/adherence, the patients' attitude towards disease has been studied since more then twenty years by means of the health belief model [12,13]. Today, new concepts of patient involvement, participation and real partnership are introduced [14]. Therapeutic interactions with patients should not longer be viewed simply as opportunities to reinforce instructions around treatment: rather, they should be seen as a space where the expertise of patients and health professionals can be pooled to arrive at mutually agreed goals [15]. In primary care, patients strongly want a patient centered approach, with communication, partnership and health promotion [16]. Evidence is increasing that involving patients more in consultations can increase compliance/adherence to treatment.
Patients and caregivers interpret signs and signals in a different way [17]. These differences in perspective are not inherently problematic. They frequently become so when patients do not meet the goals and expectations of their health care providers. In caregiver's point of view there is an expectation that when they, as authoritative experts, make recommendations, the patient only has the obligation to carry them out [18]. Furthermore, caregivers are increasingly under pressure from health policy makers or managed care organizations to reduce the costs of complications by rigorously complying with treatment guidelines.
Physicians estimate their patients' compliance/adherence to medications, and base decisions about treatment on these estimates. Misjudgment of patient compliance/adherence can have adverse consequences, including withholding of therapy or unnecessary changes in therapy. A review of the literature demonstrates that physicians are often inaccurate in estimating patient compliance/adherence with antiretroviral therapy [19]. In the field of diabetes, medical anthropologists evaluated an analytical framework for contrasting patient and provider goals and strategies [20]. This framework was designed to allow examination of how a chronic illness fits into both clinical and life-world contexts. By comparing patients' and providers' goals, strategies and evaluation criteria, a better understanding could be developed of how recommended behavioral changes are understood and applied differentially by each group. Using this approach, certain shared outward semantic similarities (i.e. "control") were found between patients' and providers' perspectives that obscured the presence of deeper differences between those perspectives. The differences had important implications for the long-term enactment of diabetes self-care regimes.
The clinical literature on "non-compliance/non-adherence" tends to problematize only the patient's perspective, treating the provider's perspective as an uncontroversial point of departure. Equal consideration is seldom given to the role of provider perspectives and the broader institutional and economic contexts of illness management [21]. Our work on compliance/adherence focusing on the providers' perspective, illustrates the difficulties Belgian GPs experience in trying to help people with diabetes type 2 being adherent to treatment plans.
Methods
This descriptive qualitative study explores the thoughts and feelings of GPs on patients' compliance/adherence. In five focus groups (FG), GPs were asked how they think their type 2 diabetes patients adhere to life style and dietary advices, medicine taking and management of the disease. We sounded out how GPs cope with conflicts concerning these topics. In our analysis, we tried to explore the barriers to compliance/adherence and possible solutions, GPs experience in diabetes management. We also explored some coping mechanisms they use to handle the conflict arising when patients do not heed their advices.
The focus group technique used was based on a standardized procedure, described in earlier publications [22-24].
After approval of the different local societies for primary care and of the ethics committee of the study centre, all GP members (n = 173) working in the surrounding municipalities of the study centre received a letter with some general information on our project. Afterwards, physicians were phoned to assess their interests in participating and subscription in one of the discussion groups planned. Forty interested physicians could be recruited and five focus group sessions were organized. In compiling the different groups, no distinctions were made on age, gender or number of years in practice in the hope of maximizing interaction and outcome. Interested physicians could make their choice between an afternoon and an evening session. They also could choose between a venue on the university campus or not. If after the five focus groups content saturation was not achieved, additional focus groups were planned.
A previous qualitative study on adherence to treatment recommendations in patients living with diabetes [25] revealed five different themes on the problem of compliance/adherence. Together with the results of a comprehensive review of recent literature on adherence [26], these themes were used to generate five open-ended core questions (table 1) exploring the ideas and thoughts of the participant GPs. The precise formulation of these questions was discussed thoroughly by the research team until mutual agreement and piloted in a first focus group.
Table 1 Focus group questions
Research questions:
FG Question 1 Introduction
• Do you as a physician have some insight in how your diabetes patients think about their illness and the treatment necessary to them?
FG Question 2 Health beliefs and advice giving
• In your treatment, you give many advices as on diet, walking, smoking cessation etc. Perhaps there is a good intention for it. However, can you tell us what you think your patients do with these advices?
• It might be possible that physicians do not understand enough the health beliefs of their patients when they give their advices. What do you think about this?
FG Question 3 Health beliefs on drug intake
• In addition, when you prescribe your medicines you have some expectations about your patients. What is your experience as to patients fulfill these expectations or not?
• Probably, patients have some alternative expectations. Do you notice that?
• It might be possible that physicians do not understand enough the health beliefs of their patients when they prescribe their medicines. What do you think about this?
FG Question 4 Health beliefs on follow up
• The guideline level on follow up of diabetes patients is not easy at all. Many examinations have to be performed; appointments have to be made to monitor the continuous changing risk profile of the diabetes patient. What do you think the diabetes patient expects from all this?
• It might be possible that physicians do not understand enough the health beliefs of their patients when they plan their follow up. What do you think about this?
FG Question 5 Coping with conflicts
• Sometimes, you certainly may not reach common grounds with your patients. How do you express these conflicts?
• How do you handle them?
• What solutions can you suggest?
A diabetes nurse, well-trained and experienced in conduction focus group discussions, moderated all focus groups, guaranteeing a standardized procedure. She observed that the level of involvement was as non-directive as possible, but meanwhile trying to collect as much data as possible, ensuring that the desired set of topics was covered and encouraging the participation of everyone [27,28].
An observer (JW) gathered information on the non-verbal communication and on the interaction between participants. In this way, special attention could be given in the analysis, to those pronouncements and text fragments where verbal and non-verbal consensus existed. The duration of the discussions was limited to two hours.
At the end of every focus group, there was a debriefing between moderator and observer to discuss the most important themes and possible differences with other focus groups. Since participants of FG 5 added no new information anymore, content saturation was supposed to be reached.
The first focus group served as a pilot. Since no indistinctness with regard to the questioning or other procedures were allocated, and since the results were very similar to the other focus group interviews we decided to keep these first results in the analyzing process. In no focus group, any striking nuisance cases were present.
All focus groups were audio taped and transcribed for content analysis.
Two researchers (JW and EV) independently performed the initial coding. In the case of disagreement, a solution was found by clarifying individually the meaning of a code and discussing until mutual consent was reached. We defined the different themes and performed cross-referencing within the different coding categories, and by text word search in the document or selected sections of it. In doing so, we clarified ideas, discovered further themes, searched for patterns and explored participants' explanatory models.
A first draft with results was sent to all participants for agreement on content and comprehensiveness (response rate 72.5 %). This participant checking leaded to only a few interpretative comments. We added or corrected most of them in the draft after discussion and agreement with the moderator.
Results
Of the 173 GPs contacted, 61 agreed to participate (35.3 %), and 40 GPs (23.1 %) attended the focus groups. Physician and practice characteristics are summarized in table 2.
Table 2 physician and practice characteristics
FG 1 FG 2 FG 3 FG 4 FG 5 Total
N 10 7 9 8 6 40
♂ / ♀ 7 / 3 4 / 3 6 / 3 4 / 4 5 / 1 26 / 14
Mean age (S.D.) 44.8 (9.8) 47.9 (11.8) 48.6 (9.0) 38.4 (5.8) 47.5 (15.3) 45.3 (10.5)
mean years in practice (S.D.) 18.1 (9.9) 21.9 (11.4) 20.5 (8.4) 12.5 (5.9) 19.2 (16.2) 18.4 (10.3)
Solo (%) 70 71 67 88 67 73
N = number of participants
S.D. = standard deviation
Content analysis of the text fragments resulted in primary codes, which after an inductive interpretation and categorization process could be structured in three themes (table 3): barriers to patient's compliance/adherence as experienced by the GPs; solutions as used by the GPs to increase their patients' compliance/adherence and coping mechanisms used by the GPs.
Table 3 Themes, categories and codes.
Theme code
Barriers to adherence
Patient Categorical approach
Social isolation
Deficient knowledge of diabetes
Minimising of the disease
Opposition to change
'Modern' medicine
Physician Choosing the easiest way
Supposed lack of respect
Model of health care No repayment for self-monitoring
Lack of multidisciplinary support
Solutions
Communication Evaluate knowledge
Give information
Repeat
Control how it can be achieved
In the patient's language
Work in a structured way Make appointments
Standardized educational packages
Structured file administration
Shared care Multidisciplinary teams
Identical messages
Coping mechanisms
Physicians Directing /paternalistic attitude
Induction of guilt and fright
Patients Discussed decision-making
Barriers to compliance/adherence
With regard to compliance/adherence, GPs recognized two kinds of patients: "motivated" patients and "not-to-be-motivated" patients. The "good" patients were very motivated, they did their best and they almost meticulously followed their regimens. The "bad" patients were hardly to be motivated, they thought they have everything under control but they were neglectful which was frustrating for the GP.
M. I think you can hardly speak about "the" diabetes patients; at least this is what I see in my surgery. You have people who are very conscious of their illness, they really take care and on the other hand, there are the ones who neglect everything. I bet everybody here knows these kinds of patients [general agreement]. So, all these different people have a different attitude towards their disease and their treatment in general.
Though they know it is an important challenge and duty, GPs were convinced that it is very difficult to motivate their patients to adhere to treatment regimens. They considered barriers for compliance/adherence on three different levels: the patient, the GP and the care model in which the treatment is offered.
The patient
At the patient's level, their GPs identified important barriers such as social isolation and deficient knowledge of diabetes. They also remarked that diabetes patients tend to minimize their disease. This really is in contrast with the GP's objectives.
F. I believe patients often do underestimate diabetes [general agreement of the group]. The patients think it is an infirmity of old age, an age disease. Consequently, they are not motivated to follow strictly their diet and therapy. That is my conclusion after 30 years of practice.
GPs acknowledged the kind of problems patients have with the treatment regimen and realized that a lifestyle change is not that simple. Yet, they felt the patient's opposition by lots of excuses for preservation of non-healthy habits.
On the other hand, some patients have learned to rely on modern medicine that offers medication for every complaint. To take that medicine is easier and fits better in our way of life than to part with dietary habits linked to the patient's cultural environment.
P. If you feel ill, then you take a pill... anyway, there are drugs for all complaints.
The general practitioner
GPs quickly begin to feel powerless once they experience that patients do not attain the EBM treatment goals. This has a negative impact on further promotion of compliance/adherence to proposed regimens. GPs cherish high expectations concerning metabolic results in their diabetes patients, but do not succeed in communicating this to their patients. Few GPs liked to repeat time and again the same messages, they didn't like to "nag" or "tease" as it is mentioned in one of the focus groups, as this simply leads to frustration.
W. Of course we must ask ourselves if we are doing the right thing. Is it really so bad to give a pill when the patient is satisfied, when the sugar levels are OK and the HbA1c is good, but when the patient doesn't strictly follow his diet. What do we have to do then?
B. When the patient persists in neglecting his treatment, I give up...
The care model
With regard to diabetes care, the Belgian health care system uses a convention system in which intramural (secondary care) centers arrange with the National Institute for Disease and Disability Insurance (RIZIV/INAMI) to refund well-defined diabetes care to specific patients. To be considered for this care, at least two insulin injections a day are required. On this condition, patients have a right to free dietary advice and diabetes education and free (but limited) materials for self-monitoring.
For all other diabetes patients, mostly treated in primary care, currently there is no reimbursement for education or self-monitoring of blood glucose. The GP felt that his good relationship with the patient is threatened by this inequality, because others can offer their patients advantages that they as GPs cannot give. In the GPs opinion, this interference in the relationship of trust could hinder the compliance/adherence. The superior position of the free accessible "specialist", who may have another message for the patient, also gave many GPs a bad feeling.
P. Our experience points out that some of these centers take over the care of patients and that we never see those patients again. We still get reports and results of blood tests on a regular base but we do not see the patient any more...
Poor co-operation between intra-mural diabetes centers and the GPs divides the care process between different partners ignoring each other's efforts and responsibilities. Moreover, in the system of fee-for-service the GP often has to go along with the expectations of the patient in order to keep him in his surgery and to protect his income. This made the physician thinks his patients do not appreciate his specific expertise and skills, which again may result in frustration and even anger.
G. Get angry ... [sigh], but afterwards I feel so frustrated. Get angry with somebody to whom you could have explained in an adult way, "this is becoming impossible", "I do it this way, that's my routine and we'll try stick to it." When I get angry, it is on my mind for the rest of the day...
Solutions
Solutions were identified in three categories. Core elements hereby are: communication, tailored care and shared care.
Communication
GPs realized they have opportunities for communication with the patient. They knew it is wise to check first what the patient already knows, in order to give further relevant information, taking into account the patient's ability to assimilate the messages. GPs should repeat information, check the understanding; explore the patients' own thoughts about and the willingness to apply them.
Doing so, GPs should be careful not to overload patients with information and not to keep "harping on" about health advice that does not (yet) interest the patient.
A. I wonder if you haven't to be careful with those explanations. You see, if each time you give a flow of information, you'll make the patient sick with it ... He'll get demotivated, especially when, after a blood sample, showing a bad result while the patient himself thinks he has done his best. You must not exaggerate in trying to give advice.
Tailored care
GPs considered that the way in which a consultation is carried out could also stimulate compliance/adherence. Important factors seen as able to increase compliance/adherence were scheduled consultations and teamwork. It should give the patient a better notion of "care giving" when a structured diabetes consultation model is adopted. This, together with consultation hours by appointment give patients a feeling of more time being spent and of more experience being applied on his/her behalf. GPs thought that standardized education tools and a structured diabetes file could improve compliance/adherence compared with an "ad hoc" consultation in which the GP only tends to go along with patients' complaints and questions.
B. I think it's partly our fault. I have the impression I cannot follow up so easily and I feel you do it in a more structured way. It all depends on that. So "How does one handle it, how can one work in a structured way on his own?" If you do so, the patient follows. In my case, it is not very structured so far and I don't follow up very well. Therefore, my patients don't do it either. I think that can make a difference when you handle it professionally and structured.
Shared care
GPs considered that working in a multi disciplinary team might encourage better compliance/adherence. They believed that dieticians could give more adequate and more varied food advice than they could themselves. When they succeeded in referring the patients to the dietician, the majority of patients were very satisfied, but the first step remained difficult.
T. I find out that it's not sufficient when we give advice. That's why I usually send the patients, as soon as possible, to the diabetic dietician who gives them explanations. When 2 or 3 people co-operate, it works better. The patients follow up more strictly then when a doctor works on his own. A doctor might also give a leaflet about diabetes, the patients read it but next time, they have forgotten everything about it. They put it somewhere, stuffed away in a drawer. When a dietician sees the patient, he asks what he eats. All that practical information makes a stronger impression than the physician's advice. Besides, I don't have enough time to give such detailed explanations.
Coping mechanisms
GPs often become directing and paternalistic in order to cope more easily with the above-mentioned barriers. This attitude may induce guilt and anxiety in the patients. Physicians assumed that shocking the patients, pressuring and threatening to send them to the hospital were often the only effective methods to improve compliance/adherence.
W. It all depends on the characters, you see. You have people who never feel responsible; they always put the blame on somebody else, despite all your explanations. If something bad happens, they'll tell you you've never explained enough, never emphasized enough or never frightened them enough. Because, you say sometimes "You should not frighten them." However, for some people the only option is to make them afraid. Otherwise, they won't do it and then they blame you for not doing it...
In this way, GPs tighten a net around the patient, restricting his/her freedom of movement. An indication of this restriction can be found in the explicit request for a more tailored care system. In emphasizing this, doctors may minimize their own responsibility and try to shift the blame to the liberal marginal conditions in which they are functioning.
Physicians also considered that patients take refuge behind different reasons in an argued decision, such as cultural motives (for example dietary tradition), financial problems (lack of reimbursement) and strict personal reasons to avoid compliance/adherence. They viewed the motives of patients as deep-rooted and difficult to modify.
Discussion
Many patients for whom diabetes medication is prescribed are poor compliers with treatment, including both oral medication and insulin [29]. In these exploratory focus groups, we discover barriers and solutions for optimal compliance/adherence in people living with diabetes type 2, seen by Belgian GPs.
From the dynamic discussions in the focus groups, we may conclude that this subject was of substantial concern to GPs. There was a constructive mutual interaction and good co-operation in all five focus groups. Open discussions went along with a positive willingness to participate in the debate.
From the GPs' point of view, an inventory could be made of the different barriers and solutions that matter for compliance/adherence of their diabetes patients. This inventory is far from complete. Besides, it was not the intention of this study to make out an exhaustive list of all kinds of compliance/adherence barriers. In this article, we confine ourselves to the viewpoint on compliance/adherence by diabetes patients, as experienced by the GP.
We learn that GPs take a shot at the patients, complaining their deficient knowledge, the fact they minimize diabetes related problems and their blind confidence in medicine. GPs also blame the health care system in which they feel as inferior doctors in comparison with specialists, widely accessible for patients. Taken together, these externalizing factors seem to lead to a profound defeatism. Where the GP tries to give "good" diabetes care by strictly promoting evidence-based diabetes objectives, the patients' "failing" frustrates him. Besides, they recognize the lack of effective communication tools for making the patient a real partner in their decisions.
Probably, GPs feel inclined to allocate the extent of someone's responsibilities too fast. Perhaps, they take too much personal responsibility and focus too much on evidence based medical treatment options wherefore they are real experts and responsible for. A failing in reaching these treatment goals then is a failing of the doctor-expert himself who is doing the best he can. Lifelong guidance of chronic ill patients as is the case in diabetes, asks for shared responsibilities where going about with the disease (behavioral changes, compliance/adherence attitudes, ...) is the patient's responsibility. New essential skills and attributes for a health care provider to promote behavioral change and risk reduction are skills for assessing readiness for behavioral change, relationship building skills, and skills in considering the patient's attitudes and beliefs about the disease or treatment involved [30]. Motivational interviewing, characterized by its empathic non-confrontive style, assists movement through the stages of change to the "action" stage where engaging in change behavior begins. It is designed to assist clients (patients) in exploring and resolving ambivalence to increase motivation for change. Physicians can use the model to intervene successfully at all different levels of behavioral change, where patients are helped by the model to take responsibility for changing habits [31].
Lacking these skills of motivational counseling and shared decision making, the best method to promote patient's compliance/adherence is assumed to be using shock tactics. Identical conclusions were found in a Polish study were doctors, when confronted with non-compliance, also applied certain tactics of doubtful effectiveness which might worsen the patient's emotional state or harm his as a person. Valuable techniques of proven efficacy were not used [32].
Tailored and shared care is considered as a possible solution for better compliance/adherence of diabetes patients. From GPs point of view, this long for teamwork fits with emerging international evidence stating that structured [33] and patient oriented [34] care result in better outcomes. This new model of care fits quite well with our medical paradigms, but might be too much adaptedto the caregivers, without taking into account the views and priorities of (chronic) patients. It remains to be seen if structured care can also deliver a better long term care of patients with chronic illnesses. Don't we forget the patient in organizing shared care?
The cornerstones of health care to support active patient participation are to guarantee the continuity of care, to integrate education in health care and to encourage the patient's attendance [35]. An increased patient satisfaction may be contributing to improved clinical outcomes [36]. However, the relation between patients' satisfaction and GPs patient-centered behaviors remains unclear [37]. There is also limited and mixed evidence on the effects of patient centered interventions on patient health care behaviors or health status; or on whether these interventions might be applicable to providers other than physicians as diabetes-specialized nurses [38]. Professionals committed to achieving the benefits of patient centered consulting should take care not to lose the focus on disease while paying attention to the unique experience of illness of each patient [39]. Enhancing patient-provider agreement on both overall treatment goals and specific strategies to meet these goals may lead to improved patient outcomes [40].
Until now, little research has being done on the relationship between health care provider and compliance/adherence. For HIV patients, better treatment adherence is reported in those patients who perceived themselves to be more engaged with their health care provider [41]. However, since self-reported measures were used, this finding should be viewed with caution. Also in antidepressant treatments, the physician's attitude about the medication is a key factor for improving compliance [42].
A Canadian study on the role of patient, physician and system factors in the management of type 2 diabetes patients [43] shows very striking similarities with our study on compliance/adherence. Canadian GPs also believe that early educational interventions for patients with diabetes resulted in better outcomes. They also described the diabetes educational centers as a valuable resource and stressed the importance of referring the patient soon after diagnosis. In Canada, time and physician remuneration were identified as the main health system barriers to optimal diabetes management.
Qualitative research is often deemed to have lower reliability compared with experimental research. Yet, it has strong face validity, especially when it includes an observational component that enables the researcher to compare oral statements with actual practice [44]. In all our focus groups, all participating GPs spoke very spontaneous in an open atmosphere without any verbal aggression. However, in answering the questions, there was a clear sound of frustration in all five focus groups. Probably, that is a reason why some of the quotes sound rather punitive and blaming of patients. We never had the feeling that GPs were "set up" to respond in such a way. They react defensively because they feel powerless. They are afraid of decreasing income and standing by loosing patients in a fee-for-service payment with free entrance to all levels of care.
At least some limitations to our study have to be considered.
First, the topic of interest. Research in such a complicated matter as compliance/adherence is extremely difficult and always fragmentary, especially given the lack of a specific model or general theory. Different authors use "compliance", "concordance" and "adherence" haphazardly, complicating a comprehensive study of the literature [26]. Additional, for instance in Dutch, the three different meanings of compliance, concordance and adherence are interpreted only by one word. Though we presume in this context that "compliance" (doing what is told to be done) is the best translation, we prefer to use the combined term "compliance/adherence".
Second, the sampling. The participating GPs in this study were all volunteering and interested in a discussion on "problems diabetes patients might have to follow treatment recommendations", as mentioned in the introduction letter. Perhaps, volunteering GPs are more willing to discuss these difficult problems in the patient provider interaction. Since our sample fits well with age, gender and practice organization (table 2) of the Flemish GP population, we assume having described a maximal variation of answers. Besides, the main goal of our study was not to generate conclusions that could be generalized, but to explore and gain a deeper understanding of the GPs perspective of patient compliance/adherence. This limitation should also be taken into account by interpreting the results and before transferring the conclusions to other contexts. It could be argued that selecting GPs in different focus groups according to gender or experience could have been a more appropriate approach. The number of years in practice might be influence frustration. However, the provoked diversity of the group was meant to elicit a rich dynamic interaction in the conducted conversation.
Third, the analysis. While looking for barriers and solutions in patients' adherence GPs' frustration is an important issue in our results. GPs feel as if the patient does not consider them capable to treat their illness. However, a recent and extensive health inquiry in Belgium confirms the people's confidence in general medicine [45]. More, patients' own choice of a primary care physician seems to be associated with better adherence to their treatment regimes [46]. This relationship even remained significant after controlling for possible confounding factors as physician gender, patient gender or length of relationship. Perhaps GPs underestimate their potential value and the effect they can achieve with their patients. Rather than reflecting on themselves and making themselves familiar with newer communication skills, GPs shift the blame of failing on an inappropriate system or other extrinsic factors (e.g. cultural and social problems).
Patients, especially those with chronic illness, make decisions about treatments that fit with their own beliefs and personal circumstances. They may have clear reasons, narrowly related to their health beliefs, for adhering or not adhering to treatment recommendations.
The shared decision making process is rather complex. A joint patient-provider perspective is needed. Health care professionals should assess patients' reasons and beliefs in order to achieve agreement on attainable treatment goals. They need to shift the emphasis away from attempting to directing patients into take the medication they prescribe, towards learning how they can contribute to the decisions that patients currently make about their medications [47].
Adherence to treatment is therefore the shared responsibility of patients and GPs. In complex treatment plans, as is the case in diabetes, a possible way of helping people being adherent perhaps is helping them making more explicit their own attainable priorities. Exploring thepatient's expectations towards the disease and its treatment and translating these individual expectations to realizable and realistic objectives for the patient and in concert with the patient, is an important communication task for the GP. In fact, it is patients who should be the primary actors in medical decision-making, and the health professionals should adopt a supportive role. This joint patient-provider interaction could be a real contribution to a more positive approach on diabetes care. It will help patients to set their own desirable goals and give the GPs' consult again more value and depth. . In essence, then, compliance is an elusive, flexible goal.
Conclusion
The patient's own expectations with regard to illness and health not always correspond to the objectives and expectations of the physician's treatment proposals. The motivation of the physician to achieve a good result may be in conflict with the patient's own motivation to lead his own life. GPs seem to be in need of communication skills to integrate the various expectations of physicians and patients regarding diabetes care.
Our findings suggest a necessary shift to a model of patient-provider-partnership with mutual agreement on shared decisions. A closer relationship wherein the patient is more engaged to his treating GP possibly could reduce frustration.
Shared care, referred by these GPs as a possible solution for better compliance/adherence, only can fulfill these purpose if, from the beginning, the patients' perspectives and concerns are not forgotten.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JW carried out the design of the study; he is responsible for the phrasing of the focus group questioning, compounding the codebook and analysis of all transcripts. He drafted and finalised the manuscript; EV participated in discussing the concept of the project, question phrasing and the coding process. He made a major contribution to rewriting all draft versions of the manuscript; PVR also participated in the definitive phrasing of the questions and critically revised the manuscript as did BS and JD. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors wish to thank all participant GPs for their willingness to participate. Special thanks are also expressed to Ria Patteet, diabetes nurse, who moderated all the five focus groups, to Betty De Strooper for the translation of the manuscript and to Dr. Colin Greaves, health psychologist at the Exeter University Peninsula Medical School (U.K.), for revising the manuscript.
==== Refs
Wens J Van Casteren V Vermeire E Van Royen P Pas L Denekens J Diagnosis and treatment of type 2 diabetes in three Belgian regions. Registration via a network of sentinel general practices Eur J Epidemiol 2001 17 743 50 12086092 10.1023/A:1015627912556
American Diabetes Association Screening for diabetes Diabetes Care 2003 26 S21 S24 12502615
UK Prospective Diabetes Study Group Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33) Lancet 1998 352 837 853 9742976 10.1016/S0140-6736(98)07019-6
UK Prospective Diabetes Study Group Effect of intensive blood glucose control with metformin on complications in overweight patients with type 2 diabetes. (UKPDS 34) Lancet 1998 352 854 865 9742977 10.1016/S0140-6736(98)07037-8
UK Prospective Diabetes Study Group Tight blood pressure control and risk of macrovascular and microvascular complications in type 2 diabetes (UKPDS 38) BMJ 1998 317 703 713 9732337
UK Prospective Diabetes Study Group Efficacy of atenolol and captopril in reducing risk of both macrovascular and microvascular complications in type 2 diabetes (UKPDS 39) BMJ 1998 317 713 720 9732338
Anon Working Party from Complinace to concordance. Achieving goals in medicine taking London, Royal Pharmaceutical Society of Great Britain 1997
Morris LS Schulz RM Patient compliance: an overview J Clin Pharm Ther 1992 17 283 295 1464632
Sackett DL Snow JC Haynes RB, Taylor DW, Sackett DL The magnitude of compliance and non-compliance Compliance in Health Care 1979 Baltimore, The John Hopkins University Press
Griffith S A review of the factors associated with compliance and the taking of prescribed medicines British Journal of General Practice 1990 40 114 116 2112014
Roberson MHB The meaning of compliance: patient perspectives Qualitative Health Research 1992 2 7 22
Alogna M Perceptions of severity of disease and health locus of control in compliant and non compliant diabetes patients Diabetes Care 1980 3 533 534 7460724
Boom CAB Hart LK The relationship between the health belief model and compliance of persons with diabetes mellitus Diabetes Care 1980 3 594 598 7002514
Cahill J Patient participation: a concept analysis J Advanced nursing 1996 24 561 571 10.1046/j.1365-2648.1996.22517.x
Bissell P May CR Noyce PR From compliance to concordance: barriers to accomplishing a re-framed model of health care interactions Social Science & Medicine 2004 58 851 62 14672598 10.1016/S0277-9536(03)00259-4
Little P Everitt H Williamson I Preferences of patients for patient centred approach to consultation in primary care: observational study BMJ 2001 322 468 472 11222423 10.1136/bmj.322.7284.468
Lagro-Janssen ALM Van der Werf G Th Hoe spreken huisartsen en hun patiënten over ziekte? Een zoektocht in de medische literatuur naar het ziektebegrip van huisartsen en patiënten Huisarts en Wetenschap 1999 42 539 544
Anderson RM Robins LS How do we know? Reflections on qualitative research in diabetes Diabetes Care 1998 21 1387 1388 9727880
Murri R Antinori A Ammassari A Nappa S Orofino G Abrescia N Mussini C D'Arminio Monforte A Wu AW AdICoNA Study Group Physician estimates of adherence and the patient-physician relationship as a setting to improve adherence to antiretroviral therapy J Acquir Immune Defic Syndr 2002 31 S158 62 12562042
Hunt LM Arar NH An analytical framework for contrasting patient and provider views of the process of chronic disease management Med Anthropol Q 2001 15 347 67 11693036 10.1525/maq.2001.15.3.347
Trestle JA Medical compliance as an ideology Soc Sci Med 1988 27 1299 308 3070761 10.1016/0277-9536(88)90194-3
Morgan DL Krueger RA "Focus Group Kit" parts 1 to 6 1997 SAGE Publications ISBN 0-7619-0760-2
Mays N Pope C Qualitative Research. Rigour and qualitative research BMJ 1995 311 109 112 7613363
Murphy E Dingwall R Greatbatch D Parker S Watson P Qualitative research methods in health technology assessment: a review of the literature Health Technology Assessment 1998 2 iii-ix, 1-274
Vermeire E Van Royen P Coenen S Wens J Denekens The adherence of type 2 diabetes patients to their therapeutic regimens: patients' perspective. A qualitative study Pract Diabetes Int 2003 20 209 214 10.1002/pdi.505
Vermeire E Hearnshaw H Van Royen P Denkens J Patient adherence to treatment: three decades of research. A comprehensive review J Clin Pharm Ther 2001 26 1 12 11286602 10.1046/j.1365-2710.2001.00363.x
Krueger RA Focus groups: a practical guide for applied research 1988 Newbury Park, Calif: Sage Publications
Morgan DL Stewart M, Tudiver F, Bass MJ, Dunn EV, Norton PG Designing focus group research Tools for primary care research 1992 Newbury Park, Calif: Sage Publication
Cramer JA A systematic review of adherence with medications for diabetes Diabetes Care 2004 27 1218 1224 15111553
Burke LE Fair J Promoting prevention: skill sets and attributes of health care providers who deliver behavioral interventions Journal of Cardiovascular Nursing 2003 18 256 266 14518601
Botelho RJ Novak S Dealing with substance misuse, abuse, and dependency Prim Care 1993 20 51 70 8464948
Heszen-Klemens I Patients' non-compliance and how doctors manage this Soc Sci Med 1987 24 409 416 3576257 10.1016/0277-9536(87)90213-9
Griffin S Kinmonth AL Systems for routine surveillance for people with diabetes mellitus (Cochrane Review) The Cochrane Library 2003 Oxford: Update Software
Renders CM Valk GD Griffin S Wagner EH Eijk JThM van Assendelft WJJ Interventions to improve the management of diabetes mellitus in primary care, outpatient and community settings. (Cochrane Review) The Cochrane Library 2003 Oxford: Update Software
Van den Arend IJ Stolk RP Krans HM Grobbee DE Schrijvers AJ Management of type 2 diabetes: a challenge for patient and physician. [Review] Patient Education & Counseling 2000 40 187 94 10771372 10.1016/S0738-3991(99)00067-1
Alazri MH Neal RD The association between satisfaction with services provided in primary care and outcomes in Type 2 diabetes mellitus Diabetic Medicine 2003 20 486 90 12786685 10.1046/j.1464-5491.2003.00957.x
Mead N Bower P Hann M The impact of general practitioners' patient-centredness on patients' post-consultation satisfaction and enablement Social Science & Medicine 2002 55 283 299 12144142 10.1016/S0277-9536(01)00171-X
Lewin SA Skea ZC Entwistle V Zwarenstein M Dick J Interventions for providers to promote a patient-centred approach in clinical consultations. [Systematic Review] Cochrane Consumers & Communication Group Cochrane Database of Systematic Reviews 2004
Kinmonth AL Woodcock A Griffin S Spiegal N Campbell MJ Randomised controlled trial of patient centred care of diabetes in general practice: impact on current wellbeing and future disease risk BMJ 1998 317 1202 1208 9794859
Heisler M Vijan S Anderson RM Ubel PA Bernstein SJ Hofer TP When do patients and their physicians agree on diabetes treatment goals and strategies, and what differences does it make? Journal of General Internal Medicine 2003 18 893 902 14687274 10.1046/j.1525-1497.2003.21132.x
Demmer C Relationship with health care provider and adherence to HIV medications Psychological reports 2003 93 494 496 14650680
Demyttenaere K Risk factors and predictors of compliance in depression. Review European Neuropsychopharmacology 2003 13 S69 75 14550579 10.1016/S0924-977X(03)00095-6
Brown JB Harris SB Webster-Bogaert S Wetmore S Faulds C Stewart M The role of patient, physician and systemic factors in the management of type 2 diabetes mellitus Family Practice 2002 19 344 349 12110552 10.1093/fampra/19.4.344
Pope C Van Royen P Baker R Qualitative methods in research on healthcare quality Qual Saf Health Care 2002 11 148 152 12448807 10.1136/qhc.11.2.148
Health Survey Interview 36 Gezondheidsenquête door middel van interview
Krupat E Stein T Selby JV Yeager CM Schmittdiel J Choice of a primary care physician and its relationship to adherence among patients with diabetes Am J Manag Care 2002 8 777 84 12234018
Donovan JL Patient decision making. The missing ingredient in complex research Int J Technol Assess Health Care 1995 11 443 455 7591546
| 15890071 | PMC1156882 | CC BY | 2021-01-04 16:29:13 | no | BMC Fam Pract. 2005 May 12; 6:20 | utf-8 | BMC Fam Pract | 2,005 | 10.1186/1471-2296-6-20 | oa_comm |
==== Front
BMC GenetBMC Genetics1471-2156BioMed Central London 1471-2156-6-231589289310.1186/1471-2156-6-23Research ArticleNaturally occurring antisense RNA of histone H2a in mouse cultured cell lines Nishida Hiromi [email protected] Yasuhiro [email protected] Yuko [email protected] Yoshihide [email protected] Laboratory for Genome Exploration Research Group, RIKEN Genomic Sciences Center (GSC), RIKEN Yokohama Institute, 1-7-22 Suehiro-cho, Tsurumi-ku, Yokohama, Kanagawa 230-0045, Japan2005 14 5 2005 6 23 23 1 12 2004 14 5 2005 Copyright © 2005 Nishida et al; licensee BioMed Central Ltd.2005Nishida et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
An antisense transcript of histone H2a that has no significant protein-coding region has been cloned from a mouse full-length cDNA library. In the present study, we evaluated this transcript by using RT-PCR and compared the expression patterns of the sense and antisense transcripts by using quantitative RT-PCR (qRT-PCR).
Results
This antisense RNA was expressed in three mouse cell lines. We call it ASH2a. ASH2a includes not only the complementary sequence of the transcript of Hist2h2aa2 (a replication-dependent histone H2a gene), but also that of the promoter of Hist2h2aa2. The upstream genomic sequence of the transcription start site of the ASH2a-coding gene (ASH2a) lacks both CCAAT and TATA boxes. This absence suggests that the regulation of ASH2a is different from that of the replication-dependent histone H2a genes. Findings from qRT-PCR indicated that the expression pattern of ASH2a was different from that of Hist2h2aa2. Expression of Hist2h2aa2 peaked at 2 to 4 h during S-phase, but that of ASH2a peaked at 1 h.
Conclusion
We showed the existence of ASH2a, a histone H2a antisense RNA, in mouse cultured cells. The expression pattern of ASH2a is different from that of the sense RNA.
==== Body
Background
A comprehensive search of the Functional Annotation of Mouse (FANTOM) database revealed about 30000 full-length cDNA clones without a significant protein-coding region [1]. Indeed, antisense transcripts seem to be present in 10% to 20% of genes in human and mouse genomes [2-6]. These findings suggest that many biological reactions related to antisense transcripts and/or protein-noncoding transcripts are still unrevealed [7]. On the other hand, the cDNA database sometimes includes reverse complements of real transcripts. These artifacts are excluded from the database on the basis of the sequences of intron-splicing sites. Therefore, transcripts without any introns need more experimental evaluation than computational annotation to ascertain their validity.
Histone mRNAs regulated by the cell cycle increase at the beginning of S-phase and decrease at the end of S-phase [8]. Among 20 histone H2a-coding genes, 18 are replication-dependent; the other 2 are replication-independent [9]. The replication-dependent genes lack introns and a poly (A) signal and contain a highly conserved stem-loop structure at the 3' end of the mRNA. This stem-loop structure plays an important role in mRNA processing and stability [10-12]. The promoters of the replication-dependent histone genes contain CCAAT and TATA boxes [13].
As far as we know, Drosophila histone H3 antisense [14] and Leishmania histone H1 antisense [15] transcripts have been reported, but no histone H2a antisense RNA or mammalian histone antisense RNA has been reported. FANTOM 2 [1] contains an antisense transcript (FANTOM clone ID 2210403F13; accession number AK028129) of histone H2a. We call this antisense transcript ASH2a. Comparison of the nucleotide sequence of ASH2a and the mouse genome sequence showed that the ASH2a-coding gene (ASH2a) lies on chromosome 6 without introns. The sequence of ASH2a is exactly complementary to that of the coding region of Hist2h2aa2 (a replication-dependent histone H2a gene) and that of the promoter. In the present study, we evaluated ASH2a transcript by using RT-PCR and compared the expression patterns of Hist2h2aa2 and ASH2a by using qRT-PCR.
Results
Detection of ASH2a
First, cDNAs were synthesized by using the random hexamer oligonucleotide for RNAs from the Hepa 1–6, 3T3, and LLC cell lines. RT-PCR was performed for two targets, the sense-antisense overlap region (between F1 and R1 in Fig. 1) and the overlap region plus the antisense-unique region (between F1 and R2 in Fig. 1). All PCR products for both targets had expected sizes (Fig. 2a). To check the PCR products, we digested them with PstI (Fig. 2b). All digests of the PCR products for the region had the expected sizes (251 and 267 bp for the sense-antisense overlap regions; 251 and 858 bp for the overlap region plus the antisense-unique regions).
Figure 1 (a) Nucleotide sequence of ASH2a. ASH2a is encoded from positions 1 to 2427. Bold characters indicate overlap with the Hist2h2aa2 transcript (italic = protein-coding region). Arrows indicate primers used in this study. (b) Relationship between Hist2h2aa2 and ASH2a RNAs. Arrows indicate the locations of the primers.
Figure 2 (a) RT-PCR products from cDNAs obtained by priming total RNA with random hexamers. Lanes: 1, DNA ladder (100-bp ladder, TOYOBO); 2–7, RT-PCR products amplified with primers F1 and R1 (upper) and those amplified with primers F1 and R2 (lower). RNA was extracted from Hepa 1–6 (lanes 2 and 5), 3T3 (lanes 3 and 6), and LLC (lanes 4 and 7) cells. Superscript III was not added in the reaction of lanes 2–4. Lane 8, PCR product of genomic DNA amplified with primers F1 and R1 (upper) and that amplified with primers F1 and R2 (lower). Arrows indicate the expected products. (b) Patterns of digestion of PCR products by PstI. Lanes: 1, DNA ladder; 2–4, digests of PCR products amplified with primers F1 and R1. PCR product was produced from Hepa 1–6 (lane 2), 3T3 (lane 3), and LLC (lane 4). Lanes 5–7, digests of PCR products amplified with primers F1 and R2. PCR product was produced from Hepa 1–6 (lane 5), 3T3 (lane 6), and LLC (lane 7). Arrows indicate the expected products. (c) RT-PCR products and the EcoRI-digest patterns of cDNAs obtained by priming total RNA with the specific primer R3. Lanes: 1 and 8, DNA ladder; 2–7, RT-PCR products amplified with primers F2 and R3. RNA was extracted from Hepa 1–6 (lanes 2 and 5), 3T3 (lanes 3 and 6), and LLC (lanes 4 and 7). Superscript III was not added in the reaction of lanes 2–4. Lanes 9–11, EcoRI-digests of PCR products amplified with primers F2 and R3. PCR product was produced from Hepa 1–6 (lane 9), 3T3 (lane 10), and LLC (lane 11). Arrows indicate the expected products.
Second, to elucidate the expression of ASH2a, we transcribed the first-strand cDNA with primer R3, which hybridizes specifically to ASH2a. A product of the expected size (685 bp) was obtained (Fig. 2c). In addition, EcoRI digested the PCR product to 269- and 416-bp fragments (Fig. 2c). These results are consistent with the nucleotide sequence of ASH2a.
Quantitative RT-PCR
Because the sequence of the Hist2h2aa2 transcript is not unique, being completely overlapped by the ASH2a sequence, the expression level detected by qRT-PCR using random primers is the sum of the Hist2h2aa2 and ASH2a levels (Fig. 1). Because ASH2a has a unique sequence, the expression level of this antisense transcript can be detected separately. Observation using qRT-PCR indicated that the sum of the Hist2h2aa2 and ASH2a expression levels was always much higher than the level of ASH2a (Fig. 3). The difference between both CT values is more than 4. Thus, we estimated the expression of Hist2h2aa2 and ASH2a as that of Hist2h2aa2 in the following study. Along with cell cycle progression from S-phase, the expression of Hist2h2aa2 increased and peaked at 2 to 4 h, and then decreased (Fig. 4). After that, it increased again, but the expression level was lower than the S-phase peak. Thus, the expression is the highest in the middle of S-phase. On the other hand, that of ASH2a peaked at 1 h, and fell to basal level thereafter (Fig. 4). The rate of the increase of ASH2a RNA was lower than that of Hist2h2aa2 RNA.
Figure 3 Representative amplification plot. Curves indicate amplification from transcripts of GAPDH (first group of rising curves), Hist2h2aa2 (second), and ASH2a (third). Different colors indicate that each result from 0 h to 12 h (13 points). X-axis, cycle numbers; Y-axis, ΔRn.
Figure 4 Transcript expression patterns. (a) Transcripts of Hist2h2aa2. (b) Transcripts of ASH2a. X-axis, time (hours); Y-axis, relative expression level, adjusted to 1.0 at 0 h. The qRT-PCR analyses were performed 4 times (indicated by different colors).
Discussion
An antisense transcript of histone H2a (ASH2a) was clearly expressed in three mouse cell lines, Hepa 1–6, 3T3, and LLC. This result strongly suggests that ASH2a is not regulated with tissue specificity. We checked the upstream regions of ASH2a and the promoter regions of mouse histone H2a genes. All promoters of replication-dependent histone H2a genes include CCAAT and TATA boxes [13]. On the other hand, the upstream region of ASH2a lacks such a structure (Fig. 1). This region has a G+C-rich sequence but lacks both CCAAT and TATA boxes (Fig. 1). Therefore, the regulation of ASH2a is strongly suggested to be different from those of replication-dependent histone H2a genes.
To check the synchronization of the cells, we compared the expression patterns of the replication-dependent histone gene Hist2h2aa2 and the replication-independent histone gene H2afz [16]. Hepa 1–6 cells used in the present study were well synchronized. In fact, the expression pattern of ASH2a was different from that of Hist2h2aa2. The amount of sense RNA was always much higher than that of ASH2a RNA at each time point. Three general functions of antisense transcripts have been identified: transcriptional interference, RNA masking, and dsRNA-dependent mechanisms, including RNA interference [17]. These functions are related to inhibition and/or degradation of sense RNAs. If ASH2a is related to the degradation of the sense RNA through dsRNA formation, ASH2a would be expressed when the sense RNA decreases. However, ASH2a is expressed during the early increase of the sense RNA.
On the other hand, ASH2a could hybridize not only to the Hist2h2aa2 transcript, but also to the transcripts of the other histone H2a-coding genes, because of the high similarity of protein coding sequences. A recent article showed that a small modulatory dsRNA can function as an activator of related genes [18], and the mechanism of action appears to be mediated through a dsRNA/protein interaction, rather than through siRNA or miRNA. Interestingly, ASH2a includes not only a sequence complementary to that of the Hist2h2aa2 transcript, but also a sequence complementary to the promoter region of Hist2h2aa2. Additional work is needed to elucidate the function of the ASH2a-related dsRNA.
Experiments using the high-density oligonucleotide arrays show that a large population of noncoding RNAs are expressed and regulated by similar molecular mechanisms to those involved in the control of protein-coding RNAs [19] and that many transcripts appear to be at very low abundance [20]. It is so important for genome research to elucidate the functions and regulation of noncoding RNAs and antisense RNAs at very low abundance such as ASH2a.
Materials and Methods
Cell lines
Murine hepatoma cell line Hepa 1–6, fibroblast cell line Flp-In-3T3 (Invitrogen), and lung carcinoma cell line LLC were cultured in DMEM supplemented with 10% fetal calf serum.
Cell cycle synchronization
Hepa 1–6 cells were synchronized at the end of G1 phase by the addition of thymidine-hydroxyurea. The cell cycle arrest was released by washing out the thymidine-hydroxyurea, then the cells were harvested at intervals of 1 h from 0 h to 12 h.
RT-PCR
Total RNA fractions extracted from mouse cells were pre-treated with DNase I and used for RT-PCR. Reaction mixture containing the RNA (approximately 0.5 μg) and the strand-specific primer (3.3 pmol) or random hexamer primers was denatured at 70°C, and then reverse-transcription reaction was done with Superscript III (Invitrogen) according to the manual. Then the cDNA was amplified by PCR under the condition of 35 or 40 cycles of 30 s at 95°C, 30 s at 54°C, and 1 or 2 min at 72°C. The sequences of primers are shown in Fig. 1. For quantitative RT-PCR (qRT-PCR), approximately 12.5 ng of total RNA was used for reverse transcription followed by PCR amplification with primers qF1 and R1, or qF2 and R2 (Fig. 1) in reaction mixture containing SYBR premix Ex Taq (Takara) in an ABI PRISM 7700 sequence detection system (Applied Biosystems). The PCR conditions were an initial step of 30 s at 95°C, followed by 40 cycles of 5 s at 95°C and 30 s at 60°C. Expression was assessed by evaluating threshold cycle (CT) values. The relative amount of expressed RNA was calculated using Livak and Schmittgen's method [21]. Quantification of GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA (primers 5'-TGTGTCCGTCGTGGATCTGA-3' and 5'-CCTGCTTCACCACCTTCTTGA-3'; product size 76 bp) was used as a control for data normalization. The qRT-PCR analyses were performed four times.
Authors' contributions
HN designed this study, carried out the molecular biological studies. YT carried out synchronization of cells and quantitative PCR. YO and YH participated in the design of this study.
Acknowledgements
We thank the anonymous reviewer for their helpful comments. We are grateful to Yap Chan Choo, Misato Nakanishi, Taku Tanaka, and Masanori Suzuki for helpful comments. This work was supported in part by a Research Grant for the RIKEN Genome Exploration Research Project, a Research Grant for Advanced and Innovational Research Program in Life Science, and a Research Grant for the Genome Network Project from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of the Japanese Government, and in part by a Research Grant for CREST of Japan Science and Technology Corporation to Y.H., and supported in part by grant 15770055 from the MEXT to H.N.
==== Refs
Okazaki Y Furuno M Kasukawa T Adachi J Bono H Kondo S Nikaido I Osato N Saito R Suzuki H Yamanaka I Kiyosawa H Yagi K Tomaru Y Hasegawa Y Nogami A Schonbach C Gojobori T Baldarelli R Hill DP Bult C Hume DA Quackenbush J Schriml LM Kanapin A Matsuda H Batalov S Beisel KW Blake JA Bradt D Brusic V Chothia C Corbani LE Cousins S Dalla E Dragani TA Fletcher CF Forrest A Frazer KS Gaasterland T Gariboldi M Gissi C Godzik A Gough J Grimmond S Gustincich S Hirokawa N Jackson IJ Jarvis ED Kanai A Kawaji H Kawasawa Y Kedzierski RM King BL Konagaya A Kurochkin IV Lee Y Lenhard B Lyons PA Maglott DR Maltais L Marchionni L McKenzie L Miki H Nagashima T Numata K Okido T Pavan WJ Pertea G Pesole G Petrovsky N Pillai R Pontius JU Qi D Ramachandran S Ravasi T Reed JC Reed DJ Reid J Ring BZ Ringwald M Sandelin A Schneider C Semple CA Setou M Shimada K Sultana R Takenaka Y Taylor MS Teasdale RD Tomita M Verardo R Wagner L Wahlestedt C Wang Y Watanabe Y Wells C Wilming LG Wynshaw-Boris A Yanagisawa M Yang I Yang L Yuan Z Zavolan M Zhu Y Zimmer A Carninci P Hayatsu N Hirozane-Kishikawa T Konno H Nakamura M Sakazume N Sato K Shiraki T Waki K Kawai J Aizawa K Arakawa T Fukuda S Hara A Hashizume W Imotani K Ishii Y Itoh M Kagawa I Miyazaki A Sakai K Sasaki D Shibata K Shinagawa A Yasunishi A Yoshino M Waterston R Lander ES Rogers J Birney E Hayashizaki Y Analysis of the mouse transcriptome based on functional annotation of 60,770 full-length cDNAs Nature 2002 420 563 573 12466851 10.1038/nature01266
Herbert A The four Rs of RNA-directed evolution Nature Genet 2004 36 19 25 14702037 10.1038/ng1275
Kiyosawa H Yamanaka I Osato N Kondo S Hayashizaki Y Antisense transcripts with FANTOM2 clone set and their implications for gene regulation Genome Res 2003 13 1324 1334 12819130 10.1101/gr.982903
Shendure J Church GM Computational discovery of sense-antisense transcription in the human and mouse genomes Genome Biol 2003 3 research0044.1 0044.14 12225583 10.1186/gb-2002-3-9-research0044
Yelin R Dahary D Sorek R Levanon EY Goldstein O Shoshan A Diber A Biton S Tamir Y Khosravi R Nemzer S Pinner E Walach S Bernstein J Savitsky K Rotman G Widespread occurrence of antisense transcription in the human genome Nat Biotechnol 2003 21 379 386 12640466 10.1038/nbt808
Carmichael GG Antisense starts making more sense Nat Biotechnol 2003 21 371 372 12665819 10.1038/nbt808
Mattick JS RNA regulation: a new genetics? Nat Rev Genet 2004 5 316 323 15131654 10.1038/nrg1321
Harris ME Böhni R Schneiderman MH Ramamurthy L Schümperli D Marzluff WF Regulation of histone mRNA in the unperturbed cell cycle: evidence suggesting control at two posttranscriptional steps Mol Cell Biol 1991 11 2416 2424 2017161
Marzluff WF Gongidi P Woods KR Jin J Maltais LJ The human and mouse replication-dependent histone genes Genomics 2002 80 487 498 12408966 10.1016/S0888-7543(02)96850-3
Dominski Z Marzluff WF Formation of the 3' end of histone mRNA Gene 1999 239 1 14 10571029 10.1016/S0378-1119(99)00367-4
Graves RA Pandey NB Chodchoy N Marzluff WF Translation is required for regulation of histone mRNA degradation Cell 1987 48 615 626 3028643 10.1016/0092-8674(87)90240-6
Lanzotti DJ Kaygum H Yang X Duronin RJ Marzluff WF Developmental control of histone mRNA and dSLBP synthesis during Drosophila embryogenesis and the role of dSLBP in histone mRNA 3' processing in vivo Mol Cell Biol 2002 22 2267 2282 11884612 10.1128/MCB.22.7.2267-2282.2002
Osley MA The regulation of histone synthesis in the cell cycle Annu Rev Biochem 1991 60 827 861 1883210 10.1146/annurev.bi.60.070191.004143
Akhmanova A Kremer H Miedema K Hennig W Naturally occurring testis-specific histone H3 antisense transcripts in Drosophila Mol Reprod Dev 1997 48 413 420 9364435 10.1002/(SICI)1098-2795(199712)48:4<413::AID-MRD1>3.0.CO;2-T
Belli SI Monnerat S Schaff C Masina S Noll T Myler PJ Stuart K Fasel N Sense and antisense transcripts in the histone H1 (HIS-1 locus of Leishmania major Int J Parasitol 2003 33 965 975 12906880 10.1016/S0020-7519(03)00126-7
Nishida H Suzuki T Tomaru Y Hayashizaki Y A novel replication-independent histone H2a gene in mouse BMC Genetics 2005 6 10 15720718 10.1186/1471-2156-6-10
Lavorgna G Dahary D Lehner B Sorek R Sanderson CM Casari G In search of antisense Trends Biochem Sci 2004 29 88 94 15102435 10.1016/j.tibs.2003.12.002
Kuwabara T Hsieh J Nakashima K Taira K Gage FH A small modulatory dsRNA specifies the fate of adult neural stem cells Cell 2004 116 779 793 15035981 10.1016/S0092-8674(04)00248-X
Cawley S Bekiranov S Ng HH Kapranov P Sekinger EA Kampa D Piccolboni A Sementchenko V Cheng J Williams AJ Wheeler R Wong B Drenkow J Yamanaka M Patel S Brubaker S Tammana H Helt G Struhl K Gingeras TR Unbiased mapping of transcription factor binding sites along human chromosomes 21 and 22 points to widespread regulation of noncoding RNAs Cell 2004 116 499 509 14980218 10.1016/S0092-8674(04)00127-8
Kampa D Cheng J Kapranov P Yamanaka M Brubaker S Cawley S Drenkow J Piccolboni A Bekiranov S Helt G Tammana H Gingeras TR Novel RNAs identified from an in-depth analysis of the transcriptome of human chromosomes 21 and 22 Genome Res 2004 14 331 342 14993201 10.1101/gr.2094104
Livak KJ Schmittgen TD Analysis of relative gene expression data using real-time quantitative PCR and the 2-ΔΔCT method Methods 2001 25 402 408 11846609 10.1006/meth.2001.1262
| 15892893 | PMC1156883 | CC BY | 2021-01-04 16:38:18 | no | BMC Genet. 2005 May 14; 6:23 | utf-8 | BMC Genet | 2,005 | 10.1186/1471-2156-6-23 | oa_comm |
==== Front
BMC GenetBMC Genetics1471-2156BioMed Central London 1471-2156-6-251594388310.1186/1471-2156-6-25Research ArticleEstimating haplotype frequencies in pooled DNA samples when there is genotyping error Quade Shannon RE [email protected] Robert C [email protected] Katrina AB [email protected] Department of Epidemiology and Biostatistics, Case Western Reserve University, 2103 Cornell Rd, Cleveland, Ohio 44106-7281, USA2005 19 5 2005 6 25 25 3 9 2004 19 5 2005 Copyright © 2005 Quade et al; licensee BioMed Central Ltd.2005Quade et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Maximum likelihood estimates of haplotype frequencies can be obtained from pooled DNA using the expectation maximization (EM) algorithm. Through simulation, we investigate the effect of genotyping error on the accuracy of haplotype frequency estimates obtained using this algorithm. We explore model parameters including allele frequency, inter-marker linkage disequilibrium (LD), genotyping error rate, and pool size.
Results
Pool sizes of 2, 5, and 10 individuals achieved comparable levels of accuracy in the estimation procedure. Common marker allele frequencies and no inter-marker LD result in less accurate estimates. This pattern is observed regardless of the amount of genotyping error simulated.
Conclusion
Genotyping error slightly decreases the accuracy of haplotype frequency estimates. However, the EM algorithm performs well even in the presence of genotyping error. Overall, pools of 2, 5, and 10 individuals yield similar accuracy of the haplotype frequency estimates, while reducing costs due to genotyping.
==== Body
Background
Association studies offer several advantages to linkage analysis for mapping susceptibility loci in complex diseases. They may be more powerful than linkage analysis for loci with a small effect, since the excess sharing across families is expected to be greater than the excess sharing within a family (identity-by-descent (IBD)) [1]. In addition, association studies are expected to provide greater precision in pinpointing the location of susceptibility loci. Finally, association studies do not require the collection of groups of relatives or extended pedigrees, which can be challenging – particularly for late onset diseases.
However, even for association studies, the large sample sizes necessary to study the genetics of complex disease appear unavoidable, so recent interest has focused on methods to reduce the cost. One approach is to use diallelic nucleotide bases, or single nucleotide polymorphisms (SNPs), to help identify susceptibility genes [2]. SNPs are abundantly available in the human genome (approximately 1 per kb of DNA) [3], providing a plentiful source in the genome from which to choose. Additionally, SNP genotyping can be completely automated, and recent technologies have decreased the time necessary to perform the genotyping (as reviewed by Syvanen 2001) [4]. As a result, SNPs are relatively easy, fast, and inexpensive to genotype compared to other existing technologies, such as microsatellite markers (e.g. [5,6]). A second approach to reduce the cost of genotyping is to use DNA pooling, where equal amounts of DNA from each of a group of individuals are combined and then genotyping is performed on the pool instead of on each individual's DNA separately. This procedure has the potential to substantially reduce the genotyping costs, since, if the pools are formed from k individuals, the genotyping costs will be reduced to (100/k)% of the cost of genotyping each individual.
Unfortunately, SNPs are relatively uninformative individually, i.e. more than one is required to obtain an amount of information equivalent to more informative markers, such as microsatellites. One way to increase the information from SNPs is to use haplotypes constructed from multiple SNPs, which is more powerful for detecting an association than using all SNPs individually [7]. The expectation-maximization (EM) algorithm has been implemented to obtain maximum likelihood estimates of haplotype frequencies for pooled data [8,9]. These studies show that the algorithm provides accurate estimates of the haplotype frequencies when no genotyping error is present.
More realistically, genotyping errors do occur, which can have implications for the accuracy of haplotype frequency estimates from pooled samples. In this paper we investigate the effect of genotyping error on 2-SNP haplotype frequency estimates obtained using the EM algorithm for pooled data. We show that the algorithm performs well even in the presence of genotyping error, compared to estimates obtained when there is no genotyping error present.
Results
To evaluate the performance of the EM algorithm, the following parameters were examined: number of individuals per pool, sample size, marker allele frequency, and strength of inter-marker linkage disequilibrium. All parameters were evaluated for scenarios with and without genotyping error.
Number of individuals (k) per pool
For every scenario simulated, the accuracy of the estimates as measured by the similarity index (see below) decreased as the number of individuals per pool increased. Figure 1 shows the accuracy of the estimation procedure under different levels of genotyping error averaged over all simulated genetic models. As the genotyping error increases, the overall accuracy of the estimates slightly decreases. For pools with one individual per pool (i.e. no pooling) the error in the haplotype frequency estimates under no genotyping error is due to sampling error. Similarly, the variance of the similarity index increased with both pool size and genotyping error. That is, the variance of the similarity index for a pool of size 1 was 0.00005 and for a pool of 10 it was 0.008. Within the pool size the variance also increased with genotyping error. For example, for a pool of 10 the variance of the similarity index in the cases of no error, intermediate error, and maximum error were 0.0081, 0.0082, and 0.0084, respectively. When evaluating pool sizes greater than 1, for a given level of genotyping error the largest decrease in accuracy was between k = 2 and k = 5 individuals. There is a smaller decrease in accuracy between k = 5 and k = 10. Even though there was a larger difference in the accuracy between the pools of k = 2 and k = 5, it was minimal. Therefore, these results suggest the accuracy is approximately the same for pool sizes of 10 (92%) as it is for pool sizes of 2 (96%). That is, much less genotyping yields virtually the same amount of accuracy as would be achieved with smaller pool sizes. The figure shown is for a total sample size of 1000; the results are similar for a sample size of 500 (data not shown).
Figure 1 Accuracy of the estimation procedure under different levels of genotyping error averaged over all simulated genetic models. Accuracy of estimation procedure measured by the similarity index as a function of the level of genotyping error simulated for k = 1, 2, 5, and 10 individuals per pool.
Marker allele frequency
To assess the effect of allele frequency on the accuracy of the estimation procedure, haplotype frequencies were computed for simulated samples with one marker allele frequency ranging from 0.01–0.99. Figure 2 shows the plots of the similarity index as a function of marker 2 allele frequency, when marker 1 allele frequency is fixed at 0.2, for a given level of genotyping error. Overall, we see the same shape of the graphs regardless of how many individuals per pool were evaluated. That is, the allele frequencies that yield the best estimates are 0.01 and 0.99, cases when a rare allele is present. These results are for the scenario with no LD present. The graphs also peak at 0.5, but this is due to the nature of the similarity index, a measure that is automatically closer to 1 for allele frequencies close to 0.5. Therefore, as with individual data, when dealing with common allele frequencies the EM algorithm will not perform as efficiently. The figure shown is for a total sample size of 1000; the results are similar for a sample size of 500 (data not shown).
Figure 2 Effect of allele frequency on accuracy of haplotype frequency estimation. Accuracy of estimation procedure measured by the similarity index as a function of the marker 2 allele frequency for k = 1, 2, 5, and 10 individuals. The marker 2 allele frequencies range from 0.01–0.99. The black line represents no genotyping error, the blue line represents intermediate genotyping error, and the red line represents maximum level of genotyping error.
Linkage disequilibrium
The amount of linkage disequilibrium that is present can vary among different populations, and thus could have an effect on the accuracy of the haplotype frequency estimate. To assess the effect of LD on the haplotype frequency estimate, three levels of the LD parameter (0, δmax/2, and δmax) were investigated at different levels of genotyping error for allele frequencies ranging from 0.1–0.99 (Figure 3). Overall, the different levels of LD between the markers resulted in comparable levels of accuracy for the EM algorithm when evaluated across different pool sizes and levels of genotyping error. Figure 3 shows that the difference between the true and estimated frequencies decreases (i.e., the similarity index increases) as the amount of LD becomes stronger. Interestingly, for pooled samples the graphs in Figure 3 show that estimates achieve better accuracy if the amount of LD present is greater than 0.1 for sample sizes of 1000 individuals. We observed using sample sizes of 500 that this critical number is about 0.15 (results not shown). Therefore, as the sample sizes decreases more LD is needed to achieve the higher accuracy for pooled samples.
Figure 3 Effect of LD on accuracy of haplotype frequency estimation for different levels of genotyping error. Accuracy of estimation procedure measured by the similarity index as a function of the LD values for k = 1, 2, 5, and 10 individuals. LD ranges from 0.00–0.25. The black line and circles represent no genotyping error, the blue line and circles represent intermediate genotyping error, and the red line and circles represent maximum level of genotyping error. Smooth line fitted using the Lowess function in Splus.
Discussion
Interest continues to increase in association analysis for complex genetic traits; however, this study design is still not without shortcomings. Therefore, we evaluated the effects of genotyping error on the estimates of haplotype frequencies when pooling DNA for association studies and, more specifically, we assessed the benefits (if any) to be gained from it. Additionally, we investigated the effects of pool size, marker allele frequencies, and LD on the accuracy of the haplotype frequency estimates.
We have shown that accuracy of the haplotype frequency estimates decreases as the level of genotyping error increases. However, this decrease is small and, even in the presence of genotyping error, the estimates of the haplotype frequencies are accurate. Ideally, it would be most beneficial to design studies with a large number of individuals per pool to minimize the genotyping costs. We observed, under all genotyping error levels, that pools of size 2, 5, and 10 all achieve about the same level of accuracy. This suggests that pool sizes of 10 individuals could be used to obtain accurate estimates. Using a larger number of individuals per pool and still obtaining the same level of accuracy allows for an even greater reduction in the cost of genotyping compared to situations that utilize a smaller number of individuals per pool. Additionally, we observed that a sample size of 500 is just as accurate as a sample size of 1000 for the models simulated. Therefore this further supports the notion that less genotyping can be performed and yet the same level of accuracy obtained for haplotype frequency estimates.
We observed that marker allele frequencies and the amount of LD can have an effect on the accuracy of the haplotype frequency estimates. If a rare allele frequency is present in the pool and/or in cases of stronger LD, in the absence of genotyping error more accurate estimates are obtained. Similarly, this same pattern was observed in the presence of genotyping error. For the case of individual genotyped unrelated samples, Kirk and Cardon (2002) [10] evaluated the EM algorithm to estimate haplotype frequencies in the presence of genotyping errors using a much smaller sample size of 50. These authors observed in situations of high LD and/or when rare alleles are present that the EM algorithm offered a high degree of accuracy even in the presence of genotyping errors. In situations with low LD and/or common alleles, the EM algorithm performs more poorly for both individual and for pooled designs.
To date, the most common pooling strategy has been to create one large pool for each condition (e.g., case and control status), and to compare allele frequencies among pools (e.g. [11,12]). This strategy would certainly result in the greatest efficiency in genotyping, but at the cost that individual haplotype frequencies cannot be estimated. Under this strategy, Le Hellard et al. (2002)[13] evaluated several quantitative SNP genotyping methods for pooled samples and compared the true allele frequencies, obtained by genotyping each sample individually, to those estimated from the pooled sample. Although errors are present when estimating the allele frequencies from pooled samples, pooling provided reasonably accurate estimates, even for these very large pools. In a comprehensive review of DNA pooling, Sham (2002) [14] concludes that pooling can be considered both cost and time effective. It remains to be determined whether the cost efficiency gained by forming large pools to reduce the amount of genotyping outweighs the statistical efficiency gained by performing haplotype analysis using smaller pools.
In this analysis we chose to introduce error into genotypes because we were evaluating pooled samples. However, there are several other types of error models available for individual genotyped samples (e.g., [15,16]). Therefore, under a different error model it is possible that the conclusions reached in this analysis might have differed. Ultimately we would like to account for the genotyping error when estimating haplotype frequencies for pooled samples, just as Zou & Zhao (2003) [16] have done for individual genotyped samples.
To assess the accuracy of our results we chose the similarity index because this measure sums across all haplotypes frequencies. However, we could have chosen to evaluate the estimated haplotype frequencies individually, as in Zou & Zhao (2003) [16]. Therefore, evaluating haplotypes using a different measure could result in different conclusions. For example, in Zou & Zhao (2003) [16] the authors report four estimated haplotype frequencies to be 0.366, 0.126, 0.132, and 0.376 where the true frequencies are 0.4, 0.1, 0.1, and 0.4, respectively. Based on this, the authors note the highest change in haplotype frequency estimates to be 30% (this is from an estimated frequency of 0.132 where the true frequency is 0.1). However, for this example the similarity index, which takes all four haplotypes into account, is 0.94.
For this analysis we only chose to evaluate two-marker loci; however, our method can be extended to accommodate many marker loci. For individual samples, as the number of markers increases there is a loss of accuracy in the haplotype frequency estimates. It is possible that this loss of accuracy could be even more severe for pooled samples.
Genotyping error may have an impact on the detection of false positive or false negative signals in genetic association studies, or on the sample size needed to detect an association when using DNA pools. Gordon et al (2002) [17] quantify the effects that individual genotyping errors have on power and required sample size for case-control genetic association studies. They report that genotyping errors increase the likelihood of missing a real effect. Similarly, Zou & Zhao (2004) [18] evaluate the impact of genotyping errors on false discovery rates for individual genotyping and the impact of measurement errors or pool formation errors for pooled genotyping. They report that genotyping errors can lead to a higher rate of false positives for individual genotyping and even higher measurement errors for pooled samples.
Here we only consider the accuracy of the EM algorithm for pooled samples to estimate the haplotype frequencies in the presence of genotyping error and do not evaluate the sample size necessary to detect an association. Therefore, even though we find pool sizes as large as 10 and a sample size of 500 to be efficient for estimating haplotype frequencies, we cannot comment on the effect of genotyping errors on the ability to find false positive or negative associations.
Conclusion
When using the EM algorithm for pooled samples, we found that genotyping error slightly decreases the accuracy of haplotype frequency estimates. However, the EM algorithm still performs well even in the presence of genotyping error. Overall, pools of 2, 5, and 10 individuals yield similar accuracy of the haplotype frequency estimates, likewise for sample sizes of 500 and 1000 individuals. Therefore, we can conclude that the overall amount of genotyping can be reduced by using 10 individuals per pool with sample sizes as small as 500 individuals.
Methods
Genotype simulation
Data were simulated both with and without genotyping error for each pool size (k) under 198 genetic models using combinations of different allele frequencies at each locus (0.01, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 and 0.99) and strength of LD between the markers (between 0 and 0.25). For simplicity, we only considered the case with two loci and two alleles, Ai and Bi for locus i, but this method can be extended to accommodate more marker loci. LD was measured as δmax, calculated as the smaller of pAqB or qApB, where pA and pB represent allele frequencies ranging from 0.01–0.99. For each combination of allele frequencies, 3 situations were simulated with δ equal to 0, δmax/2, and δmax. Throughout our study, we assume Hardy-Weinberg equilibrium. We considered pool sizes (k) of 1,2,5, and 10 individuals, and total sample sizes (N) of 500 and 1000 individuals.
We used a binomial distribution as the error distribution in our simulation, with a variance (σ2) that depends on the parameter α. Let nij be the number of times allele A at locus i is present in pool j, with possible values ti = 0, ..., 2k, and qij be the allele frequency of allele A at marker i in pool j. Then, for each locus, the conditional probability that allele A is observed ti times given the true genotype was modelled as where the summation is over values of y between 0 and 2kα where y/α rounds to ti. The parameter α is the genotyping error parameter. If α = 1, this represents the maximum amount of genotyping error, and as α becomes large, this distribution becomes equivalent to a having no genotyping error, as demonstrated in an example in Table 1. There are several desirable properties that this binomial distribution has, including 1) nij can only take on values between 0 and 2k, 2) the variance of nij will depend on the allele frequency, and 3) σ2 can be adjusted to have a range of values.
Table 1 Conditional probabilities* of observed genotypes, given the true genotype of pool j. For k = 2 individuals with SNP locus AB in pool j.
Observed Unordered Genotypes/Number of A Alleles in Pool j | True Unordered Genotype AABB
α BBBB/ 0 ABBB/ 1 AABB/ 2 AAAB/ 3 AAAA/ 4
1 0.063 0.250 0.375 0.250 0.063
5 0.0002 0.131 0.737 0.131 0.0002
25 0.000 0.006 0.988 0.006 0.000
125 0.000 0.000 1.000 0.000 0.000
* conditional probabilities computed from the binomial distribution
Based on the value of α we introduce error into genotypes by simulating three levels of genotyping error, which are defined as no genotyping error (σ2 = 0), intermediate genotyping error (σ2 = 0.01) which corresponds to a realistic level of genotyping error based on the results of LeHellard et al. (2002) [13], and the maximum possible genotyping error given our error distribution (σ2 = 0.50–5.18, depending on the value of k)
Estimation of haplotype frequencies via the EM algorithm for pooled samples
Wang et al. (2003) [8] and Ito et al. (2003) [9] independently developed algorithms to estimate haplotype frequencies utilizing an EM algorithm for pooled data. Both of these algorithms infer the estimates utilizing a maximum likelihood approach that is identical to the approach we have used when analyzing the data under no genotyping error. To investigate whether a global maximum was found, four sets of starting values were used for k = 1,2,5 individuals per pool, and two sets for k = 10, to determine if they obtain the same maximum likelihood estimate. The results we present are those from whichever starting values gave the largest maximum for the haplotype frequency estimates.
Evaluation of haplotype frequency estimates
We compared the estimated haplotype frequencies to the true haplotype frequencies using the similarity index (IF) [19]. If is the estimated haplotype frequency for haplotype i, h is the total number of haplotypes, and pi is the true haplotype frequency, then IF is defined as . The similarity index takes on values between 0 and 1 and is close to 0 when none of the estimated haplotype frequencies are close to the true haplotype frequencies, and 1 when all of the estimated haplotype frequencies equal the true haplotype frequencies.
Authors' contributions
S.Q. performed the simulations, statistical analysis, and programming of the method. R.E. contributed to modeling the genotyping error. K.G. conceived the study, participated in its design and coordination, and participated in programming of the method.
Acknowledgements
The authors thank two anonymous reviewers for their insightful comments. This work was partially supported by grant HG01577 from the National Human Genome Research Institute, grant GM2835 from the National Institute of General Medical Sciences, and grant RR03655 from the National Center for Research Resources.
==== Refs
Risch N Merikangas K The future of genetic studies of complex human diseases Science 1996 273 1516 1517 8801636
Nowotny P Kwon JM Goate AM SNP analysis to dissect human traits Curr Opin Neurobiol 2001 11 637 641 11595500 10.1016/S0959-4388(00)00261-0
Wang DG Fan JB Siao CJ Berno A Young P Sapolsky R Large-scale identification, mapping, and genotyping of single- nucleotide polymorphisms in the human genome Science 1998 280 1077 1082 9582121 10.1126/science.280.5366.1077
Syvanen AC Accessing genetic variation: genotyping single nucleotide polymorphisms Nat Rev Genet 2001 2 930 942 11733746 10.1038/35103535
Perlin MW Lancia G Ng SK Toward fully automated genotyping: genotyping microsatellite markers by deconvolution Am J Hum Genet 1995 57 1199 1210 7485172
Deloukas P Schuler GD Gyapay G Beasley EM Soderlund C Rodriguez-Tome P A physical map of 30,000 human genes Science 1998 282 744 746 9784132 10.1126/science.282.5389.744
Fallin D Schork NJ Accuracy of haplotype frequency estimation for biallelic loci, via the expectation-maximization algorithm for unphased diploid genotype data Am J Hum Genet 2000 67 947 959 10954684 10.1086/303069
Wang S Kidd KK Zhao H On the use of DNA pooling to estimate haplotype frequencies Genet Epidemiol 2003 24 74 82 12508258 10.1002/gepi.10195
Ito T Chiku S Inoue E Tomita M Morisaki T Morisaki H Estimation of haplotype frequencies, linkage-disequilibrium measures, and combination of haplotype copies in each pool by use of pooled DNA data Am J Hum Genet 2003 72 384 398 12533787 10.1086/346116
Kirk K Cardon L The impact of genotyping error on haplotype reconstruction and frequency estimation European Journal of Human Genetics 2002 10 616 622 12357332 10.1038/sj.ejhg.5200855
Breen G Harold D Ralston S Shaw D St Clair D Determining SNP allele frequencies in DNA pools Biotechniques 2000 28 464 6 468,470 10723558
Norton N Williams NM Williams HJ Spurlock G Kirov G Morris DW Universal, robust, highly quantitative SNP allele frequency measurement in DNA pools Hum Genet 2002 110 471 478 12073018 10.1007/s00439-002-0706-6
Le Hellard S Ballereau SJ Visscher PM Torrance HS Pinson J Morris SW SNP genotyping on pooled DNAs: comparison of genotyping technologies and a semi automated method for data storage and analysis Nucleic Acids Res 2002 30 e74 12140336 10.1093/nar/gnf070
Sham P Bader JS Craig I O'Donovan M Owen M DNA Pooling: a tool for large-scale association studies Nat Rev Genet 2002 3 862 871 12415316 10.1038/nrg930
Douglas J Skol A Boehnke M Probability of Detection of Genotyping Errors and Mutations as Inheritance Inconsistencies in Nuclear-Family Data Am J Hum Genet 2002 70 487 495 11791214 10.1086/338919
Zou G Zhao H Haplotype Frequency Estimation in the Presence of Genotyping Errors Human Herd 2003 56 131 138 10.1159/000073741
Gordon D Finch SJ Nothnagel M Ott J Power and Sample Size Calculations for Case-Control Genetic Association Tests when Errors Are Present: Application to Single Nucleotide Polymorphisms Human Herd 2002 54 22 33 10.1159/000066696
Zou G Zhao H The Impacts of Errors in Individual Genotyping and DNA Pooling on Association Studies Genetic Epidemiology 2004 26 1 14691952 10.1002/gepi.10277
Excoffier L Slatkin M Maximum-likelihood estimation of molecular haplotype frequencies in a diploid population Mol Biol Evol 1995 12 921 927 7476138
| 15943883 | PMC1156884 | CC BY | 2021-01-04 16:38:18 | no | BMC Genet. 2005 May 19; 6:25 | utf-8 | BMC Genet | 2,005 | 10.1186/1471-2156-6-25 | oa_comm |
==== Front
BMC GenetBMC Genetics1471-2156BioMed Central London 1471-2156-6-271591889610.1186/1471-2156-6-27Research ArticlePAX6 mutations: genotype-phenotype correlations Tzoulaki Ioanna [email protected] Ian MS [email protected] Isabel M [email protected] School of Molecular and Clinical Medicine, University of Edinburgh, Molecular Medicine Centre, Western General Hospital, Crewe Road, Edinburgh, EH4 2XU, UK2 School of Biological Sciences, Institute of Evolutionary Biology, Ashworth Laboratories, West Mains Road, Edinburgh, EH9 3JT, UK2005 26 5 2005 6 27 27 22 2 2005 26 5 2005 Copyright © 2005 Tzoulaki et al; licensee BioMed Central Ltd.2005Tzoulaki et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The PAX6 protein is a highly conserved transcriptional regulator that is important for normal ocular and neural development. In humans, heterozygous mutations of the PAX6 gene cause aniridia (absence of the iris) and related developmental eye diseases. PAX6 mutations are archived in the Human PAX6 Allelic Variant Database, which currently contains 309 records, 286 of which are mutations in patients with eye malformations.
Results
We examined the records in the Human PAX6 Allelic Variant Database and documented the frequency of different mutation types, the phenotypes associated with different mutation types, the contribution of CpG transitions to the PAX6 mutation spectrum, and the distribution of chain-terminating mutations in the open reading frame. Mutations that introduce a premature termination codon into the open reading frame are predominantly associated with aniridia; in contrast, non-aniridia phenotypes are typically associated with missense mutations. Four CpG dinucleotides in exons 8, 9, 10 and 11 are major mutation hotspots, and transitions at these CpG's account for over half of all nonsense mutations in the database. Truncating mutations are distributed throughout the PAX6 coding region, except for the last half of exon 12 and the coding part of exon 13, where they are completely absent. The absence of truncating mutations in the 3' part of the coding region is statistically significant and is consistent with the idea that nonsense-mediated decay acts on PAX6 mutant alleles.
Conclusion
The PAX6 Allelic Variant Database is a valuable resource for studying genotype-phenotype correlations. The consistent association of truncating mutations with the aniridia phenotype, and the distribution of truncating mutations in the PAX6 open reading frame, suggests that nonsense-mediated decay acts on PAX6 mutant alleles.
==== Body
Background
The PAX6 gene was cloned during the search for genes underlying the WAGR syndrome (Wilms tumor, aniridia, genitourinary abnormalities and mental retardation; MIM 194072) [1]. WAGR syndrome is caused by hemizygous deletions of 11p13 that remove one copy of PAX6 and one copy of WT1 [1,2]. Intragenic PAX6 mutations were subsequently identified in numerous non-syndromic aniridia patients, confirming PAX6 as the aniridia gene (MIM 106210) [3-6]. Studies on WAGR patients and aniridia patients with chromosomal rearrangements clearly demonstrated that aniridia could be caused by deletion of one copy of the PAX6 gene [1,2]. Thus it was proposed that aniridia results from PAX6 haploinsufficiency and is caused by loss-of-function of one allele [1,5-7].
The PAX6 gene encodes a highly conserved transcriptional regulatory protein that is expressed in the developing eye, brain, spinal cord and pancreas [8]. The 5' two-thirds of the open reading frame (ORF) encode two DNA binding domains, a paired domain and a homeodomain (Figure 1) [9,10]. The DNA binding domains are separated by a 79-amino acid linker peptide. The 3' third of the ORF encodes a proline, serine and threonine-rich (PST) domain that has transcriptional trans-activation function [11] (Figure 1). The last 40 amino acids of the PST domain constitute a highly conserved C-terminal peptide that has been implicated in modulation of DNA binding by the homeodomain [12].
Figure 1 The PAX6 cDNA and protein. Top: the PAX6 cDNA is represented as a horizontal rectangle with the different coding regions indicated: PB, paired box, LNK, linker region, HB, homeobox, PST, proline/serine/threonine-rich region. Exon boundaries are indicated by vertical black lines. 5a is alternatively spliced exon in the paired box. Thick horizontal lines indicate untranslated regions (not to scale). Bottom: cartoon of the PAX6 protein showing the different functional domains. N, N-terminus of protein; C, C-terminus; PB(N), N-terminal subdomain of paired domain; PB(C), C-terminal subdomain of paired domain; HD, homeodomain; PST, PSTdomain.
PAX6 mutations are archived in the Human PAX6 Allelic Variant Database [13,14]. The database now contains 309 records, each reporting independently ascertained sequence variations in the PAX6 gene. Two hundred and eighty-six of these are associated with pathological mutations that cause congenital eye malformations. The most common of these malformations is aniridia, which is chiefly characterised by congenital absence of the iris, but which also affects the cornea, lens and retina. PAX6 mutations also cause a range of non-aniridia phenotypes such as optic nerve defects, keratitis, microphthalmia, and foveal hypoplasia [15-17]. This database records allow analysis of the distribution of mutations in the gene and the relationship between genotype and phenotype.
Although a comprehensive review of the mutations in the database has been published previously [18] we wanted to re-analyse the data for two reasons. First, the last review was published seven years ago and the number of records has greatly increased, from 87 to 309. Second, it is instructive to consider the effect of emerging molecular mechanisms that act on mutant alleles, such as nonsense-mediated decay. Nonsense-mediated decay (NMD) is the process by which mRNAs that contain premature termination codons (PTCs) are degraded before they produce large quantities of truncated proteins [19,20]. NMD is of relevance to the PAX6 mutation spectrum because mutations that introduce a PTC into the PAX6 open reading frame are very common [18]. Before the discovery of NMD it was widely thought that PTC-containing mutant alleles generated truncated proteins [5,6]. Some researchers speculated that PAX6 proteins truncated after the homeodomain might have dominant negative activity [21-23]. The two intact DNA binding domains, divorced from the normal trans-activation domain, could theoretically bind to target DNA sequences without activating downstream genes and could potentially interfere with the function of the normal PAX6 protein. A variety of experimental assays showed that PAX6 proteins with C-terminal deletions do indeed have potent dominant negative activity [21,22,24]. It might be expected therefore that truncating mutations after the homeodomain could potentially lead to phenotypes more severe than, or markedly different from, truncating mutations in the DNA binding domains. Alternatively, if nonsense-mediated decay acts on PTC mutations in vivo, all PTC alleles will effectively be null alleles, and no phenotypic difference should be observed. This idea can be explored by examining the records in the PAX6 Allelic Variant Database
In this paper we review the mutations archived in the PAX6 Allelic Variant Database. We show that over three-quarters of aniridia cases are caused by mutations that introduce a PTC into the PAX6 open reading frame. In contrast, most non-aniridia phenotypes are associated with missense mutations. We also show that four CpG dinucleotides are major mutational hotspots, and account for half of all nonsense mutations in the database. Finally we attempt to reconcile the observed PAX6 mutation spectrum with the work done on truncated PAX6 proteins. We suggest that the PAX6 mutation spectrum is consistent with the idea that nonsense-mediated decay is a major mechanism acting on PAX6 mutant alleles, and consequently that most truncated proteins are unlikely to be produced at significant levels in vivo. Among the existing records, there are no truncating mutations in the 3' part of the coding region where RNA surveillance would not be predicted to act. This suggests that 3' mutations do in fact yield dominant negative alleles that may cause severe phenotypes, but these have not yet been ascertained.
Results and discussion
Truncating mutations in the PAX6 gene are predominantly associated with aniridia
The PAX6 Allelic Variant Database contains 309 records of which 286 refer to pathological mutations in the PAX6 coding region (exons 4–13, Figure 1) or the consensus splice sites directly flanking the coding exons. The remaining records describe polymorphisms.
Each of the 286 disease-associated mutations was classified into one of six categories according to the apparent effect of each genomic change. The six categories are nonsense mutations, splicing mutations, frame-shifting insertions or deletions, in-frame insertions or deletions, missense mutations and run-on mutations. Details of these are given in Table 1.
Table 1 Categories of mutation in the PAX6 Allelic Variant Database
Category Definition
Nonsense Single nucleotide substitution creates a stop codon in the open reading frame.
Splicing Nucleotide substitution, deletion or insertion in consensus splice site.
Frame-shifting insertion or deletion Deletion or insertion of nucleotides in the open reading frame – total number not divisible by three.
In-frame insertion or deletion Deletion or insertion of nucleotides in the open reading frame – total number divisible by three.
Missense Single nucleotide substitution changes one amino acid codon to another in the open reading frame.
Run-on Nucleotide substitution, insertion or deletion changes the termination codon to an amino acid codon. Translation is predicted to continue into the 3'UTR.
Each of the 286 disease-associated mutations in the database was assigned to one of these categories. Twelve compound mutations (each involving more than one mutational event) were categorised according to the final consequence of the mutation (see Methods). 3'UTR, 3' untranslated region.
The exon-by-exon distribution of mutation type for 286 pathological mutations is shown in Table 2. Of the 286 mutations, 102 (35.7%) are nonsense mutations, 36 (12.6%) are splice mutations, 68 (23.8%) are frame-shifting insertions or deletions, 16 (5.6%) are in-frame insertions or deletions, 50 (17.5%) are missense mutations and 14 (4.8%) are run-on mutations (Figure 2a).
Table 2 Exon-by-exon distribution of 286 disease-associated mutations in the PAX6 Allelic Variant Database
4 5 5a 6 7 8 9 10 11 12 13
Mutation type Total
N 0 6 0 9 7 24 23 14 16 3 0 102
S 2 4 0 4 1 1 4 6 5 9 0 36
FS 0 14 0 19 11 4 4 6 1 6 3 68
IF 0 3 0 13 0 0 0 0 0 0 0 16
M 3 13 5 14 3 2 1 1 0 6 2 50
RO 0 0 0 0 0 0 0 0 0 0 14 14
Total 5 40 5 59 22 31 32 27 22 24 19 286
N, nonsense mutation; S, splicing mutation; FS, frame-shifting insertion or deletion; IF, in-frame insertion or deletion; M, missense mutation; RO, run-on mutation. For a full definition of mutation types, see Table 1.
Figure 2 Distribution of different mutation types in the PAX6 Allelic Variant Database. (a) All disease-associated mutations in the database; (b) mutations causing aniridia; (c) mutations causing non-aniridia phenotypes. Mutation definitions are given in Table 1.
Nonsense mutations, splice mutations and frame-shifting insertions or deletions typically result in the introduction of a PTC into the open reading frame. In the PAX6 database, these three categories account for 72% of all disease-associated mutations.
Of the 286 pathological mutations, 257 (89.9%) are associated with aniridia and 29 (10.1%) are associated with other phenotypes, including isolated foveal hypoplasia, microphthalmia and optic nerve defects.
The exon-by-exon distribution of mutation type for 257 aniridia-associated mutations is shown in Table 3. Of the 257 mutations, 100 (38.9%) are nonsense mutations, 34 (13.2%) are splice mutations, 65 (25.3%) are frame-shifting insertions or deletions, 16 (6.2%) are in-frame insertions or deletions, 30 (11.7%) are missense mutations and 12 (4.7%) are run-on mutations (Figure 2b). The proportion of missense mutations has decreased from 17.5% of all cases, to 11.7% of aniridia cases while mutations that introduce a PTC (nonsense, splicing and frame-shifting mutations) have increased from 72% of all cases to 77% of aniridia cases.
Table 3 Exon-by-exon distribution of 257 mutations that cause aniridia.
Exon 4 5 5a 6 7 8 9 10 11 12 13
Mutation type Total
N 0 6 0 9 7 23 23 14 16 2 0 100
S 1 4 0 4 1 1 4 6 4 9 0 34
FS 0 13 0 19 11 4 4 6 1 5 2 65
IF 0 3 0 13 0 0 0 0 0 0 0 16
M 3 11 0 10 0 2 1 0 0 2 1 30
RO 0 0 0 0 0 0 0 0 0 0 12 12
Total 4 37 0 55 19 30 32 26 21 18 15 257
N, nonsense mutation; S, splicing mutation; FS, frame-shifting insertion or deletion; IF, in-frame insertion or deletion; M, missense mutation; RO, run-on mutation. For a full definition of mutation types, see Table 1.
The exon-by-exon distribution of mutation type for 29 mutations in non-aniridia cases is shown in Table 4. Of the 29 mutations, 2 (6.9%) are nonsense mutations, 2 (6.9%) are splice mutations, 3 (10.3%) are frame-shifting insertions or deletions, 20 (69%) are missense mutations and 2 (6.9%) are run-on mutations (Figure 2c). Missense mutations account for over two-thirds of non-aniridia phenotypes, while mutations that introduce a PTC are much less common than in the database as a whole, accounting for just 7 of the cases (24%).
Table 4 Exon-by-exon distribution of 29 mutations that cause phenotypes other than aniridia.
Exon 4 5 5a 6 7 8 9 10 11 12 13
Mutation type Total
N 0 0 0 0 0 1 0 0 0 1 0 2
S 1 0 0 0 0 0 0 0 1 0 0 2
FS 0 1 0 0 0 0 0 0 0 1 1 3
IF 0 0 0 0 0 0 0 0 0 0 0 0
M 0 2 5 4 3 0 0 1 0 4 1 20
RO 0 0 0 0 0 0 0 0 0 0 2 2
Total 1 3 5 4 3 1 0 1 1 6 4 29
N, nonsense mutation; S, splicing mutation; FS, frame-shifting insertion or deletion; IF, in-frame insertion or deletion; M, missense mutation; RO, run-on mutation. For a full definition of mutation types, see Table 1.
This analysis shows that the aniridia phenotype is predominantly associated with mutations that introduce a PTC, while non-aniridia phenotypes are predominantly associated with missense mutations. The missense mutations that cause non-aniridia phenotypes may do so by generating hypomorphic proteins that are able to carry out some but not all of the normal functions of PAX6, such as the correct regulation of downstream target genes [15,28]. The missense mutations that cause non-aniridia phenotypes are predominantly located in the paired domain (exons 5, 5a, 6 and 7, Table 4), suggesting that partially impaired DNA binding may be a major mechanism by which variant phenotypes arise [10,28,29]. Missense mutations can cause full-blown aniridia (Table 3), presumably by creating PAX6 proteins with little or no function [29]. Missense mutations associated with non-aniridia phenotypes typically affect a subset of the ocular tissues involved in full aniridia, such as the fovea, the optic nerve or the iris [15,29].
Mutation hotspots in the PAX6 coding region
Nonsense mutations are the single most common mutation type in aniridia patients (and in the whole database) while missense mutations are the most common cause of other phenotypes (Figure 2). Both nonsense and missense mutations are caused by single nucleotide substitutions. To learn more about how these mutations might arise, we focussed on the distribution of CpG dinucleotides in the PAX6 open reading frame, since CpG transitions are the most common single nucleotide substitutions in the human genome [30].
There are 45 CpG dinucleotides in the PAX6 coding region (Figure 3). CpG deamination can give rise to two new dinucleotides, TpG and CpA, depending on whether the C>T conversion takes place on the forward strand or the reverse strand. In the existing records, transitions at 10 of the 45 CpG's have been reported (Figure 3, Table 5). Eight CpG's have been mutated on one strand only, while two have been mutated on both strands (Figure 3, Table 5). Of the twelve changes, six cause nonsense mutations, five cause missense mutations and one causes a synonymous (neutral) substitution (Table 5).
Figure 3 Distribution of CpG dinucleotides in the PAX6 open reading frame. The PAX6 cDNA is represented as a horizontal rectangle with the different coding regions indicated: PB, paired box, LNK, linker region, HB, homeobox, PST, proline/serine/threonine-rich region. Exon boundaries are indicated by vertical black lines. Exon numbers are shown beneath the cDNA, with the number of CpG's in brackets. Above the cDNA, each of the forty-five CpG dinucleotides in the PAX6 ORF is indicated by an arrow. Red arrows indicate the ten CpG's at which a nucleotide transition has occurred. Single-headed arrows indicate that the CpG deamination has occurred only on the forward strand (CpG > TpG). Double-headed arrows indicate that deamination has occurred both on the forward strand (CpG > TpG) and the reverse strand (CpG > CpA). Elongated red arrows indicate those CpG's that have been hit more than once on the forward strand; the resultant mutation is shown together with the number (in brackets) of independent records in the database.
Table 5 CpG transitions in the PAX6 open reading frame.
Nucleotide change Codon change Location Mutation type Reports
414G>A ACG>ACA (G18R) PD exon 5 Missense 1
494C>T CGA>TGA (R44X) PD exon 5 Nonsense 1
493G>A CGA>CAA (R44Q) PD exon 5 Missense 1
669C>T CGA>TGA (R103X) PD exon 6 Nonsense 3
744C>T CGC>TGC (R128C) PD exon 7 Missense 2
782C>T GAC>GAT (D140D) LNK exon 7 Neutral 1
969C>T CGA>TGA (R203X) LNK exon 8 Nonsense 15
984C>T CGG>TGG (R208W) LNK exon 8 Missense 1
985G>A CGG>CAG (R208Q) LNK exon 8 Missense 1
1080C>T CGA>TGA (R240X) HD exon 9 Nonsense 21
1143C>T CGA>TGA (R261X) HD exon 10 Nonsense 8
1311C>T CGA>TGA (R317X) PST exon 11 Nonsense 16
PD, paired domain; LNK, linker region; HD, homeodomain; PST, proline/serine/threonine-rich region. 'Reports' indicates how many times the mutation has been reported in the PAX6 Allelic Variant Database.
Sense-strand deamination of CpG in an arginine codon CGA creates a termination codon TGA. The PAX6 ORF contains six CGA codons and all of these have been 'hit' at least once to give nonsense mutations (Table 5). It is noticeable that four CpG's in CGA codons in exons 8, 9, 10 and 11 (R203X, R240X, R261X and R317X) are a major source of nonsense mutations (Table 5, Figure 3). Together these four CpG's have been hit 60 times. These hits all cause aniridia and account for 21% of all mutations in the database and 59% of all nonsense mutations.
The observation that CpG's in exons 8, 9, 10 and 11 are a major source of mutations can be explained at least in part by the nucleotide composition and methylation status of the genomic PAX6 gene. The 5' two-thirds of the gene (from the promoter up to and including exon 7) are part of an unusually large CpG island. This region of the gene is very GC-rich and has a high frequency of CpG dinucleotides, most of which are unmethylated [31]. The last third of the gene, containing exons 8–13, is more similar to bulk genomic DNA, with a lower GC content. The frequency of CpG dinucleotides is low, but those that exist tend to be methylated, and methylation greatly increases the frequency of spontaneous deamination of cytosine, resulting in C>T transition [30,31]. Although only 13 of the 45 CpG's are in exons 8–13 (Figure 3), these are in the methylated region of the gene and are therefore much more likely to be 'hit'.
C-terminal truncating mutations are not associated with more severe phenotypes
When the first PAX6 mutations were discovered in aniridia patients, it quickly became apparent that mutations that introduce a PTC into the open reading frame are common [3-6]. Speculation arose that mutations causing translational termination after the homeodomain might yield dominant negative forms of the PAX6 protein, because truncated proteins containing only the DNA binding domains could theoretically bind to target DNA sequences without activating downstream genes and hence interfere with the function of the normal PAX6 protein [21-23]. A variety of studies then demonstrated that PAX6 proteins with C-terminal deletions do indeed have dominant negative activity [21,22,24]. It might therefore be expected that individuals with truncating mutations in the PST domain would have very low levels of normal PAX6 activity and this could result in a phenotype more severe than, or markedly different from, individuals with truncating mutations before the PST domain.
Mutations that introduce a PTC – nonsense mutations, splicing mutations and frame-shifting insertions and deletions – are scattered throughout the PAX6 open reading frame (Table 2). We examined the database records for evidence that late-terminating mutations (in the PST domain) cause different phenotypes.
Of 151 mutations that introduce a PTC before the PST domain (ie in the paired box, the linker region or the homeobox), 150 are associated with aniridia or closely related variants such as partial aniridia and iris hypoplasia. The remaining mutation is associated with optic nerve hypoplasia [15].
Of 43 mutations that introduce a PTC into the PST domain, 41 are associated with aniridia or closely related phenotypic variants. The remaining two mutations are associated with keratitis [16] and congenital cataracts [11], phenotypes that clearly overlap with aniridia. Therefore there is no evidence from the existing records that truncating mutations in the PST domain are associated with more severe phenotypes. Rather the data suggest that truncating mutations are overwhelmingly associated with aniridia regardless of their location in the gene.
How can the uniformity of patient phenotypes be reconciled with experimental data demonstrating dominant negative effects? One explanation is that the dominant negative tests may not be physiologically relevant. The experiments were performed on intronless cDNA constructs that terminate at an engineered stop codon and are designed to produce large quantities of the truncated protein. In contrast the patients have PTCs that occur in the context of an intact PAX6 gene. Once transcribed, PTC-containing RNAs are likely to be degraded by nonsense-mediated decay, a universal mechanism for preventing the accumulation of truncated proteins [19,20]. Nonsense mediated decay is inextricably linked to the synthesis, processing, splicing and preliminary translation of mRNAs derived from genomic genes [19,20]. Experimental cDNA constructs bypass these mechanisms to direct the synthesis of high levels of proteins that would not normally be made in the cell [19].
We propose that the dominant negative tests do not accurately reflect the in vivo consequence of truncating mutations. The simplest interpretation of the phenotypic data is that nonsense-mediated decay acts on most PTC-containing RNAs to generate null alleles.
The existing data are entirely consistent with the hypothesis that aniridia is a true haploinsufficiency phenotype, caused by loss of function of one allele, either by deletion or intragenic mutation. As mentioned previously, 77% of all aniridia-associated mutations in the database result in the introduction of a PTC. Therefore nonsense-mediated decay may be the major mechanism by which PAX6 null alleles are generated.
Absence of nonsense mutations at the 3' end of the PAX6 coding region
Any hypothesis concerning the role of nonsense-mediated decay must take into account the observation that it does not act on the 3' extreme of a coding region. The surveillance mechanism uses intron/exon boundaries as cues for detection of PTCs and typically does not operate on the last exon, or the last 50 bases of the penultimate exon [19,32]. In the PAX6 gene, the zone that would escape NMD would encompass the last 50 bp of exon 12 (from base 1496 onwards) and the first 83 bases of exon 13 up to the normal stop codon. This corresponds to the last 44 codons of the open reading frame (Figure 4). Truncating mutations in this region of the PAX6 gene should not be acted on by NMD and could theoretically generate dominant negative forms of the protein. In experimental assays, even a short reduction of 37 amino acids gave potent dominant negative effects [21].
Figure 4 Absence of nonsense mutations at the 3' end of the PAX6 coding region. The PAX6 open reading frame is represented as a horizontal rectangle; the untranslated regions are shown as thick black lines (not to scale). Exon boundaries are shown as vertical lines. PB, paired box; LNK, linker region; HB, homeobox; PST, PST region. Above the cDNA, thick double-headed arrows divide the coding region into two parts. Between bases 363–1495, nonsense-mediated decay is predicted to act on truncating mutations. The region from bases 1496–1628 is predicted to escape nonsense-mediated decay. The number of potential and observed nonsense mutations in the different zones of the coding region is shown in the lower part of the figure. No nonsense mutations have been observed in the region that escapes NMD, even though 14 codons could potentially give rise to nonsense mutations.
We inspected the database records to see what kinds of mutations are present in the region that is predicted to escape NMD (base 1496 onwards). Strikingly there are no nonsense mutations in this region (Table 6). There are four splicing mutations and five frame-shifting deletions, but the predicted consequence of all of these is run-on translation into the 3'UTR rather than introduction of a PTC [5,14,33,35,36]. This is in sharp contrast to the 5' part of exon 12 (up to base 1495), which contains a variety of nonsense and frame-shifting mutations, all of which are predicted to introduce a PTC (Table 6).
Table 6 Outcome of potential PTC-creating mutation in exons 12 and 13.
Mutation Location Consequence of mutation Type NMD predicted
1399delC Exon 12 PTC at TGA starting at base 1453 FSD Yes
1410delC Exon 12 PTC at TGA starting at base 1453 FSD Yes
1420C>G (S353X) Exon 12 Immediate PTC N Yes
1424C>G (Y354X) Exon 12 Immediate PTC N Yes
1426ins5 Exon 12 PTC at TGA starting at base 1469 FSI Yes
1452delG Exon 12 PTC at TGA starting at base 1453 FSD Yes
1469T>G (Y369X) Exon 12 Immediate PTC N Yes
1513del4 Exon 12 Run-on into 3'UTR FSD No
1520delGGinsT Exon 12 Run-on into 3'UTR FSD No
1545G>C Exon 12 Run-on into 3'UTR S No
IVS12+1del3 Intron 12 Run-on into 3'UTR S No
IVS12+4A>G Intron 12 Run-on into 3'UTR S No
IVS12+5G>A Intron 12 Run-on into 3'UTR S No
1562delT Exon 13 Run-on into 3'UTR FSD No
1601delT Exon 13 Run-on into 3'UTR FSD No
1615del10 Exon 13 Run-on into 3'UTR FSD No
This table shows the outcome of every known PAX6 mutation from the start of exon 12 onwards that could potentially introduce a premature termination codon. Mutations up to and including nucleotide 1495 are within the region in which nonsense-mediated decay is predicted to occur, while mutations from nucleotide 1496 onwards are predicted to escape nonsense-mediated decay. IVS, intervening sequence (intron); FSD, frame-shifting deletion; FSI, frame shifting insertion; N, nonsense mutation; S, splice mutation. PTC, premature termination codon; 3' UTR, 3' untranslated region.
Thus there appears to be an absence of truncating mutations in the part of the gene that is predicted to escape NMD. To investigate this further, we looked at the distribution of potential nonsense codons in the PAX6 coding region. The PAX6 open reading frame contains 132 codons that could be mutated to stop codons by a single base change. 102 nonsense mutations have been observed in patients and these occur at 34 of the possible 132 sites (Figure 4).
Fourteen of the potential nonsense codons lie within the region predicted to escape NMD (base 1496 onwards) but none of the resultant mutations has been observed to date (Figure 4). Assuming a random distribution of all unique nonsense mutations along the potential sites at which a nonsense mutation could occur, the probability of observing zero mutations in the non-surveillance zone is 0.012 calculated using Fisher's exact test (Figure 4). Thus the absence of mutations in exon 13 and the last 50 bp of exon 12 is unlikely to have arisen by chance.
As mentioned above, 9 different frame-shifting and splice mutations have been reported in the region predicted to escape NMD but none of these introduces a PTC. Rather they are all predicted to cause run-on translation into the 3' untranslated region (Table 6) [5,7,14,33-37].
The phenotypes associated with run-on mutations are well documented because the database contains details of 14 run-on mutations in which the normal termination codon is altered to a coding codon. All of these patients have aniridia or ocular defects within the aniridia spectrum such as iris hypoplasia, foveal hypoplasia, cataracts and nystagmus.
Thus translation beyond the normal stop codon is consistently associated with an aniridia-like phenotype, which suggests that run-on mutations generate simple loss-of-function alleles. Nonsense-mediated decay would not be predicted to act on such alleles because there is no PTC; therefore the proposed loss of function may result from the addition of an extra peptide at the C-terminal end of the PAX6 protein. The C-terminus of PAX6 is highly conserved and appears to play a role in the stabilisation of DNA binding by the homeodomain [12], so any disruption to the structure of the C-terminal region could have profound effects on the function of the PAX6 protein. It should however be emphasised that the function of run-on PAX6 proteins has not yet been tested; therefore confirmation of the hypothesis that run-on proteins show loss of function rather than dominant negative activity awaits further experimentation.
The known PAX6 mutation spectrum is devoid of mutations that introduce a PTC into the non-surveillance zone. Such mutations must surely arise, yet clearly they are not associated with aniridia. Given the evidence that even short truncations of the PAX6 protein cause strong dominant negative effects [21], we propose that termination mutations in the last part of the gene cause phenotypes significantly more severe than aniridia. These phenotypes may resemble that of the only confirmed case of an individual with a lethal compound heterozygous PAX6 mutation and may include anophthalmia, arhinia and severe central nervous system defects [11].
Conclusion
We have reviewed the mutations in the PAX6 Allelic Variant Database. Aniridia is typically caused by mutations that introduce a PTC, while non-aniridia phenotypes are cause by missense mutations. Transitions at four CpG dinucleotides in the methylated part of the gene make a major contribution to the burden of PAX6 nonsense mutations.
We have reasoned that nonsense-mediated decay acts on PAX6 mutant alleles. Mutations that introduce a PTC are consistently associated with aniridia or closely related phenotypes, regardless of where they occur in the gene. There is a statistically significant absence among the existing records of nonsense mutations in exon 13 and the last 50 bp of exon 12, where NMD would not be predicted to act.
Mutations that introduce a termination codon before the last 50 bases of exon 12 are likely to be acted on by nonsense-mediated decay and are probably functionally null. However mutations that introduce a termination codon within the last 50 bases of exon 12, or within exon 13, are likely to generate proteins with significant dominant negative activity. These mutations are not associated with aniridia and may cause very severe phenotypes that have not yet been ascertained. The effect of NMD on phenotypic severity has been experimentally demonstrated for SOX10 and MPZ, mutations of which cause neurochristopathies and myelinopathies respectively [38]. In SOX10 and MPZ, truncating mutations at the 3' end of the open reading frame escape NMD and generate dominant negative proteins that cause much more severe phenotypes than more 5' mutations [38].
The mutation spectrum of a gene can yield important insights into the molecular mechanisms that act on mutant alleles and the phenotypes that are likely to be associated with mutations in that gene.
Methods
PAX6 cDNA reference sequence and numbering
The PAX6 cDNA sequence used in this paper is taken from the PAX6 Allelic Variant Database [25]. The coding region runs from base 363 (in exon 4) to base 1628 (in exon 13). The PST region extends from base 1169 to base 1628.
PAX6 mutations and phenotypes
Data on the PAX6 mutation spectrum were collected from the PAX6 Allelic Variant Database. Only pathological mutations were considered, either within the coding region of the gene (bases 363–1628) or within the consensus splice acceptor and donor sequences of introns. Coding region polymorphisms and intronic changes outside the splice consensus sequences were not considered. Each mutation was placed into one of six categories (nonsense, splicing, frame-shifting insertion/deletion, in-frame insertion/deletion, missense and run-on – see Table 1 for definitions). There are twelve compound mutations in the database, each apparently involving more than one mutational event, such as an insertion and a deletion. These were categorised according to the final consequence of the mutation. For example the compound mutation 495delAinsCAT has a net effect of inserting two bases and is therefore categorised as a frame-shifting mutation.
Distribution of CpG dinucleotides and potential termination codons
The Mutability program determines the number and distribution of CpG dinucleotides and the number and distribution of potential nonsense and missense mutations arising from single nucleotide substitutions in a cDNA sequence. It is freely available through the PAX6 Allelic Variant Database web site [26]. We used the Mutability program to determine the location of CpG dinucleotides in the PAX6 open reading frame, and to determine the total number of codons that could potentially be mutated to a stop codon by a single nucleotide change [26]. The analysis was carried out on the complete coding region of PAX6, including exon 5a. For exons 4 and 13, which contain the initiation and termination codons respectively, only the coding DNA was considered.
Fisher's exact test
The number of observed nonsense mutations in the coding region was obtained from the PAX6 Allelic Variant Database. The number of nonsense mutations that were not observed was calculated by subtracting the number of observed mutations from the number of potential mutations (calculated using the Mutability program, above). The 'non-surveillance zone' of the coding region, in which RNA surveillance should not act was defined as the coding region of exon 13 and the last 50 bp of exon 12, ie from nucleotide c.1496 onwards. Application of Fisher's exact test to the 2 × 2 table shown in Figure 4 [34, 84, 0, 1] gives a one-tailed probability of p = 0.012, which is significant beyond the 5% level and allows rejection of the null hypothesis that the observed nonsense mutations are randomly distributed throughout the coding region. The Fisher's exact test calculation was carried out using the tool at [27]
Abbreviations
NMD, nonsense-mediated decay; ORF, open reading frame; PTC; premature termination codon.
Authors' contributions
IT carried out the bioinformatic analyses. IMSW advised on and verified the statistical analysis. IMH conceived of and supervised the study, and prepared the manuscript. All authors read and approved the final version.
Acknowledgements
IMH gratefully acknowledges the support of the UK Medical Research Council in the form of a Career Development Award. We thank Dr Alastair Brown for the Mutability program.
==== Refs
Ton CCT Hirvonen H Miwa H Weil MM Monaghan P Jordan T van Heyningen V Hastie ND Meijers-Heijboer H Drechsler M Royer-Pokora B Collins F Swaroop A Strong LC Saunders GF Positional cloning and characterization of a paired-box and homeobox-containing gene from the aniridia region Cell 1991 67 1059 1074 1684738 10.1016/0092-8674(91)90284-6
Crolla JA van Heyningen V Frequent chromosome aberrations revealed by molecular cytogenetic studies in patients with aniridia Am J Hum Genet 2002 71 1138 1149 12386836 10.1086/344396
Jordan T Hanson I Zaletayev D Hodgson S Prosser J Seawright A Hastie N van Heyningen V The human PAX6 gene is mutated in two patients with aniridia Nat Genet 1992 1 328 332 1302030 10.1038/ng0892-328
Glaser T Walton DS Maas RL Genomic structure, evolutionary conservation and aniridia mutations in the human PAX6 gene Nat Genet 1992 2 232 239 1345175 10.1038/ng1192-232
Hanson IM Seawright A Hardman K Hodgson S Zaletayev D Fekete G van Heyningen V PAX6 mutations in aniridia Hum Mol Genet 1993 2 915 920 8364574
Martha A Ferrell RE Mintz-Hittner H Lyons LA Saunders GF Paired box mutations in familial and sporadic aniridia predicts truncated aniridia proteins Am J Hum Genet 1994 54 801 811 7909985
Vincent MC Pujo AL Olivier D Calvas P Screening for PAX6 gene mutations is consistent with haploinsufficiency as the main mechanism leading to various ocular defects Eur J Hum Genet 2003 11 163 169 12634864 10.1038/sj.ejhg.5200940
Simpson TI Price DJ Pax6; a pleiotropic player in development Bioessays 2002 24 1041 1051 12386935 10.1002/bies.10174
Wilson DS Guenther B Desplan C Kuriyan J High resolution crystal structure of a paired (Pax) class cooperative homeodomain dimer on DNA Cell 1995 82 709 719 7671301 10.1016/0092-8674(95)90468-9
Xu HE Rould MA Xu W Epstein JA Maas RL Pabo CO Crystal structure of the human Pax6 paired domain-DNA complex reveals specific roles for the linker region and carboxy-terminal subdomain in DNA binding Genes Dev 1999 13 1263 1275 10346815
Glaser T Jepeal L Edwards JG Young SR Favor J Maas RL PAX6 gene dosage effect in a family with congenital cataracts, aniridia, anophthalmia and central nervous system defects Nat Genet 1994 7 463 471 7951315 10.1038/ng0894-463
Singh S Chao LY Mishra R Davies J Saunders GF Missense mutation at the C-terminus of PAX6 negatively modulates homeodomain function Hum Mol Genet 2001 10 911 918 11309364 10.1093/hmg/10.9.911
Brown A McKie M van Heyningen V Prosser J The human PAX6 mutation database Nucleic Acids Res 1998 26 259 264 9399848 10.1093/nar/26.1.259
The PAX6 Allelic Variant Database
Azuma N Yamaguchi Y Handa H Tadokoro K Asaka A Kawase E Yamada M Mutations of the PAX6 gene detected in patients with a variety of optic-nerve malformations Am J Hum Genet 2003 72 1565 1570 12721955 10.1086/375555
Mirzayans F Pearce WG MacDonald IM Walter MA Mutation of the PAX6 gene in patients with autosomal dominant keratitis Am J Hum Genet 1995 57 539 548 7668281
van Heyningen V Williamson KA PAX6 in sensory development Hum Mol Genet 2002 11 1161 1167 12015275 10.1093/hmg/11.10.1161
Prosser J van Heyningen V PAX6 mutations reviewed Hum Mutat 1998 11 93 108 9482572 10.1002/(SICI)1098-1004(1998)11:2<93::AID-HUMU1>3.0.CO;2-M
Byers PH Killing the messenger: new insights into nonsense-mediated mRNA decay J Clin Invest 2002 109 3 6 11781342 10.1172/JCI200214841
Culbertson MR RNA surveillance: unforeseen consequences for gene expression, inherited genetic disorders and cancer Trends Genet 1999 15 74 80 10098411 10.1016/S0168-9525(98)01658-8
Singh S Tang HK Lee JY Saunders GF Truncation mutations in the transactivation region of PAX6 result in dominant-negative mutants J Biol Chem 1998 273 21531 21541 9705283 10.1074/jbc.273.34.21531
Duncan MK Cvekl A Li X Piatigorsky J Truncated forms of Pax-6 disrupt lens morphology in transgenic mice Invest Ophthalmol Vis Sci 2000 41 464 473 10670477
Yasuda T Kajimoto Y Fujitani Y Watada H Yamamoto S Watarai T Umayahara Y Matsuhisa M Gorogawa S Kuwayama Y Tano Y Yamasaki Y Hori M PAX6 mutation as a genetic factor common to aniridia and glucose intolerance Diabetes 2002 51 224 230 11756345
Chow RL Altmann CR Lang RA Hemmati-Brivanlou A Pax6 induces ectopic eyes in a vertebrate Development 1999 126 4213 4222 10477290
PAX6 Composite cDNA Sequence
Mutability
Fisher's Exact Test
Tang HK Chao LY Saunders GF Functional analysis of paired box missense mutations in the PAX6 gene Hum Mol Genet 1997 6 381 386 9147640 10.1093/hmg/6.3.381
Hanson I Churchill A Love J Axton R Moore T Clarke M Meire F van Heyningen V Missense mutations in the most ancient residues of the PAX6 paired domain underlie a spectrum of human congenital eye malformations Hum Mol Genet 1999 8 165 172 9931324 10.1093/hmg/8.2.165
Nachman MW Crowell SL Estimate of the mutation rate per nucleotide in humans Genetics 2000 156 297 304 10978293
Cross SH Clark VH Bird AP Isolation of CpG islands from large genomic clones Nucleic Acids Res 1999 27 2099 2107 10219082 10.1093/nar/27.10.2099
Hentze MV Kuzolik AE A perfect message: RNA surveillance and nonsense-mediated decay Cell 1999 96 307 310 10025395 10.1016/S0092-8674(00)80542-5
Hanson I Brown A van Heyningen V A new PAX6 mutation in familial aniridia J Med Genet 1995 32 488 489 7666404
Gronskov K Rosenberg T Sand A Brondum-Nielsen K Mutational analysis of PAX6: 16 novel mutations including 5 missense mutations with a mild aniridia phenotype Eur J Hum Genet 1999 7 274 286 10234503 10.1038/sj.ejhg.5200308
Heyman I Frampton I van Heyningen V Hanson I Teague P Taylor A Simonoff E Psychiatric disorder and cognitive function in a family with an inherited novel mutation of the developmental control gene PAX6 Psychiatr Genet 1999 9 85 90 10412187
Chao LY Mishra R Strong LC Saunders GF Missense mutations in the DNA-binding region and termination codon in PAX6 Hum Mutat 2003 21 138 145 12552561 10.1002/humu.10163
Mitchell TN Free SL Williamson KA Stevens JM Churchill AJ Hanson IM Shorvon SD Moore AT van Heyningen V Sisodiya SM Polymicrogyria and absence of pineal gland due to PAX6 mutation Ann Neurol 2003 53 658 663 12731001 10.1002/ana.10576
Inoue K Khajavi M Ohyama T Hirabayashi S Wilson J Reggin JD Mancias P Butler IJ Wilkinson MF Wegner M Lupski JR Molecular mechanism for distinct neurological phenotypes conveyed by allelic truncating mutations Nat Genet 2004 36 361 369 15004559 10.1038/ng1322
| 15918896 | PMC1156885 | CC BY | 2021-01-04 16:38:18 | no | BMC Genet. 2005 May 26; 6:27 | utf-8 | BMC Genet | 2,005 | 10.1186/1471-2156-6-27 | oa_comm |
==== Front
BMC GenetBMC Genetics1471-2156BioMed Central London 1471-2156-6-281591891110.1186/1471-2156-6-28Methodology ArticleUse of SNPs to determine the breakpoints of complex deficiencies, facilitating gene mapping in Caenorhabditis elegans Kadandale Pavan [email protected] Brian [email protected] Melissa [email protected] Andrew [email protected] Waksman Institute of Microbiology, Rutgers University, 190 Frelinghuysen Road, Piscataway, NJ 08854, USA2005 26 5 2005 6 28 28 17 2 2005 26 5 2005 Copyright © 2005 Kadandale et al; licensee BioMed Central Ltd.2005Kadandale et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Genetic deletions or deficiencies have been used for gene mapping and discovery in various organisms, ranging from the nematode Caenorhabditis elegans all the way to humans. One problem with large deletions is the determination of the location of their breakpoints. This is exacerbated in the case of complex deficiencies that delete a region of the genome, while retaining some of the intervening sequence. Previous methods, using genetic complementation or cytology were hampered by low marker density and were consequently not very precise at positioning the breakpoints of complex deficiencies. The identification of increasing numbers of Single Nucleotide Polymorphisms (SNPs) has resulted in the use of these as genetic markers, and consequently in their utilization for defining the breakpoints of deletions using molecular biology methods.
Results
Here, we show that SNPs can be used to help position the breakpoints of a complex deficiency in C. elegans. The technique uses a combination of genetic crosses and molecular biology to generate robust and highly reproducible results with strong internal controls when trying to determine the breakpoints of deficiencies. The combined use of this technique and standard genetic mapping allowed us to rapidly narrow down the region of interest in our attempts to clone a gene.
Conclusion
Unlike previous methods used to locate deficiency breakpoints, our technique has the advantage of not being limited by the amount of starting material. It also incorporates internal controls to eliminate false positives and negatives. The technique can also easily be adapted for use in other organisms in which both genetic deficiencies and SNPs are available, thereby aiding gene discovery in these other models.
==== Body
Background
Chromosomal deletions (also called genetic deficiencies) have been used to help narrow down the location of genes in organisms ranging from C. elegans to human beings [1-4]. In the invertebrate model system, Drosophila melanogaster, deficiencies have been particularly useful for gene mapping, since the existence of polytene chromosomes and cytogenetics in this organism allows the actual visualization of the breakpoints of the deficiencies. In other organisms, however, mapping the breakpoints of deficiencies is not so straightforward. In C. elegans, for example, deficiencies are defined as deletion mutations that fail to complement multiple loci with the underlying assumption that the deletion would also remove the regions between these loci. Although this is generally true, deficiencies exist that are complex deletions, eliminating two distant loci, but retaining the intervening sequence (Figure 1). An example of such a complex deficiency is shown in Table 1. Genetic complementation tests [5] show that although the deficiency sDf62 deletes a region from map position 3.99 to 5.78 of chromosome IV, it fails to delete some genes that lie within this map interval (e.g. unc-43 and lin-3). In fact, our searches on WormBase [5] reveal that a surprising number of deficiencies in C. elegans are of a complex nature. When we examined a random sampling of 20 deficiencies from WormBase [5] we found that 9 of these were complex deficiencies. Although 20 is not, by any means, a large sample, it indicates that a sizeable number of deficiencies in C. elegans might be complex. This, coupled with the relative scarcity of usable genetic markers can confound the interpretation of deficiency data.
Figure 1 Difference between simple and complex deficiencies. Black bars show normal chromosomes, while grey bars represent hypothetical deficiency chromosomes. A complex deficiency deletes a large chromosomal region, but retains a small part within the larger deleted region, complicating the analysis of deficiency mapping data.
Table 1 Genetic complementation results for sDf62, showing it to be complex
Position Gene Result
3.99 ± 0.003 eat-1 +
4.58 ± 0.001 unc-43 -
4.66 ± 0.002 him-8 -
4.67 ± 0.033 let-71 -
4.81 ± 0.011 lin-3 -
4.81 ± 0.024 let-70 +
4.81 ± 0.035 let-64 -
4.82 ± 0.017 sem-3 +
4.82 ± 0.017 let-59 +
4.82 ± 0.033 let-98 +
4.82 ± 0.051 let-100 +
4.82 ± 0.069 let-73 +
4.83 ± 0.042 dif-2 +
4.86 ± 0.069 let-311 +
4.96 ± 0.069 let-655 +
4.97 ± 0.069 let-91 +
5.00 ± 0.046 rib-1 +
5.18 ± 0.003 dpy-20 +
5.43 ± 0.012 unc-22 +
5.51 ± 0.038 pha-3 +
5.73 ± 0.035 let-68 +
5.75 ± 0.053 let-656 +
5.78 ± 0.037 let-309 +
6.01 ± 0.016 let-99 -
6.13 ± 0.083 let-97 -
7.97 ± 0.010 unc-30 -
A "+" result indicates non-complementation, while a "-" indicates complementation. Loci in bold are cloned genes. Table derived from WormBase [5].
In order to address the problem of deletion breakpoint mapping in C. elegans, previous work [6] has attempted to use PCR to determine the breakpoints of such deficiencies (Figure 2). Large deletions in the homozygous state are usually lethal to the developing C. elegans embryo. Thus, from a plate of worms heterozygous for the deletion, one can pick individual dead eggs that have been laid, assuming that these must be homozygous for the deletion. One can then use PCR to test whether a particular region of interest is present in such samples, thereby determining the extent of the deletion. This method, which relies on a negative result (i.e., lack of a PCR product), has several drawbacks: (a) The amount of starting material is restricted to a single, or few dead eggs. (b) Chitinase treatment is required to remove the egg shell. (c) False positives can result if the PCR merely fails to work properly on the limited sample material. (d) False negatives can result if an embryo dies for reasons other than that it was homozygous for the deletion.
Figure 2 PCR can be used to identify deletion breakpoints. Schematic showing previous methods to identify the breakpoints of deletions using molecular biology tools. A heterozygous deficiency strain is allowed to fertilize itself. A fourth of the resulting embryos will be dead, since the homozygous deficiency is lethal. These dead eggs are collected and used for PCR to test for the presence of regions of interest.
Another method that could be used to achieve the same result is Southern Blotting [7]. However, this method is tedious, involving the preparation of high molecular weight genomic DNA, running gels and probe hybridization, and requires a larger starting amount of DNA.
In human research, taking advantage of the growing number of SNPs identified in the human genome, approaches are being developed to use SNPs to identify regions of loss of heterozygosity (LOH), and/or deletions in chromosomes [8]. The availability of a strain of C. elegans, CB4856, which is predicted to harbor about 1 SNP per kilobase of genomic sequence [9,10] when compared to the wild-type strain (N2), led us to believe that we could adopt similar methods for use in C. elegans, thereby overcoming the limitations of using deficiencies in this model system. In fact, the C. elegans SNP database already contains about 6,000 confirmed and predicted SNPs between the N2 and CB4856 strains [11], which corresponds to a SNP density of about 1 per 15 Kb of sequence. As the use of SNPs to map genes increases [12,13], we expect that many more SNPs will be discovered, making their use even more attractive. We report here the use of SNP mapping, in conjunction with traditional genetic deletion mapping to obtain a better map of the breakpoints of a complex deficiency, thereby greatly assisting the process of gene mapping and characterization.
Results and Discussion
Identification of ozDf2 as a complex deficiency
We encountered a complex deficiency while trying to map the gene spe-19 in C. elegans. Initial mapping put the gene within the interval deleted by the large deficiency, ozDf1 (spanning the region from +13.18 to +25.11 of chromosome V). Standard SNP mapping analysis [12] placed spe-19 within the region bounded by SNP M162 (map position +23) on the left and SNP ZC15 (map position +25) on the right (BG, manuscript in review). According to data from WormBase [5], this region is expected to be deleted (Figure 3) by the smaller deficiency, ozDf2 (which deletes the region from +23.5 to +25.11 of chromosome V by genetic complementation tests). When we performed the genetic complementation test, we found, to our surprise, that ozDf2/spe-19 animals were fertile suggesting that ozDf2 does not delete spe-19. We therefore hypothesized that ozDf2 was a complex deficiency (Figure 4).
Figure 3 Schematic representation of the chromosomal region around spe-19. The grey bar represents a normal chromosome. The relative positions of the various SNPs and loci mentioned in the text are indicated. Note that the figure is not drawn to scale. The region expected to be deleted by the deletion ozDf2 is indicated by the white box labelled "ozDf2."
Figure 4 Genetic complementation test of ozDf2. (A) ozDf2 was tested by a standard genetic complementation test to check whether it uncovers spe-19. The ozDf2/spe-19 worms were expected to be sterile, but were fertile. Thus, the deletion complements spe-19 although it is predicted to delete spe-19. (B) This implies that ozDf2 must be a complex deficiency, retaining the spe-19 containing region, within a deletion that surrounds this locus.
Use of SNPs to map the breakpoints of ozDf2
In order to test whether ozDf2 was, indeed, a complex deficiency, and to map its breakpoints, we used a series of SNPs in the region [11] to obtain a higher resolution map of the deficiency. For this purpose, we used SNPs whose sequence change results in the loss of a restriction enzyme recognition site in the CB4856 animals. Thus, if we see any product that is digested, even partially, we can confirm the presence of the wild-type (strain N2) sequence. We generated the appropriate lines (see Methods and Figure 5) to use SNPs to map the deficiency breakpoints. We obtained animals that are heterozygous for the deficiency ozDf2, the homologous chromosome being from strain CB4856. Thus, in these animals, in chromosomal regions lost by the deletion ozDf2, we should see sequences corresponding only to the CB4856 SNP sequence. In regions not deleted by ozDf2, we expect both the wild-type (strain N2) sequence as well as CB4856 sequence (Figure 6). We show here the results from two SNPs – F38A6:14347 and Y113G7A:29658 – both of which are predicted to be deleted by ozDf2 (Figure 7). As expected, the heterozygous ozDf2/CB4856 animals showed only the uncut CB4856 band for SNP F38A6:14347. However, SNP Y113G7A:29658 shows both the uncut CB4856 band as well as the cut N2 bands, indicating that at this region, the N2 sequence was present (Figure 7). Since we expected the N2 sequence of Y113G7A:29658 to be absent in the ozDf2/CB4856 worms (the ozDf2 deletion is expected to also have lost the Y113G7A:29658 region – see Figure 3), the retention of N2 sequence around Y113G7A:29658 confirms that ozDf2 is a complex deficiency. Thus, ozDf2 deletes the region from +23.5 to +25.11 of chromosome V, but retains a part of this interval, around the SNP Y113G7A:29658. This result indicates that spe-19 must localize to the region around SNP Y113G7A:29658, since this area is retained in the complementing deficiency, ozDf2.
Figure 5 Generation of lines for SNP mapping of deletion breakpoints. The cross performed to obtain the lines used for the SNP mapping of the deficiency breakpoints is outlined. The details of the crosses are described in the Methods.
Figure 6 Schematic representation of the results expected for the SNP analysis. Schematic showing what banding patterns would be expected in SNPs that are deleted by the ozDf2 deficiency (hypothetical SNP 1) and in SNPs that are not (hypothetical SNP 2). The boxed region shows the hypothetical banding pattern expected on an agarose gel for the two SNPs. Note that SNPs are chosen so that for a given SNP, the wild-type (N2) sequence is digested by an appropriate restriction enzyme, whereas the CB4856 SNP sequence is not digested by the same enzyme.
Figure 7 Representative results of the SNP analysis. Representative results of SNP mapping of ozDf2's breakpoints. Lanes are labeled with the genotype of the strain used. Marker used was New England Biolabs 2-Log ladder. Note that SNP F38A6:14347 is deleted in ozDf2 (no N2 bands in ozDf2/CB4856), whereas SNP Y113G7A:29658 is not deleted (N2 bands seen in ozDf2/CB4856).
We have shown here that using SNPs as genetic markers in model systems greatly eases the problems of defining the breakpoints of deficiencies, both simple and complex. This allows us to use them as powerful tools in narrowing down the region of interest when cloning genes.
Conclusion
Our use of a combination of genetic complementation data along with SNP mapping of deficiency breakpoints has allowed us to rapidly narrow down the region of interest for cloning spe-19. Deficiency mapping in a number of organisms, including C. elegans, has been beset by the lack of a convenient method to correlate genetic complementation data with the physical map, confounding our ability to accurately define the breakpoints of deficiencies. Previous methods used to overcome this obstacle depended on a negative result [6,7], which is often untrustworthy. Using SNPs, not only is the correlation between the genetic and physical maps easily achieved, but we can also, potentially, generate higher resolution maps of the breakpoints of both simple, as well as complex deficiencies. Additionally, the amplification of the identical locus from the CB4856 chromosome serves as a good internal control, eliminating false positives and false negatives and allowing us to interpret our data with a very high degree of confidence. The fact that we can use heterozygous animals, without the need to generate homozygous, dead embryos overcomes the limitations of previous methods which relied exclusively on negative results, such as using PCR to detect the presence or absence of a sequence. We chose to use bulk worm lysates from a heterozygous line as our source of DNA. However, the technique would work just as well if single worms (after confirming their genotype) were used as the source of DNA. We believe that using lines has the following advantages:
1. The line can be frozen away and then rethawed at a later date if further analysis is required.
2. Multiple SNPs can be tested using the same lysate, as opposed to requiring multiple, single worm lysates.
3. Lysates can be prepared in bulk, in replicates, for independent and robust confirmation of all results obtained by this method.
4. No special care needs to be taken for preparation of DNA for SNP detection. DNA is made from the lines using a simple, scaled up version of the method used to make worm lysates for single worm PCR.
We initially chose to use SNPs that caused an abrogation of a restriction enzyme recognition site in the CB4856 sequence, thereby eliminating any artefacts due to incomplete restriction enzyme digestion. However, by running a control reaction to ensure that enzyme digestion has progressed to completion, one could also use SNPs that cause the loss of a restriction enzyme recognition site in the N2 sequence. Additionally, this method allows the use of SNPs that are only different in sequence, causing no changes in restriction enzyme recognition sites. A simple sequencing reaction can be done to check whether the Deficiency/CB4856 worms had only CB4856 sequence (which would be the case if the deficiency deletes the region of the SNP), or whether both N2 and CB4856 sequence was present (indicating that the SNP region is not lost in the deletion being tested). In addition to the CB4856 strain which we have used, other wild isolates of C. elegans which harbour SNPs [9] can also be used, expanding the possibilities of this technique.
While we have shown this method to work in the model system C. elegans, a similar approach can be applied in any organism where deficiencies/deletions exist and where SNP mapping is possible.
Methods
Culturing of C. elegans strains
Worms were grown and cultured according to standard protocols described elsewhere [14]. The strains used for this work were: spe-19, ozDf2/dpy-21 par-4, N2 (the wild-type) and CB4856.
Survey of deficiencies on WormBase
We looked at the complementation data for 20 deficiencies chosen at random. We ensured that deficiencies from all chromosomes were included in the analysis and that the deficiencies chosen had their left and right boundaries defined by cloned genes (this allowed us to set the other boundaries of the deficiency with certainty). We then looked at the complementation data and scored the deficiency as "complex" if at least one of the genes that were within the outer boundaries of the deficiency was complemented by the deficiency. While scoring the complex deficiencies, we made sure that the internal locus that was complemented was either cloned or had multiple, corroborating mapping data. If the deficiency failed to complement all the loci that were within the outer boundaries, it was scored as a simple deficiency. The deficiencies that we scored as simple are: nDf20, sDf10, mnDf16, nDf3, tDf1, mnDf1, ccDf1,eDf2, tDf5, ctDf2 and ctDf3. The complex deficiencies we found are: hDf10, hDf17, nDf19, nDf9, eDf18, qDf8, eDf3, maDf4 and nDf27.
Genetic complementation test of ozDf2
Since spe-19 males are fertile, we mated ozDf2/dpy-21 par-4 with spe-19 males. Of the outcross F1 progeny, we would expect half to be of the genotype ozDf2/spe-19 and consequently, sterile. The other half would be fertile, spe-19/dpy-21 par-4 worms. However, all the F1 progeny were fertile, indicating that the deletion ozDf2 complemented spe-19, belying our expectations. The presence of the deletion and of spe-19 was confirmed by looking at the F2 progeny. Half the F1 progeny segregated wild-type animals, dead eggs and sterile worms, indicating that their genotype must be ozDf2/spe-19. Thus, ozDf2 complements spe-19.
Generation of lines for SNP mapping of the breakpoints of ozDf2
We crossed ozDf2/dpy-21 par-4 hermaphrodites to CB4856 males. These ozDf2/dpy-21 par-4 heterozygotes are wild-type, but ozDf2 homozygous animals are embryonic lethal and dpy-21 par-4 homozygotes are short and fat – a phenotype referred to as Dumpy, or Dpy. F1 animals that segregated Dpy progeny were discarded (these would have been dpy-21 par-4/CB4856), and only the remaining, heterozygous animals (which are all ozDf2/CB4856) were used to set up lines that were further analyzed for their SNP sequences (Figure 5). Individual ozDf2/CB4856 worms were picked to separate culture plates. These worms are fertile and generate independent lines which were used for further analysis. Each line was generated within 6 days of picking the founder worm, ensuring that the N2 deficiency chromosome would not be underrepresented significantly in the line. Thus, each individual line would contain a mixture of CB4856 and ozDf2 chromosomes, which provide abundant material for the analysis of multiple SNP sequences. We established three independent heterozygous lines for analysis.
SNP mapping of the breakpoints of ozDf2
Crude genomic DNA was prepared from the three ozDf2/CB4856 lines by methods described previously [15]. We used the Vector NTI suite to design primers that amplified ~500 bp upstream and downstream of the SNP, resulting in a PCR product that was approximately 1 Kb. The details of the primers used and the PCR products generated are as follows:
SNP Y113G7A:29 – The expected product size is 1033 bp using the forward primer (GTAGGAAGTGGTGCCAGATGACGAGG) and the reverse primer (GTCTCAATTTGTGAGCGCATGTC).
SNP F38A6:14347 – The expected product size is 961 bp using the forward primer (TGGTTCCCCTCTACCAGATGCC) and the reverse primer (CCAACACAACTCAATCTAATGTTTACC).
We used these primers to amplify the corresponding product from each of the three genomic DNA preparations of the ozDf2/CB4856 lines using the following PCR conditions: denature at 94°C for 30 seconds, anneal at 60°C for one minute, extend at 72°C for 1 minute and repeat for 30 cycles. The PCR product was then digested with an appropriate restriction enzyme, and the products analysed on a 1.5% agarose gel run using the SB buffer. For SNP Y113G7A:29, XmnI was used to digest the PCR product and for SNP F38A6:14347, DraI was used.
Authors' contributions
PK conceived of using SNPs to analyse deletion breakpoints, carried out the SNP analysis and drafted the manuscript. All of the genetic complementation mapping was carried out by BG. MH aided with the molecular biology. AS participated in designing the experiments and in their coordination and helped to draft the manuscript. All work was conducted in the laboratory of AS.
Acknowledgements
This work was supported by grants from the NIH (R01 GM63089-01) and the Johnson & Johnson Discovery Award. PK was supported by the Charles and Johanna Busch pre-doctoral fellowship. The Caenorhabditis Genetics Center provided some nematode strains and it is funded by the NIH National Center for Research Resources (NCRR).
==== Refs
Bergstrom DE Bergstrom RA Munroe RJ Lee BK Browning VL You Y Eicher EM Schimenti JC Overlapping deletions spanning the proximal two-thirds of the mouse t complex Mamm Genome 2003 14 817 829 14724736 10.1007/s00335-003-2298-4
Demerec M Hoover ME Three related X-chromosome deficiencies in Drosophila Journal of Heredity 1936 27 206–12
Iovine MK Johnson SL A genetic, deletion, physical, and human homology map of the long fin region on zebrafish linkage group 2 Genomics 2002 79 756 759 12036288 10.1006/geno.2002.6769
Sigurdson DC Spanier GJ Herman RK Caenorhabditis elegans deficiency mapping Genetics 1984 108 331 345 6500256
Wormbase
Ellis RE Kimble J The fog-3 gene and regulation of cell fate in the germ line of Caenorhabditis elegans Genetics 1995 139 561 577 7713418
Browning H Berkowitz L Madej C Paulsen JE Zolan ME Strome S Macrorestriction analysis of Caenorhabditis elegans genomic DNA Genetics 1996 144 609 619 8889524
Wong KK Tsang YT Shen J Cheng RS Chang YM Man TK Lau CC Allelic imbalance analysis by high-density single-nucleotide polymorphic allele (SNP) array with whole genome amplified DNA Nucleic Acids Res 2004 32 e69 15148342 10.1093/nar/gnh072
Koch R van Luenen HG van der Horst M Thijssen KL Plasterk RH Single nucleotide polymorphisms in wild isolates of Caenorhabditis elegans Genome Res 2000 10 1690 1696 11076854 10.1101/gr.GR-1471R
Swan KA Curtis DE McKusick KB Voinov AV Mapa FA Cancilla MR High-throughput gene mapping in Caenorhabditis elegans Genome Res 2002 12 1100 1105 12097347
C. elegans Single Nucleotide Polymorphism Data
Jakubowski J Kornfeld K A local, high-density, single-nucleotide polymorphism map used to clone Caenorhabditis elegans cdf-1 Genetics 1999 153 743 752 10511554
Wicks SR Yeh RT Gish WR Waterston RH Plasterk RH Rapid gene mapping in Caenorhabditis elegans using a high density polymorphism map Nat Genet 2001 28 160 164 11381264 10.1038/88878
Brenner S The genetics of Caenorhabditis elegans Genetics 1974 77 71 94 4366476
Barstead RJ Hope IA Reverse genetics C elegans: A practical approach 1999 1 Oxford , Oxford University Press 97 118
| 15918911 | PMC1156886 | CC BY | 2021-01-04 16:38:18 | no | BMC Genet. 2005 May 26; 6:28 | utf-8 | BMC Genet | 2,005 | 10.1186/1471-2156-6-28 | oa_comm |
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-621587174510.1186/1471-2164-6-62Methodology ArticleIn vitro identification and in silico utilization of interspecies sequence similarities using GeneChip® technology Grigoryev Dmitry N [email protected] Shwu-Fan [email protected] Brett A [email protected] Rafael A [email protected] Shui Q [email protected] Joe GN [email protected] Center for Translational Respiratory Medicine, Division of Pulmonary and Critical Care Medicine, Johns Hopkins University, 5501 Hopkins Bayview Circle, JHAAC/4A.24, Baltimore, MD 21224, USA2 Department of Anesthesiology and Critical Care Medicine, Johns Hopkins University, 600 North Wolfe Street, Tower 711, Baltimore, MD 21287, USA3 Department of Biostatistics, Johns Hopkins University,615 N. Wolfe Street, E3035, Baltimore, MD 21205, USA2005 4 5 2005 6 62 62 9 8 2004 4 5 2005 Copyright © 2005 Grigoryev et al; licensee BioMed Central Ltd.2005Grigoryev et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Genomic approaches in large animal models (canine, ovine etc) are challenging due to insufficient genomic information for these species and the lack of availability of corresponding microarray platforms. To address this problem, we speculated that conserved interspecies genetic sequences can be experimentally detected by cross-species hybridization. The Affymetrix platform probe redundancy offers flexibility in selecting individual probes with high sequence similarities between related species for gene expression analysis.
Results
Gene expression profiles of 40 canine samples were generated using the human HG-U133A GeneChip (U133A). Due to interspecies genetic differences, only 14 ± 2% of canine transcripts were detected by U133A probe sets whereas profiling of 40 human samples detected 49 ± 6% of human transcripts. However, when these probe sets were deconstructed into individual probes and examined performance of each probe, we found that 47% of human probes were able to find their targets in canine tissues and generate a detectable hybridization signal. Therefore, we restricted gene expression analysis to these probes and observed the 60% increase in the number of identified canine transcripts. These results were validated by comparison of transcripts identified by our restricted analysis of cross-species hybridization with transcripts identified by hybridization of total lung canine mRNA to new Affymetrix Canine GeneChip®.
Conclusion
The experimental identification and restriction of gene expression analysis to probes with detectable hybridization signal drastically increases transcript detection of canine-human hybridization suggesting the possibility of broad utilization of cross-hybridizations of related species using GeneChip technology.
==== Body
Background
Genome-wide analyses of multiple organisms in a variety of experimental biological systems are powerful tools employed in biomedical research. However, microarray platforms for numerous species, particularly large mammals, have yet to be fully developed. Large mammals, often preferred species for modeling of various pathophysiological conditions and environmental responses, demonstrate better concordance to humans with regard to toxicological effects compared to small animal models (rodents) [1,2]. Further, large animal models are indispensable in disease modeling that requires longer live span [3], in organ transplant research [4], and in studies where multiple samples are required to be collected [5]. For organ- and system-specific applications, such as the respiratory system, canine and ovine lungs are physiologically and hemodynamically more closely related to human lungs than rodent lungs and allow regional evaluation of air, blood, and exudate distribution. Thus, canine [6,7] and ovine [8,9] models are the most commonly utilized in pathophysiological modeling of the respiratory system.
Despite these obvious advantages of the large mammal model, the lack of large mammal genomic sequence data presents significant limitation for genome-wide analysis of gene expression profiles by microarray techniques. To overcome this limitation, cross-species RNA hybridization has been utilized as one potential solution [10-13]. This approach is based on the notion that evolutionarily-conserved genetic sequences between mammals will be efficiently detected during cross-species hybridization. Despite a paucity of literature focused on the use of Affymetrix GeneChip oligonucleotide array for cross-species hybridizations [14,15], we speculated that this array platform could potentially serve as a powerful tool for cross-species genome analysis. GeneChips carry 11–16 oligonucleotide probes per target gene, increasing the opportunity to potentially identify genes with stretches of high homology to specific probes. To test this hypothesis, we cross-hybridized total RNA obtained from canine lung tissues to the human HG-U133A GeneChip. Analysis of generated expression profiles using Affymetrix MicroArray Suite (MAS 5.0) showed that approximately 14% of canine mRNA hybridizes to the human probe sets and generates hybridization signal similar to that of human mRNA. Analysis of available canine orthologues corresponding to these well performing human probes revealed high sequence similarities between human and canine counterparts. Based on these findings we speculated that even canine genes with only partial sequence similarities to their human counterparts can be identified by MAS 5.0 if the analysis is restricted to probes with high sequence similarity to their canine counterparts. Antipova et al. recently reported that utilization of a single well performing Affymetrix probe pair can be sufficient for detection of its corresponding target [16], further supporting our hypothesis.
A straight forward approach to test this hypothesis would be to align the target sequences from the human HG_U133A GeneChip to the known canine sequences using Basic Local Alignment Search Tool (BLAST), followed by identification of human probes with high sequence similarity to their canine counterparts. Unfortunately, due to limited canine genomic information, only 3% of human target sequences from the HG_U133A GeneChip have been matched to known canine orthologues, which renders this approach inadequate. However, we speculated that since canine orthologues with high sequence similarities to human probes generate detectable hybridization signals, the reverse should be also true: the detectable hybridization signal should reflect high sequence similarity. Therefore, we examined individual probes instead of complete probe sets. Probes with detectable hybridization signals were considered adequate representatives of their corresponding canine orthologues and were thus employed for final signal intensity calculations. The MAS 5.0 analysis when restricted to probes with detectable hybridization signals, demonstrated a 4-fold increase in canine transcript detection as compared to unrestricted analysis. Our results strongly suggest that identification of probes with detectable cross-species hybridization signal and subsequent restriction of gene expression analysis to these probes can be applied to multiple related species for which gene arrays are not commercially available.
Results
The hybridization pattern of the protein translocation complex beta gene involved in cell maintenance (SEC61B, Gene Ontology #8151) is assumed to be constantly expressed in the majority of tissues and was selected for demonstration of the probe level analysis. This analysis was conducted using signal intensity values generated by hybridization of human and canine SEC61B mRNAs to HG-U133A GeneChip. The probe level analysis deconstructs the entire Affymetrix probe set into its individual components and allows one to evaluate the performance of each individual probe. Detailed output of human SEC61B hybridization to 203133_at probe set (Figures 1A and 2A) shows that signal intensity value of PM (perfect match or total signal) obtained for 11th probe pair is higher than that of MM (mismatch or background signal). Therefore, correction for the background will produce positive signal intensity value, and this probe pair will contribute to the "Present" (detectable transcript) classification by MAS 5.0. However, cross-species hybridization of canine mRNA to the same probe pair does not produce sufficient signal intensity value (Figures 1C and 2B), possibly due to differences in target sequence between canine and human. This will, therefore, shift the overall probe set classification towards "Absent" (non-detectable transcript) classification. At the same time, there are probes that failed to detect both human and canine SEC61B. This could be due to either secondary structure formation (Figure 1B and 1D) [17] or the presence of sequences in the hybridizing mRNA mixture that possess homology to the 11th MM probe, thus raising the background signal over the specific signal. The pattern of SEC61B hybridization shows that, among 11 probe pairs of the 203133_at probe set, 3 probes (#2, #3, and #5) failed to detect its target in human tissues but the detectable hybridization of the remaining 8 probes prevailed (Figure 2A), resulting in classification of the SEC61B transcript by MAS 5.0 as "Present". In contrast, a reduction in the number of probes that were able to detect canine SEC61B down to only 6 probes (#1, ##6–10, Figure 2B) caused MAS 5.0 to classify this gene as "Absent". As shown in Figure 2C, the failure of the 5 non-hybridizing canine SEC61B probes was mainly attributed to differences in gene sequences between these two species.
Figure 1 Hybridization pattern of human (panels A and B) and canine (panels C and D) mRNAs to 2nd and 11th probe pairs of 203133_at Affymetrix probe set. The 203133_at probe set contains 11 probe pairs of which 2nd and 11th represented by anchored oligonucleotide sequences. Corresponding probe pairs identified at the bottom and comprised of perfect match probe (PM) and mismatch probe (MM). The purposely mutated 13th nucleotide in MM sequences is shown as "X". The target mRNAs represented by solid lines aligned to oligonucleotide sequences and mismatched regions depicted as humps. Numbers on the top of each target mRNA represent intensity values generated by MAS 5.0 (Figure 2 A and B). Dotted lines depict nucleotide bond involved in formation of predicted hairpin structure of the 2nd probe of the 203133_at probe set.
Figure 2 Hybridization patterns of human and canine SEC61B gene to Affymetrix 203133_at probe set and alignment of 203133_at probe pairs to canine SEC61B gene sequence. Panels A and B represent screen shots of MAS 5.0 generated intensities for "perfect match" probes (grey columns) and "mismatched' (open columns), respectively. Panel C depicts CLUSTALW generated alignment of canine SEC61B gene (accession number L25052) and 11 "perfect match" probes of Affymetrix 203133_at probe set. Star signs represent matched base pairs and double-headed arrows show locations of 203133_at probes (the corresponding probe number attached at the 5' end of probes). Panel D represent human SEC61B gene hybridization patterns throughout 40 GeneChip arrays. The order numbers of 203133_at probes are shown on the left. GeneChips were grouped by tested tissues (8 tissue groups with 5 chips per group) and are labeled on the top of each profile. Probe pairs were divided into three groups: poorly performing (open blocks, (PM - MM) ≤ 20), moderately performing (grey blocks, 20 < (PM - MM) ≤ 200), and well performing (solid blocks, (PM - MM) > 200). Panel E represent canine SEC61B gene hybridization patterns throughout 40 GeneChip arrays.
This observation held true for most of the evaluated targets, leading us to speculation that among the 11 human probes of a given probe set there will be probes with high sequence similarity to its canine target and that identification and utilization of these probes will increase sensitivity of cross-species hybridization. Moreover, when we evaluated performance of all 9,871,960 HG-U133A probes from 40 cross-species hybridizations, we found that almost half of these probes (4,688,566) generated a detectable hybridization signal (PM-MM>20). Therefore, we speculated that excellent and poorly performing probes should be equally distributed in the probe sets as well, and that masking of the poorly performing half will increase transcript detecting ability of these modified probe sets. The test run of MAS 5.0 analysis using masking of the 5 poorly performing probes of 203133_at probe set converted the initial "Absent" call for canine SEC61B gene to a "Present" call and will formulate the basis for further investigation. Since the canine genome is not yet completed, we were unable to select probes with high canine/human sequence similarity solely based on gene sequence information. Instead, we evaluated the overall performance of each individual probe based on their hybridization signal. We speculated that sequences of human probes with a reasonable cross-species hybridization signal (PM-MM>20) most likely possess a high similarity to their canine targets.
The probe level analysis was performed for 40 expression profiles generated by hybridization of 8 different human tissues and 40 expression profiles generated by hybridization of canine lung tissues to the human HG-U133A GeneChip. The resulting hybridization signal pattern of the 203133_at probe set for both species is presented in Figure 2D and 2E. As expected, differences in the canine SEC61B mRNA sequence region corresponding to the 11th probe (Figure 2C) did not allow canine target detection in 39 out of 40 hybridizations (Figure 2E). These data suggest that once a probe poorly hybridizes to its target in several initial experiments, it will most likely fail to detect its target in the substantial number of consequent hybridizations. To investigate whether this pattern persists throughout whole gene array, we analyzed hybridization patterns of all 246799 probes of the HG-U133A GeneChip. The resulting histogram for the same species hybridization (Figure 3A) demonstrates that a large fraction (~44% cases) of probes performed well during 38 or more hybridizations (95%). This trend was also true for poorly performing probes, where a considerable fraction (~11%) probes failed to detect their target during 38 or more hybridizations. In canine/human hybridization the number of poorly performing probes was higher (24%) while fraction of well performing probes lowered to 38% (Figure 3B). The analysis of the overall net effect of canine/human sequence differences on the hybridization pattern was conducted by correction for poorly performing probes during human samples hybridization. This revealed a 7% decrease in well performing and 13% increase in poorly performing probe fractions (Figure 3C). For detailed evaluation of direct effects of interspecies sequential differences on the cross-species hybridization, we selected probe sets where all probes generated detectable hybridization signal during human sample hybridization. Interestingly, among these probes, the fraction of probes that detected their corresponding canine targets during 38 or more cross-species hybridization was surprisingly high (~72%), with a very low (~3%) fraction of probes that failed to detect their targets in canine tissues (Figure 4). To evaluate correlation between performance of an individual probe and its alignment to the corresponding canine target, the number of mismatches were identified and counted for 20 randomly selected canine genes. This analysis demonstrated that, on average, probes in both excellent and poorly performing fractions had 1 and 2.5 mismatched bases, respectively, with the mismatch rate significantly lower in excellent performing probes (Figure 4, insert). The presence of a mismatch in probes with detectable hybridization signal was not unexpected and correlates with previously reported observation that relatively long (≥ 16 bp) stretch of the perfect match could produce stable hybridization and signal generation [15].
Figure 3 Histogram of probe performance during human and canine samples hybridizations to HG-U133A. The number of hybridizations that generated detectable signal for individual probe throughout 40 GeneChips was computed and histograms for 246799 analyzed probes were generated for human (panel A) and canine (panel B) samples. Panel C represents the net effect of canine sample on probe performance.
Figure 4 Effects of human/canine genomic differences on probe performance and correlation of probe performance with sequence similarity. Only well performed during human/human hybridization probes were employed for analysis of the effects of genetic sequence differences between human and canine. Two pools of transcripts were probes with non-detectable hybridization signal in 95% cases (generated detectable signal during 0–2 hybridizations) and probes that generated detectable signal in 95% cases (38–40 hybridizations). Randomly selected probes from each pool were aligned to their canine counterparts and number of mismatched base pairs was identified. The difference in detected mismatches between these probe pools was evaluated using Mann-Whitney test with p < 0.05 used as a significance cut-off (insert).
The criterion for masking of a particular probe was based on the performance of a given probe throughout 40 GeneChips. We found that masking of probes which poorly performed (PM-MM≤20) in 15 or more hybridizations (~40% of all hybridizations) produced the highest number of "Present" calls (Figure 5A) therefore this percentage was selected as a cut off value for masking. We speculated that probes that failed to detect its target in more than 40% of hybridizations lack homology to their canine counterparts at least in part due to sequence divergence. To test this hypothesis we conducted a sequence analysis of randomly selected samples and find that out of the 42 tested probes 27 probes aligned against canine sequence with one or more mismatches and 22 of these probes (approximately 82%) were masked by our procedure.
Figure 5 Effects of masking of poorly performed probes during human and canine hybridizations to HG-U133A on "Present" calls, average probe set size and composition. Panel A. MAS 5.0 was applied to human (open triangles) and canine (open circles) hybridization profiles without masking (NM) and with masking of probes that poorly performed in 40, ≥35, ≥30, etc. hybridizations (x axis) and resulting numbers of total "Present" calls generated for each array were averaged and expressed in percentage values (y axis). Average sizes of human (solid triangles) and canine (solid circles) probe sets were computed for corresponding masking conditions and expressed in percentage values (y axis). Panel B. Distribution of probe set sizes after masking probes that failed to generate detectable hybridization signal during 15 or more hybridizations (~40%) of human (triangles) and canine (circles) samples. Panel C. Distribution of masked probes according to location on the target sequence.
We also analyzed the effects of our procedure on the probe set size (Figure 5B). The main fractions of probe sets adjusted to human (77% of all probe sets) or canine (81% of all probe sets) hybridizations are represented by 5–10 and 4–8 probe pairs, respectively. Our masking procedure left 2172 (10%) probe-sets unaltered and completely eliminated 7 (0.03%) probe sets based on the human hybridization pattern. Analysis of canine hybridization identified 184 (0.8%) highly homologous canine targets, which hybridize to the entire corresponding human probe sets, and eliminated 11 probe-sets (0.05%) where none of the probes recognizes its target in canine tissues.
We next examined relative location of poorly performing human probes on the canine genes. It has been reported previously that human/canine homology is prone to be higher within the coding region [14]. Our findings are consistent with this report and show that our approach masked more probes towards the 3' untranslated region of target sequences (Figure 5C).
The effect of our masking procedure was also evaluated for transcript abundance detection. The combined (40 GeneChips) output of regular MAS 5.0 analysis was compared with outputs produced using masking files generated for human and canine hybridizations (Figure 6). In human hybridization, our algorithm reduced "Absent" calls by 3 fold and increased "Present" calls by more than 1.5 fold. The application of the masking file generated for canine hybridization had a more modest effect than human-specific masking. The transcript detection during the hybridization of canine mRNA to the human HG-U133A chip was much less sensitive (14%) than that of hybridization of human mRNA (48%). Our canine-specific masking recovered approximately 60% (Figure 6) of transcripts which, otherwise, were called "Absent" by MAS 5.0. The anticipated increase in transcript detection by application of human-specific masking is attributed to the masking of imperfectly designed probes. As shown in Figure 1, the 11th probe of the 203133_at probe set failed to detect its target in both canine and human tissues. Therefore, masking of such a probe will improve transcript detection in both canine and human sample analyses. Similarly, the canine-specific masking of human hybridization also increased the number of detectable transcripts as compared to non-masking results (Figure 6A).
Figure 6 Comparison of transcript-detecting ability before and after masking procedure. Solid bars represent percentage values of "Absent" transcript abundance calls. Open and grey bars represent "Marginal" and "Present" transcript abundance calls, respectively.
Finally, we compared transcripts detected by our masking procedure with those identified by hybridization of canine lung tissues to the newly developed Affymetrix canine GeneChip. We have identified 996 orthologous transcripts common to both HG-U133A and canine GeneChips and compared their detection in both arrays using "Present"/"Absent" classifications by MAS 5.0 or GCOS 1.1, respectively. This analysis showed that our masking approach increased the fraction of correctly identified transcripts by 46% and decreased the false negative fraction by 53% as compared to unmasked transcript detection (Figure 7).
Figure 7 Comparison of transcripts detected by cross- and same-species hybridization. The common orthologues presented on HG-U133A and Affymetrix canine GeneChips were identified (996 orthologous pairs) using MegaBLAST. Transcripts identified by hybridization of canine sample to Affymetrix canine GeneChips were compared with those identified during hybridization of canine sample to human HG-133A GeneChips with and without application of masking procedure. Correct (solid bars), false positive (open bars) and false negative (grey bars) signal intensity calls were identified and expressed in percentage values.
Discussion
The widespread use of large mammal models in respiratory research, coupled to the limited availability of corresponding gene microarrays, motivated us to investigate the feasibility of utilizing gene arrays designed for related organisms. In this work, we hybridized canine lung tissues to human Affymetrix HG-U133A GeneChip, and demonstrated that oligonucleotide probes designed for identification of human targets were able to generate detectable hybridization signal for canine tissues (Figure 6B). However, the number of identified transcripts was 3–4 times lower than that of the transcripts which were identified during human sample hybridization (Figure 6A). Our analysis indicated that the low sensitivity of the majority of the HG-U133A probe sets to canine transcripts was due to partial genomic differences between human and canine sequences (Figure 2B and 2C). The abundance of probes in Affymetrix probe sets (11–16 probes per set) presented an opportunity for reduction of probe sets to probes that generate detectable cross-species hybridization signal and therefore, improving transcript detecting ability of such modified probe sets. Our probe level analysis allows evaluation and identification of probes that failed to detect their targets in canine tissues. We demonstrated that this poor performance was not random, but rather linked to either imperfect probe design (Figure 1B) or interspecies differences in genomic sequences (Figure 4). Another indirect linkage of probe/target sequence similarity and probe performance was demonstrated by inclination of poorly performing probes toward the 3' end of a target (Figure 5C), which is more prone to diversity [14]. The masking of this poorly performing probes resulted in an increase of "Present" calls in both species (Figure 5A), where the increase in human transcript detection was attributed mainly to masking of imperfectly designed probes. Interestingly, the increase of "Present" calls was more prominent in canine hybridizations when we masked the most poorly performing probes that were not able to detect their targets in more that 30 hybridizations. However, thereafter the difference in number of "Present" calls between canine and human remains constant (Figure 5A). This comparable increase in the amount of "Present" calls is most likely species-independent and attributed to probe set optimizations, which is reflected by changes in probe set sizes. We demonstrated that masking probes that were not detecting their targets in 40% cases generated maximum of "Present" calls in both species (Figure 5A) and speculated that probe pairs that poorly performed in more than 40% of hybridizations lack homology to their canine counterparts. This was confirmed by alignment analysis of these probes to their canine counterparts (Figure 2), which showed that 82% of masked probes produced one or more mismatches with their canine counterparts. The remaining 18% can be attributed to a sequencing error, a single nucleotide polymorphism, or longer than 16 bp stretches of perfect matches that can produce detectable hybridization signal [15]. We also identified probes that were perfectly aligned to their canine counterparts yet perform poorly. These probes represented approximately 7% of tested probe pairs, a value similar to that previously reported [17] and can be attributed to the secondary structure or to the presence of yet unknown sequence that recognizes MM probe as a perfect match and the resulting increase in background signal obscure the hybridization signal.
We demonstrated that modification of probe sets by masking poorly performing probes drastically increased detection of canine transcripts by the human HG-U133A GeneChip (Figure 6B). To validate these results, we compared transcripts identified by our masking approach with those identified by hybridization of canine lung tissues RNA to the Affymetrix canine-specific GeneChip. This comparison demonstrated that modification of probe sets prior to MAS 5.0 analysis drastically increases correlations between transcripts identified by same- and cross-species hybridizations (Figure 7). The slight increase (from 3% to 10%) in false positive transcript fraction introduced by our procedure is a common feature of high density oligonucleotide array analyses [18] and is well compensated by the approximately 10-fold decrease in the false negative transcript fraction from 59% to 6% (Figure 7).
Conclusion
The studies presented here show that the Affymetrix microarray platform can be successfully used for cross-species hybridization of related organisms. A simplified version of this method (probes that failed to generate detectable signal in all hybridizations were masked) has already been successfully applied to small (7 GeneChips) data sets [19] and produced biologically sensible results. We expect that method presented here will be applicable not only for canine/human hybridization but also for other combinations of cross-species hybridizations. Whenever there is homology between species, there is an opportunity for identification probes with detectable hybridization signal using probe level analysis described here. This method will accelerate genome-wide analysis in experimental models for which species-specific gene arrays are not currently available.
Methods
Animal preparation and sample collection
All experimental procedures were approved by the Johns Hopkins University Animal Care and Use Committee. Four mongrel dogs (weight 21.3 ± 1.5 kg) were anesthetized with 25 mk/kg pentobarbital i.v. and instrumented with femoral arterial and venous catheters. Anesthesia was maintained with additional pentobarbital (5 mg/kg iv every hour and when indicated) and muscle relaxation provided by pancuronium (3 mg bolus and 0.5 mg hourly iv). A 39 or 41 French double-lumen endobronchial tube (Mallinkrodt, St. Louis, Missouri) was placed via a tracheostomy and left lung was then mildly injured by repeated lavage with warmed saline. One animal was sacrificed by exsanguination after supplemental pentobarbital (10 mg/kg i.v.) right after lavage, the chest opened, and 10 control tissue samples were taken from 5 corresponding regions in both lungs (apex- dependent and non-dependent, base- non dependent, mid, and dependent). Other 3 animals were ventilated for 5 hours, then sacrificed as described above and 30 samples (10 from each animal) were collected following the control animal pattern. The lung tissue samples were immersed in RNAlater (Ambion, Austin, Texas), snap-frozen and stored at -80°C until processed for RNA isolation. The protocol for cross-species hybridization sample preparation has been described by our group previously [20,21]. Briefly, total RNA was isolated from individual lung samples described above using the Trizol reagent and Qiagen RNeasy columns. Individual cDNAs were prepared from each RNA isolate using reverse transcriptase (GIBCO-BRL SuperScript). Each cDNA was subsequently used as a template to make biotin-labeled cRNA using an in vitro transcription reaction (Enzo), resulting in a single cRNA for each lung sample. Each cRNA was hybridized with an individual Affymetrix HG-U133A oligonucleotide array, which was subsequently processed and scanned according to the manufacturer's instructions. All arrays were hybridized in the same batch to avoid variability in hybridization conditions. The sample collection and preparation for the same specie hybridization was adapted from the described above protocol and applied for Affymetrix Canine GeneChip hybridization by McVerry et al [22].
Human tissues and cell lines sampling and RNA isolation
All human samples and analyses posted on the HopGene server were generated with IRB approval (#B0112210102). Pathologic samples of human myocardium 1 from unused donor heart, 2 from patients with ischemic cardiomyopathy, and 2 from patients with nonischemic cardiomyopathy were collected at time of transplantation. Peripheral blood mononuclear cells were collected form 1 healthy person, 2 smokers, and 2 patients with allergic asthma and separated by Ficoll-hypaque density gradient centrifugation. Airway brushing cells were collected in normal saline form 1 healthy person, 3 smokers, and 1 patient with allergic asthma and separated by centrifugation. Bronchoalveolar lavage (BAL) was performed during bronchoscopy of 2 healthy individuals, 2 smokers, and 2 patients with allergic asthma by instilling 100 ml of normal saline, BAL fluid was aspirated into tubes and cells collected by centrifugation. The muscle tissue was collected from 5 individuals with different stages of obesity. All human tissues were snap-frozen in liquid nitrogen upon collection. These frozen samples (~50 mg from each sample) were directly solubilized in chaotropic solubilization buffer using a Brinkman Polytron tissue disruptor. Larger tissue fragments (>100 mg) were pulverized into frozen powder with a mortar and pestle, pre-chilled to liquid nitrogen temperature, and then the frozen powder was solubilized with the Polytron. RNA was purified using Trizol LS (Life Technologies) and an additional RNA purification step was conducted using the RNAeasy purification kit (Qiagen Inc., Valuencia, CA).
Human smooth muscle cell cultures were treated with ApoC-1 for 10 min (n = 2), 24 h (n = 2) or left intact (n = 1); human pulmonary artery endothelial cells were exposed to hyperoxia for 6 h (n = 1), 24 h (n = 1), 48 h (n = 1), or normoxia (n = 2); IB3-1 bronchial epithelial cells were treated with 1 mM PBA for 24 h (n = 1) or left intact (n = 4). Cells were rinsed twice with ice cold Hank's buffer then scraped into 3ml of ice cold TRIzol (Gibco BRL). Total RNAs was purified using Trizol LS (Life Technologies) and an additional RNA clean-up step was conducted using the RNAeasy purification kit (Qiagen Inc., Valuencia, CA). Approximately 10 μg of purified, total RNA was used for analyses.
RNA preparation and hybridization
Purified total RNA was reverse transcribed to first-strand cDNA using a hybrid primer consisting of oligo-dT and T7 RNA polymerase promoter sequences. The single-stranded cDNA was then converted to double-stranded cDNA. Complementary DNA corresponding to 5–10 μg of total RNA was used in a cRNA amplification step using T7 RNA polymerase and two biotinylated nucleotide precursors. The resulting biotinylated cRNA was fragmented to a size of approximately 50 bp. Approximately 20–30 μg of cRNA from each tissue was hybridized to corresponding (HG_U133A or Canine) GeneChips (Affymetrix, Santa Clara, CA). The bound cRNA was visualized by binding of streptavidin/phycoerythrin conjugates to the hybridized GeneChip, followed by laser scanning of bound phycoerythrin. These scan results are posted on HopGene ftp server and are freely accessible at
Expression Data Analysis
Data analysis was conducted using Affymetrix MicroArray Suit 5.0 (MAS 5.0) and GeneChip Operating Software (GCOS 1.1) (Affymetrix Inc., Santa Clara, CA). Signal intensity values for hybridization of human samples or heterohybridization of canine samples to HG-U133A (40 GeneChips for each specie), and hybridization of canine lung tissues to Affymetrix canine GeneChips (12 GeneChips) were generated using user-definable parameters set to Affymetrix default values and total chip fluorescence intensities scaled to 150. Each GeneChip was analyzed with and without application of custom generated probe masking files. The probe masking function is imbedded into MAS 5.0 and GCOS 1.1 and can be easy evoked and linked to a list of probes selected for masking. The overall schema of analysis of 40 human HG-U133A GeneChips, which were hybridized with canine mRNA is outlined in Figure 8.
Figure 8 Schema of Affymetrix data analysis restricted to excellent performing probe pairs.
Creating "Perfect match"/"Mismatch" probe level matrix
Each HG-U133A GeneChip contains 22,215 (excluding internal controls) probe sets, each representing one target sequence. The majority of the probe sets (98 %), contain 11 probe pairs with each probe pair consisting of two 25-mer oligonucleotides. One of these oligonucleotides is a "perfect match" (PM) to a target sequence and another is a "mismatch" (MM) with an intentionally mutated middle nucleotide (Figure 1A). Therefore, each probe set contains 11 PM probes and 11 corresponding MM probes for calculation of total and nonspecific hybridization signals, respectively (reader is referred to Affymetrix website for detailed explanations, ). The MAS 5.0 program analyzes the differences in the amount of hybridized mRNA to each PM (total signal) and MM probes (background signal) across entire probe set, and computes the relative abundance of a transcript and its significance (p value). Transcripts with p < 0.04 are called "Present," with 0.04 ≤ p ≤ 0.06 called "Marginal," and with p > 0.06 called "Absent". The evaluation of the effects of our masking approach on the change in the transcript detection call was based solely on the Affymetrix transcript abundance classification without modification of the original algorithm, which makes employment of our masking files universal among Affymetrix users.
Expression profiles analyzed in this report were generated using total mRNA extracted from different parts of canine lungs exposed to 15 different experimental conditions. Signal intensities files (CEL files) generated by MAS 5.0 and GCOS 1.1 for HG-U133A and Canine GeneChip, respectively were imported into the R-based Bioconductor analytical package and processed using affy library [23]. Intensity signals for each experimental group were corrected for background and normalized using default mas5 package settings, which closely simulate the MAS 5.0 algorithm. The raw intensities of PM and MM values for each of 246799 probes per chip (excluding internal Affymetrix controls) were extracted from mas5 generated probe level matrix [24] and converted into the flat text file with 80 by 246799 entries (Figure 8). The R-language scripts for extraction of PM and MM values are available upon request. Further manipulations and analyses of extracted data were conducted using HopGene designed algorithms written in Python 2.2. Statistical parts of the script employed the Python Statistical Module developed by NSG group
Correlation of the human-probe/canine-target similarity and hybridization signal
To evaluate relation between probe/target sequence similarity and hybridization signal all probe pairs were classified as efficient (PM-MM>20, the minimum practical signal value [25]) or inefficient (PM-MM≤20). This classification was applied to individual probe pairs throughout 40 canine/human hybridizations and 40 human/human hybridizations. Probe level analysis of human/human hybridization was aimed at identification of non-functional due to poor design or inappropriate hybridization conditions probe pairs (Figure 1B). These probes were used for identification of the net effect of canine/human genetic differences, that is probes that were inefficient in both human/human and canine/human hybridizations (Figure 4) were eliminated from further analysis. Target sequences (FASTA format) representing canine net effect probes (Figure 4) were retrieved from NetAffx and queried against NCBI non-redundant (nr) database for their corresponding canine orthologues using MegaBLAST . Since the MegaBLAST batch query tool limits the number of submitting sequences, we divided this human dataset into 3 batches (~7000 sequences each) and used Canis familiaris term as the search restriction. BLASTN 2.2.8 [Jan-05-2004] analysis employed sequences form GenBank, EMBL, DDBJ, PDB databases, excluding EST, STS, GSS, or phase 0, 1 or 2 HTGS sequences and showed that 662 target sequences on the HG_U133A GeneChip had 1122 canine counterparts with >75% sequence homology. For manual evaluation 20 target sequences were randomly selected from the pool of sequences that contained probes efficient in less that 5% hybridizations (hybridized only to 0–2 chips, Figure 4). Since each target sequence contains 11 probes the total of 220 probes was evaluated and 22 probes that efficiently hybridized only to 0–2 GeneChips (inefficient group) and 31 probes that efficiently hybridized to 38–40 (95%) chips (efficient group) were aligned to their canine counterparts. Alignments of these 53 probes to their canine counterparts were generated using CLUSTALW and nucleotide mismatches were counted and accordingly assigned to the "efficient" or "inefficient" group. Since mismatch distribution was heavily skewed towards small numbers of mismatches (1–2) the correlation between probe performance and target sequence homology (number of identified mismatches) was evaluated using nonparametric Mann-Whitney test with p < 0.05 used as a significance cut-off (Figure 4, insert).
Creating probe pair masking files
Hybridization of individual probe pairs (246799total) to each GeneChip (40 total) was evaluated and inefficient probe pairs that produced hybridization signal below minimum practical signal value [25] (PM-MM≤20) were identified. The flat text file containing two columns was generated, with the first column containing probe set ID entry and the second column containing order numbers of probe pairs to be masked. The empty template masking file was generated using MAS 5.0 masking tool and subsequently populated with data from the flat text file. The first masking file was built from probe pairs that were inefficient in all 40 hybridizations. The second masking file contained probe pairs that were inefficient in ≥35 hybridizations. The subsequent masking files were generated with 5 chosen as an increment of inefficient hybridizations with the last masking file comprised of probe pairs that were inefficient in the only one hybridization. Resulting 9 masking files (40, 35, 30, 25, 20, 15, 10, 5, 1) were applied during MAS 5.0 analysis of canine and human mRNA hybridizations to the HG-U133A GeneChips. The highest percentages of the "Present" call (Figure 5) were generated around the point at which probes that were inefficient in ≥15 hybridizations (approximately 40%) were masked and was considered an optimum condition for a given dataset. In further discussions the percentages of the described above masking points were conveniently approximated.
Evaluating effects of masking approach
To simplify the evaluation of our masking procedure, only probe-sets containing 11 probe pairs, which comprises ~98% of all HG_U133A probe-sets, were assessed. The remaining 2% of probe-sets served as internal controls (20 probe pairs per set) and probe sets built from 14–16 probe pairs were inherited from the HG_U95AVs2 GeneChip. Our masking probe pair list for 11-member probe-sets contained 156,174 probe pairs (~65% of total). All counting and poorly performing probe selecting protocols were scripted in Python 2.2 and are available upon request.
Concordance of the masking HG_U133A probes with their poor homology to the corresponding canine counterparts was evaluated using analysis of 5 human/canine alignments including one presented in Figure 2C and four randomly selected alignments from three representative pools of probe sets: probe sets where all probes were detectable, probe sets where none of the probes were detectable and probe sets represented by both detectable and non-detectable probes .
Affymetrix Canine and HG_U133A GeneChip comparison
Identification of homologous to Affymetrix Canine GeneChip target sequences on human HG_U133A GeneChip was conducted using MegaBLAST. The Affymetrix Canine GeneChip's target sequences in FASTA format were retrieved from NetAffx, split into 12 batches (~2000 sequences each) to overcome MegaBLAST limitations and queried against nr database using Homo sapiens term for filtering. GeneBank accession numbers of human orthologues identified by MegaBLAST for canine chip targets were matched against those on HG-U133A GeneChip using Microsoft Access 2000. Detection calls for common orthologues between 40 cross-hybridized HG-U133A GeneChips and 12 same-species hybridized canine GeneChips were compared. The transcripts that were efficiently detected in ≥ 40% (described above optimum condition) of tested GeneChips were considered as expressed. Then detection of these expressed in canine lung tissues transcripts was compared with their corresponding detection in original and masked cross-hybridization of canine lung tissues to human HG-U133A GeneChip. Transcripts detected by Canine GeneChip but missed by HG-U133A cross-hybridization were "false negative" and those not detected by Canine GeneChip but shown by HG-U133A heterohybridization were "false positive".
Authors' contributions
DNG originated the probe level analysis of cross-species hybridization, carried out the bioinformatics studies and drafted the manuscript. SFM led the human component of the project and conducted human sample selection and hybridization analysis. BAS led the canine component of the project, originated and designed canine ventilation model, provided canine samples for hybridization, and revised the manuscript. RAI participated in the design of the study, established Bioconductor database for new canine GeneChip and consulted on the statistical analysis. SQY optimized protocols for canine samples hybridization to human and canine GeneChips, and assured the quality of generated microarray data. JGNC conceived of the study, supervised and coordinated the project, revised and advised on the manuscript preparation. All authors read and approved the final manuscript.
Acknowledgements
We thank James MacDonald for helpful suggestions on Bioconductor application, and Nicholas Shank for organizing and publishing, pertinent to this paper, materials on the HopGene website. This work was supported by the NHLBI – sponsored HopGene Program in Genomics Application (HL-69340), SCCOR (HL-073994), and individual (DNG) NRSA grant (F32 HL74590-01A1).
==== Refs
Olson H Betton G Robinson D Thomas K Monro A Kolaja G Lilly P Sanders J Sipes G Bracken W Dorato M Van Deun K Smith P Berger B Heller A Concordance of the toxicity of pharmaceuticals in humans and in animals Regul Toxicol Pharmacol 2000 32 56 67 11029269 10.1006/rtph.2000.1399
Olson H Betton G Stritar J Robinson D The predictivity of the toxicity of pharmaceuticals in humans from animal data--an interim assessment Toxicol Lett 1998 102-103 535 538 10022308 10.1016/S0378-4274(98)00261-6
Hein WR Griebel PJ A road less travelled: large animal models in immunological research Nat Rev Immunol 2003 3 79 84 12511878 10.1038/nri977
Pirenne J Contribution of large animal models to the development of clinical intestinal transplantation Acta Gastroenterol Belg 1999 62 221 225 10427786
Blache D Adam CL Martin GB The mature male sheep: a model to study the effects of nutrition on the reproductive axis Reprod Suppl 2002 59 219 233 12698984
Gu MN Xiao JF Huang YR Zhou W Fu WJ Chen YM Piao YJ Chen PY Zeng FY Study of direct lung injury by seawater in canine models Di Yi Jun Yi Da Xue Xue Bao 2003 23 201 205 12657559
Hubloue I Biarent D Abdel Kafi S Bejjani G Melot C Naeije R Leeman M Endothelin receptor blockade in canine oleic acid-induced lung injury Intensive Care Med 2003 29 1003 1006 12734647
Enkhbaatar P Murakami K Shimoda K Mizutani A McGuire R Schmalstieg F Cox R Hawkins H Jodoin J Lee S Traber L Herndon D Traber D Inhibition of neuronal nitric oxide synthase by 7-nitroindazole attenuates acute lung injury in an ovine model Am J Physiol Regul Integr Comp Physiol 2003 285 R366 72 12763743
Murakami K Bjertnaes LJ Schmalstieg FC McGuire R Cox RA Hawkins HK Herndon DN Traber LD Traber DL A novel animal model of sepsis after acute lung injury in sheep Crit Care Med 2002 30 2083 2090 12352045 10.1097/00003246-200209000-00022
Chismar JD Mondala T Fox HS Roberts E Langford D Masliah E Salomon DR Head SR Analysis of result variability from high-density oligonucleotide arrays comparing same-species and cross-species hybridizations Biotechniques 2002 33 516 8, 520, 522 passim 12238761
Moody DE Zou Z McIntyre L Cross-species hybridisation of pig RNA to human nylon microarrays BMC Genomics 2002 3 27 12354330 10.1186/1471-2164-3-27
Wang Z Dooley TP Curto EV Davis RL VandeBerg JL Cross-species application of cDNA microarrays to profile gene expression using UV-induced melanoma in Monodelphis domestica as the model system Genomics 2004 83 588 599 15028282 10.1016/j.ygeno.2003.10.007
Medhora M Bousamra M Zhu D Somberg L Jacobs ER Upregulation of collagens detected by gene array in a model of flow-induced pulmonary vascular remodeling Am J Physiol Heart Circ Physiol 2002 282 H414 22 11788387
Higgins MA Berridge BR Mills BJ Schultze AE Gao H Searfoss GH Baker TK Ryan TP Gene expression analysis of the acute phase response using a canine microarray Toxicol Sci 2003 74 470 484 12773757 10.1093/toxsci/kfg142
Ji W Zhou W Gregg K Yu N Davis S A method for cross-species gene expression analysis with high-density oligonucleotide arrays Nucleic Acids Res 2004 32 e93 15247326 10.1093/nar/gnh084
Antipova AA Tamayo P Golub TR A strategy for oligonucleotide microarray probe reduction Genome Biol 2002 3 RESEARCH0073 12537562 10.1186/gb-2002-3-12-research0073
Mei R Hubbell E Bekiranov S Mittmann M Christians FC Shen MM Lu G Fang J Liu WM Ryder T Kaplan P Kulp D Webster TA Probe selection for high-density oligonucleotide arrays Proc Natl Acad Sci U S A 2003 100 11237 11242 14500916 10.1073/pnas.1534744100
Rajagopalan D A comparison of statistical methods for analysis of high density oligonucleotide array data Bioinformatics 2003 19 1469 1476 12912826 10.1093/bioinformatics/btg202
Grigoryev DN Ma SF Irizarry RA Ye SQ Quackenbush J Garcia JG Orthologous gene-expression profiling in multi-species models: search for candidate genes Genome Biol 2004 5 R34 15128448 10.1186/gb-2004-5-5-r34
Simon BA Easley BR Grigoryev DN Ma SF Ye SQ Lavoie T Garcia JGN Microarray analysis of regional cellular responses to local mechanical stress in experimental acute lung injury (Submitted)
Ye SQ Simon BA Maloney JP Zambelli-Weiner A Gao L Grant A Easley RB McVerry BJ Tuder RM Standiford T Brower RG Barnes KC Garcia JG Pre-B-Cell Colony-enhancing Factor as a Potential Novel Biomarker in Acute Lung Injury Am J Respir Crit Care Med 2005 171 361 370 15579727 10.1164/rccm.200404-563OC
McVerry BJ Peng X Hassoun PM Sammani S Simon BA Garcia JG Sphingosine 1-phosphate reduces vascular leak in murine and canine models of acute lung injury Am J Respir Crit Care Med 2004 170 987 993 15282202 10.1164/rccm.200405-684OC
Irizarry RA Gautier L Cope L Parmigiani G, Garrett ES, Irizarry RA and Zeger SL An R package for analyses of Affymetrix oligonucleotide arrays The Analysis of Gene Expression Data: Methods and Software 2003 New York, Springer
Irizarry RA Bolstad BM Collin F Cope LM Hobbs B Speed TP Summaries of Affymetrix GeneChip probe level data Nucleic Acids Res 2003 31 e15 12582260 10.1093/nar/gng015
Doniger SW Salomonis N Dahlquist KD Vranizan K Lawlor SC Conklin BR MAPPFinder: using Gene Ontology and GenMAPP to create a global gene-expression profile from microarray data Genome Biol 2003 4 R7 12540299 10.1186/gb-2003-4-1-r7
| 15871745 | PMC1156887 | CC BY | 2021-01-04 16:32:46 | no | BMC Genomics. 2005 May 4; 6:62 | utf-8 | BMC Genomics | 2,005 | 10.1186/1471-2164-6-62 | oa_comm |
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-721588820810.1186/1471-2164-6-72Research ArticleHyperspectral microarray scanning: impact on the accuracy and reliability of gene expression data Timlin Jerilyn A [email protected] David M [email protected] Michael B [email protected] Anthony D [email protected] M Juanita [email protected] Margaret [email protected] Sandia National Laboratories*, P.O. Box 5800, Albuquerque, NM, 87185, USA2 Lovelace Respiratory Research Institute, 2435 Ridgecrest SE, Albuquerque, NM, 87108, USA2005 11 5 2005 6 72 72 15 2 2005 11 5 2005 Copyright © 2005 Timlin et al; licensee BioMed Central Ltd.2005Timlin et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Commercial microarray scanners and software cannot distinguish between spectrally overlapping emission sources, and hence cannot accurately identify or correct for emissions not originating from the labeled cDNA. We employed our hyperspectral microarray scanner coupled with multivariate data analysis algorithms that independently identify and quantitate emissions from all sources to investigate three artifacts that reduce the accuracy and reliability of microarray data: skew toward the green channel, dye separation, and variable background emissions.
Results
Here we demonstrate that several common microarray artifacts resulted from the presence of emission sources other than the labeled cDNA that can dramatically alter the accuracy and reliability of the array data. The microarrays utilized in this study were representative of a wide cross-section of the microarrays currently employed in genomic research. These findings reinforce the need for careful attention to detail to recognize and subsequently eliminate or quantify the presence of extraneous emissions in microarray images.
Conclusion
Hyperspectral scanning together with multivariate analysis offers a unique and detailed understanding of the sources of microarray emissions after hybridization. This opportunity to simultaneously identify and quantitate contaminant and background emissions in microarrays markedly improves the reliability and accuracy of the data and permits a level of quality control of microarray emissions previously unachievable. Using these tools, we can not only quantify the extent and contribution of extraneous emission sources to the signal, but also determine the consequences of failing to account for them and gain the insight necessary to adjust preparation protocols to prevent such problems from occurring.
==== Body
Background
Since their introduction in 1995 [1], DNA-based microarrays (also known as genechips) have driven an explosion in functional genomic analyses. All varieties of microarrays have in common the ability to perform binary comparisons of gene expression for a large number of genes simultaneously in a microchip format [2-6].
In theory, biological changes should define the limitations of microarray technology, but unfortunately, technological issues have frequently limited the usefulness of microarray data. Non-biological factors including printing artifacts, dye-gene interactions, background emissions, and slide-to-slide variations significantly reduce the ability to accurately monitor changes in gene expression in microarray experiments [4,7-10]. These experimental factors are common and often laboratory dependant due to the complicated multi-step procedures used in the production, hybridization, and analysis of microarrays. In attempts to minimize the effect of variability due to non-biological sources, a variety of statistical analyses [11,12], normalization techniques [13-15], and metrics for image quality [16] have been proposed. However, all these analysis techniques have two assumptions: 1) that the only contributors to the signal within a printed spot are the labeled DNA and a uniform background due to emission of the glass substrate and 2) that background fluorescence surrounding the spot is the same as background fluorescence under the spot, despite evidence that this assumption may not be valid [7,12,17].
Neither of these assumptions can be validated with current commercial microarray scanners. Commercial scanners are univariate instruments; that is they use filters to pass all photons emitted in a specific wavelength range to a single point detector. This mode of operation can be fast, but it does not allow discrimination of photons by emission source. Thus, it is not possible to distinguish two photons of the similar wavelength that arise from different emitting species if they are passed through the filter for that channel. Many problems that plague microarrays (inaccuracies in background correction, dye-gene effects, skew toward one channel, dye crosstalk, and contaminating fluorescence) cannot be accurately assessed in data from filter-based microarray scanners due to this limitation and this can lead to erroneous data [9,17].
To address these issues, we have developed a hyperspectral-imaging microarray scanner [18] that allows the simultaneous quantification of all fluorescent species, including the spot-localized background leading to a significant improvement in the accuracy of microarray data. The hyperspectral scanner (HSS) coupled with multivariate data analysis provides in-depth understanding of the signal detected by traditional microarray scanners and can promote improvement in microarray technology and actually improve the quality of microarray data. The benefits of an additional dimension of spectral information for material science, cytogenetic, and histological applications [19] and live-cell microscopy [20] have been reviewed. However, until our report, hyperspectral imagers have not demonstrated the sensitivity or speed of commercial microarray scanners or the multivariate data analysis capabilities necessary to extract sufficient information from the complex data [21-23]. This paper presents the use of the HSS and multivariate data analysis to understand three anomalies commonly seen in microarray data: skew toward the green channel, dye separation, and high, variable background signal. The unique capability of HSS technology to identify and correct for the presence of these phenomena improves the reliability and accuracy of gene expression data.
Results and Discussion
Hyperspectral imaging and multivariate data analysis
The HSS we have developed is optimized for imaging printed DNA microarrays and excites a sample with a single laser, typically at 532 nm, while recording the emission over a wavelength range from 550–900 nm in approximately 0.75 nm increments to create a hyperspectral data cube (Fig. 1a). Details of the optical design and characterization of this line-imaging system have been published elsewhere [18]. Additional lasers are available and the wavelength range and spatial and spectral resolution are adjustable. The sensitivity and dynamic range of this HSS is the same as or better than the commercial microarray scanners we have tested for dyes emitting in the green channel of commercial systems such as Cy3 [18]. Typically, red emitting dyes like Cy5 are not optimally excited by our 532 nm laser, but based on its excitation spectrum, we are achieving ~5–8% excitation of Cy5. In comparison studies between an Axon 4000B microarray scanner exciting Cy5 with 633 nm light and the HSS exciting Cy5 with 532 nm light (data not shown) the HSS was found to be a factor of 6 less sensitive for Cy5 than commercial microarray scanner. However, the signal acquired from Cy5 is sufficient for quantitation by the multivariate algorithms. In addition, for the studies reported in this publication the focus is predominantly on the green channel emissions. In its current configuration the HSS scanner operates at a slightly slower speed of data acquisition (scanning at a maximum rate of 0.07 mm/s for 10 μm resolution) but this disadvantage is compensated for by the more accurate quantification and increased specificity of the fluorescent signal, especially in the presence of contaminants. The CCD readout rate is the limiting factor in the current HSS and newly available charge-coupled device (CCD) detector electronics could allow the HSS to scan at speeds up to twice the speed of commercial scanners.
Figure 1 Illustration of hyperspectral data cube and multivariate analysis results. (a) Three-dimensional hyperspectral data cube for an idealized two-component microarray containing emission only from labeled DNA within the printed spots. Each pixel in the x-y image plane contains an entire fluorescence emission spectrum from 550–900 nm. (b-c) Results of multivariate curve resolution on two-component sample hyperspectral data cube shown in A. (b) Pure component spectra identify "what" species are fluorescing in an image. (c) Corresponding component concentration maps show "where" and "how much" of each component in Fig. 1b is present.
A typical hyperspectral data cube contains tens of thousands to millions of spectra and spectral and spatial relationships cannot be readily visualized (Fig. 1a). To reduce the data to a simpler representation of the important features, HSS data are analyzed using multivariate curve resolution (MCR), a powerful factor analysis technique based on a constrained alternating least squares procedure [24]. MCR algorithms, when applied to fluorescence emission images, use correlations between variables to extract 1) representative pure-component spectra of the all the emitting species (Fig. 1b) and 2) independent concentration maps depicting the spatial location of these species and their relative concentrations from the highly overlapped spectral data (Fig. 1c). The MCR analysis assumes that the number of non-noise components is known or can be estimated and requires initial estimates to either the spectra or concentrations.
Figure 2 MCR analysis results of cDNA microarray exhibiting contamination fluorescence. Fluorescence of all species was excited simultaneously in a single scan with 532 nm laser. (a) MCR extracted pure component spectra representing glass, Cy3, Cy5 and a contaminant. Inset shows ~30000 raw spectra in the image data cube used to determine the pure component spectra. (b) Extracted concentration maps corresponding to each of the pure component species in Fig. 2a. Resulting HSS intensities have been scaled to match the individual channel intensities seen on the Axon 4000B microarray scanner.
Principal component analysis, a closely related and popular multivariate analysis technique for biological data [25], typically provides an accurate determination of the number of non-noise components, but often cannot provide easily interpretable pure component spectra underlying complex spectral image data due to rotational ambiguity. For these studies involving microarray data, principal component analysis was used to determine the number of non-noise components and to provide initial estimates of the spectral shapes for the MCR algorithm to iterate upon. The application of non-negativity and equality spectral constraints within the MCR algorithms facilitates physically meaningful solution of spectral components. The benefits of MCR over more traditional univariate (band integration, band ratios) and other multivariate (least squares, linear unmixing) analysis methods include the ability to separate overlapping spectral emissions, discover the pure-component spectra with little or no information given a priori, and model unknown spectral contributions such as interferrents, backgrounds, and instrument artifacts. Applications of MCR for vibrational spectroscopic data have been outlined in a recent review [26]. Recently, members of our group have developed efficient, automated multivariate statistical analysis algorithms to analyze large X-ray hyperspectral image data using desktop computers [27]. We have optimized these algorithms for fluorescence image data from the HSS [28,29].
Throughout this paper, emissions from within the area of the printed DNA spots are referred to as spot-specific emissions and emissions that are not spot-specific are referred to as background emissions. Also, ratio images generated by commercial microarray scanners are referred to as R/G (Red/Green) images and images generated by HSS are referred to as Cy5/Cy3 images, because the commercial scanner image is constructed from the intensities in the red and green channels regardless of source, whereas the HSS and data analysis produces true images of the Cy5 and Cy3-labeled cDNA that are uncontaminated by emission from other species.
Spot-localized emissions: Skew to toward the green channel
Earlier work has shown the presence of contaminant fluorescence in the green channel of many common types of microarray slides [17]. The HSS was used to confirm the presence of the green contaminant in slides from a variety of laboratories (four different commercial sources and in-house printed arrays from at least four independent laboratories) after hybridization. This contaminant is introduced to the slide during printing and can be recognized in the raw images as a variable, weak green channel signal that may be observed in all spots. In our investigations we examined slides from at least three laboratories printed using DMSO printing methods and did not identify a green contaminant in any of the spots on these slides.
Gtotal = GCy3 + Gglass + Gcont Eq. 1
Commercial filter-based microarray scanners will confound signal from a spot-localized green contaminant with signal from the labeled cDNA and glass substrate in the green channel. This is because the green signal intensity recorded for a pixel with a spot-localized green contaminant is the sum of all the green fluorescence intensities, including the glass emission (a relatively constant small value), the contaminant (a variable value), and the labeled hybridized cDNA. This extra green intensity has no affect in the red channel and little effect on spots with high green channel signal, but contributes significantly to the signal from spots with weak and medium green-channel intensities. As illustrated in Equation 1 data obtained from slides with spot-localized contaminating fluorescence will report green channel intensities that are falsely high (Gtotal is the total signal acquired in the green channel; GCy3, Gglass, and Gcont are the signals arising from the Cy3 labeled DNA, glass substrate and contaminant, respectively.) Unfortunately the effect of spot-localized contaminant emission cannot be corrected in commercial scanner microarray data using standard background correction procedures that estimate background signal for a spot from the signal around the spot and thus it leads to errors in the calculated expression ratios (R/G). The detrimental effect of the contaminant decreases with increasing green-channel intensities but can easily account for skews toward the green channel, dye-gene effects, unsuccessful dye-flip experiments, and highly variable low intensity data observed in many microarray experiments.
Figure 2 shows the results of multivariate data analysis of a hyperspectral image from a microarray slide with contaminating fluorescence. Because the HSS can isolate the contaminant emission spectrum, true pure-component concentration maps can be generated and the resulting images of the Cy3 labeled cDNA are contaminant-free. Using the optical filter functions of the Axon 4000B microarray scanner, HSS images can be scaled to match the total intensity of each of the commercial scanner channels, thus providing a direct comparison to the commercial scanner results. These scaled concentration maps are used to calculate an accurate Cy5/Cy3 image that gives spot intensity values without contributions from the glass and contaminant emissions. The R/G ratio constructed from the commercial scan (Fig. 3a) and the more accurate Cy5/Cy3 ratio image of the same area on the slide (Fig. 3b) show the difference is dramatic, with 75% of the spots having ratios in error of a factor of 2 or more due to the presence of the green channel contaminant. These errors would change the basic conclusion from the data that most genes in the test sample were down-regulated relative to the control when, in fact, the correct conclusion should be that most genes in the test sample were either up regulated or not differentially expressed. The fluorescence contribution of the contaminant cannot be corrected with current commercial technology due to extreme spectral overlap and the variability of the spot-localized contaminant concentration. Image thresholding could be useful if the contaminant contribution were known and fairly constant but even low levels of contaminant fluorescence would require high thresholds be set. For example, a maximum of 100 arbitrary fluorescence units of spot-localized contaminant signal would require a threshold of 500 arbitrary fluorescence units be set in an attempt to maintain errors in the 20% range.
Figure 3 Green contaminant presence alters resulting image from commercial scanner. Comparison of resulting images from two scanners with green contaminant present. (a) Red/Green ratio image generated from Axon 4000B microarray scanner depicting most spots as green indicating the control gene was expressed to larger extent than the treatment. (b) Cy5/Cy3 ratio image generated from HSS concentration maps depicting only a few spots as differentially expressed.
Spot-localized emissions: Dye separation
"Dye separation" is a phenomenon referred to in the microarray literature as a ring of one color surrounding the other, typically red around green [16]. Although this phenomenon is seen in published images, there is no satisfactory explanation for what causes the labeled cDNA to hybridize with a dye-specific spatial pattern. Other spatial patterns of hybridization, e.g., doughnuts and coffee rings are well documented, but spatial anomalies theoretically should affect both dyes within a spot to the same extent.
Apparent dye separation was visible in some spots on the microarray slides we examined with the green contaminant. Figure 4a shows a R/G ratio image from a typical spot exhibiting dye separation scanned on an Axon 4000B scanner. From this image, the Cy5-labeled cDNA appears to have hybridized in a circle around the area where the Cy3-labeled cDNA hybridized. HSS analysis revealed that the spot-specific signal was from three sources: Cy5-labeled cDNA, Cy3-labeled cDNA, and green contaminant. The individual concentration maps generated from the multivariate analysis of the HSS image data demonstrate that, for this spot, the diameters of Cy5 and Cy3 hybridization are equal, although little Cy3 is present (Fig. 4c–d). The contaminant emission, which is brighter than the Cy3, is present in a smaller diameter circle (Fig. 4e) and the result of this localization difference is a spot with a bright green center and a red ring on commercial scanners. This size difference most likely occurs because of differences in surface tension of the printed cDNA and contaminant when drying or differences in charge of the cDNA and contaminant. Figure 4b shows an accurate Cy5/Cy3 image of this spot created from the HSS concentration maps. The original Axon data produced a ratio of Cy5/Cy3 medians of 3.0 and ratio of Cy5/Cy3 means of 3.0 while the more accurate HSS data gives rise to ratios of medians and means of 7.7 and 7.5, respectively, for this particular spot.
Figure 4 Images illustrating dye separation. (a) Red/Green ratio image generated from Axon 4000B microarray scanner showing the red ring characteristic of apparent dye separation. (b) Cy5/Cy3 ratio image generated from HSS concentration maps showing uniform Cy5 and Cy3 distribution throughout the spot. (c-e) Individual component concentration maps resulting from the multivariate analysis of the HSS image data highlighting the difference in spot diameter of the brighter contaminant that results in the apparent dye separation when imaged with a filter-based microarray scanner. All fluorescence emissions in the hyperspectral images were excited simultaneously in a single scan. (c) Cy5 concentration map. (d) Cy3 concentration map. (e) Green contaminant concentration map.
Background emissions
In microarray experiments, mean or median intensities are calculated per spot to determine expression ratios. These spot intensity values are typically corrected for background emission using methods that subtract local or global estimates of background contributions. Various background correction methods exist in data analysis software, but all of these methods make one critical assumption about the data – that the background emissions are the same outside the spot as they are under the spot. This assumption is valid in an ideal situation where a perfectly homogeneous glass slide is the sole source of background emissions and the printed DNA spot is sufficiently thin and non-scattering so as not to interfere with the excitation of the glass beneath it. Unfortunately this assumption is rarely valid. Researchers have shown the results of microarray experiments are very dependant on background subtraction methods used and have theorized that local background values are not representative of the true background emission in a spot, leading to erroneous values and negative spot intensities [30]. Options such as not correcting for background, using a global background value for every spot, or using negative control values as a background can be successful in some cases, but are not robust.
Using the HSS, we have explored background emissions on many printed microarrays spanning a variety of preparation protocols. The unique ability to identify and isolate all emission sources and model the background for each pixel directly from the spectra allows us to generate pure-concentration maps of the dyes of interest without contributions from background sources. In every microarray we have scanned (9+ different labs, in-house and commercially printed slides) the background was different under the spot than around the spot. This difference can vary from a subtle decrease in intensity under the printed spot to a much more predominant intensity variation that seriously affects the accuracy of the data. Understanding the background emissions is essential to ensuring that appropriate background correction techniques are used when scans are to be performed with commercial scanners. This increased understanding of background emissions also provides the feedback necessary to alter the preparation process to minimize the background emissions present on microarray slides.
Figure 5a shows the R/G ratio image of a portion of a microarray with spatially variable, high background emissions and poor spot-specific signal in the red channel. Although only a small area is shown, the entire array contains significant variations in background intensities and patterns. Multivariate analysis of the spectra from a HSS image determines that the emission outside of the spots is predominantly from Cy5-labeled cDNA, but also includes contributions from Cy3-labeled cDNA as well as the glass substrate. In this case the non-specific interaction of the labeled cDNA with the glass substrate is most likely caused by inadequate blocking procedures. Non-specifically bound cDNA is not expected to contribute to the intensities within a spot and therefore local or even global subtraction methods would lead to spot intensities that are too low and even negative. The critical advantage of HSS analysis in this situation is that pure emission spectra are quantified and spot intensities from these images do not need separate background correction.
Figure 5 Images illustrating background problems. All images are of same area on the microarray. (a) Red/Green ratio image generated from Axon 4000B microarray scanner. The green channel intensity dominates background and spot signal. (b-d) Individual component concentration maps extracted from multivariate analysis of HSS image data. All fluorescence emissions in the hyperspectral images were excited simultaneously in a single scan. (b) Glass concentration map. Note that the spots all appear slightly dimmer than the background. (c) Cy3 concentration map. (d) Cy5 concentration map.
Two additional points should be noted about Figure 5. First, in the HSS Cy3 concentration map, Cy3 is present both in and outside of the printed cDNA spots. This is in contrast to the Cy5-HSS image, in which Cy5 is absent from many spots. We have observed a slight spectral shift in the Cy5 emission maximum under these conditions compared to other slides with successful Cy5 hybridization, suggesting that the Cy5 outside the spots may contain significant amounts of residual, unincorporated dye. Cy5 and Cy3 both exhibit a spectral shift in emission maxima upon incorporation into cDNA. This effect is illustrated in Additional File 1.
Additional File 1 compares the MCR extracted spectra from a spotted array containing only Cy5-dCTP in the spots and a second spotted array containing only Cy5-cDNA in the spots showing the effect the molecule Cy5 is attached to can have on the emission maxima of Cy5. Second, the HSS image of the glass concentration shows another phenomenon: the glass intensity is lower under very bright spots (spots with intensities > 6000 arbitrary fluorescence units on the commercial scanner we utilized) than outside of the spots. This difference is slight but consistent. We believe this is due to scattering or absorption of the laser light by the printed spot, decreasing the irradiance of the glass beneath the DNA spot.
We also observed several microarrays with high background in smears across the slide. (data not shown) In those cases, the spectra from the smears did not resemble Cy3 or Cy5, but, instead, were from a green contaminant. The smears persist across the surface of the spot and contribute to additional signal intensity in the Cy3 channel. No background correction method can adequately correct for this emission and unless the contribution from this contaminant is modelled and removed from the data, the analyzed microarray data will not be reliable.
Conclusion
The microarrays utilized in this study consisted of commercial preprinted slides or slides printed in-house using widely accepted techniques, and thus are representative of a wide cross-section of the microarrays currently employed in genomic research. The results presented in this paper demonstrate that three common anomalies cited in the microarray literature can be explained by extraneous emission sources, reinforcing the urgent need for careful attention to detail to recognize the presence of extraneous emissions that reduce accuracy of existing microarray data.
Our work illustrates the impact hyperspectral scanning can have on the accuracy and reliability of genomic data. Hyperspectral imaging, together with multivariate data analysis, is a critical tool for understanding and characterizing the sources of microarray emissions after hybridization. Our state-of-the-art HSS and multivariate algorithms permit the simultaneous identification and quantification of contaminant and background emissions in microarrays, thus markedly improving the reliability and accuracy of these data. Using these tools, we can obtain accurate concentration maps that are corrected for the effect of contaminant emissions. Perhaps more importantly, this approach provides the capability to quantify the extent and contribution of extraneous emission sources to the signal, the consequences of failing to account for them, and the insight necessary to adjust preparation protocols to prevent such problems from occurring. The utility of the HSS technology coupled with multivariate data analysis is not limited to microarrays, but could improve the understanding of data from a variety of applications based of fluorescence imaging and microscopy. In addition to providing more accurate measurements, the HSS also has potential to increase the throughput and improve analysis of microarray experiments by providing quantitation of multiple overlapping dyes. Future work will demonstrate this capability with microarrays and multiple fluorescence imaging applications.
Methods
Microarray slide preparation
The data presented represent timepoints from a large experiment performed to study gene expression in yeast cells exiting the stationary phase. Wild-type MATαS288c cells (ATCC) were grown in 100 ml YPD+A at 30°C for 7 days (stationary phase) with aeration. Cells of a 7-day culture were resuspended in fresh media prewarmed to 30°C and samples were taken at various timepoints. 7-Day cells sampled before refeeding served as the reference timepoint 0. Approximately 4 × 109 cells were collected to yield 50–200 μg total RNA. The RNA protocol is described in the web supplement. . Twenty μg of RNA was used for each labeling reaction. Cy3 and Cy5 (Amersham) labels were incorporated into reference and experimental cDNA (green and red, respectively) via first strand cDNA synthesis.
For the slides shown in this paper commercially printed CMT S288C yeast v. 1.32 arrays (Corning) were used although additional slides from other laboratories were investigated and are consistent with our findings for these arrays. The slides were pre-hybridized, hybridized, and post-processed according to procedures described in previously published methods [17].
Microarray scanning and analysis using commercial microarray scanner
Hybridized glass microarray slides were scanned with 10 μm spatial resolution using an Axon 4000B microarray scanner. Instrument settings were as follows: 100% laser power and PMT settings of 600–650 for both the red and green channels. Resulting images were analyzed using GenePix 3.0 or 4.0 (Axon) to locate and quantify the printed spots. Grids used to define spot diameters were aligned manually and diameters were adjusted to match actual spot diameters (180–240 μm). Care was taken to flag spots that were visibly compromised (dust, scratches, smears, etc.) as "bad' and these spots were not used in calculating statistics or in any future analyses. Background levels were calculated locally for each spot using the median of the background pixels. The spot intensities reported are the median values after local background subtraction.
Microarray scanning using hyperspectral microarray scanner
Following the scans with the commercial scanner, areas on the same slides were scanned using the HSS [18]. All fluorescence was excited with a 532 nm laser (<20 mW total power dispersed over a 1 mm × 0.010 mm area for line imaging) and the emission spectrum from 550 to 900 nm was collected in line scanning (push-broom) mode with a scan speed of 0.04 mm/s and a CCD rate of 4 frames/sec with EMCCD gain set to 128. This scanning protocol produced images with 10 μm spatial resolution in the motion direction. A bin by 2 configuration was used in the vertical dimension of the CCD in order to assure a spatial resolution of 10 μm in that dimension using a 10× objective (Nikon PlanApochromat) and taken into consideration our overall system magnification (6× with 10× objective). Instrument control/data collection was accomplished using custom software written in the C++ environment.
All image preprocessing and multivariate analysis was performed using in-house written Matlab software. (Matlab v 6.5, Mathworks, Inc) HSS images were preprocessed to locate and remove cosmic events using a modified, 5-point median filter over the immediate area of a cosmic spike only. To correct for spectral image curvature and to calibrate the wavelength axis reference images of neon and krypton gas discharge lamps were used. Signals from both lamps were combined for accurate calibration over the large spectral range. Principal component analysis was performed on the resulting preprocessed images and the number of spectral components was determined by inspecting the eigenvalues in a Scree plot [31]. A constrained, alternating least squares MCR analysis was performed using the number of components from the PCA analysis plus an additional offset (linear baseline) component. The MCR algorithm also requires initial estimates for the spectra and/or the concentrations. The algorithm was initialized with either the positive values of the principal components or spectra identified from previous image analyses. Non-negativity constraints were applied to all components for both concentrations and spectra and spectral equality constraints were applied to the offset component throughout the analysis and to portions of other component spectra as needed to prevent physically unrealistic solutions (i.e. Cy5 emission signal at wavelengths <590 nm or spectral or concentration nonnegativity). All constraints were applied using rigorous least squares methods. One hundred to five hundred iterations were performed in the MCR analysis, although the solution often converged in fewer iterations. Resulting component concentration maps were scaled to be directly comparable to the commercial scanner results by convolving the hyperspectral data recreated by the concentration maps of the emissions of interest with the optical filter function of each of the commercial scanner channels. The results were integrated and scaled to achieve total integrated intensities equal to the per channel values obtained with the commercial scanner. These scaled concentration maps were then written as individual tif files and read into the GenePix software where spots were identified and quantitated for each of the components using the methods presented above for commercial microarray data.
Authors' contributions
JAT performed the hyperspectral scanning, multivariate data analysis, spectral interpretation, and prepared the manuscript. MBS designed and constructed the hyperspectral microarray scanning and wrote all data acquisition software. DMH conceived the study, developed the multivariate data analysis tools and techniques and aided in the interpretation of the results. ADA performed commercial microarray experiments. MJM performed commercial microarray experiments and data analysis, and assisted in drafting the manuscript. MWW conceived the study, coordinated microarray experiments, provided biological expertise and interpretation, and assisted in manuscript writing. All authors read and approved the final manuscript.
List of Abbreviations
hyperspectral scanner (HSS)
multivariate curve resolution (MCR)
charge-coupled device (CCD)
electron multiplier CCD (EMCCD)
Red/Green ratio (R/G)
Supplementary Material
Additional File 1
Spectral shifts Cy5 upon incorporation into cDNA MCR extracted spectral profiles of Cy5-dCTP (blue trace) and Cy5-cDNA (green trace). Spectra profiles were obtained from MCR analysis of hyperspectral images from two individual spotted arrays manufactured in-house containing spots of only Cy5 bound to dCTP and Cy5 incorporated into cDNA. Spectral traces are normalized for maximum intensity equal to one
Click here for file
Acknowledgements
The authors thank Gabriel Quinones at the University of New Mexico for his assistance in printing and hybridizing the microarrays. The authors also thank Michael R. Keenan and Mark Van Benthem at Sandia National Laboratories for their development of rigorous, efficient multivariate curve resolution algorithms and analysis software.
Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000. This work was funded in part by the US Department of Energy's Genomics: GTL program under project, "Carbon Sequestration in Synechococcus Sp.: From Molecular Machines to Hierarchical Modeling," and in part by a grant from the W. M. Keck Foundation to DMH. The work was also funded by NSF (MCB-0092364) and NIH (HG02262 and GM67593) to MWW and NSF(DBI-0314435) to MJM.
==== Refs
Schena M Shalon D Davis RW Brown PO Quantitative monitoring of gene expression patterns with a complementary DNA microarray Science 1995 270 467 470 7569999
Schena M Hames BD DNA Microarrays: A Practical Approach The Practical Approach Series 1999 New York , Oxford University Press, Inc. 210
Schena M Microarray Biochip Technology 2000 Natick , Eaton Publishing 298
Murphy D Gene expression studies using microarrays: Principles ; problems ; and prospects Advances in Physiology Education 2002 26 256 270 12443997
Hegde P Qi R Abernathy K Gay C Dharap S Gaspard R Hughes JE Snesrud E Lee N Quackenbush J A concise guide to cDNA microarray analysis Biotechniques 2000 29 548 562 10997270
Hughes TR Shoemaker DD DNA microarrays for expression profiling Current Opinion in Chemical Biology 2000 5 21 25 11166643 10.1016/S1367-5931(00)00163-0
Tseng GC Oh M Rohlin L Liao JC Wong WH Issues in cDNA microarray analysis: quality filtering, channel normalization, models of variations and assessment of gene effects Nucleic Acids Research 2001 29 2549 2557 11410663 10.1093/nar/29.12.2549
Wang WJ Ghosh S Guo SW Quantitative quality control in microarray image processing and data acquisition Nucleic Acids Research 2001 29 U32 U39 10.1093/nar/29.15.e75
Kerr MK Churchill GA Experimental design for gene expression microarrays Biostatistics 2002 2 183 201 12933549 10.1093/biostatistics/2.2.183
Balazsi G Kay KA Barabasi AL Oltvai ZN Spurious spatial periodicity of co-expression in microarray data due to printing design Nucleic Acids Research 2003 31 4425 4433 12888502 10.1093/nar/gkg485
Lee MLT Kuo FC Whitmore GA Sklar J Importance of replication in microarray gene expression studies: statistical methods and evidence from repetitive cDNA hybridizations Proceedings of the National Academy of Sciences 2000 97 9834 9839 10.1073/pnas.97.18.9834
Kerr MK Churchill GA Statistical design and the analysis of gene expression microarray data Genetical Research 2001 77 123 128 11355567 10.1017/S0016672301005055
Schuchhardt J Beule D Malik A Wolski E Eickhoff H Lehrach H Herzel H Normalization strategies for cDNA microarrays Nucleic Acids Research 2000 28 e47 10773095 10.1093/nar/28.10.e47
Kerr MK Martin M Churchill GA Analysis of variance for gene expression microarray data Journal of Computational Biology 2000 7 819 837 11382364 10.1089/10665270050514954
Yang YH Dudoit S Luu P Lin DM Peng V Ngai J Speed TP Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation Nucleic Acids Res 2002 30 e15 11842121 10.1093/nar/30.4.e15
Brown CS Goodwin PC Sorger PK Image metrics in the statistical analysis of DNA microarray data Proc Nat Acad Sci USA 2001 98 8944 8949 11481466 10.1073/pnas.161242998
Martinez MJ Aragon AD Rodriguez AL Weber JM Timlin JA Sinclair MB Haaland DM Werner-Washburne M Identification and removal of contaminating fluorescence from commercial and in-house printed DNA microarrays Nucleic Acids Research 2003 31 e18 12582263 10.1093/nar/gng018
Sinclair MB Timlin JA Haaland DM Werner-Washburne M Design, construction, characterization, and application of a hyperspectral microarray scanner Applied Optics 2004 43 2079 2089 15074416 10.1364/AO.43.002079
Schultz RA Nielsen T Zavaleta JR Ruch R Wyatt R Garner HR Hyperspectral imaging: A novel approach for microscopic analysis Cytometry 2001 43 239 247 11260591 10.1002/1097-0320(20010401)43:4<239::AID-CYTO1056>3.0.CO;2-Z
Zimmerman T Rietdorf J Pepperkok R Spectral imaging and its applications in live cell microscopy Federation of European Biochemical Societies Letters 2003 546 87 92 12829241
Bogdanov V In-line complete hyperspectral fluorescent imaging of nucleic acid molecules 2001 USA , Orchid BioSciences, Inc.
Garner HR Hyperspectral slide reader 2000 USA , Board of Regents, The University of Texas System
Huebschman ML Schultz RA Garner HR Characteristics and capabilities of the hyperspectral imaging microscope IEEE Engineering In Medicine And Biology Magazine 2002 21 104 117 12222107 10.1109/MEMB.2002.1032647
Bro R DeJong S A fast non-negativity-constrained least squares algorithm Journal of Chemometrics 1997 11 393 401 10.1002/(SICI)1099-128X(199709/10)11:5<393::AID-CEM483>3.3.CO;2-C
Benigini R Giuliani A Quantitative modeling and biology American Journal of Physiology 1994 266 R1697 R1704 8203653
Schoonover JR Marx R Zhang SL Multivariate curve resolution in the analysis of vibrational spectroscopy data files Applied Spectroscopy 2003 57 154A 170A 14658670 10.1366/000370203321666461
Kotula PG Keenan MR Michael JR Automated analysis of SEM X-Ray spectral images: a powerful new microanalysis tool. Microscopy & Microanalysis 2003 9 1 17 12597783 10.1017/S1431927603030058
Van Benthem MH Keenan MR Haaland DM Application of equality constraints on variables during alternating least squares procedures Journal of Chemometrics 2002 16 613 622 10.1002/cem.761
Haaland DM Timlin JA Sinclair MB Van Benthem MH Martinez MJ Aragon AD Werner-Washburne M Multivariate curve resolution for hyperspectral image analysis:applications to microarray technology: San Jose, CA.
2003 4959 International Society for Optical Engineering 55 66
Yang YH Buckley MJ Dudoit S Speed TP Comparison of methods for image analysis on cDNA microarray data Journal of Computational and Graphical Statistics 2002 11 108 136 10.1198/106186002317375640
Cattell RB The Scree test for the number of factors Multivariate Behavioral Research 1966 1 245
| 15888208 | PMC1156888 | CC BY | 2021-01-04 16:32:45 | no | BMC Genomics. 2005 May 11; 6:72 | utf-8 | BMC Genomics | 2,005 | 10.1186/1471-2164-6-72 | oa_comm |
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-731590449310.1186/1471-2164-6-73SoftwareSoftware for optimization of SNP and PCR-RFLP genotyping to discriminate many genomes with the fewest assays Gardner Shea N [email protected] Mark C [email protected] Pathogen Bio-Informatics, Lawrence Livermore National Laboratory, Livermore, CA, USA2005 16 5 2005 6 73 73 27 7 2004 16 5 2005 Copyright © 2005 Gardner and Wagner; licensee BioMed Central Ltd.2005Gardner and Wagner; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Microbial forensics is important in tracking the source of a pathogen, whether the disease is a naturally occurring outbreak or part of a criminal investigation.
Results
A method and SPR Opt (SNP and PCR-RFLP Optimization) software to perform a comprehensive, whole-genome analysis to forensically discriminate multiple sequences is presented. Tools for the optimization of forensic typing using Single Nucleotide Polymorphism (SNP) and PCR-Restriction Fragment Length Polymorphism (PCR-RFLP) analyses across multiple isolate sequences of a species are described. The PCR-RFLP analysis includes prediction and selection of optimal primers and restriction enzymes to enable maximum isolate discrimination based on sequence information. SPR Opt calculates all SNP or PCR-RFLP variations present in the sequences, groups them into haplotypes according to their co-segregation across those sequences, and performs combinatoric analyses to determine which sets of haplotypes provide maximal discrimination among all the input sequences. Those set combinations requiring that membership in the fewest haplotypes be queried (i.e. the fewest assays be performed) are found. These analyses highlight variable regions based on existing sequence data. These markers may be heterogeneous among unsequenced isolates as well, and thus may be useful for characterizing the relationships among unsequenced as well as sequenced isolates. The predictions are multi-locus. Analyses of mumps and SARS viruses are summarized. Phylogenetic trees created based on SNPs, PCR-RFLPs, and full genomes are compared for SARS virus, illustrating that purported phylogenies based only on SNP or PCR-RFLP variations do not match those based on multiple sequence alignment of the full genomes.
Conclusion
This is the first software to optimize the selection of forensic markers to maximize information gained from the fewest assays, accepting whole or partial genome sequence data as input. As more sequence data becomes available for multiple strains and isolates of a species, automated, computational approaches such as those described here will be essential to make sense of large amounts of information, and to guide and optimize efforts in the laboratory. The software and source code for SPR Opt is publicly available and free for non-profit use at .
==== Body
Background
Microbial forensics and epidemiology is important in tracking the source of a pathogen, whether the disease is a naturally occurring outbreak or part of a criminal investigation. Polymorphisms among isolates or strains provide information as to the origin, phylogenetic relationships, or transmission patterns of those isolates [1]. The 2001 anthrax attacks highlight the importance of rapid forensic identification of the source of an agent used in a bioterrorism event. Sequencing HIV fragments indicated that a Florida dentist probably infected at least six of his patients with HIV [2]. A series of court cases in Scotland center on accusations that hospital staff are transmitting methicillin resistant staphylococcus aureus to patients [3]. The Centers for Disease Control and Prevention has also developed a network for molecular subtyping, or fingerprinting, of foodborne pathogens [4]. Because SNPs, insertion/deletion mutations, or sequence repeats may affect or be linked with phenotypic traits such virulence or antibiotic resistance, analysis of variance in polymorphic markers may also contribute to improvements in the diagnosis and treatment of infectious diseases[5].
Increasing availability of genomic sequence data makes it possible to predict regions of a genome that display variation among strains or isolates [1]. In the event of a suspected biothreat outbreak, the agent would be completely sequenced; however, full genome sequencing may require weeks or more. Ideally, there should be information immediately available about the hotspots of variation, the key sequence regions or assays that can discriminate among the possible sources of the agent, based on existing sequence data. This demands a full-genome analysis of available sequence data to predict discriminating markers among strains or isolates. Knowledge of and validated assays to query these variable regions for genotyping analyses could then be used to rapidly classify an unknown isolate in terms of its relationship to the already-characterized strains. These results could be available within hours, long before full sequence information becomes available. Once full-sequence information is generated, reliable, automated tools are required to find how this sequence differs or is similar to other strains. Recently, Budowle and colleagues stated that there is a "need for an infrastructure with analytical tools and knowledge bases to rapidly provide investigative leads..." [6].
The needs described above demand a full knowledge of all the SNPs and fragment length polymorphisms (e.g. detectable by PCR-RFLP analyses) that distinguish known isolates. SNPs and PCR-RFLP analyses have been used extensively in genotyping for forensic and epidemiological applications [7-10]. Although extensive experimental bench work or human examination of multiple sequence alignments can illuminate such variations, non-automated analysis is tedious and error-prone, especially for long sequences or when more than a few sequences are available. Existing software programs related to forensics focus on human crime forensics, paternity investigation, and so on, and do not enable full-genome prediction of marker regions or predict the combinations of variable regions that facilitate maximal isolate discrimination with a minimal number of assays [11].
To address this need, we have developed an automated forensic pipeline for SNP and PCR-RFLP optimization, called SPR Opt (SNP PCR-RFLP Optimization). To our knowledge, this is the first comprehensive, automated, computational tool that performs the following four steps:
1) identifies all SNPs or PCR-RFLP variations in a set of input sequences that can be as long as whole microbial genomes,
2) groups the co-segregating markers into haplotypes,
3) performs combinatoric analysis on the haplotypes to generate multi-locus solutions to maximally discriminate each of the input sequences from the others,
4) uses simulated annealing to find the best solutions using the smallest total number of haplotypes (that is, the fewest total assays) to discriminate all the input genomes to the maximum degree possible.
Both SNP and PCR-RFLP solutions computed by the SPR Opt software include multiple loci when necessary for maximal sequence discrimination. In this paper, input sequences may also be referred to as genomes. However, the input need not be complete genomes; it can be gene sequences or other fragments from a number of isolates. Here a SNP is considered to be a single polymorphic base surrounded by conserved upstream and downstream sequence [9]. The length of the conserved sequence surrounding a SNP is specified by the user. In future versions of the software under development, a less strict SNP definition will be allowed in which requirements for conservation surrounding the SNP position are relaxed, allowing polymorphisms and indels in the region immediately surrounding the SNP position.
SPR Opt also predicts PCR-RFLP assays to discriminate the input sequences. Those primer pair and restriction enzyme combinations to maximally discriminate the input sequences are identified. For the PCR step, it is assumed that primer pairs must be conserved among all the input sequences. The resulting amplicons are examined for length polymorphisms or sequence variations that would result in differences in fragment length distributions after restriction digest among the multiple input sequences (RFLP). This software elucidates all PCR-RFLP detectable variations, whether they are caused by insertions or deletions of non-repeated sequence, tandem repeats, non-tandem repeats, microsatellites, mutations that alter a restriction site, and so on.
For both SNP and PCR-RFLP analyses, some conservation is required among the input sequences surrounding the variable site, whether it be conserved bases up and downstream of the SNP, or conserved primers surrounding the fragment length polymorphism or RFLP. This ensures that there are not false negatives because the surrounding sequence was not amplified or otherwise detected. For example, detection of SNPs may be performed using a microarray with oligos chosen so that the central base is a SNP, and there are 4 oligos for each conserved surrounding sequence, each representing the SNP position filled with A, T, C, or G, respectively. If there is not a hybridization signal from any of the four oligos, then this serves as an alert that there may be a problem with the reaction conditions or the sample. In contrast, if conserved oligos (sequences surrounding the SNPs) are not used, then it is unclear whether the sample is a variant without that particular SNP, or whether the sample is degraded, the reaction conditions unsuitable, or the target species is not present.
Here it is assumed that a haplotype is a group of markers (SNP or PCR-RFLP) that segregate in the same pattern across the genomes. Any single marker in a given haplotype indicates the identities of the other markers in this haplotype, so that every marker within the haplotype need not be examined. Here it is not assumed that the markers within a given haplotype cluster spatially in any pattern within a given sequence. That is, markers in the same haplotype may be distributed evenly or randomly, and do not necessarily occur within blocks of contiguous sequence. Haplotypes do not necessarily correlate with "hotspots" of mutation. Markers in the same haplotype may be likely to be linked, located in close proximity and inherited as a group, but it is not assumed that this is the case. Indeed, if markers in the same haplotype are separated in the genome, it may indicate that a recombination event has occurred. Alternatively, this pattern could be a result of selection or chance. Such an investigation of how positional information of markers in the same haplotype may indicate recombination is beyond the scope of this paper.
Although the first part of this software depends on locating the SNP or PCR-RFLP sites (step 1 above), the "backend" (steps 2–4 above) of this computational forensics pipeline depends only on the haplotypes. Thus, while the tools are currently coded to work for the SNP or PCR-RFLP analyses described here, any type of marker such as microarray hybridization patterns, microsatellite markers, or even chemical or physical features that differ among isolates could be categorized into haplotypes and fed into the back end of the combinatorial analyses described below.
One question that is investigated here is that of how well phylogenetic trees based on the haplotype splits of PCR-RFLPs or SNPs correspond with those generated from full genome sequence alignments or phenotypic differences. The results presented indicate that trees built solely on forensic markers do not closely match those based on full genome alignments. This finding has important implications for tree generation and interpretation in the absence of sufficient genomic sequence information, since empirical forensic techniques are often used for tree prediction. A second analysis presented is the frequency that different restriction enzymes result in fragment length polymorphisms for the SARS and mumps data.
The impacts of this software are the following: 1) The software provides computational guidance as to an optimal set of assays for genotyping the isolates, particularly helpful when large numbers of sequences or genomes are available. 2) All the SNP or PCR-RFLP variations in the available data are found and grouped into haplotypes. This includes identification of the sequence surrounding a SNP that will be useful in designing the assay (for example, the oligos or primers for an array, ligation, or single base extension reactions), or primer prediction and restriction enzyme selection. If the available sequence data is limited, there is no guarantee that the primers identified will be found in unsequenced strains, although requiring conservation among the sequenced isolates increases the chances of conservation among unsequenced isolates as well.
The application of SPR Opt is illustrated for two viruses for which multiple genomes are publicly available. The software and source code is publicly available and free for non-profit use at .
Results
The output files listed in Table 2 containing the details of these analyses are provided as supplementary information for the web . The main results are summarized in Tables 3, 4, 5. The large number of genomes, variable sites, and character haplotypes illustrate the utility of this software for focusing in on the most informative combinations of sites. Only some of the mumps virus or SARS virus genomes can be uniquely discerned from other genomes of the same species. Unresolved clusters are given in Additional file 2. Computational prediction of PCR amplicon length variation (without restriction digest) for mumps virus indicates there are no variable length amplicons using the parameter values and definitions used here. Therefore, this method when restriction digest is omitted is not appropriate for genotyping mumps virus.
Table 2 Description of the output files created. Examples for the organisms analyzed here can be found at
File Name Content Description
SNPs_all, FLPs_all list of all the SNPs or PCR-RFLPs found in the input genomes
genome_groups lists the genome groupings that correspond to each of the character haplotypes
character_haplotypes lists the co-segregating SNPs or PCR-RFLPs that characterize each haplotype
all_discriminating_sets for each genome, lists haplotype combinations ("sets") that will discriminate the specified genome to the maximum degree possible
sim_anneal_results_summary lists combinations of sets of haplotypes to discriminate all the input genomes to the maximum degree possible using the fewest haplotypes. Each row is a unique combination that has the best score found, where the score is the number of haplotypes required.
Table 3 Summary of predictions for SNPs
Organism Number Sequences Number Sites Number Haplotypes Number Unresolved clusters Minimum Number Assays
Mumps 17 171 85 11 10
SARS 102 218 164 65 114
The minimum number of assays indicates the fewest haplotypes that need to be queried to discriminate all the input genomes to the maximum degree possible, that is, the number of assays that would need to be done to classify an unknown isolate in terms of its similarity to the currently sequenced strains using this type of forensic marker.
Table 4 Summary of predictions for PCR without restriction digest
Organism Number Variable Amplicons No. Combinations of Amplicon X Unique Fragment Length Distributions No. Haplo-types No. Unresolved clusters Min. No. Assays
Mumps 0 0 0 1 NA
SARS 39 178 51 26 31
The number combinations of Amplicon X Unique Fragment Length Distributions is the number of unique fragment length distributions summed over all variable amplicons.
Table 5 Summary of predictions for PCR-RFLP
Organism Number Variable Amplicon X Enzyme Combinations Number Combinations of Amplicon X Enzyme X Unique Fragment Length Distributions No. Haplo-types No. Unresolved clusters Min. No. Assays
Mumps 1,070 3,113 288 13 13
SARS 1,694 7,344 440 75 99
The number of combinations of Amplicon X Enzyme X Unique Fragment Length Distributions is summed over all variable amplicon, enzyme combinations in the PCR-RFLP analyses.
A comparison of the SARS phylograms created using full genome sequences, SNPs, or PCR-RFLPs (Figures 2, 3, 4, 5) illustrate that although there are similarities within the fine branch structure showing the relationship of very similar genomes, the basic structure of the trees differs. Genetic distances, indicated by the lengths of the branches, vary among the trees, and the particular genome clusters predicted to be the least similar (longest branches) contrast among the trees. A tree created from an optimal solution for SNP forensics has little branching structure and homogeneous branch lengths, and does not appear very similar to or to provide as much information as trees created from all the SNP data or from the multiple genome alignment.
Figure 2 Phylogram of SARS created based on a multiple sequence alignment.
Figure 3 Phylogram of SARS created based on all SNP haplotypes.
Figure 4 Phylogram of SARS created based on all PCR-RFLP haplotypes.
Figure 5 Phylogram of SARS created based on only SNP haplotypes in one optimal solution set (that maximally discriminates each of the input genomes using the fewest total haplotypes).
Table 6 shows that the ranking of enzymes that most frequently generate RFLP variations in SARS virus are similar but not identical to the rankings for generating RFLP variations in mumps virus.
Table 6 Enzyme frequency for creating fragment length polymorphisms in SARS and mumps viruses
Enzyme SARS Mumps Enzyme SARS Mumps
Tru9I 335 149 NciI 54 32
Hsp92II 320 98 EcoRV 52 36
Rsal 318 101 TthlllI 52 15
DdeI 317 104 EclHKI 52 11
Alul 270 135 Csp45I 51 29
Bsp1286I 227 48 Bbul 47 14
TaqI 213 121 SphI 47 14
CfoI 195 54 Bst98I 44 20
HhaI 195 54 XbaI 41 28
MboI 190 135 XmnI 39 29
Sau3AI 190 135 NheI 39 14
HinfI 179 105 AccB7I 39 8
AccI 170 45 NcoI 35 8
VspI 161 62 HpaI 34 8
BstOI 160 77 EcoRI 32 23
MspAI 140 55 AvaI 30 32
Styl 133 39 NruI 24 4
Sau96I 132 74 BclI 23 33
AvalI 128 57 HindIII 22 27
SinI 128 57 Sall 21 15
HincII 126 24 BssHII 20 4
BsrSI 119 49 Eco72I 19 6
RvuII 114 40 Smal 17 12
Dral 113 26 Xmal 17 12
Pstl 113 13 XhoI 13 14
XhoII 112 69 Mlul 13 0
Haell 112 18 Clal 9 12
HaelIl 110 83 SnaBI 8 4
Scal 103 16 Nael 6 8
HpaII 91 82 NgoMI 6 8
MspI 91 82 Eco47III 3 18
Sspl 91 33 Stul 3 17
BanII 91 18 BamHl 0 24
EcoICRI 91 8 AcyI 0 18
Spel 85 21 BalI 0 18
Ndel 83 20 Hsp92I 0 18
Nsil 82 8 Bgll 0 12
Acc65I 74 9 Agel 0 11
KpnI 74 9 BsaOI 0 9
BstXI 73 6 SacI 0 6
Bsu36I 72 16 AatII 0 4
BsrBRI 65 41 Apal 0 4
BglIl 64 35 Pvul 0 4
BanI 62 17 BstZI 0 3
BstEII 61 8 Eco52I 0 3
Alw441 59 6 SacII 0 2
Discussion
PCR without restriction digestion was inferior to PCR-RFLP and SNPs for genome discrimination of the viruses examined, although in additional analyses of bacterial genomes, PCR without digestion was adequate for complete genome discrimination (unpublished). When it is possible to choose from a very large number of restriction enzymes, PCR-RFLP may enable a greater level of discrimination than SNP analysis, illustrated by the result that the genomes can be subdivided into more, smaller unresolved clusters using PCR-RFLP than using SNPs. However, this is not always the case: for example the two SARS genomes PUMC03 and Sino1-11 can be discriminated using SNPs but not using PCR-RFLP. If only a small number of restriction enzymes are available, then analyses using SPR Opt indicate that SNPs outperform PCR-RFLP (unpublished). In most cases, either SNPs or PCR-RFLPs can discriminate most genomes, with minor differences in the exact unresolved clusters predicted. In these cases, one may need to use both techniques in order to discern different sets of similar genomes.
Contrasting the phylogenetic trees generated from full genomes, SNPs, or PCR-RFLPs suggests that caution is needed in assessing genome divergence based on forensic/epidemiological data in lieu of full genome sequences, since the isolates predicted to be most divergent may differ across the three measures of variation. In addition, if only the subset of SNPs determined to be optimal for forensic discrimination are queried, phylogenetic trees based on this information may not be representative of the true phylogenic relationships, and may result in particularly poor estimation of branching structure and branch lengths.
Highly heterogeneous viruses like many of the single-stranded RNA viruses (e.g. human immunodeficiency virus, hepatitis virus species, poliovirus, etc.) must be subdivided into clades or other sub-groupings of genetically similar strains even before this software can be applied. Lack of conserved sequence upstream and downstream of a SNP, or lack of conserved primers for PCR amplification, may thwart attempts at SNP or PCR-RFLP discovery at the species level. Instead, one requires subgroups of genomes that are sufficiently similar to locate conserved regions surrounding the forensically informative sites. A future version of the software under development will employ a less restrictive definition of SNPs, allowing some variation surrounding the SNP position.
In addition to SNP and PCR-RFLP analysis like the examples presented here for a given target species, there are many questions one could address with the aid of this software. For example, the character haplotypes and genome groups generated by SPR Opt contain a wealth of information. If there are differences in virulence, host range, or other interesting phenotypic traits, one may examine the list of genome groups to see if there are any whose genome membership corresponds with the phenotypic variation. If so, these may be interesting regions for further biological investigation. For example, in which genes do the corresponding character haplotypes land? Do the nucleotide variations translate into protein sequence differences? One could also take a complementary approach, asking whether SNPs or PCR-RFLPs are clustered in certain genes or intergenic regions. It would also be interesting to examine whether or not SNP locations within a species correspond to regions of relative inter-specific conservation or variation. As mentioned in the introduction, the distribution of co-segregating markers in a given haplotype across the genome sequence might be used to look for evidence of recombination events or correlated selection on multiple genes, as might be observed if genes are in the same pathway or affected by the same environmental factors.
Conclusion
In conclusion, bioinformatic software called SPR Opt is described to optimize SNP and PCR-RFLP analyses in order to provide the maximum amount of genotyping information from the fewest assays at the bench. SPR Opt requires as input a set of sequences and their multiple sequence alignment. This software not only predicts the variable sites based on input sequence data, but also groups these into co-segregating haplotypes and provides guidance as to the ways in which these may maximally discriminate genomes using the fewest possible assays. These are computationally challenging problems that are solved using a bit vector intersection approach to determine sets of haplotypes to maximally discriminate each input genome, as well as parallel simulated annealing to select a subset of the many possible solutions that will enable users to extract the most information from the fewest forensic tests. Analyses of two viruses were presented, and a number of potential investigations using this software are suggested. This is the first comprehensive tool to optimize the selection of forensic markers to maximize information gained from the fewest assays, accepting whole or partial genome sequence data as input. As more sequence data becomes available for multiple strains and isolates of a species, automated, computational approaches such as those described here will be essential to make sense of large amounts of information, and to guide and optimize efforts in the laboratory.
Methods
The input required
The required input to the software is a fasta-formatted file of all the input genomes and a multiple sequence alignment. A consensus "gestalt" is automatically constructed from the multiple sequence alignment of the input genomes, in which conserved bases are indicated as letters (A, C, G, or T) and dots indicate positions where all input sequences do not agree [12]. Where there are insertions or deletions in the alignment, the corresponding positions in the conservation gestalt contain dots.
SNPs
A SNP is defined to be a single polymorphic base surrounded by conserved upstream sequence of length min_len_upstream bases and conserved downstream sequence of length min_len_downstream bases. By conserved sequence, it is meant that the sequence is identical among all input genomes. If the sequence surrounding a SNP occurs more than once or is absent in any of the input genomes (checking both the plus and minus strands), the position is deleted from consideration as a SNP. A polymorphic base differs in one or more of the input genomes. If min_len_downstream and min_len_upstream are greater than zero, then this is a strict definition of SNPs. Insertions or deletions of even a single base are not considered SNPs. The parameters min_len_upstream and min_len_downstream are user-specified. The software finds SNPs using the conservation gestalt; in the conservation gestalt, any dot surrounded by at least min_len_upstream and min_len_downstream letters is a candidate SNP. The sequences indicated by the letters upstream and downstream of the candidate SNP are stored. Using regular expression pattern matching in PERL, the location of the SNP is determined in each of the input genomes. If an exact match to the conserved sequence surrounding the SNP does not occur exactly once in each of the input sequences, then the SNP and its surrounding sequence is eliminated from further consideration. Thus, the SNP is defined by its conserved surrounding sequence, and the position of a SNP within a genome may differ among the genomes.
If a less restrictive definition of a SNP is required, it is possible to set either (but not both) min_len_downstream or min_len_upstream equal to 0, and thus pick up more regions, including those that may be the beginning of insertion/deletion variations among genotypes. This might be a desired approach for the Single Base Extension assay, since only conservation immediately 5' of the variable position matters [9].
The choice of the length requirement for the surrounding conserved sequence is very important. If min_len_upstream and min_len_downstream are too short, then sequence surrounding the SNP is likely to be repeated within a given genome. Since each of the repeats could have a different character at the SNP position, if these candidates are not eliminated then it is not possible to distinguish the case of a sample containing a mixture of multiple genotypes from the case of a sample containing a single strain that has multiple occurrences of the sequence surrounding a candidate SNP position with different bases filling the variable position. In contrast, if min_len_upstream and min_len_downstream are too long, then SNP variation may be overlooked because candidate SNPs are surrounded by an insufficient number of conserved bases. If a microarray chip platform using oligos of length 25 bases is to be used for the assay, then setting min_len_upstream and min_len_downstream equal to 12 will predict 25-mers with the central base position the SNP.
This software excludes degenerate bases (e.g. R = G or A, Y = T or C, etc.) indicated within a given genome from consideration as SNPs. Degenerate bases may be due to polymorphic populations within a given strain or to low quality sequencing, and thus do not deliver a confident characterization of differences between strains.
PCR-RFLPs
A PCR-RFLP variation occurs if PCR amplicons or the fragments that result after restriction digest of such amplicons have a different length distribution among the input genomes. Fragment length distributions differ if they have different numbers of fragments or if any of the fragments differ in length.
The fragments are generated by first determining amplicons in which the forward and reverse primers are conserved among all the input genomes. If no restriction digest is to be applied, then an amplicon must differ in length among the input genomes. With restriction digest, the distribution of fragment lengths after cutting the amplicon with restriction enzymes must differ among the genomes. The parameter num_restriction_enzymes may be set equal to 0, 1, 2, or 3, as will be described below. Length variation must be detectable by electrophoresis, since there is a limit to the precision to which lengths can be determined. Here, if all the fragments in each of the genomes are less than 50 bases long, or if there are no differences in fragment lengths among any of the genomes that are at least precision bases, then the amplicon+enzyme combination is not considered to be sufficiently variable.
In order to find candidate regions for PCR-RFLP variation, sequence fragments that may contain a PCR-RFLP marker are selected from the conservation gestalt. These are chosen to be just over the max_amplicon_length, to have at least one variable position (a dot in the gestalt file), and are chosen using a sliding window moving at least jump bases from the start of the previous window. If jump is too large, PCR-RFLP sites may be missed, and if jump is too small, the same insertion/deletion variation may be counted more than once if it is contained within more than one pair of primers. However, all instances of counting the same variable site will appear within the same haploblock, so a simple examination of the markers within a given haploblock sorted by position should indicate those that query the same site as a result of using too small a value for the jump parameter. The software may also be run several times with different values of jump, and the largest value of jump that still provides the greatest discrimination among genomes may be selected.
Second, conserved primer pairs are selected from these fragments using MIT's primer3 software with user-specified parameters. Third, genomes are searched for an exact match to each primer pair on both the plus and minus (reverse complemented) strands, and those primer pairs are discarded in which 1) one or both primers are absent from any of the genomes, 2) one or both primers occur more than once in any of the genomes (with no mismatches), or 3) the forward and reverse primers are too far apart to reliably generate an amplicon in one or more of the genomes. In the analyses here, it is assumed that any distance longer than 1200 bases is too long for amplification, which is reasonable if a short elongation time in the PCR thermocycle is to be used. This value can be changed in the source code.
Amplicons can be cut by 0, 1, 2 or 3 enzymes simultaneously (or sequentially before being run through electrophoresis), as a user-specified option num_restriction_enzymes. Simulated cutting is performed computationally using the regular expression pattern matching function in Perl, and it is assumed that cuts occur in all locations where a given enzyme sequence occurs (that is, the DNA is exposed to the enzyme for a sufficient duration to cut at all the sequence-specific sites).
If the number of restriction enzymes is set to 0, then in the output files the restriction enzyme is indicated as "NONE", and the PCR amplicons must vary in length among the sequences without any restriction digest. If num_restriction_enzymes = 1, then a given PCR product may be digested by only one enzyme at a time before electrophoresis or other empirical measurement of fragment lengths. That is, digestion of the original PCR products with many alternative enzymes may be performed as long as each digestion is followed by its own fragment length measurement. The final solution guiding how to discriminate all the input sequences may involve a number of different enzymes. But when num_restriction_enzymes = 1, digestion is always done only with one enzyme at a time before fragment lengths are measured.
If num_restriction_enzymes = 2 or 3, then digestion can occur with combinations of 2 or 3 enzymes, respectively, before the measurement of fragment lengths occurs after each digestion with a given combination of enzymes. Enzyme combinations should be examined by the user to make sure the buffers and reaction conditions are compatible for all the enzymes in the combination, as the software does not assess enzyme compatibility. Although setting num_restriction_enzymes = 2 or 3 is allowed, it is not clear that users ever need to use these options, since with num_restriction_enzymes = 1, many different enzymes may contribute to the total solution of discriminating all the input genomes, as long as they are not applied simultaneously. All results reported here were computed using num_restriction_enzymes = 1.
The restriction enzymes to be considered are specified by the user. Currently, the software is not implemented for non-palindromic restriction enzymes, although this could be added as a minor modification if required.
Procedures implemented by the software
1) Find all SNP or PCR-RFLP sites
First, the software calculates all SNPs or all PCR-RFLPs, as defined above. These are listed in the files SNPs_all or FLPs_all, respectively. "FLP" refers to Fragment Length Polymorphisms, whether the polymorphisms result from PCR amplification or restriction digest length differences. The total number of SNPs is counted as the number of positions that are variable. For PCR-RFLPs, the software reports the total number of unique combinations of primer pair sequences and restriction enzyme(s) that yield variation in fragment length distributions among the genomes, as well as the number of unique fragment length distributions (summed across primer pairs and restriction enzyme combinations).
2) Cluster SNPs or PCR-RFLPs into co-segregating groups called haplotypes
The second computation the software performs is to divide the SNPs (or PCR-RFLPs) into co-segregating groups of markers called character haplotypes. That is, any two variable markers that distinguish the same set of genomes, and thus provide equivalent forensic/epidemiological information, are considered members of the same character haplotype. The genomes that co-segregate for a particular character haplotype is called the genome group. Genomes are members of multiple genome groups that may overlap in membership, and there may be genome groups that are proper subsets of other genome groups. For easy association, the genome group identification number is the same as the character haplotype identification number. The algorithm to generate the list of character haplotypes and genome groups works in the following way: at each variable site, genomes are grouped by the marker identity ("allele", e.g. SNP character) each genome contains at that locus. Every time a new clustering pattern of genomes occurs, the cluster of genomes defines a new genome group, and the marker is stored in the character haplotype associated with that genome group. If a clustering pattern has already been observed at a locus previously examined, then that allele is added to the list of markers for the associated character haplotype.
The file character_haplotypes contains a listing of the SNPs or PCR-RFLPs contained in each character haplotype, and the file genome_groups gives the associated genome groups. Inspection of these files is useful in additional analyses of sequence clusters (e.g. phylogenetic tree construction based on number of SNPs or PCR-RFLPs supporting a given relationship) or group-level assays (find a haplotype that will distinguish any of genomes A, B, or C from all the others, e.g. discriminate virulent strains from vaccine strains).
3) Find sets of haplotypes that maximally discriminate each genome or unresolved cluster of genomes
The third part of the software computes all sets of 1 or more character haplotype(s) that will maximally discriminate each genome. Each genome requires the testing of one or more polymorphic sites to pull it out from the other genomes. Thus, each solution set to resolve a given genome contains one to many haplotypes (a multi-locus solution). There may be many alternative solution sets for every genome, each of which provide the same information, and it is up to the user to select the one that works the best on the chosen platform.
To find a set of character haplotypes to uniquely discriminate one genome from the rest requires that the intersection of the genomes across the associated genome groups is that single target genome. For example, if genome group 1 (associated with haplotype 1) contains genomes A, B, and C, and genome group 2 (associated with haplotype 2) contains genomes A and D, then the set of haplotypes 1 and 2 can uniquely discriminate genome A. That is, if in an unknown sample, one found a SNP that was included in haplotype 1 and another SNP that was included in haplotype 2, then it would be concluded that the unknown sample was like A, and not B, C, or D.
The combinatoric demands of this step may be substantial. For example, if there are 100 genome groups, then examining the intersections of all possible combinations of 3 genome groups requires 100C3 = 161,700 tests. We represent each genome group as a bit vector of 1's and 0's indicating the membership or exclusion, respectively, of each of the genomes in that genome group. Then all genome groups containing only a single genome (each of the markers in the associated character haplotype uniquely identifies that genome) are reported, as well as all genome group combinations of size = number_combinations_to_test and fewer in which the bitwise AND (intersection) is a single genome. The bit vector approach was the fastest method that was tested by the authors. Thus, all intersections of sets of genome groups are examined, and for each of the input genomes, a search is made for those in which the intersection is uniquely that genome. The maximum number of genome groups per combination that is tested in a multi-locus solution set is the parameter number_combinations_to_test. Thus, combinatorics are performed on number haplotypessCnumber_combinations_to_test total combinations.
If there is not an intersection that contains uniquely the target genome, the intersections that contain the fewest other genomes in addition to the target genome is reported as the solution set for that target. Thus, if the target in question cannot be uniquely discriminated, the most specific level to which it can be discerned is output. The other genomes in solution set i for genome A is represented as the list others(i,A).
When there are many genomes and many genome groups, it may be necessary to take two additional steps to find the combinations to maximally discriminate sequences. For example, there may be hundreds of genome groups, so that the number_combinations_to_test must be set at the low value of 2 in order for the combinatoric step to finish in a reasonable amount of time (hours or less). First, if there are two preliminary solutions i ≠ j where others(i,A) ≠ others(j,A), new combinatorics are performed on only the union of the genome groups that comprise all preliminary solutions for the genome A under consideration. Since the number of genome groups making up all preliminary solutions for the specific genome A is a small subset of the total number of genome groups, testing all possible combinations (not just combinations up to a given size) computes rapidly. This generates new solution sets, each of which contains more genome groups than the preliminary solution sets. Each of these new solution sets for the given genome has an identical list of others(A), so that the index i may be dropped.
A second step is sometimes required to find maximally discriminating multi-locus combinations in situations with large numbers of genomes and genome groups. For genome B that is in the list of others(A), if the size of others(B) < size of others(A), then one should be able to discriminate genome A to a higher level (with fewer others(A)), if more genome groups are included in the solution set. In this case, the program cycles through all genome groups, adding the first genome group encountered to the solution sets for A that contain genome A but not genome B. Although taking the first acceptable genome group may not be the best (that is, may not cause the greatest reduction possible in the list others(A)), this method is fast and does give a good solution in the many test cases examined. The list of others(A) is then recomputed with the new genome group included in the solution set, and this process is repeated for all genomes in the list others(A). We repeat this procedure for all the genomes that cannot be uniquely discriminated. This generates final solution set(s) for each genome that may contain many more genome groups than the original combinatoric size number_combinations_to_test.
The first step described above is performed automatically based on results of comparing lists of others(i,A) with others(j,A) for all i ≠ j. The second step is computed automatically based on results of comparing others(A) with others(B) for all genomes A and B. These extra steps are rarely needed, since in most cases, there are few enough genome groups to select a sufficiently large value of number_combinations_to_test to find optimal solution sets with the fewest possible genome groups for each genome, rather than to find acceptable (but possibly sub-optimal) solution sets containing more than number_combinations_to_test genome groups. These steps were required for analyses of SARS virus. All solution sets for each genome are given in the file all_discriminating_sets.
There is a third option for speeding up the combinatorics of the process of finding sets of genome groups that maximally discriminate each genome: this is to exclude consideration of those genome groups with many genomes from the combinatoric calculations. This step yields a set of "pared groups". Since genome groups with many genomes provide the poorest discrimination among those genomes, it is reasonable to only consider the genome groups with the fewest genomes. This is an optional parameter pare_groups that can be set to 1. If pare_groups = 1 then the algorithm finds the cutoff number of genomes per genome group below which there are genome groups that together contain all the input genomes. Then only the haplotypes that contain fewer than this cutoff number of genomes are considered in the combinatoric steps described above. This option may result in orders of magnitude improvement in algorithm speed and memory for situations in which there are a large number of genome groups.
When genomes cannot be uniquely distinguished at the single genome level, the maximally discriminating genome groups are referred to as unresolved clusters. For complete isolate-level discrimination of all the input genomes, the number of unresolved clusters must equal the number of genomes.
4) Find combinations of solution sets containing the fewest total haplotypes
Finally, the combinations of solution sets for each genome are found that enable the testing of the fewest total character haplotypes (i.e. that are associated with the fewest genome groups). For the example shown in Figure 1, either set 1 or set 2 can differentiate genome C. If set 1 is used, then membership in four haplotypes (Ht 2, 3, 4, and 5) must be queried to discern both C and D, but if set 2 is chosen instead, then only 3 haplotypes (Ht 3, 4, and 5) need to be examined. A similar analysis would take place to determine the minimum number of character haplotypes to maximally discriminate all the genomes A-F shown in Figure 1.
Figure 1 Combinations, or sets, of haplotypes can uniquely distinguish an individual genome if the intersection of the genomes across that set is uniquely the one genome in question.
When many of the genomes can be discriminated by multiple alternative solution sets, the number of possible set combinations across all the genomes is the product of the number of solution sets for each genome. Unpublished analyses show that the number of possible set combinations can skyrocket to over 1030 combinations, far too many to do an exhaustive search for the global optimum solution that minimizes the number of character haplotypes to be examined. Therefore, simulated annealing was used to search for approximately optimal solutions using the Metropolis algorithm. Forty parallel processes of simulated annealing were run, and from these the best solutions were selected. The best score is the minimum number of haplotypes that must be examined. The file sim_anneal_results_summary lists the combinations of set numbers (each set is a set of haplotypes) for each genome that enable the fewest haplotypes to be tested overall. Each row is a different solution from simulated annealing that has the same score, that is, each row contains a list of set numbers, one set for each genome, characterizing the combination of sets. The set numbers correspond to the set numbers given in the file all_discriminating_sets, and are not the same as the genome group or character haplotype numbers. Many of these simulated annealing solutions are very similar to one another. For test cases examined in which there were fewer than 107 possible set combinations, we verified that simulated annealing predicted the true global optimum.
Parameters for analyses described here
For the analyses described below, the following parameter values were used for SNP analyses: min_len_upstream = 7 and min_len_downstream = 7. The minimum length of conserved upstream and downstream bases was selected to be 7 because it enabled a finer level of genome discrimination than did 12-mers (which would correspond with Affymetrix chip 25-mers with the central base being either a perfect match or a mismatch).
For the PCR-RFLP analyses described here, the max_amplicon_length = 1000. Amplicons of length 900–1000 bases are specified as preferred, although shorter amplicons are allowed if longer ones cannot be found. Other parameters are: jump = 500 when num_restriction_enzymes = 1, jump = 200 when num_restriction_enzymes = 0, and precision = 5. Primer3 parameters in the file p3.params.pcr.primers as well as a file listing the restriction enzymes used in the computations are available for download at .
SNP and PCR-RFLP analyses were performed using SPR Opt on 102 genomes of severe acute respiratory syndrome (SARS) virus (~30 Kb) and 17 genomes of mumps virus (~15 Kb) that were publicly available at the time the analyses were done. The Genbank genome identification information for these is given in Additional file 1. The multiple sequence alignments were generated using Multiple Genome Aligner [13].
Phylogenetic analyses
Three phylogenetic trees for SARS virus were constructed based on 1) a full genome multiple sequence alignment, 2) SNPs, and 3) PCR-RFLPs. Newick trees were created using the PHYLIP software package [14,15], using the neighbor program to generate the trees based on a distance matrix of pairwise distances between the genomes. For the multiple sequence alignment, the dnadist program using maximum likelihood was used to generate the distance matrix. For SNPs and PCR-RFLPs, distance matrices were created by summarizing the data contained in the genome groups and character haplotypes. This was required because the raw listing of SNPs and PCR-RFLPs necessarily differed in format, while the genome group formatting was consistent between the SNP and the PCR-RFLP data so the same algorithm could be used to generate phylogenetic trees from each method. For the SNP data, the algorithm described below was validated by comparing the resulting phylogenetic tree with that created from a standard SNP matrix, also described below.
To calculate a distance matrix from the genome group data, first, each genome group was weighted by the number of SNPs or the number of PCR primers (for PCR-RFLP) contained in the associated character haplotype. For PCR-RFLPs, the weight was not increased if two or more alternative enzymes cutting the same amplicon (i.e. from the same set of primers) gave the same pattern of genome segregation, since in the runs described, there were many possible enzymes that provided the same information about genome relationships after cutting a single amplicon, and this might artificially increase the weight. After the weights were calculated, then for all possible pairs of genomes, each genome group was examined, and each time a genome group contained both of the genomes, the weight of that genome group was subtracted from the total distance score between that pair of genomes. After all the pairwise distances were calculated, a constant equal to the minimum (negative) distance was subtracted from each pairwise distance, so that none of the distances would be negative. The distance matrices for each of the SNP and the PCR-RFLP analyses were used in the PHYLIP neighbor program. The resulting phylogenetic trees (phylograms) were drawn using a web interface [16]. A phylogenetic tree was also created in a similar manner to that described above except based on only the SNP genome groups in one randomly chosen optimal set solution listed in the results from simulated annealing (that is, a set of the minimum number of haplotypes to maximally discriminate all the input genomes). Such a tree would represent a purported phylogeny if only the data from a set chosen for maximum forensic discrimination is used to predict phylogenetic relationships.
To verify that this method of creating a distance matrix from the genome group data provides an accurate picture for the SNP data, a distance matrix was created from a traditional SNP matrix. In the SNP matrix, columns correspond to the SNPs, and rows correspond to the isolates. This matrix, containing every SNP in the data set, was used in the dnadist algorithm of PHYLIP using the maximum likelihood option. The tree generated was identical to that created using the algorithm described above calculating distances from the genome groups.
Restriction enzyme frequencies for creating RFLPs
The total number of times that each restriction enzyme created a different distribution of fragment lengths was summed across all amplicons in the PCR-RFLP analyses. This was done simply by counting the frequency of occurrence of each enzyme in the file FLPs_all (Table 6).
Abbreviations
SPR Opt: SNP and PCR-RFLP Optimization
SNP: Single Nucleotide Polymorphism
PCR-RFLP: Polymerase Chain Reaction-Restriction Fragment Length Polymorphism
Authors' contributions
SNG conceived of the project, developed much of the software, and drafted the manuscript. MCW created the code for performing set combinatorics using bit vectors, assisted with the code for distributed simulated annealing, and bundled the code for distribution.
Table 1 Description of the user-specified parameters for SNP and PCR-RFLP analyses
User Specified Parameters Description
min_len_upstream minimum length of conserved sequence upstream (5') of a SNP
min_len_downstream minimum length of conserved sequence downstream (3') of a SNP
max_amplicon_length maximum amplicon length allowed for PCR-RFLP analysis
jump the series of sequences input to primer3 for amplicon generation for PCR-RFLP analyses are chosen by sliding a window along the consensus gestalt. Each new window must start at least jump bases from the start of the previous window.
precision there must be at least one difference in fragment lengths among all input genomes that is at least this long for the given amplicon+enzyme combination to be considered variable enough for further PCR-RFLP examination.
num_restriction_enzymes number of restriction enzymes used in an PCR-RFLP analysis before a single electrophoretic determination of fragment length distributions is made. This number may range from 0–3.
number_combinations_to_test maximum number of haplotypes per combination tested in a multi-locus solution set to maximally discriminate all the input sequences. Thus, combinatorics are performed on number haplotypessCnumber_combinations_to_test total combinations.
Supplementary Material
Additional File 2
Lists of unresolved clusters. Description of additional data files provided at Digest*organism_name NoDigest*organism_name SNP*organism_name The * refers to the files indicated in Table 2. Digest* is for PCR-RFLP with num_restriction_enzymes = 1, and NoDigest* is for PCR-RFLP with num_restriction_enzymes = 0. The NoDigest* results are not given for mumps, since there was not adequate variation using this method for forensic discrimination of these input sequences, as indicated in Table 4. The multiple sequence alignment files used in these analyses for SARS and mumps viruses are also available for download. There are a total of 27 files containing all the microbial forensic results and data described above. All are in text format, and can be found at .
Click here for file
Additional File 1
List of genomes used in the analyses. Description of additional data files provided at Digest*organism_name NoDigest*organism_name SNP*organism_name The * refers to the files indicated in Table 2. Digest* is for PCR-RFLP with num_restriction_enzymes = 1, and NoDigest* is for PCR-RFLP with num_restriction_enzymes = 0. The NoDigest* results are not given for mumps, since there was not adequate variation using this method for forensic discrimination of these input sequences, as indicated in Table 4. The multiple sequence alignment files used in these analyses for SARS and mumps viruses are also available for download. There are a total of 27 files containing all the microbial forensic results and data described above. All are in text format, and can be found at .
Click here for file
Acknowledgements
This work was performed under the auspices of the U.S. Department of Energy by the University of California, Lawrence Livermore National Laboratory under Contract No. W-7405-Eng-48. This work was supported by the Intelligence Technology Innovation Center. We gratefully acknowledge discussion and input from Tom Slezak throughout this work. We also thank Steve Velsko, Tom Slezak, and Sharon Messenger for critically reading previous versions of the manuscript and for their helpful comments.
==== Refs
Cummings CA Relman DA Microbial Forensics-"Cross-examining pathogens" Science 2002 296 1976 1979 12004075 10.1126/science.1073125
Anonymous Forensics - bringing bacteria into the courtroom http://wwwbrightsurfcom/news/feb_03/AAAS_news_022503html 2003
Thompson T Our hospitals have to clean up their act The Scotsman, http://wwwthescotsmancouk/indexcfm?id=592232003 2003
CDC U.S. National Molecular Subtyping Network for Foodborne Disease Surveillance (the Centers for Disease Control and Preventions's PulseNet site) 2003 www.cdc.gov/pulsenet/,
Pyrosequencing http://www.hvdlifesciences.com/products_pyroseq_genomics_print.htm
Budowle B Schutzer SE Einseln A Kelley LC Walsh AC Smith JAL Marrone BL Robertson J Campos J Building microbial forensics as a response to bioterrorism Science 2003 301 1852 1853 14512607 10.1126/science.1090083
Wolf C Rentsch J Hubner P PCR-RFLP analysis of mitochondrial DNA: a reliable method for species identification J Agric Food Chem 1999 47 1350 1355 10563979 10.1021/jf9808426
Meyer R Hofelein C Luthy J Candrian U Polymerase chain reaction-restriction fragment length polymorphism analysis: a simple method for species identification in food J AOAC Int 1995 78 1542 1551 8664595
Kwok PY Methods for genotyping single nucleotide polymorphisms Annu Rev Genomics Hum Genet 2001 2 235 258 11701650 10.1146/annurev.genom.2.1.235
Wang D Gao H Zhang R Ma X Zhou Y Cheng J Single nucleotide polymorphism discrimination assisted by improved base stacking hybridization using oligonucleotide microarrays BioTechniques 2003 35 300 308 12951771
Brenner CH Forensic mathematics http://dna-viewcom/ 2004
Slezak T Kuczmarski T Ott L Torres C Medeiros D Smith J Truitt B Mulakken N Lam M Vitalis E Zemla A Zhou CE Gardner S Comparative genomics tools applied to bioterrorism defence Brief Bioinform 2003 4 133 149 12846395
Hohl M Kurtz S Ohlebusch E Efficient multiple genome alignment Bioinformatics 2002 18 S312 S320 12169561
Felsenstein J PHYLIP - Phylogeny Inference Package (Version 3.2) Cladistics 1989 5 164 166
Felsenstein J PHYLIP (Phylogeny Inference Package) version 3.6 Distributed by the author Department of Genome Sciences, University of Washington, Seattle 2004
http://www.phylodiversity.net/~rick/drawtree/
| 15904493 | PMC1156889 | CC BY | 2021-01-04 16:39:53 | no | BMC Genomics. 2005 May 16; 6:73 | utf-8 | BMC Genomics | 2,005 | 10.1186/1471-2164-6-73 | oa_comm |
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-751590720610.1186/1471-2164-6-75Research ArticleControl of gene expression during T cell activation: alternate regulation of mRNA transcription and mRNA stability Cheadle Chris [email protected] Jinshui [email protected] Yoon S [email protected] Thomas [email protected] Jill [email protected] Lana [email protected] Myriam [email protected] Kevin G [email protected] Cellular Biochemistry Section, Basic Research Laboratory, Center for Cancer Research, National Cancer Institute, NIH, 9000 Rockville Pike, Bethesda, MD 20892 USA2 Laboratory of Cellular and Molecular Biology, National Institute on Aging-Intramural Research Program, NIH, 5600 Nathan Shock Drive, Baltimore MD 21224-6825 USA3 Genomatix Software GmbH, Landsberger Str. 6, D-80339 München, Germany4 Capital Genomix, 9290 Gaither Road, Gaithersburg, MD 20877 USA5 DNA Array Unit, National Institute on Aging-Intramural Research Program, NIH, 5600 Nathan Shock Drive, Baltimore MD 21224-6825 USA2005 20 5 2005 6 75 75 18 2 2005 20 5 2005 Copyright © 2005 Cheadle et al; licensee BioMed Central Ltd.2005Cheadle et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Microarray technology has become highly valuable for identifying complex global changes in gene expression patterns. The effective correlation of observed changes in gene expression with shared transcription regulatory elements remains difficult to demonstrate convincingly. One reason for this difficulty may result from the intricate convergence of both transcriptional and mRNA turnover events which, together, directly influence steady-state mRNA levels.
Results
In order to investigate the relative contribution of gene transcription and changes in mRNA stability regulation to standard analyses of gene expression, we used two distinct microarray methods which individually measure nuclear gene transcription and changes in polyA mRNA gene expression. Gene expression profiles were obtained from both polyA mRNA (whole-cell) and nuclear run-on (newly transcribed) RNA across a time course of one hour following the activation of human Jurkat T cells with PMA plus ionomycin. Comparative analysis revealed that regulation of mRNA stability may account for as much as 50% of all measurements of changes in polyA mRNA in this system, as inferred by the absence of any corresponding regulation of nuclear gene transcription activity for these groups of genes. Genes which displayed dramatic elevations in both mRNA and nuclear run-on RNA were shown to be inhibited by Actinomycin D (ActD) pre-treatment of cells while large numbers of genes regulated only through altered mRNA turnover (both up and down) were ActD-resistant. Consistent patterns across the time course were observed for both transcribed and stability-regulated genes.
Conclusion
We propose that regulation of mRNA stability contributes significantly to the observed changes in gene expression in response to external stimuli, as measured by high throughput systems.
==== Body
Background
Virtually all microarray studies to-date have measured changes in steady-state mRNA levels by harvesting total cellular RNA and using it to generate probes through a variety of strategies including end-labeling of purified mRNA [1], incorporating a label into the first strand cDNA made from mRNA [2], or attaching a T7 RNA polymerase promoter during cDNA synthesis, then labeling of the resulting RNA [3]. More recently, several groups have demonstrated the feasibility of hybridizing metabolically labeled mRNAs directly from nuclear run-on (NRO) reactions to nylon filter microarrays in order to investigate nascent transcripts [1,4-6]. Schuhmacher et al. [5], in particular, used a B cell line carrying a conditional, tetracycline-regulated myc gene, and found that myc induction resulted in only a small overlap in regulated mRNAs at 4 hours post-induction when comparing polyA mRNA and NRO RNA on microarrays. This early work provided evidence that transcriptional activation of genes does not necessarily lead to a corresponding increase of their steady-state mRNA levels.
More recently, our laboratory has examined the relationship between newly transcribed (NRO) RNA and polyA mRNA in a stress model using human non-small lung carcinoma H1299 cells. In response to a variety of stresses (ultraviolet light, heat shock, or prostaglandin), we found that approximately half of the observed changes in mRNA levels of stress-regulated genes were accompanied by a corresponding increase or decrease in gene transcription as measured by NRO. The remaining half of stress-altered changes in gene expression was largely attributable to changes in mRNA turnover, thus suggesting that, on a global level, changes in mRNA turnover profoundly influence gene expression patterns [6].
Several questions, however, remained to be answered from these earlier studies. Since in both the myc induction and stress experiments, as mentioned above, measurements of both newly transcribed and polyA mRNAs were made at a single time point, there existed a reasonable possibility of temporal disjunctions between the timing of mRNA new gene synthesis and the rates of accumulation of mRNA in the cell. The second question remaining unanswered is whether or not highly significant levels of mRNA stability regulation (> 50% of all measured gene expression in the stress example) is common to different biological model systems. In order to begin to address these questions we investigated changes at the levels of transcription and total cellular mRNA abundance simultaneously across a time course of activation using Jurkat T cells.
T-cell activation is one of the most widely studied models of cellular response to exogenous stimulation. The initial events include rapid signaling via protein-protein interactions, phosphorylation/dephosphorylation of target signaling molecules, and release of Ca2+ from intracellular stores. Subsequent activation of signal transduction cascades culminates in the implementation of gene expression patterns characteristic of the immune response. Initial microarray studies using T cells have focused on gene expression changes occurring several hours after activation [7-13], even though earlier work using more traditional methods had defined the commitment period for T-cell activation, including alteration in gene expression patterns, as occurring between 1–2 h after exposure to the activating agent [14]. In order to investigate these earlier gene expression changes we chose to examine a time course of activation spanning the first hour after stimulation. While a recent study by Garcia-Martinez et al. [15] in yeast using a similar approach demonstrated large shifts in mRNA stability following a glucose-to-galactose shift, the work presented here is the first systematic accounting of the changes in both gene transcription and mRNA stability in response to a major cellular activation event over a defined time period in higher eukaryotes.
Results
T-cell commitment is believed to occur early during activation and therefore changes in gene expression during the earliest stages of induction are of particular interest. An experimental model using human Jurkat T cells activated with PMA and ionomycin was used to investigate early changes in gene expression (up to one hour of stimulation), focusing on both changes in mRNA transcription rates as well as polyA mRNA levels. One hour was chosen in order to examine gene regulatory events occurring immediately after activation and to avoid, for the most part, the influence of secondary gene regulatory mechanisms taking place at later time points (e.g., increased synthesis of transcription factors). NRO RNA was prepared from isolated cell nuclei (Methods) and polyA mRNA from intact cells. Figure 1 shows an example of filter images obtained after hybridization of arrays using either NRO RNA or polyA mRNA. Cells stimulated for 30 minutes exhibited moderate changes in gene expression in the mRNA arrays. In contrast, NRO RNA arrays revealed rapid and robust changes in transcription that were evident as early as 5 minutes following induction. Unexpectedly, a careful analysis of all significant changes in gene expression across the time course revealed that these early-response genes (as identified by NRO) were a relatively small subset of all of the genes shown to be regulated (see below).
Figure 1 Hybridization images of polyA mRNA and nuclear run-on (NRO) RNA. Depicted are signals in fields of arrays corresponding to untreated (time 0), as well as 5, and 30 min after induction of Jurkat cells using 40 ng/ml PMA and 1 μM Ionomycin (P+I). Two-color overlays contrast either 5 min or 30 min (red) versus 0 (green) for polyA mRNA and NRO RNAs.
In all, a total of 4608 genes, including sets of genes enriched for immune response and signal transduction function, were polled (Fig. 2). Of these genes, 2386 showed significant regulation (p < 0.001, or Z ratio > ± 1.5) during the time course of one hour post activation by either changing gene transcription or polyA mRNA levels. These significantly regulated genes were chosen for further analysis.
Figure 2 Distributions of significantly regulated genes in both polyA mRNA and nuclear run-on (NRO) RNA. For this analysis, a gene was considered to be up- or down-regulated in either polyA mRNA RNA or NRO RNA (Altered Gene Expression) if it was significantly different from the baseline at any point during the time course of activation; all other genes are in the 'Unaltered Gene Expression' category. The number and per cent of genes in each of 8 possible regulatory categories are displayed.
The distribution and direction (increase, decrease, or no change) of significant changes in gene expression either at the transcriptional level (NRO) or at the level of polyA mRNA (whole-cell) are displayed in the table in Figure 2. The single most common expression event (55.2 %) was an up or down regulation at one or more time points as measured in mRNA without a corresponding (either up or down) regulation as measured by transcription at any time point. The second largest group of regulated genes (27.4%) showed changes in transcription with no corresponding change in polyA mRNA. Examples of genes dramatically up-regulated in this second class included CD69, a type II transmembrane receptor involved in lymphocyte proliferation and a classic marker of early T cell activation; PPP3C, the catalytic subunit of the calmodulin-activated phosphatase, calcineurin, which plays a central role in signal transduction from the T cell receptor to the nucleus; as well as several members of the JUN family of transcription factor immediate early response genes (Fig. 3A).
Figure 3 Comparison between polyA mRNA and nuclear run-on RNA of immediate early gene activation in Jurkat T cells. A.1 Heatmap of relative gene expression intensities (Z scores). A.2 Graphical representation of the same data illustrating an immediately apparent up-regulation of gene expression in NRO but not polyA mRNA. B. Single end-point PCR validation of up-regulation in polyA mRNA by 1 hour of a subset of genes shown to be activated within 5 minutes by nuclear run-on RNA. C. Patterns of polyA mRNA and nuclear run-on RNA levels for NFKB1 and its inhibitor (NFKBIA).
The final, relatively minor groups of regulated genes included genes which were regulated at both the transcriptional and the polyA mRNA levels in either the same (8.4%) or opposite directions (9%). The relatively low concordance between transcriptional production of mRNA and its measured appearance in polyA mRNA levels was somewhat surprising, although clear examples of coordinated step-wise production were noted for some key genes, as for example, the early response genes EGR1 and ETR101 (previously shown to be induced at 30 minutes by phorbol ester treatment of a human promyelocytic leukemia cell line [16] and, the apoptosis-related genes DAP (death-associated protein, mediator of interferon-gamma-induced apoptosis) and CASP3, as well as the immune response signal transducer and activator of transcription (STAT) 6 (Fig. 3A).
Dramatic activation of immune response, immediate-early response genes, and apoptosis-related genes was observed in the nuclear run-on RNA as early as 5 minutes following activation (Fig. 3A). This group of genes included immediate-early response genes commonly up-regulated during cellular activation (ETR101, Myb, Myc, and genes of the JUN and EGR families), genes specifically associated with an early response in immune cells (IL6, IL8, STAT4, STAT6, PPP3C, NFKBIA, IRF5, and CD69), as well as genes involved in regulating apoptosis (BCL2A1, CASP3, CASP9, CASP10, and DAP). Many of these same genes were eventually up-regulated in polyA mRNA later in the time course and their increase in expression after one hour was independently validated by single end-point PCR validation using GS320 technology (Fig. 3B).
A particularly interesting example of the dichotomy between transcription and changes in polyA mRNA levels was seen in the production of the mRNAs encoding NFKB1 (NF-kappa B), a key mediator of the transcriptional control of genes involved in the immune response and acute phase reactions, and its inhibitor, NFKBIA (NF-kappa B inhibitor A). Both NFKB1 and NFKBIA have previously been shown by microarray analysis to be significantly induced in polyA mRNA between 3–4 hrs following phorbol or lectin activation of either Jurkat or human peripheral blood lymphocytes [7,8,11,17]. As demonstrated here (Fig. 3C), the production of NFKB1 mRNA clearly increases between 30 minutes and one hour at the transcriptional level without a detectable corresponding increase in polyA mRNA during that time (subsequent PCR analysis did show some increase at the steady-state level between 0 and 60 minutes for the NFKB1 gene but this increase failed to meet the significance thresholds set for the microarray analysis). NFKBIA, on the other hand, is rapidly induced transcriptionally to a maximum level by 30 minutes, returning essentially to baseline within one hour. Meanwhile, NFKBIA steady-state levels can be seen to gradually rise across the first hour of the time course. Analysis of the dynamics of gene expression for NFKB1 and its inhibitor as deduced from conventional microarrays might suggest that by one hour NFKB1 production had not yet begun (contradicted by the NRO data here), and also that the mRNA for the inhibitor of NKFKB1 is steadily increasing (when, in fact, it is clear from NRO data that virtually all increases in the production of NFKBIA have concluded by one hour). These data provide a clear example of how information from nuclear run-on microarrays can enhance studies of gene activation and feedback mechanisms.
Consistency at each time point for genes regulated by activation-induced changes in mRNA stability can be seen in a graph of the Z ratio differences (Methods) between NRO and polyA mRNA calculated for all genes at all time points (Fig. 4A). These putative stability-regulated genes exhibited a very high degree of consistency at all the time points measured. This replicability is further illustrated in the heat map of clustered gene expression in Figure 4B in which the data for each gene has been independently normalized to its baseline (0 time) level. Large numbers of genes as measured in the polyA mRNA are consistently and strongly regulated following activation. Some of these changes are mirrored by changes in gene transcription (NRO) but most are not.
Figure 4 Global comparison of gene expression changes in polyA mRNA and NRO RNA. A. At each time period indicated columns correspond to the values derived by subtracting the Z ratio of NRO RNA from the Z ratio of polyA mRNA for every gene. Equivalency between calculations varies around 0, data is aligned using a combined average index, and is displayed from left to right to represent the highest average positive Z ratio difference to the lowest average negative Z ratio difference. Columns in red correspond to genes exhibiting significant differences (Z ratio difference values greater or less than ± 1.5) in gene expression changes when comparing polyA mRNA and NRO RNA for each gene at each time point. B. Hierarchical clustering of all significant changes in gene expression in either NRO or polyA mRNA across the time course of activation. The median Z score data for each gene is individually normalized to its baseline value in for the sake of this comparison.
In order to compare changes in gene expression patterns at the transcriptional and polyA mRNA levels in a systematic fashion, a simple barcode of 1, -1, or 0 was applied to all significant changes in gene expression indicating up, down, or no change, respectively. In addition, a value of -1, 0, or 1 (low, moderate, or high) was assigned to each gene according to its relative intensity at baseline (0 time). An analysis of the distribution of these gene expression patterns (Fig. 5) revealed that both at the transcriptional and the steady-state levels two thirds of all genes were restricted to just one of 20 patterns (out of a possible number of 729) and that half of these patterns were shared between the two groups. An interesting distinction between the two groups was that whereas up-regulation from a moderately high level of baseline expression was highly favored for new gene synthesis, down-regulation from a moderately high level of baseline expression was very highly favored during polyA mRNA regulation (Fig. 6). In fact, down-regulation, as a consistent trend, was much less common among transcriptional-regulated than steady-state-regulated genes, with implications (see below) as to the roles these modes of regulation play in concert for the control of gene expression.
Figure 5 Frequency distribution of gene expression patterns generated from polyA mRNA or NRO RNA. The top 20 patterns for each method is shown. Significant changes in gene expression were assigned a 1, -1, or 0 for up, down, or no change, respectively. In addition, a value of -1, 0, or 1 (low, moderate, or high) was assigned to each gene according to its relative intensity at baseline (0 time). The absolute numbers of genes in each pattern are reported in the column labeled Top 20 and the percentage of those genes relative to all significantly regulated genes can be found in the column labeled % Sig. Totals are as indicated at the bottom of each column. Filled-in boxes denote patterns equally shared in the top 20 between both methods
Figure 6 Differential distribution of transcriptional and polyA mRNA up- (A) or down- (B) regulated gene expression patterns during Jurkat T cell activation. The number of genes consistently regulated (up or down at every time point) are correlated with their relative expression baselines (Z score: high>1, 1>medium>-1, low<-1).
In order to confirm that stability regulation was in fact a reasonable explanation for the observation that the expression levels of large numbers of genes were changing at the whole cell but not the transcriptional level, a series of experiments were carried out in which activation of Jurkat cells was carried out in the presence or absence of the transcription inhibitor Actinomycin D. Analysis of polyA mRNA demonstrated the strong inhibition of mRNA levels for genes previously shown to be transcriptionally up-regulated (Fig. 7A). In contrast, large numbers of genes which were significantly regulated in polyA mRNA but not in NRO RNA were not affected by Actinomycin D treatment (Fig. 7B &7C). Among these unaffected genes there was a bias towards presumptively de-stablized genes (Fig. 7C) consistent with the earlier conclusion (Fig. 6B) that down-regulation is the predominant motif in overall polyA mRNA levels.
Figure 7 Persistence of stability-regulated changes in gene expression in the presence of Actinomycin D. A. Effect of Actinomycin D on the (P+I)-induced changes in the expression patterns of gene deemed to be transcriptionally regulated. Z ratio comparisons are made to the baseline or unactivated state. B. Lack of an effect of Actinomycin D on the (P+I)-induced changes in the expression patterns of genes deemed to be stability regulated. The absolute differences (Z diff) between the activated and un-activated state for both polyA mRNA and NRO RNA for a subset of these genes is shown. C. Comparison of the Z ratios for all significantly regulated genes in the presence or absence of Actinomycin D showing no corresponding regulation in NRO RNA. Data is sorted by polyA mRNA (without ActD) values and a significance threshold of Z ratios > ± 1.5 was used for these calculations.
Although an examination of the functional classifications of stability regulated versus transcriptionally regulated genes yielded no obvious trends, some biological pathways appear to be differentially, and sometimes, exhaustively regulated by each type of expression event. One example of this can be seen in the apoptotic pathways, which are comprehensively regulated during the Jurkat activation scenario [16,18,19]. As can be seen from the pathway schematic illustrated in Figure 8, some major effectors of apoptosis such as CASP 3, 4, 8, 9, &10 (up-regulated) and BCL2 (down-regulated), are controlled by new gene synthesis while other factors such as CASP 1, 6, &7 (up-regulated) and BCL2L2 (down-regulated) appear to be regulated by stability processes alone. Regardless of the cellular rationale for regulating at the level of new gene synthesis or mRNA stability, it is clear from these data that there is a strong internal coherence: genes regulated by one mechanism do not crossover to the other within the time frame investigated.
Figure 8 Regulation of apoptotic pathways during T cell activation involves changes in gene expression by both mRNA transcriptional and mRNA stability mechanisms. Genes colored in red and blue were up- or down-regulated in polyA mRNA only. Genes colored in yellow and purple were up- or down-regulated in NRO RNA. Pathway is from a Kegg map modified in Genmapp2.0 [27].
Common patterns of transcription factor binding sites in the upstream promoter regions of groups of genes whose transcription was significantly up-regulated were detected. An enrichment of transcription factor-binding sites for both NFAT and NFκB was found in the promoter regions of genes most significantly up-regulated at 1 hour as shown in Figure 9. The frequency with which both NFAT- and NFκB-binding sites were found in the promoters of this group of genes was significantly greater than that seen for other genes in the array [e.g., genes down-regulated at this time point or genes selected at random [supplemental data]. The discovery that NFAT- and NFκB-binding sites were enriched in the promoters of these genes was not unexpected, since these transcription factors, along with AP-1 components Fos and Jun, constitute the major transcription factors involved in the early stages of T-cell activation. Indeed, the frequency of genes significantly up-regulated in NRO RNA and enriched for AP-1 binding sites peaks during the time course (supplemental data). Many of the genes that are transcriptionally upregulated at 1 hour, such as CD69, are elevated throughout the time examined, although a few genes including PTEN, DUSP5, and NFκB1 itself, show elevated transcription only after 60 minutes. The simultaneous increase of NFκB1 gene transcription (between 30 minutes and 1 hour) combined with a noticeable increase in the transcription of genes containing NFκB-binding sites (at one hour) demonstrates the synchronous relationship between the appearance of this transcription factor and its downstream targets at the indicated time points.
Figure 9 Enrichment of NFAT and NFκB transcription factor-binding sites in the promoter regions of genes upregulated after 60 min of treatment of human Jurkat T cells with PMA + Ionomycin. Color gradient from bright red to bright green is directly proportional to Z ratios and indicates the increase (red) or decrease (green) of gene expression relative to baseline at each of the indicated time points.
Discussion
The absence of significant regulation at the transcriptional level during a time in which gene expression is strongly perturbed at the polyA mRNA level during T cell activation reveals that many genes are being regulated through changes in stability. The lack of detectable transcriptional regulation of large numbers of steady-state mRNA gene changes is particularly striking insofar as the NRO measurements are likely to be even more sensitive to changes in gene expression than polyA mRNA measurements since they are a direct measure of newly synthesized mRNA. This finding is consistent with the work of Raghavan et. al. [12,20] who also noted the presence of numerous transcripts exhibiting stimulus-dependent changes in mRNA decay in human T lymphocytes treated for 3 hours with anti-CD3 and/or anti-CD28 antibodies. These experiments were carried out by arresting transcription with Actinomycin D (in the presence or absence of activation), and mRNA turnover rates were globally measured by applying polyA mRNA to microarrays. Similarly, in this current work, patterns of Actinomycin D resistant changes in gene expression among groups of genes significantly regulated only at the polyA mRNA level supports the observation that large numbers of genes are regulated primarily by stability during T cell activation.
The smaller, but still substantial, group of genes which displayed significant transcriptional regulation without any corresponding changes in polyA mRNA, may either be a reflection of the enhanced sensitivity of transcriptional detection or, perhaps, result from a persistent lag between changes in transcriptional output and their reflection in steady-state mRNA levels. Such a lag might arise from differences in scale of the absolute size of mRNA pools between newly transcribed and polyA mRNA. In this scenario, changes in the amounts of mRNA that are readily detectable by nuclear run-on may be too small to have an immediate detectable impact on the steady-state RNA pools, possibly for long periods of time. Another possibility is that the nascent mRNAs of some of these genes are so rapidly degraded that they never significantly impact on polyA mRNA levels at all.
As previously noted (Fig. 6), up-regulation of gene expression was a dominant trend for transcriptionally regulated genes while down-regulation was dramatically favored for genes regulated at the polyA mRNA level. From the standpoint of an overall cellular economy of gene expression, effective control at the level of transcription can be achieved primarily by turning on new gene transcription, while at the whole-cell level, rapid and effective regulation can be achieved by massive shifts in the stability of existing mRNA pools. In fact, the single most dramatic regulatory event experimentally observed was the rapid clearance of large pools of steady-state mRNA, presumably as result of a sudden and demanding change in cellular conditions.
Questions remain as to the extent that this type of regulation may occur under differing biological scenarios. Global regulation of mRNA stability has thus far been systematically studied under conditions of stress and cellular activation. It remains to be determined whether or not changes in mRNA stability are responsible for altering gene expression programs in response to other biological conditions. The existence of these regulatory paradigms may require a reevaluation of the common model of the control of gene expression which essentially invokes the turning on and off of gene transcription in order to explain changes in polyA mRNA levels.
Conclusion
Although stability regulation might reasonably be considered as a cellular measure to bridge the gap between the very rapid events of signal transduction and the longer term induction of cellular programs involving the coordinated transcription of batteries of new gene synthesis, the current data actually suggests otherwise. One of the goals in this work was to examine changes in gene expression due to altered transcription and those due to altered turnover across a time course allowing sufficient time for the resolution of time disparities between new mRNA synthesis and their appearance as measurable cellular pools. Sufficient time for effective transcription and even translation was clearly demonstrated by the example of NFKB1 mRNA induction followed by the transcription of its downstream targets. Patterns of stability regulated genes were, in this model system, non-random and surprisingly persistent throughout the entire time course suggesting that stability regulation was a major component in the control of gene expression for significant periods of time. The presence of highly active and global regulation of polyA mRNA levels by stability-altering mechanisms suggests a new significant level of regulation in the control of gene expression which spans wide phylogenetic distances from yeast [15] to humans, emphasizes a possible parsimonious role for new gene synthesis in response to changing cellular environments, and may help to explain some of the difficulties encountered in attempts to comprehensively correlate clustered changes in polyA mRNA with common promoter regulatory elements.
Methods
Cell culture and treatment
The human Jurkat E6-1 T cell line was cultured in RPMI-1640 medium (GIBCO-BRL, Gaithersburg, MD) supplemented with 10% FCS (Hyclone, Logan, UT) and antibiotics, and incubated at 37°C, 5% CO2. Cells were stimulated with 40 ng/ml phorbol 12-myristate 13-acetate (PMA) and 1 μM ionomycin (Sigma, St. Louis, MO) (P+I). Actinomycin D (A.G. Scientific, San Diego, CA) was added at 10 ug/ml 30 minutes prior to activation with PMA + I.
Preparation and purification of RNA
Isolation of nuclei and preparation of nuclear run-on (NRO) RNA were as described [6,21] with some modifications. Briefly, ~108 cells per sample were pelleted, washed with PBS, resuspended in ice-cold lysis buffer (20 mM Tris-HCl [pH 7.4], 20 mM NaCl, 5 mM MgCl2, 0.25% [v/v] NP-40), and incubated on ice for 5 min. Nuclei were spun down, resuspended in storage buffer (50 mM Tris-HCl [pH 8.3], 5 mM MgCl2, 0.1 mM EDTA-NaOH [pH 8.0], 40% [v/v] glycerol) and stored at -80°C until use. Thawed nuclei (200-μl aliquots) were mixed with 200 μl of reaction buffer (5 mM Tris-HCl [pH 8.0], 0.15 mM KCl, 2.5 mM MgCl2, 2.5 mM DTT, 0.5 mM of each ATP, UTP, GTP) plus 500 μCi of [α-33P]UTP (3,000 Ci/mmol, 10 mCi/ml; ICN), and incubated for 30 min at 30°C with shaking. Samples were then incubated with DNase I (100U, RNase-free; Roche Diagnostics) for 20 min at 37°C, and with proteinase K (1 μg/μl) for 1 h at 37°C. Finally, nascent RNA was purified by Sephadex G-50 column filtration. For the preparation of polyA mRNA cellular RNA, resting or stimulated Jurkat cells were lysed in STAT60™ (TEL-TEST, INC., Friendswood, TX) as described [6]. RNA concentration and quality were assessed spectrophotometrically and by agarose gel electrophoresis. NRO RNA was stored at -80°C until use.
PolyA mRNA samples were radiolabeled and hybridized as previously described [22]. In brief, 5 μg of total mRNA for each sample was annealed, in 16 μl H2O, with 1 μg of 24-mer poly(dT) primer (Research Genetics, Alabama), by heating at 65°C for 10 min and cooling on ice for 2 min. The RT reaction was performed by adding 8 μl of 5X first strand RT buffer (Life Technologies, Rockville, MD), 4 μl of 20 mM dNTPs minus dCTP) (Pharmacia, Piscataway, NJ), 4 μl of 0.1 M DTT, 40 U of RNAseOUT (Life Technologies), 6 μl of 3000 Ci/mmol α-33P dCTP (ICN Biomedicals, Costa Mesa, CA) to the RNA/primer mixture to a final volume of 40 μl. Two μl (400 U) of Superscript II reverse transcriptase (Life Technologies) was then added, and the sample was incubated for 60 min at 42°C. The reaction was stopped by the addition of five μl of 0.5 M EDTA. The samples were incubated at 65°C for 30 min after addition of 10 μl of 0.1 M NaOH in order to hydrolyze and remove RNA. The samples were pH neutralized by the addition of 25 μl of 0.5 M Tris, pH 8.0, and purified using Bio-Rad 6 purification columns (Hercules, CA). An aliquout of each labled sample was quantitated by liquid scintillation counting using a Beckman LS 6500. The entire remaining sample was stored at -20°C until use.
Array construction and hybridization
Microarray construction and hybridization were previously described [23]. Briefly, NIA Human Focused Arrays consisting of a set of 4600 spotted cDNAs, arrayed in duplicate, representing a set of 2742 non-redundant genes (enriched in genes involved in immune function and signal transduction), were printed on Nytran + Supercharge nylon membranes (Schleicher & Schuell, Keene, NH), and hybridized with [α-33P]dCTP-labeled cDNA or [α-33P]UTP RNA probes overnight at 50°C as previously described [22], protocols available at . Hybridized arrays were rinsed in 2 X SSC and 0.1% SDS twice at 55°C followed by washes in 2 X SSC and 0.1% SDS at 55°C. Microarrays were exposed for 1–3 d and scanned using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) at a 50-μm pixel resolution. ArrayPro software (MediaCybernetics, Silver Spring, MD) was used to convert the hybridization signals into raw intensity values; data generated were transferred into Microsoft Excel for further analysis.
PCR analysis
Single endpoint PCR was carried out on polyA mRNA for gene detection and relative quantitative comparison between baseline (0 time) and I hour activation (with PMA +I). In brief, GS320 libraries for both control and activated polyA mRNA were prepared as described [24], normalized using a panel of ribosomal protein genes, and specific genes were amplified in duplicate using predefined GS320 primers.
Array data analysis
RNA samples (typically n = 3 to n = 5) were prepared from multiple experiments, each of which consisted of a consecutive series of time points. Equal amounts of total RNA (5 μg) or NRO RNA (108 cell-equivalents) were used in each hybridization. Raw intensity data for each experiment was transformed to log10, then used for the calculation of Z scores as described [22] [see Additional file 1]. Significant changes in gene expression were calculated in the form of Z ratios and/or Z test values ([25]), using Z score values in all calculations. Z ratios constitute a measure of the change in gene expression of a given gene from its baseline value (in this case – time 0), expressed in units of standard deviation from the average change of all genes for that comparison. Z ratios are a direct measure of the likelihood that an observed change is an outlier in an otherwise normal distribution and, as such, are independent from underlying intensity values. Since the contents of the population of nuclear run-on and polyA mRNA are different in both complexity and number, care was taken not to compare Z score normalized intensities directly. Comparisons between Z ratios, however, test for equivalence of significant changes between the transcriptional and steady-state changes in gene expression each relative to its own population. All gene expression changes were assessed through comparison with untreated cells (time 0). A Z ratio value of ± 1.50 and/or a Z test value p < 0.0001 were the significance thresholds used in this study.
Hierarchical clustering was performed using the Cluster and TreeView software programs, developed at Stanford University [26]. The clustering algorithm was set to complete linkage clustering using the uncentered Pearson correlation.
Mapping of transcription factor binding sites in selected genes was performed using software from Genomatix Software Gmbh, Munich, Germany . See Additional file 2 for a complete description of the analysis.
Abbreviations
NRO, nuclear run-on; P+I, PMA plus Ionomycin; ActD, Actinomycin D
Authors' contributions
CC and JF performed the microarray assays (JF-NRO RNA, CC-polyA mRNA). CC carried out statistical analysis, and drafted the manuscript (with MG & KGB). JR and LD performed the PCR validation assays. TW assisted with promoter analysis and YSC-C participated in the design of the study. All authors read and approved the final manuscript.
Supplementary Material
Additional File 1
Cheadle et al – BMC Genomics – basic data 05-27-05.xls, contains all original data (normalized), significance testing, and barcode assignments.
Click here for file
Additional File 2
Cheadle et al – BMC Genomics TF binding sites frequency analysis 05-27-05.xls, contains all calculations for TF binding site analysis.
Click here for file
Acknowledgements
We are grateful to R.L. Wange (NIA, NIH) for kindly providing Jurkat cells.
==== Refs
Legen J Kemp S Krause K Profanter B Herrmann RG Maier RM Comparative analysis of plastid transcription profiles of entire plastid chromosomes from tobacco attributed to wild-type and PEP- deficient transcription machineries Plant J 2002 31 171 188 12121447 10.1046/j.1365-313X.2002.01349.x
Schena M Shalon D Davis RW Brown PO Quantitative monitoring of gene expression patterns with a complementary DNA microarray Science 1995 270 467 470 7569999
Eberwine J Amplification of mRNA populations using aRNA generated from immobilized oligo(dT)-T7 primed cDNA Biotechniques 1996 20 584 591 8800675
Meininghaus M Chapman RD Horndasch M Eick D Conditional expression of RNA polymerase II in mammalian cells. Deletion of the carboxyl-terminal domain of the large subunit affects early steps in transcription J Biol Chem 2000 275 24375 24382 10825165 10.1074/jbc.M001883200
Schuhmacher M Kohlhuber F Holzel M Kaiser C Burtscher H Jarsch M Bornkamm GW Laux G Polack A Weidle UH Eick D The transcriptional program of a human B cell line in response to Myc Nucleic Acids Res 2001 29 397 406 11139609 10.1093/nar/29.2.397
Fan J Yang X Wang W Wood WH Becker KG Gorospe M Global analysis of stress-regulated mRNA turnover by using cDNA arrays Proc Natl Acad Sci U S A 2002 99 10611 10616 12149460 10.1073/pnas.162212399
Ellisen LW Palmer RE Maki RG Truong VB Tamayo P Oliner JD Haber DA Cascades of transcriptional induction during human lymphocyte activation Eur J Cell Biol 2001 80 321 328 11432721
Cristillo AD Bierer BE Identification of novel targets of immunosuppressive agents by cDNA- based microarray analysis J Biol Chem 2002 277 4465 4476 11694517 10.1074/jbc.M108598200
Marrack P Mitchell T Hildeman D Kedl R Teague TK Bender J Rees W Schaefer BC Kappler J Genomic-scale analysis of gene expression in resting and activated T cells Curr Opin Immunol 2000 12 206 209 10712937 10.1016/S0952-7915(99)00075-8
Rogge L Bianchi E Biffi M Bono E Chang SY Alexander H Santini C Ferrari G Sinigaglia L Seiler M Neeb M Mous J Sinigaglia F Certa U Transcript imaging of the development of human T helper cells using oligonucleotide arrays Nat Genet 2000 25 96 101 10802665 10.1038/75671
Feske S Giltnane J Dolmetsch R Staudt LM Rao A Gene regulation mediated by calcium signals in T lymphocytes Nat Immunol 2001 2 316 324 11276202 10.1038/86318
Raghavan A Ogilvie RL Reilly C Abelson ML Raghavan S Vasdewani J Krathwohl M Bohjanen PR Genome-wide analysis of mRNA decay in resting and activated primary human T lymphocytes Nucleic Acids Res 2002 30 5529 5538 12490721 10.1093/nar/gkf682
Grolleau A Bowman J Pradet-Balade B Puravs E Hanash S Garcia-Sanz JA Beretta L Global and specific translational control by rapamycin in T cells uncovered by microarrays and proteomics J Biol Chem 2002 277 22175 22184 11943782 10.1074/jbc.M202014200
Crabtree GR Contingent genetic regulatory events in T lymphocyte activation Science 1989 243 355 361 2783497
Garcia-Martinez J Aranda A Perez-Ortin JE Genomic run-on evaluates transcription rates for all yeast genes and identifies gene regulatory mechanisms Mol Cell 2004 15 303 313 15260981 10.1016/j.molcel.2004.06.004
Shimizu N Ohta M Fujiwara C Sagara J Mochizuki N Oda T Utiyama H Expression of a novel immediate early gene during 12-O-tetradecanoylphorbol-13-acetate-induced macrophagic differentiation of HL-60 cells J Biol Chem 1991 266 12157 12161 2061303
Schena M Shalon D Heller R Chai A Brown PO Davis RW Parallel human genome analysis: microarray-based expression monitoring of 1000 genes Proc Natl Acad Sci U S A 1996 93 10614 10619 8855227 10.1073/pnas.93.20.10614
Kizaki H Tadakuma T Odaka C Muramatsu J Ishimura Y Activation of a suicide process of thymocytes through DNA fragmentation by calcium ionophores and phorbol esters J Immunol 1989 143 1790 1794 2506269
Ruiz-Ruiz MC Oliver FJ Izquierdo M Lopez-Rivas A Activation-induced apoptosis in Jurkat cells through a myc-independent mechanism Mol Immunol 1995 32 947 955 7477000 10.1016/0161-5890(95)00059-N
Raghavan A Bohjanen PR Microarray-based analyses of mRNA decay in the regulation of mammalian gene expression Brief Funct Genomic Proteomic 2004 3 112 124 15355594
Gorospe M Wang X Holbrook NJ p53-dependent elevation of p21Waf1 expression by UV light is mediated through mRNA stabilization and involves a vanadate-sensitive regulatory system Mol Cell Biol 1998 18 1400 1407 9488455
Cheadle C Vawter M Freed WJ Becker KG Analysis of microarray data using Z score transformation Journal of Molecular Diagnostics 2003 5 73 81 12707371
Barrett T Cheadle C Wood WB Teichberg D Donovan DM Freed WJ Becker KG Vawter MP Assembly and use of a broadly applicable neural cDNA microarray Restor Neurol Neurosci 2001 18 127 135 11847435
Wang A Pierce A Judson-Kremer K Gaddis S Aldaz CM Johnson DG MacLeod MC Rapid analysis of gene expression (RAGE) facilitates universal expression profiling Nucleic Acids Res 1999 27 4609 4618 10556317 10.1093/nar/27.23.4609
Nadon R Woody E Shi P Rghei N Hubschle H Susko E Ramm P Geschwind D and Gregg J Statistical inference in array genomics. Microarrays for the Neurosciences 2002 Cambridge, MIT Press
Eisen MB Spellman PT Brown PO Botstein D Cluster analysis and display of genome-wide expression patterns Proc Natl Acad Sci U S A 1998 95 14863 14868 9843981 10.1073/pnas.95.25.14863
Dahlquist KD Salomonis N Vranizan K Lawlor SC Conklin BR GenMAPP, a new tool for viewing and analyzing microarray data on biological pathways Nat Genet 2002 31 19 20 11984561 10.1038/ng0502-19
| 15907206 | PMC1156890 | CC BY | 2021-01-04 16:39:54 | no | BMC Genomics. 2005 May 20; 6:75 | utf-8 | BMC Genomics | 2,005 | 10.1186/1471-2164-6-75 | oa_comm |
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-411592153010.1186/1472-6963-5-41Research ArticleOrganisation and financing of the health care systems of Bulgaria and Greece – what are the parallels? Exadaktylos Nikolaos M [email protected] Higher Technological Educational Institute of Thessaloniki (A.T.E.I.T.), Vasilis Olgas 6, 54640, Thessaloniki, Greece2005 28 5 2005 5 41 41 12 5 2004 28 5 2005 Copyright © 2005 Exadaktylos; licensee BioMed Central Ltd.2005Exadaktylos; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Bulgarian and Greek Medical Care systems have been reformated the last fifteen years. The aim of this study was an examination and comparison of the Bulgarian and Greek Medical Care Systems.
Methods
This study was prepared by using all the published data related to both Bulgarian and Greek Medical Care systems. Besides, personal communications with related offices such as administration offices of hospitals and Ministries of Health were made.
Results
In both countries, besides the compulsory insurance there is also additional voluntary insurance which is provided by private companies. The most important difference is the family doctor (specialist in general medicine) existing in Bulgaria. Every insured person needs a 'referral form' completed by the family doctor before visiting a hospital for medical attention (except emergencies). In contrast, in Greece an insured person can directly visit any hospital without needing any forms and independent of the severity of their health problem. An important disadvantage of the Greek health system is the low number of hospitals (139), in relation to population. In contrast, there are 211 hospitals in Bulgaria, although its population is lower than in Greece.
Conclusion
In both Greek and Bulgarian health systems changes must be done to solve the problems related to informal payments, limited financing, large debts, lack of appropriate investment policy, lack of an objective method for the costing of medical activities and inefficient management.
==== Body
Background
Bulgaria provided universal health care free at the point of use prior to its transition from communism in 1989. Throughout the 1990s, the Soviet-style model in central and eastern Europe that provided free health services has been subject to radical reforms [1]. The medical care system in Bulgaria, which consisted of two levels (pre-hospital and hospital), is undergoing "medical reform" which officially began in 1999 [2] (based on three laws adopted by the National Assembly of the Republic of Bulgaria: the Health Insurance Law (1998), the Law on the Professional Organizations of Physicians and Dentists (1998) and the Law on Health Care Establishments (1999), because the state funded health care system started to deteriorate with severe shortages and out of pocket expenditure reaching unaffordable levels. Reform was clearly needed, but there was no consensus on its direction, in particular, whether financing should be through taxation or social insurance.
Following the enactment of health insurance legislation in 1998, social insurance contributions (split between employer and employee) began to be deducted by employers in 1999. The amount of revenue collected initially was limited by the low tax base (given low incomes and high unemployment) and tax evasion. In 2000 the National Health Insurance Fund (NHIF) covered 13% of all public health care expenditures. It is expected that the state and municipal budgets' share of total public financing will gradually decrease over the years as the NHIF assumes an increasingly important financing role. These basic rights were developed and specified in Bulgarian Law about medical insurance were published in the Government Gazette [3].
Not much data have been published about the Greek Medical Care System. However, improvement to the Medical Care System constitutes one of the main aims of Greek Governments, as is obvious from the many reforms and modifications conducted the last twenty years [4-7].
In both Greece and Bulgaria, medical insurance can be separated into compulsory and voluntary (there are a number of non-governmental organizations in the health sector. These include organizations in Bulgaria that existed during the communist period, such as those for the blind, the deaf and the disabled. In addition, a number of newer organizations have developed, representing people with multiple sclerosis, diabetes and cancer). Voluntary medical insurance is additional to compulsory and is implemented by limited private companies holding a licence (based on the Law). This regulates the establishment of commercial and operation companies within the voluntary insurance sector. Voluntary health insurance has been limited in both Bulgaria [1] and Greece, so far taken out only by high-income groups. Voluntary health insurance can provide extra insurance (to be 'bought') on a voluntary basis by any individual. Beyond the basic package, citizens are free to buy different insurance packages on the market. Private insurance may also cover those services included in the basic package and negotiated by the National Framework Contract. Voluntary health insurance funds are also legally entitled to own hospitals and pharmacies.
Methods
This study is the first afford to examine and compare the Bulgarian and Greek Medical Care systems. The main problem to conducting this study was the lack of much published data, especially for the Greek Medical Care System.
Therefore, all information published in journals, magazines, newspapers, Government Gazettes, books and teaching notes were used. In addition, personal communications with the administration offices of the hospitals in both countries were made. Finally, data were obtained from the Bulgarian and Greek Ministry of Health.
Results
Structure of the health care systems of Bulgaria and Greece
The Bulgarian health care system, in contrast to Greek system, was highly centralized and some decentralization has taken place since 1991. First, ownership of most health care facilities was devolved to locally elected municipalities from 1992. Following a 1997 amendment to the Law on Health, health facilities can become independent juridical entities. Second, the Ministry of Health decentralized much administration to the 28 regional health centres in 1995, allowing a flatter management structure. Third, there has been extensive privatization of pharmacies and physicians' practices.
As reported above, there are two main systems in both countries: private and public health system. These two systems have a tendency for approaching each other [8]. At present, private practice involves mainly dental offices and physicians' surgeries and consulting rooms, pharmacies, laboratories, and outpatient clinics and polyclinics. In addition there are inpatient health care establishments. Boonekamp [9] reported that the changes aim to adapt the new health systems to the requirements of the market. The health systems in both Greece and Bulgaria, after their re-organization, have a tendency for moving from the public to the private system.
Medical care in Bulgaria is based on the recent law and dependant on the target and the volume of the implemented medical activity, is divided into two basic groups:
a. Medical centers of hospital treatment.
b. The medical centers of non-hospital medical treatment are distinguished as follows:
1. Clinics of primary treatment:
- Individual clinics of primary medical treatment – served by a doctor specialized in general medicine.
- Corporate clinics of primary medical treatment – organized by a commercial company or a corporation served by doctors specialized in general medicine.
2. Clinics of special medical treatment:
- Individual clinic of special non-hospital medicine – served by a doctor with a recognized specialty (not general medicine).
- Individual clinic of special medical treatment – organized by a commercial company or corporation served by doctors with the same speciality (except general medicine).
- Medical center, dental center and medical-dental center – in which, at least, three doctors / dentists with different recognized specialties provide special non-hospital aid. These centers are headed by a doctor or a dentist with recognized specialty.
- Diagnostic-consultative center (multi-clinic) – in which, at least, ten doctors with different specialties work. This center has the necessary medical equipment (there is, at least, one medical-diagnostic laboratory and apparatus for imaging diagnostics). It is probably headed by a doctor with the required specialty in each case and a specialist in medical management having postgraduate studies in economics or medical information science or economics of medical care etc.
- Autonomous medical-diagnostic laboratory – in which a doctor is jointed by a specialist on medical consultations after a referral from a doctor or dentist. Depending on the specialization of the laboratory, at least, one doctor works there.
- Autonomous medical-technical laboratory – in which well – trained specialists perform special technical activities and produce special medical and auxiliary aids after a referral from a doctor or dentist. It is headed by a doctor or dentist or a specialist, depending on the specialization of the laboratory.
As reported above, political and socioeconomic impacts have caused reforms of the Greek medical care system the last twenty five years. The first reform happened in 1983 with the development of National System Health (NSH) following from five amends [4-7,10]. All the amends were based on the same rationalities and philosophy and aimed to improve and update Greek medical care [11].
Medical care is distinguished into: a) open or non – hospital care, comprising of medical services concerning the diagnosis and treatment out of hospital and b) the closed or hospital care, concerning medical and hospital services provided in the hospital for diagnosis and treatment [12].
Today, there are three levels of medical care [13]:
a. First-degree level of care – refers to reception centers, where the patient has first contact with the health system (doctor, midwife, nurse etc.). The organization of the first-degree care services are clinics, health centers, out-patients of hospitals, multi-clinics of an insurance organization. According to the manifesto of Alma Ata, as published by Dimoliatis et al., [14], the first-degree of care was based on the scientifically proved data which are easily applied with low cost. It mainly provides services of prevention, curative and recovery of patient health.
b. Second-degree level of care – refers to the medical care provided by the hospital of a region (local or prefecture). This hospital can handle basic health problems, which require hospital care. The provided services are the hospital treatment, laboratory checks for covering of hospital treatment required and general operations [13].
c. Third-degree level of care – refers to the provision of complicated or specialized health problems. The provided services are high hospitalization and curative which require well-experienced staff and well-equipped laboratories. This medical care is usually found close to population centers.
Both Greek and Bulgarian medical systems have similar hospital treatment. The basic difference is the family doctor existing in Bulgarian non-hospital treatment. In Greece, the capitulary of the family doctor system was lawed in 1997 (N.2519/97) [5], but it has not be applied yet.
In Bulgaria, the first contact of the patient with the health system is a visit to family doctor, in contrast to Greece, in which the first contact of patient with the health system is the 'out patients' department of hospital or with private doctors having a contract with the public insurance or with the polyclinics of Greek Foundation of Federal Health Insurance (Greek abbreviation, I.K.A.) or with the agricultural clinics or with the health centers. Because most of the patients visit the out patients' department of hospitals, a problem arises from the crowd of patients resulting in abnormalities in the function of hospitals. As a result of this problem it is the policy of Greek Governments to concentrate on the hospital treatments at the expense of the non-hospital treatments.
Medical care of insured people in Bulgaria and Greece
The National Health Insurance Fund (NHIF) is an autonomous institution for compulsory health insurance that was established in accordance with Bulgarian legislation. The Health Insurance Law adopted by the Bulgarian parliament in 1998, introduced a Bismarckian type of health insurance system, with only one health insurance agency and mandatory health insurance payments deducted from personal income. Parliament decides the size of health insurance payments and each year determines the budget of the National Health Insurance Fund [15]. The medical centers of non-hospital and hospital care have individual contracts with the National Health Insurance Fund (represented by the directors of the Regional Funds) and become the providers of medical care in the system of the compulsory medical insurance.
The provided hospital care of compulsory insurance is done in the form of agreement "packages" or "clinical paths" paid by the national fund without any economic participation of the insured individual. Today, there are in total 40 clinical paths. The "clinical path" is a system of demands and behavior instructions of the different medical specialties for the hospital treatment of patients with certain diseases, included within the legal framework, which are paid by the National Fund. The "clinical path", in its own way, acts as a tool to ensure the quality of the medical activity, for the reason that the payment in hospitals is directly related to the quality of the provided medical assistance according to the relative clinical path.
Every compulsorily medically insured person can choose by his / her own personal doctor specialist in general medicine (family doctor) from a medical center (having a contract with the National Health Insurance Fund), which provides a primary non-hospital aid. The family doctor is responsible for every health problem of his patient. Depending on the severity of the health problem, the family doctor decides the treatment or refers the patient with a relative document "medical instruction consultation or performance of common cure" to the executor of the non-hospital medical aid if support from a more specialized doctor is required. The insured person can choose for himself the medical center of special non-hospital assistance with the only limitation that the chosen medical center must have a contract with the regional fund and be within the region. If there is not a medical center of required non-hospital assistance to carry out the appropriate diagnostic and therapeutic procedures within this region, the family doctor can forward the insured patient to a medical center of another region. Finally, in cases where no non-hospital foundation can provide the required treatment, the doctor-executors of this foundation prepare for the admittance of the insured to the hospital by filling a form called "Instruction for Admittance to the Hospital".
There are big economical problems in the Greek hospitals due to the low day payments from the public insurance foundations which have been in the same levels since 1993. This problem is very complicated because, if the day payments from the public insurance foundations increased, then other economical problems would be raised in the public insurance foundation. Balabanova and McKee [16] suggest that a health financing system under public control that fits well with values and population preferences is likely to improve compliance and be more sustainable. A field which must be investigated is the collaboration of the public hospitals with the private insurance companies considering the increasing number of persons with private insurance the last fifteen years [11].
Public insurance is compulsory while private insurance is not compulsory (subsidiary) in Greece. In the last few years, the percentage of private insurance has been dramatically increased as an addition to public insurance. Users pay for services not included in the packages. These can be paid for by voluntary (private) health insurance provided by private shareholding companies for additional health insurance. Citizens have the right to purchase packages of additional services from the private health insurance funds, thus guaranteeing a mixed system of public-private financing. In addition they are entitled to purchase packages offering a full range of health care services.
The basic package for primary health care contains the following services:
• ambulatory care (examination)
• surveillance, home visits, consultations
• health promotion and health prophylactics
• immunizations
• referrals for medical and diagnostic tests
• prescription of drugs, etc.
For the performance of services included in the basic package, general practitioners are paid by capitation on the basis of the number of patients on their list. In addition to the basic package of services general practitioners participate in special health programmes, called Management of Health
Priorities, including:
• maternal and infant health care
• adolescent health care
Hospital care in Bulgaria and Greece
In contrast to the therapeutical institutions of non-hospital or pre-hospital care, these foundations have a more mutli-complex form of organization and way of practicing their medical activity.
Depending on the kind of medical care provided and the duration of treatment and stay in hospital, the medical centers of hospital care in Bulgaria are divided as follow:
a. Hospitals of active treatment – these hospitals treat patients with acute diseases, maltreatment, severe chronic diseases, conditions requiring operating treatment in a hospital, obstetric services and medical-cosmetic services.
b. Hospitals for completion of treatment and prolonged treatment – these hospitals treat patients with needs for a prolonged rehabilitation of health, people with chronic diseases requiring care.
c. Hospitals for rehabilitation – these hospitals treat people with needs for physical treatment, kinetic and psychic rehabilitation, physiotherapy, bath-therapy and sea-therapy.
d. Hospitals for completion of treatment, prolonged treatment and rehabilitation – In these hospitals, activities referring to the above b and c hospitals are carried out.
Depending on the specializations of medical activity applied in these hospitals, they are divided into specialized (with one specialization) and general (with many specializations).
Depending on the area served by a hospital, they can be divided as:
a. District – when the hospitalized patients come from one or neighboring municipalities.
b. Regional – when the hospitalized patients come from municipalities from one region.
c. Tertiary – when the hospitalized patients come from different regions.
d. National – where diagnostic and therapeutical activities and scientific research on the application of modern medical technologies are carried out (unique for the country) or duties for the processing and accomplishment of the national health policy are performed.
Finally, depending on the property status, the hospitals are divided into public (with government or municipal participation) and private.
According to evidence given by the Ministry of Health [17], the total number of hospitals in Bulgaria was 218 (included the newly-established) with a total of 47.602 beds in 2001, with a tendency of reduction at rates of 12.65% in 2002 (recorded at 30 June). The average use of beds in hospitals of active treatment with many specialties was 251 days and in the specialized hospitals of active treatment was 239 days in 2000, while for 2001 was 241 and 214 days respectively [17]. The average duration of a patient attending in a hospital for active treatment for the year 2000 was 10.2 days and for the year 2001 it was 9.6 days [17].
The different kinds of medical centers of hospital care, their establishment, organization and closing down are regulated in detail in the Bulgarian Law for the medical centers [2].
The planning and the allocation of hospitals is taken from the National Health Map as implementing of the National Health Policy.
The National Health Map of Bulgaria was sanctioned by the decision no. 688/4-11-1999 of the Council of Ministers and was published in the Government Gazette [18]. The kind, the number and the allocation of hospitals into regions are also included in the annex no. 2 [18]. From the above, it can be seen that of the 218 hospitals of Bulgaria, 127 are hospitals with many specializations for active treatment, 73 are specialized and 18 are private-owned.
The hospital care in Greece is distinguished into three categories [18]:
a. First-degree hospital care – comprises the services provided in the out-patients of a hospital, such as diagnosis and treatment of patients and emergencies cases.
b. Second-degree hospital care – refers to the admittance of a patient in the hospital for diagnosis and includes medical attendance, laboratory checking and general operations.
c. Third-degree hospital care – refers to the admittance of the patient in the hospital but in addition, it provides:
• Highly specialized knowledge
• Highly specialized skills of accession
• Highly specialized equipment
Cooperation and support of other, except the major, medical specializations [19].
Currently, Greece has 139 hospitals, of which 114 are general and the remaining 25 are specialist. Military hospitals are not included.
The hospital care is provided by different types of hospitals, distinguished as follow:
a. Depending on the extent of the rendered services, into general and specialized hospitals. The general ones have departments for treatment in more than one specialization, while the specialized have one specialization such as psychiatric, anti-cancer, obstetrics etc. The specialized hospitals cover the needs of more regions.
b. Depending on the duration of medical attendance, they are distinguished into hospitals of "acute" treatment, where the duration is less than one month and into hospitals of "chronic" diseases, where the duration is longer e.g. psychiatric, geriatric etc.
c. Depending on their legal form they are distinguished as State or L.P.P.J., Municipal, Public-Benefit, Hospitals of Insurance Organizations, and Private.
d. Depending on their geographical range and the size of the population served they are distinguished as follows:
• Local, with an area of responsibility up to 50.000 residents
• Prefectural, serving an area up to 200.000 residents. They operate in every prefecture and have departments, at least, in basic specializations providing medical training only in some specialized fields.
• Regional, operating in chair of each health region, covering the needs of people of the region. They have departments for all or most medical specialties, providing medical training for all or the most specialties contributing to the promotion of the medical research. The regional hospitals are considered as units of tertiary degree care, while the Local and Prefectural as units of secondary degree care.
e. Depending on their the type of training provided they are distinguished into:
• university hospitals
• hospitals having a limited training role
• hospitals that do not perform training work.
As in other central and eastern European countries, informal payments by patients for health care services were common in Bulgaria during the 1980s, although not officially sanctioned by the communist authorities. Such payments became increasingly common during the 1990s [1,15]. Informal payment is also a problem for the Greek Medical Care System. This problem is widely observed in the most of the Greek hospitals in which out of pocket payments are given to the hospital doctors mainly to perform operations. Balabanova and McKee [20] reportrd that informal payments stem from the low income of staff, patients seeking better treatment, acute funding of shortages and from tradition. Another problem, especially in Greece, is the creation of artificial demand for medical services. A good example is the recent estimation for both public and private medical sectors where 70% of the heart operations are performed often without necessary indications [21]. Besides, Greece is the only European country in which embryoctomy accounts for 75% of childbirths.
Discussion
The main aim of this study was to investigate the basic differences and problems in hospital care between Greece and Bulgaria. So far, there are very few reports of both systems and a comparison could help to improve the medical services in both countries.
In both countries, besides the compulsory insurance, there is also the additional voluntary sector which is provided by private companies. The most important difference in health systems between the two countries is the family doctor (specialist in general medicine) existing in Bulgaria. Although the Greek government legislated for the family doctor with the modification of NSH in 1997 [5], the application of this was cancelled four years later [14]. A reason could be the low number of general doctors (560) [22] existing in Greece, which were not enough to cover the needs for applying this system. The reason for the low number of general doctors was that they were considered as second class doctors [23].
A basic difference between the two systems is also that, in Bulgaria, every insured person needs a 'referral form' completed by the family doctor before visiting a hospital for medical attendance (except emergencies). Therefore, by this system, patients do not get crowded in the hospitals of Bulgaria whereas in Greece patients can directly visit a hospital independent of the severity of their health problem. Other causes of the overcrowding in the Greek hospitals are the low number of available beds, the low number of hospitals (139) (in comparison to population) and the ineffectiveness of the first-degree level of care. In contrast, there are 218 hospitals in Bulgaria, although its population is much lower than in Greece. Probably, an increase in the number of hospitals in Greece could improve the medical services.
Some of the common problems for both systems are:
• Limited financing
• Big debts
• Lack of appropriate investment policy
• Lack of objective method for the costing of medical activities
• Inefficient management.
The liability insurance of Greeks in the last few years has provided of a new income for hospitals in addition to the financial support given from Government. However, in both countries problems are caused by the low sums, which are paid by the insurance organizations and from the delay of payment. Thus, the incomes from the insurance organizations can not cover the expenses of hospitals. Solutions to these problems must come from the Government. Increases in the financial support for the public medical care, reduction of expenses, and distribution of the incomes are some of the suggested solutions. Also, essential is the re-planning of the hospitalization cost and maybe, the "closed hospitalization" in the hospitals should be abolished. Especially for Greek medical care, the first-degree level of care must be improved to solve the problem of overcrowding in hospitals. In Greece, effords by the responsible ministers to improve the NSH have been ineffective.
Conclusion
It is important to report that Bulgaria has made significant steps of progress in respect to medical care in the last few years and this is only the beginning as the medical reform has recently started (1999) and has not been completely applied. With regard to hospital care, Greece is far ahead in the technological equipment and the scientific field. However, Bulgaria is better than Greece on the viewpoint of admission to hospital. Another characteristic that should be pointed out regarding the improvement of hospital care in Bulgaria is the potential of competition among hospital foundations, according to the recent law, something that has not yet happened in Greece.
Competing interests
The author(s) declare that they have no competing interests.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
I am grateful to Dr. Thomas Thomidis for their constructive comments on earlier versions of this manuscript.
==== Refs
Balabanova D McKee M Access to health care in a system transition: the case of Bulgaria International Journal of Health Planning and Management 2002 17 377 395 12476643 10.1002/hpm.687
Government Gazette of the Republic of Bulgaria 1999 62
Government Gazette of the Republic of Bulgaria 1998 70
N.2194/ FEK 34 (A), 16/03/1994 Athens 1
N.2519/ FEK 165 (A), 21/08/1997 Athens 6429 6463
N.2889/ FEK 37 (A), 02/03/2001 Athens 1079 1104
N.3172/ FEK 197 (A), 06/08/2003 Athens 3811 3834
Brommels M The challenge of two markets Lectures during the conference "Health Management in an International Perspective", Department of Health Policy and Management, Erasmus University Rotterdam, The Netherlands 23 October 1992.
Boonekamp L Marketing for the health-care organizations: An introduction to network management Journal of Management in Medicine 1994 8 11 24 10140626 10.1108/02689239410073321
N.3204/ FEK 296 (A), 23/12/2003 Athens 4997 5018
Souliotis K The role of the private insurance on the Greek Medical Care 2000 Papazisis Publisher, Athens 93
Sigalas I Administration of the Hospitals 1996 University Notes. University Lectures. Thessaloniki 28 31
Sigalas I Organization and operation of the Regional hospital 1990 Thessaloniki. AHEPA Training Center 5 17
Dimoliatis G Kyriopoulos G Laggos D Filalithis T The public medical care in Greece 2002 Themelio Publisher 141
Koulaksazov S Todorova S Tragakes E Hristova S Health Care Systems in Transition: Bulgaria The European Observatory on Health Care Systems 2003 (accessed 25 November 2004)
Balabanova D McKee M Reforming health care financing in Bulgaria: the population perspective Social Science and Medicine 2004 58 753 765 14672591 10.1016/S0277-9536(03)00241-7
Ministry of Health of Bulgaria. Statistical Service 2003
Government Gazette of the Republic of Bulgaria 1999 98
Theodorou M Sarris M Soulis S Health systems and Greek reality Athens 1995 64
Balabanova D McKee M Understanding informal payments for health care: the example of Bulgaria Health Policy 2002 62 243 273 12385850 10.1016/S0168-8510(02)00035-0
Tsantos T Problems of the Greek Medical Care Systems Newspaper (Greek) 2002 2371 11
Kyriopoulos G Filolithis T First-degree level of care in Greece 1996 Themelio Publisher 169
Liaropoulos L Organization of health services 1993 Teaching notes. Department of Nursing, University of Athens 75
| 15921530 | PMC1156891 | CC BY | 2021-01-04 16:31:50 | no | BMC Health Serv Res. 2005 May 28; 5:41 | utf-8 | BMC Health Serv Res | 2,005 | 10.1186/1472-6963-5-41 | oa_comm |
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-311587174810.1186/1471-2334-5-31Research ArticleInvestigation of Staphylococcus strains with heterogeneous resistance to glycopeptides in a Turkish university hospital Nakipoglu Yasar [email protected] Sengul [email protected] Atahan A [email protected] Handan [email protected] Department of Microbiology and Clinical Microbiology, Faculty of Medicine, University of Istanbul, Istanbul, Turkey2 Department of Infectious Diseases and Clinical Microbiology, Faculty of Medicine, University of Istanbul, Istanbul, Turkey2005 5 5 2005 5 31 31 3 12 2004 5 5 2005 Copyright © 2005 Nakipoglu et al; licensee BioMed Central Ltd.2005Nakipoglu et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The hetero-glycopeptide intermediate staphylococci is considered to be the precursor of glycopeptide intermediate staphylococci especially vancomycin intermediate Staphylococcus aureus (VISA). For this purpose, we aimed to investigate the heterogeneous resistance to glycopeptide and their frequencies in 135 Staphylococcus strains.
Methods
Heterogeneous resistance of Staphylococcus strains was detected by inoculating the strains onto Brain Heart Infusion agar supplemented with 4 mg/L of vancomycin (BHA-V4). Agar dilution method was used for determining MICs of glycopeptides and population analysis profile was performed for detecting frequency of heterogeneous resistance for the parents of selected strains on BHA-4.
Results
Eight (6%) out of 135 Staphylococcus strains were exhibited heterogeneous resistance to at least one glycopeptide. One (1.2%) out of 81 S. aureus was found intermediate resistance to teicoplanin (MIC 16 mg/L). Other seven strains were Staphylococcus haemolyticus (13%) out of 54 coagulase negative staphylococci (CoNS). Six of the seven strains were detected heterogeneously reducing susceptibility to vancomycin (MICs ranged between 5–8 mg/L) and teicoplanin (MICs ranged between 32–64 mg/L), and one S. haemolyticus was found heterogeneous resistance to teicoplanin (MIC 32 mg/L). Frequencies of heterogeneous resistance were measured being one in 106 – 107 cfu/ml. MICs of vancomycin and teicoplanin for hetero-staphylococci were determined as 2–6 folds and 3–16 folds higher than their parents, respectively. These strains were isolated from six patients (7%) and two (4%) of health care wokers hands. Hetero-VISA strain was not detected.
Conclusion
Heterogeneous resistance to glycopeptide in CoNS strains was observed to be significantly more emergent than those of S. aureus strains (vancomycin P 0.001, teicoplanin, P 0.007). The increase MICs of glycopeptide resistance for subpopulations of staphylococci comparing with their parents could be an important clue for recognizing the early steps in the appearance of VISA strains. We suggested to screen clinical S. aureus and CoNS strains, systematically, for the presence of heterogeneously resistance to glycopeptide.
==== Body
Background
Since the first report published from Japan in 1997 on vancomycin-intermediate (MIC 8–16 mg/L) Staphylococcus aureus (VISA)[1] and studies from different countries, suggested that the emergence of glycopeptide resistance among staphylococci is a global issue [2]. Survey of resistance to glycopeptide indicates that there are two types of resistances. The first type is homogeneous resistance which occurs as a result of thickening in the cell wall, such as being in VISA strain [3]. The other homogeneous resistance was reported from Michigan in June 2002 [4] in which MIC of vancomycin for the S. aureus strain was found ≥ 32 mg/L and regarded as the first vancomycin resistance S. aureus (VRSA). VRSA strain has vanA gene of enterococci which explains the origin of VRSA strain [5].
The main difference between the VISA and VRSA strains is that the later could transfer the glycopeptide resistance among the strains by plasmids. Homogeneous resistance could be determined by conventional methods as described by the National Committee for Clinical Laboratory Standards (NCCLS) [6]. The second type is heterogeneous resistance, which was described firstly by Hiramatsu et al [1] from Japan in 1997, and it is represented with strain Mu3 (hetero-VISA). Mu3 strain has been isolated from an old man with pneumonia who did not give any response to vancomycin therapy, although MIC of Mu3 was 4 mg/L it has been containing a mutant subpopulation of cells having intermediate resistance to vancomycin (MIC 8 mg/L).
It is difficult to detect heteroresistance by NCCLS methods because the concentration of bacterial population which was advised by NCCLS (5 × 105 cfu /mL) is not sufficient to detect heteroresistance (1 in 106 cfu/mL). Population analysis method is the most common screening method for the detection of heteroresistant staphylococci [1]. Depending on many studies hetero-VISA is precursor strain for VISA [7]. A growing concern over recent reports from many countries suggests that the hetero-VISA is prevalent and responsible for the failure of vancomycin therapy [5]. Clonal spread of these strains via hands of Health Care Workers (HCW) has also been reported [8]. Beacuse of low virulance of CoNS, they have been dismissed as being culture contaminants in the past, however, in recent years they have been assumed of greater importance as true pathogens and the emergence of strains with decreased levels of susceptibility to vancomycin and teicoplanin have been noticed in many studies reviewed by Cercenado et al [9]. Istanbul University hospital is a large hospital with 1500 bed capacity, we aimed to determine heterogeneously resistance to vancomycin and teicoplanin in S. aureus and CoNS strains for the first time in Turkey.
Methods
Strains
A nonrepetitive 135 Staphylococcus strains were collected from September 2001 to April 2002 from the Faculty of Medicine of the Department of Microbiology and Clinical Microbiology of the University of Istanbul in Turkey. These were 87 strains (73 S. aureus and 14 CoNS) which were clinical or colonized isolates of the skin of different body sites of the hospitalized patients, and 48 strains (40 CoNS and 8 S. aureus) colonized on the skin of the hands of HCWs. CoNS strains were identified to the species level by standard biochemical procedures [10]. The species distribution of CoNS were as follows: S. haemolyticus 18 (33.3%), S. epidermidis 10 (18.5%), S. simulans 6 (11.1%), S. schleiferi 6 (11.1%), S. lugdunensis 3 (5.5%), S. arlettae 3 (5.5%), S. xylosus 3 (5.5%), S. hominis 3 (5.5%), S. capitis 1 (1.8 %), and S. warneri 1 (1.8%). Staphylococcus aureus ATCC 29213 was used as control strain for determining MICs of glycopeptides. Heterogeneous VISA strain Mu3 was used as control in population analysis profile method.
MICs criteria of Glycopeptides
MICs were evaluated according to the NCCLS breakpoints for Staphylococcus strains [11] and were as follows: For vancomycin ≤ 4 mg/L as susceptible, 8 – 16 mg/L as intermediate, and ≥ 32 mg/L as resistance. For teicoplanin: ≤ 8 mg/L as susceptible, 16 mg/L as intermediate and ≥ 32 mg/L as resistance.
Detection of Staphylococcus strains with hetero-reduced susceptibility to vancomycin
This method was applied as described previously by Hiramatsu et al [1]. Briefly, overnight cultures were adjusted to 0.5 McFarland turbidity and 10 μl of the suspension was inoculated onto Brain Heart Infusion agar (BHA) (BBL; Becton Dickinson and Company, Cockeyvilles, USA) plates containing 4 mg/L of vancomycin (Sigma, chemical Inco, Germany) (BHA-V4), and incubated at 37°C for 48 hours. If a countable number (one to 30) of colonies grew, the strain was considered to have potential heteroresistance to glycopeptide and was confirmed with population analysis profile method. The strain was accepted as susceptible to vancomycin if an obvious growth was not seen on the inoculated plate.
Population analysis profile method
This method was used for detecting of staphylococci subclones intermediate or resistant to glycopeptide. For this purpose, the parents of the selected colonies on BHA-V4 were used and the method was performed as previously defined by Hiramatsu et al [1]. A 50-μl aliquot of the overnight isolates and control strain (Mu3) in Brain Heart Infusion (BHI) broth were adjusted to optical density of 578 (108 cfu/ml) and serial 10-fold dilutions were spread over two different BHA plates, one containing vancomycin at concentrations ranging from 1 to 10 mg/L (ten dilutions), and the other containing teicoplanin (Gruppo Lepetit S.P.A, Italy) at concentrations ranging from 1 to 64 mg/L (two fold dilutions). After incubation at 37°C for 48 hours, the number of viable cells were calculated and plotted on a semi-logarithmic scale. Heterogeneous Resistant Staphylococci was defined as any screen-positive strain which contained subpopulations with MIC > 4 mg/L for vancomycin or ≥ 16 μg/ml for teicoplanin at a frequency of 1 in 106 cfu/mL or higher.
Detection MICs of vancomycin and teicoplanin
MICs were determined for the parents of strains grown on BHA-V4 by using cation-adjusted MHA (Oxoid, Unipath LTD, Hampshire, England) and two fold increments of vancomycin at concentrations ranging from 0.25–32 mg/L and teicoplanin at concentrations ranging from 0.25–128 mg/L and the agar dilution method was performed according to the NCCLS [6].
Patients Medical Record
Records of patients who had colonized or infected with Staphylococcus strains with decreased susceptibility to glycopeptides were reviewed and data related to the patients age, gender, ward, underlying disease, specimen, duration of hospitalization and previous treatment with glycopeptides were evaluated.
Statistical analysis
Fisher's test was used for comparing of glycopeptide resistance between S. aureus and CoNS strains.
Results
Detection of staphylococci with reduced susceptibility to glycopeptides
Nine (7%) out of 135 Staphylococcus strains grew on BHA-V4 screening plates. Two of these strains were found as being methicillin susceptible S. aureus (MSSA, 4502 and 4503) and other seven strains were found as being methicillin resistant CoNS (MRCoNS) (strains from 4504 to 4510).
MICs of vancomycin and teicoplanin of strains grew on BHA-V4
MICs of vancomycin and teicoplanin for the parents of subclones were as follows: for each of 4502 and 4503 strains were found being susceptible (1 mg/L), for seven MRCoNS strains were found susceptible to vancomycin and ranged between 1–4 mg/L. MICs of teicoplanin were found being intermediately susceptible (16 mg/L) for two MRCoNS (4508, 4510) strains and susceptible (1 to 8 mg/L) for another five strains.
Population analysis profile
Although S. aureus no 4502 grew on BHA-4, its sublones were susceptible to vancomycin (MIC 4 mg/L) and teicoplanin (MIC 8 mg/L). This strain was not accepted as heterogeneous strain, whereas those of S. aureus 4503 were susceptible to vancomycin (MIC 3 mg/L) and resistant to teicoplanin (MIC 32 mg/L) and the strain was hetero-teicoplanin resistant S. aureus (TRSA). MICs for subclones of seven MRCoNS (all were S. haemolyticus) of vancomycin (except strain 4506: MIC 4 mg/L) were reduced susceptibility and varied between 5–8 mg/L, and all were resistant to teicoplanin (MICs were ranged between 32–64 mg/L). Heterogeneous frequencies in staphylococci strains were observed 1 in 106 – 107cfu/ml (Fig. 1, 2, 3, 4). Therefore, heteroresistance to glycopeptide antibiotics was found to be one (1.2%) out of 81 S. aureus strains and seven (13%) out of 54 CoNS strains. Heteroresistance to glycopeptide was higher in S. haemolyticus (39%) and it was found as being the only species among CoNS that has showed heteroresistance to glycopeptide.
Figure 1 Population analysis of two S. aureus strains (4502,4503) grew on BHA-V4. MICs of vancomycin for subclones were susceptible (3–4 mg/L).
Figure 2 Population analysis of two S. aureus strains grew on BHA-V4. MICs of teicoplanin for subclones of MSSA-4502 were susceptible (8 mg/L) and for MSSA-4503 were resistant (32 mg/L).
Figure 3 Population analysis of MRCoNS strains grew on BHA-V4. MICs of vancomycin for six MRCoNS strains subclones (4504, 4505, 4507, 4508,4509,4510) were reduced (5–8 mg/L) and for one strain (4506) subclones were susceptible (4 mg/L).
Figure 4 Population analysis of seven MRCoNS strains (4504–4510) grew on BHA-V4. MICs of teicoplanin for these strains subclones were resistant (32–64 mg/L).
Glycopeptide resistance in clinical and colonized strains
Six (7%) out of 87 patients and two (4%) out of 48 HCWs hands isolates displayed heterogeneous resistance to at least one glycopeptides. Only one strain (4503) out of eight S. aureus strains colonizing HCWs hand's skin was hetero-intermediate resistance to teicoplanin, and none of the 73 screened clinical S. aureus strains had reduced susceptibilty to glycopeptides. For CoNS strains, six out of 14 patients isolates and one out of 40 HCWs hand skin strains were hetero-resistant to vancomycin and teicoplanin with the exception of 4506 strain; it was susceptible to vancomycin and resistant to teicoplanin (Table 1).
Table 1 Clinical features of patients infected or colonized with MRCoNS with reduced susceptibility to glycopeptides
Patients Strain No Age/sex Ward Clinical feature Underlying disease Specimen Duration of hospitalization (days) Previous treatment with glycopeptide (days) Treatment with glycopeptide (days) Clinical outcome
4504 46/F Infectious Diseases Pneumonia Trauma tracheal aspiration fluid 20 No vancomycin (14) Improved
4505 33/M Haematology Septicemia CML Groin skin 27 No No NT
4506 44/M Haematology Septicemia AML Abdominal wall skin 46 vancomycin (20) No NT
4507 46/F Haematology Septicemia ALL Axilla skin 60 vancomycin (10) No NT
4508 26/M SICU Septicemia Trauma and Fracures Of mandibula and maxilla Blood 31 No vancomycin (10) Improved
4510 75/F Infectious Diseases Septicemia Surgical operation of Colon Blood 6 No vancomycin (14) Death not related to staphylococcal infection
Abbreviations: M, male; F, female; CML, chronic myelogenous leukemia; AML, acute myelogenous leukemia; AML, acute lymphocytic leukemia; SICU, surgical intensive care unit; NT, staphylococci skin colonization not treated.
Discussion
Heterogeneous resistance to teicoplanin among S. aureus strains was very low (1.2%), whereas this resistance among MRCoNS was higher (13%). Using vancomycin was licensed in 1994 in Turkey and preceded teicoplanin about two years. From this point of view, the expectation was to detect Staphylococcus strains with vancomycin resistance more than teicoplanin resistance. But our results have shown that seven S. haemolyticus were resistant to teicoplanin (MIC ≥ 32 mg/L) and one S. aureus strain has an intermediate resistance to teicoplanin (MIC 16 mg/L), whereas resistance to vancomycin in the same eight strains subclones was increased mildly above the susceptible range (4 m/L) in six strains, and intermediate range (8 mg/L) in two strains. It is hard to link between the glycopeptide use and appearance of resistance to this antibiotic group in our strains, because except in two strains (4506 and 4507), no previous treatment with any glycopeptide was found through viewing the medical reports of the patients colonized or infected with these strains. On the other hand, the MICs of vancomycin even for these two strains were 4 mg/L for the former and 5 mg/L for late strain. The selective pressure, independently from the treatment with glycopeptide, could be the major factor in the heterogeneous resistance among strains (Table 1).
Historically, S. aureus acquired teicoplanin resistance before it acquired vancomycin resistance. Teicoplanin resistance is frequently accompanied with a small increase in vancomycin resistance [12]. Shlaes et al [13], demonstrated that PBP2 is overproduced in a TRSA mutant strain (MIC 16 mg/L) compared with its parent strain. Over-production of PBP2 is also observed in the hetero-VISA strain Mu3 which is resistant to teicoplanin. Most of the studies on glycopeptide resistance mechanisms were done with S. aureus strains and it may be also true for CoNS strains; but, more studies are needed.
Species variation has been detected in vancomycin susceptibility of CoNS, according to many studies [14-16]. This resistance was more common in S. haemolyticus and S. epidermidis. In this study, we also obtained similar results, in which S. haemolyticus was the only species revealed with high resistance (39%) to glycopeptide among ten different species. Indeed there is not a sufficient explanation which relates directly the glycopeptide resistance and species phenomenon.
A survey of hetero-VISA revealed differences in the prevalence of hetero-VISA between the countries and even between university and nonuniversity hospitals in the same country. In Japan after the first report on hetero-VISA in 1997, Hiramatsu et al [1] have reported a result of hetero-VISA as 3%, however, no hetero-VISA strain was found in a study conducted in the same year by Ike et al [17] on 6,625 S. aureus of university and nonuniversity hospitals strains. Prevalence of heteroVISA strains in the other European countries like Italy, Germany, France, Netherland were reported as 1.1%, 0.21%, 0.6%, 6%, respectively, [18-21] and in Thailand and Korea as 1.93% and 0.54%, respectively [22,23]. We did not isolate heteroVISA strain, but we detected heteroTRSA. Heterogeneous resistances to vancomycin and teicoplanin among CoNS were found to be 11% and 13%, respectively, and higher than those obtained by Gruneberg et al [24], a European collaborative study on 1594 CoNS from bloodstream infections, and they found that all of the isolates were susceptible to vancomycin and only 0.7% of them were resistant to teicoplanin.
Examination of six heteroMRCoNS showed that three strains colonized the skin of different body sites of immunocompromised patients and three others were isolated from clinical specimens (Table 1). The MICs of vancomycin for these strains were varied between the susceptible and intermediate ranges, whereas teicoplanin was in the resistance ranges. The strain no 4509 which was isolated from the hand skin of HCW was intermediate resistant to vancomycin (7 mg/L) and resistant to teicoplanin (64 mg/L). According to the Spanish studies, 0.55 % of CoNS strains exhibiting either decreased levels of susceptibility or true resistance to teicoplanin [9].
Two (4%) out of 48 HCWs hands were colonized with two types of strains, S. aureus 4503, which was the only intermediate resistant to teicoplanin (MIC 16 mg/L) as well as S. haemolyticus 4509. The last strain was an important strain because vancomycin MIC for its subpopulations was closely to intermediate (7 mg/L) and also resistant to teicoplanin. The role of these two HCW for spreading of the strains was not investigated.
The treatment failures due to the selection of teicoplanin resistant mutants have been reported following treatment of S. aureus infections with teicoplanin [12,13,25]. A study made by Wong et al [26], showed that the patients infected by heteroresistant strains have higher mortality rates than the patients infected by sensitive isolates. Other studies on treatment failure associated with heterogeneous VISA have been reported as well [1,5,27]. Our three infected patients with hetero-CoNS have been treated successfuly with vancomycin, on the other hand, death of the third patient was not related to the infection (Table 1).
Although the number of S. aureus and CoNS strains included in this study were low, and it is not enough to predict the prevalence. This first study from one of the four largest university hospitals from Turkey may give an idea regarding the status of resistance to glycopeptide among Staphylococcus strains, and as a preliminary study need to be followed up with other studies.
Conclusions
Our study showed that CoNS strains have an emerging heteroresistance for both vancomycin and teicoplanin more than those of S. aureus strains and this difference was found to be significant (P = 0.001, P = 0.007 for vancomycin and teicoplanin, respectively). We suggested to screen clinical S. aureus and CoNS strains, systematically, in case of the presence of heterogeneously resistance to glycopeptide to avoid treatment failure or any nosocomial outbreaks in the future.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
YN participated in the study design and was principal writer of the manuscript. YN, SD, and HK carried out the laboratory studies; AC was responsible for collecting the data from patients medical records and treatment of the patients.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Hiramatsu K Aritaka N Hanaki H Kawasaki S Hosoda Y Hori S Fukuchi Y Kobayashi I Dissemination in Japanese hospitals of strains of Staphylococcus aureus heterogenously resistant to vancomycin Lancet 1997 350 1670 1673 9400512 10.1016/S0140-6736(97)07324-8
Srinivasan A Dick JD Perl TM Vancomycin Resistance in Staphylococci Clin Microbiol Rev 2002 15 430 438 12097250 10.1128/CMR.15.3.430-438.2002
Hiramatsu K Ito T Hanaki H Vancomycin-Resistant Staphylococcus aureus (VRSA) 1999
CDC Staphylococcus aureus resistant to vancomycin-United States Morb Mortal Wkly Rep 2002 51 565 567
Liu C Chambers HF Staphylococcus aureus with heterogeneous resistance to vancomycin: Epidemiology, clinical significance, and critical assessment of diagnostic methods Antimicrob Agents Chemother 2003 47 3040 3045 14506006 10.1128/AAC.47.10.3040-3045.2003
National Committee for Clinical Laboratory Standards Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically, 5th ed. Approved standard M7-A5 National Committee for Clinical Laboratory Standards Wayne, Pa 2000
Guerin F Buu-Hoi A Mainardi JL Kac G Colardelle N Vaupre S Gutmann L Podglajen I Outbreak of Methicillin-Resistant Staphylococcus aureus with reduced Susceptibility to Glycopeptides in a Parisian Hospital J Clin Microbiol 2000 38 2985 2988 10921964
Mulligan ME Murray-Leisure KA Ribner BS Standiford HC John JF Korvick JA Kauffman CA Yu VL Methicillin-resistant Staphylococcus aureus : a consensus review of the microbiology, pathogenesis, and epidemiology with implications for prevention and management Am J Med 1993 94 313 328 8452155 10.1016/0002-9343(93)90063-U
Cercenado E Garcia-Leoni ME Diaz MD Sanchez-Carrillo C Catalan P De Quiros JC Bouza E Emergence of teicoplanin-resistant coagulase-negative staphylococci J Clin Microbiol 1996 34 1765 1768 8784585
Kloos WE Bannerman TL Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH Staphylococcus and Micrococcus Manual of clinical microbiology 1995 6 Washington, D.C: American Society for Microbiology 282 298
National Committee for Clinical Laboratory Standards Performance standards for antimicrobial susceptibility testing, 11th informational supplement, Approved standard M100-S11 National Committee for Clinical Laboratory Standards Wayne, Pa 2001 21
Hiramatsu K Vancomycin-resistant Staphylococcus aureus : a new model of antibiotic resistance Lancet Infect Dis 2001 1 147 11871491 10.1016/S1473-3099(01)00091-3
Shlaes DM Shlaes JH Vincent S Etter L Fey PD Goering RV Teicoplanin-resistant Staphylococcus aureus expresses a novel membrane protein and increases expression of penicillin-binding protein 2 complex Antimicrob Agents Chemother 1993 37 2432 2437 8285629
Herwaldt L Boyken L Pfaller M In vitro selection of resistance to vancomycin in bloodstream isolates of Staphylococcus haemolyticus and Staphylococcus epidermidis Eur J Clin Microbiol Infect Dis 1991 10 1007 1012 1802693
Schwalbe RS Ritz WJ Verma PR Barranco EA Gilligan PH Selection for vancomycin resistance in clinical isolates of Staphylococcus haemolyticus J Infect Dis 1990 161 45 51 2295858
Schwalbe RS Stapleton JT Gilligan PH Emergence of vancomycin resistance in coagulase-negative staphylococci N Engl J Med 1987 316 927 931 3821839
Ike Y Arakawa Y Ma X Tatewaki K Nagasawa M Tomita H Tanimoto K Fujimoto S Nationwide survey shows that meticillin-resistant Staphylococcus aureus strains heterogeneously and intermediately resistant to vancomycin are not disseminated throughout Japanese hospitals J Clin Microbiol 2001 39 4445 4451 11724859 10.1128/JCM.39.12.4445-4451.2001
Marchese A Balistreri G Tonoli E Heterogeneous vancomycin resistance in methicillin-resistant Staphylococcus aureus strains in a large Italian hospital J Clin Microbiol 2000 39 866 869 10655401
Bierbaum G Fuchs K Lenz W Szekat C Sahl HG Presence of Staphylococcus aureus with reduced Susceptibility to Vancomycin in Germany Eur J Clin Microbiol Infect Dis 1999 18 691 696 10584894 10.1007/s100960050380
Reverdy ME Jarraud S Bobin-Dubreux S Burel E Girardo P Lina G Vandenesch F Etienne J Incidence of Staphylococcus aureus with reduced susceptibility to glycopeptides in two French hospitals Clin Microbiol Infect 2001 7 267 272 11422254 10.1046/j.1198-743x.2001.00256.x
Griethysen AV Veen AV Buiting A Walsh T Kluytmans J High percentage of methicillin-resistant Staphylococcus aureus isolates with reduced susceptibility to glycopeptides in the Netherlands J Clin Microbiol 2003 41 2487 2491 12791870 10.1128/JCM.41.6.2487-2491.2003
Trakulsomboon S Danchaivijitr S Rongrungruang Y First report of Methicillin-resistant Staphylococcus aureus with reduced susceptibility to vancomycin in Thailand J Clin Microbiol 2001 39 591 595 11158112 10.1128/JCM.39.2.591-595.2001
Kim MN Hwang SH Pyo YJ Mun HM Pai CH Clonal Spread of Staphylococcus Heterogenously Resistant to Vancomycin in a University Hospital in Korea J Clin Microbiol 2002 40 1376 1380 11923359 10.1128/JCM.40.4.1376-1380.2002
Gruneberg RN Hryniewiez W Clinical relevance of a European collaborative study on comporative susceptibility of Gram-positive clinical isolates to teicoplanin and vancomycin Int J Antimicrob Agents 1998 10 271 277 9916900 10.1016/S0924-8579(98)00050-8
Moellering RC The spector of glycopeptide resistance: current trends and future consideration Am J Med 1998 104 3S 6S 9684651 10.1016/S0002-9343(98)00148-X
Wong SS Ho PL Woo PC Bacteremia caused by staphylococci with inducible vancomycin heteroresistance Clin Infect Dis 1999 29 760 767 10589883
Moore MR Perdreau-Remington F Chambers HF Treatment faliure associated with heterogeneous vancomycin-intermediate Staphylococcus aureus in a patient with endocarditis and in the rabbit model of endocarditis Antimicrob Agents and Chemother 2003 47 1262 1266 12654656 10.1128/AAC.47.4.1262-1266.2003
| 15871748 | PMC1156892 | CC BY | 2021-01-04 16:28:16 | no | BMC Infect Dis. 2005 May 5; 5:31 | utf-8 | BMC Infect Dis | 2,005 | 10.1186/1471-2334-5-31 | oa_comm |
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-331590450110.1186/1471-2334-5-33Research ArticlePrevalence and risk factors of syphilis infection among drug addicts Scherbaum Norbert [email protected] Bernhard T [email protected] Rafael [email protected] Thomas [email protected] Gerhard [email protected] Martin [email protected] Department of Psychiatry and Psychotherapy, University Hospital Essen, Germany2 Mental Health Epidemiology, Department of Psychiatry and Psychotherapy, University of Muenster, Albert-Schweitzer-Str. 11, 48129 Muenster, Germany3 Department of Public Health Medicine; School of Public Health, University of Bielefeld, Germany4 Hospital of Psychosomatic Medicine, Bergisch-Gladbach, Germany5 Centre for Psychiatry, Psychotherapy, and Psychosomatics, Dortmund, Germany6 Centre for Psychiatry and Psychotherapeutic Medicine at Gilead Hospital, Bethel, Bielefeld, Germany2005 17 5 2005 5 33 33 16 3 2005 17 5 2005 Copyright © 2005 Scherbaum et al; licensee BioMed Central Ltd.2005Scherbaum et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Recent epidemiological data show an increased trend of official estimates for syphilis infection in the general population. Many of the infected cases remain undetected leaving an underestimation of the true prevalence of syphilis in the general population, but also among subpopulations such as illicit drug users. There is limited epidemiological data published on the proportion and risk factors of syphilis infections associated with illicit drug abuse.
Methods
Illicit drug addicts (n = 1223) in inpatients units in Germany were screened (2000–01) for syphilis and interviewed regarding patterns of drug use and sexual behaviour. TPHA-test for initial screening and FTA-ABS-IgM test in TPHA-positive patients were used.
Results
In total, TPHA-tests were positive in 39 (3.3%) and 7 patients (0.6%) were IgM positive. The prevalence rate for syphilis in males was 1.9% and for women it was 8.5%. Female patients were 4.56 (CI 95% 2.37–8.78) times more likely to have a positive TPHA test than males. Sexual behaviours such as high number of sexual partners, sex for drugs/money, sex on the first day were associated with syphilis infection only in women. Females with frequent sex for drugs or money had 4.31 (CI 95% 2.32–8.52) times more likely a reactive TPHA test than remaining patients. Neither the sociodemographic factors nor sexual behaviour were statistically significant associated with syphilis infection among men at all.
Conclusion
Our data suggest the need for screening for syphilis among these illicit drug users in inpatient settings, in particular among sexual active women. This conclusion is corroborated by the finding of increasing numbers of syphilis infections in the general population. The identification of syphilis cases among drug addicts would give treatment options to these individuals and would help to reduce the spread of infection in this population, but also a spread into heterosexual populations related to prostitution.
==== Body
Background
World-wide, the incidence and prevalence of syphilis differ due to region, ethnic factors, gender and socio-economic factors [1-5]. In the 90s the yearly incidence of syphilis in Germany decreased to 1.4 cases / 100,000 inhabitants according to official numbers. However, a sentinel surveillance conducted in 1994 showed that 84% of the syphilis cases were not reported and thus, the real incidence could be as much as six times higher [5]. More recently an inverse trend of official estimates was observed with incidence increased to 3.1 cases / 100,000 inhabitants in 2002, but the increase was related only to infections acquired in the sample of men who have sex with men [6].
There are few epidemiological data published on the proportion of syphilis infections associated with illicit drug abuse. According to reports from the USA, drug addicts, especially injecting drug users (IDU), are a high-risk group for syphilis infection [7]. Gourevitch et al. reported 1996 a six year follow-up investigation of 790 opiate addicts in methadone maintenance treatment. In their study, 4.4% of the study subjects had antibodies against syphilis, and the yearly incidence of that infection was 5.7 per 1000 [7]. The prevalence of syphilis in the sample of opiate drug addicts in Germany is not reported so far. Along with abandoned general screening strategies for syphilis at admission to general psychiatric departments [8], screening for syphilis at admission to detoxification wards is no longer clinical routine in most German hospitals, despite severe complications of untreated syphilis [9,10].
Objectives
The goal of our study was to estimate the prevalence of syphilis in the population of illicit drug addicts, mainly opiate addicts, admitted to detoxification wards in psychiatric hospitals. Furthermore, risk factors associated with syphilis infection were investigated to identify individuals at high risk for syphilis infection in this population.
Methods
Study population
Our study was performed at detoxification wards for drug addicts in eight hospitals in the federal state of North Rhine-Westphalia, Germany. Overall, from 1529 eligible patients admitted to any of these detoxification wards during a two years period (2000–2001) 1309 patients participated in the study. Of these patients 86 (7%) individuals left hospital within less than 24 hours or were recurrent admissions and were excluded. Finally, there were 1223 patients in the study, which accounted for an overall response rate of 80%.
All patients were interviewed by physicians using a standardised questionnaire and subsequently tested for syphilis infection. The topics of the questionnaire were: patterns of drug use, sexual risk behaviour and known infection status for Treponema pallidum (syphilis), HIV, Hepatitis B-virus (HBV) and Hepatitis-C-virus (HCV), or Neisseria gonorrhoeae (Gonorrhoea). In addition, sociodemographic data were obtained including country of birth (or ethnic origin) and time of immigration to Germany. The study was approved by the local ethics committee of the University of Essen, Germany, and all patients gave informed consent for participation in the study. Patients identified with active syphilis obtained a standard treatment for syphilis. Penicillin G was applied intravenously due to unknown state and duration of the syphilis infection [11] and potential simultaneous HIV-infection. [12].
Microbiological testing
German syphilis screening guidelines allow for a two stage testing procedure [11]. Firstly, we used the TPHA as a screening test for Treponema pallidum (Mast TPHA kit, Mast Laboratories, UK). According to producer information the sensitivity of this test is 98.2%, the specificity is 99.7% [13,11]. Secondly, patients with reactive TPHA test were tested using a FTA-ABS IgM-test for confirmation of positive Treponema pallidum. Thus, patients with positive TPHA and FTA-ABS IgM were considered in need of treatment (excluding cases with already initiated treatment before admission). Patients with positive TPHA but negative FTA-ABS IgM were considered either as successfully treated (if there was history of treatment) or as spontaneously recovered from the infection (if there was no history of treatment). VDRL tests were carried out in treated patients to evaluate treatment effects after 3, 6, and 12 months [11]. These follow-up results were not considered in the study.
Data analysis
Data were entered and analysed with SPSS, Version 11.0 [14]. The prevalence of syphilis infection was determined by calculating the ratio of TPHA-test positives and all tested drug addicts stratified for gender. Risk factors associated with syphilis infection were analyzed by the use of logistic regression analysis adjusted for age and gender.
Results
Characteristics of the sample
In total, 1223 illicit drug addicts with a mean age of 30.1 years participated in the study of which 951 were males (77.8%) and 272 females (22.2%). 991 patients (81.6%) were born in Germany.
Nearly all (n = 1186; 97%) of the study population were opiate addicts. The length of opiate addiction in the sample ranged between 1 month and 37 years with the median of 8 years, and there was a relatively equal proportion of users with 1, 2 or 3 (up to 10) years history of use.
425 patients (34.8%) were in a methadone maintenance treatment program at admission with a duration of 2 years (SD 2.1)) on average. 1068 (87.5%) patients reported heroin abuse in the last three months, the majority of them (n = 654) by i.v.-injection. Recent cocaine abuse was reported by 612 patients (55.4%), mostly intravenously (n = 403). A stable partnership over the last year before admission was reported by 265 (21.7%) patients.
Self-reported infection status and sexual behaviour
A known Hepatitis-C-infection was reported by 65.4% of the subjects, 26.9% knew of a previous Hepatitis-B-infection, 2.3% reported being HIV-positive, 2.4% admitted having had at some time gonorrhoea, and 2.5% reported a former or current syphilis-infection. According to the self-reported sexual behaviour in our sample in table 1, a higher proportion of female (22.7%) than male (4.5%) participants had frequently sex in exchange for drugs or money. Frequent sex on the first day of meeting a partner was also more frequent in females (18.9%) than in males (2.2%). This type of sexual behaviour indicates a higher proportion of prostitutes among females than males (table 1).
Table 1 Sero-prevalence of syphilis antibodies (according to TPHA-Test) among inpatient drug addicts by gender, sociodemographic factors and sexual behaviour
Male Female
N with test % with reactive TPHA -Test P N with test % with reactive TPHA -Test p 1
Overall 890 1.9 259 8.5
Age groups 0.55 0.05
below 21 years 56 1.8 23 13.0
21–30 years 406 1.7 128 6.3
31–40 years 356 1.7 80 6.3
over 40 years 72 4.2 28 21.4
Illicit drug abuse 0.28 0.13
no or <2 y 91 4.4 33 0
2 to 4 y 193 1.0 45 6.7
5–7 y 106 1.9 31 16.1
>7 y 500 1.8 150 9.3
Migration background 0.184 0.374
Native German 694 1.6 233 9.0
Migrants 196 3.1 26 3.8
Stable partnership 0.60 0.60
no or <1 year 558 1.8 142 9.9
1 to 4 years 171 1.2 72 8.3
≥ 5 years 150 2.7 41 4.9
Number of sexual partners in the last 5 years 0.61 0.00
0–1 247 1.6 49 2.0
2–5 315 2.2 86 3.5
6–20 192 1.0 33 0
>20 136 2.9 91 19.8
Sex for drugs/money 0.52 0.00
never 826 1.8 147 2.7
seldom 22 0 30 0
frequent 22 4.5 75 22.7
Sex on the first day 0.89 0.00
never 502 1.6 120 1.7
seldom 215 1.9 41 2.4
frequent 137 2.2 90 18.9
Condom use with risky partner 0.53 0.01
never 232 1.3 73 1.4
up to half of the cases 166 1.2 34 2.9
in 3/4 of the cases 33 0 14 21.4
nearly always 182 1.6 73 15.1
no risky intercourse 236 3.0 52 7.7
p-value of chi square test or Fisher's exact test, respectively;
Syphilis-infection status based on testing
TPHA-testing was carried out in 1181 patients of which 39 (3.3%) tests were positive (female n = 22; male n = 17). The distribution of TPHA positive cases ranged from 0% to 8.6 across the 8 participating hospitals. In 24 out of the 39 TPHA-positive cases (61.5%), patients were aware of a former syphilis infection. In 6 cases a former syphilis infection was reported by the patient but the TPHA-test was not reactive. The specific IgM-test for current infection was positive in 7 out of the 39 TPHA-positive cases (0.6% of all tested patients), indicating current syphilis infection (female n = 4; male n = 3). Five out of these 7 patients were not aware of their current infection. Nine out of 32 TPHA positive but FTA-ABS IgM-negative patients did not know about an earlier infection (29%, 3 of 14 men and 6 of 18 women).
Prevalence of syphilis
The proportion of patients with reactive TPHA is presented in Table 1 showing results stratified for gender. Women had a 4.56 (95% CI 2.37–8.78) times higher risk for a reactive TPHA-test than men. Furthermore, women's sexual behaviour, such as high numbers of sexual partners, sex for drugs/money, and sex on the first day, was associated with syphilis infection. Contrary, none of the figures for men in Table 1 showed any statistically significance. Condom use in men indicated no clear trend for risky or healthy practice, in contrast to women who mostly used condoms at sexual intercourse. The duration of the methadone maintenance program (MMTP) showed no statistically significant difference (p = 0.41) between patients with reactive TPHA (mean duration of MMTP 2.4 y, SD 2.8) and those without reactive TPHA (mean duration of MMTP 2.0 y, SD 2.1).
Risk factors for syphilis
Table 2 presents a logistic regression analysis for the association between sociodemographic and behavioural parameters and syphilis infection. Apart from gender, no other demographic factor such as age or ethnic background showed any statistically significant associations with syphilis infection. Sexual behaviour was related to infection status as regular sex on the first day meeting a partner and high numbers of sexual partners were statistically significant associated with syphilis infection. Since risk factors differed between women and men, we stratified the results by gender. There were no consistent effects of age or drug abuse history on risk for reactive TPHA apart from an increased risk for older women. Sexual behaviour was associated with an increased risk for syphilis infection in women but not in men; however, even within women the increased risk was limited to women with regular sex for drugs. Females with frequent sex for drugs or money had 4.31 (CI 95% 2.32–8.52) times more likely a reactive TPHA test than remaining patients. Frequent sex on the first day of meeting a partner was statistically significant associated with syphilis infection in females (OR 17.8; 95%CI, 3.8–83.38), but not in males (OR 1.43; 95%CI, 0.37–5.5). There was an increased risk in women who used condoms more consistently, but after adjustment for regular sex for drugs this effect was not significant any more (not shown). No gender differences were found for the association of syphilis infection and migration or the existence of a stable partnership, respectively.
Table 2 Association of sociodemographic factors, clinical parameters and sexual behaviour with syphilis infection among 1223 drug addicts
Syphilis infection (positive for TPHA)
OR* 95%CI p-value
Gender
Male 1 [reference group] <0.00
Female 4.56 [2.37–8.78]
Age groups
below 21 y 1 [reference group]
21–30 y 0.59 [0.19–1.85] 0.363
31–40 y 0.59 [0.18–1.94] 0.383
over 40 y 1.94 [0.56–6.71] 0.295
Illicit drug abuse
no or <2 y 1 [reference group]
2 to 4 y 0.89 [0.23–3.53] 0.872
5–7 y 1.88 [0.52–6.79] 0.334
>7 y 0.96 [0.32–2.92] 0.945
Migration background
native German 1 [reference group]
Migrants 1.51 [0.63–3.62] 0.36
Stable partnership
> 5 years 1 [reference group]
1 to 5 years 1.17 [0.46–2.97] 0.94
<1 year 1.04 [0.35–3.16] 0.74
Number of sexual partners in the last 5 years
0–1 1 [reference group]
2–5 1.71 [0.57–5.17] 0.34
6–20 0.69 [0.13–3.67] 0.67
> 20 5.85 [2.07–16.5] 0.00
Sex for drugs/money
never 1 [reference group]
frequent 8.81 [3.66–21.19] 0.00
Sex on the first day of meeting a partner
never 1 [reference group]
seldom 1.31 [0.44–3.9] 0.63
frequent 4.98 [2.22–11.19] 0.00
OR denotes Odds ratio; CI denotes confidence interval; *adjusted for sex and age (where appropriate)
Discussion
The sero-prevalence of syphilis (according to TPHA) of 3.3% in our study is comparable to the findings of studies carried out in the USA among drug abusers in treatment facilities [7,15,16], but it is much lower than 23% among injecting drug users in Bangladesh [17]. However, only 1 in 5 of the TPHA positive patients in our study (0.6% of all screened patients) was in need of treatment, all other patients were already treated or recovered from the disease; these proportions were not reported for the other countries and may significantly differ.
According to our results, women were more often affected than men; the risk was especially increased in women having commercial sex. A study among IDUs in Spain showed a 4 times higher prevalence of syphilis and other STDs among women than men [16]. Similar to our results, syphilis was associated in this study with commercial sex in women.
We could not identify any significant risk factors for syphilis in male IDUs in our sample. We neither found a higher risk in immigrants nor in HIV positive patients. An increased risk for syphilis among immigrants especially from East European countries has been expected based on the reports of a higher prevalence of syphilis in these countries [18]. Although in our study about one third of immigrants (75 out of 205) came originally from the former USSR, no increased risk for syphilis was detected for this group. In our study intravenous drug use (heroin or cocaine) was not a proven risk factor for previous or current syphilis infection, whereas it was observed in the study by Rolfs et al. [19].
Contrary to our results an Italian study on heroin addicts with a 6-year follow-up period did not find any cases of syphilis infection [20]. The possible explanation for this difference is the higher proportion of patients with past or current commercial sex activity in our study. Additionally, the differences in prevalence can be caused by the different recruitment procedures used in both studies. While we performed a cross-sectional study, the Italian group used a cohort design. The follow-up of drug addicts over years usually requires patients in long-term contact with the support system. Such patients in follow-up or treatment settings might have a lower risk for the syphilis infection. However, this hypothesis could be not corroborated by our data, because we could not demonstrate a protective effect of the participation in substitution programs on the syphilis sero-status. An additional aspect is related especially to the infectious nature of syphilis. The Italian data were based on an investigation in one city only. In our study the prevalence of syphilis (TPHA positive tests) ranged between 0% and 8.9% for the eight study centres. This shows that syphilis may not spread efficiently between clusters of drug addicts in different cities and it may account for the differences between the Italian data and our study.
Given the complexity of long term complications, our estimate of prevalence does not provide sufficient information for a cost-effectiveness analysis. An analysis by Banger et al. supporting to abandon the general screening strategy at admission to psychiatric hospitals was based on 0.07% screening tests being positive [8]. Our data show higher numbers of screening positives in this high risk group setting compared to the study results of Banger [8]. Thus, syphilis screening for this high risk group should be favoured.
There are some important limitations to our study. Because of the very low prevalence of positive FTA-ABS-IgM tests confirming active syphilis, we had to use TPHA positive cases for our analysis of risk factors, i.e. including all patients with a past and current syphilis infection in the same group. This requires the assumption that the characteristics or behaviour of the patients with a past infection still do reflect their risk status. With respect to gender and women's sexual risk behaviour, the assumption was corroborated by the agreement between risk prediction based on TPHA and FTA-ABS-IgM. Nevertheless, the failure to determine any further risk factors can be caused by the misrepresentation of the risk factors in the analysis based largely on patients who had syphilis in the past. Some cases might have remained undetected (false negative) but it is unlikely that this would be associated with any particular risk factors. On the other side the false positive cases would have the effect of diminishing estimate of the true increased risk.
Conclusion
In conclusion our data underline the importance to screen drug addicts in inpatient treatment settings for syphilis as these patients represent a high risk group for sexually transmitted diseases. In particular sexual active women having high numbers of sexual partners were at highest risk for syphilis infection. Especially infected prostitutes may act as a multiplier in the dynamics of spread of infection into the heterosexual population.
The decision to abandon a general screening for syphilis in many hospitals was to our knowledge neither based on a formal cost-effectiveness analysis nor on the analysis of the impact on the incidence of syphilis in a society. Taking into account increasing numbers of syphilis cases worldwide, we would suggest at least a screening for syphilis among high risk opiate drug addicts in hospitals where general screening for syphilis was abandoned. Furthermore, it might useful to suggest testing for gonorrhea and chlamydia as well as for syphilis for people in drug treatment programs as a significant amount of gonorrhea and chlamydia is asymptomatic, especially in women.
Although we studied only the situation in Germany, similar conditions may apply to other developed countries.
Declaration of competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
NSCH, BTB and RM contributed to the study design, data collection and analysis, and preparation of the draft manuscript; TK, GR, and MR were responsible for conceptualisation of the study and for coordination of data collection; all authors took part in revising the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors thank nurses and doctors who participated in the coordination and collection of data.
==== Refs
Aseffa A Ishak A Stevens R Fergussen E Giles M Yohannes G Kidan KG Prevalence of HIV, syphilis and genital chlamydial infection among women in north-west Ethiopia Epidemiology & Infection 1998 120 171 7 9593487 10.1017/S0950268897008595
Dada AJ Ajayi AO Diamondstone L Quinn TC Blattner WA Biggar RJ A serosurvey of Haemophilus ducreyi, syphilis and herpes simplex virus type 2 and their association with human immundeficiency virus among female sex workers in Lagos, Nigeria Sexually Transmitted Diseases 1998 25 237 42 9587174 10.1097/00007435-199805000-00004
Kambarami RA Manyame B Macq J Syphilis in Murewa district, Simbabwe: an old problem that rages on Central African Journal of Medicine 1998 44 229 32 10101428
Prange HW Ritter G [The epidemiology of neurosyphilis] Nervenarzt 1981 52 32 35 (German) 7219611
Epidemiological Bulletin [Current data and information on infectious diseases: current implications] Robert Koch Institute 1997 16 109 110 (German)
Epidemiological Bulletin [Current data and information on infectious diseases and Public Health: Syphilis in Germany in 2002] Robert Koch Institute 2003 36 285 289 (German)
Gourevitch MN Hartel D Schoenbaum EE Selwyn PA Davenny K Griedland GH Klein RS A prospective study of syphilis and HIV infection among injection drug users receiving methadone in the Bronx American Journal of Public Health 1996 86 1112 1115 8712270
Banger M Olbrich HM Fuchs S Gastpar M [Cost-effectiveness of syphilis screening in a clinical for general psychiatry] Nervenarzt 1995 66 49 53 (German) 7885513
Luger A Gschnait F Niebauer G Economical aspects of routine screening for syphilis in Viennese municipal hospitals Wiener Klinische Wochenschrift 1987 99 808 11 3124355
Gjestland T The Oslo study of untreated syphilis; an epidemiologic investigation of the natural course of the syphilitic infection based upon a re-study of the Boeck-Bruusgaard material Acta Derm Venereol 1955 35 3 368 Annex I-LVI 13301322
Schneede P Hofstetter A Infectious disease working group of the German Society of Urology: Guidelines for the diagnostic and treatment of STDs
French P Syphilis Guidelines Revision Group. UK national guidelines on the management of late syphilis 2002
Egglestone St Ashley I Sensitivity and specificity of TPHA test Supra Regional Treponemal Serology Lab, PHL, Bristol 1995
SPSS for Windows, Release 11.01 SPSS Inc Headquarters, 233 S Wacker Drive, 11th floor Chicago, Illinois 60606
Hwang LY Ross MW Zack C Bull L Rickman K Holleman M Prevalence of sexually transmitted infections and associated risk factors among populations of drug abusers Clin Infect Dis 2000 31 920 6 11049771 10.1086/318131
Muga R Roca J Tor J Pigem C Rodriguez R Egea JM Vlahov D Munoz A Syphilis in injecting drug users: clues for high-risk sexual behaviour in female IDUs Int J Std AIDS 1997 8 225 8 9147154 10.1258/0956462971919967
Azim T Bogaerts J Yirell DL Banerjea AC Sarker MS Ahmed G Amin MM Rahman AS Hussain AM Injecting drug users in Bangladesh: prevalence of syphilis, hepatitis, HIV and HIV subtypes AIDS 2002 16 121 3 2002; Jan 4 11741170 10.1097/00002030-200201040-00015
Atlani L Carael M Brunet JB Frasca T Chaika N Social change and HIV in the former USSR: the making of a new epidemic Soc Sci Med Jun 2000 50 1547 56 10.1016/S0277-9536(99)00464-5
Rolfs RT Goldberg M Sharrar RG Risk factors for syphilis: cocaine use and prostitution American Journal of Public Health 1990 80 853 7 2356911
Lugobini F Quaglio G Mezzelani P Lechi A No positive tests for syphilis in 6 years of observation among heroin drug users in north-eastern Italy Addiction 2002 97 104 105 11895263 10.1046/j.1360-0443.2002.0050j.x
| 15904501 | PMC1156893 | CC BY | 2021-01-04 16:28:14 | no | BMC Infect Dis. 2005 May 17; 5:33 | utf-8 | BMC Infect Dis | 2,005 | 10.1186/1471-2334-5-33 | oa_comm |
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-341590450210.1186/1471-2334-5-34Research ArticleEchinococcal disease in Alberta, Canada: more than a calcified opacity Somily Ali [email protected] Joan L [email protected] Lilly J [email protected] Ravi [email protected] Thomas J [email protected] Department of Pathology, University of Alberta, Edmonton AB, Canada2 Department of Pediatrics And Stollery Children's Hospital, University of Alberta, Edmonton AB, Canada3 Department of Medicine, University of Alberta, Edmonton AB, Canada4 Department of /Radiology, University of Alberta, Edmonton AB, Canada2005 17 5 2005 5 34 34 24 1 2005 17 5 2005 Copyright © 2005 Somily et al; licensee BioMed Central Ltd.2005Somily et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Most cases of echinococcal disease (ED) acquired in Canada are thought to be due to the sylvatic form of Echinococcus granulosus, which may be more benign than ED due to either Echinococcus multilocularis or the pastoral form of E. granulosus. There are limited descriptions of the clinical course and outcome of Canadian patients with ED in the modern era.
Methods
A retrospective chart review was performed of patients hospitalized with echinococcal disease (ED) from 1991 to 2001 in Edmonton, Alberta.
Results
Forty-two cases of ED were identified of which 19 were definite, 3 probable, and 20 possible. Further analysis was limited to the 22 definite and probable cases, of which 77% were female and 41% aboriginal, with an age range of 5 to 87 years. Nine patients (40%) had pulmonary involvement and 11 (50%) hepatic involvement. One patient had an intracardiac mass presenting as a cerebrovascular event and one had a splenic cyst. Seven of the 22 patients had combined surgical resection and medical treatment, six had surgical resection of the cyst alone, four had cyst aspiration, one had medical treatment alone and four had no specific treatment. There was no mortality attributable to ED but three patients died of unrelated illnesses.
Conclusion
Echinococcal disease in northern Alberta has a marked diversity of clinical presentations, and generally has a good prognosis despite a wide variety of therapeutic interventions.
==== Body
Background
Echinococcal disease (ED) is a zoonotic infection caused by cestodes of the genus Echinococcus. Alveolar echinococcosis results from infection with the species Echinococcus multilocularis. Disease occurs in the Northern Hemisphere including Alaska, Japan, China, Russia and most countries in continental Europe [1], and results in multilocular cystic lesions. Cystic echinococcosis (CE) results from infection with the species Echinococcus granulosus. Disease occurs worldwide with the highest incidence in Mediterranean countries, Russia and the adjacent newly independent states, China, north and east Africa, Australia, and South America [1]. Echinococcus granulosus infections have been further classified as pastoral or sylvatic variants or biotypes. In the pastoral variant, sheep serve as the most common intermediate hosts while in the sylvatic variant the usual intermediate hosts are caribou or moose.
Most cases of ED acquired in Canada are thought to be due to the sylvatic form of E. granulosus. In the Northwest Territories and the northern portion of the Canadian Prairie Provinces of Alberta, Saskatchewan and Manitoba, the barren ground caribou is the most common intermediate host [2]. Moose with hydatid cysts have been found in every province of Canada west of the Maritimes and it is estimated that 50% of moose in Ontario and British Columbia are infected with the parasite [3]. Hydatid cysts have also been found in wapiti, elk, reindeer, coastal deer, white tail deer and bison. The definitive host is a canine (dog, fox, or wolf) that harbors the tapeworm after ingestion of tissue from an infected intermediate host. Twenty-eight to fifty percent of dogs in the Northwest Territories are infected with E. granulosus [2]. Intermediate hosts or humans can ingest eggs shed in the feces of the definitive host, with humans being dead-end hosts. Ingestion of E. granulosus eggs by intermediate hosts or humans is followed by the release of an oncosphere; the larva is subsequently transported via the blood stream or lymphatics to primary target organs (the lung or liver, and rarely the central nervous system, bone, or myocardium). Within the affected organ, the oncosphere matures into a vesicle, which grows expansively by concentric enlargement. A fully mature hydatid cyst, normally fluid-filled and unilocular, represents the final stage [4].
Edmonton is a referral center for northern Alberta and the Northwest Territories, thereby affording us an opportunity to see the complications of hydatid disease. We reviewed a decade of our experience with ED to gain further understanding of the epidemiological and clinical features of this disease.
Methods
A retrospective chart review was performed of all cases identified as ED using International Classification of Disease (ICD) Ninth Edition codes [ICD-9 122.0 to 122.9] from 1991 – 2001 in all Edmonton regional hospitals. Demographic, clinical and diagnostic data were collected. Aboriginal persons were defined as self-identified native Indian or Inuit people, or individuals with addresses that were on reserve land.
Definite ED was defined by evidence of ED on pathology (hooklets, brood capsules, or parasitic degenerative elements), probable ED by positive serology for Echinococcus and possible ED by a clinical presentation compatible with ED but with no corroborative histologic or serologic evidence.
Ultrasounds of all cystic lesions from patients who met the study criteria were reviewed by a radiologist who was blinded as to whether the patients had possible, probable, or definite ED. Cysts were then classified according to the proposed international classification system [5].
Results
Illustrative case
A 28-year-old aboriginal female who was five days postpartum had productive cough, pleuritic chest pain, dyspnea, fever, chills and sweating, and had received cephalexin, ciprofloxacin and cloxacillin without clinical improvement. She lived in an aboriginal community and had a history of contact with dogs and the ingestion of meat from a variety of wild animals. Laboratory results included a white blood cell count of 13.3 × 109/L, with 1.8 × 109/L (14%) eosinophils, erythrocyte sedimentation rate of 47 mm/hr, AST 32 U/L, and ALT 112 U/L. The chest radiograph showed a cavity in the left lower lobe and lingula with an air fluid level. Computed tomography of the chest showed a solitary (4.5 × 6.3 × 7.7 cm) lingual cavity with calcification, septation and a small amount of fluid with an opacity in the posterior segment of apico-posterior segment of the left upper lobe in addition to a left pleural effusion.
The patient underwent thoracotomy and left segmental pneumonectomy, left lingulectomy and decortication of the lung. Histologic examination of the lung cysts showed occasional hooklets and degenerative parasitic elements. She had a relatively uncomplicated post operative recovery and was discharged on albendazole 400 mg PO bid for 4 weeks. Echinococcal serology revealed an antibody titer to Echinococcus of 1:1024 using an Enzyme-linked immunosorbent assay (ELISA) technique.
Review of cases
Demographics
Forty-two cases of ED were identified, of which 19 were classified as definite, 3 as probable, and 20 as possible. The demographic features are shown in Table 1. Of the 22 definite and probable cases, 77% were female and 41% were aboriginal. The mean age of the definite and probable cases was 32 years (range 5 – 87 years). Possible cases were predominantly male (55%) and older (mean age 54 years). Eight of the total 42 patients had a history of eating caribou, elk or moose meat. Seven patients had a history of contact with dogs, and none had a history of contact with foxes or wolves. Unfortunately this information was not consistently recorded in the patient records.
Table 1 Demographic features of 42 patients hospitalized in Edmonton, Alberta with echinococcal disease 1991–2001
Definite/Probable (N = 22) Possible (N = 20) Total (N = 42)
Male 5 (23%) 11 (55%) 16 (38%)
Female 17 (77%) 9 (45%) 26 (62%)
Born in Canada 14 (64%) 14 (70%) 28 (67%)
Foreign born* 4 (18%) 2 (10%) 6 (14%)
Place of birth unknown 4 (18%) 4 (20%) 8 (19%)
Edmonton residence 6 (27%) 6 (30%) 12 (29%)
NWT residence 9 (41%) 8 (40%) 17 (40%)
Rural Alberta residence 7 (32%) 6 (30%) 13 (31%)
Mean age (years) 32 ± 16 54 ± 25
Pulmonary disease 9 (40%) 2 (10%) 11(26%)
Hepatic disease 12 (55 %) 16 (80%) 28 (67%)
Spleen /kidney 0 2 (10%) 2 (5%)
Cardiac 1 (5%) 0 1 (2%)
*Patients were born in Lebanon, Ukraine, Iraq, Paraguay, Poland, or Yugoslavia
Of the 22 definite and probable cases of ED, 9 (40%) had pulmonary involvement, 11 (50%) had hepatic involvement, one (5%) had splenic involvement and one (5%) had cerebral disease. The patient with cerebral disease was a 28-year-old woman with cerebral emboli, who had a pre-operative diagnosis of cardiac myxoma based on echocardiography and magnetic resonance imaging. The cardiac mass was resected and found to be an echinococcal lesion. Ten percent of the possible ED patients had pulmonary lesions, 80% had hepatic lesions, and 10% had non-hepatic intra-abdominal lesions (an immigrant from Poland with a calcified 10 cm diameter cyst in the spleen, and a man with adenocarcinoma of the rectum who had several calcified cysts in the right kidney).
The presenting symptoms of the 38 patients with pulmonary or hepatic disease are shown in Table 2. Predominant symptoms were pleuritic chest pain (82%), cough (73%), and dyspnea (56%) in the patients with pulmonary involvement, and abdominal pain (64%) in the patients with hepatic involvement.
Table 2 Symptoms in 38 patients with echinococcal disease and pulmonary or hepatic involvement
Pulmonary Hepatic
Symptom Definite/probable (N = 9) Possible (N = 2) Definite/probable (N = 11) Possible (N = 16)
Asymptomatic 0 0 2 2
Weight loss 2 2 1 2
Pleuritic chest pain 7 2 1 2
Cough 6 2 1 4
Hemoptysis 1 1 0 1
Dyspnea 5 1 2 4
Abdominal pain 0 0 9 9
Fever 3 0 4 5
Night sweats 2 0 1 2
Anorexia 4 0 1 4
Malaise 2 1 1 2
Jaundice 2 0 1 2
Chills 1 0 3 1
Vomiting 4 0 2 4
Investigations
Results of laboratory investigations in the 38 patients with pulmonary or hepatic disease are shown in Table 3. Mild anemia was common, and 37% of the 19 patients with definite and probable ED who had a differential white blood cell count performed had eosinophilia. Of the12 patients with definite and probable pulmonary or hepatic ED who had ELISA testing performed, six had positive results. Testing was performed using a titer method (four of seven were positive) or an optical density method (two of five were positive).
Table 3 Mean values and ranges of results of laboratory investigations in 38 patients with pulmonary or hepatic echinococcal disease *
Pulmonary Hepatic
Laboratory test
Normal values Definite/probable
(N = 9) Possible
(N = 2) Definite/probable
(N = 11) Possible
(N = 16)
Hemoglobin
135 – 175 g/L 120(112–133) 114.5(113–116) 123.7 (96–180) 124 (93–161)
White blood cell count
4–11 × 109 /L 10(6.4–18) 9.7(9.6–9.8) 8.7 (4.3–14.5) 8.26 (3.8 – 14)
Eosinophilia
>= 0.1 × 103 /L 2(0.1–7) 1.95(0.3–3.6) 1.8 (0 .1–6) 1.5 (0 .1 – 4.7)
ESR
0–20 mm/hr 28.5(10 – 47) 30.5(28–33) 32.6 (18–47) 23 (11–35)
AST
<40 U/L 30.4 (14–49) 25.5(22–29) 32 (12–59) 63 (14 – 432)
Alkaline Phosphatase
30–130 u/L 122.3(48 – 231) 137(123–152) 141 (62 – 461) 168 (41 – 892)
Total Bilirubin
3.4–17.1 umol/l 11(4–27) 4.5(2–7) 15.7 (4–32) 24.4 (3.3 – 98)
ELISA
(positives/ number tested) 3/7 0/2 3/5 0/5
Legend: ELISA – enzyme-linked immunosorbent assay; ESR – erythrocyte sedimentation rate
* Results are incomplete in some patients.
Results of chest radiographs for the 38 patients with pulmonary or hepatic involvement are shown in Table 4. Six of the 11 lesions in the patients with pulmonary involvement were in the left lung with either a cystic mass (present in nine of 11 cases) or nodular appearance associated with a pleural effusion in five of nine pulmonary cases. One patient had a pneumonic presentation with no cyst. Four of the 27 patients with hepatic disease had multiple liver cysts. The right lobe of the liver was involved in about two-thirds of the hepatic cases.
Table 4 Chest radiographic findings in 42 patients with echinococcal disease
Pulmonary Hepatic Other
Chest radiograph Definite/probable N = 9 Possible N = 2 Definite/ probable N = 11 Possible N = 16 Cardiac (definite) N = 1 Kidney or spleen (probable/possible) N = 3
Normal 0 0 5 7 0 1
Abnormal 9 2 6 9 1 2
Cyst 7 2 0 2 0 0
Pleural effusion 4 1 4 3 1 0
Pneumonia 5 1 0 4 0 0
Cavity 2 1 1 0 0 0
Granuloma 0 0 1 0 0 0
Atelectasis 1 0 1 3 0 1
Calcified cyst 6 2 0 1 0 0
Ultrasounds were available for review in 15 of the 30 patients with intra-abdominal disease. Results are shown in Table 5 with seven showing active cysts (CE1 or CE2) and eight showing inactive cysts (CE 4 or CE5) [5].
Table 5 Ultrasound findings in 15 patients with intra-abdominal cystic echinococcosis graded according to the proposed international classification (5)*
Definite/probable hepatic (N = 4) Possible hepatic (n = 8) Possible kidney (N = 1) Probable splenic (N = 1) Possible splenic (N = 1)
CL - - - - -
CE1 - 3 1 - -
CE2 1 2 - - -
CE3 - - - - -
CE4 3 - - 1 1
CE5 - 3 - - -
*Ultrasounds were no longer available for review on the other 15 patients with intra-abdominal disease.
Therapy and outcome
Treatment of all patients with is shown in Table 6. Of the 22 patients with definite and probable ED, seven had surgical resections of cyst(s) combined with medical treatment (albendazole or mebendazole), six had surgical removal of the cyst alone, four had cyst aspiration, one had medical treatment alone and four were not specifically treated. Of the 20 cases with probable ED, one had surgical resections of cyst combined with medical treatment, one had surgical removal of the cyst alone, one had medical treatment alone and 17 were not specifically treated. Pre-operative complications included suspected rupture of a cyst into a bronchus in two patients, suspected bacterial infection of the cyst in two patients with pulmonary cysts and one patient with a hepatic cyst, and obstructive jaundice in one patient. Major complications after surgical resection occurred only with hepatic disease and included single cases of pleural effusion, wound infection, biliary leak, and intra-abdominal bleeding resulting in a hepatic lobectomy.
Table 6 Therapy of patients with hydatid disease
Pulmonary Hepatic Other
Therapy Definite/probable N = 9 Possible N = 2 Definite/probable N = 11 Possible N= 16 Cardiac (definite) N = 1 Kidney or spleen (probable/possible) N = 3
Medical only* 0 1 1 0 0 0
Surgical resection only 3 0 2 1 1** 0
Medical therapy and surgical resection 5 0 2 1 0 0
Aspiration of cyst only 0 0 4 0 0 0
Aspiration of cyst and medical treatment 0 0 0 0 0 0
Observation only 1 1 2 14 0 3
* mebendazole or albendazole
**echinococcal cyst was resected from left ventricle
Three patients with possible hepatic ED died of illnesses that were not directly related to ED. The first patient had cryptogenic cirrhosis and chronic cholelithiasis, as well as a left lobe liver cyst. She had an open cholecystectomy for acute cholecystitis, and died post-operatively from intra-abdominal bleeding and sepsis (pathology was not performed on the hepatic cyst). The second patient with multiple myeloma and paraplegia was incidentally noted to have a hepatic cyst and died of unexplained respiratory failure. The third patient admitted with bilateral deep vein thromboses and incidentally found to have hepatic cysts died of bronchopneumonia and heart failure.
Comparison of Canadian-born to foreign-born patients
In comparing the clinical features and outcome of Canadian-born patients (presumed to have the sylvatic variant of E. granulosus) and foreign-born patients (who could have any type of Echinococcus infection), no differences were noted (data not shown).
Discussion
Over a ten-year period there were 19 definite, 3 probable, and 20 possible cases of ED identified at a northern Alberta referral center. In North America, ED was previously considered a disease of immigrants. Of 596 cases of ED reported up to 1950, only 36 occurred in persons born in Canada or the United States [3]. In our series, only 6 of 36 cases where the country of birth was known were foreign-born. Forty-one percent of the 42 cases occurred in aboriginal Canadians, who represent only 5.3% of the population of Alberta and 3.4% of the population of Canada (2001 Canadian census data). The majority of patients were middle-aged females. Previous studies have shown a predominance of males [4] and females [6], and it has been postulated that the sex predominance may be determined by which sex has more contact with the usual definitive host in that country [3].
Two-thirds of the cases of ED in our study were pulmonary, but for the definite and probable cases, half were hepatic. A previous study from northern Canada and the United States described hepatic involvement in 71% of cases and pulmonary involvement in only 7% of cases [4]. The results of a recent study of 17 patients from Manitoba and northwestern Ontario were comparable to ours with pulmonary involvement in 47% and hepatic involvement in 47% of cases [7]. However, the true rate of hepatic versus pulmonary ED cannot be determined by these studies as detection is often on the basis of radiographic studies performed for reasons unrelated to the ED. Previous reviews suggested that pulmonary cysts are more common in children and young adults while hepatic cysts are more common in older persons [2], which is compatible with our findings (data not shown).
Laboratory tests are of limited value in the diagnosis of ED. One study showed that only 29% of cases with pulmonary cysts and 15% of cases with hepatic cysts had eosinophilia [8]. In our series, 54 % of the pulmonary cases and 22% of the hepatic cases had eosinophilia. Perhaps this low rate is explained by the chronicity of the disease with eosinophilia being present only early in the course via stimulation of Th2 lymphocytes and production of IL-4 and IL-5 subsequently generating IgG1 and IgE-secreting cells, and eliciting eosinophilia [9,10]. Another possibility is that eosinophilia is only likely after cyst rupture and release of antigenic material [11].
Serology for ED has variable sensitivity depending on the type of test. The most sensitive is IgG ELISA with an 83% sensitivity [10] although this test may only become positive after cyst rupture [11]. In our study, serology was positive in 54% of cases of definite and probable ED.
There are no clinical features that can reliably distinguish E. granulosus from E. multilocularis. However, a specific E. multilocularis antigen such as the high affinity purified Em2 antigen AE metacestodes has been reported to be able to discriminate between E. granulosus and E. multilocularis in 95% of cases [12]. This test was not available at the time of this review. Molecular methods to distinguish the pastoral from the sylvatic form of E. granulosus have recently been described [13]. Daughter cysts have not been reported in sylvatic disease [14], and this form of the disease appears to be more benign than the pastoral form with a far lower incidence of complications even when cyst rupture occurs [14]. Most cases of the sylvatic form of ED are asymptomatic and are diagnosed when cysts are found during procedures performed for other reasons [14]. It seems likely that most of our Canadian-born patients had sylvatic disease as they reside in areas where exposure to sheep is rare. Most of our patients with pulmonary cysts were minimally symptomatic on presentation with cough or chest pain, and cyst rupture was recognized in only two pulmonary cases. In addition, four patients with definite and probable ED were discharged with no therapy and had no apparent sequelae.
Traditionally, surgical resection of the cyst was the treatment for ED, but simpler surgical procedures such as hydatid cyst aspiration have been shown to result in low mortality and improved outcome [6]. However, medical management of relatively asymptomatic presumed sylvatic cases can be considered, as the natural history of the sylvatic variant is to rupture without any complications or anaphylaxis [14]. It is not clear if albendazole or mebendazole are indicated or if observation is sufficient. Surgical resection or cyst aspiration is recommended in sylvatic cases only for complications such as secondary infection or pressure symptoms.
Although the outcome of our patients was generally good complications did occur. Two patients had symptomatic rupture of a pulmonary cyst and two patients had suspected bacterial infection of pulmonary cysts. Pressure or mass effects of hepatic cysts on the bile ducts, portal and hepatic veins, or on the inferior vena cava can result in cholestasis, portal hypertension, venous obstruction, or Budd-Chiari syndrome, sometimes resulting in the need for liver transplantation [15]. Obstructive jaundice was described in 18% of 132 patients in one case series [16] and occurred in one of our 12 definite and probable hepatic cases. Rupture of cysts into the biliary tree can produce biliary colic, obstructive jaundice, cholangitis or pancreatitis [17]. Cerebral disease is more common in children than in adults, and occurs in 1.5%–2.0% of patients, with the most common presentations being seizures or signs of raised intracranial pressure secondary to the cysts in the brain [7]. In our single case, cerebral disease was due to emboli from a cardiac echinococcal lesion.
Conclusion
Prevention of ED has been attempted by distributing praziquantil bait in areas where there is a high prevalence of infected foxes [18]. Unfortunately, such strategies are not practical in a vast country such as Canada. Until our understanding of the modes of transmission of ED improves, education of the public about this relatively rare disease is unlikely to be effective. Therefore, from a public health perspective, the emphasis must be on further progress in management of this disease, which will require increasing awareness on the part of physicians regarding the multitude of presentations, improved diagnostics, and further study of long-term outcomes with different treatment modalities.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
All authors contributed to study design and to revising the manuscipt. AS collected and analyzed the data and wrote the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
McManus DP Zhang W Li J Bartley PB Echinococcosis Lancet 2003 362 1295 1304 14575976 10.1016/S0140-6736(03)14573-4
Miller MJ Hydatid infection in Canada Can Med Assoc J 1953 68 423 49 13042713
Jidejian Y Hydatid disease Surgery 1953 34 155 67 13077110
Magath TB Hydatid diseases in North America Pennsylvania Med J 1941 44 813
WHO Informal Working Group International classification of ultrasound images in cystic echinococcosis for application in clinical and field epidemiological settings Acta Tropica 2003 85 253 261 12606104 10.1016/S0001-706X(02)00223-1
Khuroo MS Wani NA Javid G Khan BA Yattoo GN Shah AH Jeelani SG Percutaneous drainage compared with surgery for hepatic hydatid cysts New Engl J Med 1997 337 881 87 9302302 10.1056/NEJM199709253371303
Al Saghier M Taylor MC Gerenberg HM Canadian-acquired hydatid disease: a case report Can J Infect Dis 2001 12 178 82
Wilson JF Cystic hydatid disease in Alaska Rev Resp Dis 1967 98 1 4
Waller PF Eosinophilia in travelers Med Clin North Am 1992 76 1413 1432 1405826
Zarzosa MP Orduna Domingo A Gutierrez P Alonso P Cuervo M Prado A Bratos MA Garcia-Yuste M Ramos G Rodriguez Torres A Evaluation of six serological tests in diagnosis and postoperative control of hydatid disease patients Diagn Microbiol Infect Dis 1999 35 255 62 10668582 10.1016/S0732-8893(99)00079-6
Sreter T Szell Z Egyed Z Varga I Echinococcus multilocularis: an emerging pathogen in Hungary and Central Eastern Europe? Emerg Infect Dis 2003 9 384 6 12643838
Lanier AP Trujillo DE Schantz PM Wilson JF Gottstein B McMahon BJ Comparison of serologic test for the diagnosis and follow up of alveolar hydatid disease Am J Trop Med Hyg 1987 37 609 15 3688314
McManus DP Zhang L Castrodale D Le T Pearson M Blair D Short report: molecular genetic characteristics of an unusually severe case of hydatid disease in Alaska caused by the cervid strain of Echinococcus granulosus Am J Trop Med Hyg 2002 67 296 298 12408670
Finlay JC Speert DP Sylvatic hydatid disease in children case reports and review of endemic echinococcus granulosus infection in Canada and Alaska Pediatr Infect Dis J 1992 11 322 6 1565558
Moreno-Gonzalez E Loinaz Segurola C Garcia Urena MA Garcia Garcia I Gomez Sanz R Jimenez Romero C Gonzalez Pinto I Corral Sanchez MA Palma Carazo F Liver transplantation for Echinococcus granulosus hydatid disease Transplantation 1994 58 797 800 7940713
Safioleas M Misiakos E Manti C Katsikas D Skalkeas G Diagnostic evaluation and surgical management of hydatid diseases of the liver World J Surg 1994 18 859 865 7846909 10.1007/BF00299087
Al-Toma AA Vermeijden RJ Van De Wiel A Acute pancreatitis complicating intrabiliary rupture of liver hydatid cyst Eur J Intern Med 2004 15 65 67 15066654 10.1016/j.ejim.2003.11.008
Hegglin D Ward PI Deplazes P Antihelminthic baiting of foxes against urban contamination with Echinococcus multilocularis Emerg Infect Dis 2003 9 1266 72 14609462
| 15904502 | PMC1156894 | CC BY | 2021-01-04 16:28:14 | no | BMC Infect Dis. 2005 May 17; 5:34 | utf-8 | BMC Infect Dis | 2,005 | 10.1186/1471-2334-5-34 | oa_comm |
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-341590450210.1186/1471-2334-5-34Research ArticleEchinococcal disease in Alberta, Canada: more than a calcified opacity Somily Ali [email protected] Joan L [email protected] Lilly J [email protected] Ravi [email protected] Thomas J [email protected] Department of Pathology, University of Alberta, Edmonton AB, Canada2 Department of Pediatrics And Stollery Children's Hospital, University of Alberta, Edmonton AB, Canada3 Department of Medicine, University of Alberta, Edmonton AB, Canada4 Department of /Radiology, University of Alberta, Edmonton AB, Canada2005 17 5 2005 5 34 34 24 1 2005 17 5 2005 Copyright © 2005 Somily et al; licensee BioMed Central Ltd.2005Somily et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Most cases of echinococcal disease (ED) acquired in Canada are thought to be due to the sylvatic form of Echinococcus granulosus, which may be more benign than ED due to either Echinococcus multilocularis or the pastoral form of E. granulosus. There are limited descriptions of the clinical course and outcome of Canadian patients with ED in the modern era.
Methods
A retrospective chart review was performed of patients hospitalized with echinococcal disease (ED) from 1991 to 2001 in Edmonton, Alberta.
Results
Forty-two cases of ED were identified of which 19 were definite, 3 probable, and 20 possible. Further analysis was limited to the 22 definite and probable cases, of which 77% were female and 41% aboriginal, with an age range of 5 to 87 years. Nine patients (40%) had pulmonary involvement and 11 (50%) hepatic involvement. One patient had an intracardiac mass presenting as a cerebrovascular event and one had a splenic cyst. Seven of the 22 patients had combined surgical resection and medical treatment, six had surgical resection of the cyst alone, four had cyst aspiration, one had medical treatment alone and four had no specific treatment. There was no mortality attributable to ED but three patients died of unrelated illnesses.
Conclusion
Echinococcal disease in northern Alberta has a marked diversity of clinical presentations, and generally has a good prognosis despite a wide variety of therapeutic interventions.
==== Body
Background
Echinococcal disease (ED) is a zoonotic infection caused by cestodes of the genus Echinococcus. Alveolar echinococcosis results from infection with the species Echinococcus multilocularis. Disease occurs in the Northern Hemisphere including Alaska, Japan, China, Russia and most countries in continental Europe [1], and results in multilocular cystic lesions. Cystic echinococcosis (CE) results from infection with the species Echinococcus granulosus. Disease occurs worldwide with the highest incidence in Mediterranean countries, Russia and the adjacent newly independent states, China, north and east Africa, Australia, and South America [1]. Echinococcus granulosus infections have been further classified as pastoral or sylvatic variants or biotypes. In the pastoral variant, sheep serve as the most common intermediate hosts while in the sylvatic variant the usual intermediate hosts are caribou or moose.
Most cases of ED acquired in Canada are thought to be due to the sylvatic form of E. granulosus. In the Northwest Territories and the northern portion of the Canadian Prairie Provinces of Alberta, Saskatchewan and Manitoba, the barren ground caribou is the most common intermediate host [2]. Moose with hydatid cysts have been found in every province of Canada west of the Maritimes and it is estimated that 50% of moose in Ontario and British Columbia are infected with the parasite [3]. Hydatid cysts have also been found in wapiti, elk, reindeer, coastal deer, white tail deer and bison. The definitive host is a canine (dog, fox, or wolf) that harbors the tapeworm after ingestion of tissue from an infected intermediate host. Twenty-eight to fifty percent of dogs in the Northwest Territories are infected with E. granulosus [2]. Intermediate hosts or humans can ingest eggs shed in the feces of the definitive host, with humans being dead-end hosts. Ingestion of E. granulosus eggs by intermediate hosts or humans is followed by the release of an oncosphere; the larva is subsequently transported via the blood stream or lymphatics to primary target organs (the lung or liver, and rarely the central nervous system, bone, or myocardium). Within the affected organ, the oncosphere matures into a vesicle, which grows expansively by concentric enlargement. A fully mature hydatid cyst, normally fluid-filled and unilocular, represents the final stage [4].
Edmonton is a referral center for northern Alberta and the Northwest Territories, thereby affording us an opportunity to see the complications of hydatid disease. We reviewed a decade of our experience with ED to gain further understanding of the epidemiological and clinical features of this disease.
Methods
A retrospective chart review was performed of all cases identified as ED using International Classification of Disease (ICD) Ninth Edition codes [ICD-9 122.0 to 122.9] from 1991 – 2001 in all Edmonton regional hospitals. Demographic, clinical and diagnostic data were collected. Aboriginal persons were defined as self-identified native Indian or Inuit people, or individuals with addresses that were on reserve land.
Definite ED was defined by evidence of ED on pathology (hooklets, brood capsules, or parasitic degenerative elements), probable ED by positive serology for Echinococcus and possible ED by a clinical presentation compatible with ED but with no corroborative histologic or serologic evidence.
Ultrasounds of all cystic lesions from patients who met the study criteria were reviewed by a radiologist who was blinded as to whether the patients had possible, probable, or definite ED. Cysts were then classified according to the proposed international classification system [5].
Results
Illustrative case
A 28-year-old aboriginal female who was five days postpartum had productive cough, pleuritic chest pain, dyspnea, fever, chills and sweating, and had received cephalexin, ciprofloxacin and cloxacillin without clinical improvement. She lived in an aboriginal community and had a history of contact with dogs and the ingestion of meat from a variety of wild animals. Laboratory results included a white blood cell count of 13.3 × 109/L, with 1.8 × 109/L (14%) eosinophils, erythrocyte sedimentation rate of 47 mm/hr, AST 32 U/L, and ALT 112 U/L. The chest radiograph showed a cavity in the left lower lobe and lingula with an air fluid level. Computed tomography of the chest showed a solitary (4.5 × 6.3 × 7.7 cm) lingual cavity with calcification, septation and a small amount of fluid with an opacity in the posterior segment of apico-posterior segment of the left upper lobe in addition to a left pleural effusion.
The patient underwent thoracotomy and left segmental pneumonectomy, left lingulectomy and decortication of the lung. Histologic examination of the lung cysts showed occasional hooklets and degenerative parasitic elements. She had a relatively uncomplicated post operative recovery and was discharged on albendazole 400 mg PO bid for 4 weeks. Echinococcal serology revealed an antibody titer to Echinococcus of 1:1024 using an Enzyme-linked immunosorbent assay (ELISA) technique.
Review of cases
Demographics
Forty-two cases of ED were identified, of which 19 were classified as definite, 3 as probable, and 20 as possible. The demographic features are shown in Table 1. Of the 22 definite and probable cases, 77% were female and 41% were aboriginal. The mean age of the definite and probable cases was 32 years (range 5 – 87 years). Possible cases were predominantly male (55%) and older (mean age 54 years). Eight of the total 42 patients had a history of eating caribou, elk or moose meat. Seven patients had a history of contact with dogs, and none had a history of contact with foxes or wolves. Unfortunately this information was not consistently recorded in the patient records.
Table 1 Demographic features of 42 patients hospitalized in Edmonton, Alberta with echinococcal disease 1991–2001
Definite/Probable (N = 22) Possible (N = 20) Total (N = 42)
Male 5 (23%) 11 (55%) 16 (38%)
Female 17 (77%) 9 (45%) 26 (62%)
Born in Canada 14 (64%) 14 (70%) 28 (67%)
Foreign born* 4 (18%) 2 (10%) 6 (14%)
Place of birth unknown 4 (18%) 4 (20%) 8 (19%)
Edmonton residence 6 (27%) 6 (30%) 12 (29%)
NWT residence 9 (41%) 8 (40%) 17 (40%)
Rural Alberta residence 7 (32%) 6 (30%) 13 (31%)
Mean age (years) 32 ± 16 54 ± 25
Pulmonary disease 9 (40%) 2 (10%) 11(26%)
Hepatic disease 12 (55 %) 16 (80%) 28 (67%)
Spleen /kidney 0 2 (10%) 2 (5%)
Cardiac 1 (5%) 0 1 (2%)
*Patients were born in Lebanon, Ukraine, Iraq, Paraguay, Poland, or Yugoslavia
Of the 22 definite and probable cases of ED, 9 (40%) had pulmonary involvement, 11 (50%) had hepatic involvement, one (5%) had splenic involvement and one (5%) had cerebral disease. The patient with cerebral disease was a 28-year-old woman with cerebral emboli, who had a pre-operative diagnosis of cardiac myxoma based on echocardiography and magnetic resonance imaging. The cardiac mass was resected and found to be an echinococcal lesion. Ten percent of the possible ED patients had pulmonary lesions, 80% had hepatic lesions, and 10% had non-hepatic intra-abdominal lesions (an immigrant from Poland with a calcified 10 cm diameter cyst in the spleen, and a man with adenocarcinoma of the rectum who had several calcified cysts in the right kidney).
The presenting symptoms of the 38 patients with pulmonary or hepatic disease are shown in Table 2. Predominant symptoms were pleuritic chest pain (82%), cough (73%), and dyspnea (56%) in the patients with pulmonary involvement, and abdominal pain (64%) in the patients with hepatic involvement.
Table 2 Symptoms in 38 patients with echinococcal disease and pulmonary or hepatic involvement
Pulmonary Hepatic
Symptom Definite/probable (N = 9) Possible (N = 2) Definite/probable (N = 11) Possible (N = 16)
Asymptomatic 0 0 2 2
Weight loss 2 2 1 2
Pleuritic chest pain 7 2 1 2
Cough 6 2 1 4
Hemoptysis 1 1 0 1
Dyspnea 5 1 2 4
Abdominal pain 0 0 9 9
Fever 3 0 4 5
Night sweats 2 0 1 2
Anorexia 4 0 1 4
Malaise 2 1 1 2
Jaundice 2 0 1 2
Chills 1 0 3 1
Vomiting 4 0 2 4
Investigations
Results of laboratory investigations in the 38 patients with pulmonary or hepatic disease are shown in Table 3. Mild anemia was common, and 37% of the 19 patients with definite and probable ED who had a differential white blood cell count performed had eosinophilia. Of the12 patients with definite and probable pulmonary or hepatic ED who had ELISA testing performed, six had positive results. Testing was performed using a titer method (four of seven were positive) or an optical density method (two of five were positive).
Table 3 Mean values and ranges of results of laboratory investigations in 38 patients with pulmonary or hepatic echinococcal disease *
Pulmonary Hepatic
Laboratory test
Normal values Definite/probable
(N = 9) Possible
(N = 2) Definite/probable
(N = 11) Possible
(N = 16)
Hemoglobin
135 – 175 g/L 120(112–133) 114.5(113–116) 123.7 (96–180) 124 (93–161)
White blood cell count
4–11 × 109 /L 10(6.4–18) 9.7(9.6–9.8) 8.7 (4.3–14.5) 8.26 (3.8 – 14)
Eosinophilia
>= 0.1 × 103 /L 2(0.1–7) 1.95(0.3–3.6) 1.8 (0 .1–6) 1.5 (0 .1 – 4.7)
ESR
0–20 mm/hr 28.5(10 – 47) 30.5(28–33) 32.6 (18–47) 23 (11–35)
AST
<40 U/L 30.4 (14–49) 25.5(22–29) 32 (12–59) 63 (14 – 432)
Alkaline Phosphatase
30–130 u/L 122.3(48 – 231) 137(123–152) 141 (62 – 461) 168 (41 – 892)
Total Bilirubin
3.4–17.1 umol/l 11(4–27) 4.5(2–7) 15.7 (4–32) 24.4 (3.3 – 98)
ELISA
(positives/ number tested) 3/7 0/2 3/5 0/5
Legend: ELISA – enzyme-linked immunosorbent assay; ESR – erythrocyte sedimentation rate
* Results are incomplete in some patients.
Results of chest radiographs for the 38 patients with pulmonary or hepatic involvement are shown in Table 4. Six of the 11 lesions in the patients with pulmonary involvement were in the left lung with either a cystic mass (present in nine of 11 cases) or nodular appearance associated with a pleural effusion in five of nine pulmonary cases. One patient had a pneumonic presentation with no cyst. Four of the 27 patients with hepatic disease had multiple liver cysts. The right lobe of the liver was involved in about two-thirds of the hepatic cases.
Table 4 Chest radiographic findings in 42 patients with echinococcal disease
Pulmonary Hepatic Other
Chest radiograph Definite/probable N = 9 Possible N = 2 Definite/ probable N = 11 Possible N = 16 Cardiac (definite) N = 1 Kidney or spleen (probable/possible) N = 3
Normal 0 0 5 7 0 1
Abnormal 9 2 6 9 1 2
Cyst 7 2 0 2 0 0
Pleural effusion 4 1 4 3 1 0
Pneumonia 5 1 0 4 0 0
Cavity 2 1 1 0 0 0
Granuloma 0 0 1 0 0 0
Atelectasis 1 0 1 3 0 1
Calcified cyst 6 2 0 1 0 0
Ultrasounds were available for review in 15 of the 30 patients with intra-abdominal disease. Results are shown in Table 5 with seven showing active cysts (CE1 or CE2) and eight showing inactive cysts (CE 4 or CE5) [5].
Table 5 Ultrasound findings in 15 patients with intra-abdominal cystic echinococcosis graded according to the proposed international classification (5)*
Definite/probable hepatic (N = 4) Possible hepatic (n = 8) Possible kidney (N = 1) Probable splenic (N = 1) Possible splenic (N = 1)
CL - - - - -
CE1 - 3 1 - -
CE2 1 2 - - -
CE3 - - - - -
CE4 3 - - 1 1
CE5 - 3 - - -
*Ultrasounds were no longer available for review on the other 15 patients with intra-abdominal disease.
Therapy and outcome
Treatment of all patients with is shown in Table 6. Of the 22 patients with definite and probable ED, seven had surgical resections of cyst(s) combined with medical treatment (albendazole or mebendazole), six had surgical removal of the cyst alone, four had cyst aspiration, one had medical treatment alone and four were not specifically treated. Of the 20 cases with probable ED, one had surgical resections of cyst combined with medical treatment, one had surgical removal of the cyst alone, one had medical treatment alone and 17 were not specifically treated. Pre-operative complications included suspected rupture of a cyst into a bronchus in two patients, suspected bacterial infection of the cyst in two patients with pulmonary cysts and one patient with a hepatic cyst, and obstructive jaundice in one patient. Major complications after surgical resection occurred only with hepatic disease and included single cases of pleural effusion, wound infection, biliary leak, and intra-abdominal bleeding resulting in a hepatic lobectomy.
Table 6 Therapy of patients with hydatid disease
Pulmonary Hepatic Other
Therapy Definite/probable N = 9 Possible N = 2 Definite/probable N = 11 Possible N= 16 Cardiac (definite) N = 1 Kidney or spleen (probable/possible) N = 3
Medical only* 0 1 1 0 0 0
Surgical resection only 3 0 2 1 1** 0
Medical therapy and surgical resection 5 0 2 1 0 0
Aspiration of cyst only 0 0 4 0 0 0
Aspiration of cyst and medical treatment 0 0 0 0 0 0
Observation only 1 1 2 14 0 3
* mebendazole or albendazole
**echinococcal cyst was resected from left ventricle
Three patients with possible hepatic ED died of illnesses that were not directly related to ED. The first patient had cryptogenic cirrhosis and chronic cholelithiasis, as well as a left lobe liver cyst. She had an open cholecystectomy for acute cholecystitis, and died post-operatively from intra-abdominal bleeding and sepsis (pathology was not performed on the hepatic cyst). The second patient with multiple myeloma and paraplegia was incidentally noted to have a hepatic cyst and died of unexplained respiratory failure. The third patient admitted with bilateral deep vein thromboses and incidentally found to have hepatic cysts died of bronchopneumonia and heart failure.
Comparison of Canadian-born to foreign-born patients
In comparing the clinical features and outcome of Canadian-born patients (presumed to have the sylvatic variant of E. granulosus) and foreign-born patients (who could have any type of Echinococcus infection), no differences were noted (data not shown).
Discussion
Over a ten-year period there were 19 definite, 3 probable, and 20 possible cases of ED identified at a northern Alberta referral center. In North America, ED was previously considered a disease of immigrants. Of 596 cases of ED reported up to 1950, only 36 occurred in persons born in Canada or the United States [3]. In our series, only 6 of 36 cases where the country of birth was known were foreign-born. Forty-one percent of the 42 cases occurred in aboriginal Canadians, who represent only 5.3% of the population of Alberta and 3.4% of the population of Canada (2001 Canadian census data). The majority of patients were middle-aged females. Previous studies have shown a predominance of males [4] and females [6], and it has been postulated that the sex predominance may be determined by which sex has more contact with the usual definitive host in that country [3].
Two-thirds of the cases of ED in our study were pulmonary, but for the definite and probable cases, half were hepatic. A previous study from northern Canada and the United States described hepatic involvement in 71% of cases and pulmonary involvement in only 7% of cases [4]. The results of a recent study of 17 patients from Manitoba and northwestern Ontario were comparable to ours with pulmonary involvement in 47% and hepatic involvement in 47% of cases [7]. However, the true rate of hepatic versus pulmonary ED cannot be determined by these studies as detection is often on the basis of radiographic studies performed for reasons unrelated to the ED. Previous reviews suggested that pulmonary cysts are more common in children and young adults while hepatic cysts are more common in older persons [2], which is compatible with our findings (data not shown).
Laboratory tests are of limited value in the diagnosis of ED. One study showed that only 29% of cases with pulmonary cysts and 15% of cases with hepatic cysts had eosinophilia [8]. In our series, 54 % of the pulmonary cases and 22% of the hepatic cases had eosinophilia. Perhaps this low rate is explained by the chronicity of the disease with eosinophilia being present only early in the course via stimulation of Th2 lymphocytes and production of IL-4 and IL-5 subsequently generating IgG1 and IgE-secreting cells, and eliciting eosinophilia [9,10]. Another possibility is that eosinophilia is only likely after cyst rupture and release of antigenic material [11].
Serology for ED has variable sensitivity depending on the type of test. The most sensitive is IgG ELISA with an 83% sensitivity [10] although this test may only become positive after cyst rupture [11]. In our study, serology was positive in 54% of cases of definite and probable ED.
There are no clinical features that can reliably distinguish E. granulosus from E. multilocularis. However, a specific E. multilocularis antigen such as the high affinity purified Em2 antigen AE metacestodes has been reported to be able to discriminate between E. granulosus and E. multilocularis in 95% of cases [12]. This test was not available at the time of this review. Molecular methods to distinguish the pastoral from the sylvatic form of E. granulosus have recently been described [13]. Daughter cysts have not been reported in sylvatic disease [14], and this form of the disease appears to be more benign than the pastoral form with a far lower incidence of complications even when cyst rupture occurs [14]. Most cases of the sylvatic form of ED are asymptomatic and are diagnosed when cysts are found during procedures performed for other reasons [14]. It seems likely that most of our Canadian-born patients had sylvatic disease as they reside in areas where exposure to sheep is rare. Most of our patients with pulmonary cysts were minimally symptomatic on presentation with cough or chest pain, and cyst rupture was recognized in only two pulmonary cases. In addition, four patients with definite and probable ED were discharged with no therapy and had no apparent sequelae.
Traditionally, surgical resection of the cyst was the treatment for ED, but simpler surgical procedures such as hydatid cyst aspiration have been shown to result in low mortality and improved outcome [6]. However, medical management of relatively asymptomatic presumed sylvatic cases can be considered, as the natural history of the sylvatic variant is to rupture without any complications or anaphylaxis [14]. It is not clear if albendazole or mebendazole are indicated or if observation is sufficient. Surgical resection or cyst aspiration is recommended in sylvatic cases only for complications such as secondary infection or pressure symptoms.
Although the outcome of our patients was generally good complications did occur. Two patients had symptomatic rupture of a pulmonary cyst and two patients had suspected bacterial infection of pulmonary cysts. Pressure or mass effects of hepatic cysts on the bile ducts, portal and hepatic veins, or on the inferior vena cava can result in cholestasis, portal hypertension, venous obstruction, or Budd-Chiari syndrome, sometimes resulting in the need for liver transplantation [15]. Obstructive jaundice was described in 18% of 132 patients in one case series [16] and occurred in one of our 12 definite and probable hepatic cases. Rupture of cysts into the biliary tree can produce biliary colic, obstructive jaundice, cholangitis or pancreatitis [17]. Cerebral disease is more common in children than in adults, and occurs in 1.5%–2.0% of patients, with the most common presentations being seizures or signs of raised intracranial pressure secondary to the cysts in the brain [7]. In our single case, cerebral disease was due to emboli from a cardiac echinococcal lesion.
Conclusion
Prevention of ED has been attempted by distributing praziquantil bait in areas where there is a high prevalence of infected foxes [18]. Unfortunately, such strategies are not practical in a vast country such as Canada. Until our understanding of the modes of transmission of ED improves, education of the public about this relatively rare disease is unlikely to be effective. Therefore, from a public health perspective, the emphasis must be on further progress in management of this disease, which will require increasing awareness on the part of physicians regarding the multitude of presentations, improved diagnostics, and further study of long-term outcomes with different treatment modalities.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
All authors contributed to study design and to revising the manuscipt. AS collected and analyzed the data and wrote the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
McManus DP Zhang W Li J Bartley PB Echinococcosis Lancet 2003 362 1295 1304 14575976 10.1016/S0140-6736(03)14573-4
Miller MJ Hydatid infection in Canada Can Med Assoc J 1953 68 423 49 13042713
Jidejian Y Hydatid disease Surgery 1953 34 155 67 13077110
Magath TB Hydatid diseases in North America Pennsylvania Med J 1941 44 813
WHO Informal Working Group International classification of ultrasound images in cystic echinococcosis for application in clinical and field epidemiological settings Acta Tropica 2003 85 253 261 12606104 10.1016/S0001-706X(02)00223-1
Khuroo MS Wani NA Javid G Khan BA Yattoo GN Shah AH Jeelani SG Percutaneous drainage compared with surgery for hepatic hydatid cysts New Engl J Med 1997 337 881 87 9302302 10.1056/NEJM199709253371303
Al Saghier M Taylor MC Gerenberg HM Canadian-acquired hydatid disease: a case report Can J Infect Dis 2001 12 178 82
Wilson JF Cystic hydatid disease in Alaska Rev Resp Dis 1967 98 1 4
Waller PF Eosinophilia in travelers Med Clin North Am 1992 76 1413 1432 1405826
Zarzosa MP Orduna Domingo A Gutierrez P Alonso P Cuervo M Prado A Bratos MA Garcia-Yuste M Ramos G Rodriguez Torres A Evaluation of six serological tests in diagnosis and postoperative control of hydatid disease patients Diagn Microbiol Infect Dis 1999 35 255 62 10668582 10.1016/S0732-8893(99)00079-6
Sreter T Szell Z Egyed Z Varga I Echinococcus multilocularis: an emerging pathogen in Hungary and Central Eastern Europe? Emerg Infect Dis 2003 9 384 6 12643838
Lanier AP Trujillo DE Schantz PM Wilson JF Gottstein B McMahon BJ Comparison of serologic test for the diagnosis and follow up of alveolar hydatid disease Am J Trop Med Hyg 1987 37 609 15 3688314
McManus DP Zhang L Castrodale D Le T Pearson M Blair D Short report: molecular genetic characteristics of an unusually severe case of hydatid disease in Alaska caused by the cervid strain of Echinococcus granulosus Am J Trop Med Hyg 2002 67 296 298 12408670
Finlay JC Speert DP Sylvatic hydatid disease in children case reports and review of endemic echinococcus granulosus infection in Canada and Alaska Pediatr Infect Dis J 1992 11 322 6 1565558
Moreno-Gonzalez E Loinaz Segurola C Garcia Urena MA Garcia Garcia I Gomez Sanz R Jimenez Romero C Gonzalez Pinto I Corral Sanchez MA Palma Carazo F Liver transplantation for Echinococcus granulosus hydatid disease Transplantation 1994 58 797 800 7940713
Safioleas M Misiakos E Manti C Katsikas D Skalkeas G Diagnostic evaluation and surgical management of hydatid diseases of the liver World J Surg 1994 18 859 865 7846909 10.1007/BF00299087
Al-Toma AA Vermeijden RJ Van De Wiel A Acute pancreatitis complicating intrabiliary rupture of liver hydatid cyst Eur J Intern Med 2004 15 65 67 15066654 10.1016/j.ejim.2003.11.008
Hegglin D Ward PI Deplazes P Antihelminthic baiting of foxes against urban contamination with Echinococcus multilocularis Emerg Infect Dis 2003 9 1266 72 14609462
| 15904503 | PMC1156895 | CC BY | 2021-01-04 16:28:15 | no | BMC Infect Dis. 2005 May 17; 5:35 | latin-1 | BMC Infect Dis | 2,005 | 10.1186/1471-2334-5-35 | oa_comm |
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-361590450410.1186/1471-2334-5-36Research ArticleHepatitis E virus infection in hemodialysis patients: A seroepidemiological survey in Iran Taremi Mahnaz [email protected] Manouchehr [email protected] Latif [email protected] MohammadJavad [email protected] MohammadReza [email protected] National Research Department of Foodborne Diseases (NRDFD), The Research Center for Gastroenterology and Liver Diseases, Shaheed Beheshti University of Medical Sciences, Taleghani Hospital, Tabnak St., Evin, Tehran, Iran2 Drug Applied Research Center (DARC), Tabriz University of Medical Sciences, Tabriz, Iran3 Infectious Diseases and Tropical Medicine Research Center (IDTMRC), Shaheed Beheshti University of Medical Sciences, Tehran, Iran4 Gastroenterology and Liver Diseases Department, Shaheed Beheshti University of Medical Sciences, Tehran, Iran5 Research Center for Gastroenterology and Liver Diseases (RCGLD), Shaheed Beheshti University of Medical Sciences, Taleghani Hospital, Tabnak St., Evin, Tehran, Iran2005 17 5 2005 5 36 36 18 12 2004 17 5 2005 Copyright © 2005 Taremi et al; licensee BioMed Central Ltd.2005Taremi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The hepatitis E virus (HEV) has a global distribution and is known to have caused large waterborne epidemics of icteric hepatitis. Transmission is generally via the fecal-oral route. Some reports have suggested parenteral transmission of HEV. Anti-HEV prevalence data among chronic hemodialysis (HD) patients are few and give conflicting results.
Methods
This cross-sectional study was conducted in August of 2004. We tested 324 chronic HD patients attending three different units in the city of Tabriz, northwestern part of Iran, for anti-HEV antibody. A specific solid- phase enzyme-linked immunoassay (Diapro, Italy) was used.
Results
The overall seroprevalence of hepatitis E was 7.4 %(95% CI: 4.6%–10.6%). The prevalence rate of HBV and HCV infection were 4.6% (95% CI: 2.3%–6.9%) and 20.4% (95% CI: 16%–24.8%), respectively. No significant association was found between anti-HEV positivity and age, sex, duration of hemodialysis, positivity for hepatitis B or C virus infection markers and history of transfusion.
Conclusion
We observed high anti-HEV antibody prevalence; there was no association between HEV and blood borne infections (HBV, HCV, and HIV) in our HD patients. This is the first report concerning seroepidemiology of HEV infection in a large group of chronic HD individuals in Iran.
==== Body
Background
The causative agent of hepatitis E, hepatitis E virus (HEV) had provisionally been classified into the caliciviridae family, but in the most recent ICTV (International Committee on Taxonomy of Viruses) classification, HEV has been placed in its own taxonomic group within the class IV (+) sense RNA viruses, "Hepatitis E-like viruses" [1,2].
Hepatitis E is an important public health concern in many developing countries of Southeast and Central Asia, the Middle East, northern and western parts of Africa, and Mexico, where the occurrence of outbreaks have been reported [3-7]. Although the overall mortality rate associated with HEV infection is low, it is reportedly as high as 20% in infected pregnant women [8,9].
Transmission of HEV occurs primarily by the fecal-oral route through contaminated water supplies in developing countries. Recent studies have indicated that zoonosis is involved in the transmission of HEV, especially in industrialized countries where hepatitis E had been believed to be non-endemic [10,11]. Further, vertical transmission of HEV from infected mothers to their children has been observed [12]. Also dental treatments were suspected as risk factors for HEV contamination [13]. It has been reported that a substantial proportion of blood donors (3/200 or 1.5%) were positive for HEV RNA and viremic blood donors are potentially able to cause transfusion-associated hepatitis E in areas of high endemicity [14,15]. Patients on chronic hemodialysis have an increased risk of exposure to nosocomially-transmitted agents, and the possibility of transmission of HEV in this group of patients had been raised. Reports on the prevalence and possible nosocomial transmission of HEV in patients on hemodialysis (HD) are scant. Some authors observed a high prevalence of anti-HEV antibody in their hemodialysis patients and hypothesized that the fecal-oral route may not be the only route of transmission of HEV among their patients [16]. Other investigators, in contrast, found few anti-HEV-positive patients in their HD populations [17,18]. Moreover, in most of these reports a small number of patients were studied.
To our knowledge, the seroprevalence rate of anti- HEV among HD patients in Iran has not been examined. Iran is a country with few suspected outbreaks of HEV [19]. These findings prompted the present study to establish the prevalence of anti-HEV antibody among HD patients in the city of Tabriz in the northwestern part of Iran.
Methods
Patients
This study was carried out in Tabriz, northwestern part of Iran with about 3.5 million residents. We studied all patients (n = 324) on chronic hemodialysis treatment at three different dialysis units in Tabriz in August 2004. Routine HD techniques were performed with 3 or 4-hours treatments three times a week. Blood samples were obtained from all the patients for serological testing. Information was obtained from the medical records of the patients on the duration of hemodialysis and etiology of renal failure. In addition, a questionnaire was completed for each subject detailing the age, sex, and history of blood transfusion.
Laboratory assay
Blood samples were taken from each patient before the hemodialysis session, and the serum was separated without delay. Sera were stored at -20°C, coded and further tested at the laboratory of Research Center for Gastroenterology and Liver Diseases (Taleghani Hospital, Tehran) for anti-HEV IgG by enzyme immunometric assay (Diapro, Italy HEV EIA) according to the manufacturer's instruction. Cut-off was defined with positive and negative control sera that were included in each assay, according to manufacturer's instruction. Samples were considered positive if the optical density (OD) value was above the cut-off value and all positive samples were retested in duplicate with the same EIA assay to confirm the initial results. All patients were previously tested for HBs Ag (Diasorin, USA), anti-HCV (Third generation assay, Diasorin, USA) and anti-HIV (Biotest, Germany) by EIA assay.
Statistical analysis
Statistical analysis was performed with SPSS (version 11) statistical software (Scientific Package for Social Sciences, Chicago, IL). Descriptive statistics were reported. Continuous variables were summarized as mean ± SD, and comparisons were performed with use of t-test for unpaired samples. The chi-square test, Fisher's exact test and Mann-Whitney test were used to compare the proportions between groups. The level of significance was set at a P value of <0.05. The study was endorsed by the responsible ethics committee.
Results
Three hundred and twenty four patients were tested. There were 190 (59%) males, and 134 (41%) females; the mean age (± SD) was 53.5 ± 15.1 years. The median duration of HD treatment was 27 months (range1–261 months). The chronic renal failure of the patients was due to glomerulonephritis (n = 113), diabetic nephropathy (n = 73), nephroangiosclerosis (n = 28), polycystic kidney disease (n = 22), chronic interstitial nephritis (n = 5) and other etiologies (n = 83). No patient admitted a history of intravenous drug use.
There were twenty-four of 324 chronic hemodialysis patients showing anti-HEV antibody, the anti-HEV prevalence in our population was 7.4% (95%CI: 4.6%–10.6%). There were 15/324(4.6%) patients with persistent HBV infection (HBs Ag positive) and 66/324 (20.4%) patients were anti-HCV antibody positive. All the patients were seronegative for anti-HIV antibody. Table 1 shows the patient characteristics in relation to the presence of HEV coinfection. No statistically significant difference was seen between anti-HEV positive and negative patients with respect to age, sex, incidence of blood transfusion and other blood borne infections (HBV, HCV, and HIV).
Table 1 Characteristics of anti-HEV positive and negative patients in relation to HEV coinfection.
Characteristics Anti-HEV positive (n = 24) Anti-HEV negative (n = 300) P
Age (year)a 58.7 ± 12 53 ± 15.3 NS
Sex (M/F) 15/9 175/125 NS
Incidence of blood transfusion 3.5 ± 6.6 2.5 ± 7 NS
HBV-positive patients 0/24 (0%) 15/300 (5%) NS
Anti-HCV-positive patients 2/24 (8.3%) 64/300 (21.3%) NS
HBV, hepatitis B virus; HCV, hepatitis C virus; HD, hemodialysis; NS, not significant.
aExpressed as the mean ± SD.
The duration of hemodialysis ranged between1 and 121 months (median 26) for anti-HEV positive patients and between 1 and 261 months (median 27) for the anti-HEV negative one with no significant difference (Mann-Whitney P = 0.73).
Discussion
Studies concerning HEV epidemiology among chronic HD patients are few and give conflicting results. Difference of HEV prevalence in the general population, the criteria for inclusion of patients, and the routes of HEV transmission could partially explain the diverse results found.
HEV is usually associated with the fecal-oral route, and blood is not considered an important cause of HEV transmission as the virus dose not produces a chronic carrier state [3]. However, experimental transmission of HEV to man showed a transient phase of viremia proceeding the onset of clinical symptoms and prolonged viremia has been observed in some patients [20-22]. Therefore, a theoretical possibility of HEV transmission in endemic areas via a parenteral route has been suggested. Such a possibility is supported by the observation that anti-HEV antibody is more frequent in transfusion recipients than in the same number of non-transfused controls [23].
We found that the prevalence of anti-HEV IgG was 7.4 % in chronic HD patients attending three different units in the city of Tabriz, northwestern part of Iran. There were few reports on suspected outbreaks of HEV in Iran [19] and a population-based study indicated that the prevalence rate of IgG anti-HEV among healthy population was 9.6% [24]. The observation that anti-HEV prevalence rate in HD patients in our study (7.4%) is lower than in the previous general population-based study (9.6%)might be due to differences in the population, the size of the sample studied or situation of public health services. In Iran, large cities have better public health services, such as clinics, municipal water and sewage systems, possibly explaining the reduced risk of infection. These factors need to be further evaluated.
The detection of IgG anti-HEV in 7.4% of chronic HD provides additional evidence that HEV is endemic in Iran. Similar findings have been reported from Saudi Arabia [25], where 7.2% of HD patients were detected for IgG anti- HEV. In contrast, the prevalence of IgG anti-HEV in patients at the hemodialysis unit of non- endemic countries was significantly low [26-28]. The prevalence of HEV vary in different hemodialysis units, relatively independent of the prevalence in the general population and probably reflect the difference in the epidemiological characteristics of HEV in areas of low and high endemicity [27,29]. Apart from socioeconomic and environmental factors, intra-unit factors that may be associated with HEV transmission in some hemodialysis units needs to be evaluated further [29].
No statistically significant association was observed between HEV seropositivity and blood-borne viruses (HBV, HCV and HIV). One study has shown a striking association between hepatitis C and HEV, pointing to similar or overlapping routes of transmission [30]. Such an observation is in contrast with our finding. HEV infection, as detected by anti-HEV antibody, was associated with no risk factor in most patients.
Conclusion
This cross-sectional study showed a high prevalence of anti-HEV antibody in our HD Patients; we didn't find association between HEV and blood-borne viruses.
A careful surveillance in the general population is required and further appropriate investigations are needed to identify the exact mode of transmission and risk groups for this infection in Iran.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MT raised the original idea and design of the study and prepared the first draft of the manuscript. MK conceived the study, and participated in its design and coordination.
LG participated in the design of the study and performed the statistical analysis. MJE and MRZ revised the draft manuscript. All authors have read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This study was funded in part by the Drug Applied Research Center (DARC), Tabriz University of Medical Sciences. We wish to express our gratitude and thanks to Dr. T. Shahnazi, Dr. M. M. Arabi and Mr. K. Zolfagharian for their excellent technical assistance, Dr. S. Abediazar and Dr. H. Argani, Department of Nephrology, Tabriz University of Medical Sciences for supplying the patients' sera.
==== Refs
Pringle CR Virus taxonomy-1999. The universal system of virus taxonomy, updated to include the new proposals ratified by the International Committee on Taxonomy of Viruses during 1998 Arch Virol 1999 144 421 429 10470265 10.1007/s007050050515
Virus Taxonomy: VIIIth Report of the International Committee on Taxonomy of Viruses
Viswanathan R A review of the literature on the epidemiology of infectious hepatitis Indian J Med Res 1957 45 145 155 13438550
Khuroo MS Study of an epidemic of non-A, non-B hepatitis. Possibility of another human hepatitis virus distinct from post-transfusion non-A, non-B type Am J Med 1980 68 818 824 6770682 10.1016/0002-9343(80)90200-4
Arora NK Panda SK Nanda SK Ansari IH Joshi S Dixit R Bathla R Hepatitis E infection in children: study of an outbreak J Gastroenterol Hepatol 1999 14 572 577 10385067 10.1046/j.1440-1746.1999.01916.x
Naik SR Aggarwal R Salunke PN Mehrotra NN A large waterborne viral hepatitis E epidemic in Kanpur, India Bull World Health Organ 1992 70 597 604 1464145
Velazquez O Stetler HC Avila C Ornelas G Alvarez C Hadler SC Bradley DW Sepulveda J Epidemic transmission of enterically transmitted non-A, non-B hepatitis in Mexico, 1986–1987 JAMA 1990 263 3281 3285 2112204 10.1001/jama.263.24.3281
Balayan MS Epidemiology of hepatitis E virus infection J Viral Hepat 1997 4 155 165 9181524 10.1046/j.1365-2893.1997.00145.x
Hussaini SH. Skidmore SJ Richardson P Sherratt LM Cooper BT O'Grady JG Severe hepatitis E infection during pregnancy J Viral Hepat 1997 4 51 54 9031065 10.1046/j.1365-2893.1997.00123.x
Meng XJ Novel strains of hepatitis E virus identified from humans and other animal species: is hepatitis E a zoonosis? J Hepatol 2000 33 842 845 11097496 10.1016/S0168-8278(00)80319-0
Halbur PG Kasorndorkbua C Gilbert C Guenette D Potters MB Purcell RH Emerson SU Toth TE Meng XJ Comparative pathogenesis of infection of pigs with hepatitis E viruses recovered from a pig and a human J Clin Microbiol 2001 39 918 923 11230404 10.1128/JCM.39.3.918-923.2001
Khuroo MS Kamili S Jameel S Vertical transmission of hepatitis E virus Lancet 1995 345 1025 1026 7723501 10.1016/S0140-6736(95)90761-0
Tassopoulos NC Krawczynski K Hatzakis A Katsoulidou A Delladetsima I Koutelou MG Trichopoulos D Case report: role of hepatitis E virus in the etiology of community-acquired non-A, non-B hepatitis in Greece J Med Virol 1994 42 124 128 8158106
Arankalle VA Chobe LP Hepatitis E virus: can it be transmitted parenterally? J Viral Hepat 1999 6 161 164 10607228 10.1046/j.1365-2893.1999.00141.x
Arankalle VA Chobe LP Retrospective analysis of blood transfusion recipients: evidence for post-transfusion hepatitis E Vox Sang 2000 79 72 74 11054043 10.1159/000031215
Halfon P Ouzan D Chanas M Khiri H Feryn JM Mangin L Masseyef MF Salvadori JM High prevalence of hepatitis E virus antibody in haemodialysis patients (letter) Lancet 1994 344 746 7802786 10.1016/S0140-6736(94)92232-2
Courtney MG O'Mahoney M Albloushi S Sachithanandan S Walshe J Carmody M Donoghue J Parfrey N Shattock AG Fielding J Hepatitis E virus antibody prevalence (letter) Lancet 1994 344 1166 7802828 10.1016/S0140-6736(94)90678-5
Psichogiou M Vaindirli E Tzala E Voudiclari S Boletis J Vosnidis G Moutafis S Skoutelis G Hadjiconstantinou V Troonen H Hatzakis A Hepatitis E virus (HEV) infection in hemodialysis patients. The Multicentre Hemodialysis Cohort Study on Viral Hepatitis Nephrol Dial Transplant 1996 11 1093 1095 8671974
Ariyegan M Amini S Hepatitis E epidemic in IRAN J Med Council IRIrn 1998 15 139 143
Chauhan A Jameel S Dilawari JB Chawla YK Kaur U Ganguly NK Hepatitis E virus transmission to a volunteer Lancet 1993 341 149 150 8093748 10.1016/0140-6736(93)90008-5
Nanda SK Ansari IH Acharya SK Jameel S Panda SK Protracted viremia during acute sporadic hepatitis E virus infection Gastroenterology 1995 108 225 230 7806046
Schlauder GG Dawson GJ Mushahwar IK Ritter A Sutherland R Moaness A Kamel MA Viremia in Egyptian children with hepatitis E virus infection. (Letter) Lancet 1993 341 78 10.1016/0140-6736(93)90187-L
Mannucci PM Gringeri A Santagostino E Romano L Zanetti A Low risk of transmission of hepatitis E virus by large-pool coagulation factor concentrates. (Letter) Lancet 1994 343 597 598 7906342 10.1016/S0140-6736(94)91548-2
Zali MR Taremi M Arabi SMM Ardalan A Alizadeh AHM Ansari SH Seroprevalence of Hepatitis E in Nahavand, Iran: A population-based study proceeding of the Digestive Diseases week, 15–19 May 2004; New Orleans, USA
Ayoola EA Want MA Gadour MOEH Al-Hazmi MH Hamza MKM Hepatitis E Virus Infection in Hemodialysis Patients: A Case-Control Study in Saudi Arabia J Med Virol 2002 66 329 334 11793384 10.1002/jmv.2149
Fabrizi F Lunghi G Bacchini G Corti M Pagano A Locatelli F Hepatitis E virus infection in hemodialysis patients: a seroepidemiological survey Nephrol Dial Transplant 1997 12 133 136 9027787 10.1093/ndt/12.1.133
Staffan PES Stefan HJ Brith C Prevalence of Antibodies to Hepatitis E Virus Among Hemodialysis Patients in Sweden J Med Virol 1998 54 38 43 9443107 10.1002/(SICI)1096-9071(199801)54:1<38::AID-JMV6>3.0.CO;2-Q
Trinta KS Liberto MI Paula VS Yoshida CF Gaspar AM Hepatitis E virus infection in selected Brazilian populations Mem Inst Oswaldo Cruz 2001 96 25 29 11285473 10.1590/S0074-02762001000100004
Dalekos GN Zervou E Elisaf M Germanos N Galanakis E Bourantas K Siamopoulos KC Tsianos EV Antibodies to hepatitis E virus among several population in Greece: increased prevalence in an hemodialysis unit Transfusion 1998 38 589 595 9661693 10.1046/j.1537-2995.1998.38698326339.x
Pisanti FA Coppola A Galli C Association between hepatitis C and hepatitis E viruses in southern Italy Lancet 1994 344 746 747 7915787 10.1016/S0140-6736(94)92233-0
| 15904504 | PMC1156896 | CC BY | 2021-01-04 16:28:14 | no | BMC Infect Dis. 2005 May 17; 5:36 | utf-8 | BMC Infect Dis | 2,005 | 10.1186/1471-2334-5-36 | oa_comm |
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-381594388610.1186/1471-2334-5-38Case ReportFatal meningitis in a previously healthy young adult caused by Streptococcus pneumoniae serotype 38: an emerging serotype? Baker Carolyn I [email protected] Christopher P [email protected] Margaret AK [email protected] Lisa A [email protected] Kevin L [email protected] Department of Defense Center for Deployment Health Research, Naval Health Research Center, San Diego, California, USA2 Armed Forces Institute of Pathology, Rockville, Maryland, USA2005 19 5 2005 5 38 38 8 11 2004 19 5 2005 Copyright © 2005 Baker et al; licensee BioMed Central Ltd.2005Baker et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In December 2001, a fatal case of pneumococcal meningitis in a Marine Corps recruit was identified. As pneumococcal vaccine usage in recruit populations is being considered, an investigation was initiated into the causative serotype.
Case presentation
Traditional and molecular methods were utilized to determine the serotype of the infecting pneumococcus. The pneumococcal isolate was identified as serotype 38 (PS38), a serotype not covered by current vaccine formulations. The global significance of this serotype was explored in the medical literature, and found to be a rare but recognized cause of carriage and invasive disease.
Conclusion
The potential of PS38 to cause severe disease is documented in this report. Current literature does not support the hypothesis that this serotype is increasing in incidence. However, as we monitor the changing epidemiology of pneumococcal illness in the US in this conjugate era, PS38 might find a more prominent and concerning niche as a replacement serotype.
==== Body
Background
In December 2001, the Department of Defense Center for Deployment Health Research at the Naval Health Research Center (NHRC) was consulted regarding a case of fatal meningitis caused by Streptococcus pneumoniae (pneumococcus) in a Marine Corps recruit.
The pneumococcus is a common cause of fatal bacterial meningitis in the US [1]. A 23-valent polysaccharide vaccine and a 7-valent conjugate pneumococcal vaccine are available and potentially effective in protecting against serotype-specific invasive infections. For this reason, it is important to determine whether clinically significant pneumococcal infections are vaccine-covered serotypes, particularly in settings like military training camps where epidemic spread can occur.
Case presentation
On December 22, 2001, an 18-year-old male in his eighth week of Marine Corps basic training presented to the field medical station with headache and an episode of vomiting after physical training.
His neurological symptoms progressed over the next several hours until he became disoriented and unresponsive. He was evaluated in the local hospital emergency department where a fever of 39.1C (102.5F) and disorientation to person, place, and time were observed. Lumbar puncture performed during medical evaluation revealed an opening pressure greater than 500 mm; cerebrospinal fluid (CSF) was milky white, with a glucose concentration of 3 mg/dl and total protein of 269 mg/dl. White count in the CSF was 4444 with 91% segmented neutrophils, and microscopic evaluation of the CSF revealed Gram-positive diplococci. A presumptive diagnosis of pneumococcal meningitis was made; intravenous ceftriaxone, ampicillin, and vancomycin were initiated. Despite these measures, the patient's condition remained unimproved. He was transferred to the intensive care unit of a tertiary care hospital on ventilator support. Intravenous dexamethasone was initiated; antibiotics and supportive care were continued. He died approximately 30 hours after his initial presentation to the field medical station.
The patient had no previous history of meningitis, neurologic abnormalities or other medical problems. Prior to onset of his illness, there was no history of head trauma, and he was taking no medications. He received a meningococcal vaccination (Menomune A/C/Y/W135, Aventis Pasteur) on October 31, 2001. He had no history of receiving pneumococcal vaccination.
Laboratory evaluation
Blood and CSF cultures were positive for Streptococcus pneumoniae, sensitive to all antibiotics tested. Streptococcal culture of the throat was negative.
A pneumococcal isolate from the blood was sent to the Respiratory Disease Laboratory at NHRC for additional evaluation. Serotyping was performed using a modified version of the latex agglutination typing method [2]. Samples were tested against antisera for vaccine-covered serotypes (Statens Serum Institut, Denmark) and results confirmed using the classic Quellung reaction and multi-locus sequence typing (MLST) technique.
For MLST, chromosomal DNA was extracted using spin DNA extraction columns (Qiagen, Valencia, CA). A subsequent PCR reaction was carried out in 100-μl volumes, using primer pairs developed by Enright and Spratt [3]. Each of the primer pairs amplify an internal fragment of 7 housekeeping genes: aroE, gdh, gki, recP, spi, xpt, and ddl. Amplified DNA fragments were purified using QIAquick purification columns (Qiagen, Valencia, CA) and directly sequenced in each direction on an automated sequencer with Big Dye terminator chemistry (Applied Biosystems, Foster City, CA). Allelic matches at each loci were determined through a search engine at the pneumococcal MLST Web site .
Using the classic serotyping methods outlined, the isolate did not react with any of the 23-valent vaccine-specific typing antisera. Molecular investigation utilizing MLST demonstrated an allelic profile that matched sequence type 393. Using the MLST database, this sequence type matched a likely pneumococcal serotype 38 (PS38). The latex agglutination was repeated using antisera against this serotype, and results were confirmatory.
PS38 is not covered by the current 23-valent polysaccharide or the 7-valent conjugate pneumococcal vaccine formulations, which include pneumococcal serotypes that are considered the most common causes of invasive and antibiotic-resistant infections in the US. Given the fatal outcome in this case and the importance of vaccine formulations being appropriately targeted, the medical literature was explored to evaluate pneumococcal disease caused by PS38 in the US and outside the US. Particular emphasis was placed on temporal trends and documented cases of meningitis caused by this serotype.
PS38 in the US
Within the US, relatively few cases of PS38 disease have been reported. A few reported PS38 carriage isolates were found (3/235, 1.3%) from Alabama 1975 to 1978, but no cases of PS38 were among the 105 invasive isolates tested [4]. From 1985 to 1989 in Alabama, no PS38 was noted among 303 cases of meningitis or bacteremia, and only 2 cases were found among 228 cases of otitis media [5]. Black and colleagues reported two cases of invasive PS38 disease in children attending clinics in northern California from October 1995 to August 1998 [6]. A comprehensive analysis of pneumococcal isolates recovered from patients with invasive disease in the US in 1999 was performed by Gertz and colleagues. Among 1,168 isolates examined, only 5 were identified as PS38 (0.4%) [7]. More recently, three of 185 (1.6%) invasive isolates, and only 8/1379 (0.6%) carriage isolates were found to be PS38 among isolates from Alaska, 1998 to 2002 [8]. Given the invasive:carriage ratio of this and other analyzed studies, this report concluded that although the invasive potential of PS38 was similar to other vaccine serotypes, the carriage was so low that the significance of serious infections caused by PS38 was minimal at that time. Although not specified, meningitis was likely among these few documented invasive illnesses caused by PS38 within the US.
In a large review of the Active Bacterial Core Surveillance (ABCs) performed by Robinson and colleagues from 1995 to 1998, serotype 38 was not mentioned among the top 14 serotypes [9]. Likewise, numerous other studies failed to reveal PS38 among the reported contributing serotypes [1,10].
PS38 outside the US
Cases of invasive PS38 disease have been reported in Australia, Spain, England, and Canada in recent years [11-16]. One Australian study identified 7 (12.4%) cases of invasive PS38 disease in non-indigenous adolescents, making PS38 the second most common non-vaccine covered serotype identified in this population [11]. Of particular interest is the study of pneumococcal isolates causing local and invasive disease in Spain from January 1997 to June 2001. Unlike the previous studies, a strain of PS38 was identified that was resistant to penicillin (MIC > 0.06 mg/L) [14].
In addition to causing invasive pneumococcal infections, PS38 has been found in isolates from children with acute otitis media in the Czech Republic [17] and from nasopharyngeal swabs in a study of pneumococcal carriage in Canadian children attending daycare from 1995 to 1996 [18]. In the study of pneumococcal carriage, PS38 ranked the 14th most common serotype, with 8 out of 589 (1.36%) isolates positive for PS38.
In contrast to these small studies with recognized PS38 disease or carriage, numerous other comprehensive reports from outside the US failed to demonstrate a burden. A review of over 70 global data sets performed by Hausdorff and colleagues failed to reveal PS38 among the top 25 serotypes identified among children; however, PS38 did rank #23 out of the top 24 isolates identified from adults [19]. In a paired report, Hausdorff did not assign significance to PS38 as he evaluated the disease manifestations of the top 16 global pneumococcal serotypes [20]. Likewise, in subsequent studies, PS38 was not mentioned among isolates of importance for otitis media or for invasive disease in the literature outside the US [8,21,22]. More recently, among invasive isolates considered: 150 from studies in Oxford, England, 1995–2001 [23]; 92 from Iceland, 1992–2001 [8]; 69 from Toronto, 1995 [24]; 112 from Kenya, 1992–1996 [25], and 56 from Papau New Guinea, 1981–1987 [26], PS38 was again not among the serotypes identified.
Among reports of studies outside the US that specifically address meningitis cases caused by PS38, Table 1 summarizes reports from Africa, Asia, Europe, and South America from the late 1970s to the mid-1990s. The prevalence of PS38-associated meningitis ranged from 0.2% in South Africa to 5.6% in Southern India. Of the cases with reported antibiotic sensitivities, all were penicillin-susceptible [27,28]. In four of the five reports, PS38 meningitis was among the 23 most common serotypes reported [28-31].
Table 1 Epidemiology of PS38 Meningitis Cases Outside the US. Studies outside the US of Streptococcus pneumoniae serotype 38 (PS38) isolated from the cerebrospinal fluid (CSF) of individuals being considered for the clinical diagnosis of meningitis.
Location Time Period No. of Cases Prevalence of PS38 Antibiotic Susceptibility Patient Age Group
Cairo, Egypt [29] Late 1970s 3 2.3% NP Infants, Children, adults
South Africa [27] 1979–1986 2 0.2% Pan-sensitive NP
São Paulo, Brazil [30] 1977–1988 13 1.3% NP Children, adults, elderly
Southern India [31] 1992–1993 1 5.6% NP Children
Netherlands [28] 1994 3 2.0% Penicillin-susceptible NP
NP = Not provided
Of the PS38 disease reported outside the US, invasive pneumococcal disease was not uncommon. Additionally, even though most PS38 isolates tested were pan-sensitive to broad-spectrum antibiotics, antibiotics can be ineffective in a rapidly fulminating infection. This suggests considering primary measures such as vaccination for PS38 disease prevention. In some countries, PS38 was implicated in enough cases of invasive disease to be considered for inclusion into a 14- or 23-valent vaccine formulated for that region [28-31]. Although PS38 seems to be a sporadic cause of pneumococcal disease worldwide, temporal trends of increasing incidence were not appreciated.
Conclusion
Streptococcus pneumoniae serotype 38 has been an uncommon cause of pneumococcal disease in the US and thus, is not included in commercial vaccine formulations. The significance of this serotype might be more important outside the US. Although the invasive potential of PS38 is documented, cases are rare. Prior to this report, a definitive meningitis death in the US caused by PS38 was not found in our review of the literature; nevertheless, there is little current evidence that PS38 is gaining in incidence or importance.
As we monitor the changing epidemiology of pneumococcal illness in the US in this conjugate era, PS38, with its documented invasive potential, might find a more prominent and concerning niche as a replacement serotype. Continued laboratory-based surveillance with a focus on emerging strains and the changing epidemiology is critical for optimal development of future prevention efforts in these dynamic times.
List of abbreviations
NHRC – Naval Health Research Center
CSF – cerebrospinal fluid
MLST – multi-locus sequence typing
PS38 – pneumococcus serotype 38
ABC – Active Bacterial Core Surveillance
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
CIB performed all coordination of sample processing, researched for the case report, and wrote the initial draft. CPB was responsible for all laboratory processing. MAKR provided oversight and expert consultation. LAP provided information on the clinical course of the patient during the hospital stay. KLR conceived of the case report, guided the laboratory processing and writing of the manuscript, and coordinated all aspects.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We thank Dr. Lorraine Nadkarni and Dr. Dipak Nadkarni (Naval Hospital Beaufort), and Mr. Frank Stroncheck and Mr. Alexandre Dubovtsev (Henry M. Jackson Foundation) for assistance in obtaining information, contacts, and specimen samples. We also thank Ms. Rebecca Gunnill of the DoD Center for Deployment Health Research for assistance in editing and manuscript preparation.
This represents NHRC report 04-11, supported by the U.S. Department of Defense under Work Unit Number 61102A.M0101.BAX-6609. The views expressed in this article are those of the authors and do not reflect the official policy or position of the Department of the Navy, Department of Defense, or the U.S. Government. Approved for public release; distribution unlimited. This work has been conducted in compliance with all applicable federal regulations governing the protection of human subjects in research under protocol #NHRC.2001.0007.
==== Refs
Stanek RJ Mufson MA A 20-year epidemiological study of pneumococcal meningitis Clin Infect Dis 1999 28 1265 72 10451164
Facklam RR Washington JA Balows A, Hausler WJ, Herman KL, Isenberg HD, Shandomy HJ From Streptococcus and related catalase-negative Gram-positive cocci Manual of Clinical Microbiology 1991 5 Washington, DC: American Society for Microbiology 238 57
Enright M Spratt BG A multi-locus sequence typing scheme for Streptococcus pneumoniae : identification of clones associated with serious invasive disease Microbiology 1998 144 3049 60 9846740
Gray BM Converse GM Dillon HC Serotypes of Streptococcus pneumoniae causing disease J Infect Dis 1979 140 979 83 44310
Orange M Gray BM Pneumococcal serotypes causing disease in children The Ped Infect Dis J 1993 12 244 6
Black S Shinefield H Fireman B Lewis E Ray P Hansen JR Elvin L Ensor KM Hackell J Siber G Malinoski F Madore D Chang I Kohberger R Watson W Austrian R Edwards K Efficacy, safety, and immunogenicity of heptavalent pneumococcal conjugate vaccine in children Pediatr Infec Dis J 2000 19 187 195 10749457 10.1097/00006454-200003000-00003
Gertz RE McEllistrem C Boxrud DJ Li Z Sakota V Thompson TA Facklam RR Besser JM Harrison LH Whitney CG Beall B Clonal distribution of Invasive pneumococcal isolates from children and selected adults in the United States prior to 7-valent conjugate vaccine introduction J Clin Microbiol 2003 41 4194 4216 12958247 10.1128/JCM.41.9.4194-4216.2003
Brueggemann AB Peto TEA Crook DW Butler JC Kristinsson KG Spratt BG Temporal and geographic stability of the serogroup-specific invasive disease potential of Streptococcus pneumoniae in children J of Infect Dis 2004 190 1203 11 15346329 10.1086/423820
Robinson KA Baughman W Rothrock G Barrett NL Pass M Lexau C Damaske B Stefonek K Barnes B Patterson J Zell ER Schuchat A Whitney CG Epidemiology of invasive Streptococcus pneumoniae infections in the United States, 1995–1998 JAMA 2001 285 1729 1735 11277827 10.1001/jama.285.13.1729
Joloba ML Windau A Bajaksouzian S Appelbaum PC Hausdorff WP Jacobs MR Pneumococcal conjugate vaccine serotypes and the antimicrobial susceptibility of such isolates in children with otitis media Clin Infect Dis 2001 33 1489 94 11588694 10.1086/323027
Watson M Bayley K Bell JM Gilbert GL Hogg G Keil AD Krause V Murphy D Roche P Smith HV Stewart MG Stylianopoulos J Turnidge J Laboratory surveillance of invasive pneumococcal disease in Australia in 2001 to 2002 – implications for vaccine serotype coverage Commun Dis Intell 2003 27 478 487 15508501
Communicable Disease Unit Invasive Pneumococcal Disease in Children Aged Under Five Years in Queensland, in 2002 Queensland Health 2003
Hills SL Hanna JN Denise Murphy D Invasive pneumococcal disease in North Queensland, 2001 Commun Dis Intell 2002 26 520 4 12549517
Fenoll A Asensio G Jado I Berrón S Camacho MT Ortega M Casal J Antimicrobial susceptibility and pneumococcal serotypes J Antimicrob Chemother 2002 13 19 12556429 10.1093/jac/dkf509
CDSC Invasive pneumococcal infection England and Wales 1999 Commun Dis Rep 2001 11 CDR Wkly 6 12
National Centre for Streptococcus Annual Report for April 1, 2002 to March 31, 2003 Retrieved October 2, 2004
Prymula R Motlova J Kriz P Comparison of Streptococcus pneumoniae serotypes causing acute otitis media & invasive disease in young children in the Czech Republic Indian J Med Res 2004 119 168 170 15232187
Kellner JD Ford-Jones EL Members of the Toronto Child Care Centre Study Group Streptococcus pneumoniae carriage in children attending 59 Canadian child care centers Arch Pediatr Adolesc Med 1999 153 495 502 10323630
Hausdorff WP Bryant J Paradiso PR Siber GR Which pneumococcal serogroups cause the most invasive disease: implications for conjugate vaccine formulation and use, part I Clin Infect Dis 2000 30 100 21 10619740 10.1086/313608
Hausdorff WP Bryant J Kloek C Paradiso PR Siber GR The contribution of specific pneumococcal serogroups to different disease manifestations: implications for conjugate vaccine formulation and use, part II Clin Infect Dis 2000 30 122 40 10619741 10.1086/313609
Hausdorff WP Yothers G Dagan R Kilpi T Pelton SI Cohen R Jacobs MR Kaplan SL Levy C Lopez EL Mason EO JrSyriopoulou V Wynne B Bryant J Multinational study of pneumococcal serotypes causing acute otitis media in children Pediatr Infect Dis 2002 21 1008 16 10.1097/00006454-200211000-00007
Hausdorff WB Invasive pneumococcal disease in children: geographic and temporal variations in incidence and serotype distribution Eur J Pediatr 2002 161 S135 S139 12494260 10.1007/s00431-002-1066-x
Brueggemann AB Griffiths DT Meats E Peto T Crook DW Spratt BG Clonal Relationships between invasive and carriage Streptococcus pneumoniae and serotype- and clone-specific differences in invasive disease potential J Infect Dis 2003 187 1424 32 12717624 10.1086/374624
Kellner JD McGeer A Cetron MS Low DE Butler JC Matlow A Talbot J Ford-Jones EL The use of Streptooccus pneumoniae nasopharyngeal isolates from healthy children to predict features of invasive disease Pediatr Infect Dis J 1998 17 279 86 9576381 10.1097/00006454-199804000-00004
Scott JAG Hall AJ Hannington A Serotype distribution and prevalence of resistance to bezylpenicillin in tree representative populations of Streptococcus pneumoniae isolates from the coast of Kenya Clin Infect Dis 1998 27 1442 50 9868658
Smith T Lehmann D Montgomery J Gratten M Riley ID Alpers MP Acquisition and invasiveness of different serotypes of Streptococcus pneumoniae in young children Epidemiol Infect 1993 111 27 39 8348930
Klugman KP Koornhof HJ Drug resistance patterns and serogroups or serotypes of pneumococcal isolates from cerebrospinal fluid or blood, 1979–1986 J Infect Dis 1988 158 956 64 3183429
Overweg K Bogaert K Sluijter M Yother J Dankert J Groot R Hermans PW Genetic relatedness within serotypes of penicillin-susceptible Streptococcus pneumoniae isolates J Clin Microbiol 2000 38 4548 53 11101594
Guirguis NI Helmy MF Mohamed MR Ali RH A suggested vaccine formulation for the control of pneumococcal meningitis in Egypt J Egypt Public Health Assoc 1990 65 291 303 2133903
Taunay AE Austrian R Landgraf IM Vieira MFP Melles CEA Sorotipos de Streptococcus pneumoniae isolados de líquido cefalorraquidiano no período de 1977–1988 na cidade de São Paulo, Brasil Rev Inst Med Trop Sao Paulo 1990 32 11 5 2259828
Sridharan G John TJ Lalitha MK Harrison LH Steinhoff MC Serotypes of Streptococcus pneumoniae causing meningitis in southern India: use of new direct latex agglutination antigen detection tests in cerebrospinal fluid Diagn Microbiol Infect Dis 1994 18 211 4 7924216 10.1016/0732-8893(94)90022-1
| 15943886 | PMC1156897 | CC BY | 2021-01-04 16:28:14 | no | BMC Infect Dis. 2005 May 19; 5:38 | utf-8 | BMC Infect Dis | 2,005 | 10.1186/1471-2334-5-38 | oa_comm |
==== Front
BMC Mol BiolBMC Molecular Biology1471-2199BioMed Central London 1471-2199-6-121589289210.1186/1471-2199-6-12Research ArticleCharacterization of TRZ1, a yeast homolog of the human candidate prostate cancer susceptibility gene ELAC2 encoding tRNase Z Chen Yang [email protected] Audrey [email protected] Christina [email protected] Yuan [email protected] Donna [email protected] Sean V [email protected] Myriad Genetics, Inc. 320 Wakara Way, Salt Lake City, UT 84108, USA2 International Agency for Research on Cancer, World Health Organization, 150 Cours Albert-Thomas, 69372 Lyon Cedex 08, France2005 13 5 2005 6 12 12 12 11 2004 13 5 2005 Copyright © 2005 Chen et al; licensee BioMed Central Ltd.2005Chen et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In humans, mutation of ELAC2 is associated with an increased risk of prostate cancer. ELAC2 has been shown to have tRNase Z activity and is associated with the γ-tubulin complex.
Results
In this work, we show that the yeast homolog of ELAC2, encoded by TRZ1 (tRNase Z 1), is involved genetically in RNA processing. The temperature sensitivity of a trz1 mutant can be rescued by multiple copies of REX2, which encodes a protein with RNA 3' processing activity, suggesting a role of Trz1p in RNA processing in vivo. Trz1p has two putative nucleotide triphosphate-binding motifs (P-loop) and a conserved histidine motif. The histidine motif and the putative nucleotide binding motif at the C-domain are important for Trz1p function because mutant proteins bearing changes to the critical residues in these motifs are unable to rescue deletion of TRZ1. The growth defect exhibited by trz1 yeast is not complemented by the heterologous ELAC2, suggesting that Trz1p may have additional functions in yeast.
Conclusion
Our results provide genetic evidence that prostate cancer susceptibility gene ELAC2 may be involved in RNA processing, especially rRNA processing and mitochondrial function.
==== Body
Background
With a cumulative incidence by age 75 of 14% among Caucasian-Americans and 25% among African-Americans, prostate cancer has become the most common non-cutaneous malignancy in the US [1]. Despite its high incidence rate, there is little known about the genetics and etiology of prostate cancer. We recently identified a candidate prostate cancer susceptibility gene through linkage analysis and positional cloning [2]. This gene, ELAC2, is located on chromosome 17p and encodes a protein of 826 amino acids with tRNase Z activity (tRNA 3' endoribonuclease activity) [3]. ELAC2 belongs to an incompletely characterized family of proteins that is conserved among eukaryotes, archaea and bacteria. ELAC2 is ubiquitously expressed in all tissues with the highest expression levels found in testes. Elac2p is detected in several cancer cell lines and binds the γ-tubulin complex [4]. At present the role that sequence variants in ELAC2 may play in the genesis of prostate cancer is unclear.
TRZ1 (tRNase Z 1, YKR079c), the yeast ortholog of ELAC2, encodes an essential protein of 838 amino acids. Like ELAC2 and its other eukaryotic orthologs, TRZ1 can be divided roughly into amino- and carboxy-halves, or domains, which are structurally homologous [2]. The carboxy domain possesses tRNase Z activity; the amino domain is enzymatically inactive but may modulate the specificity of carboxy domain. The most strongly conserved sequence element is the histidine motif, located near the front of the carboxy domain. In the amino domain, the structural homolog of the histidine motif is easily recognizable; however, most of the key zinc coordinating histidines have changed to some other polar amino acid, which is entirely consistent with the observation that this domain lacks tRNase Z activity. PROSITE scans revealed that the vertebrate ELAC2s contain a sequence element near the center of their amino domain that matches the nucleotide binding P-loop motif [2]. Although not exactly matching the canonical P-loop sequence, this sequence element is well conserved in TRZ1 and its structural homolog is well conserved near the center of the carboxy domain of all gene family members. It has recently been shown that both of the two ELAC family members (tRNase ZL and tRNase ZS) from the Animals, Plants, Fungi, Bacteria, and Archaea possess tRNase Z activity [3,5-9]. The amino acid and enzymatic conservation across ELAC family members suggests strongly that they all share some basic cellular functions. Since Saccharomyces cerevisiae is a well-characterized genetic system, we initiated a genetic analysis of the functions of the yeast member of this gene family, TRZ1, in order to better understand those of ELAC2.
RNA processing, one of the post-transcriptional controls on gene expression, plays an important role in regulating cell growth and fate. Pre-mRNA processing is a multi-step process coupled to transcription through the involvement of the unique C-terminal domain (CTD) of the large subunit of RNA polymerase II. The processing involves RNA binding proteins, snRNAs (small nuclear RNAs), endo- and exo-nucleases, and other specific factors. The function of these proteins is subject to regulation by different signaling pathways, according to the developmental stage of the cell [10].
Transfer RNAs (tRNAs) undergo extensive post-transcriptional modifications, including: 5' processing; 3' processing; nucleotide modification; and, in the case of about 20% of yeast tRNA, intron removal in the nucleus before export to the cytosol. RNase P (ribonuclease P), which consists of a RNA subunit and nine proteins, is responsible for 5' tRNA processing (reviewed in [11]). Two proteins, Trz1p [3] and Lhp1p [12], have been shown to be involved in 3' tRNA maturation in yeast; Trz1p has tRNase Z activity while Lhp1p stabilizes pre-tRNAs as an RNA chaperone [3,12]. It is known that there are more than 55 proteins involved in tRNA splicing, modification, degradation and export.
Ribosomal RNAs (rRNAs) are transcribed as large precursors by RNA polymerases I and III in nucleoli [13]. These RNAs also undergo processing, including 5' and 3' cleavages, removal of introns, and nucleotide modifications through the action of a series of RNA/protein complexes. The mature rRNAs complex with about 80 proteins to form ribosomes. The RNase MRP (ribonuclease MRP) complex is responsible for the majority of 5' maturation steps and the exosome complex is responsible for 3' rRNA maturation. Many of the modifications and splicing events require small nucleolar RNAs (snoRNAs) and other specific protein factors. Notably, some proteins are shared between the tRNA and rRNA maturation complexes (reviewed in [14]).
In this paper we show that, although Trz1p is required for growth, yeast are largely insensitive to the absolute level of its expression. ELAC2 cannot functionally replace Trz1p, indicating that Trz1p has additional or specific functions. In the carboxy domain of the protein, the conserved histidine motif and the putative P-loop are important for Trz1p biological function because mutations of the conserved residues in these motifs render the protein inactive in yeast. Through a high copy suppressor screen, we found that TRZ1 genetically interacts with REX2, suggesting a role for TRZ1 in RNA processing and mitochondrial maintenance.
Results
Effect of altered TRZ1 expression on yeast cell growth
We generated three yeast strain types in order to study the consequences of altered TRZ1 expression in yeast. In the first (YL09-01 and YL09-02), TRZ1 is expressed from a Gal1 promoter on a low-copy plasmid in a wild type TRZ1 background. In the second (YL10-04 and YL10-05), the GAL1 promoter is inserted upstream of TRZ1 between the translational start site and its endogenous 5'UTR at the chromosomal locus (no deletion of the endogenous locus). In the third strain type (YL10-01, YL10-02 and YL10-03), endogenous TRZ1 is deleted and the gene is expressed from a GAL1 promoter on a low-copy plasmid. The latter two strain types were made to investigate the effect of up- and down- regulation of TRZ1 expression on growth. The genotypes of all strains used in this study are listed in Table 1. Also listed in Table 1 is the ρ +/ρ - status of each strain, assayed as the ability of the strain to grow on media containing only non-fermentable carbon sources. The significance of the ρ +/ρ - status, which represents the wild type (ρ+) or impaired (ρ-) (petite mutant) mitochondrial respiratory function, will become clear below.
As shown in Table 2, when TRZ1 is over-expressed in a wild type background (YL09-01), the yeast doubling time is 3.7 hours (average of 2 assays, 3.85 hours in one experiment and 3.55 hours in another experiment), not significantly longer than that of YPH499, the parental yeast control (3.4 hours; average of 2 assays, 3.11 hours in one experiment and 3.75 hours in another experiment) or that of wild type yeast that over-express β-galactosidase (3.0 hours, one assay only). Therefore, over-expression of TRZ1 in wild type yeast does not significantly slow yeast growth. Overexpression of FLAG-TRZ1 in YL09-02 was verified by Western blot with anti-FLAG antibody. As shown in Figure 1a, FLAG -TRZ1 is rapidly induced in YL09-02 grown on galactose-containing medium. Notably, both YL09-01 and YL09-02 are ρ+.
We then investigated the consequence of over-expression of TRZ1 from the chromosomal locus. In YL10-04 a GAL1 promoter is inserted immediately upstream of the translational start codon of FLAG-TRZ1. As shown in Figure 1b, FLAG-TRZ1 is promptly induced after addition of galactose in the YL10-04 strain. Liquid growth assays in a variety of different media were done to determine the doubling time of strains expressing varying levels of Trz1p (Table 2). No growth was observed for YL10-04 and YL10-05 in galactose medium, respectively, compared to 3.75 hr doubling time forYPH499. However, these GAL1 promoter insertion strains also grew more slowly in non-inducing medium. The doubling time for YL10-04 and YL10-05 in glucose medium is 3 and 2 times longer than that of parental YPH499 cells (Table 2). We experienced difficulties when trying to isolate Gal1 promoter insertion strains and found that the isolated strains were petite mutants; therefore, we decided not to study these strains further.
We observed the growth characteristics of trz1 strains with TRZ1 expressed from a GAL1 promoter on a low-copy plasmid (YL10-01, YL10-02 and YL10-03). As shown in Table 2, when TRZ1 expression is repressed in glucose medium, YL10-01 grows marginally slower (doubling time = 2.95 hours) than YPH500, the isogenic wild type yeast (doubling time = 2.27 hours). The same is true for YL10-02 and YL10-03 with respect to YPH499, their isogenic wild type strain. Thus, although TRZ1 is an essential gene, even low levels of its transcript are sufficient for near normal growth. When TRZ1 expression is induced in galactose medium, however, YL10-01 and YL10-02 fail to grow, while YL10-03 grows at about the same rate as wild type yeast. This is intriguing, particularly when comparing YL10-02 and YL10-03, because the strains share the same genetic background. They differ in two ways: YL10-02 is ρ- (petite phenotype) while YL10-03 is ρ+ (table 1); and YL10-02 expresses untagged TRZ1 while YL10-03 expresses TRZ1 tagged with a FLAG epitope. The phenotypic discrepancy between YL10-02 and YL10-03 is not likely due to a lack of expression of Trz1p or lack of functional Trz1p in YL10-03, because we could detect FLAG•TRZ1p in YL10-03 (data not shown) and p414/Gal1p-FLAG•TRZ1 could functionally complement the trz1 deletion (Table 4). The difference between the strains is more likely due to their ρ status.
We studied these strains further by testing the effect of different carbon sources on their growth. Table 3 summarizes these results. When TRZ1 expression from a Gal1p-driven TRZ1 plasmid is suppressed on SC-glucose, wild type yeast (YL09-01) and trz1 yeast (YL10-03) grow at a rate similar to that of YPH499 at room temperature, 30°C and 37°C. We could not detect any significant difference in growth rate among YPH499, YL09-01 and YL10-03 on sucrose, raffinose or galactose at any temperature tested (Table 3). This result agrees with the results from liquid growth assays in that over-expression of TRZ1 in these two strains (YL09-01 and YL10-03) does not result in any discernable change in growth rate. When spotted onto glucose media and incubated at room temperature, YL10-01 and YL10-02 grow at a rate equivalent to that of YPH499. As the incubation temperature is increased, however, YL10-01 and YL10-02 grow significantly slower than YPH499. When spotted on sucrose media, these strains grow slower than YPH499 at all tested temperatures, with the most significant difference in doubling time occurring at 37°C. YL10-01 and YL10-02 barely grow on raffinose and do not grow at all on galactose. As argued above, the phenotypic discrepancies among the strains is not likely due to differences in expression of functional TRZ1p. Instead, the different phenotypes resulting from galactose-induced expression of Trz1p is more likely attributable to the ρ+/ρ- status of the strains, consistent with the known inability of ρ- cells to grow on galactose-containing media [15]. Thus, the ρ- stains (YL10-01, YL10-02, YL10-04 and YL10-05) show sugar and temperature sensitivity; the ρ+ strains (YL09-01 and YL10-03) do not. In summary, yeast are largely insensitive to the level of Trz1p, provided there is some minimal level of the protein; the apparent sensitivity of some strains to galactose-induced Trz1p expression is attributable to the non-specific galactose-sensitivity of ρ- yeast.
ELAC2 does not functionally substitute for TRZ1; motifs required for Trz1p function
Since TRZ1 is the yeast homolog of ELAC2, we tested the ability of ELAC2 to functionally complement deletion of TRZ1 in yeast. YL03-47 was transformed with an array of ELAC2 constructs (full length ELAC2, FLAG tagged ELAC2, N-domain ELAC2, C-domain ELAC2, and four full length ELAC2 variants found in prostate patients) (Table 4). Transformants were plated onto FOA to select against the TRZ1 expression plasmid and to determine if any construct could complement trz1. As shown in Table 4, neither wild type nor prostate cancer-associated ELAC2 mutants could substitute for wild type TRZ1 function in YL03-47. The human paralog of ELAC2, ELAC1, likewise failed to rescue trz1 (Table 4).
We next incorporated three specific mutations into the histidine motif region of TRZ1: (i) M535T, which corresponds to the missense substitution A541T in ELAC2 that is associated with human prostate cancer; (ii) H542ter, which corresponds to the frameshift 1641insG associated with human prostate cancer; and (iii) G548R, which corresponds to a ts mutant, G256R, that was isolated in the homologous yeast gene PSO2 [16] (Figure 2) (trz1 M535T, trz1 H542Ter, trz1 G548R; Table 4). Notably, in no case did the fusion of two copies of the FLAG epitope to TRZ1 alleles influence results (Table 4). The missense mutant TRZ1 allele (trz1 M535T) complements trz1 as well as the wild type gene. However, neither trz1 G548R in which a conserved glycine in the histidine motif is altered, nor the mutant TRZ1 truncated after His542 in the same motif (trz1 H542Ter) could complement trz1 in yeast.
Motif searches identified a putative P-loop (A/GxxxxGKS/T) at amino acids 276–283 of human ELAC2 (AxxxxGKS). This sequence element is located near the center of the N-domain and is clearly conserved among the eukaryotic orthologs. The corresponding sequence in Trz1p (AxxxxGQT, residues 224–231) falls within the range of variation for the structural P-loop recently explored by Brakoulias et al [17]. In the homology between the N- and C-domains of the protein, this candidate P-loop corresponds to a strongly conserved element near the middle of the C-domain (residues 693–700 of Trz1p), which also has features reminiscent of the structural P-loop [17]. For simplicity, we will refer to these as the first and second putative P-loops. In so doing, however, we recognize that they may be examples of some other conserved structural element. We mutated the key residues in both putative P-loops individually and tested the mutants' abilities to complement trz1. The first putative P-loop is dispensable for survival because mutants trz1 G229D and trz1 Q230A, which should destroy its function if it is a P-loop [18,19] (and are far outside the cross-species range of variation of the sequence element if it is not a P-loop), can functionally substitute for the essential function of Trz1p. However, the second putative P-loop is required for yeast survival because mutants here (trz1 G698D and trz1 D699A) cannot complement TRZ1 deletion. Sequence analysis shows that TRZ1 may be the result of ancient tandem duplication/fusion that produced a gene with homologous amino and carboxyl halves. We therefore tested whether either the N- or C- domain of TRZ1 alone is capable of rescuing trz1. Neither half could rescue (FLAG.TRZ1.C-term and FLAG.TRZ1.N-term), even when fused to the complementary domain of ELAC2 (FLAG.TRZ1-ELAC2 and ELAC2-TRZ1) (Table 4).
Screen for and characterization of temperature sensitive (ts) TRZ1 mutants
We screened for ts mutants of Trz1p with PCR-based random mutagenesis and bacterial mutator approaches; neither of these attempts identified ts mutants. To circumvent technical difficulties, we used a targeted approach instead. We designed degenerate oligos targeting sequence encoding amino acids 535, 536, 537, 538, 541 and 548 in the Trz1p histidine motif (Figure 2). These amino acids were selected for mutagenesis because, with the exception of residue 548, they are less conserved across species than the canonical residues of the histidine motif; therefore, mutation may be less likely to completely inactivate protein function (Fig. 2). G548 was targeted because, as mentioned before, a ts mutant had been identified at the corresponding amino acid in S. cerevisiae PSO2. We generated 6 mini-libraries, each of which targets one of these residues in YL03-47, and isolated the clones that appeared at 30°C but that grew slowly or not at all at 37°C. To confirm the ts phenotype plasmids from these isolates were purified and used to transform naïve YL03-47. The screen identified two ts mutants of Trz1p, Y537L and L538K (Figure 3). We also found mutations at position 535, 536, 537, 538, 541 and 548 that are lethal at both 30°C and 37°C. Observation of lethal missense substitutions at M535 is intrinsically interesting because it is consistent with the idea that the corresponding residue A541 of human ELAC2 is under some functional constraint. More than 60% of the lethal mutations were concentrated at G548, indicating that this residue is more functionally constrained than the corresponding G256 of S. cerevisiae PSO2 (data not shown).
We integrated TRZ1, trz1-537 (encodes Trz1p Y537L) and trz1-538 (encodes Trz1p L538K) into the LEU2 locus in a trz1 background to yield YL13-01, YL13-02, and YL13-03, respectively. As shown in Figure 4, the ts mutants grew more slowly on glucose medium than did the controls at all temperatures tested. The inhibition of growth was more obvious at higher temperature. Growth of the ts mutants was further slowed on sucrose. When plated on raffinose, ts mutants grow only at 30°C. YL13-02 and YL13-03 are also petite mutants (i.e., ρ-) while YL13-01 is not (Table 1). The ability of YL13-01 but not YL13-02 or YL13-03 to grow on galactose can therefore be explained in terms of the known inability of ρ- yeast to grow on galactose [15]. Table 5 summarizes the doubling time for each ts mutant at different temperatures in SC-Leu medium containing glucose. At 30°C YL13-02 and YL13-03 grow at about the same rate in liquid medium containing glucose as does YL13-01. However, at 37°C the ts strains grow significantly more slowly in glucose (2–3 fold increase in doubling time) than does YL13-01.
We integrated the trz1-537 allele at the TRZ1 locus to yield YL12-01. YL12-01 exhibits a slow-growth phenotype like that of the LEU2 integrants (YL13-02 and YL13-03). YL12-01 grows slower on glucose media compared with the isogenic wild type YPH499 at all temperatures tested and this growth difference increases at 37°C (data not shown). The trz1-537 integrants exhibit severe ts growth on sucrose containing medium and they do not grow on raffinose and galactose containing medium (data not shown). The doubling time of these mutants in glucose-supplemented liquid medium is in agreement with results from the plate assay (Table 5). To verify that the growth phenotype and sugar sensitivity observed with trz1-537 integrants do not result from a cloning artifact, we did the following. YPH499 and YL12-01 were transformed with empty vector (p416) or p416/TRZ1 and plated onto SC-Ura containing glucose. Once grown, colonies were replica-plated onto SC-Ura supplemented with glucose, sucrose, raffinose or galactose. Providing a wild type TRZ1 on a CEN vector restored the growth of YL12-01 to the level of YPH499 on all media tested (Figure 5). YL12-01 transformed with an empty vector grew more slowly than YPH499 on all media tested, with the decrease in growth rate in the order of glucose, sucrose, raffinose and galactose media, in agreement with the previous plate assay. In addition, we checked the ρ status for these transformants. Providing a copy of wild type TRZ1 on a CEN vector (YL12-02) converted YL12-01 from ρ- to ρ+ (Table 1), indicating that the petite phenotype of YL12-01 may not be due to the loss of mitochondrial DNA. Instead, it may result from the interference of important mitochondrial functions, suggesting that Trz1p also plays a role inside mitochondria. However, normal growth on a non-fermentable carbon source was not restored in diploid strains derived from crosses between ts mutant strains and a haploid ρ0 tester strain (Table 1 and data not shown). Given the fact that the tester strain should have a wild type copy of TRZ1 and ts mutant haploid strains should have mitochondrial DNA, we speculate that ts alleles of TRZ1 may have pleiotropic effects on certain important activities in mitochondria. The effect on some of these activities may be dominant, and mating with a wild type haploid strain lacking mitochondrial DNA may not fully restore the mitochondrial functions.
To determine if the trz1-537 allele causes any cell cycle arrest phenotype, we analyzed YL12-01, with YPH499 as control, by FACS; both strains were grown at RT or 37°C for this study. Allele trz1-537 caused a 1 hour delay in cells entering 2N (M phase) at 37°C. When yeast reached stationary phase, YL12-01 had more cells in 2N (M phase) than YPH499 grown under the same conditions (data not shown). When viewed under a microscope the trz1-537 integrants were observed to aggregate when grown at 37°C while parental YPH499 cells did not. With YL12-01 we also observed enlarged cells and cells with elongated buds; the percentage of these abnormally shaped cells increased with the length of time that the cells were grown at 37°C (data not shown). Therefore, trz1-537 allele does not cause any cell cycle arrest phenotype and the abnormal shape we observed is most likely the result of cell death in yeast.
High copy suppressor screen
During the characterization of YL12-01, we identified a window of two days wherein it would be possible to screen for high copy suppressors of trz1-537. As shown in Figure 6, at 30°C, the appearance of YL12-01/p416Gal1 transformants takes 4 days on SC-URA. However, colonies appear after only 2 days when the same strain is transformed with p416/TRZ1 (YL12-02). YPH499 transformed with vector or TRZ1 resulted in observable colonies after 2 days of incubation at 30°C on SC-URA. Therefore, we screened for high copy suppressors of trz1-537 by using yeast genomic fragments to transform YL12-01 and selecting transformants that had grown out in 2-3 days. We screened approximately 34,000 genomic fragments and isolated 4 suppressor clones. Two of the four clones yielded an identical sequence from a 8057 bp genomic region on chromosome 12 (253,621-261,678) that contains four genes (ERG3, YLR057w, SHM2 and REX2). One of the remaining clones includes a 2594-bp genomic sequence from chromosome 11 that contains the full-length coding region of TRZ1 plus 25 bp of 5'UTR and 53 bp of 3'UTR. The 4th clone contains chromosome 12 sequences at one end (from 794,633 down) and Ty3 LTR at the other. This last clone was not pursued further.
We cloned each of the four genes on the 8057 bp region of chromosome 12 and inserted these independently into CEN and 2 μ expression vectors containing the CYC1 promoter. We tested each gene for an ability to rescue the ts phenotype of trz1-537 yeast. Only REX2 rescued the ts phenotype and rescue was observed with both CEN and 2 μ vectors. REX2 encodes a RNA 3' processing enzyme that localizes to mitochondria [20,21].
Discussion
In this paper, we report investigation of TRZ1, the yeast homolog of ELAC2. Altered expression of TRZ1 in yeast does not cause any altered growth phenotype. Conservation of the histidine motif and the C-domain putative P-loop is important for the essential function of Trz1p in yeast. Trz1p (Y537L), a ts mutant, exhibits a slow growth phenotype on a variety of yeast media. Additional copies of REX2 rescue the slow growth phenotype associated with Trz1p (Y537L), providing genetic confirmation that the product of TRZ1 plays a role in RNA processing.
1. Altered expression of TRZ1 is not detrimental to yeast growth
Although TRZ1 is an essential gene, yeast can tolerate altered expression of TRZ1. We down-regulated TRZ1 message by deleting the endogenous gene and expressing TRZ1 from a Gal1 promoter on a CEN plasmid. We could not detect any apparent growth defect when TRZ1 expression was suppressed on glucose-containing medium. We also over-expressed exogenous TRZ1 message and protein from a Gal1 promoter on backgrounds of either an intact or deleted endogenous TRZ1 locus. In neither of these backgrounds does over-expression of TRZ1 cause any significant alteration of yeast doubling time. We also attempted to decrease TRZ1 mRNA by integration of a thermal labile cassette (Degron approach; data not shown) or by insertion of a Gal1 promoter just upstream of the endogenous TRZ1 coding sequence. In these attempts, however, we did not obtain any integrants that were not petite mutants.
In conclusion, under or overexpression of TRZ1 does not grossly affect growth or cell cycle; however, the tendency for alteration in this gene to cause a petite phenotype confounds our ability to detect small perturbations.
2. ELAC2 cannot substitute for TRZ1 function and the C-domain putative P-loop is important for TRZ1 essential function
Human ELAC2 cannot rescue TRZ1 deletion in yeast. We observed this with both full-length ELAC2 and with TRZ1/ELAC2 and ELAC2/TRZ1 fusions. The same lack of complementation was observed with ELAC1, the paralog of ELAC2. These results could be explained if ELAC1 and ELAC2 have overlapping but not totally redundant functions. TRZ1 is more similar to ELAC2 than to ELAC1, but Trz1p may accomplish functions in yeast performed by both ELAC1 and ELAC2 in humans. Therefore, substituting TRZ1 with either one of these genes may not replace all of the essential functions of Trz1p in yeast. It would be interesting to see whether supplying both human ELAC1 and ELAC2 could rescue TRZ1 deletion in yeast.
A motif search found two putative nucleotide-binding motifs (P-loops) in TRZ1 and ELAC2, one in the N-domain and one in the C-domain of each gene [2]. The finding that yeast tolerates mutation of the N-domain but not the C-domain P-loop of TRZ1 suggests the putative C-domain nucleotide-binding motif plays an important role in Trz1p function. The fact that the C-domains of TRZ1 and ELAC2 contain the most conserved sequence across other organisms also supports the notion that the vital functions of TRZ1 require that this sequence element is intact. However, the N-domain of Trz1p is also necessary for function because expression of an N-terminally deleted TRZ1 does not rescue trz1 yeast, nor did the expression of the fusion protein between the N-domain of human ELAC2 and the C-domain of yeast TRZ1. This observation is consistent with the results from Nashimoto's lab [9]. Their data suggest that the N-terminal half of ELAC2 (human tRNase ZL), to which family yeast TRZ1 also belongs, is responsible for substrate specificity and confers RNase 65 activity that is not present in ELAC1 (human tRNase ZS). And the activities of tRNase Z and RNase 65 toward a set of substrates between human ELAC2 and yeast TRZ1 are different, suggesting the N-domain of tRNase ZLs is essential for certain functions in the cell and they may not functionally equivalent among different species [9].
3. Potential role of TRZ1 in rRNA biogenesis, mRNA splicing and mitochondrion maintenance in addition to its role in tRNA processing
The majority of the RNA in growing yeast is noncoding [13] including rRNA, tRNA, snRNA, snoRNA as well as the RNA components of RNase P and RNase MRP. These RNAs are involved in processes that are vital to cell survival. The biogenesis of rRNA is a complicated process that involves a large number of proteins and RNAs that catalyze or modulate transcription, translocation, and various processing reactions, including exonucleolytic and endonucleolytic cleavages, ligations, terminal additions (5S), and nucleoside modifications. The details of yeast rRNA processing and some of the complexes involved have been described [13,14,22,23].
REX proteins (RNA exonucleases) belong to a family of well-conserved RNA 3' → 5' exonucleases [24]. Rex1p is required for 5S and tRNA-Arg3 maturation. Deletion of Rex2p results in a defect in the processing of the U4 snRNA that is involved in mRNA splicing. Rex3p is required for the maturation of MRP RNA, an important endonucleolytic complex involved in processing both rRNA and tRNA. Rex1p and Rex2p have been shown to be involved in the final maturation of 5.8S rRNA. Rex1p, Rex2p and Rex3p are also involved in the maturation of U5 snRNA and the RNA subunit of RNase P, the RNase that is involved in both tRNA 5' and rRNA maturation. In addition to participating in tRNA and rRNA processing, Rex2p has been implicated in spliceosome assembly by association with the snRNP U2 component PRP11, in DNA repair by association with Rad50p, and in cytoskeleton reorganization by association with the actin cortical patch component Cof1p [25]. Rex2p localizes to the mitochondria and interacts genetically with YME1 and YME2, components of the inner mitochondrial membrane that participate in mitochondrion organization and biogenesis [20,21]. In summary, the REX proteins function in both RNA maturation and mitochondrial maintenance. The finding that REX2 rescues the slow growth phenotype of yeast expressing Trz1p (Y537L), suggests that TRZ1 may also participate in those same processes. Further support for this hypothesis comes from the finding that TRZ1 exists in a complex with NUC1 [26,27], an RNase and DNA endo/exonuclease that localizes to the mitochondria and is involved in mitochondrial DNA recombination. In addition, we observed that supplying an extra copy of wild type TRZ1 on a CEN plasmid alters YL12-01, converting the ts mutant Trz1p (Y537L) from ρ- to ρ+ (Table 1). The mitochondrial localization of Trz1p [26] together with the suppression of the Trz1p (Y537L) phenotype by REX2 and the effect of expression of wt TRZ1 on the petite mutant status of YL12-01 suggest that Trz1p may participate in the activities of mitochondria.
The human homolog of TRZ1, ELAC2, was identified as a candidate prostate cancer susceptibility gene [2]. Schiffer et al. [5] purified a nuclear tRNase Z from wheat germ and found two homologs of tRNase Z in A. thaliana, NUZ (nuclear RNase Z) and CPZ (chloroplast RNase Z). Both Arabidopsis homologs have tRNase Z activity in vitro. These authors also showed that NUZ is a homolog of ELAC1 and represents a family of proteins that is conserved from bacteria to humans. Subsequently, Takaku et al. showed that human ELAC2, its paralog ELAC1, and the yeast homolog TRZ1 have tRNase Z activity in vitro [3]. The finding in this study that extra copies of REX2 rescue a ts allele of TRZ1 confirms a role for Trz1p in the processing of RNA, placing TRZ1 in particular in rRNA and tRNA maturation pathways. This is consistent with previous results. First, rRNA biogenesis involves processes both in the nucleolus and cytoplasm, and TRZ1 protein localizes to both of these cellular compartments [28]. Second, many proteins and some complexes, for example, MRP, RNase P and the RNA polymerase III complex, participate both in tRNA and rRNA processing and synthesis. Therefore it is possible that TRZ1 participates in both tRNA and rRNA processing. Third, using a tet-promoter TRZ1 allele and a custom DNA oligo array for noncoding RNA biogenesis, Peng et al. demonstrated that TRZ1 has an undefined role in the processing of 35S rRNA [29]. The high-copy suppressor result could also be explained more straightforwardly if we assume that Trz1p (Y357L) has diminished or abolished tRNase Z activity, which could result in low level of mature tRNAs, and Rex2p could remove the 3' trailers from pre-tRNAs exonucleolytically, that could result in restored mature tRNA levels.
Since ELAC2 cannot functionally complement TRZ1 deletion, we cannot directly test in yeast the functional effect of the human ELAC2 missense substitution A541T, which appears to confer modest risk of prostate cancer [30]. Although this variant does not seem to affect the tRNase Z activity of human ELAC2 in vitro [3], we did find that some missense substitutions at the corresponding yeast residue, M535, are lethal, demonstrating that this position is under functional constraint. Recent demonstrations that hoe-1, the C. elegans homolog of ELAC2, plays a role in germline proliferation [31], and human ELAC2 maybe involved in cell cycle regulation [4], may help better illuminate the contribution of ELAC2 to the etiology of human prostate cancer.
Conclusion
We have undertaken a genetic study of yeast TRZ1 gene, the homolog of prostate cancer susceptibility gene ELAC2. Our data suggest that the absolute level of TRZ1 transcript is not important for the fitness of yeast, the vital functions of TRZ1 require an intact C-terminal P-loop and some conserved residues in Histidine motif. More importantly, our high copy suppressor screen suggests that TRZ1 is involved in RNA processing, especially rRNA processing and mitochondrial biogenesis.
Methods
Strains and Reagents
The strains used in this study are listed in Table 1. The petite status of each strain, determined by growth on a non-fermentable carbon source, is listed in Table 1. YPH499, YPH500, and YPH501 were purchased from Strategene (La Jolla, CA). Genetic manipulations, yeast growth conditions and preparation of yeast media and plates were as described [32]. Yeast transformation and gene disruption were carried out using lithium acetate [32]. Cloning and manipulation of yeast plasmids were done in bacterial strain DH5α. Yeast medium was purchased from Qbiogene (Carlsbad, CA). 5'-Fluoroorotic acid (5'-FOA), used at a concentration of 0.1%, was purchased from Toronto Research Chemicals, Inc. (Ontario, Canada).
Plasmid and strain construction
A 3.7 kb fragment containing TRZ1 coding sequence and intergenic regions was amplified from yeast genomic DNA (primer sequences available upon request) and cloned into pBluescript-KS (BSKS/TRZ1). To express epitope-tagged protein, TRZ1 N- and C-termini were re-engineered using PCR. We generated BSSK TRZ1•FLAG (2 copies of a FLAG epitope fused to the C-terminus), BSSK/TRZ1•V5 (1 copy of a V5 epitope fused to the C-terminus), BSSK/TRZ1•3XHA or BSSK/TRZ1•3Xmyc (3 copies of HA or Myc epitopes fused to the C-terminus, respectively) and BSSK/FLAG•TRZ1 (2 copies of a FLAG epitope fused to the N-terminus). Site-directed mutagenesis was performed with QuikChange Site-Directed Mutagenesis kit from Stratagene (La Jolla, CA) following their protocols.
YL03-47 is a trz1 haploid strain in which the genomic trz1 is complemented by high-copy episomal TRZ1. To construct this strain, first a TRZ1/trz1 heterozygote, was generated using a DNA fragment (TRZ1•hisG•URA3•hisG•TRZ1) that contained in succession: the TRZ1 promoter region; the 5' 37 codons of TRZ1; URA3 flanked by hisG cassettes; the 3' 123 codons of TRZ1; and 108 bp of sequence 3' of the TRZ1 stop codon. YPH501 was transformed with this fragment, and TRZ1/trz1 heterozygotes were selected as uracil prototrophs. Next, one such TRZ1/trz1 heterozygote was plated onto FOA to select a URA3-deficient derivitive of the TRZ1/trz1 heterozygote. This TRZ1/trz1 uracil auxotroph was then transformed with a 2 μ URA3 TRZ1 expression plasmid (pRS426/TRZ1). Finally, to obtain YL03-47, tetrads from the resulting transformants were dissected, and URA+ colonies were screened by PCR for the presence of hisG sequence at the endogenous TRZ1 locus.
YL12-01, in which the chromosomal TRZ1 is replaced with trz1-537 allele, is constructed from YPH499 following a two-step gene replacement strategy. DNA fragment encoding the ts mutant trz1-537 allele was cloned into an URA3-marked integrating vector (p406/trz1-537). The resulting plasmid was linearized with Bgl II and transformed into YPH499 to integrate into TRZ1 chromosomal locus. Uracil prototrophs were selected from the transformation and plated on FOA to select cells that had undergone excision of the plasmid. The replacement of wild type TRZ1 with the ts mutant trz1-537 allele was confirmed with sequencing.
YL13-01, YL13-02 and YL13-03, in which the chromosomal TRZ1 is deleted and a wild type TRZ1 or a trz1-537 allele or a trz1-538 allele is integrated at the LEU2 locus, were generated as follows. DNA fragments encoding TRZ1, or the ts mutant alleles trz1-537 or trz1-538 were cloned into an integrating LEU2-marked vector (p405/TRZ1). The resultant plasmids were linearized with BstXI, and YL03-47 was transformed with the fragments for integration into the leu2-Δ 1 locus. Leucine prototrophs were selected from the transformation and plated on FOA to select cells that had lost pRS426/TRZ1. The identity of the integrated alleles was confirmed by PCR and sequencing.
Western blot
Yeast cells were grown as indicated in the text. Usually, about 1.5 ml yeast culture with an A600 of about 0.6 was centrifuged and the pellet resuspended in 150 μL boiling buffer (50 mM Tris pH8.0, 0.1% Bromophenol blue, 2% SDS, 10% glycerol and fresh 0.1 M DTT). The yeast suspension was heated to 95°C for 20 minutes followed by vortexing and centrifugation at 14000 rpm for 1 minute. The resulting cellular extract was loaded onto a SDS-PAGE gel. Electrophoresis, transfer and Western blot were carried out using Invitrogen's XCell SureLock system following their protocol. Anti-myc and anti-V5 antibodies were purchased from Invitrogen (Calsbad, CA) and used at 1:2000 dilution. Anti-FLAG antibody was purchased from Sigma (St. Louis, MO) and used at 1:400 dilution.
Temperature-sensitive (ts) mutant screen
To prepare for a screen for TRZ1 mutations that rendered yeast growth temperature-sensitive, the codons of specific TRZ1 amino acids were mutagenized. We targeted amino acids 535, 536, 537, 538, and 541 for ts screening. To mutagenize amino acid 535, a pair of oligos with degenerate nucleotides encoding residue 535 were synthesized. The primer sequences were as following: 079c.535.F (5'-CAG GAT TTG AAA NNN ATA TAT CTG AGT CAC-3'), 079c.535.R (5'-ACT CAG ATA TAT NNN TTT CAA ATC CTG AAA TAT TG-3'), 079c.NdeI (5'-GAT TAT GAT TGT GCT GAG CTT GGC-3'), 079c.BglII (5'-TTC TTC ACG GCA TCC TCC AGT AGC-3'). Two fragments were generated by primary PCR with wild type TRZ1 coding sequence as template and 079c.NdeI/079c.535.R and 079c.535.F/079c.BglII as primers. About 100 ng of each of the resulting fragments were mixed and used as templates for secondary PCR using 079c.NdeI/079c.BglII as primer. The resulting PCR products were purified by Qiagen PCR column, mixed with p415/TRZ1 (Δ NdeI-NcoI) that lacks TRZ1 sequence between Nde I and NcoI sites, and the mixture used to transform competent YL03-47. Transformants were plated onto SC-LEU. For other targeted amino acids, PCR-generated mini-libraries were made by following a similar strategy to that outlined above. About 500 colonies were screened for each targeted amino acid. After two days' growth on SC-Leu media at 30°C, colonies were replica-plated onto two SC-Leu+FOA plates, one of which was incubated at 30°C, while the other was incubated at 37°C. After 24 hours colonies that grew at 30°C but failed to grow at 37°C were isolated from the original SC-LEU plate. The LEU2 marked plasmids were isolated, sequenced, and used to retransform YL03-47. Transformants were plated on FOA to select yeast that had lost p426/TRZ1 and streaked onto SC-LEU+FOA plates to confirm the temperature sensitive phenotype.
High copy suppressor screen
A 2 μ URA3-marked yeast genomic library (gift from Mike Forte and Beth Lachlyd, Vollum Institute, Portland, OR) was used to transform YL12-01. Transformants were plated onto SC-URA and after four days' incubation at 30°C, colonies were picked. About 90 colonies were picked out of a total of 70,000 colonies screened. The library plasmids were isolated and used to retransform YL12-01. The suppressor phenotype was confirmed by comparing the growth of YL12-01 transformed with each library plasmid with that of YL12-01 transformed with p416/Gal1 and p416/TRZ1. The sequence of the high copy suppressors was determined.
List of abbreviations used
TRZ1: tRNase Z 1
tRNase Z: transfer RNA (tRNA) 3' processing endoribonuclease
CTD: C-terminal domain
REX: RNA exonuclease
tRNA: transfer RNA
rRNA: ribosomal RNA
snRNA: small nuclear RNA
snoRNA: small nucleolar RNA
Authors' contributions
YC (Yang Chen) designed most of the experiments, carried out most of the genetic studies, interpreted the data and wrote the manuscript. AB carried out most of the high copy suppressor screens and ts screens. CD carried out most of the growth assays and confirmation of high copy suppressors. YC (Yuan Chen) carried out some of the ts screen and growth assays. DS and SVT have made substantial contributions to conception and design of the experiment and drafting of the manuscript.
Acknowledgements
We thank Dr. Mike Forte and Dr. Beth Lachlyd (Vollum Institute) for sharing their high copy genomic yeast library. We thank Dr. John Manfredi for his intellectual input. We are especially grateful to Dr. Christine Klein for her critical reading and editing of the manuscript.
Figures and Tables
Figure 1 Induction of exogenous FLAG•Trz1p. A. Growth of YL09-02 was initiated in raffinose complete medium. After overnight growth at 30°C, the culture was diluted with raffinose complete medium to O.D. 0.2–2.3 at 600 nm, grown to O.D. 0.4, then induced in exponential growth phase with 2% galactose for varying periods of time before processing for Western blot. B. YPH499 and YL10-04 were grown as described above and induced with 2% galactose for 4 hours before harvesting for Western blot. Arrow indicates FLAG•Trz1p.
Figure 2 Protein multiple sequence alignment across the core histidine motif of selected ELAC/ tRNase Z family members. The ELAC/tRNase Z family members were selected from Saccharomyces cerevisiae (Sce), Schizosaccharomyces pombe (Spo), Arabidopsis thaliana (Ath), Drosophila melanogaster (Dme), Homo sapiens (Has), Escherichia coli (Eco) and Methanothermobacter thermautotrophicum (Mth). The boxed residues in the S. cerevisiae sequence are the amino acids that we mutagenized in our effort to create a ts mutant. 1, S. cerevisiae M535 corresponds to human A541. 2, Y537L is ts. 3, L538K is ts. 4, H542 corresponds to human H548, the position of a prostate cancer associated frameshift. 5, G548 corresponds to G256 of S. cerevisiae PSO2; although PSO2 G256R is a ts mutant [16], TRZ1 G548R is lethal. * denotes identical amino acid, and : /. indicates homologous conservation across species.
Figure 3 Temperature sensitive growth of yeast expressing alleles trz1-537 and trz1-538. Isolates of yeast expressing trz1-537 (A1, B1, C1, D1) or trz1-538 (B3, C3, D3, E3) (YL03-47 with ts alleles on a LEU2 plasmid) were streaked onto SC-LEU plates and incubated at RT or 37°C. YPH499 and YL03-47 were used as controls.
Figure 4 Growth at different temperatures of yeast expressing TRZ1, trz1-537 or trz1-538 alleles on media with various carbon sources added. Yeast bearing TRZ1 alleles integrated at the LEU2 locus [YL13-01 (WT), YL13-02 (tr1-537) and YL13-03 (trz1-538)] were serially diluted and spotted onto SC-LEU with different sugars added. Plates were incubated at RT (upper), 30°C (middle) and 37°C (lower).
Figure 5 Wild type TRZ1 restores growth at 37°C of YL12-01 on media supplemented with different sugars. YPH499 (A) and YL12-01 (B) were transformed with either an empty vector (upper plates) or a vector bearing the wild type TRZ1 gene (lower plates). Transformants were replica-plated onto SC-URA containing different sugars and incubated at 37°C for 2 days. Wild type TRZ1 restores the growth of ts strain YL12-01 on all carbon sources.
Figure 6 Growth difference exploited for suppressor screen. YPH499 and YL12-01 were transformed with either an empty vector (p416) or with p416/TRZ1 expressing a wild type TRZ1 gene. Transformants were plated onto SC-URA and incubated at 30°C for 3 days.
Table 1 Strains used in this study
YPH499 MATa ura3-52 lys2-801amber ade2-101orchre trp1-Δ63 his3-Δ200 leu2-Δ1 ρ+
YPH500 Same as YPH499 except the mating type is α ρ+
YPH501 Same as YPH499 except the mating type is a/α ρ+
YL03-47 YPH499 trz1Δ::hisG (TRZ1 URA3 2 μ) ρ+
YL09-00 YPH499 transformed with Gal1p-β-gal (TRP1, CEN) NA
YL09-01 YPH499 transformed with Gal1p-TRZ1 (TRP1, CEN) ρ+
YL09-02 YPH499 transformed with Gal1p-FLAG•TRZ1 (TRP1, CEN) ρ+
YL10-01 YPH500 trz1Δ::URA3 (GAL1p-TRZ1 (TRP1, CEN)) ρ-
YL10-02 YPH499 trz1Δ::hisG (GAL1p-TRZ1 (TRP1, CEN)) ρ-
YL10-03 YPH499 trz1Δ::hisG (GAL1p-FLAG•TRZ1 (TRP1, CEN)) ρ+
YL10-04 YPH499 trz1::Gal1-FLAG•TRZ1 ρ-
YL10-05 YPH499 trz1::Gal1p-TRZ1 ρ-
YL12-01 YPH499 trz1Δ::trz1-537 ρ-
YL12-02 YL12-01 transformed with TRZ1 (URA3, CEN) ρ+
YL13-01 YPH499 trz1Δ::hisG TRZ1 (WT)::LEU2 ρ+
YL13-02 YPH499 trz1Δ::hisG trz1-537::LEU2 ρ-
YL13-03 YPH499 trz1Δ::hsiG trz1-538::LEU2 ρ-
Table 2 doubling time (hour)
Strains SC glucose SC galactose
YPH499 2.18 3.11
YL09-00 2.11 2.99
YL09-01 2.20 3.85
YPH499 2.62 3.75
YL10-04 7.70 139
YL10-05 5.64 49.5
YL09-01 2.22 3.55
YPH500 2.27 5.54
YL10-01 2.95 63.0
YPH499 1.90 3.25
YL10-01 2.96 105
YL10-02 2.75 431
YL10-03 1.83 3.17
YL03-47 2.27 3.25
Yeasts were inoculated into 2 mls of SC-raffinose medium and grew overnight at 30°C. The overnight culture was then diluted with either SC-glucose or SC-galactose to O.D. 600 nm of 0.1 and return to 30°C shaking. The O.D. 600 nm reading on these cultures were monitored throughout the day until they reached more than 1. The doubling time equals to ln (2)/k, where k is the slope of the least-squares linear fit of ln (O.D.600 nm) versus time. Spaces between lines in the table separate experiments done on different days.
Table 3 growth comparison on different sugar
Glucose Sucrose
RT 30°C 37°C RT 30°C 37°C
YPH499 +++++ ++++ +++++ ++++ +++++ +++++
YL09-01 +++++ ++++ +++++ ++++ +++++ +++++
YL10-01 +++++ +++ +++ ++ ++++ ++
YL10-02 +++++ +++ +++ ++ ++++ ++
YL10-03 +++++ ++++ +++++ ++++ +++++ +++++
YL03-47 +++++ ++++ +++++ ++++ +++++ +++++
Raffinose Galactose
RT 30°C 37°C RT 30°C 37°C
YPH499 +++ +++++ +++ ++ +++++ ++
YL09-01 +++ +++++ +++ ++ +++++ ++
YL10-01 + ++ + - - -
YL10-02 + ++ + - - -
YL10-03 +++ +++++ +++ ++ +++++ ++
YL03-47 +++ +++++ +++ ++ +++++ ++
SC-glucose: 3 days at RT, 2 days at 30 and 37°C.
SC-sucrose: 3 days at RT, 2 days at 30 and 37°C.
SC-raffinose: 3 days at RT, 2 days at 30 and 37°C.
SC-galactose: 4 days at RT, 2 days at 30 and 37°C.
+++++ fully grown spot with pink color
++++ fully grown spot with weak pink color
+++ fully grown spot with white color
++ clearly seen by eye
+ barely seen by eye
Yeasts were inoculated into 2 mls of SC-raffinose medium and grew overnight at 30°C. The resultant overnight cultures were diluted with SC-raffinose medium to O.D. at 600 nm 0.2 and returned to growth at 30°C for another 2–4 hours to get to the exponential growth phase. Then the cultures were diluted to O.D. 600 nm 0.2. These freshly diluted cultures were serially diluted and spotted onto SC plate with different sugars added.
Table 4 Complementation of trz1
Constructs Complement Δ trz1
ELAC2 No
ELAC2.FLAG No
FLAG. ELAC2.C-term No
ELAC2.N-term No
ELAC2 (S217L) No
ELAC2 (A541T) No
ELAC2 (R781H) No
ELAC2 (1641insG) No
ELAC1 No
TRZ1 WT Yes
FLAG.TRZ1 WT Yes
trz1 M535T Yes
FLAG.trz1 M535T Yes
trz1 H542Ter No
FLAG.trz1 H542Ter No
trz1 G548R No
FLAG.trz1 G648R No
trz1 G229D Yes
trz1 Q230A Yes
trz1 G698D No
trz1 D699A No
FLAG.TRZ1. C-term No
FLAG.TRZ1. N-term No
FLAG.TRZ1- ELAC2 No
ELAC2-TRZ1 No
The chimeric proteins are made by fusion of either the ELAC2 N-domain with TRZ1 C-domain or the reverse. Trz1p N-domain ends at a.a. 463 and the C-domain starts at a.a. 465. The ELAC2 N-domain ends at a.a. 481 and the C-domain starts at a.a. 481.
The shuffle experiment is done in strain YL03-47 according to protocols described in [32]. The constructs are TRP -marked CEN vectors.
Table 5 Doubling time of ts allele integrants
Strain LEU2 locus RT 37°C
YL13-01 TRZ1 WT 5.54 ± 0.26 1.44 ± 0.04
YL13-02 trz1-537 6.87 ± 0.85 2.88 ± 0.01
YL13-03 trz1-538 6.31 ± 1.03 4.87 ± 1.15
Strain TRZ1 locus RT 37°C
YPH499 WT 3.17 1.85
YL12-01 trz1-537 4.75 ± 0.06 3.67 ± 0.02
Yeasts were inoculated into 2 mls of SC-glucose medium and grew at 30°C overnight. The overnight cultures were diluted with SC-glucose medium to O.D. at 600 nm 0.1 and returned to 30°C with shaking. The O.D. 600 nm reading was monitored throughout the day until it was more 1. Doubling time was calculated as described in table 2.
==== Refs
Ferlay J Whelan SL Parkin DM, Whelan SL, Feraly J, Teppo L, Thomas DB Cancer Incidence in Five Continents 2002 VIII International Agency for Research on Cancer, Lyon, France 91 92 and 208–210
Tavtigian S Simard J Teng DHF Abtin V Baumgard M Beck A Camp NJ Carillo AR Chen Y Dayananth P Desrochers M Dumont M Farnham JM Frank D Frye C Ghaffari S Gupte J Hu R Iliev D Janecki T Kort EN Laity KE Leavitt A Leblanc G McArthur-Morrison J Pederson A Penn B Peterson K Reid JE Richards S Schroeder M Smith R Snyder SC Swedlund B Swensen J Thomas A Tranchant M Woodland A Labrie F Skolnick MH Neuhausen S Rommens J Cannon-albright LA A candidate prostate cancer susceptibility gene at chromosome 17p Nat Genet 2001 27 172 180 11175785 10.1038/84808
Takaku H Minagawa A Takagi M Nashimoto M A candidate prostate cancer susceptibility gene encodes tRNA 3' processing endoribonuclease Nucleic Acids Res 2003 31 2272 2278 12711671 10.1093/nar/gkg337
Korver W Guevara C Chen Y Neuteboom S Bookstein R Tavtigian S Lees E The product of the candidate prostate cancer susceptibility gene ELAC2 interacts with the γ-tubulin complex Int J Cancer 2003 104 283 288 12569551 10.1002/ijc.10945
Schiffer S Rösch S Marchfelder A Assigning a function to a conserved group of proteins: the tRNA 3'-processing enzymes EMBO J 2002 21 2769 2777 12032089 10.1093/emboj/21.11.2769
Pellegrini O Nezzar J Marchfelder A Putzer H Condon C Endonucleolytic processing of CCA-less tRNA precursors by RNase Z in Bacillus subtilis EMBO J 2003 22 4534 4543 12941704 10.1093/emboj/cdg435
Dubrovsky EB Dubrovskaya VA Levinger L Schiffer S Marchfelder A Drosophila RNase Z processes mitochondrial and nuclear pre-tRNA 3' ends in vivo Nucleic Acids Res 2004 32 255 262 14715923 10.1093/nar/gkh182
Minagawa A Takaku H Takagi M Nashimoto M A Novel Endonucleolytic Mechanism to Generate the CCA 3' Termini of tRNA Molecules in Thermotoga maritime J Biol Chem 2004 279 15688 15697 14749326 10.1074/jbc.M313951200
Takaku H Minagawa A Takagi M Nashimoto A The N-terminal half-domain of the long form of tRNase Z is required for the RNase 65 activity Nucleic Acids Res 2004 32 4429 4438 15317868 10.1093/nar/gkh774
Herbert A Rich A RNA processing and the evolution of eukaryotes Nat Genet 1999 21 265 269 10080177 10.1038/6780
Xiao S Houser-Scott F Engelke DR increased complexity to cope with the nuclear pre-tRNA pathway J Cell Physiol 2001 187 11 21 11241345 10.1002/1097-4652(200104)187:1<11::AID-JCP1055>3.0.CO;2-K
Yoo CJ Wolin SL The yeast La protein is required for the 3' endonucleolytic cleavage that matures tRNA precursors Cell 1997 89 393 402 9150139 10.1016/S0092-8674(00)80220-2
Perry RP RNA processing comes of age J Cell Biol 1981 91 28s 38s 6172431 10.1083/jcb.91.3.28s
Kressler D Linder P de La Cruz J Protein trans-acting factors involved in ribosome biogenesis in Saccharomyces cerevisiae Mol Cell Biol 1999 19 7897 912 10567516
Carlson M Osmond BC Botstein D Mutants of yeast defective in sucrose utilization Genetics 1981 98 25 40 7040163
Niegmann E Brendel M A single amino acid change in SNM1 -ecoded protein leads to thermoconditional deficiency for DNA cross-link repair in Saccharomyces cerevisiae Mutat Res 1994 315 275 279 7526204
Brakoulias A Jackson RM Towards a structural classification of phosphate binding sites in protein-nucleotide complexes: an automated all-against-all structural comparison using geometric matching Proteins 2004 56 250 260 15211509 10.1002/prot.20123
Wu TH Marinus MG Dominant negative mutator mutations in the mutS gene of Escherichia coli J Bacteriol 1994 176 5393 5400 8071216
Shen H Yao B Mueller DM Primary structural constraints of P-loop of mitochondrial F1-ATPase from yeast J Biol Chem 1994 269 9424 9428 8144526
Hanekamp T Thorsness PE Inactivation of YME2/RNA12, which encodes an integral inner mitochondrial membrane protein, causes increased escape of DNA from mitochondria to the nucleus in Saccharomyces cerevisiae Mol Cell Biol 1996 16 2764 2771 8649384
Hanekamp T Thorsness PE YNT20, a bypass suppressor of yme1 yme2, encodes a putative 3'-5' exonuclease localized in mitochondria of Saccharomyces cerevisiae Curr Genet 1999 34 438 48 9933355 10.1007/s002940050418
Lewis JD Tollervey D Like attracts like: getting RNA processing toether in the nucleus Science 2000 288 1385 1389 10827942 10.1126/science.288.5470.1385
Faber AW Dijk MV Raue HA Vos JC Ngl2p is a Ccr4p-like RNA nuclease essential for the final step in 3'-end processing of 5.8S rRNA in Saccharomyces cerevisiae RNA 2002 8 1095 1101 12358428 10.1017/S1355838202021027
van Hoof A Lennertz P Parker R Three conserved members of the RNase D family have unique and overlapping functions in the processing of 5S, 5.8S, U4, U5, RNase MRP and RNase P RNAs in yeast EMBO J 2000 19 1357 1365 10716935 10.1093/emboj/19.6.1357
Ho Y Gruhler A Heilbut A Bader GD Moore L Adams SL Millar A Taylor P Bennett K Boutilier K Yang L Wolting C Donaldson I Schandorff S Shewnarane J Vo M Taggart J Goudreault M Muskat B Alfarano C Dewar D Lin Z Michalickova K Willems AR Sassi H Nielsen PA Rasmussen KJ Andersen JR Johansen LE Hansen LH Jespersen H Podtelejnikov A Nielsen E Crawford J Poulsen V Sorensen BD Matthiesen J Hendrickson RC Gleeson F Pawson T Moran MF Durocher D Mann M Hogue CW Figeys D Tyers M Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry Nature 2002 415 180 183 11805837 10.1038/415180a
Hazbun TR Malmström L Anderson S Graczyk BJ Fox B Riffle M Sundin BA Aranda JD McDonald WH Chiu C Snydsman BE Bradley P Muller EGD Fields S Baker D Yates JR Davis TN Assigning function to yeast proteins by integration of technologies Mol Cell 2003 12 1353 1365 14690591 10.1016/S1097-2765(03)00476-3
Gavin AC Bösche M Krause R Grandi P Marzioch M Bauer A Schultz J Rick JM Michon AM Cruciat CM Remor M Höfert C Schelder M Brajenovic M Ruffner H Merino A Klein K Hudak M Dickson D Rudi T Gnau V Bauch A Bastuck S Huhse B Leutwein C Heurtier MA Copley RR Edelmann A Querfurth E Rybin V Drewes G Raida M Bouwmeester T Bork P Seraphin B Kuster B Neubauer G Superti-Furga G Functional organization of the yeast proteome by systematic analysis of protein complexes Nature 2002 415 141 147 11805826 10.1038/415141a
Huh WK Falvo JV Gerke LC Carroll AS Howson RW Weissman JS O'Shea EK Global analysis of protein localization in budding yeast Nature 2003 425 686 691 14562095 10.1038/nature02026
Peng WT Robinson MD Mnaimneh S Krogan NJ Cagney G Morris Q Davierwala AP Grigull J Yang X Zhang W Mitsakakis N Ryan OW Datta N Jojic V Pal C Canadien V Richards D Beattie B Wu LF Altschuler SJ Roweis S Frey BJ Emili A Greenblatt JF Hughes TR A panoramic view of yeast noncoding RNA processing Cell 2003 113 919 933 12837249 10.1016/S0092-8674(03)00466-5
Camp NJ Tavtigian SV Meta-analysis of associations of the Ser217Leu and Ala541Thr variants in ELAC2 (HPC2) and prostate cancer Am J Hum Genet 2002 71 1475 1478 12515253 10.1086/344516
Smith M Levitan DJ The Caenorhabditis elegans homolog of the putative prostate cancer susceptibility gene ELAC2, hoe-1, plays a role in germline proliferation Dev Biol 2004 266 151 160 14729485 10.1016/j.ydbio.2003.10.016
Guthrie C Fink GR Guide to Yeast Genetics and Molecular Biology, in Method in Enzymology 1991 194 San Diego, Academic Press, Inc
| 15892892 | PMC1156898 | CC BY | 2021-01-04 16:22:25 | no | BMC Mol Biol. 2005 May 13; 6:12 | utf-8 | BMC Mol Biol | 2,005 | 10.1186/1471-2199-6-12 | oa_comm |
==== Front
BMC MedBMC Medicine1741-7015BioMed Central London 1741-7015-3-81589288310.1186/1741-7015-3-8Research ArticlePsychological illness is commonly associated with functional gastrointestinal disorders and is important to consider during patient consultation: a population-based study Ålander Ture [email protected]ärdsudd Kurt [email protected] Sven-Erik [email protected]éus Lars [email protected] Family Medicine, Stockholm, Karolinska Institute, Sweden2 Ålander Family Practice, Uppsala, Sweden3 Department of Public Health and Caring Sciences, Unit of Family Medicine, University Hospital, Uppsala, Sweden2005 13 5 2005 3 8 8 16 9 2004 13 5 2005 Copyright © 2005 Ålander et al; licensee BioMed Central Ltd.2005Ålander et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Some individuals with functional gastrointestinal disorders (FGID) suffer long-lasting symptoms without ever consulting their doctors. Our aim was to study co-morbidity and lifestyle differences among consulters and non-consulters with persistent FGID and controls in a defined adult population.
Methods
A random sample of the general adult Swedish population was obtained by a postal questionnaire. The Abdominal Symptom Questionnaire (ASQ) was used to measure GI symptomatology and grade of GI symptom severity and the Complaint Score Questionnaire (CSQ) was used to measure general symptoms. Subjects were then grouped for study by their symptomatic profiles. Subjects with long-standing FGID (n = 141) and subjects strictly free from gastrointestinal (GI) symptoms (n = 97) were invited to attend their local health centers for further assessment.
Results
Subjects with FGID have a higher risk of psychological illness [OR 8.4, CI95(4.0–17.5)] than somatic illness [OR 2.8, CI95(1.3–5.7)] or ache and fatigue symptoms [OR 4.3, CI95(2.1–8.7)]. Subjects with psychological illness have a higher risk of severe GI symptoms than controls; moreover they have a greater chance of being consulters. Patients with FGID have more severe GI symptoms than non-patients.
Conclusion
There is a strong relation between extra-intestinal, mental and somatic complaints and FGID in both patients and non-patients. Psychological illness increases the chance of concomitantly having more severe GI symptoms, which also enhance consultation behaviour.
==== Body
Background
Gastrointestinal (GI) symptoms are common and are reported within three months by almost every other adult in Western countries (UK, Sweden, USA, Australia) [1-5]. (The abbreviations used in this manuscript are explained in the Additional File.) The most common symptoms are gastroesophageal reflux symptoms/disease (GERS) and dyspepsia originating from the upper GI tract, and Irritable Bowel Syndrome (IBS) from the lower GI tract. Prevalence rates are reported to be 25% for both GERS and dyspepsia and about 12% for IBS [6]. Although intermittent, these disorders may linger in many sufferers [2,7]. The total cost of dyspepsia to society, with and without peptic ulcer disease, GERS and IBS, is considerable [8,9]. Thus, these disorders constitute a major public health problem.
Dyspepsia without peptic ulcer disease (PUD) (or any other organic GI disease) is called functional dyspepsia [10]. In a Swedish population-based upper endoscope study, the prevalence of dyspepsia was 38% and the prevalence of PUD 4%, with about a quarter of the PUD subjects having no GI symptoms [48]. Thus, the vast majority of those reporting dyspepsia in the population can be expected to have functional dyspepsia. IBS is a functional gastrointestinal disorder [11]. Although functional dyspepsia and IBS are considered separate disorders [10,12], many subjects report concomitant symptoms thought to originate from both the upper and the lower the GI tract [1,13]. Also, sufferers may experience a change in predominant symptom profile over time between dyspepsia and IBS, but much less towards GERS [2]. This means that dyspepsia and IBS sufferers probably have common aspects of both pathophysiology and health care seeking behavior. Furthermore GERS, in a majority of cases, has a proven (non-functional) acid-related etiology [11]. Accordingly, it seems reasonable to exclude subjects with only GERS symptoms from the functional definition and to investigate the burden on society of functional dyspepsia and IBS taken together. We thus label these conditions as functional gastrointestinal disorders or FGID.
A minority of subjects with FGID (5–20%) are reported to visit a doctor quarterly for their complaints [14,15]. On the other hand, up to half of those with FGID never see a doctor despite prolonged symptoms [14,16]. For IBS, the proportion is even higher, up to 80 % [14]. Thus, GI patient-based data will not cover all aspects of morbidity. Moreover, it has been reported previously that most patients' sick leave is due to disorders other than their abdominal complaints. Only 23% of those subjects with abdominal pain who were on sick leave stated that abdominal pain was the cause of their absenteeism [17].
There are prior cross-sectional population-based point prevalence studies on GI co-morbidity [18-20]. However, there is a lack of knowledge of the co-morbidity of subjects with proven chronic/long lasting/persistent FGID and its impact on health care seeking behaviour due to GI symptoms. The aim of this study was to compare the co-morbidity and its relation to health care seeking behaviour among non-patients with persistent FGID with those who are persistently GI symptom free, in a randomly-selected adult Swedish population.
Methods
Setting
In January 1995 there were a total of 21,545 Swedish citizens in the Östhammar community, 14,932 of whom were born between 1909 and 1974. Data from three prior studies[2], which included random samples of the Östhammar population, were used to provide symptom-free controls in the current study. We called this new study the Gastrointestinal Consult Study (GiCon). A sample was drawn from the National Swedish Population Registry 1995, now involving all Swedish citizens born between 1909 to 1974 and thus 20 to 87 years of age, born on day 3, 12 or 24 of each month (n = 1537), i.e. using the same date of birth criteria as in previous studies on the same population in 1988 and 1989. Thus, we increased the upper age limit to 87 years in order to include all prior participants. As a consequence of the sampling strategy, all subjects younger than 27 years of age were included for the first time (n = 86). Most the participants were now being approached for the third time. Of the 1537 subjects in the study base (see Figure 1), 105 were excluded due to having declined further participation in the 1988 study and 4 were excluded due to unidentified ID. Thus, 1428 subjects were still eligible. They were sent a validated postal questionnaire, called "The Abdominal Symptom Questionnaire" (ASQ) [21], described below. Two reminders were sent out when necessary. Drop-out reasons are shown in Figure 1, together with the mean age and sex distributions of the 911 responders who constituted the population sample of the current GiCon study.
Figure 1 Sampling procedure of the Gastrointestinal Consult Study (GiCon study). Sampling procedure of the Gastrointestinal Consult Study (GiCon study) with relation to the prior study of 1995.
Formation of the study groups and symptom classification
All 911 responders in the population sample were classified according to their response in the ASQ (see Figure 1) as having either Functional Gastrointestinal Disorder (FGID, n = 244), i.e. functional dyspepsia or irritable bowel syndrome, or being Strictly Symptom Free (SSF, n = 219). Moreover, subjects reporting minor symptoms that did not fulfil the criteria for dyspepsia or irritable bowel syndrome, or reporting isolated symptoms of gastroesophageal reflux symptoms (GERS, n = 271), were excluded. Subjects who were free from abdominal symptoms in this 1995 survey but had reported symptoms in the previous studies (1988–1989) were labelled as Prior Symptomatic (n = 177) and were also excluded.
The FGID and SSF sample groups (n = 244 + 219 = 463) were then invited by post to participate in the current investigation and to visit one of the six local health centres. Subjects who did not respond to the first letter were sent a reminder, and non-responders to the second letter were finally contacted by phone. In total, 187 (77%) with FGID and 156 (71%) with SSF accepted the invitation. There were 117 (57 FGID, 60 SSF) subjects with incomplete (n = 12) or no (n = 105) response. At the health centres, the participants filled in an ASQ and a Complaint Score Questionnaire (CSQ), as described below, and the height and weight of each was measured. The time period between the original 1995 ASQ and the second ASQ survey in 1996 ranged from 7 to 15 months. Of the FGID group of 187 subjects, 8 subjects who now reported no symptoms and 38 with minor symptoms were excluded; and out of the 156 SSF, 13 who now reported FGID and 46 with minor symptoms were also excluded. Thus, data from 141 FGID subjects (mean age 45.7 y, 34% men) and 97 SSF subjects (mean age 52.4 y, 48% men) remained in the analysis. The sampling procedure for retrospective data – twice over one year for those with, and up to four times over eight years for those without, GI symptoms – assured persistent symptomatic status.
Questionnaires
The abdominal symptom questionnaire
The Abdominal Symptom Questionnaire (ASQ) has been validated previously and found to be reliable and reproducible [2,13,22]. In the questionnaire, subjects were asked if they had been troubled (Yes/No) by any of the 29 listed general gastrointestinal symptoms over the previous three months. They were also asked if they had been troubled by any of 11 listed descriptors of abdominal pain and where the symptom was located (upper, centre or lower abdominal, right and left flank, as shown in Figure 2). A similar ASQ was used for the postal survey and the surgery visit, with the latter extended to include a symptom severity Likert scale graded from zero to seven for each symptom asked for. In the analysis, the data were pooled into a three-grade scale (1, 1–4, 5–7)
Figure 2 GI symptoms, pain modalities and location. The symptoms inquired are shown in Part 1, and the pain and discomfort modalities with the master sketch for indicating their abdominal location in Part 2. The asterisks are explained in the text. Swedish laymen terms were used in the questionnaire. The master sketch shows the eligible pain locations.
Definitions of symptoms from the ASQ
(See Figure 2) [23].
Dyspepsia was defined as one or more of the symptoms (reflux episodes, heartburn, retrosternal pain, nausea, vomiting, early satiety, uncomfortable feeling of fullness after meals, abdominal distension), and one or more of the pain modalities (burning sensation, aching, pain, tenderness, sinking feeling, "butterflies", cramp, twinge, stitch, colic, gripes) with any abdominal location except the lower part, but no concomitant IBS.
Irritable Bowel Syndrome (IBS) was defined as one or more of the symptoms (abdominal distension, abdominal discomfort or pain on defecation, abdominal discomfort or pain relieved by defecation, feeling of incomplete defecation, mucous stools) and one or more of the symptoms (diarrhea, constipation, alternating diarrhea and constipation) together with one or more of the pain modalities (burning sensation, aching, pain, tenderness, sinking feeling, "butterflies", cramp, twinge, stitch, colic, gripes), with any location.
When using these definitions of dyspepsia and IBS, responders with only reflux symptoms, i.e. heartburn and/or retrosternal pain but no abdominal pain or discomfort, were classified as having GERS and not dyspepsia, while those with such symptoms and abdominal pain or discomfort fell into the dyspepsia group. Also, IBS, as defined below, was given priority over dyspepsia. Thus, concomitant occurrence of both led to a diagnosis of IBS. This definition is in accordance with published guidelines [24].
Functional Gastrointestinal Disorder, FGID
FGID was defined as either dyspepsia or IBS.
Minor symptoms refer to symptoms not fulfilling the criteria of dyspepsia or IBS. By definition, those with GERS were included in this group
"Strictly Symptom Free" (SSF) is defined as having no reported symptoms at all in the Abdominal Symptom Questionnaire in the 1995 survey, and subjects stating that they had not even been troubled previously with abdominal symptoms. Those subjects who had participated in the two former surveys in 1988 and 1989 (n = 265) should also have reported no symptoms in each of those two investigations.
"Prior symptomatics" were those with no symptoms in the 1995 survey but with GI symptoms reported in the prior studies of 1988 and 1989.
The complaint score questionnaire
The Complaint Score Questionnaire contains 30 questions, as shown in Figure 3, indicating the presence or absence of 30 different symptoms [25]. The questions are dichotomous and can be categorized into six domains: abdominal/urinary, ache/fatigue, muscular-skeletal, nutritional, cardio-vascular and depressive.
Figure 3 Proportion of subjects with FGID and with SSF reporting complaints during the previous three months in the CSQ. Proportions (0–1.0) of subjects with FGID and with SSF (n = 238) reporting complaints during the previous three months in the Complaint Score Questionnaire (CSQ). There were significance differences *(p < 0.0016) on an age and sex adjusted logistic regression for all variables except coughing (ns) and excessive weight (ns).
Other questions
The participants were asked to state their coffee, alcohol and tobacco consumption, and whether they had ever had peptic ulcer disease or had ever consulted a doctor for GI complaints. Also, the past three months of general pain and all GI medication were indicated. Educational background was registered at five levels (elementary, comprehensive, secondary, upper secondary, university) and medical knowledge was evaluated by means of a self-explanatory questionnaire (see Additional File). The answers were scored with a sum of 0–15.
Statistical power and analysis
One hundred and twenty three subjects were required (in each the FGID and SSF groups) in order to have a 90% power at the P < 0.05 level to detect a 20% absolute difference in exposure variables. This assumed a 24% prevalence of FGID in the population, equal numbers of subjects in the FGID and SSF groups, 15 and 20% absolute difference in the exposure variables within the two steps of the study, 75% response rate on the ASQ, and 25% exclusion from each group in the last step. The variables analysed are presented in the Additional File. Univariate analyses were performed using Student's t-test, Wilcoxon's ranksum test (Mann Whitney) and Pearson's chi-2 test. Multivariate analysis with FGID as the dependent variable was performed using a logistic regression, and with dyspeptic severity as the dependent variable with the ordinal logistic regression technique. Age and BMI were linear to outcome and were thus handled as continuous variables. To test the symptoms in the CSQ, we made an age and sex adjusted logistic regression model for each symptom and accepted a P value less than 0.05/(number of symptoms) = 0.05/30 ≈ 0.0016 as significant. The explanatory variables were evaluated by a logistic regression in a sex and age adjusted model.
The association between each potential determinant obtained from the questionnaires and the presence of FGID was quantified by using odds ratios and 95% confidence intervals. All exposure variables with P < 0.25 were then entered together in a multivariate logistic regression model to evaluate which was independently associated with the presence of FGID. A factor analysis was performed using all 30 complaints and factors with eigenvalues greater than 1.3. Four factors were obtained by a principal components analysis with varimax rotation. The 30 variables were dichotomized with the highest 1/3 given the value 1 and the rest the value 0. A logistic regression was performed with the four factors age and sex adjusted, and factors with p > 0.05 were excluded. To test whether the model fitted the data, a Pearson goodness-of-fit test with p values greater than 0.05 was performed. When the number of covariates approximated the number of observations, the Hosmer-Lemeshow test was performed to determine whether the model fitted the data. For the ordinal logistic model, comparison with a multinomial model made an approximate fit test. No interactions were found between the variables in the main model. Ninety five per cent confidence intervals (CI) were computed using parametric methods. A p-value of 0.05 or less was generally regarded as statistically significant. All tests were two-tailed and the statistical package Stata 8 was used for analysis [27].
Ethics
This study was approved by the Ethics Committee of the Medical Faculty, Uppsala University, June 5th 1996.
Results
Representativeness of study sample
In order to investigate the effect of the drop-outs during the sampling procedure, the mean ages, sex ratios and education levels in the corresponding FGID and SSF groups were analysed as shown in Table 1, for the samples from 1988, 1995 and 1996 (see Table 1) [13]. There were no significant differences in any of these aspects (see Table 1).
Table 1 Comparison between previous population studies in Östhammar, Sweden. Age, sex and education level at different stages of the sampling process. From the first population sample in 1988 to the present study 1996. ns = p > 0.05.
Group (G) Sample Year n Age years mean (SD) Sex % males Education median level (range)
1 First population sample 1988 1156 48.9 (16.0) 50.4 3 (1–4)*
2 Eligible sample 1995 1428 49.9 (17.2) 49.9 3 (1–4)**
3 Population sample 1995 911 49.2 (16.46) 47.0 3 (1–4)
4 Sample group FGID 1995 244 45.5 (15.3) 36.1 3 (2–4)
5 Sample Group SSF 1995 219 51.7 (17.6) 51.6 3 (1–4)
6 Study Group FGID 1996 141 45.7 (14.3) 34.0 3 (2–4)
7 Study Group SSF 1996 97 52.4 (15.3) 48.0 3 (1–4)
G2 vs G3 ns ns
G2 vs G3 ns ns ns
G4 vs G6 ns ns ns
G5 vs G7 ns ns ns
*responders = 1156 **responders n = 1384
Study group characteristics
As shown in Table 2, there were more females among those with FGID. However, there were no intergroup differences in education, medical knowledge, BMI, intake of coffee, alcohol and smoking. The age difference was irrelevant, as those with SSF had been largely excluded due to prior study results. Disease-related variables were significantly different between the study groups. These variables were introduced to further modelling, as shown below. The variables 'intake of GI medicine' and 'previous PUD' were not included due to sparse data.
Table 2 Comparison between explanatory variables for subjects with FGID and SSF. Comparison between explanatory variables for subjects with FGID and SSF adjusted for sex and age. Ordinal variables are presented as median (range), dichotomous variables as proportion %, and continuous variables as mean (SD).
Variable FGID
n = 141 SSF
n = 97 P
AGE 45.7 (14) 52.4 (15) not relevant
SEX (female %) 66 53 0.026 ***
BMI 26.3 (4.7) 26.2 (4.2) Ns *
Education 3 (2–4) 3 (1–4) Ns **
Medical knowledge 10 (4–15) 11 (5–15) Ns **
GI sympt severity 4 (3–5) 0 (0–0) <0.0005 **
GI Consultation (ever) 72% 8% <0.0005 ***
Pain medicine (3 month) 77% 34% <0.0005 ***
GI medicine (3 month) 32% 1% <0.0005 ***
Previous PUD (ever) 12% 1% 0.006 ***
Coffee 2 (2–3) 2 (2–3) Ns **
Alcohol 4 (3–4) 3 (2–4) Ns **
Smoking 1 (1–2) 1 (1–2) Ns **
* = Student's t-test ** = Mann-Whitney test *** = Pearson Chi-2 test
CSQ and FGID vs. SSF
The results from the 30 CSQ complaints for the FGID and SSF study groups are presented in Figure 3. Those with persistent FGID scored statistically higher on all variables except difficulties in passing urine, excessive weight, coughing and impaired hearing.
Risk modelling
Risks of reported CSQ complaints for FGID vs. SSF, expressed as age and sex adjusted OR, are presented in Table 3. GI complaints (abdominal pain, nausea, diarrhea and constipation) were excluded as we aimed to analyze the co-morbidity with GI symptoms. The OR was significant for all except four complaints. After adjusting for alcohol and pain and GI drug intake (Table 3), 20 complaints remained significant. A factor analysis was performed including the 26 non-GI complaints. After a varimax rotation of the four factors with eigenvalues > 1.3, we found four factors representing psychological illness, somatic illness, ache/fatigue and one "miscellaneous" (Table 3). Each factor was then introduced in a logistic regression model adjusted for sex and age (Table 3), and the "miscellaneous" factor was shown to be non-significant. In the last sequential analysis, the three factors that remained significant in the prior analysis were introduced together into a main effect model, adjusted for age and sex. The OR for these factors remained significant, as shown in Table 3.
Table 3 Odds ratios of FGID/SSF for complaints in the Complaints Score Questionnaire (CSQ). Odds ratios (OR, with 95% confidence intervals (CI)) of FGID/SSF (n = 238) for complaints elicited by the CSQ. Logistic regression is presented in different models. A factor analysis extracted four factors: A = psychological illness factor, B = somatic illness factor, C = miscellaneous factor, D = ache/fatigue factor. These were used in the modelling in the right two columns.
I II III IV V
Symptom OR (CI) by Models adjusted for sex and age OR (CI) by Models adjusted for sex, age, alcohol, pain tablets, GI-tablets FACTOR OR (CI) by Models adjusted for sex and age OR (CI) by a main effect model adjusted for sex and age
SSF (all variables) 1 (Reference) 1 (Reference)
FGID
Cries easily 6.7 (2.3–19.9) 9.8 (2.0–47) A 8.0 (4.1–15.8) 1) Psychological illness 8.4 (4.0–17.5) 1) Psychological illness
Sleeping disturbance 6.2 (2.7–14.0) 3.2 (1.3–8) A
General fatigue 14.5 (7.4–28.7) 12.6 (5.3–30) A, D
Irritability 8.8 (4.1–17.8) 5.6 (2.3 – 13.7) A, D
Nervousness 18.4 (4.2–80.3) 14.3 (2.8 – 72) A
Impaired concentration 19.0 (5.7–63.8) 15.3 (4.0 – 58) A
Difficulty to relax 15.7 (6.0–41.5) 10.9 (3.7–32) A
Restlessness 40.0 (9.4–170) 32.2 (6.7–154) A
Depression 8.6 (4.1–18) 4.7 (2.0 – 11) A
Exhaustion 12.7 (4.4–37) 9.1 (2.7–30) A
Chest pain 40.0 (5.3–300) B 3.7 (2.0–7.1) 1) Somaticillness 2.8 (1.3–5.7) 1) Somaticillness
Pain in the joints 6.2 (2.8–13.6) 7.5 (2.6–21) B
Pain in the legs 4.4 (2.2–8.9) 3.8 (1.6–9.3) B
Overweight 2.0 (1.0–4.0) 1.4 (0.6–3.6) B
Breathlessness 8.9 (3.2–25) 12.1 (3.3–45) B
Dizziness 10.1 (4.2–24) 11.4 (3.8–34) B
Impaired hearing 3.0 (1.3–6.8) 3.1 (1.0–9.5) B
Eye problem 4.2 (1.9–9.1) 3.4 (1.3–9.0) B
Loss of weight - - C 1.7 (0.9–3.0) 1) Miscellaneous
Bad appetite - - C
Feeling cold 7.3 (3.1–17) 7.0 (2.6–19) C
Difficulty in passing urine 9.6 (2.0–47) 9.1 (1.4–59) C
Back ache 4.4 (2.4–8.2) 2.0 (0.9–4.3) D 2.9 (1.6–5.2) 1) Ache/fatigue 4.3 (2.1–8.7) 1) Ache/fatigue
Headache 6.3 (3.4–12) 4.1 (1.9 – 9.1) D
Sweating 3.6 (1.7 – 7.4) 3.3 (1.3 – 8.5)
Coughing 2.0 (0.98–4.2) 1.7 (0.6–4.4)
1) Reference group (OR = 1) is those coded 0 in each factor
Consulters versus non consulters
Consulters and non-consulters among those with persistent FGID were compared regarding their complaints, as reported in the CSQ. The proportion (0–1.0) per complaint is shown in Figure 4. There were no statistically significant differences between the consulters and non-consulters for any of the complaints (adjusted for age and sex, p > 0.0016).
Figure 4 Proportion of complaints from the Complaint Score Questionnaire among those with FGID, divided into Consulters and Non-Consulters. Proportion (0–1.0) of complaints from the Complaint Score Questionnaire (CSQ) among those with FGID, divided into Consulters and Non-Consulters (n = 141). None of the variables showed a significant difference between Consulters and Non-Consulters for P values less than 0.0016, tested by a sex and age adjusted logistic regression.
ASQ symptom severity and consulting behaviour vs CSQ factors
From the ASQ, the mean grades of GI symptom severity for affirmative symptoms were analyzed against the three final CSQ factors from Table 3 (psychological illness, somatic illness, ache/fatigue) and for age, sex and consulting behaviour, as shown in Table 4. The analysis showed an obviously higher risk of increased GI symptom severity for consulters (OR 12.3) and for psychological illness (OR 4.5), while somatic illness and ache/fatigue had a low risk, with the 95% CI close to 1.0. From the ASQ, consulting behaviour was analyzed for the CSQ factors psychological illness, abdominal illness, age and sex (see Table 5). The analysis showed a greater chance of being a consulter for abdominal illness (OR 2.0) and psychological illness (OR 2.2).
Table 4 Odds ratios of graded GI symptom severities in the ASQ for consulting, psychological illness, somatic illness and ache/fatigue factors. Odds ratios (OR, with 95% confidence intervals (CI)) of graded (0,1,2) GI symptom severity in the ASQ for consulting, psychological illness, somatic illness and ache/fatigue factors, age and sex, for both FGID and SSF (n = 232). Ordinal logistic regression.
Variable OR (CI)
Psychological illness low 1
high 4.5(2.4–8.4)
Somatic illness low 1
high 2.0(1.1–3.8)
Ache/fatigue low 1
high 2.1(1.1–4.2)
Consulters no 1
yes 12.3(6.3–23.9)
Age (continuous) 0.96 (0.94–0.99)
Sex female 1
male 0.9 (0.5–1.7)
Table 5 Odds ratio of consulting for abdominal complaints in the ASQ, 1995, for psychological illness, abdominal illness, age and sex, for both FGID and SSF. Odds ratio (OR, with 95% confidence intervals (CI) of consulting for abdominal complaints in the ASQ for 1995 for psychological illness, abdominal illness, age and sex for both FGID and SSF (n = 232). Logistic regression. Somatic illness and ache/fatigue were excluded in the final model because they showed no significance in the prior step.
Variable OR (CI)
Psychological illness low 1
High 2.2 (1.2–4.0)
Abdominal illness low 1
High 2.0 (1.1–3.8)
Age (continuous) 1.0 (0.99–1.0)
Sex female 1
Male 0.7 (0.40–1.3)
Discussion
This study shows that among subjects with longstanding FGID, there is a remarkably high prevalence of psychological illness and also of non-GI somatic complaints. These are present regardless of whether the subjects have consulted their doctor about their GI problems, and are more severe in subjects with persistent FGID. Although FGID was more common in women, the consultation rates in sufferers were similar for males and females and were not age-related. Only about a quarter of the sufferers had never consulted their doctor.
We consider that our findings can be generalized to the whole population, as the study groups were sampled from a well-defined and thoroughly investigated population, most of whom had participated in prior studies [28,29]. The original study base was formed in 1988 from the Swedish National Population register, which guarantees complete coverage of all citizens. There were no differences in age, gender or education level between the study and the sample groups or between the population samples from 1988 and 1995. Also, the proportions of those reporting symptoms explicable in terms of an organic disease have been shown to be insignificant [13]. Thus, the FGID subjects and the symptom-free subjects would seem to be representative.
The validity of the research tool is a potential source of bias. However, for symptom reporting, only well-validated questionnaires were used. Both the ASQ and the CSQ have been adequately validated [28-30]. The psychological illness factor identified in this study embraces symptoms of both "neurotic" and "personality" dimensions, listed in Table 3, and the outcome seems plausible. The questionnaire used to assess medical knowledge was simple and straightforward, with kappa values per question 1 = 0.70, 2 = 0.89, 3 = 0.47, 4 = 0.78, 5 = 0.80, 6 = 0.70 when repeated within a week, considered acceptable for all with some reservation in the 0.47 case [53].
The definitions of dyspepsia and IBS used in this study were those used in the original study from 1988 [31], when the Rome II criteria [11] were unavailable. We opted to retain our original study definitions despite ongoing changes.
Our definition of dyspepsia was more restricted in terms of symptoms than the Rome II definition, but wider in terms of abdominal location, as not only epigastric but also mid abdominal symptoms were included. The IBS definition used was in accordance with the Rome II criteria [22]. Consequently, we consider the overall prevalence of FGID in this study and the concordance on an individual level to be applicable within today's definitions [11].
Subjects with FGID were on average younger than controls without FGID, which may be expected as the prevalence of dyspepsia and IBS is generally higher in younger age groups [32]. This study was not a case control study, however, but a study of all subjects with FGID and SSF within the cohort. Any differences caused by this grouping strategy were controlled for in the analysis by gender- and age-adjusted logistic regression. There was a particular association between FGID and psychological illness, although "fibromyalgia-like" symptoms (ache and fatigue) [49] and other somatic complaints were also common. Previous outpatient studies have shown that IBS is associated with psychiatric illnesses such as depression [34], dysthymia [35] and anxiety [36], and similar findings are reported for dyspepsia [50,51]. Greater mental pathology has been reported for consulters, i.e. patients, than non-consulters with IBS [44,45,52], but this has not been demonstrated convincingly in dyspepsia subjects [53].
Healthcare seeking behaviour is complex and until now it has been studied largely in patient samples. Nyrén et al. found that patients with non-ulcer dyspepsia had an excessive need for sick leave compared with ulcer patients, but that the predominant reason for leave was related to musculoskeletal rather than abdominal symptoms [41]. Kettell et al. [42] found that severity of abdominal pain and anxiety about the seriousness of their condition were important factors in patients consulting for IBS [42]. There seem to be no differences in the use of healthcare services or co-morbidity status between the year before and the year after diagnosis of IBS [43,46].
Although this study was focused on total co-morbidity, in terms of consultation, other factors associated with FGID consultation were also considered. The total consultation pattern for the subjects will be published elsewhere. In essence, effective treatment of patients with FGID involves not only addressing GI discomfort, but also considering mental and somatic complaints such as depression and exhaustion.
Conclusion
In the present study, psychological illness proved to be an important co-morbidity factor among subjects with FGID, and the severities of the two were linked. We cannot conclude anything about the cause of the relationship. The presence of psychological illness was also associated with a greater need for medical consultation. Fear of life-threatening illness has been reported to be an important reason for consultation in FGID [1], and this worry and anxiety needs to be taken into account when attempting to manage FGID successfully. Somatic co-morbidity was found to be a less important reason for consultation, although a high proportion of subjects with FGID (77%) in our study were taking analgesics while only 32% used specific gastrointestinal medication.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
TÅ participated in the study design and checking the coding of the data, carried out the construction of the "medical knowledge questionnaire" including a test-retest survey, distributed and collected the questionnaires, called the participants to his health centre, supported the other health centers personnel, performed all the statistical analysis, participated in writing the manuscript, and submitted the final manuscript. LA carried out the study design and contributed to the statistical analysis and the writing of the manuscript. KS contributed to checking the coding of the data and participated in the study design. SJ contributed help with the statistical analysis.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This work was supported by the personnel of the Primary Health Care units of Östhammar, the Uppsala County Council and Anna Hedin Ph D, National Institute of Public Health, Sweden. We also thank Madeline Frame for assistance with the manuscript preparation.
==== Refs
Jones R Lydeard SE Hobbs FD Kenkre JE Williams EI Jones SJ Jones Repper JA Caldow JL Dunwoodie WM Bottomley JM Dyspepsia in England and Scotland Gut 1990 31 401 5 2338264
Agreus L Svardsudd K Talley NJ Jones MP Tibblin G Natural history of gastroesophageal reflux disease and functional abdominal disorders: a population-based study Am J Gastroenterol 2001 96 2905 14 11693325 10.1111/j.1572-0241.2001.04680.x
Agreus L Natural history of dyspepsia Gut 2002 50 iv2 9 11953337
Talley NJ Zinsmeister AR Schleck CD Melton LJ III Dyspepsia and dyspepsia subgroups: a population-based study Gastroenterology 1992 102 1259 68 1551533
Talley N Boyce P Jones M Identification of upper and lower gastrointestinal symptom groupings in an urban population Gut 1998 42 690 95 9659166
Agréus L Talley NJ Dyspepsia: current understanding and management Annu Rev Med 1998 49 475 93 9509276 10.1146/annurev.med.49.1.475
McDougall NI Johnston BT Kee F Collins JS McFarland RJ Love AH Natural history of reflux oesophagitis: a 10 year follow up of its effect on patient symptomatology and quality of life Gut 1996 38 481 6 8707073
Agreus L Borgquist L The cost of gastro-oesophageal reflux disease, dyspepsia and peptic ulcer disease in Sweden Pharmacoeconomics 2002 20 347 55 11994044
Leong SA Barghout V Birnbaum HG Thibeault CE Ben-Hamadi R Frech F Ofman JJ The economic consequences of irritable bowel syndrome: a US employer perspective Arch Intern Med 2003 163 929 35 12719202 10.1001/archinte.163.8.929
Drossman DA Corazziari E Thompson WG Talley NJ Whitehead W The Functional Gastrointestinal Disorders 2000 Second Degnon Associates, McLean, Va, USA
Talley N Stanghellini V Heading R Koch K Malagelada J Tytgat G Functional Gastroduodenal Disorders Gut 1999 45 1137 42
Whitehead WE Gibbs NA Li Z Drossman DA Is functional dyspepsia just a subset of the irritable bowel syndrome? Baillieres Clin Gastroenterol 1998 12 443 61 9890081 10.1016/S0950-3528(98)90017-3
Agreus L Svardsudd K Nyren O Tibblin G Irritable bowel syndrome and dyspepsia in the general population: overlap and lack of stability over time Gastroenterology 1995 109 671 80 7657095
Agreus L Socio-economic factors, health care consumption and rating of abdominal symptom severity. A report from the abdominal symptom study Fam Pract 1993 10 152 63 8359604
Haycox A Einarson T Eggleston A The health economic impact of upper gastrointestinal symptoms in the general population: results from the Domestic/International Gastroenterology Surveillance Study (DIGEST) Scand J Gastroenterol Suppl 1999 231 38 47 10565622
Westbrook JI McIntosh J Talley NJ Factors associated with consulting medical or non-medical practitioners for dyspepsia: an Australian population-based study Aliment Pharmacol Ther 2000 14 1581 8 11121905 10.1046/j.1365-2036.2000.00878.x
Nyren O Lindberg G Lindstrom E Marke LA Seensalu R Economic costs of functional dyspepsia Pharmacoeconomics 1992 1 312 24 10146996
Stanghellini V Relationship between upper gastrointestinal symptoms and lifestyle, psychosocial factors and comorbidity in the general population: results from the Domestic/International Gastroenterology Surveillance Study (DIGEST) Scand J Gastroenterol Suppl 1999 231 29 37 10565621
Haug TT Mykletun A Dahl AA Are anxiety and depression related to gastrointestinal symptoms in the general population? Scand J Gastroenterol 2002 37 294 8 11916191 10.1080/003655202317284192
Koloski NA Talley NJ Boyce PM Epidemiology and health care seeking in the functional GI disorders: a population-based study Am J Gastroenterol 2002 97 2290 9 12358247 10.1111/j.1572-0241.2002.05783.x
Agreus L Svardsudd K Nyren O Tibblin G Reproducibility and validity of a postal questionnaire. The abdominal symptom study Scand J Prim Health Care 1993 11 252 62 8146509
Agreus L Talley NJ Svardsudd K Tibblin G Jones MP Identifying dyspepsia and irritable bowel syndrome: the value of pain or discomfort, and bowel habit descriptors [In Process Citation] Scand J Gastroenterol 2000 35 142 51 10720111 10.1080/003655200750024308
Agréus L Svärdsudd K Tibblin G Nyrén O The epidemiology of dyspepsia: Symptom clusters and demographic characteristics Gastroenterology 1991 100 A1 2007463
Talley NJ Working Team Report. Functional Dyspepsia: A Classification with Guidelines for Diagnosis and Management Gastroenterology International 1991 4 145 160
Tibblin G Tibblin B Peciva S Kullman S Svardsudd K "The Goteborg quality of life instrument" – an assessment of well-being and symptoms among men born 1913 and 1923. Methods and validity Scand J Prim Health Care Suppl 1990 1 33 8 2100362
World Health Organization WHO/46 Press Release: World Health Organization 1997
StataCorp Stata Statistical Software: Release 80 2003 College Station, TX: Stata Corporation
Agréus L Svärdsudd K Nyrén O Tibblin G Reproducibility and validity of a postal questionnaire. The abdominal symptom study Scand J Prim Health Care 1993 11 252 62 8146509
Agréus L The Abdominal Symptom Study. An Epidemiological Survey of Gastrointestinal and Other Abdominal Symptoms in the Adult Population of Östhammar, Sweden PhD thesis 1993 Uppsala University, Department of Family Medicine
Tibblin G Cato K Svärdsudd K Goteborg quality of life study of men born in 1913 and 1923 – age, sex, job satisfaction and cardiovascular diseases Scand J Prim Health Care 1990 1 39 45
Agréus L Nyrén O Svärdsudd K Tibblin G Ont i magen-En epidemiologisk studie om bukbesvär i Östhammars kommun Svenska Läkarsällskapets Hygea 1990 99 90
Agréus L Svärdsudd K Talley NJ Jones MP Tibblin G Natural history of gastroesophageal reflux disease and functional abdominal disorders: a population-based study Am J Gastroenterol 2001 96 2905 14 11693325 10.1016/S0002-9270(01)03242-7
Gupta S Masand PS Kaplan D Bhandary A Hendricks S The relationship between schizophrenia and irritable bowel syndrome (IBS) Schizophr Res 1997 23 265 8 9075306 10.1016/S0920-9964(96)00099-0
Masand PS Kaplan DS Gupta S Bhandary AN Nasra GS Kline MD Margo KL Major depression and irritable bowel syndrome: is there a relationship? J Clin Psychiatry 1995 56 363 7 7635853
Masand PS Kaplan DS Gupta S Bhandary AN Irritable bowel syndrome and dysthymia. Is there a relationship? Psychosomatics 1997 38 63 9 8997118
Lydiard RB Fossey MD Marsh W Ballenger JC Prevalence of psychiatric disorders in patients with irritable bowel syndrome Psychosomatics 1993 34 229 34 8493304
Sullivan G Blewett AE Jenkins PL Allison MC Eating attitudes and the irritable bowel syndrome Gen Hosp Psychiatry 1997 19 62 4 9034814 10.1016/S0163-8343(96)00106-5
Walker EA Gelfand AN Gelfand MD Green C Katon WJ Chronic pelvic pain and gynecological symptoms in women with irritable bowel syndrome J Psychosom Obstet Gynaecol 1996 17 39 46 8860885
Woodman CL Breen K Noyes R JrMoss C Fagerholm R Yagla SJ Summers R The relationship between irritable bowel syndrome and psychiatric illness. A family study Psychosomatics 1998 39 45 54 9538675
Sperber AD Atzmon Y Neumann L Weisberg I Shalit Y M Abu-Shakrah Fich A Buskila D Fibromyalgia in the irritable bowel syndrome: studies of prevalence and clinical implications Am J Gastroenterol 1999 94 3541 6 10606316 10.1111/j.1572-0241.1999.01643.x
Nyrén O Adami HO Gustavsson S Lööf L Excess sick-listing in nonulcer dyspepsia J Clin Gastroenterol 1986 8 339 45 3760516
Kettell J Jones R Lydeard S Reasons for consultation in irritable bowel syndrome: symptoms and patient characteristics Br J Gen Pract 1992 42 459 61 1472391
Ruigomez A Wallander MA Johansson S Garcia Rodriguez LA One-year follow-up of newly diagnosed irritable bowel syndrome patients Aliment Pharmacol Ther 1999 13 1097 102 10468687 10.1046/j.1365-2036.1999.00576.x
Ballenger JC Davidson JR Lecrubier Y Nutt DJ Lydiard RB Mayer EA Consensus statement on depression, anxiety, and functional gastrointestinal disorders Irritable bowel syndrome, anxiety, and depression: what are the links? J Clin Psychiatry 2001 62 48 51 12108822
Lydiard RB Irritable bowel syndrome, anxiety, and depression: what are the links? J Clin Psychiatry 2001 62 38 47 12108820
Whitehead WE Palsson O Jones KR Systematic review of the comorbidity of irritable bowel syndrome with other disorders: what are the causes and implications? Gastroenterology 2002 122 1140 56 11910364
Jarrett ME Burr RL Cain KC Hertig V Weisman P Heitkemper MM Anxiety and depression are related to autonomic nervous system function in women with irritable bowel syndrome Dig Dis Sci 2003 48 386 94 12643620 10.1023/A:1021904216312
Aro P Ronkainen J Storskrubb T Bolling-Sternevald E Talley NJ Agréus L Prevalence of symptoms and upper endoscopic findings in a random adult population. A report from Kalixanda study Gastroenterology 2001 120 A-231
Wolfe F Smythe HA Yunus MB Bennett RM Bombardier C Goldenberg DL Tugwell P Campbell SM Abeles M Clark P The American College of Rheumatology 1990 Criteria for the Classification of Fibromyalgia. Report of the Multicenter Criteria Committee Arthritis Rheum 1990 33 160 72 2306288
Talley NJ Fung LH Gilligan IJ McNeil D Piper DW Association of anxiety, neuroticism, and depression with dyspepsia of unknown cause. A case-control study Gastroenterology 1986 90 886 92 3949118
Kane F JrStrohlein J Harper RG Nonulcer dyspepsia associated with psychiatric disorder South Med J 1993 86 641 6 8506484
Drossman DA McKee DC Sandler RS Mitchell CM Cramer EM Lowman BC Burger AL Psychosocial factors in the irritable bowel syndrome. A multivariate study of patients and nonpatients with irritable bowel syndrome Gastroenterology 1988 95 701 8 3396817
Sackett DL A primer on the precision and accuracy of the clinical examination JAMA 1992 267 2638 44 1573753 10.1001/jama.267.19.2638
| 15892883 | PMC1156899 | CC BY | 2021-01-04 16:27:57 | no | BMC Med. 2005 May 13; 3:8 | utf-8 | BMC Med | 2,005 | 10.1186/1741-7015-3-8 | oa_comm |
==== Front
BMC Med EducBMC Medical Education1472-6920BioMed Central London 1472-6920-5-171590451710.1186/1472-6920-5-17Research ArticleThe transition from learner to provider/teacher: The learning needs of new orthopaedic consultants McKinstry Brian [email protected] Malcolm [email protected] Katy [email protected] Stuart [email protected] General Practice Division, Community Health Sciences, University of Edinburgh, 20 West Richmond St., Edinburgh UK2 Department of Orthopaedics, Royal hospital for Sick Children, Sciennes Road, Edinburgh, UK3 NHS Education, Lister Institute, 11, Hill Square Edinburgh UK2005 17 5 2005 5 17 17 30 1 2005 17 5 2005 Copyright © 2005 McKinstry et al; licensee BioMed Central Ltd.2005McKinstry et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Given the relatively sudden change from learner to teacher-provider that new consultants experience and the likely clinical and managerial challenges this may pose, there is a relative dearth of research into the problems they may have in relation to their new roles, or how supported they feel by senior colleagues acting in a mentoring role. This research sought to determine new consultants views on the quality and relevance of their training, its relationship to their confidence in clinical and managerial skills and their views on mentorship by senior colleagues.
Methods
Detailed postal questionnaire to new consultants using open and closed questions.
Open questionnaire to established consultants to validate new consultant responses.
Results
Respondents felt their clinical training was good and were generally confident in most clinical skills although some perceived deficiencies in more complex procedures and specialist areas. Most lacked confidence in many managerial skills. These perceptions were verified by established consultants. Although no relationship was found between total training time or quality of training with confidence, extra training in specific sub-specialities improved confidence in these areas. While most established consultants thought that mentorship would be useful for new consultants, only 52% of them shared this view.
Conclusion
Training and experience in management should be given greater emphasis. There may be a need for specific, targeted training in complex procedures for doctors who experience lack of confidence in these areas. Mentorship should be offered to new consultants and recognised in the job-plan of the new consultant contract.
==== Body
Background
Every year in the UK over a thousand doctors move from training grades to consultant posts. Educational research tends to focus on training grades and there is a relative dearth of information on the problems these new consultants have in relation to their new roles particularly their perceived learning needs in relation to clinical and management issues. Considering the relatively sudden change from learner to teacher-provider and the likely challenges this may pose, it is not clear to what extent new consultants in their early years feel supported by senior colleagues acting in a mentor or teaching role to aid them with this transition. The recent shortening of training for UK consultant posts following the Calman report [1] in the mid 1990s and the reduction in working hours during training leading up to the full implementation of the European Working Time Directive have led to concerns that new consultants, particularly surgeons, may be less well prepared than their predecessors to face the challenges of their new role [2,3].
Against this background we chose to investigate the problems new consultants may face. We chose to use orthopaedic surgery as a model for the craft specialties because it is considered to have a consistent training program with a relatively unspecialised end-point. It is also one of the best organised surgical specialties, with a careful, annual census and lastly has a significant number of new consultants appointed annually.
National Health Service Education for Scotland (NES) collaborated with the British Orthopaedic Association (BOA) to survey consultants registered with the BOA who had been in post for less than five years. We sought to determine their self-perceived learning needs and how these were related to their training experience. Their views were sought on the quality and relevance of their training experience, the degree to which they felt supported by senior colleagues in their new posts and their views on formal mentorship. In an attempt to triangulate the new consultant responses on self-perceived learning needs, questionnaires were sent to a small number of established consultants for their views on the problems faced by new consultants.
Aims
To determine:
• the self-perceived learning needs of new orthopaedic consultants
• if there is a correlation between perceived length and supervision of training or experience abroad and subsequent self-perceived competence
• whether new consultants feel they were given adequate support from senior colleagues in the early years of consultantship and their attitude to formal mentoring
• the views of new consultants and established consultants on the length and quality of current training
• what changes these doctors believe should be made to orthopaedic training
Methods
Designing the questionnaires
The content of the questionnaires was decided after a series of face-to-face interviews with specialist registrar surgeons within one year of the end of training (SpR) and established and recently appointed consultants. These surgeons showed a preference for a detailed check list questionnaire which allowed space for free text. The senior consultants interviewed were very clear that for them a short more open questionnaire would be preferred. As the questions to new consultants included the sensitive areas of self-assessed competence interviewees stressed that responses should be anonymous.
The survey for new consultants (see additional file 1) covered the following areas:
• extent and perceived quality of their training
• content and quality of time spent abroad
• access to courses and advice about choice
• perceived competency in the main areas of elective and trauma surgery, managerial and communication skills, teaching and research skills
• the availability of advice and help including attitudes to formal mentoring
• perceived learning/developmental needs
• perception of how well training has prepared them for consultantship
• their main learning needs and when these might have been best addressed
The survey of established consultants (see additional file 2) was an open unstructured questionnaire which asked;
• What they thought were the main problems faced by new consultants
• Their views on the current breadth and depth of training
• How they thought training might be improved
• Their views on the support readily available for new consultants
• Their views on the utility of formal mentoring to new consultants
The survey was piloted on a group of ten new consultants and SpRs. They found the questionnaire relatively straightforward and easy to complete.
Distributing the questionnaire
Using the database of the BOA the questionnaire was sent to a randomly selected group of 150 consultants who had taken up their first consultant post in the last five years. The decision to use the 5-year cut off was pragmatic in that the BOA has a separate register of this group. In addition we sent the established consultant questionnaire to 100 randomly chosen consultants of greater than five years standing selected from the BOA database. One reminder was sent.
Analysis
The new consultant questionnaire responses were entered into SPSS. Results were analysed using non-parametric tests (Mann-Whitney or Chi-Square as appropriate). The established consultant responses from open questionnaire were analysed for themes by BM and checked by KE independently.
Results
Response rates
90/150 (60%) 'new' consultants and 30/100 (30%) 'established' consultants replied. Questionnaires were generally very well completed with frequent comments made.
New consultants training experience
The majority of this group graduated between the years 1986 and 1990 (range 1978–1990). Most graduated in the UK (67%), spent 4 years (range 1–7) as a Senior House Officer (SHO) (a junior training grade in the UK), and 6 years (range 3.5–12) as Specialist Registrar (SpR) (a senior training grade in the UK established in the mid 1990s). The majority of these were spent in 'mixed' orthopaedic/trauma posts (58%), with smaller numbers in specialist adult elective orthopaedics (16%), specialist trauma (11%), paediatrics (10%) and research (10%).
Length of training and quality of supervision
Consultants generally thought that the amount of time gaining experience was 'just right'. A very small number (4.4%) thought that there had been too little time in part of their training. Most consultants thought they had received good supervision in their posts although 18% said they 'could have done with more' and 15% described their supervision as 'often insufficient'. None described it as 'grossly insufficient'.
Time spent abroad
59 (67%) of the consultants had spent time abroad during their training. The most common destinations were Australia (22), followed by the USA (12), Ireland (7) and Canada (6). Most spent between 6 and 18 months abroad. The majority (77%) spent their time in practical tasks such as operating; 20% mainly observed; and 4% did some research. All but one described the experience as very positive. However we found no relationship between having experienced time training abroad and current confidence in any clinical, teaching or managerial skill.
Guidance on and availability courses
SpRs in the UK are encouraged to undertake thirty study days a year most of which is spent on organised courses run by the Royal Colleges of Surgeons, BOA or similar organisations. The courses are very varied including sub-specialist subjects such as hand surgery or management topics such as appraisal.
New consultants were unhappy with the guidance they had been given during training with regard to what type of postgraduate courses available both in the UK and abroad would be most appropriate for them to attend. Particularly some felt that advice on which point in their career path to attend a particular type of course would have been useful. Most described the advice as poor or non-existent (73%). Although the majority (72%) felt that they were able to attend courses that were of value to their future career, a sizeable proportion were not, mainly due to either lack of time (19%) or finance (12%). Several commented that courses should be more clearly targeted for specific groups at different stages of training.
When asked; 'Did your clinical training prepare you adequately for your current post?' 77 (90%) concurred. Almost all additional comments rated the clinical training as good.
How do new consultants rate their skills?
Consultants generally were confident in most practical procedures (table 1) with a few exceptions (largely in specialist areas) in which a proportion reported that they were 'not very' or 'not at all' competent (spinal surgery (79%), paediatric surgery (56%), shoulder surgery (46%), sports medicine (36%) and hand surgery 48%)). However, there was clear lack of perceived skill in a wide range of management areas (see table 2) with half or more of all respondents reporting they were 'not very ' or 'not at all' competent in negotiation, appraisal, business planning, medicolegal skills and managing private practice and, perhaps less importantly for most consultants, some research skills. However most regarded their teaching skills (table 3) and all regarded their communication skills as very or quite good.
Table 1 How surgeons responded to the question; 'How would rate your competence in the following skills?'
Degree of competence (valid%)
Skill Very Quite Not very Not at all missing
Arthroplasties
Knee 48 (55) 37 (42) 3 (3) 0 (0) 2
Hip 45 (51) 43 (49) 0 (0) 0 (0) 2
Other elective procedures
Knee 34 (42) 43 (53) 3 (3) 1 (1) 9
Hip 25 (32) 45 (57) 7 (9) 1 (1) 12
Shoulder 10 (12) 37 (43) 35 (41) 4 (5) 4
Spinal 2 (22) 16 (19) 40 (48) 26 (31) 6
Hand 15 (7) 53 (27) 17 (48) 0 (0) 5
Amputation 15 (17) 48 (56) 19 (22) 4 (5) 4
Arthroscopy 49 (54) 37 (41) 2 (2) 0 (0) 2
Sports medicine 21 (24) 34 (40) 28 (33) 3 (3) 1
Trauma 58 (66) 29 (33) 1 (1) 0 (0) 2
Paediatric 11 (13) 26 (30) 44 (51) 5 (51) 4
Table 2 How surgeons responded to the question; 'How would rate your competence in the following skills?'
n = 90 Perceived degree of competence (%)
Skill Very Quite Not very Not at all missing
Management
Negotiation 7 (8) 37 (42) 38 (43) 6 (7) 2
Business planning 2(2) 17(19) 50 (56) 20 (22) 1
Financial skills 3 (3) 14 (16) 48 (54) 24 (27) 1
Leadership 20 (23) 54 (60) 12 (14) 3 (3) 1
Appraisal 6 (7) 24 (27) 42 (48) 16 (18) 2
Presentations 37 (42) 46 (52) 5 (6) 1 (1) 1
Medico-legal (court work, reports) 8 (9) 28 (32) 36 (40) 17 (19) 1
Risk management 3 (3) 33 (38) 41 (47) 11 (13) 2
Managing private practice 4 (5) 14 (16) 39 (44) 32 (36) 1
Research Skills
Applying for grants 6 (7) 18 (21) 33 (38) 31 (35) 2
Running research projects 10 (11) 33 (37) 29 (33) 17 (19) 1
Literature review 26 (29) 51 (57) 8 (9) 4 (5) 1
Writing up 23 (26) 40 (45) 19 (21) 7 (8) 1
Table 3 How surgeons responded to the question; 'How would rate your competence in the following skills?'
n = 90 Perceived degree of competence (%)
Skill Very Quite Not very Not at all missing
Teaching skills
Small group teaching 47 (53) 39 (44) 3 (3) 0 (0) 1
Mentoring skills 18 (20) 58 (65) 11 (12) 2 (2) 1
Appraising students 18 (20) 49(55) 20 (23) 2 (2) 1
Neither the total time spent in training nor the perceived supervision and guidance in training was related to self-rated competence. However, those who had spent longer (>12 months) in a sub-specialist area such as paediatric orthopaedics were more likely to perceive themselves as very or quite skilled in that area (16/18 (89%) v 11/19 (58%) p < 0.05). Those who had reported more difficulty in attending 'suitable' courses showed a trend significant at the p < 0.05 level for less confidence in arthroscopy and sports medicine. Those doctors (n = 9) who felt unprepared for consultantship after their training showed a trend for scoring their confidence lower across most clinical skills. They were significantly more likely to describe their communication skills with colleagues as not very good or poor (4/9(44%) v 1/76(1%) p < 0.001).
When asked what were their main learning needs at the moment, the commonest response was in the area of management skills, particularly negotiation skills. Many commented that they had not been at all prepared for the amount of administration and negotiation they encountered as a new consultant. The clinical skills identified as learning needs were most frequently revision arthroplasty (10% of all respondents), followed by spinal surgery, shoulder arthroscopy, foot, ankle and hand surgery. New consultants thought these management skills and specialised clinical learning needs should be best addressed in late SpR training or during fellowship years.
Availability of support when appointed as a new consultant
Most new consultants (86%) thought that help was readily available when they took up their posts although 44% did have concerns regarding this when they first became a consultant. Some suggested that such help needed to be more structured and that consultants themselves needed to recognise that they should occasionally ask for help. When asked if they would have liked some form of mentoring from a senior colleague when they began working as a consultant, 52% replied that they would have. Others felt that mentoring might be like 'being a Senior Registrar again' or that some senior colleagues were 'too out of date to be of help'. A small number was concerned that it might diminish them in front of senior colleagues. Many felt that the support they received informally meant mentoring was unnecessary.
The views of established consultants
Established consultants (EC) confirmed that, while generally well trained clinically, new consultants lacked managerial and administrative skills. They particularly pin-pointed the importance of negotiation skills, for problems as diverse as getting an office, negotiating suitable theatre and out-patient clinic time with powerful senior colleagues, or negotiating a balance between teaching, private practice and their National Health Service (NHS) clinical commitments:-
"Suddenly thrust into having to negotiate for all they want or need to carry out their plans with no previous experience of management" (EC)
"The new consultants who, in my experience, have shown greatest stress and difficulty have been those who over-stretched themselves with non-NHS work. Sadly only the NHS is blamed for the stress and is expected to provide the solution."(EC)
Established consultants drew a distinction between training which they generally thought good (although occasionally too specialised) and experience, which they were concerned new consultants lacked. Like the new consultants they identified complex clinical problems (e.g. revision arthroplasty) and rare problems as difficult areas. In these cases advice from more experienced surgeons was seen as important:-
"Although many consultants may be knowledgeable in the general sense, they have little experiential knowledge and little practical sense. This is mainly due to reduction in training time, and, in addition, reduction in hours of training."(EC)
"They are expected (in DGHs) to be masters of all specialties within orthopaedics, and generally their training will be deficient in one or more sub-specialties."(EC)
Like the new consultants, experienced consultants felt that more time should be spent on management training in the final years of a program:-
"Introduce training on time management, paper management and teaching techniques."
"5th and 6th year SpRs should have to attend directorate meetings to give them a taste of what is to come." (EC)
Established consultants, however, felt that training needed to be longer than currently, some suggesting clinical fellowships, junior consultant grades and proleptic appointments in addition to current training.
In terms of support, most felt that generally this was good for new consultants. Several perceived that a bigger problem was that of the new consultant who doesn't seek help when it is required:-
"Yes. Those that know to ask for help and seek it are not the problem. Those that don't ask are." (EC)
But two commented:-
"No, it is very rare that newly appointed consultants have access. Even if there is access one tends to be hesitant. We should give them a lot of support, guidance and make them feel very comfortable so they can come forward and ask for help." (EC)
".....there is a requirement that a 'no blame' culture exists for this to happen fairly, and without fear or favour. Unfortunately, this is not always the case." (EC)
All but one established consultant thought that some form of mentoring was desirable. Many thought that it happened informally already. Again it was suggested that proleptic appointments could be the basis of a mentored transition to independent practice. Some saw appraisal by a senior colleague as fulfilling this role. One saw advantages to both established and new consultants in this process. There were some caveats, however:
"Yes. Mentors would have to be carefully chosen. A consultant in the same subspecialty is best placed to give advice on difficult clinical cases, but can be less than helpful because he sees a rival and wants to defend his own territory" (EC)
"...better to encourage support of all colleagues" (EC)
Discussion
A 60% return, while good for a postal questionnaire, begs the question as to how representative the results may be. In particular although well completed and containing much valuable information, the 30% return from the established consultants must be interpreted with particular caution as they may represent the more attentive and professional element. However the main purpose of this part of the survey was to provide some validation of the self perceived skills of the new consultants and generally this was provided by the replies we received.
Several studies suggest that self-assessment of skills is not often corroborated by external observation [4]. There is, however, evidence that highly detailed competencies such as we employed, are more likely to be more accurately self-rated than global skills [4,5]. Still, some results (for example 100% describing their confidence in their patient communication skills as very or quite good when there is general evidence to the contrary [6]) raise concerns about the validity of some of these self-assessments. The results should therefore be viewed with some caution. There are undoubtedly more effective ways of assessing the practice of clinical and managerial skills such as audit, 360 degree assessment or perhaps even ethnographic studies. It may be that with the introduction of regular appraisal and audit in the UK such data may become available for future studies.
While new orthopaedic consultants generally feel their clinical training has been good (apart from poor guidance on suitable courses), the transition from teacher to learner/provider clearly presents them with new challenges for which they feel they are not well prepared. This has also been described by others in the UK and abroad [7-9]. These challenges are largely in the area of management although, understandably, complex procedures such as revision arthroplasty represent another area of concern. New consultants were clear that they required additional training in management and some clinical areas, but both they felt that the timing of this was important. They felt training should occur immediately prior to becoming a consultant or shortly afterwards when it was most relevant to them. They also felt that some involvement in management issues near the end of their training period would be helpful.
These observations were largely confirmed by their more experienced colleagues. While established consultants felt that more time was needed in training, this was not the perception of the newer consultants who believed that targeted training and actual experience (for example attending management meetings) within the current training time could address these areas.
The small number of new consultants who reported that they felt unprepared for their role were much more likely to report that their communication with other colleagues was poor. Interpersonal skills have been described as vital for success as a new consultant [8]. This is an area which requires further research to determine if this relationship is causal and if it has the potential for intervention during training.
Although most had felt they had been well supported informally a majority of new consultants and senior consultants were in favour of formal mentorship in the early years of posts, some were very negatively disposed towards the concept. The introduction of consultant appraisal may help in this area although it will not help with the day-to-day problems faced by new consultants. Where mentorship has been made available it has been found to be helpful [8]. Possibly the best approach is to offer and promote mentorship to those who wish it.
Internationally, health care systems are trying to come to grips with reduced working hours for doctors in training [10-12]. There is some reassurance in these results that shortened training has not resulted in serious self-perceived skill deficiencies and this is in keeping with a study of trainee satisfaction with training after the Calman reductions in training time [13]. However, it will be interesting to repeat this survey in the future to see if the further reduction in working time brought about by the European Working Time Directive has the consequences so many fear [14].
While this study focused on orthopaedic surgery, there is some evidence that the generic management skills, identified as being problematical, challenge consultants in other specialties [15]. Clearly this is an area which needs to be addressed in training.
Conclusion
Apart from some concerns about complex and specialist procedures new orthopaedic consultants are generally confident in their clinical competence. However, they perceive important learning needs in management skills. Senior colleagues corroborated these observations. New consultants were generally satisfied with their training but some felt supervision was less than adequate at times and that they had insufficient advice and finacial support for instructional courses. Training in management skills, medicolegal matters and experience of management procedures during training should be given greater emphasis during later SpR years. There may be a need for specific, targeted training in complex orthopaedic procedures for doctors who identify lack of confidence in these areas. Mentorship by senior colleagues should be offered to new consultants, but in a way that is neither prescriptive or demeaning.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BMcK designed the study, carried out preliminary interviews, analysed the data and wrote the paper. MM helped design the study and the questionnaires and helped write the paper. Katy Elliot dispatched the questionnaires and entered the data. SM helped design the study and contributed to the writing of the paper.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional File 1
New consultants' questionnaire.
Click here for file
Additional File 2
Established consultants' questionnaire.
Click here for file
Acknowledgements
We are grateful to Peter Driscol for his help in designing the questionnaire, David Adams for his help in recruitment and to the doctors who took part in helping construct and who responded to the questionnaire.
The study was supported by NHS education and the British Orthopaedic Association
==== Refs
Department of Health A guide to specialist registrar training 1995 London, HMSO
Bates T Curricular training and the new deal Ann R Coll Surg Engl 1996 78 61 62 8687069
Skidmore D Junior surgeons are becoming deskilled as a result of Calman proposals BMJ 1997 314 1281 9154052
Gordon M A review of the validity and accuracy of self assessments in health professions training Academic Medicine 1991 66 762 769 1750956
McKinstry B Peacock H Blaney D Can trainers accurately assess their training skills using a detailed checklist? A questionnaire-based comparison of trainer self-assessment and registrar assessment of trainers' learning needs Education for Primary Care 2004 14 426 431
Cartwright A Windsor J Outpatients and their Doctors: a Study of Patients Potential Patients, General Practitioners and Hospital Doctors 1992 London, HMSO
Johnson G Brown R Howell M Higher specialist training in accident and emergency medicine – Past, present and future Journal Of Accident & Emergency Medicine 1997 14 104 106 9132183
Houghton A Peters T Bolton J What do new consultants have to say? BMJ 2002 325 S145 12424185 10.1136/bmj.325.7372.S145a
Itani KM Liscum K Brunicardi FC Physician leadership is a new mandate in surgical training American Journal of Surgery 2004 187 328 321 15006559 10.1016/j.amjsurg.2003.12.004
Sprangers F The Dutch experience of implementing the European Working Time Directive BMJ 2002 325 S71 12202350 10.1136/bmj.325.7362.S71
Watson PY The working and learning environment for physicians in training in the US BMJ USA 2002 2 76 77
Warshaw A Sarr M The now and future world of restricted work hours for surgeons Surgery 2003 134 1 2 12874571 10.1067/msy.2003.200
Paice E Aitken M Cowan G Heard S Trainee satisfaction before and after the Calman reforms of specialist training: questionnaire survey BMJ 2000 320 832 836 10731174 10.1136/bmj.320.7238.832
Chikwe J de Souza AC Pepper JR No time to train surgeons BMJ 2004 328 418 419 14976074 10.1136/bmj.328.7437.418
Johnson G Brown R Howell M Higher specialist training in accident and emergency medicine – past, present and future J Accid Emerg Med 1997 14 104 106 9132183
| 15904517 | PMC1156900 | CC BY | 2021-01-04 16:30:57 | no | BMC Med Educ. 2005 May 17; 5:17 | utf-8 | BMC Med Educ | 2,005 | 10.1186/1472-6920-5-17 | oa_comm |
==== Front
BMC Med ImagingBMC Medical Imaging1471-2342BioMed Central London 1471-2342-5-21589007210.1186/1471-2342-5-2Research ArticleMolecular mapping of periodontal tissues using infrared microspectroscopy Hynes Allan [email protected] David A [email protected] Angela [email protected] David L [email protected] Michael G [email protected] Kan-Zhi [email protected] Dental Diagnostic and Surgical Sciences, University of Manitoba, Winnipeg, Canada2 Oral Health and Systemic Disease Research Group, University of Louisville School of Dentistry, Louisville, KY, USA3 Institute for Biodiagnostics, National Research Council, Winnipeg, Canada2005 12 5 2005 5 2 2 27 1 2005 12 5 2005 Copyright © 2005 Hynes et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Chronic periodontitis is an inflammatory disease of the supporting structures of the teeth. Infrared microspectroscopy has the potential to simultaneously monitor multiple disease markers, including cellular infiltration and collagen catabolism, and hence differentiate diseased and healthy tissues. Therefore, our aim was to establish an infrared microspectroscopy methodology with which to analyze and interpret molecular maps defining pathogenic processes in periodontal tissues.
Methods
Specific key cellular and connective tissue components were identified by infrared microspectroscopy and using a chemical imaging method.
Results
Higher densities of DNA, total protein and lipid were revealed in epithelial tissue, compared to the lower percentage of these components in connective tissue. Collagen-specific tissue mapping by infrared microspectroscopy revealed much higher levels of collagen deposition in the connective tissues compared to that in the epithelium, as would be expected. Thus inflammatory events such as cellular infiltration and collagen deposition and catabolism can be identified by infrared microspectroscopy.
Conclusion
These results suggest that infrared microspectroscopy may represent a simple, reagent-free, multi-dimensional tool with which to examine periodontal disease etiology using entirely unprocessed tissue sections.
==== Body
Background
Periodontitis is defined by the inflammatory destruction of the supporting structures of teeth, including the periodontal ligament and alveolar bone. Periodontal diseases are generally chronic in nature and usually persist in the absence of treatment [1,2]. These diseases are the result of exposure of the periodontium to dental plaque biofilms that accumulate on the teeth to form bacterial masses at or below the gingival margin [3].
Biomedical utilization of the electromagnetic spectrum of light has revolutionized the practice of medicine over the centuries, most recently through the dramatic healthcare advances afforded by the development of magnetic resonance imaging (see Figure 1). The last region of the spectrum to be applied to the practice of medicine is the infrared region. Infrared (IR) spectroscopy is now being increasingly utilized in multiple biomedical settings [4].
IR spectroscopy can distinguish differences in the characteristics of diverse molecules by probing chemical bond vibrations and use these molecular and sub-molecular profiles to define and differentiate "diseased" and "healthy" tissues [4]. As covalent bonds vibrate, they absorb energy in the form of IR light (see Figure 2). The wavelength of light that is absorbed depends on the nature of the covalent bond (e.g. C=O, N-H), the type of vibration (bending, stretching, etc.), and the environment of the bond. The IR spectrum of a tissue sample can be regarded as molecular fingerprint of the tissue. If this molecular fingerprint is modified by a disease process, then IR spectroscopy can be used to detect and monitor the disease process.
IR microspectroscopy is a relatively new technique in which infrared spectra are observed through a microscope that transmits and detects infrared radiation. IR microspectroscopy has been previously utilized to monitor variations in the catabolism and anabolism of collagen and other components within cardiac tissues and oral cancer [5,6]. Collagens exhibit a series of unique IR absorption bands between 1000 and 1300 cm-1. Specifically, the strong band at 1204 cm-1 has been identified as typical of collagen deposition [5,6]. By integrating the intensity of this absorption band, one can readily plot molecular contour maps that clearly delineate areas of collagen deposition. Inflammation-driven collagen degradation is a hallmark of periodontitis [7,8], with strong dissolution of collagen types I and III observed [9,10].
Particular absorptions are assigned to various functional groups in an attempt to extract biochemical information. Absorptions between 1620–1680 cm-1 are usually attributed to amide I vibration of proteins, while absorptions at 1080 and 1240 cm-1 are attributed to PO2- symmetric and asymmetric stretching vibrations of DNA phosphodiester groups [4]. Using this data qualitative and semi-quantitative information can be extracted from the spectra.
Therefore, we set out to establish an IR microspectroscopic methodology that would allow the analysis and interpretation of molecular maps defining pathogenic processes in unprocessed periodontal tissues. This task is much more difficult in periodontal tissues in comparison to most other tissues due to the following factors; (1) Periodontal soft tissues are very thin (to a few mm); (2) Periodontal tissues excised at surgery are generally small (around 5 mm2); and (3) Excised periodontal tissues are particularly friable. Each of these factors present problems in embedding, sectioning, and orientating unfixed periodontal samples.
Methods
Subjects
Patients referred to the Graduate Periodontics Clinic, University of Manitoba for periodontal therapy and giving written, informed consent were recruited. Inclusion criteria were a clinical diagnosis of chronic periodontitis and a requirement for periodontal surgery at a diseased site (probing pocket depth ≥ 5 mm, bleeding on probing, and clinical attachment loss ≥ 3 mm). Exclusion criteria were tobacco smoking, pregnancy, a requirement for antibiotic prophylaxis prior to periodontal probing, prolonged anti-inflammatory medications within the past 3-months (e.g. NSAIDs, steroids, antibiotics, or immunosuppressants), any systemic condition that may interfere with the study, such as inflammatory diseases, diabetes or blood dyscrasias, and lesions of the gingiva unrelated to plaque-induced periodontal disease. All subjects had received oral hygiene instruction and root planing prior to surgery.
Expired-air Carbon Monoxide measurement
Self-reported non-smoking status was validated at the time of recruitment by analysis of expired-air CO concentrations (PiCO meter, Bedfont Sci., UK and associated PiCO chart software), calibrated according to the manufacturers instructions. Non-smokers were required to exhibit expired-air CO concentrations <10 ppm.
Periodontal tissue collection and processing
On the day of surgery, prior to any anesthesia being administered, clinical measurements were obtained from the surgical site: probing depths, bleeding on probing and clinical attachment level (in that order). Twenty periodontal tissue samples were obtained from patients in good general health who were undergoing surgical treatment for chronic periodontitis. Mid-interproximal tissue was preferred, as these sites show increased clinical and histological signs of inflammation compared to other gingival sites [11]. Periodontal tissue samples were snap frozen in isopentane supercooled in liquid nitrogen and then stored at -80°C.
A small amount of optimal cutting tool embedding media (Miles Inc., Elkhart, IN.) was applied to the tissue samples in order to facilitate attachment to the cryotome. A series of 10 μm sections were then cut, mounted onto an IR transparent barium fluoride window, and air-dried. Adjacent 10 μm samples were mounted and processed for conventional histology including visualization of tissue integrity and inflammatory foci by H & E staining.
Infrared microspectroscopy
Infrared microspectroscopy was performed using a Bruker FTIR spectrometer and IR microscope system equipped with a liquid nitrogen cooled mercury cadmium telluride detector. For IR spectral acquisition, the microscope aperture was closed to allow the IR beam to illuminate an area of tissue measuring 50 μm × 50 μm, thereby masking all other regions of the tissue section. For each 50 μm × 50 μm demarcation, 64 interferograms were collected. The signal was averaged against a blank area as background and Fourier-transformed to generate IR spectra with a nominal resolution of 4 cm-1. After each acquisition, the stage was stepped 50 μm under computer control and the next spectrum acquired. This process was repeated until the complete area of interest was mapped. General biochemical mappings regarding the specific components of proteins, lipid, and DNA were obtained using a chemical imaging method. Specifically, the bands used for protein mapping were the amide I band (1654 cm-1); lipid mapping used the lipid CH stretching vibrations at 2800–3000 cm-1; and DNA mapping employed the PO2- symmetric stretching vibration of DNA phosphodiester groups at 1080 cm-1. To highlight collagen deposition, a unique collagen IR band at 1204 cm-1 was used to generate tissue maps, as we have described previously in cardiac tissues [5]. A digital CCD video camera was coordinated to the IR microscope to record each mapped area for future band and area correlation. Microphotography of the identical mapping position in the adjacent H and E-stained tissue section allowed the comparison of general histological features with the IR mapping data.
Data processing for IR microspectroscopy
Because most IR bands are broad and are composed of overlapping components, it is necessary to pre-process the original spectra by applying a band-narrowing algorithm that separates the individual bands. All data processing was performed using the Cytospec V software package . To permit a useful comparison of the cluster analysis, uniformly pre-treated data was used. All the original IR spectra were converted into second derivative spectra using the Savitzky/Golay algorithm with a 9-point window for the multivariate statistical analysis. Derivative spectra were scaled before the cluster analysis where the sum squared deviation over the indicated wavelengths equals unity (vector normalization) [12]. The unsupervised cluster analysis used Ward's minimum variance algorithm and Euclidean distances as distance measure. The Euclidean distance between spectra is calculated and the pair of spectra with the least distance is grouped to create a cluster. Then the separation between this cluster and all other spectra is calculated and another two closest spectra/clusters are joined to form a new cluster. This procedure continues until all spectra/clusters are combined. In this unsupervised classification no information about the disease state of the samples is needed, only the similarity or dissimilarity of their infrared spectra is used for this classification.
Results
IR microspectroscopic imaging generated molecular tissue maps that provide a spectral signature of the intensity and spatial location of the chemical components of the diseased periodontal tissues examined. A representative microphotograph of periodontal tissue stained with H and E is shown in the left panel of Figure 3, while the spectra generated from defined area by IR microspectroscopy indicated by the square box are shown in the right panel. The H and E staining reveals typical histological constitution of periodontal tissue, i.e. the gingival epithelium comprising the epithelial layer that covers the external surface of the gingival and underlying connective tissue composed of gingival fibers, ground substance, and cells, including neural and vascular elements. Spectra from both the epithelium and connective tissues display the distinctive infrared spectral signatures of specific molecular structures, such as DNA, protein and lipids, as shown in the right panel.
In order to reveal more subtle molecular differences in the periodontal tissue, the average spectra of these areas were generated using cluster analysis, as shown in Figure 4. As expected, both spectra show the signature of typical tissue components, i.e. total protein and collagen-specific signals, lipids (membranes), and nucleic acids. For instance, the absorptions at 1654 cm-1 arise from the amide I C=O stretching vibrations of the peptide groups in proteins [13]. The band at 1080 cm-1 is due to vibration of the phosphodiester groups in DNA that can be used to identify the content of cellular nucleus. The information regarding relative lipid concentration in both tissues can be found at the 1740 cm-1 band that originates from the ester C=O group while collagen substance in the tissue can be traced by looking at the specific band at 1204 cm-1, as marked in the figure [5]. The top panel in Figure 4 is the difference spectrum, generated by subtracting the average spectrum of connective tissue from that of epithelium. The difference spectra help identify the specific molecular components that differ most between the two groups of spectra. Upwardly pointing peaks (positively shaded areas) are indicative of the presence of more of a certain molecular component in the epithelium tissue compared to that in the connective tissue, and vice-versa for the downwardly pointing peaks (negatively shaded areas). There are several important changes in the IR spectra that involve molecular vibrations of proteins, including collagen, lipids, and DNA. As expected, a higher cellular constituent content (protein, lipids and DNA) was found in the epithelial tissue, compared to the lower percentage of these components in the connective tissue. However, the collagen content was much higher in the connective tissue than in the epithelium.
Based upon these molecular differences in periodontal tissues, IR chemical mappings were established and compared with conventional H and E-stained sections, as shown in Figure 5. Specifically, regions of interest in the periodontal tissue were identified using the H and E stained section and a high quality microscope in the visible range. Then, the corresponding features were identified by anatomical landmarks on the unstained adjacent tissue sections; subsequently, the tissue section was transferred to the IR microscope and aligned using the anatomical landmarks, as shown in Figure 5A and 5B. The numbers in the X- and Y-axes indicate the step scans employed in the acquisition of IR mapping. Therefore, a total of 962 spectra in this section were obtained and stored in the hypercube. To visualize the spectral data from the hypercube, the easiest way is to present "horizontal" slices through the hypercube, which displays intensity values at a given wavenumber for all spectral vectors as a false color representation [14]. The intensity value of each spectrum is assigned a color code and displayed against the X and Y coordinates of the spectral element. By examining different "color slices" (i.e., the intensities at different spectral elements), variations in the chemical composition can be detected for various pixels in the hypercube. As demonstrated in Figure 5C, the chemical IR mapping based on protein concentrations in the periodontal tissue revealed highest protein content in the epithelium, which was well correlated with tissue components observed by conventional H and E staining. The cell density in the epithelial tissue is significantly higher than that in the connective tissue (Figure 3). The epithelial cells provide the major source of protein noted in the IR spectra, especially at the level of functional group mapping. However, although the epithelial tissue contains more total protein than the connective tissue, the latter contains much more collagen. Other important cellular chemical mappings were also generated from the IR maps, as shown in Figure 6A–D. These false color IR molecular have the familiar appearance of traditional histological sections.
Finally, we employed a multivariate statistical method, CLA (cluster analysis), to verify that the epithelial cells can be readily separated from connective tissues, as shown in Figure 7. The advantage of this computational approach is that it operates unsupervised: no prior knowledge of reference spectra of any tissue type is required, and characteristic spectra and membership of the families of spectra identified by the characteristic spectra are established, regardless of the number of different spectral families and the differences between spectral families [14]. This cluster analysis uses a large portion of the spectrum, rather than a few selected points, regions, or integrated regions, to determine whether spectra are related or not. Because the differences between spectra of normal and diseased tissue are generally small, this method offers the advantage of emphasizing differences in the overall shape, rather than selected intensities, for the discrimination of spectra. In this study, the entire range from 900–1800 cm-1 was used and nine groups based on their spectral features were generated.
In the final step of cluster analysis, all spectra in the same family are assigned a color code, and small, colored squares are drawn at the pixel coordinates of all spectra belonging to the same family to produce our false color maps. As shown in Figure 7B, the blue color (line II) indicates that the spectra from this group were in the same family, originating from the epithelium, like the IR protein mapping of the same tissue shown in Figure 7C. The yellow color in Figure 7B (line I) represents those similar spectra generated from connective tissue. Panel A in Figure 7 displays the average spectrum produced by the cluster analysis with color codes. In other words, the average spectrum from epithelium and connective tissues are in blue and yellow, marked II and I, respectively, and also correspond to the spectra analyzed in full detail in Figure 4.
Discussion
IR microspectroscopy possesses numerous advantages over traditional approaches to pathology. Fixation and staining of tissues are not required before histological viewing, little or no sample preparation is necessary and only minimal technical expertise is required by the operator. The method lends itself readily to rapid, high-volume repetitive measurements. IR microspectroscopy is non-destructive, meaning that the sample may be saved and passed on for further measurements if required. Furthermore, IR microspectroscopy provides information concerning the molecular structure of the tissue and multiple analytes may be measured simultaneously from a single spectrum. This combination of features is simply unavailable from visual microscopy.
Our initial data suggest that infrared microspectroscopy represents a suitable tool with which to simultaneously monitor multiple disease markers in periodontal biopsies, including cellular infiltration, collagen catabolism, and other differences in the molecular profile of diseased tissues. We have established a methodology by which IR microspectroscopy is capable of revealing several major biochemical components and specific features, including collagen content, in the studied tissue using "digital staining", without the need for any chemical reagents or probes. IR maps of inflammation-driven collagen degradation in periodontal tissue sections can therefore be constructed and analyzed. The promising preliminary results, obtained in establishing this IR microspectroscopic methodology, suggest a potential role for infrared microspectroscopy in understanding the inflammatory processes underlying the progression of periodontitis.
The major problem in adapting IR technology to the study of inflammatory processes in gingival tissues was the small size and fragility of the periodontal biopsies themselves. However, we are able to process these small tissue samples and can clearly differentiate predominantly cellular and predominantly acellular areas of tissue, and visualize areas of collagen deposition and degradation. The next steps will be to compare IR microscopic maps of diseased and healthy tissues; to correlate NIR-based definitions of pathological tissue changes with disease severity; and to correlate IR maps with classical clinical signs of periodontal diseases, such as edema, gingival bleeding, and periodontal pocket depths on probing, and clinical attachment loss.
Conclusion
As inflammatory cell infiltration and subsequent collagen degradation are hallmarks of periodontitis [7-10], and because IR spectral analyses can determine such tissue events, in addition to multiple other tissue changes at the molecular and sub-molecular level, then this IR microspectroscopic methodology can now be applied to hypothesis-driven research that aims to identify disease-related pathological changes in periodontal tissues.
List of abbreviations
FTIR, Fourier-transform infrared; IR, infrared
Competing interests
The authors have no conflicting interests or financial implications related to the publication of this review article. The research activities described herein are humanitarian (non-profit making) in nature. However, certain application of infrared technology may be patentable.
Authors' contributions
The original hypothesis was developed by DAS, K-ZL, and DLS. DAS performed the literature review included in the manuscript. AGH and DLS identified suitable clinical cases. AGH performed the surgeries and prepared tissue sections for sectioning. AGH and AM performed most of the infrared microspectroscopy experiments. K-ZL, DAS and MGS performed the extrapolation of data from the IR spectra and the spectral data analysis. All authors made editorial contributions to the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This study was supported by funding from the National Research Council, Canada and by the Health Sciences Research Foundation, Canada.
Figures and Tables
Figure 1 Biomedical applications of the electromagnetic spectrum. The classic electromagnetic spectrum is shown aligned with common, established biomedical applications. The infrared region of the electromagnetic spectrum lies between the visible and microwave regions, as indicated by the red arrow.
Figure 2 Representative spectrum in the mid-IR region. The region below 1500 cm-1 is defined as the fingerprint region. The region above 1500 cm-1 is the functional group region.
Figure 3 Comparison of periodontal tissues visualized by IR microspectroscopic mapping and by conventional H and E staining. Major periodontal tissue components (DNA, protein, and lipid), as determined by IR microspectroscopic mapping, are correlated with conventional H&E staining in a representative periodontal tissue section. IR spectra obtained from within the square boxes at different areas of the tissue section (periodontal epithelium and nearby connective tissue) are shown.
Figure 4 Comparison of the mean infrared absorption spectra* of epithelial (lower spectrum) and connective tissue (upper spectrum) and the epithelial-connective tissue difference spectrum. *Spectra are baseline corrected.
Figure 5 Comparison of unstained periodontal tissue section with H and E staining and the associated IR mapping. The original periodontal tissue used for IR mapping (A), as defined by the adjacent tissue stained with H&E (B), correlates very well with the IR chemical map based on protein band intensity (C). The color bar indicates the relative intensity of the specific IR band in the tissue.
Figure 6 IR chemical mappings in periodontal tissue. Four major molecular distributions (total protein; lipid; DNA; collagen) in periodontal tissue (A-D) were generated based on the corresponding IR molecular signals. Higher concentrations of the major cellular components were observed in the periodontal epithelium (A-C) whereas higher amounts of collagen exist in the connective tissue (D), as expected.
Figure 7 Cluster mapping of periodontal tissue. Average spectra were generated from nine groups based on their overall spectral features (A); each representing one class in the cluster map of periodontal tissue (B) produced by the cluster analysis. Good correlation can be observed between the cluster mapping (B) and corresponding chemical mapping (protein) (C) based on the differential spectral features of the epithelial and connective tissues.
==== Refs
Page RC Shroeder HE Pathogenesis of inflammatory periodontal disease. A summary of current work Lab Invest 1976 34 235 265 765622
Page RC The role of inflammatory mediators in the pathogenesis of periodontal disease J Periodont Res 1991 26 230 242 1679130
Loesche WJ The antimicrobial treatment of periodontal disease: changing the treatment paradigm Crit Rev Oral Biol Med 1999 10 245 75 10759408
Jackson M Sowa MG Mantsch HH Infrared spectroscopy: a new frontier in medicine Biophys Chem 1997 68 109 25 9468614 10.1016/S0301-4622(97)80555-8
Liu KZ Dixon IM Mantsch HH Distribution of collagen deposition in cardiomyopathic hamster hearts determined by in microscopy Cardiovasc Pathol 1999 8 41 47 10722247 10.1016/S1054-8807(98)00024-6
Schultz CP Liu KZ Kerr PD Mantsch HH In-situ infrared histopathology of keratinization in human oral/or pharyngeal aqueous cell carcinoma Oncol Res 1998 10 277 286 9802063
Ejeil AL Gaultier F Igondjo-Tchen S Senni K Pellat B Godeau G Gogly B Are cytokines linked to collagen breakdown during periodontal disease progression? J Periodontol 2003 74 196 201 12666708 10.1902/jop.2003.74.2.196
van der Zee E Everts V Beertsen W Cytokines modulate routes of collagen breakdown J Clin Periodontol 1997 24 297 305 9178108
Hillmann G Krause S Ozdemir A Dogan S Geurtsen W Immunohistological and morphometric analysis of inflammatory cells in rapidly progressive periodontitis and adult periodontitis Clin Oral lnvestig 2001 5 227 35 11800435 10.1007/s00784-001-0134-3
Feldner BD Reinhardt RA Garbin CP Seymour GJ Casey JH Histological evaluation of interleukin-1 beta and collagen in gingival tissue from untreated adult periodontitis J Periodontal Res 1994 29 54 61 7509386
Abrams K Caton J Polson AM Histologic comparison of interproximal gingival tissues related to the presence or absence of bleeding J Periodontol 1984 55 629 632 6334156
Lasch P Haensch W Naumann D Diem M Imaging of colorectal adenocarcinoma using FT-IR microspectroscopy and cluster analysis Biochim Biophys Acta 2004 176 186 14990348
Liu KZ Jia J Newland AC Kelsey SM Mantsch HH Quantitative determination of apoptosis on leukemic cells by infrared spectroscopy Apoptosis 2001 6 269 278 11445669 10.1023/A:1011383408381
Diem D Chiriboga L Yee H Infrared spectroscopy of human cells and tissue. VIII. Strategies for analysis of infrared tissue mapping data and applications to liver tissue Biopolymers 2000 57 82 90 10.1002/1097-0282(2000)57:5<282::AID-BIP50>3.0.CO;2-R
| 15890072 | PMC1156901 | CC BY | 2021-01-04 16:03:35 | no | BMC Med Imaging. 2005 May 12; 5:2 | utf-8 | BMC Med Imaging | 2,005 | 10.1186/1471-2342-5-2 | oa_comm |
==== Front
BMC MicrobiolBMC Microbiology1471-2180BioMed Central London 1471-2180-5-261590449510.1186/1471-2180-5-26Methodology ArticleSelecting representative model micro-organisms Holland BR [email protected] J [email protected] Allan Wilson Centre for Molecular Ecology and Evolution, Massey University, Palmerston North, New Zealand2 Institute of Molecular BioSciences, Massey University, Palmerston North, New Zealand2005 17 5 2005 5 26 26 23 2 2005 17 5 2005 Copyright © 2005 Holland and Schmid; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Micro-biological research relies on the use of model organisms that act as representatives of their species or subspecies, these are frequently well-characterized laboratory strains. However, it has often become apparent that the model strain initially chosen does not represent important features of the species. For micro-organisms, the diversity of their genomes is such that even the best possible choice of initial strain for sequencing may not assure that the genome obtained adequately represents the species. To acquire information about a species' genome as efficiently as possible, we require a method to choose strains for analysis on the basis of how well they represent the species.
Results
We develop the Best Total Coverage (BTC) method for selecting one or more representative model organisms from a group of interest, given that rough genetic distances between the members of the group are known. Software implementing a "greedy" version of the method can be used with large data sets, its effectiveness is tested using both constructed and biological data sets.
Conclusion
In both the simulated and biological examples the greedy-BTC method outperformed random selection of model organisms, and for two biological examples it outperformed selection of model strains based on phylogenetic structure. Although the method was designed with microbial species in mind, and is tested here on three microbial data sets, it will also be applicable to other types of organism.
==== Body
Background
To gain insight into biological processes, biologists often rely on model organisms. Focusing the funding and research efforts of many research groups on one model organism is considered more likely to advance the field than scattering the same resources over a large number of organisms. A recent application of this philosophy is in genome sequencing: rather than simultaneously initiating the sequencing of the genomes of many individuals from a species, typically a single representative is chosen. In microbiology the initial choice of model strain is frequently a well-characterized laboratory strain, often selected at a time when the ability to determine population structure and to measure genetic distances was limited (see for example [1,2]). There is merit in choosing a physiologically well-characterised laboratory strain for genome sequencing, as it facilitates the interpretation and annotation of sequence data. However, it has often become apparent that the model strain initially chosen does not represent important features of the species [3-5]. For micro-organisms, the diversity of their genomes is such that even the best possible choice of initial strain for sequencing may not assure that the genome obtained adequately represents the species [5-7]; additional strains may need to be studied [7]. To acquire information about a species' genome as efficiently as possible, we require a method to choose strains for analysis on the basis of how well they represent the species.
The problem of choosing representative model strains is complicated by the fact that the value of an organism as a representative depends on the features it is meant to represent. As an example illustrating this problem, assume that we require model organisms that represent a larger group of organisms in terms of amino acid sequence at a particular open reading frame (ORF). Suppose that we are able to find a single model organism which represents the entire group; i.e. BLAST searches find significant homologies between the amino acid sequence of the model protein and proteins from the other organisms. Let us assume that homologies between the protein of the model organism and the more distant members of the group are low, albeit still significant. We would conclude, using the BLAST homology criterion that we have represented the group, and have done so efficiently, using only one model organism. However, if we were then to try to use this same organism to design PCR primers for amplifying the ORF in all members of the group, the model organism would fail for the more distant members. Likewise if we were to investigate another, less conserved ORF across the group, the model organism may fail to show significant homology to a large number of members of the group even at the amino acid sequence level. Another difficulty in choosing model organisms is that our selection will be based on limited and biased information obtained in a screen of the group we wish to represent, such as sequences from a few genes or restriction fragment length polymorphism (RFLP) data which may not represent, or only be loosely linked to, diversity across the entire genome.
In this paper we describe and evaluate approaches for the rational selection of model organisms, which take these problems into consideration.
Results and discussion
Finding the best model organism
As illustrated in the introduction, there will be some threshold distance, i.e. level of genetic difference, between the model organism and other organisms, above which the model organism will no longer be a useful representative, and different applications will have different threshold values. If we need a model organism for some specific application where a threshold distance, T, is known, the problem of selecting the best model organism can be solved by choosing the organism for which the greatest number of other organisms lie within this genetic distance, T. We refer to this criterion as Best Coverage.
For the purpose of illustration we display a constructed example where distances between organisms are the distances on a plane (Figs. 1A and 1B). Intuition suggests that model organisms chosen according to the Best Coverage criterion will be central to the group of interest. For example, in Figure 1A organism a is the most central, it has the best coverage for any choice of threshold distance. Conversely, a non-central organism such as organism b has worse coverage for any value of T. Choosing a model organism in the example in Figure 1A is trivial because the members of the group are distributed in a highly symmetrical fashion. However, in many cases members of a group will not be symmetrically distributed and there will be no obvious central organism. In such a case, shown in Figure 1B, the value of the threshold distance, T, will affect the decision as to which organism is best. Depending on the choice of T, organism a, b, or c could be preferred. Figures 1C and 1D display Coverage for increasing values of T for the marked organisms in 1A and 1B respectively. Note that in Figure 1D organism a is best for low threshold values and that organism c is best for high threshold values, however, taken as an aggregate over a range of T values organism b is superior.
Finding the best threshold distance for selecting model organisms from a group poses a number of difficulties. Firstly, as noted, different applications require different thresholds and so no one threshold distance will be ideal as a basis of model strain selection, unless the model strain is only to be used for one particular application. Secondly, even if we knew in advance the exact feature which we want the model organism to represent, there will only be limited information on the variation in this feature for the entire group of organisms -after all, if the organisms were exactly characterised there would be no need to pick model organisms for further study. Lastly, if we want to make a good choice of model organisms we will need to initially sample a large number of members of the group the model organisms are supposed to represent. This favours the use of methods that produce "quick and dirty" estimates of relationships between members such as short sequence alignments, fragment length polymorphism, biotyping or enzymatic activity. The resulting distances will only be a rough guide to the true relationship between organisms, and their relationship to T for specific applications will be unknown.
These considerations suggest that a good criterion for choosing a model organism is that it be representative for a range of T values. This can be implemented by normalising the available distances within a group to be represented to have a maximum of 1, and summing the Coverage criterion for T at steps between 0 and 1, e.g. {0 0.05, 0.10... 0.95, 1.00}. We refer to this score as being the Total Coverage for the model organism, as it is an aggregate over a range of T values. The Best Total Coverage (BTC) method picks the organism for which this score is largest. Relating this to the coverage graphs (Figs. 1C and 1D) the BTC method picks the organism with the largest area beneath the curve, for 1C this is organism a, and for 1D it is organism b.
Determining the number of model organisms required
We can see in Figure 1A, and especially 1B, that for small threshold values, even the best possible model organism, as selected by the best total coverage method, cannot represent the entire group of organisms. Rather than choosing a single model organism we would prefer, if possible, to choose a set of model organisms so that as many organisms as possible lie within the threshold distance of at least one model organism for a particular application. Obviously coverage will continue to improve, or at least not get worse, as more and more model organisms are allowed. Indeed, in the ideal case every unique organism would be chosen as a model organism. However, financial and time constraints usually mean that such a solution is not feasible. Plotting the increase in coverage for different numbers of models organisms allows one to judge when no significant improvement results from adding another model organism, or whether the improvement outweighs the cost of dealing with an extra model organism. For example, Figure 1E shows the BTC score (as a percentage of the maximum possible) for the data set in Figure 1B with from one up to thirty model organisms allowed. In this example the graph suggests that only marginal improvements in coverage can be obtained by having more than one or two model organisms. In the case where there are distinguishable clusters of genetically similar strains, the value of k could be predetermined based on the number of clusters.
Computational strategies for implementing the methods
For practical application of the strategies outlined above, we must consider the computational complexity of finding the best set of model organisms. For a group of n isolates and a predetermined number of model organisms k there are n!/k!(n - k)! possible sets of model organisms to be tested, and for each set of model organisms it requires nk operations to calculate Coverage, because for each of the n isolates, at worst k putative model organisms must be checked to determine if the distance between the isolate and the model organism is less than T. So, for a fixed k, it will take time proportional to nk + 1 to compute the best set of model organisms. This means that for large n and k it will become computationally intractable to test all possible sets of model organisms. For example, for k = 3 model organisms from a set of 100 organisms there are 161,700 possible sets of model organisms to consider, for k = 5 model organisms there are 75,287,520 sets to consider, and for k = 10 there are 1.73 × 1013.
To avoid this problem a greedy approximation to the exact method can be used. Initially one model organism is chosen, the one that gives the largest Total Coverage score, then model organisms are added one at a time, at each stage picking the organism which gives the largest improvement in the Total Coverage score, until some predefined number of model organisms have been selected. We name this the greedy-BTC method.
Test of the effectiveness of the greedy-BTC method for choosing model organisms
We tested the performance of the greedy method in several ways. Firstly we compared the greedy-BTC method against the exact-BTC method using a range of simulated sequence data (see the Methods section for details on data generation). The greedy method gave similar results to the exact method for problems with small (computationally feasible) numbers of model sequences (Table 1; for comparison the Total Coverage scores for randomly selected model sequences are also included). For larger numbers of model organisms we would expect that the greedy method would perform less well relative to the exact method. However, in this context we are not so concerned with finding an optimal solution – a good solution will do.
Secondly we compared the greedy-BTC method against both a large number of random selections of model organisms and researchers' selections based on phylogenetic evidence. We used two MLST data sets: 139 strains of Enterococcus faecium [8], and 121 strains of Candida albicans [9]. Six researchers (four phylogenetics researchers from the Allan Wilson Centre for Molecular Ecology and Evolution, and the two authors) were given copies of the two trees and asked to select three and five model strains respectively that they felt would be good representatives of the group of strains based on the phylogenetic tree. (The reasons we fixed k = 5 for the E. faecium data and k = 3 for the C. albicans data are detailed in a later section.) The trees in figures 2A and 3A show the researchers' selections of model organisms and the greedy-BTC choice of model organisms. Figures 2B and 3B compare the performance of these selections along with 1000 random selections of model organisms. The greedy-BTC strains performed better than random selection in all but 4/2000 cases and always outperformed the human selections.
An interesting outcome of this experiment is that strain selections which seem almost identical based on the phylogenetic tree, for example those selections represented by red and green circles in figure 3A, performed quite differently. This is probably because trees cannot (except in the case of perfectly treelike distances) reflect all of the information in the distance matrix [10].
Assessing usefulness of the method in choosing model organisms representing characters unknown at the time of selection
As pointed out in the introduction, selection of model organisms will often be based on characters different from the ones which will later be studied using the model organism. Since different loci may evolve according to different processes in different lineages [10], selecting a model organism using one set of characters does not guarantee that it is representative for other characters. The exchange of genetic material between organisms of different species (horizontal gene transfer) will produce additional problems of a similar nature. The true test of a method for selecting model organisms is therefore if it can arrive at a model organism which is reasonable representative for characters not used in its selection.
We therefore tested the BTC method on three examples of micro-biological data sets in which we could assess the representativeness of model organisms for characters not used in their selection. In each case we began by using one character set to choose the greedy-BTC model strains for k in the range 1–10. As in the previous example with simulated data we compared the greedy-BTC score to the Total Coverage score attained by random sets of model organisms for each data set and value of k. Then, for a fixed value of k, we compared our choice of model organisms with random choices of k organisms, in terms of how well they represented the characters not used in the selection of the model organisms.
The first example was a sample of 22 Pseudomonas aeruginosa strains, for which both RFLP typing data and sequence data (pvdS gene) are available [11]. We used the RFLP data for model strain selection. As expected, the plot of greedy BTC score versus k for the RFLP based distances (Fig. 4A) showed a gradual improvement in Total Coverage as increasing numbers of model strains are used. For each value of k (k = 1–10) we compared the Total Coverage of the greedy-BTC model strains to the Total Coverage of 1000 randomly chosen sets of k strains. For all values of k tested the greedy-BTC score is in the top quartile of the distribution of scores for randomly chosen strains.
To assess the representativeness of the BTC model strain on the pvdS data (not used in model strain selection) we fixed k = 1, as the total number of strains was fairly small, and Figure 4A did not suggest any obvious clustering in the data (clustering would be indicated by a sharp jump in coverage for some value of k). We then measured the Total Coverage of the selected model strain for the pvdS gene and compared it to the Total Coverage of the remaining 21 strains (Fig 4B). (In order to compute the Total Coverage score for the pvdS data we constructed Hamming distances from the sequence alignment.) In 55% of cases random selection did equally well (this was unsurprising as 12/22 of the strains had identical pvdS sequences), but in 45% of the cases random choice gave a worse representation than the BTC selected model strain.
The second data set consisted of partial sequences of seven genes (adk, atpA, ddl, gyd, gdh, purK, and pstS) of 139 Enterococcus faecium strains [8]. We constructed seven test data sets, for each a distance matrix was generated by taking the Hamming distances from the sequences for six of the loci (i.e. each time we omitted one locus). Using these distances we carried out selection of model organisms separately for each of the seven test data sets. We then tested the performance of the greedy-BTC method by assessing how well the model organisms represented the seventh locus (omitted when constructing the distance matrix).
For each value of k (k = 1..10) we compared the Total Coverage of the greedy-BTC model strains to the Total Coverage of 1000 randomly chosen sets of k strains. The results for one of these data sets (adk omitted) are shown in Figure 5A. The greedy-BTC model strains were always in the top quartile of the distribution of scores for random strains.
To assess the representativeness of the greedy-BTC model strains on the omitted gene (not used in model strain selection) we fixed k = 5, and compared the five greedy-BTC model strains to random selections of five strains. Figure 5B shows the performance on each data set. In one case the greedy-BTC model strains outperformed the random strains and in six cases the greedy-BTC model strains performed about equally to the median random score. However, note that the distribution of random scores is skewed – i.e. the best random sets of strains had scores about 5% higher than the greedy-BTC model strains but the worst random sets had scores about 10% lower than the model strains. Figure 5C shows the aggregate performance over all seven data sets. Picking strains systematically using the BTC method did better than picking a random set of five strains in 844 out of 1000 cases.
The third data set consisted of partial sequences of seven genes (AAT1a, ACC1, ADP1, MPIb, SYA1, VPS13, and ZWF1b) of 122 Candida albicans strains [9]. Similarly to the previous example, we generated seven distance matrices, each based on six of the loci, and, using these distances, carried out selection of model strains. We then tested our choice by assessing how well the model strains represented the alleles at the seventh locus (omitted when constructing the distance matrix).
For each value of k (k = 1..10) we compared the Total Coverage of the greedy-BTC model strains to the Total Coverage of 1000 randomly chosen sets of k strains. The results for one of these sequence sets (AAT1a omitted) are shown in Figure 6A. The greedy-BTC model strains were always in the top quartile of the distribution of scores for random strains, and for k = 5, 6, 7, 8, 9, and 10 the scores for the greedy-BTC model strains were outside the range of scores for random sets of strains.
To assess the representativeness of the greedy-BTC model strain on the omitted gene (not used in model strain selection) we fixed k = 3, and compared the three greedy-BTC model strain to random selections of three strains. Figure 6B shows the performance on each data set. In four out of seven cases the greedy-BTC model strains outperform randomly chosen strains at representing the omitted gene, but for AAT1a and ACC1 random selection does better. Figure 6C shows the aggregate performance over all seven data sets, in 759 cases out of 1000 the greedy-BTC model strains outperformed the randomly chosen strains.
Conclusion
The greedy-BTC method outperformed selections based on phylogenetic evidence made by researchers with experience in phylogenetics. By testing the greedy-BTC choice against the exact-BTC where feasible or against a large number of random selections of model strains we also demonstrated that the method comes close to achieving optimal representation of organisms (according to the BTC criterion).
The greedy-BTC also performed well in the more challenging situation where the characters to be represented were not used in model strain selection; for most of the biological data sets, with the exception of two loci in the Candida albicans data, the greedy-BTC model organisms did an equal or better job than random organisms of representing data not used in model strain selection.
One interesting feature of the distributions of Total Coverage scores for random organisms is its skewness – many sets produce good scores but there is a long tail of poor scores. This feature seemed to be consistent across a range of data sets. This means that picking organisms at random may occasionally do a little better than using the greedy-BTC method but it may also do a lot worse.
The three biological data sets analysed suggest that the BTC method we have developed should facilitate selection of model organisms that will be representative of the group of interest for a wide range of applications. Nevertheless, differences in the rate of evolution at different loci plus other phenomena such as horizontal gene transfer place limits on the degree of reliability of selected model organisms in terms of representativeness in regard to other, unknown loci. We would therefore suggest that if the BTC method were used to select organisms for a project associated with major expenses (in terms of time or money), such as a genome sequencing project, it would be advisable to begin by eliminating part of the input data and testing how well the method works at selecting model organisms that capture the diversity of the omitted part of the data. Such an analysis will not only provide a general idea of how well the selected organisms will represent the sample, it will also reveal if some of the markers intended for selection may give misleading information. If, for instance one of them had been acquired by horizontal gene transfer, this should become apparent as poor representation of allelic diversity at that marker locus by model organisms selected on the basis of the remaining markers.
We note that one problem our method does not address is how to obtain a suitable collection of the group of organisms on which the BTC method is to be used to choose representatives. The BTC method is aimed at optimally representing a group defined by its user, and highly prevalent genetically similar subgroups in this user-defined group will carry more weight than low prevalence groups. Therefore, if the group is biased so that particular groups are over represented, the BTC choice of model strains will be biased as well.
Methods
The greedy-BTC and exact-BTC methods described for selecting model organisms have been implemented in a C program that has been tested using both UNIX and WindowsXP. The exact method is only feasible for k ≤ 3 unless the total number of strains is also fairly small; the program automatically defaults to the greedy method if k ≥ 4 and n ≥ 10. The program also works for user-defined threshold values. All code is available on request from [email protected].
The method described can be expressed mathematically as follows. Firstly let X be the set of all organisms, and d(m,x) a measure of the distance between organism m and x. Define y(m,x) = 1, if d(m,x) <T,
y(m,x) = 0, otherwise.
In the case of choosing a single model organism m
To solve the multiple model organism case we choose a set M of fixed size k, such that
The test data used to generate Table 1 was generated using the software Treevolve version 1.3.2 [12] to simulate DNA sequences along a large (1000 taxon) random tree according to a simple model of nucleotide substitution (the Jukes Cantor model [13]). One hundred data sets of size 20 were then sampled from this large data set, and distance matrices were generated by taking the uncorrected distance between sequences (this is the proportion of sites that differ between the two sequences).
List of abbreviations
BTC – best total coverage
MLST – multi-locus sequence typing
ORF – open reading frame
PCR – polymerase chain reaction
RFLP – restriction fragment length polymorphism
Authors' contributions
B.R.H. developed and implemented the methods, performed simulations and wrote the bulk of the manuscript. J.S. developed the methods, wrote the introduction section of the manuscript, edited the manuscript, and sourced the example data sets.
Acknowledgements
B.R.H. acknowledges the help of Mike Hendy and David Penny during her PhD when this work originated. B.R.H. and J.S. acknowledge Jamie Riden for the C implementation of the method. B.R.H. and J.S. acknowledge Richard D. Cannon for his comments on a draft of this ms. B.R.H. and J.S. acknowledge Frank Odds for the use of the Candida MLST data. This work was supported in part by Marsden grant MAU902 from the royal Society to J.S.
Figures and Tables
Figure 1 Coverage plots for example data sets. A and B) Artificial data sets where organisms are represented by dots with Euclidean distances. The concentric circles around the labelled organisms indicate how many organisms would be well represented by this choice of model strain for different threshold values. C and D) Coverage plots for the data sets in A and B respectively, showing how many organisms are well represented for increasing threshold values. Key for D: a – x, b – o, c – Δ. E) Best Total Coverage score (as a percentage of maximum) for 1–30 model organisms, for data set B, the model organisms are selected using a greedy implementation of the BTC method as described later in the text.
Figure 2 Comparing greedy-BTC model strains to researchers' selections for the Enterococcus data. Neighbor-joining [10] tree based on MLST data for 139 E. faecium strains. Each MLST scheme has seven genes, these genes were concatenated together and a distance matrix (used by both neighbor-joining and the greedy-BTC method) was constructed based on uncorrected distances between pairs of sequences (i.e. the proportion of sites that differ). The blue triangles indicate the five model strains selected using the greedy-BTC method. Coloured circles indicate model strains selected by six researchers. The inset histogram compares the BTC-scores of the greedy and human selection to 1000 random selections of five model strains.
Figure 3 Comparing greedy-BTC model strains to researchers' selections for the Candida data. Neighbor-joining tree based on MLST data for 121 C. albicans strains (see figure 2 caption for details of construction). The blue triangles indicate the three model strains selected using the greedy-BTC method. Coloured circles indicate model strains selected by six researchers. The inset histogram compares the BTC-scores of the greedy and human selection to 1000 random selections of three model strains.
Figure 4 Performance of the greedy-BTC method for Pseudomonas data. A) The greedy Best Total Coverage score for k = 1–10 model organisms, for the Pseudomonas RFLP distances, is shown by the crosses. For each value of k the box and whisker plot indicates the quartiles (i.e. the box indicates those 50 % of the values closest to the median) and range for 1000 random choices of model organisms. Total Coverage scores were produced by adding up the percentage of organisms represented at each of 20 threshold value intervals (0, 0.05, 0.10,...,0.95 and 1.0); they are reported as a percentage of the maximum possible Total Coverage score. B) Histogram showing the Total Coverage scores of 1000 random choices of model organism, the arrow indicates the Total Coverage score for the BTC choice of model strain.
Figure 5 Performance of the greedy-BTC method for Enterococcus data. A) The greedy Best Total Coverage score for k = 1–10 model organisms, for the distances generated from all loci excluding adk, is shown by the crosses. For each value of k the box and whisker plot indicates the quartiles (i.e. the box indicates those 50 % of the values closest to the median) and range for 1000 random choices of model organisms. Total Coverage scores were produced by adding up the percentage of organisms represented at each of 20 threshold value intervals (0, 0.05, 0.10,..., 0.95 and 1.0); they are reported as a percentage of the maximum possible Total Coverage score. B) For each omitted locus the box and whisker plot indicates the quartiles and range for 1000 random choices of 5 strains. The score of the five greedy-BTC model strains are shown by crosses. C) Histogram showing the aggregate Total Coverage scores of 1000 random choices of five strains summed over the seven data sets, the arrow indicates the aggregate Total Coverage score for the BTC choices of model strains.
Figure 6 Performance of the greedy BTC method for Candida data. A) The greedy Best Total Coverage score for k = 1–10 model organisms, for the distances generated from all loci excluding AAT1a, is shown by the crosses. For each value of k the box and whisker plot indicates the quartiles (i.e. the box indicates those 50 % of the values closest to the median) and range for 1000 random choices of three strains. Total Coverage scores were produced by adding up the percentage of organisms represented at each of 20 threshold value intervals (0, 0.05, 0.10,..., 0.95 and 1.0); they are reported as a percentage of the maximum possible Total Coverage score. B) For each omitted loci the box and whisker plot indicates the quartiles and range for 1000 random choices of three strains. The score of the three greedy-BTC model strains are shown by crosses. C) Histogram showing the aggregate Total Coverage scores of 1000 random choices of three strains summed over the seven data sets, the arrow indicates the aggregate Total Coverage score for the BTC choices of model strains.
Table 1 Performance of the greedy algorithm in selecting model organisms from simulated DNA sequence data sets.
Total coverage (percent of exact method)
k = 1 k = 2 k = 3
Greedy Score 100.0% 99.6% 99.6%
Random Score 77.7% 76.5% 76.4%
The scores are the mean Total Coverage over the 100 simulated data sets for k = 1, 2 or 3 with the greedy and random scores (scores obtained when randomly chosen sequences are used as model strains) expressed as a percentage of the exact scores. All data sets were generated as described in the methods section, distances were normalised to have a maximum value of 1, and Total Coverage was evaluated in steps of 0.05 between 0 and 1.
==== Refs
Stover CK Pham XQ Erwin AL Mizoguchi SD Warrener P Hickey MJ Brinkman FS Hufnagle WO Kowalik DJ Lagrou M Complete genome sequence of Pseudomonas aeruginosa PA01, an opportunistic pathogen Nature 2000 406 959 964 10984043 10.1038/35023079
Jones T Federspiel NA Chibana H Dungan J Kalman S Magee BB Newport G Thorstenson YR Agabian N Magee PT Davis RW Scherer S The diploid genome sequence of Candida albicans Proc Natl Acad Sci USA 2004 101 7329 7334 15123810 10.1073/pnas.0401648101
Liu HC Styles A Fink GR Saccharomyces cerevisiae S288C has a mutation in FLO8, a gene required for filamentous growth Genetics 1996 144 967 978 8913742
Takagi H Shichiri M Takemura M Mohri M Nakamori S Saccharomyces cerevisiae Sigma 1278b Has Novel Genes of the N-Acetyltransferase Gene Superfamily Required for L-Proline Analogue Resistance J Bacteriol 2000 182 4249 4256 10894734 10.1128/JB.182.15.4249-4256.2000
Fitzgerald JR Musser JM Evolutionary genomics of pathogenic bacteria Trends Microbiol 2001 9 547 553 11825715 10.1016/S0966-842X(01)02228-4
Lan R Reeves PR Intraspecies variation in bacterial genomes: the need for a species genome concept Trends Microbiol 2000 8 396 401 10989306 10.1016/S0966-842X(00)01791-1
Boucher Y Nesbo CL Doolittle WF Microbial genomes: dealing with diversity Curr Opin Microbiol 2001 4 285 289 11378480 10.1016/S1369-5274(00)00204-6
Homan WL Tribe D Poznanski S Li M Hogg G Spalburg E van Embden JD Willems RJ Multilocus Sequence Typing Scheme for Enterococcus faecium J Clin Microbiol 2002 40 1963 1971 12037049 10.1128/JCM.40.6.1963-1971.2002
Odds F Unpublished Candida albicans MLST data
Swofford DL Olsen GJ Waddell PJ Hillis DM Hillis DM, Moritz C, Mable BK Phylogenetic Inference Molecular Systematics 1996 2 Sinauer Associates, Inc. Publishers, Sunderland, Massachusetts 407 514
Al-Samarrai TH Zhang N Lamont I Martin L Kolbe J Wilsher M Morris AJ Schmid J Simple and inexpensive but highly discriminating method for computer-assisted DNA fingerprinting of Pseudomonas aeruginosa J Clin Microbiol 2000 38 4445 4452 11101578
University of Oxford, Department of Zoology. Evolutionary biology group, software page
Jukes TH Cantor CR Munro HN Evolution of protein molecules Mammalian Protein Metabolism 1969 Academic Press 21 132
| 15904495 | PMC1156902 | CC BY | 2021-01-04 16:03:39 | no | BMC Microbiol. 2005 May 17; 5:26 | utf-8 | BMC Microbiol | 2,005 | 10.1186/1471-2180-5-26 | oa_comm |
==== Front
BMC MicrobiolBMC Microbiology1471-2180BioMed Central London 1471-2180-5-281590449710.1186/1471-2180-5-28Methodology ArticleFungal-specific PCR primers developed for analysis of the ITS region of environmental DNA extracts Martin Kendall J [email protected] Paul T [email protected] Dynamac Corporation, National Health and Environmental Effects Research Laboratory, Corvallis, OR USA2 USEPA National Health and Environmental Effects Research Laboratory, Corvallis, OR, USA2005 18 5 2005 5 28 28 30 11 2004 18 5 2005 This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Internal Transcribed Spacer (ITS) regions of fungal ribosomal DNA (rDNA) are highly variable sequences of great importance in distinguishing fungal species by PCR analysis. Previously published PCR primers available for amplifying these sequences from environmental samples provide varying degrees of success at discriminating against plant DNA while maintaining a broad range of compatibility. Typically, it has been necessary to use multiple primer sets to accommodate the range of fungi under study, potentially creating artificial distinctions for fungal sequences that amplify with more than one primer set.
Results
Numerous sequences for PCR primers were tested to develop PCR assays with a wide range of fungal compatibility and high discrimination from plant DNA. A nested set of 4 primers was developed that reflected these criteria and performed well amplifying ITS regions of fungal rDNA. Primers in the 5.8S sequence were also developed that would permit separate amplifications of ITS1 and ITS2. A range of basidiomycete fruiting bodies and ascomycete cultures were analyzed with the nested set of primers and Restriction Fragment Length Polymorphism (RFLP) fingerprinting to demonstrate the specificity of the assay. Single ectomycorrhizal root tips were similarly analyzed. These primers have also been successfully applied to Quantitative PCR (QPCR), Length Heterogeneity PCR (LH-PCR) and Terminal Restriction Fragment Length Polymorphism (T-RFLP) analyses of fungi. A set of wide-range plant-specific primers were developed at positions corresponding to one pair of the fungal primers. These were used to verify that the host plant DNA was not being amplified with the fungal primers.
Conclusion
These plant primers have been successfully applied to PCR-RFLP analyses of forest plant tissues from above- and below-ground samples and work well at distinguishing a selection of plants to the species level. The complete set of primers was developed with an emphasis on discrimination between plant and fungal sequences and should be particularly useful for studies of fungi where samples also contain high levels of background plant DNA, such as verifying ectomycorrhizal morphotypes or characterizing phylosphere communities.
PCRprimersbasidiomycetesascomycetesInternal Transcribed Spacer (ITS)X/T rapid extraction method
==== Body
Background
Studies of fungi (Kingdom Eumycota) in natural environments often require simultaneous analysis of a broad taxonomic range. For instance, the fungi forming ectomycorrhizal symbioses number over 5000 species [1]. Analysis of natural ectomycorrhizal fungal communities traditionally has been a laborious, highly-skilled process with heavy reliance on gross morphological characterization of the ectomycorrhizal root-tips. The laborious nature of microscopic analysis and identification is the driving force behind developing such methods as molecular verification of identified morphotypes by means of PCR-RFLP [2]. Analysis of natural populations of ectomycorrhizas requires high rates of sampling because of the high number of fungal species involved and the high spatial variability observed in natural systems [3,4]. Primers allowing simultaneous analysis of all the fungal phyla which are involved in ectomycorrhizal symbioses would be a useful tool in studying the ecology of these fungi. Of particular interest for such work is the division Dikaryomycota, which includes the subdivisions Basidiomycotina and Ascomycotina, and encompasses all ectomycorrhizal fungi [1].
The earliest PCR primers to gain wide acceptance for work with fungal Internal Transcribe Sequences (ITS) were "ITS1" and "ITS4" which amplify the highly variable ITS1 and ITS2 sequences surrounding the 5.8S-coding sequence and situated between the Small SubUnit-coding sequence (SSU) and the Large SubUnit-coding sequence (LSU) of the ribosomal operon [5]. These primers amplify a wide range of fungal targets and work well to analyze DNA isolated from individual organisms, but do not exclude effectively the plant host sequences in mixed, phytosphere DNA extracts typical of studies of plant-associated microbiota. Subsequently, the plant-excluding primers ITS1-F and ITS4-B came into wide use for analyses of fungal ITS sequences, but these primers were "intended to be specific to fungi and basidiomycetes, respectively" [6]. We present here a suite of primers designed to amplify Dikaryomycota efficiently, requiring little optimization for use in the hands of even relatively untrained operators.
In addition to the increased efficiency needed for well-established methods such as PCR-RFLP analyses of single fungal species [7], robust primers are needed for newer molecular methods currently used to characterize microbial communities: Length Heterogeneity PCR (LH-PCR) [8], high-throughput sequencing [9], and Terminal Restriction Fragment Length Polymorphism (T-RFLP) [10]. LH-PCR is a reliable and effective approach to analyze targets with high variability in overall length. In T-RFLP analyses, as with LH-PCR, a fluorescent label on the PCR primer is used for detection, but in this case only the primer-terminal fragments of restriction digested PCR products are detected. These fragments contain the labelled primer and extend to the first instance of a restriction site for the enzyme used. With increased access to capillary-electrophoresis instruments capable of high resolution discrimination of oligonucleotide lengths, these methods have become both rapid and reliable. Such methods enable rapid analysis of environmental samples and can provide extensive data on microbial communities as defined or restricted by the specificity range of primers used. These data include both relative abundances of dominant microbial phylotypes and characteristic PCR-product or TRF (terminal restriction fragment) sizes for these phylotypes. For comparison of fungal communities, either method provides a relatively complete, culture-independent analysis. Additionally, where the effort is justified, identified phylotypes subsequently can be taxonomically characterized (by applying sequence analysis to amplified targets) with the advantage that resources can be focused on phylotypes that are most important to the study at hand. Terminal restriction fragment sizes also can be compared to a database of theoretical restriction fragments derived from sequence information to approximate taxonomic identity [11,12].
While LH-PCR or T-RFLP gives an estimate of relative abundance of phylotypes in a community, quantitative PCR (QPCR) gives an overall quantification for the target sequence which can then be subdivided mathematically to provide an estimate of absolute population level for individual phylotypes. Higuchi et al. [13,14] developed a method for real-time detection of PCR products to quantify sample target sequences based on fluorescence of the intercalating dye ethidium bromide, and numerous companies now offer kits for such quantification using SYBR Green I (Invitrogen Life Technologies, Carlsbad, CA) as the intercalating dye. Most of these instruments now include the ability to perform a melting curve analysis on the PCR product after quantification [15], which partially ameliorates uncertainty arising from the inability of the dye to distinguish target amplicons from non-target PCR products.
We describe a set of primers which performs well in all of these analytical approaches. This characteristic of wide applicability to studies of Dikaryomycota from diverse environmental samples, particularly those with significant plant tissue content, lends a degree of interoperability between analytical approaches. For instance, real-time PCR analysis of samples can guide the choice of cycle numbers for LH-PCR or T-RFLP reactions which are sensitive to late stage PCR. This is important for comparing data between more established approaches and newer methods coming into popular use.
Results and discussion
Initial work with fungal PCR
We began investigating fungal ITS primers using ITS1-F as the forward primer situated at the 3' end of SSU and ITS4-B as the reverse primer in the 5' section of LSU [6]. These primers amplify the entire ITS region (Figure 1). The reverse primer, ITS4-B, was not intended to amplify ascomycete targets, however, and based on sequence comparisons, it appears that it can be a poor match to many basidiomycetes. We also investigated reverse primers in the 5' section of LSU developed by Egger [16]. These primers separately amplify Ascomycetes and Basidiomycetes. All of these primers have strong positive attributes, but since our morphological analysis of ectomycorrhizal root tips did not distinguish between Ascomycetes and Basidiomycetes, we developed primers that were capable of amplifying all Dikaryomycota simultaneously.
Initially, we approached our goal of a reliable PCR amplification broadly targeting Dikaryomycota by developing a primer (NLB3) which works well with ITS1-F and is close to the annealing site for ITS4-B. The ITS1-F/NLB3 primer pair effectively excludes plant sequences, amplifies basidiomycete targets, and additionally, amplifies ascomycete targets. We found, however, that ITS1-F apparently caused spurious product bands from low concentration (single root tip) DNA extracts of our ectomycorrhizas. The source of these bands was verified by reamplification with ITS1-F alone. The ITS1-F/NLB3 primer pair has been shown to be suitable for mycorrhizal applications in another laboratory [17].
Fungal ITS primer development
In order to obtain a cleaner product for restriction, we developed an alternative to ITS1-F, resulting in a primer pair which shows greater specificity to the intended fungal ribosomal target (Figure 1). We also developed a robust pair of Dikaryomycota-specific primers (NSA3/NLC2) that could serve as first round primers in nested-PCR reactions with annealing sites outside those for a second pair of Dikaryomycota-specific primers (NSI1/NLB4). We have used this nested reaction extensively for PCR-RFLP work with ectomycorrhizal root tips. The NLB4 primer is identical to the NLB3 primer except that it has one additional base at the 3' end (see Table 1) and has a slightly higher calculated melting temperature.
Additional primer development allowing for separate amplification of ITS1 and ITS2 produced a final suite of 6 primers (see Table 1, Figure 1) specific to Dikaryomycota: outer nested-PCR primers (NSA3/NLC2), a pair of primers that amplify both ITS regions (NSI1/NLB4) and forward (58A1F, 58A2F) and reverse (58A2R) primers in the 5.8S sequence (overlapping with ITS3/ITS2 of White et al. [5]) for amplification of internal transcribed spacers ITS2 and ITS1, respectively. The outer nested-PCR primers (NSA3/NLC2) work well with all other fungal primer pairs (Figure 1) in this suite and provide greater sensitivity and specificity (and can be used to verify the identity of the product). Two primers specific to Plantae were also developed (NSIP/NLBP), and for our research can be used to verify that fungal PCR product is not contaminated by plant sequences (Table 1, Figure 2 and see Additional File 1). These primers are situated at the best estimate of the positions homologous to the fungal primers, NSI1 and NLB4, by multiple sequence alignment. The plant primers have been applied successfully to LH-PCR and PCR-RFLP analyses of forest plant tissues from above- and below-ground samples and work well at discriminating a representative sampling of plants from a forest in the Cascade Mountains in Oregon USA to the species level. Length heterogeneity was sufficient to distinguish more than 20 understory species (data not shown) in an old-growth Douglas-fir forest (Pseudotsuga menziesii (Mirbel) Franco), but the conifers [Douglas-fir, silver fir (Abies amabilis (Dougl.) Forbes), and hemlock (Tsuga heterophylla (Raf.) Sarg.)] required PCR-RFLP to obtain species level identification.
To ensure that the suite of fungal primers was amplifying the intended target sequences, fungal PCR products were verified in three ways. Firstly, PCR-RFLP patterns obtained under the current PCR protocol were compared to known patterns derived from sequences for the Saccharomyces cerevisiae ribosomal operon (SCYLR154C, Z73326, S. cerevisiae chromosome XII) using three restriction enzymes to ensure that fidelity was maintained. Secondly, PCR-RFLP patterns for ITS1 or ITS2 PCR reactions were checked against the same patterns derived from a nested PCR starting with the NSA3/NLC2 first-round primer pair. In this case, insufficient genomic DNA was carried over to the second round to amplify on its own so the product of the inner, second-round primers must have arisen from target sequences within the first product and not from another part of the genome. Thirdly, for DNA extracts with high percentages of plant DNA, the plant primers NSIP and NLBP were used to characterize the product and/or patterns that would result from amplifying plant sequences because of insufficient stringency. These methods were also incorporated into a quality assurance protocol.
Efficacy of the new fungal PCR system
The primer pair NSI1/NLB4 has successfully amplified 32 species of fungal sporocarps in 15 genera collected from geographically diverse locales in Oregon (all basidiomycetes but Tuber). As can be seen in Figure 3, the high diversity of the ITS sequences is evident from the variation in PCR product lengths. These amplifications were used to perform PCR-RFLP cluster analysis (Figure 4). The primer pair NSI1/NLB4 also has successfully amplified 26 species in 15 genera of dilution-plate isolated ascomycetous soil microfungi. No Zygomycetes glomus species were tested and from sequence data it appears unlikely that they would amplify (Figure 2). The mycorrhizal Ascomycetes, Cenoccocum (identified visually and by the presence of the ITS1 intron [18]) and Tuber melanosporum, have amplified well and given restriction patterns consistent with their published sequences (Table 2). The ITS1 intron is also found in Hymenoscyphus ericae [19]. Occurrence of this ITS1 intron may be widespread enough to affect some studies where the large size of the amplicons (800–1000 base pairs) causes technical difficulties. If necessary, the forward primer, ITS1 [5], may be used with NLB4 as an alternative since this non-specific primer is downstream of the intron and can be used to avoid amplifying the intron.
The capacity of these primers to amplify fungal targets has encouraged us to develop numerous collaborations where the primers have been applied to a wide range of techniques. Examples of applying the primers in current projects include:
1) amplified ectomycorrhizal fungi from more than 2000 single-root-tip DNA extracts [20],
2) the primer pair NSI1/NLB4 has successfully amplified 30 species in 18 genera of plate-isolated ascomycetous soil microfungi (identified by Drs. Lidia Watrud and Jeffrey Stone, personal communication),
3) PCR-RFLP, T-RFLP and direct-sequencing analyses of individual morphotypes and whole-core extracts of ectomycorrhizal fungi on Loblolly pine grown in North Carolina USA [21] (Burke et al, personal communication),
4) characterized ectomycorrhizal fungi found on roots of eucalypts grown in Uruguay [17],
5) analysis of epiphytic fungi on agricultural crops by QPCR and LH-PCR [22].
Of the two primers situated in 5.8S, 58A1F gives slightly more robust PCR reactions than 58A2F, but based on sequence comparisons, is more likely to have difficulty amplifying Cenococcum targets. Both work well with NLB4, the LSU reverse primer used for ITS2 analysis. They both lay within the sequence for ITS3 [5] and are shorter by two bases: 58A1F being shorter on the 5' end and 58A2F being shorter on the 3' end.
Xanthogenate/Tween rapid extraction method
In this study we also evaluated possible increases in efficiency of analyses by using the Xanthogenate/Tween (X/T) rapid extraction method instead of the CTAB/chloroform extraction. We initially tested the X/T method with additions of DNase to confirm the ability of the solution to protect DNA (commercially available S. cerevisiae genomic DNA, Promega cat. no. G3101). Comparing agarose gels showed that DNA degradation or loss attributable to DNase were not detectable; the same result was obtained with the CTAB extraction. DNA yields from single root tips using X/T were sufficient to give useful PCR reactions for more than half of the samples analyzed. Around 80% of single root tips extracted by the CTAB method yielded amplifiable fungal DNA using the nested amplification presented in this study. There are approximately half as many transfers in the X/T method as in the CTAB method and fewer centrifugations, so we are able to process four times as many samples in the same time. The fungal tissue of the ectomycorrhizas we were studying is concentrated mostly on the outside of the ectomycorrhizal root tips. By not grinding the samples, and therefore extracting relatively more from the surface than the interior of the ectomycorrhizal root tip, we also decrease the amount of plant DNA and phenolics in the extracts. Consequently, for the analyses where the goal was PCR-RFLP verification of ectomycorrhizal roots first classified using gross morphological traits, only the more healthy mycorrhizas would likely produce strong patterns for the primary fungal symbionts. This was seen potentially as a positive attribute of the X/T extraction procedure. Those root tips that failed to yield amplifiable fungal DNA by the X/T procedure may have done so after CTAB extraction, but also may be senescent roots colonized by saprobic fungi. In other words, the high sensitivity of the CTAB extraction in conjunction with the nested PCR carries an increased risk of identifying saprobic fungi as ectomycorrhizal fungi where the ectomycorrhizal fungi are absent, senescent or otherwise poorly amplifiable. It may be that the decreased sensitivity of the X/T extraction procedure along with potential preferential extraction of DNA from healthy mycorrhizal fungi would help to overcome this risk. Researchers should bear in mind, though, that this bias towards root-surface fungal tissue could lead to underestimating fungal types that primarily reside within the root. The Xanthogenate/Tween extraction procedure should be seen as a complement to extractions involving tissue homogenization. The combination of these approaches could allow for some interesting possibilities such as localizing fungal distribution on and in the root.
Conclusion
The new suite of primers for Dikaryomycota presented here (specified using an extension of the Gargas and DePriest [23] numbering system) is anticipated to serve as the basis for a wide ranging system to analyze microbial communities, particularly in association with plants. In addition to the new plant primers, we are developing homologous primer sets for Oomycota and Zygomycota. By using the same sequence positions for multiple primer sets, we expect to maintain a high degree of comparability between PCR reactions. This approach also allows us to use an iterative approach to overcoming difficulties inherent in defining the taxonomic range of any particular primer set by developing "homologous" primers with differing target ranges. Emergent methods for molecular analysis of microbial communities are allowing for increasingly high sampling rates. Increased sampling rates are, in turn, the driving force behind more effectively characterizing these spatially-diverse and complex groups. Methods such as the Xanthogenate /Tween DNA extraction procedure similarly allow for rapid screening of large numbers of samples. Future studies of microbial communities are likely to rely heavily on methods that can be adapted to large-scale sample processing needed to make efficient use of increasingly rapid and powerful molecular technologies.
Methods
Primer description
Primers were developed using multiple sequence alignments [24] of a representative taxonomic range of fungal and plant sequences downloaded from EMBL [25]. Nomenclature developed by Gargas and DePriest [23] for fungal primers was used with the exception that the sequence they refer to covers only the 1798 bases of 18S. We used the complementary sequence for Saccharomyces cerevisiae chromosome XII (EMBL ID SCYLR154C) to extend the numbering beyond the final base of 18S (EMBL ID SCRGEA). In this extension, the complement to base 8356 of SCYLR154C is the final base, 1798, of SCRGEA. The resulting S. cerevisiae ribosomal operon numbering system can be calculated by subtracting the SCYLR154C base-position from 8356 and adding this difference to the 1798 bases of SCRGEA.
Test samples
Soil microfungi from litter and soil in the Cascade Mountains in Oregon USA were isolated on DRB medium [26] and were subcultured to malt agar and Czapek's with yeast extract [27] for microscopic examination. Isolates were identified on the basis of conidium formation and conidiophore structure [27-30]. Sporocarps were collected and archived from numerous locations in Oregon. They were identified by gross morphology and spore color [31,32]. Ectomycorrhizal root tips were sorted based on gross morphological features (e.g., branching pattern, color, mantle texture, size, shape of tip, luster) as described elsewhere [33].
DNA extraction
For most routine extractions of fungal tissue and our early extractions of ectomycorrhizal root tips, DNA was extracted using a CTAB protocol with 0.8% mercaptoethanol and incubating at 65°C for 1 hour [34]. An alternate extraction method modified from Ross [35] based on the action of xanthogenate [36] was later employed for rapid survey analyses of large numbers of ectomycorrhizal root tips. Single ectomycorrhizal root tips (approximately 1 mg tissue dry weight) were placed in each 1.5 mL tube with 600 μL Xanthine/Tween Buffer (100 mM Tris-HCl pH 7.5, 12.5 mM potassium ethyl xanthogenate (Fluka, Buchs, Switzerland, cat. no. 60040), 10 mM EDTA pH 8.0, 10% Tween 20, 500 mM NaCl). These were sonicated for 15 seconds and incubated 90–120 min at 60°C on a shaker. Tubes were then centrifuged 5 min at 10,000 × g to pellet undissolved tissue. Supernatant was removed to a new tube containing 1/5 volume of PEG/NaCl (20% PEG-8000/2.5 M NaCl [37]) and 8 μL 1.25 mg/mL Linear Acrylamide (Ambion, Austin, Texas, cat. no. 9520). This was mixed by gently tipping the tube, and the mixture was incubated at 30°C for 15 minutes. DNA was precipitated by centrifuging 5 minutes (room temp.) at 10,000 × g, and then recovered by removing the PEG solution. The final DNA pellet was obtained by rinsing twice with 200 μL 80% ice-cold ethanol while mixing by gently tipping the tubes, followed by centrifuging 5 minutes at 10,000 × g, and lastly, by drying the pellet under vacuum for 30–45 minutes. In both protocols DNA was resuspended in a final volume of 100 μL TE buffer (pH 8, 10 mM Tris: 1 mM EDTA) after extraction by heating to 37°C for 10 minutes.
PCR amplification
PCR was carried out in 50 μL reaction volumes using 2.0 mM MgCl2, 0.2 μM each primer, 0.2 mM dNTP, 0.5 mg [mL]-1 bovine serum albumin (BSA) and 0.04 U [μL]-1 FastStart Taq DNA polymerase (Roche Diagnostics GmbH, Mannheim, Germany) on a thermal cycler equipped with a heated lid. An initial denaturation and enzyme activation step of 6–10 minutes at 95°C was followed by amplification for 35 cycles at the following conditions: 30 seconds at 95°C, 40 seconds at 60°C, 40 seconds at 72°C. A final 5-minute extension at 72°C completed the protocol. PCR typically yielded between 50–60 ng amplicon [μL]-1.
Restriction analysis
PCR-RFLP for the ITS region was performed with three restriction endonucleases: CfoI, HinfI, and TaqI. After electrophoretic separation on 3% agarose gels containing ethidium bromide, images were captured using the AlphaImager 950, version 3.24 (Alpha Innotech Corporation, San Leandro, CA), image capture system and analyzed using the Gene Profiler program (Scanalytics, Inc., Fairfax, VA). After restriction patterns were quantified using Gene Profiler, data were collected in the associated database facility to generate a presences/absence binary-data file. This file was input to TreeConW software [38] to develop dendrograms using the Link algorithm [39], {Gdxy = (Nx + Ny) / (Nx + Ny + Nxy)}, and "Unweighted Pair-Group Method using Arithmetic averages" (UPGMA) clustering.
DNA quantification
Absolute amounts of target DNA sequences extracted from samples were quantified by real-time PCR (QPCR) employing the DNA-binding fluorophore SYBR® Green 1 (Invitrogen Life Technologies, Carlsbad, CA). QPCR was performed on a RotorGene 3000 centrifugal amplification system (Corbett Research, Mortlake, Australia). PCR was performed with the primer pair, NSI1/58A2R, in 15 μL reaction volumes containing between 10 ng and 10 pg purified DNA, 0.40x SYBR® green, 0.50 mg mL-1 BSA, 3.0 mM MgCl, 0.6 μm each primer, 0.2 mM dNTP, and 0.1 U [μL]-1 FastStart Taq DNA polymerase in 100 μL tubes. A dilution series containing known amounts of S. cerevisiae genomic DNA (Promega Corporation, Madison, WI, cat. no. G3101) was used as the standard for quantification of sample DNA in each morphotype class. Upon completing PCR, melting curve analysis was used to determine whether there was detectable primer-dimer contribution to the SYBR green fluorescence measurement of amplified DNA (identified as a distinct drop in fluorescence with increasing temperature at temperatures below 82°C where low-melting-point, non-target amplicons denature). Samples compromised by primer-dimer reactions were deleted from the analysis. Using a value of 17 femtograms per haploid genome of S. cerevisiae, we calculated Saccharomyces genome equivalents (SGE) per μL of extract volume.
Abbreviations
BSA (Bovine Serum Albumin), ITS (Internal Transcribed Spacer), LH-PCR (Length Heterogeneity PCR), LSU (Large SubUnit-coding sequence), PCR-RFLP (Polymerase Chain Reaction-Restriction Fragment Length Polymorphism), QPCR (Quantitative PCR), SSU (Small SubUnit-coding sequence) TRF (Terminal Restriction Fragment), T-RFLP (Terminal Restriction Fragment Length Polymorphism), X/T (Xanthogenate/Tween rapid extraction method)
Authors' contributions
Dr. Kendall J. Martin primarily was responsible for the research direction, conducting the research, and interpreting and presenting the results. Dr. Paul T. Rygiewicz instituted and coordinated the overarching research projects within which this work was hosted and played an advisory role in pursuing this research and developing the manuscript.
Supplementary Material
Additional File 1
Multiple alignments of published sequences for the primer sites. This is the full set of sequences aligned to show the range of potential compatibility for the primers. This set was greatly reduced to create Figure 2.
Click here for file
Acknowledgements
The research was funded, in part, by the U.S. Environmental Protection Agency. The manuscript was subjected to the Agency's peer and administrative reviews, and it was approved for publication. Mention of trade names or commercial products in this paper does not constitute endorsement or recommendation of use. We thank Dr. F.S. Sanchez for identifying sporocarps while working with us as a NATO Fellow, and Dr. V.L. Oliveira for assisting in developing the X/T rapid extraction method.
Figures and Tables
Figure 1 Diagram of primer locations in the ribosomal cassette consisting of SSU, ITS1, 5.8S, ITS2, and LSU rDNA. Primers are positioned above (forward primers) or below (reverse) their sequence positions. ITS1, ITS2, ITS3, and ITS4 from White et al. [5], primers ITS8mun, ITS9mun, ITS10mun, NL5mun, NL6Amun, NL6Bmun, NL8mun from Egger [16], primers ITS1-F, ITS4-B from Gardes and Bruns [6] and the remaining primers (NSA3, NSI1, 58A1F, 58A2F, 58A2R, NLB4, NLC2) from this study. Scale is in base pairs according to the extension of the Gargas and DePriest [23] nomenclature system described in this study.
Figure 2 Multiple alignments of published sequences for the primer sites. Primer sequence alignments are depicted with consensus bases blank and mismatch bases noted. The six bases at the 3' end of the primers are highlighted yellow. "Clusters" were created by sorting the sequences in Excel and grouping identical sequences. "Species per cluster" numbers indicate the number of species with that mismatch pattern. Only one sequence per species was included in the alignment (final totals of: 584 species for 18S, 633 species for 5.8S, and 943 species for 28S). The "Cluster number" is a sequential numbering of clusters in order of descending "Species per cluster" with sequences from the EMBL Fungal database numbered before sequences from the EMBL Plant database (the latter are presented in green font). Sequences were collected using FASTA [40] for a set of 12 taxonomically representative fungal sequences for each region. Most mismatch groups containing fewer than 4 species were removed from the figure to save space. The full alignment is also available [see Additional file 1].
Figure 3 Example of the nested NSI1/NLB4 PCR reaction applied to single-species fungal DNA extracts. Ethidium Bromide fluorescence image showing electrophoresis of nested NSI1/NLB4 PCR product for a group of sporocarp extracts. Gel was run with 0.7% high melting point agarose plus 2% Synergel additive (Diversified Biotech, Boston, MA, cat. no. SYN-100). Roche DNA molecular weight marker VIII (cat. no. 1336045) was loaded at 100 ng in lanes 3, 8 and 13 (sizes for fragments to the left of the figure). Ten μL of PCR reaction was loaded in each well for each of the templates listed at the top of the gel as well as the no-template control in the last well.
Figure 4 Dendrogram and PCR-RFLP patterns for sporocarps. The upper section shows the species names and the cluster tree resulting from presences/absence analysis of the PCR-RFLP data. PCR-RFLP fragment sizes are presented below in base pairs for Taq1, HinF1 and Cfo1 restrictions as indicated at the right of the figure.
Table 1 Sequence characteristics of primers developed in this study.
Name Sequence (5' to 3') Nomenclature Bases Tm
NSA3 AAACTCTGTCGTGCTGGGGATA nu-SSU-1543-5' 22 67
NSI1 GATTGAATGGCTTAGTGAGG nu-SSU-1671-5' 20 59
58A1F GCATCGATGAAGAACGC nu-5.8S-2206-3' 17 61
58A2F ATCGATGAAGAACGCAG nu-5.8S-2208-3' 17 57
58A2R CTGCGTTCTTCATCGAT nu-5.8S-2192-5' 17 57
NLB3 GGATTCTCACCCTCTATGA nu-LSU-2754-5' 19 56
NLB4 GGATTCTCACCCTCTATGAC nu-LSU-2755-5' 20 57
NLC2 GAGCTGCATTCCCAAACAACTC nu-LSU-2821-5' 22 67
Plant control primers
NSIP GATTGAATGATCCGGTGAAG PC-nu-SSU-1671-5' 20 62
NLBP GCTGTCACCCTCTCAGGC PC-nu-LSU-2755-5' 18 64
Sequences are all listed 5' to 3' for the oligonucleotides. Nomenclature after Gargas and DePriest [23] as extended in this study. Tm is calculated by nearest neighbour method as an on-line service provided by Sigma-Genosys (The Woodlands, TX).
==== Refs
Molina R Caldwell BA Castellano MA Horton T Smith JE G Bitton Mycorrhizae: Ectomycorrhizal fungi Encyclopedia of Environmental Microbiology 2002 2124 2132
Horton TR Bruns TD The molecular revolution in ectomycorrhizal ecology: peeking into the black-box Molecular Ecology 2001 10 1855 1871 11555231 10.1046/j.0962-1083.2001.01333.x
Goodman DM Trofymow JA Distribution of ectomycorrhizas in micro-habitats in mature and old-growth stands of Douglas-fir on southeastern Vancouver Island Soil Biol Biochem 1998 30 2127 2138 10.1016/S0038-0717(98)00094-7
Taylor AFS Fungal diversity in ectomycorrhizal communities: sampling effort and species detection Plant Soil 2002 244 19 28 10.1023/A:1020279815472
White TJ Bruns T Lee S Taylor JW Innis MA, Gelfand DH, Sninsky JJ, White TJ Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics PCR Protocols: A Guide to Methods and Applications 1990 New York: Academic Press Inc 315 322
Gardes M Bruns TD ITS primers with enhanced specificity for basidiomycetes – application to the identification of mycorrhizae and rusts Mol Ecol 1993 2 113 118 8180733
Gardes M Fortin J White TJ Bruns TD Taylor JW Identification of indigenous and introduced symbiotic fungi in mycorrhizae by amplification of the internal transcribed spacer Can J Bot 1991 69 180 190
Suzuki M Rappé MS Giovannoni SJ Kinetic bias in estimates of coastal picoplankton community structure obtained by measurements of small-subunit rRNA gene PCR amplicon length heterogeneity Appl Environ Microbiol 1998 64 4522 4529 9797317
Landeweert R Leeflang P Kuyper TW Hoffland E Rosling A Wernars K Smit E Molecular identification of ectomycorrhizal mycelium in soil horizons Appl Environ Microbiol 2003 69 327 333 12514012 10.1128/AEM.69.1.327-333.2003
Liu W-T Marsh TL Cheng H Forney LJ Characterization of microbial diversity by determining Terminal Restriction Fragment Length Polymorphisms of genes encoding 16S rRNA Appl Environ Microbiol 1997 63 4516 4522 9361437
Marsh TL Liu WT Forney LJ Cheng H Beginning a molecular analysis of the eukaryal community in activated sludge Water Science Technology 1998 37 455 460 10.1016/S0273-1223(98)00145-0
Cole JR Chai B Marsh TL Farris RJ Wang Q Kulam SA Chandra S McGarrell DM Schmidt TM Garrity GM Tiedje JM The Ribosomal Database Project (RDP-II): previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy Nucleic Acids Res 2003 31 442 443 12520046 10.1093/nar/gkg039
Higuchi R Fockler C Dollinger G Watson R Kinetic PCR analysis: real-time monitoring of DNA amplification reactions BioTechnology 1993 11 1026 1030 7764001 10.1038/nbt0993-1026
Higuchi R Dollinger G Walsh PS Griffith R Simultaneous amplification and detection of specific DNA sequences BioTechnology 1992 10 413 417 1368485 10.1038/nbt0492-413
Ririe KM Rasmussen RP Wittwer CT Product differentiation by analysis of DNA melting curves during the Polymerase Chain Reaction Anal Biochem 1997 245 154 160 9056205 10.1006/abio.1996.9916
Egger KN Molecular analysis of ectomycorrhizal fungal communities Can J Bot 1994 S1415 S1422
Malvárez G Oliveira VL A PCR/RFLP technique to characterize fungal species in Eucalyptus grandis Hill ex. Maiden ectomycorrhizas Mycorrhiza 2003 13 101 105 12682832
Shinohara ML LoBuglio KF Rogers SO Group-I intron family in the nuclear ribosomal RNA small subunit genes of Cenococcum geophilum isolates Curr Genet 1996 29 377 387 8598059 10.1007/s002940050059
Egger KN Osmond G Goodier JL Sequence and putative secondary structure of group I introns in the nuclear-encoded ribosomal RNA genes of the fungus Humenoscyphys ericae Biochimica et Biophysica Acta 1995 1261 275 278 7711072
Martin KJ Rygiewicz PT Fungal-specific PCR primers developed for analysis of the ITS region of environmental DNA extracts Agronomy Abstracts, Annual Meeting, American Society of Agronomy: Salt Lake City, UT 1999
Burke D Martin K Rygiewicz P Topa M Characterization of Ectomycorrhizal Fungal Communities Using TRFLP Soil Biol Biochem 2005
Martin KJ Brooks RA Phillips LL Lepo JE Combining real-time quantitative PCR with terminal restriction fragment length polymorphism methods to detect crop-bioterrorism agents Agronomy Abstracts, American Society of Agronomy: Denver, CO 2003
Gargas A DePriest PT A nomenclature for fungal PCR primers with examples from intron-containing SSU rDNA Mycologia 1996 88 745 748
Thompson JD Higgins DG Gibson TJ CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice Nucleic Acids Res 1994 22 4673 4680 7984417
Kulikova T Aldebert P Althorpe N Baker W Bates K Browne P van den Broek A Cochrane G Duggan K Eberhardt R Faruque N Garcia-Pastor M Harte N Kanz C Leinonen R Lin Q Lombard V Lopez R Mancuso R McHale M Nardone F Silventoinen V Stoehr P Stoesser G Tuli MA Tzouvara K Vaughan R Wu D Zhu W Apweiler R The EMBL Nucleotide Sequence Database Nucleic Acids Res 2004 32 D27 D30 14681351 10.1093/nar/gkh120
King AD JrHocking AD Pitt JI Dichloran-rose Bengal medium for enumeration and isolation of molds from foods Appl Environ Microbiol 1979 37 959 996 485140
Pitt J A Laboratory Guide to the Common Penicillium Species 1988 North Ryde, NSW: CSIRO
Carmichael JW Kendrick WB Conners IL Sigler L Genera of Hyphomycetes 1980 Edmonton: University of Alberta Press
Domsch KH Gams W Anderson T-H Compendium of Soil Fungi (Reprint with Supplements) 1993 New York: Academic Press
Nelson PE Toussoun TA Marasas WFO Fusarium Species: an Illustrated Manual for Identification 1983 University Park: Penn. State University Press
Arora D Mushrooms Demystified: A Comprehensive Guide to the Fleshy Fungi 1986 Berkeley: Ten Speed Press
Smith AH A Field Guide to Western Mushrooms 1975 Ann Arbor: University of Michigan Press
Rygiewicz PT Martin KJ Tuininga AR Morphological diversity of ectomycorrhizas on Douglas-fir (Pseudotsuga menziesii Mirb. Franco) seedlings grown under elevated atmospheric CO2 and temperature Oecologia 2000 124 299 308 10.1007/s004420000385
Baker DD Mullin B Diversity of Frankia nodule endophytes of the actinorhizal shrub Ceanothus as assessed by RFLP patterns from single nodule lobes Soil Biol Biochem 1994 26 547 552 10.1016/0038-0717(94)90241-0
Ross IK Non-grinding method of DNA isolation from human pathogenic filamentous fungi using xanthogenates BioTechniques 1995 18 828 830 7619488
Jhingan AK A novel technology for DNA isolation Methods Mol Cell Biol 1992 3 15 22
Ioannou YA Chen FW Quantitation of DNA fragmentation in apoptosis Nucleic Acids Res 1996 24 992 3 8600475 10.1093/nar/24.5.992
van de Peer Y De Wachter R TREECON for Windows: a software package for the construction and drawing of evolutionary trees for the Microsoft Windows environment Comput Applic Biosci 1994 10 569 570
Link W Dixens C Singh M Schwall M Melchinger AE Genetic diversity in European and Mediterranean faba bean germ plasm revealed by RAPD markers Theor Appl Genet 1995 90 27 32 10.1007/BF00220992
Pearson WR Rapid and Sensitive Sequence Comparison with FASTP and FASTA Methods in Enzymology 1990 183 63 98 2156132
| 15904497 | PMC1156903 | CC BY | 2021-01-04 16:03:39 | no | BMC Microbiol. 2005 May 18; 5:28 | utf-8 | BMC Microbiol | 2,005 | 10.1186/1471-2180-5-28 | oa_comm |
==== Front
BMC NeurosciBMC Neuroscience1471-2202BioMed Central London 1471-2202-6-361589288910.1186/1471-2202-6-36Research ArticleHippocampal lesions facilitate instrumental learning with delayed reinforcement but induce impulsive choice in rats Cheung Timothy HC [email protected] Rudolf N [email protected] Department of Experimental Psychology, University of Cambridge, Downing Street, Cambridge CB2 3EB, UK2 Psychopharmacology Section, Division of Psychiatry, B Floor, Medical School, Queen's Medical Centre, Nottingham NG7 2UH, UK2005 13 5 2005 6 36 36 24 3 2005 13 5 2005 Copyright © 2005 Cheung and Cardinal; licensee BioMed Central Ltd.2005Cheung and Cardinal; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Animals must frequently act to influence the world even when the reinforcing outcomes of their actions are delayed. Learning with action-outcome delays is a complex problem, and little is known of the neural mechanisms that bridge such delays. When outcomes are delayed, they may be attributed to (or associated with) the action that caused them, or mistakenly attributed to other stimuli, such as the environmental context. Consequently, animals that are poor at forming context-outcome associations might learn action-outcome associations better with delayed reinforcement than normal animals. The hippocampus contributes to the representation of environmental context, being required for aspects of contextual conditioning. We therefore hypothesized that animals with hippocampal lesions would be better than normal animals at learning to act on the basis of delayed reinforcement. We tested the ability of hippocampal-lesioned rats to learn a free-operant instrumental response using delayed reinforcement, and what is potentially a related ability – the ability to exhibit self-controlled choice, or to sacrifice an immediate, small reward in order to obtain a delayed but larger reward.
Results
Rats with sham or excitotoxic hippocampal lesions acquired an instrumental response with different delays (0, 10, or 20 s) between the response and reinforcer delivery. These delays retarded learning in normal rats. Hippocampal-lesioned rats responded slightly less than sham-operated controls in the absence of delays, but they became better at learning (relative to shams) as the delays increased; delays impaired learning less in hippocampal-lesioned rats than in shams. In contrast, lesioned rats exhibited impulsive choice, preferring an immediate, small reward to a delayed, larger reward, even though they preferred the large reward when it was not delayed.
Conclusion
These results support the view that the hippocampus hinders action-outcome learning with delayed outcomes, perhaps because it promotes the formation of context-outcome associations instead. However, although lesioned rats were better at learning with delayed reinforcement, they were worse at choosing it, suggesting that self-controlled choice and learning with delayed reinforcement tax different psychological processes.
==== Body
Background
When one event or stimulus in the world reliably precedes and predicts another, animals readily learn the predictive relationship, exemplified by Pavlovian conditioning. Similarly, when an animal's own actions cause (and thus predict) some outcome, animals learn this relationship (an aspect of instrumental or operant conditioning). Frequently, however, antecedent and consequent events are separated in time. When animals act to obtain reinforcement, the final outcomes do not always follow the actions immediately; thus, animals must learn instrumental action-outcome contingencies using delayed reinforcement. Delays can hamper both Pavlovian and instrumental conditioning [1-5]: for example, although animals can bridge substantial delays to acquire instrumental responses, instrumental conditioning has long been observed to be systematically impaired as the outcome is delayed [6-11]. Furthermore, individual variation in the ability to use delayed reinforcement may determine one aspect of impulsivity: an animal able to forgo short-term poor rewards in order to obtain delayed but better rewards may be termed self-controlled, whereas an animal that cannot tolerate delays to reward may be said to exhibit impulsive choice [12-15].
There are several psychological reasons why action-outcome delays might impair learning or performance of an instrumental response [12,16]. Instrumental responding is controlled by several processes [4,17,18]; for example, rats work for outcomes that they value, using knowledge of the action-outcome contingencies in force to produce goal-directed actions. They also develop direct stimulus-response (S-R) associations, or habits. Action-outcome delays might, therefore, reduce the instrumental incentive value of the goal: valuing the goal less, animals may work less for it. Similarly, delays may hinder animals' ability to perceive the action-outcome contingency. Delayed rewards may also be less effective at reinforcing S-R habits. It is presently not known whether responses acquired with delayed reinforcement are governed by a different balance of habits and goal-directed actions than responses acquired with immediate reinforcement. However, one important factor in learning to act using delayed reinforcement may be the role of the environmental context. The animal's task is to attribute the outcome to its actions; instead, it may erroneously associate the outcome with the context, since the context is a cue that is temporally closer to the outcome than the action. The longer the delay, the more this contextual competition comes to impair the learning of the action-outcome contingency. Instrumental conditioning with delayed reinforcement can be enhanced if rats are exposed to the relevant contextual cues prior to instrumental training, and this enhancement is lessened if 'free' (non-contingent) rewards are given during the contextual pre-exposure periods [9,17]. These results are consistent with the theory that during the action-outcome delay, contextual cues compete with the action to become associated with the outcome; pre-exposing the animals to the context with no consequences reduces this contextual competition, by making the context a bad predictor of the outcome (perhaps via latent inhibition or learned irrelevance), and this in turn makes the action-outcome contingency more salient and easier to learn [9,17].
Little is known of the neural basis of instrumental learning with delayed reinforcement [16]. However, there is good evidence that the hippocampus contributes to the representation of context. Lesions of the hippocampal formation (H) have been shown to impair Pavlovian conditioning to a contextual conditioned stimulus (CS), but not to a discrete CS, in rats [19-31], at least for some processes involving contextual representation [32-34]. Since context-outcome associations are thought to hinder instrumental learning with delayed reinforcement (contextual competition) [9,17], it follows that if H lesions impair the formation of associations involving the context, such lesions might reduce contextual competition and hence facilitate instrumental conditioning when there is an action-outcome delay.
To investigate whether the hippocampus contributes to learning with delayed reinforcement, we examined the ability of rats with excitotoxic lesions of the hippocampus to acquire instrumental responding with delayed reward, comparing them to sham-operated controls. Each subject was allowed to respond freely on two levers, one of which produced reinforcement after a delay of 0, 10, or 20 s (Figure 1). We report that H-lesioned rats were slightly impaired at learning the lever-press response in the absence of delays. Delays retarded learning in sham-operated controls, but the delays did not impair the H-lesioned rats to the same extent. Thus, as the delays were increased, H-lesioned rats became better at learning relative to controls, suggesting that the presence of delays had less of an effect on H-lesioned rats. To establish whether this relative improvement in learning with delayed reinforcement would also manifest itself as improved self-control, we also trained a different group of rats on a task in which they had to choose between an immediate, small reward and a delayed, large reward (Figure 2) and made excitotoxic hippocampal lesions before retesting the rats postoperatively. Good learning with delayed reinforcement did not translate to self-controlled choice. We report that H lesions severely impaired rats' ability to choose the larger reward when it was delayed, but not when the delay preceding delivery of the large reward was removed, demonstrating that hippocampal lesions induce impulsive choice.
Figure 1 Task schematic: free-operant instrumental responding on a fixed-ratio-1 (FR-1) schedule with delayed reinforcement. Subjects are offered two levers; one (the active lever) delivers a single food pellet for every press (an FR-1 schedule) and the other (the inactive lever) has no programmed consequence. Food can either be delivered immediately (a) or after a delay (b) following responses on the active lever. The levers remain available throughout the session (hence, free-operant responding). Events of interest are lever presses, delivery of food pellets, and collection of food by the rat (when it pokes its nose into the food alcove following food delivery). To obtain food, the hungry rat must discriminate the active from the inactive lever, which is more difficult when the outcome is delayed. In these examples, the rat's response patterns (active and inactive lever presses, and collection of food) are fictional, while food delivery is contingent upon active lever pressing.
Figure 2 Task schematic: choice between small, immediate and large, delayed reward. Delayed reinforcement choice task [35, 86, 99], based on that of Evenden and Ryan [107]. Hungry rats regularly choose between two levers. Responding on one lever leads to the immediate delivery of a small food reward (1 pellet); responding on the other leads to a much larger food reward (4 pellets), but this reward is delayed for between 0 and 60 seconds. The figure shows the format of a single trial; trials begin at regular intervals (every 100 s), so choice of the small reinforcer is always suboptimal. Sessions consist of 5 blocks. In each block, two single-lever trials are given (one trial for each lever, in random order), to ensure the animals sample the options available at that time; these are followed by ten choice trials. The delay to the large reinforcer is varied systematically across the session: delays for each block are 0, 10, 20, 40, and 60 s respectively.
Results
Histology
In Experiment 1, there were five postoperative deaths. No rats were excluded after histological analysis; final group sizes were 9 (H, 0 s delay), 5 (sham, 0 s delay), 9 (H, 10 s delay), 6 (sham, 10 s delay), 9 (H, 20 s delay), and 5 (sham, 20 s delay). In Experiment 2, there was one postoperative death (H group), and one rat (sham group) fell ill five sessions after surgery and was killed. Histological analysis revealed that the lesions were incomplete or encroached significantly on neighbouring structures in 3 subjects. These subjects were excluded; final group sizes were therefore 7 (sham) and 12 (H).
A diagram of the rat hippocampus is shown in Figure 3. Lesions of the hippocampus encompassed much of the dorsal and ventral hippocampal pyramidal cell (cornu ammonis CA1-CA3) fields, the dentate gyrus, the subiculum, and the fimbriae. Neuronal loss and associated gliosis extended in an anteroposterior direction from approximately -0.8 mm to -7.8 mm relative to bregma (negative coordinates are posterior). Damage to the dorsal and ventral hippocampal commissure was occasionally seen, but damage to the overlying cortex was minimal. Schematics of the lesions are shown in Figure 4, and photomicrographs of a representative lesion are shown in Figure 5.
Figure 3 Diagram of the rat hippocampus. Drawings of the rat brain showing the three-dimensional organization of the hippocampus and related structures. Three coronal sections through the left hippocampus are shown at the bottom right of the figure, with their approximate anteroposterior coordinate relative to bregma. CA1, CA2, CA3: cornu ammonis fields 1–3; DG: dentate gyrus; EC: entorhinal cortex; f: fornix; s: septal pole of the hippocampus; S: subiculum; t: temporal pole of the hippocampus. Adapted from Figure 1 of ref. [113], copyright (1995), with permission from Elsevier.
Figure 4 Schematic of lesions of the hippocampus. Black shading indicates the extent of neuronal loss common to all subjects (and also the third and lateral ventricles); grey indicates the area lesioned in at least one subject. Coronal sections are (from top to bottom) -1.8, -2.8, -3.8, -4.8, -5.8, and -6.8 mm relative to bregma. Diagrams are modified from ref. [114]. Panels a-c show schematics for Experiment 1 (acquisition of a free-operant instrumental response with delayed reinforcement; 0 s, 10 s, and 20 s groups, respectively) while d shows schematics for Experiment 2 (choice between small, immediate and large, delayed reinforcement).
Figure 5 Photomicrographs of lesions of the hippocampus. Lesions of the hippocampus: photomicrographs of sections ~4.7 mm posterior to bregma, stained with cresyl violet. (a) Sham-operated rat, dorsal hippocampus, right hemisphere (medial to the left). CA1, cornu ammonis field 1; CA3, cornu ammonis field 3; DG, dentate gyrus; cc, corpus callosum; PtA, parietal association cortex. (b) Hippocampal-lesioned rat; same area as (a). There is tissue collapse within the lesion and the ventricle is greatly expanded. (c) Sham-operated rat, ventral hippocampus. AHiPM, amygdalohippocampal area, posteromedial part; cp, cerebral peduncle. (d) Hippocampal-lesioned rat, same area as (c). (e) Coronal diagram of the rat brain at 4.8 mm posterior to bregma [114], with scale. The upper grey box indicates approximately the region shown in (a) and (b); the lower grey box indicates approximately the region shown in (c) and (d).
Acquisition of instrumental responding (experiment 1)
As expected, response-reinforcer delays retarded the acquisition of instrumental responding in sham-operated rats (Figure 6a). However, this impairment was lessened in H-lesioned rats (Figure 6b). H-lesioned rats responded less than shams in the absence of a response-reinforcer delay (Figure 7a), but responded as well as shams when delays were imposed (Figure 7b,c); H-lesioned rats were even facilitated numerically relative to shams in the 20 s delay condition (Figure 7c), though this difference was not statistically significant on its own. These conclusions were reached statistically as follows.
Figure 6 Effects of delays to reinforcement on acquisition of free-operant responding under an FR-1 schedule. Data plotted to show the effects of delays. All groups discriminated between the active and the inactive lever, and delays retarded acquisition of the active lever response in both groups. (a) Responding of sham-operated control rats, under all three response-reinforcer delay conditions. (b) Responding of hippocampal-lesioned rats under all delay conditions. The next figure replots these data to show the effect of the lesion more clearly.
Figure 7 Effect of hippocampal lesions on acquisition of free-operant responding with delayed reinforcement. Data plotted to show the effects of hippocampal lesions (same data as in the previous figure). There was a delay-dependent impairment in H-lesioned rats (significant lesion × delay interaction, see text), who learned less well than shams only when reinforcement was not delayed. (a) With a delay of 0 s, H-lesioned rats responded less on the active lever than shams did. (b) With a 10 s delay, H-lesioned rats responded the same as shams. (c) With a 20 s delay, H-lesioned rats responded more than shams on the active lever, though this difference was not statistically significant on its own.
An overall ANOVA, using the model lesion2 × delay3 × (session14 × lever2 × S), revealed a lesion × lever × delay interaction (F2,37 = 4.16, p = .023), justifying sub-analyses, in addition to effects of delay (F2,37 = 17.9, p < .001), lever (F1,37 = 435, p < .001), delay × lever (F2,37 = 4.16, p < .001), session (F5.35,198.0 = 38.7, = .412, p < .001), delay × session (F10.7,198.0 = 3.03, p = .001), session × lever (F4.99,184.6 = 17.5, = .384, p < .001), and delay × session × lever (F10.0,184.6 = 2.30, = .384, p = .015). The differences between the groups were in their responding on the active lever (active lever, lesion × delay: F2,37 = 3.71, p = .034) rather than on the inactive lever (inactive lever, terms involving lesion: maximum F2,37 = 1.146, NS). All six groups learned to discriminate between the two levers, responding more on the active lever than on the inactive lever (p < .05, main effect of lever for each group).
Delays reduced the rate of acquisition and the final level of responding on the active lever for sham-operated rats (Figure 6a; delay, F2,13 = 58.7, p < .001; delay × session, F10.8,70.4 = 2.67, = .417, p = .007). Delays also increased responding on the inactive lever somewhat (Figure 6a; delay: F2,13 = 5.26, p = .021; delay × session, F13.1,85.2 = 1.22, = .504, NS). Similarly, in H-lesioned rats, delays reduced responding on the active lever (Figure 6b; delay: F2,24 = 12.3, p < .001; delay × session: F7.8,93.1 = 2.76, = .298, p = .009), although they did not significantly affect responding on the inactive lever (delay: F2,24 = 1.91, NS; delay × session: F12.3,147.9 = 1.37, = .474, NS).
At 0 s delay, H-lesioned rats responded significantly less than shams on the active lever (Figure 7a; lesion: F1,12 = 6.11, p = .029). There were no differences in responding on the inactive lever (Fs < 1, NS). At 10 s delay, there were no differences between sham-operated and H-lesioned rats in responding on either the active or the inactive lever (Fs < 1.35, p ≥ .266). At 20 s delay, there were also no significant differences on either lever (active lever: lesion F1,12 = 2.485, p = .141, lesion × session F < 1, NS; inactive lever: Fs < 1, NS), although the H-lesioned rats responded numerically more than shams on the active lever throughout.
Inspection of Figure 6 also suggested that delays had less of an impact on the final (asymptotic) rates of responding in H-lesioned rats than in shams. The sessions were divided by eye into an acquisition phase (sessions 1–6) and a 'stable' phase (sessions 7–14). Responding on the active lever in the 'stable' phase was analysed; this revealed a lesion × delay interaction (F2,37 = 3.44, p = .043), with delays markedly reducing stable rates of responding in shams (F2,13 = 42.3, p < .001) but less so in H-lesioned rats (F2,24 = 3.11, p = .063).
Experienced response-delivery and response-collection delays (experiment 1)
For every reinforcer delivered, the active lever response most closely preceding it in time was identified, and the time between that response and delivery of the reinforcer (the 'response-delivery delay') was calculated. This time can therefore be equal to or less than the programmed delay, and is only relevant for subjects experiencing non-zero programmed response-reinforcer delays. The response-to-reinforcer-collection ('response-collection') delays were also calculated: for every reinforcer delivered, the response most closely preceding it and the nosepoke most closely following it were identified, and the time between these two events calculated. This time can be shorter or longer than the programmed delay, and is relevant for all subjects.
H-lesioned rats experienced slightly shorter response-delivery delays than shams when the programmed delay was 10 s or 20 s (Figure 8a): there was a lesion × programmed delay interaction (F1,25 = 6.28, p = .019), and simple effects of the lesion when the programmed delay was 10 s (F1,13 = 8.49, p = .012) and when it was 20 s (F1,12 = 9.50, p = .009).
Figure 8 Programmed and experienced delays to reinforcement. H-lesioned rats experienced slightly shorter response-delivery delays (the delay between the most recent lever press and pellet delivery) than shams, and slightly shorter response-collection delays (the delay between the most recent lever press and pellet collection). (a) Mean experienced response-delivery delays (one value calculated per subject). When the programmed delay was 0 s, reinforcers were delivered immediately so no data are shown. H-lesioned rats experienced shorter response-delivery delays when the programmed delay was 10 s (* p = .012) or 20 s (** p = .009). (b) Mean experienced response-collection delays (one value calculated per subject). H-lesioned rats experienced slightly shorter delays overall (* p = .047, main effect of lesion), but the experienced delays did not differ significantly at any given programmed delay. (c) Distribution of experienced response-delivery delays. All experienced delays for a given subject were aggregated across all sessions, and the proportion falling into different 2-s ranges were calculated to give one value per range per subject; the graphs show means ± SEMs of these values. The interval notation '[a, b)' indicates that a given delay x falls in the range a ≤ x <b. H-lesioned rats experienced slightly fewer long delays and slightly more short delays in the 10 s condition (p = .019) and in the 20 s condition (p = .014). (d) Distribution of experienced response-collection delays, displayed in the same manner as (c). There were no differences in the distribution of delays experienced by H-lesioned and sham rats in the 0 s condition. In the 10 s condition, H-lesioned rats experienced a slightly lower proportion of long delays and a slightly higher proportion of short delays (p = .009), and similarly in the 20 s condition (p = .001).
H-lesioned rats also experienced slightly shorter response-collection delays across all programmed delays (Figure 8b) (lesion: F1,37 = 4.21, p = .047), though the difference was not significant at any one programmed delay (lesion × programmed delay: F2,37 = 2.35, p = .109; simple effects of the lesion at different programmed delays: maximum F1,12 = 3.08, p = .105).
These differences in the mean delay experienced by each rat were reflected in differences in the distribution of response-delivery and response-collection delays when the programmed delay was non-zero (Figure 8c,d). All experienced delays for a given subject were aggregated across all sessions, and the proportion falling into different 2-s ranges were calculated to give one value per range per subject. For response-delivery delays, H-lesioned rats experienced slightly fewer long delays and slightly more short delays in the 10 s condition (lesion × range, F3.6,47.0 = 3.40, = .723, p = .019) and in the 20 s condition (lesion × range, F1.4,16.6 = 6.54, = .138, p = .014). For response-collection delays, there were no differences in the distribution of delays experienced by H-lesioned and sham rats in the 0 s condition (lesion and lesion × range, Fs < 1, NS). In the 10 s condition, H-lesioned rats experienced a slightly lower proportion of long response-collection delays and a slightly higher proportion of short response-collection delays (lesion × range, F4.4,57.6 = 3.60, = .233, p = .009). Similarly, in the 20 s condition, H-lesioned rats experienced a slightly lower proportion of long response-collection delays and a slightly higher proportion of short response-collection delays than shams (lesion × range, F3.1,37.5 = 7.02, = .164, p = .001).
Since H-lesioned rats experienced slightly shorter delays than sham-operated rats, it was necessary to take this into account when establishing the effect of delays on learning, as follows.
Effect of delays on learning (experiment 1)
There was a systematic relationship between the acquisition rate and the programmed delay of reinforcement, and this was altered in H-lesioned rats, who were less impaired by delays (compared to their performance at zero delay) than shams were. Figure 9a replots the rates of lever-pressing on session 6, at the end of the initial 'acquisition' phase. Despite the comparatively low power of such an analysis, lever-pressing was analysed for this session only, using the model lesion2 × delay3 × S. This revealed a significant lesion × delay interaction (F2,37 = 8.67, p = .001), which was analysed further. Increasing delays significantly reduced the rate of responding in this session for shams (F2,13 = 31.4, p < .001) and H-lesioned rats (F2,24 = 8.88, p = .001). H-lesioned rats responded less than shams at zero delay (F1,12 = 8.08, p = .015), were not significantly different from shams at 10 s delay (F1,13 = 1.848, p = .197), and responded more than shams at 20 s delay (F1,12 = 6.23, p = .028).
Figure 9 Learning as a function of programmed and experienced delays to reinforcement. The imposition of response-reinforcer delays systematically retarded the acquisition of free-operant instrumental responding, but this effect was lessened in H-lesioned rats, even allowing for differences in experienced response-collection delays. (a) The rate of lever-pressing in session 6 is plotted against the programmed response-reinforcer delay. There was a lesion × delay interaction (### p = .001): H-lesioned rats responded less than shams at zero delay (* p = .015), were not significantly different from shams at 10 s delay (p = .197), and responded more than shams at 20 s (* p = .028). (b) Responding in session 6 plotted against the experienced response-to-reinforcer collection delays for sessions 1–6 (vertical error bars: SEM of the square-root-transformed number of responses in session 6; horizontal error bars: SEM of the experienced response-collection delay, calculated up to and including that session). The gradients of the two lines differed significantly (## p = .002; see text), indicating that the relationship between experienced delays and responding was altered in H-lesioned rats.
Since the H group experienced slightly shorter response-delivery and response-collection delays than shams when the programmed delay was non-zero (Figure 8), it is important to establish whether this effect alone was responsible for the lesser effect of delays on learning, or whether the effect of delays on H-lesioned rats was lessened over and above any effect to decrease the experienced delay. The mean experienced response-collection delay was calculated for each subject up to and including session 6. The square-root-transformed number of lever-presses in session 6 was then analysed using a general linear model of the form lesion2 × experienced delaycov × S; unlike a standard analysis of covariance, the factor × covariate interaction term was included in the model. This confirmed that the detrimental effects of delay upon learning were reduced in H-lesioned rats, compared to controls, over and above the differences in experienced delay (Figure 9b; lesion × experienced delay: F1,39 = 10.8, p = .002).
Experienced delays and learning on the inactive lever (experiment 1)
No such delay-dependent lesion effects were observed for the inactive lever. Experienced inactive-response-delivery delays (calculated across all sessions in the same manner as for the active lever) were much longer and more variable than corresponding delays for the active lever, because subjects responded on the inactive lever so little. Means ± SEMs were 271 ± 31 s (sham, 0 s), 241 ± 23 s (H, 0 s), 201 ± 45 s (sham, 10 s), 184 ± 45 s (H, 10 s), 127 ± 21 s (sham, 20 s), and 171 ± 36 s (H, 20 s). ANOVA of these data showed that these experienced inactive-response-delivery delays depended upon the programmed active-response-delivery delay (delay: F2,37 = 3.80, p = .032) but there was no effect of the lesion and no lesion × delay interaction (Fs < 1, NS). Experienced inactive-response-collection delays were 272 ± 31 s (sham, 0 s), 242 ± 23 s (H, 0 s), 204 ± 45 s (sham, 10 s), 186 ± 45 s (H, 10 s), 130 ± 21 s (sham, 20 s), and 174 ± 35 s (H, 20 s). Again, ANOVA that these experienced delays depended upon the programmed active-response-delivery delays (delay: F2,37 = 3.68, p = .035) but there was no effect of the lesion and no lesion × delay interaction (Fs < 1, NS). When the square-root-transformed number of responses on the inactive lever in session 6 was analysed with the experienced delays up to that point as a predictor, using the model lesion2 × experienced inactive-response-collection delaycov × S just as for the active lever analysis, there was no lesion × experienced delay interaction; neither was there an effect of lesion or experienced delay (maximum F2,37 = 1.54, NS).
Choice between an immediate, small reward and a large, delayed reward (experiment 2)
Preoperatively, subjects preferred the larger reinforcer less when it was delayed, and the groups remained matched following later histological selection (Figure 10a). Choice ratios (percent choice of the large, delayed reinforcer, calculated for all free-choice trials on which subjects responded) from the last 3 preoperative sessions were analysed using the model lesion intent2 × (delay5 × S). While there was a strong effect of delay (F2.2,36.9 = 22.9, p < .001), no terms involving lesion intent were significant (Fs < 1, NS).
Figure 10 Effects of hippocampal lesions on choice between immediate, small reward and large, delayed reward. (a) Pattern of choice in the last three sessions before surgery; the sham and lesion groups were matched for performance. Rats' preference for the large reinforcer declined with delay (p < .001). (b) Choice in the first seven postoperative sessions. Although there was a change in behaviour in the lesioned group (lesion × pre/postop., p = .021), the difference between the two groups was not significant in its own right for these sessions (p = .08). (c) Effects of omitting all delays in alternating sessions (error bar, 2 SED for the three-way interaction). H-lesioned rats remained sensitive to the contingencies, altering their behaviour in response to delay omission, as shams did. (d) Last of six further consecutive sessions in which delays were omitted. Both groups preferred the large reinforcer strongly when it was not delayed, with no differences between sham and H-lesioned rats. (e) First three sessions following reintroduction of delays. Hippocampal-lesioned rats were impulsive, choosing the large, delayed reinforcer less often than shams (* p = .027). (f) Next three sessions following reintroduction of delays. Hippocampal-lesioned rats remained impulsive (** p = .007), and generalization between trial blocks occurred, reducing their preference for the large reinforcer in the zero-delay block as well (see text).
The choice patterns of the two groups diverged following surgery, with the H-lesioned rats choosing the large, delayed reinforcer less than sham-operated controls (Figure 10b). Comparison of choice in the last 3 preoperative sessions (Figure 10a) to that in the first 7 postoperative sessions (Figure 10b), using the model lesion2 × (pre/post2 × delay5 × S) revealed a lesion × pre/post interaction (F1,17 = 6.50, p = .021). However, at this point, analysis of postoperative choice patterns on their own (Figure 10b) did not reveal a significant difference between the two groups (lesion: F1,17 = 3.46, p = .08; delay × lesion: F < 1, NS); as it did not take account of preoperative choice patterns, this analysis was less powerful. Later, H-lesioned rats diverged further from sham-operated controls and the difference between the two became significant even without taking account of preoperative information (see below). Both groups remained sensitive to the delay postoperatively (sham, effect of delay: F2.1,12.8 = 4.59, = .531, p = .03; lesion, effect of delay: F1.5,16.5 = 7.05, = .374, p = .01).
There were no differences between H-lesioned and sham-operated rats in any other measures collected, including the rate of omissions, the latency to initiate trials, the latency to choose a lever, the latency to collect food, and the rate of nosepoking in the food alcove during delays to reinforcement. Data from the 7 baseline postoperative sessions (sessions 20–26) were analysed. Omissions were very infrequent (overall, rats failed to initiate and/or to press a lever on 0.2% of trials) and there were no group differences in the rates of omission (F < 1, NS). Initiation latencies did not differ between groups (lesion: F < 1, NS; lesion × delay, F2.8,48.0 = 1.027, = .706, NS). Neither did choice latencies: an analysis using the model lesion2 × (delay5 × lever2 × S) revealed no significant terms involving lesion (Fs < 1.06, NS). Food collection latencies were analysed using the model lesion2 × (choice2 × delay5 × S). Predictably, rats were slower to collect the food following choice of the large, delayed reinforcer as the delays got longer (choice × delay: F2.4,38.5 = 19.8, = .602, p < .001; effect of delay following choice of the small, immediate reinforcer: F2.5,39.7 = 1.63, = .62, NS; effect of delay following choice of the large, delayed reinforcer: F2.2,35.8 = 18.4, = .559, p < .001) but this was not influenced by the lesion (terms involving lesion, maximum F1,16 = 1.93, NS). The proportion of the delay spent nosepoking did not alter as a function of the delay, and was not affected by the lesion (only applicable to trials on which the large reinforcer was chosen with a non-zero delay; delay, F1.3,22.3 = 2.38, = .437, p = .131; lesion × delay, F1.3,22.3 = 1.53, = .437, NS; lesion: F < 1, NS).
Effects of removing and reintroducing delays to the large reinforcer (experiment 2)
Both H-lesioned and sham-operated rats were sensitive to the removal of delays in alternating sessions, increasing their preference for the large reinforcer during sessions when it was not delayed (Figure 10c). Choice ratios from these sessions were analysed using the model lesion2 × (delays/no delays2 × trial block5 × S). This revealed a delays/no delays × block interaction (F2.5,42.6 = 15.3, = .626, p < .001). Additionally, there was a main effect of lesion (F1,17 = 7.23, p = .016), indicating a greater overall preference for the smaller reinforcer across these sessions in H-lesioned rats compared to controls, but there were no other significant terms involving lesion (Fs < 1.05, NS). In sessions when delays were present, both H-lesioned and sham-operated rats showed a within-session shift in preference as the delay increased (sham, effect of delay: F4,24 = 4.79, p = .006; H-lesioned, effect of delay: F2.6,28.1 = 20.21, = .638, p < .001), and H-lesioned rats chose the smaller, immediate reinforcer more often (lesion: F1,17 = 5.91, p = .026). In sessions when delays were not present, neither group showed a within-session shift in preference (shams: F4,24 = 2.46, p = .073; H-lesioned: F2.1,23.5 = 1.63, = .534, NS), though again the H-lesioned rats showed a stronger preference for the smaller reinforcer (F1,17 = 6.61, p = .02).
These analyses suggested that the H-lesioned rats' preference for the larger reward was less than that of shams even when it was not delayed. However, an alternative possibility is that the H-lesioned rats were loath to choose the large reinforcer when it was delayed, and that this generalized to affect preference even when it was not delayed [35,36]. Consequently, subjects were given a further six sessions with no delays present; preference on the last of these sessions is shown in Figure 10d. H-lesioned rats showed a strong preference for the large reinforcer when it was not delayed, just as shams did, with mean choice ratios >94% in all conditions. In this session, there were no group differences (Fs < 1, NS) and no within-session shift in preference (Fs < 1, overall and for H-lesioned and sham-operated groups individually).
When delays were reintroduced (sessions 37–42), preference for the larger, delayed reinforcer declined much more sharply in H-lesioned rats than in shams (Figure 10e,f). Preference for the large reinforcer declined first at long delays, then progressively at shorter delays, such that even responding in the zero-delay block was affected. In sessions 37–39, H-lesioned rats chose the large reinforcer less often than shams (lesion: F1,17 = 5.90, p = .027; lesion × delay: F1.9,32.7 = 1.16, = .482, NS), with this difference being significant for 10 s and 20 s delays (p < .05) but not 0 s (p = .075), 40 s (p = .058), or 60 s delays (p = .054). In sessions 40–42, the pattern was essentially the same (lesion: F1,17 = 9.47, p = .007; lesion × delay: F1.6,28.0 = 1.286, = .412, NS), except that individual differences were now significant at all delays (p < .05).
Locomotor activity, body mass, and food consumption
H-lesioned animals were hyperactive compared to sham-operated controls in both experiments (Figure 11a,b), as reported previously [37,38]. In Experiment 1, analysis of the square-root-transformed number of infrared beam breaks using the model lesion2 × (bin12 × S) revealed effects of lesion (F1,41 = 9.77, p = .003), reflecting hyperactivity in the H group, with additional effects of bin (F8.2,335.7 = 58.4, = .744, p < .001), reflecting habituation, and a lesion × bin interaction (F8.2,335.7 = 2.95, = .744, p = .003). In Experiment 2, hyperactivity was again observed (lesion: F1,17 = 24.1, p < .001; bin: F8.6,145.7 = 15.9, = .779, p < .001; lesion × bin: F < 1, NS).
Figure 11 Locomotor activity in a novel environment and body mass. Hippocampal-lesioned rats were significantly hyperactive compared to sham-operated controls, in both (a) Experiment 1 (p = .003) and (b) Experiment 2 (p < .001). (c) Body mass across both experiments. There were no differences between groups preoperatively in either experiment. In Experiment 1, the groups gained weight at the same rate, but in Experiment 2, which lasted longer, the H-lesioned rats weighed less at the end of the experiment (p = .01).
In Experiment 1, H-lesioned rats remained the same weight as sham-operated controls throughout, though in Experiment 2, which lasted longer, they gained less weight than shams (Figure 11c). There were no differences between groups preoperatively in either experiment (Fs ≤ 1.35, NS). In Experiment 1, the groups gained weight at the same rate (lesion × time, F < 1, NS; group difference at second time point: F < 1, NS). Data from two H-lesioned subjects in Experiment 2 were lost. In Experiment 2, the H-lesioned rats weighed less at the end of the experiment (lesion × time, F1,15 = 14.5, p = .002; group difference at second time point: F1,15 = 8.56, p = .01).
H-lesioned rats consumed their maintenance chow more quickly and consumed more of it, but they did not differ from sham-operated controls in their consumption of the sucrose pellets employed as reinforcers in the behavioural tasks. In 30 minutes, H-lesioned rats consumed more chow (11.2 ± 0.6 g) than shams (8.2 ± 0.8 g) (F1,7 = 8.36, p = .01). However, there were no differences between the mass of sucrose pellets consumed in 30 minutes by H-lesioned rats (17.5 ± 1.2 g) and by shams (18.3 ± 1.6 g) (F < 1, NS). H-lesioned rats were quicker to consume 2.5 g of chow (taking 302 ± 18 s) than shams (385 ± 24 s) (Levene's test indicated significant heterogeneity of variance; Mann-Whitney U7,12 = 18, p = .045). However, although H-lesioned rats were also slightly quicker to consume 2.5 g of sucrose pellets (taking 160 ± 9 s) than shams (who took 176 ± 12 s), this difference was not significant (F1,17 = 1.24, p = .281).
Discussion
Excitotoxic lesions of the dorsal and ventral hippocampus slightly retarded instrumental learning on a continuous reinforcement (fixed-ratio-1; FR-1) schedule in the absence of response-reinforcer delays. However, H-lesioned rats were only impaired when reinforcement was delivered immediately, and not when it was delivered after a delay. H-lesioned rats were less sensitive to the deleterious effects of response-reinforcer delays on learning, to the extent that with long (20 s) response-reinforcer delays, H-lesioned rats showed numerically better discrimination between the active and inactive levers than shams (Figure 7, Figure 9). Despite this delay-dependent facilitation of instrumental conditioning, H-lesioned rats were less able than shams to choose a delayed, large reinforcer in preference to an immediate, small reinforcer (Figure 10). That is, H-lesioned rats exhibited impulsive choice.
Pavlovian and instrumental conditioning with delayed reinforcement
Free-operant instrumental conditioning, and instrumental discrimination learning, have long been known to be impaired systematically by response-reinforcer (action-outcome) delays [6-9]. This might be for several reasons [12,16,39] because instrumental responding depends on several processes, including knowledge of the action-outcome contingency, a representation of the instrumental incentive value of the outcome, S-R habits, and the influence of Pavlovian CSs that have motivational significance through processes such as conditioned reinforcement and Pavlovian-instrumental transfer [4,17,18]. Action-outcome delays might affect several of these processes. For example, such delays may hinder the subject's ability to perceive the action-outcome contingency, so the subject is unaware that its actions will result in the outcome. Delays might reduce the value of the goal, so the subject is less willing to work for it. Delays might also impair the process by which S-R habits are reinforced; finally, they might affect the degree to which stimuli associated with reinforcement by Pavlovian conditioning are capable of motivating behaviour. In the present experiment, we did not present explicit stimuli associated with either the response or the reinforcer, to minimize the possible contribution of processes such as conditioned reinforcement. Nevertheless, it is largely an open question which processes contributing to instrumental learning and performance are the ones most affected by response-reinforcer delays; for example, it is not present known whether responses acquired with delayed reinforcement are governed by a different balance of habits and goal-directed actions than responses acquired with immediate reinforcement.
However, the environmental context has been clearly demonstrated to influence learning with delays. The effect of contextual factors on learning has been demonstrated using two Pavlovian conditioning paradigms: delay conditioning and trace conditioning. When a CS is followed by a US and the two overlap and are contiguous in time, the paradigm is known as 'delay' conditioning (the terminology is somewhat confusing). When a CS is followed by a US and the two do not overlap, so that there is a gap between the end of the CS and the start of the US, the paradigm is known as 'trace' conditioning (because the US must be associated with a 'trace' of the CS) [39]. 'Trace' Pavlovian conditioning, with a CS-US gap, results in poorer learning than 'delay' Pavlovian conditioning, even if the interstimulus interval (ISI; the time between the onset of the CS and the onset of the US) is held constant [39-48], just as insertion of an action-outcome gap retards instrumental conditioning [7-9]. Several lines of evidence suggest that contextual competition is a cause of reduced responding to the discrete CS in trace conditioning. As the trace gap is lengthened, conditioned responses (CRs) tend to occur to the context instead of to the discrete CS [20,43,49-52]. Trace conditioning can be improved by the addition of a 'filler' stimulus during the CS-US gap, which might decrease contextual competition or act as a secondary or conditioned reinforcer [48,53,54]. The smaller the ratio of the ISI to the intertrial interval (ITI; the time between the end of the US and the start of the next trial's CS), the faster conditioning proceeds [5], and one explanation of this is that long ITIs reduce the strength of context-US associations, making CS-US associations more salient. Finally, pre-exposure to the context in the absence of any US improves subsequent conditioning to a discrete CS, as would be expected under the contextual competition account since pre-exposure should produce latent inhibition of the context (see [55]).
The idea that reinforcing outcomes may be associated with either a discrete predictor (such as a Pavlovian CS or an instrumental response) or the background context, and that the two compete in some way for such association, also explains observations concerning the effect of contextual manipulations on instrumental conditioning [9,17]. Dickinson et al. [9] trained rats on a free-operant, FR-1 schedule of reinforcement very similar to the one used in the present experiments. They found that the rate of learning, and the asymptotic level of responding, declined across groups as the response-reinforcer delay was increased from 0 to 32 s; rats trained with a 64-s delay failed to learn at all, compared to yoked controls. However, when rats were exposed to the training context prior to training (in the absence of the lever or any reinforcers) their learning was improved, and successful discrimination was seen even with a delay of 64 s. This is exactly what would be expected if a process of contextual competition was operating. The subject's task is to distinguish P(outcome | action) from P(outcome | no action), or, making the contribution of the context explicit, to distinguish P(outcome | action + context) from P(outcome | context). Pre-exposure to the context would be expected to produce latent inhibition to the context, reducing the strength of context-outcome associations. Viewed another way, non-reinforced exposure to the context forces the subjects to experience a zero-response, zero-reinforcer situation, i.e. P(outcome | context) = 0. When they are then exposed to the instrumental contingency, such that P(outcome | action + context) > 0, this prior experience may enhance their ability to detect the instrumental contingency ΔP = P(outcome | action) – P(outcome | no action). This interpretation is also supported by the demonstration that delivering 'free' rewards (not contingent upon any response of the subject) during the contextual pre-exposure reduces the beneficial effect of this pre-exposure on instrumental learning [9]; by increasing P(outcome | context), this reduces the subject's ability to detect the contingency [56-58]. Thus, the formation of context-outcome associations may explain the ability of action-outcome delays to retard instrumental learning.
Contribution of the hippocampus to instrumental conditioning with immediate reinforcement
In the present study, excitotoxic hippocampal lesions impaired instrumental conditioning on an FR-1 schedule in the absence of response-reinforcer delays. This contrasts with the findings of Corbit et al. [59] that electrolytic lesions of the dorsal hippocampus did not affect the acquisition of instrumental responding on a training schedule that progressed from a fixed-interval-20-s (FI-20) schedule up to a random-ratio-20 (RR-20) schedule, consistent with earlier results [60]. Rats with excitotoxic NMDA lesions of the dorsal hippocampus also responded at similar, or greater, rates than sham-operated controls in this training regimen [61]. Obviously, the discrepancy between these findings and the present results might be due to the differences in the schedules used (FR-1 versus RR-20) or in the lesion (dorsal + ventral hippocampus versus dorsal hippocampus only). However, it is less likely that the impairment was due to a primary motivational difference: although the H-lesioned rats gained less mass than shams, the food consumption tests showed that they ate as many of the pellets used as reinforcers as shams, and as quickly, suggesting that the impairment observed was not due to reduced primary motivation for the food. It is, however, possible that it represents a rate-dependent impairment (i.e. that the H-lesioned rats in the zero-delay condition were responding at their maximum possible rate).
Furthermore, electrolytic lesions of the dorsal hippocampus have previously been shown to render rats insensitive to changes in the instrumental action-outcome contingency, but in a very specific manner [59,61]. One way to test subjects' sensitivity to this contingency is to train them to respond on two levers for two different outcomes, and then to deliver one of the outcomes non-contingently, as well as contingent upon the response. Subjects that are sensitive to the action-outcome contingency should selectively reduce their responding for the foodstuff being delivered non-contingently [62]. Electrolytic dorsal hippocampal lesions impaired this ability, though not the ability to discriminate the two foodstuffs or to respond to changes in their value [59,61]. This may have been because the lesion affected contextual conditioning: if an animal cannot associate non-contingent rewards with the context, it may erroneously associate them with its own action. However, excitotoxic lesions of the dorsal hippocampus did not produce this effect, which was reproduced instead by excitotoxic lesions of the entorhinal cortex and subiculum [61]; lesions of these structures have also been shown to impair contextual conditioning in Pavlovian tasks [29,63-67] (though see [24,68]).
Contribution of the hippocampus to instrumental conditioning with delayed reinforcement
In contrast, when a delay was imposed between responding and reinforcement in an FR-1 schedule, H-lesioned rats were not impaired at instrumental conditioning, and were even somewhat facilitated in learning, relative to shams, when the reinforcer was delayed by 20 s. Since H-lesioned rats were impaired in the absence of delays, this indicates a delay-dependent improvement in learning, relative to shams. Furthermore, asymptotic rates of responding were reduced less by the delay in H-lesioned rats than in controls. The facilitation of learning after a lesion strongly suggests that the lesion has disrupted one process or strategy that normally competes with another process involved in solving the task (e.g. [69,70]). Given the involvement of the hippocampus in contextual conditioning [20-26,28-30,33,34], the most obvious explanation is that the lesions facilitated instrumental conditioning with delayed reinforcement by reducing competition from context-reinforcer associations that normally hinder the formation or expression of response-reinforcer associations.
Certain simple explanations of the present results can be ruled out. The delay-dependent impairment makes an explanation in terms of differences in primary motivation for the food per se unlikely, and the use of a control (inactive) lever means that differences in responding were not attributable to differences in general activity levels, but instead to the contingencies in force on the active lever. Our results also indicated that when there were programmed delays to reinforcement, H-lesioned animals experienced shorter response-reinforcer collection delays, partly because they collected the reinforcer more promptly than shams. This effect probably improved learning in the delay conditions. However, in addition to this effect, there was a further delay-dependent improvement exhibited by H-lesioned rats: even allowing for the shorter response-collection delays that they experienced, their instrumental learning was impaired less by delays than that of sham-operated controls.
The role of the hippocampus in learning an instrumental response with delayed reinforcement has been examined before, though in a very different way. Port et al. [71] found that aspirative lesions of the dorsal hippocampus did not impair appetitive instrumental conditioning with delayed reinforcement. Numerically, lesioned rats were slightly faster to learn (to reach a criterion number of reinforced responses) than shams, but the difference was not significant. This is certainly consistent with the present results, in which a delay-dependent improvement was seen as a result of excitotoxic lesions of the dorsal and ventral hippocampus. However, direct comparison is difficult. Firstly, the lesion extent was different. Secondly, aspirative lesions destroy not just overlying cortex but also fibres of passage (axons of non-hippocampal neurons traversing the hippocampus) [72]. Thirdly, the task used by Port et al. [71] was quite different to that used in the present study: lever presses led to the delivery of responses after a 5-s delay, while responses during the delay were not reinforced; thus, higher rates of responding inevitably reduced the action-outcome contingency. Fourthly, no other delays were tested and no zero-delay condition was used, so any delay-dependent changes would not have been apparent. Fifthly, no control lever was present, so that responding could only be compared across conditions (inferences from a single condition being potentially confounded with general activity levels). Finally, an autoshaping procedure was used to train the rats, and autoshaping is itself known to be impaired by hippocampal lesions [32,73,74], so this may have mitigated against finding an improvement in the lesioned group.
Contrasting the effects of hippocampal lesions on instrumental and Pavlovian conditioning involving delayed reinforcement
Hippocampal lesions have also been found to affect Pavlovian conditioning involving temporally non-contiguous stimuli. Trace (non-contiguous) conditioning, described above, is clearly analogous to the instrumental conditioning task with delayed reinforcement used in the present study, which had response-reinforcer gaps [39]. Hippocampal lesions have been reported to impair trace conditioning to a discrete explicit CS, sparing delay (contiguous) conditioning [40,41,44,75-78]. Trace discrimination learning can also be impaired by hippocampal lesions [79]. Thus, hippocampal lesions appear to impair the acquisition of a Pavlovian CR when there is a gap between CS and US. Indeed, it has been suggested that the hippocampus is particularly important for associating discontiguous events – that is, when there is a temporal, or indeed spatial, discontiguity or gap between two events to be associated [80]. In the current study, however, hippocampal lesions delay-dependently facilitated (relative to shams) the acquisition of an instrumental response when there was a gap between action and outcome. What accounts for these apparently contradictory results?
Firstly, the neural differences between trace and delay conditioning may be not as great as it first seems. The fact that hippocampal lesions have often been shown to impair trace conditioning, but not delay conditioning, may be because trace conditioning is more difficult. If delay conditioning is rendered as hard as trace conditioning by extending the delay ('long-delay conditioning'), the hippocampus is required [44]; similarly, H-lesioned subjects can exhibit trace conditioning if pretrained in a delay conditioning paradigm [44,77], some trace discrimination tasks are intact after hippocampal lesions [81], and the hippocampus is not required for expression of the trace response after learning [82].
Secondly, it may be that a representation of context can help trace (and perhaps long-delay) Pavlovian conditioning, while contextual associations only hinder instrumental responding in the present task, so a lesion that disrupts contextual processing has a differential effect on the two. For example, if the CS during trace conditioning acts as an occasion-setter, signalling that the ensuing context will be followed by a US (an example of feature-positive discrimination), then hippocampal lesions, which can impair both feature-positive discrimination [83] and contextual conditioning, might be expected to impair trace conditioning more than delay conditioning (N.J. Mackintosh, personal communication, 12 October 2004). Yet in instrumental conditioning with delayed reinforcement, as examined here, context-outcome associations can only hinder the learning of response-outcome associations, so hippocampal lesions might be expected to improve learning in a delay-dependent fashion, as was observed. A study by Desmedt et al. [43] supports the hypothesis that trace and delay conditioning endow the context with qualitatively different roles and that the hippocampus contributes differentially to these roles (though they suggest instead that the context is associated directly with the US in 'trace' conditioning, and that the context and discrete CS act as occasion-setters for each other in some way in 'delay' conditioning [33], with hippocampal lesions facilitating this occasion-setting and impairing direct context processing). Thus, although contexts clearly have several associative functions, the contribution to the hippocampus to these processes is by no means clear cut [33,83,84].
Effect of hippocampal lesions on choice involving delayed reinforcement
If lesions of the hippocampus reduce the normal deleterious effects of delays on the ability to associate actions with their outcomes, it might be expected that they would also improve subjects' ability to choose a delayed, large reward in preference to an immediate, small reward. Instead, the opposite pattern of results was observed: postoperatively, H-lesioned rats made impulsive choices, preferring the immediate, small reward. This preference was flexible, responding to changes in the contingencies within the task, and all subjects readily reverted to choosing the large reward when all delays were removed, indicating that they could discriminate the large and small rewards and continued to prefer the large reward when it was not delayed. Upon reintroduction of the delays, however, the preference for the large reward collapsed in the H-lesioned group much more prominently than in shams, indicating that they were less tolerant of delays to reinforcement.
The present results are also somewhat similar to those reported by Rawlins et al. [85], who examined choice between certain and uncertain rewards. Normal rats preferred immediate certain reward to immediate uncertain reward, and also preferred delayed certain reward to immediate uncertain reward; however, rats with hippocampal or medial septal lesions were less tolerant of the delay (or more tolerant of the uncertainty), preferring immediate uncertain reward to delayed certain reward.
Some simple explanations for this effect may readily be ruled out. It is unlikely that lesioned rats' impulsive choice was caused by lower motivation to obtain the food, for two reasons. First, there were no differences in the rate at which they consumed the sucrose pellets used as the reinforcer. Second, the performance of H-lesioned rats was not similar in other respects to that of a subject with lower primary motivation, such as a sated rat [86]; for example, they did not make more omissions than sham-operated controls. It is also unlikely that H-lesioned rats' preference for the small, immediate reward were the consequence of a positional bias away from the lever producing large, delayed reward: when the delays were omitted, all the H-lesioned rats readily and consistently chose the large reinforcer, only to prefer the small reinforcer again when delays were re-introduced. Furthermore, it is difficult to see that impaired contextual conditioning could explain the pattern of results; as discussed above, the absence of context-reinforcer associations should help, rather than hinder, the ability to associate actions with delayed outcomes. Likewise, although context-response associations may influence instrumental responding and contexts may act as occasion-setters (signalling the operation of a particular action-outcome contingency), there is no a priori reason to believe that the context should differentially cue instrumental responding more when the outcome is delayed; neither is there conclusive evidence suggesting that hippocampal lesions impair such a process (see [33]).
One obvious difference between the two experiments is that in Experiment 1, lesions were made before training, whereas in Experiment 2, lesions were made after training. Hence it is possible that hippocampal lesions selectively impair the retrieval of a well-learned instrumental response or action-outcome contingency involving delayed outcomes, while sparing those involving immediate outcomes. However, there are two reasons why this scenario is unlikely. Firstly, animals must be able to perform an action in order for that action to be conditioned instrumentally. Experiment 1 demonstrated that H-lesioned rats were able to acquire instrumental responses for delayed reinforcement at least as well as, if not better than, sham-operated controls. Therefore, it is unlikely that hippocampal lesions selectively impair the performance of instrumental responses for delayed outcomes. Secondly, this idea would not explain why H-lesioned rats showed a reduced preference for the large, delayed reinforcer upon reintroduction of delays, after they had shown a strong postoperative preference for the large reinforcer when delays had been consistently omitted. This required new learning on their part, and since Experiment 1 demonstrated that learning with delayed reinforcement is normal in H-lesioned rats, this would predict self-controlled rather than impulsive choice upon reintroduction of delays.
The task used does not determine whether H-lesioned rats exhibit altered sensitivity to reinforcer magnitude or delay, for either abnormality might produce impulsive choice [87]. Although H-lesioned rats were able to discriminate in absolute terms between the large and the small reinforcer (consistent with previous studies, e.g. [88]), it is possible that they discriminated between the reinforcer magnitudes to a lesser extent than shams. In this scenario, H-lesioned rats might have exhibited impulsive choice simply because the perceived value of the large reinforcer was not subjectively big enough to compensate for the normal effects of the delay. Alternatively, H-lesioned rats may perceive reward magnitudes normally, and exhibit impulsive choice because they are specifically hypersensitive to (intolerant of) the effects of delays to reinforcement. Such evidence as exists suggests that H-lesioned rats perceive reward magnitude normally [88,89]. Experiment 1 indicated that H-lesioned rats are somewhat better than shams at instrumental conditioning with action-outcome delays of >10 s. This suggests that H-lesioned rats associated the action with the delayed outcome normally in Experiment 2, so if they also perceived its magnitude normally, then it is likely that they valued the delayed outcome less.
The present results may also be explained in terms of altered temporal perception. For example, a lesion that increased the speed of an 'internal clock' [90] might affect choice prospectively in this task (i.e. the lesioned subject perceives itself to be at a later time-point in the session than it actually is, hastening the within-session shift towards the immediate lever), or might affect retrospective choice (i.e. the lesioned subject experiences the delay to the large reinforcer as longer than it actually is, causing it to value the reinforcer less than shams). The evidence for the role of the hippocampus in temporal perception is inconclusive: some studies have found that aspirative hippocampal lesions did not affect timing behaviour [91-94], whereas others have suggested that lesions of the hippocampus or fimbria/fornix speed up an internal clock, or reduce the estimation of time periods when a stimulus being timed is interrupted [80,95-98].
Finally, the hippocampus is heavily connected to a number of other structures known to play a role in subjects' relative preference between immediate, small and delayed, large rewards. Lesions of the nucleus accumbens core (AcbC), basolateral amygdala, and orbitofrontal cortex (OFC) have all been found to produce impulsive choice [35,99-101]. OFC lesions appear to alter the processing of reward magnitude as well as delay, and lesions here have produced both impulsive and self-controlled choice under different circumstances [99-102]. AcbC lesions appear to have a more selective effect on the processing of delays, impairing both preference for, and learning with, delayed rewards in the absence of effects on reward magnitude processing [35,103]. Although the hippocampal formation projects heavily to most of the nucleus accumbens (via the subiculum) [104] and H-lesioned rats in the present study showed the impulsive choice known to be exhibited by AcbC-lesioned rats, H-lesioned rats showed the opposite effect to AcbC-lesioned rats in the simple instrumental learning task, being delay-dependently improved rather than impaired relative to shams.
Conclusion
We have demonstrated that excitotoxic lesions of the hippocampus ameliorate the deleterious effects of response-reinforcer delays on instrumental learning. Hippocampal-lesioned rats responded slightly less than controls in the absence of delays, but they became better at learning (relative to shams) as the delays increased, in a delay-dependent fashion. This may have been because the lesion hindered the formation of context-outcome associations, promoting response-outcome association instead. In contrast, lesioned rats exhibited impulsive choice, preferring an immediate, small reward to a delayed, larger reward, even though they preferred the large reward when it was not delayed. Thus, lesioned rats were better at learning with delayed reinforcement but worse at choosing it, suggesting that self-controlled choice and learning with delayed reinforcement tax different psychological processes.
Methods
Overview of experiments
Experiment 1: Effects of hippocampal lesions on acquisition of instrumental responding with delayed reinforcement
Forty-eight rats received excitotoxic lesions of the hippocampus (n = 32) or sham lesions (n = 16). Five died postoperatively. Subject were next trained in a task in which they had continuous access to two identical levers; one lever delivered a single food pellet each time it was pressed, and the other lever had no effect. For some rats, the food pellet was delivered immediately after the lever press (0 s condition; 9 H-lesioned rats and 5 shams). For others, each pellet was delayed by either 10 s (9 H, 6 sham) or 20 s (9 H, 5 sham). Subjects were trained for 14 sessions. They then had their locomotor activity assessed, and finally they were killed and perfused for histology.
Experiment 2: Effects of hippocampal lesions on choice involving delayed reinforcement
Twenty-four naïve rats were first trained to press levers for food and to nosepoke to initiate lever presentations in discrete trials. Subjects were then trained on a choice-of-delayed-reinforcement task (described below) for 19 sessions. After this, they were assigned to matched groups (as described below) to receive lesions of the hippocampus (H, n = 16) or sham lesions (sham, n = 8). Following recovery, they were retested on the basic task for 7 sessions to obtain a baseline measure of performance. After this, 4 sessions were given in which all delays were omitted in alternate sessions (DNDN design; D = delays present, N = no delays). Half of the subjects began this test with the delays present, and half with no delays (counterbalanced across groups). As a deficit was observed during testing (before histological data were available), further behavioural tests were given to elucidate the nature of the deficit. All subjects were given a further 6 sessions with no delays, in an attempt to re-equalize the two groups' performance and ensure that all animals would come to prefer the lever producing the large reinforcer. Delays were then re-introduced for a further 6 sessions. All subjects then underwent a food consumption test and had their locomotor activity assessed; finally, they were killed and perfused for histology.
Subjects and housing conditions
Subjects were male Lister hooded rats (Harlan-Olac UK Ltd) housed in a temperature-controlled room (minimum 22°C) under a 12:12 h reversed light-dark cycle (lights off 07:30 to 19:30). Subjects were approximately 15 weeks old on arrival at the laboratory and were given a minimum of a week to acclimatize, with free access to food, before experiments began. Experiments took place between 09:00 and 21:00, with individual subjects being tested at a consistent time of day. Subjects had free access to water, and were housed either in groups of four (Experiment 1) or in pairs (Experiment 2). During behavioural testing, they were maintained at 85–90% of their free-feeding mass using a restricted feeding regimen. Feeding occurred in the home cages at the end of the experimental day. All procedures were subject to UK Home Office approval (Project Licence 80/1767) under the Animals (Scientific Procedures) Act 1986.
Excitotoxic lesions of the hippocampus
Subjects were anaesthetized with Avertin (2% w/v 2,2,2-tribromoethanol, 1% w/v 2-methylbutan-2-ol, and 8% v/v ethanol in phosphate-buffered saline, sterilized by filtration, 10 ml/kg intraperitoneally) and placed in a Kopf or Stoelting stereotaxic frame (David Kopf Instruments, Tujunga, California, USA; Stoelting Co., Wood Dale, Illinois, USA) fitted with atraumatic ear bars. The skull was exposed and a dental drill was used to remove the bone directly above the injection and cannulation sites. The dura mater was broken with the tip of a hypodermic needle, avoiding damage to underlying venous sinuses. Excitotoxic hippocampal lesions targeted both the dorsal hippocampus and the ventral hippocampus. Lesions were made by injecting 0.09 M N-methyl-D-aspartic acid (NMDA; Sigma, UK) [72] through a glass micropipette (tip diameter 50–100 μm), using the coordinates, volumes, and timings shown in Table 1. The toxin had been dissolved in 0.1 M phosphate buffer (composition 0.07 M Na2HPO4, 0.028 M NaH2PO4 in double-distilled water, sterilized by filtration) and adjusted with NaOH to a final pH of 7.2–7.4. Sham lesions were made in the same manner except that vehicle was infused. At the end of the operation, animals were given 15 ml/kg of sterile 5% w/v glucose, 0.9% w/v sodium chloride intraperitoneally. Lesioned animals were given 0.2 ml of 5 mg/ml diazepam (Roche Products Ltd, UK) i.m. to prevent seizures. They were given two weeks to recover, with free access to food, and were handled regularly. Any instances of postoperative constipation were treated with liquid paraffin orally and rectally. At the end of this period, food restriction commenced or was resumed.
Table 1 Lesion coordinates. Excitotoxic lesions of the entire hippocampus were made by injecting 0.09 M NMDA at the coordinates shown (see Methods). Along the anteroposterior (AP), mediolateral (ML), and dorsoventral (DV) axes, positive coordinates are in the anterior, left, and superior directions respectively. All coordinates are in mm. DV coordinates are measured from the dura above the injection site.
Region within hippocampus Sites per hemisphere AP ML DV Volume injected per site Duration of each infusion Time allowed for diffusion after each infusion
Dorsal 2 -2.8 ± 1.6 -3.3 0.4 μl 4 min 3 min
-4.2 ± 2.6 -3.0 0.4 μl 4 min 3 min
Ventral 4 -4.8 ± 4.8 -6.0 0.2 μl 2 min 3 min
-5.3 ± 4.6 -4.2 0.2 μl 2 min 3 min
-5.3 ± 4.6 -6.0 0.2 μl 2 min 3 min
-5.8 ± 4.6 -4.2 0.2 μl 2 min 3 min
Behavioural apparatus
Behavioural testing was conducted in one of two types of operant chamber of identical configuration (from Med Associates Inc., Georgia, Vermont, USA, or Paul Fray Ltd, Cambridge, UK). Each chamber was fitted with a 2.8 W overhead house light and two retractable levers on either side of an alcove fitted with an infrared photodiode to detect head entry and a 2.8 W lightbulb ('traylight'). Sucrose pellets (45 mg, Rodent Diet Formula P, Noyes, Lancaster, New Hampshire, USA) could be delivered into the alcove. The chambers were enclosed within sound-attenuating boxes fitted with fans to provide air circulation. The apparatus was controlled by software written by RNC in C++ [105] using the Whisker control system [106].
Instrumental conditioning with delayed reinforcement (experiment 1)
A variety of free-operant schedules may be used to assess instrumental acquisition with delayed reinforcement [9]. We used the simplest possible free-operant schedule [103]: each response scheduled a reinforcer after the programmed delay (Figure 1). In such a schedule, if the subject responds during the delay, the experienced response-reinforcer delay will not match the programmed delay (as the second response is temporally close to the first reinforcer). However, this schedule has the advantage that the response-reinforcer contingency is constant (every response does in fact cause the delivery of reinforcement) and the reinforcement rate is not constrained [9]. So that responding could be attributed to the instrumental response-reinforcer contingency, rather than the effects of general activity or reinforcement itself, responding on the active lever was compared to responding on a control lever that had no programmed consequence. Different groups of lesioned and sham-operated subjects were trained using different delays; the delay was consistent for every subject. Delays of 0, 10, and 20 s were used.
Immediately after subjects were placed in the operant chamber, the sessions began. The houselight was illuminated, and remained on for each 30-min session. Two levers were extended into the chamber. All lever responses were first 'debounced' to 10 ms (i.e. if a response occurred within 10 ms of a previous valid response it was attributed to mechanical bounce and ignored). Other than this, all lever presses and nosepokes into the food alcove were recorded. Responding on the left (active) lever caused a single pellet to be delivered following a delay, under a fixed-ratio-1 (FR-1) schedule (Figure 1). To attribute acquisition of a lever-press response to the instrumental contingency, it is also necessary to control for the effects of reinforcer delivery itself [9]; therefore, responding on the active lever was compared to responding on the right (inactive) lever, which had no programmed consequence. To minimize any potential contribution of conditioned reinforcement to the task, no explicit signals were associated with pellet delivery other than the noise of the pellet dispenser apparatus.
Lever and nosepoke training prior to the delayed reinforcement choice task (experiment 2)
Subjects were first trained under an FR-1 schedule (where every lever press leads to the immediate delivery of a pellet) with only one lever present, to a criterion of a total of 50 presses on that lever across 30-min sessions, first for the left lever and then for the right. Subjects were then trained on a simplified version of the full task. The session began with the levers retracted and the operant chamber in darkness. Trials began every 40 s with the illumination of the houselight and the traylight. The subject was required to make a nosepoke response within 10 s, or the trial was aborted and the chamber returned to darkness (scored as an omission). If the subject nosepoked within the time limit, the traylight was extinguished and a single lever was presented (left/right at random). Subjects were required to respond on the lever within 10 s or the lever was retracted and the chamber darkened (scored as an omission). Upon pressing the lever, the houselight was switched off, a single pellet was delivered immediately and the traylight was illuminated until either the pellet was collected or 10 s had elapsed, whereupon the chamber was darkened, and the trial was counted as successful. Rats were trained to a criterion of 60 successful trials in one hour (the maximum possible with trials lasting 40 s being 90).
Choice between small, immediate and large, delayed rewards (experiment 2)
The task was based on Evenden & Ryan's [107] procedure and has been described before [35,86,99]. The session began in darkness with the levers retracted; this was designated the intertrial state. Trials began at 100-s intervals; the format of a single trial is shown in Figure 2. Each trial began with the illumination of the houselight and the traylight. The rat was required to make a nosepoke response, ensuring that it was centrally located at the start of the trial (latency to poke was designated the initiation latency). If the rat did not respond within 10 s of the start of the trial, the operant chamber was reset to the intertrial state until the next trial began and the trial was scored as an omission. If the rat was already nosepoking when the trial began, the next stage followed immediately. Upon a successful nosepoke, the traylight was extinguished and one or both levers were extended. One lever was designated the Delayed lever, the other the Immediate lever (counterbalanced left/right). The latency to choose a lever was recorded. (If the rat did not respond within 10 s of lever presentation, the chamber was reset to the intertrial state until the next trial and the trial was scored as an omission.) When a lever was chosen, both levers were retracted and the houselight was switched off. Choice of the Immediate lever caused the immediate delivery of one pellet; choice of the Delayed lever caused the delivery of 4 pellets following a delay. When reinforcement was delivered, the traylight was switched on. Multiple pellets were delivered 0.5 s apart. If the rat collected the pellets before the next trial began, then the traylight was switched off and time from delivery of the first pellet until a nosepoke occurred was recorded as the collection latency. If the rat did not collect the food within 10 s of its delivery, the operant chamber entered the intertrial state, though collection latencies were still recorded up to the start of the next trial. The chamber was then in the intertrial state and remained so until the next trial. There was no mechanism to remove uneaten pellets, but failure to collect the reward was an extremely rare event.
The delay was varied systematically across the session. A session consisted of 5 blocks, each comprising two trials on which only one lever was presented (one trial for each lever, in randomized order) followed by ten free-choice trials. Preferences were calculated for each block from only those trials on which the subject responded. Delays for each block were 0, 10, 20, 40 and 60 s respectively. As trials began every 100 s, the total session length was 100 minutes; subjects received one session per day.
Preoperatively, subjects were trained on this task for 19 sessions. To allocate subjects into matched groups for surgery, their degree of sensitivity to the effects of delays within each session was assessed for the last three preoperative sessions, by calculating the slope of the linear regression of percentage choice of the large delayed reinforcer against delay for each subject. Rats were ranked by this measure, and rats with equivalent levels of performance were randomized to receive sham or H lesion surgery: the ranked list was divided into ordered triplets, and from each triplet one subject was assigned to the sham group and the other two to the H group, at random.
Locomotor activity in a novel environment
Locomotor activity was measured in wire mesh cages, 25 (W) × 40 (D) × 18 (H) cm, equipped with water bottles and two horizontal infrared photocell beams situated 1 cm from the floor that enabled movements along the long axis of the cage to be registered. Subjects were placed in these cages, which were initially unfamiliar to them, and their activity was recorded for 2 h. All animals were tested in the food-deprived state.
Food consumption tests
Food consumption was assessed using four tests, conducted in subjects' home cages (always with only one rat present) on separate days under conditions of food deprivation. (1) Subjects were given free access to the 45-mg sucrose pellets used as reinforcers (Rodent Diet Formula P, Noyes, Lancaster, NH) for 30 minutes; the amount eaten was recorded. (2) This test was repeated with the chow used as the maintenance diet. (3) The time taken to consume 50 sucrose pellets was recorded. (4) The time taken to consume an equivalent mass of chow (2.25 g) was recorded.
Histology
Rats were deeply anaesthetized with pentobarbitone sodium (200 mg/ml, minimum of 1.5 ml i.p.) and perfused transcardially with 0.01 M phosphate-buffered saline (PBS) followed by 4% paraformaldehyde in PBS. Their brains were removed and postfixed in paraformaldehyde before being dehydrated in 20% sucrose for cryoprotection. The brains were sectioned coronally at 60 μm thickness on a freezing microtome and every third section mounted on chromium potassium sulphate/gelatin-coated glass microscope slides and allowed to dry. Sections were passed through a series of ethanol solutions of descending concentration (3 minutes in each of 100%, 95%, and 70% v/v ethanol in water) and stained for ~5 min with cresyl violet. This stain comprises 0.05% w/v aqueous cresyl violet (Raymond A. Lamb Ltd, Eastbourne, UK), 2 mM acetic acid, and 5 mM formic acid in water. Following staining, sections were rinsed in water and 70% ethanol before being differentiated in 95% ethanol. Finally, they were dehydrated and delipidated in 100% ethanol and Histoclear (National Diagnostics, UK) before being cover-slipped using DePeX mounting medium (BDH, UK) and allowed to dry. The sections were used to verify lesion placement and assess the extent of lesion-induced neuronal loss. Lesions were detectable as the absence of visible neurons (cell bodies of the order of 100 μm in diameter with a characteristic shape), often associated with a degree of tissue collapse (sometimes with consequent ventricular expansion when the lesion was adjacent to a ventricle) and gliosis (visible as the presence of smaller, densely-staining cells).
Data analysis
Data collected by the chamber control programs were imported into a relational database (Microsoft Access 97) for case selection and analysed with SPSS 11. Figures were created with SigmaPlot 2001/v7 and Adobe Illustrator 8. All graphs show group means and error bars are ± 1 standard error of the mean (SEM) unless otherwise stated. Count data (lever presses and locomotor activity counts), for which variance increases with the mean, were subjected to a square-root transformation prior to any analysis [108]. Homogeneity of variance was verified using Levene's test [109]. General linear models are described as dependent variable = A2 × Bcov × (C5 × Dcov × S) where A is a between-subjects factor with two levels, B is a between-subjects covariate, C is a within-subjects factor with five levels, and D is a within-subjects covariate; S denotes subjects in designs involving within-subjects factors [110]. For repeated measures analyses, Mauchly's test of sphericity of the covariance matrix was applied [111] and the degrees of freedom corrected to more conservative values using the Huynh-Feldt epsilon for any terms involving factors in which the sphericity assumption was violated [112].
List of abbreviations used
, Huynh-Feldt epsilon
AcbC, nucleus accumbens core
ANCOVA, analysis of covariance
ANOVA, analysis of variance
CA, cornu ammonis (Ammon's horn)
CR, conditioned response
CS, conditioned stimulus
FI, fixed interval
FR, fixed ratio
H, hippocampus
h, hour
i.m., intramuscular
i.p., intraperitoneal
ISI, interstimulus interval
ITI, intertrial interval
min, minute
NMDA, N-methyl-D-aspartate
OFC, orbitofrontal cortex
P(A | B), probability of A occurring, given that B has occurred
P(A), probability of event A occurring
PBS, phosphate-buffered saline
RR, random ratio
SED, standard error of the difference between means
SEM, standard error of the mean
S-R, stimulus-response
US, unconditioned stimulus
v/v, volume per unit volume
w/v, weight per unit volume
Authors' contributions
RNC conceived and designed the studies, supervised THCC, and wrote the software. THCC participated in the design of the studies, performed the surgery, tested the animals, and processed the histological material. The work contributed to THCC's M.Phil. thesis. Both authors analysed the results and wrote the manuscript.
Acknowledgements
The authors thank Mercedes Arroyo, Jeff Dalley, and Rutsuko Ito for skilled technical assistance, and Nick Mackintosh and Tony Dickinson for helpful discussion. Supported by a Wellcome Trust programme grant (to Trevor W. Robbins, Barry J. Everitt, Angela C. Roberts, and Barbara J. Sahakian); conducted within the UK Medical Research Council (MRC) Behavioural and Clinical Neuroscience Centre, Cambridge. Competing interests: none declared.
==== Refs
Mackintosh NJ Conditioning and associative learning 1983 Oxford, Oxford University Press
Dickinson A Contemporary Animal Learning Theory 1980 Cambridge, Cambridge University Press
Hall G Mackintosh NJ Pavlovian conditioning: laws of association Animal Learning and Cognition 1994 San Diego, Academic Press 15 43
Dickinson A Mackintosh NJ Instrumental conditioning Animal Learning and Cognition 1994 San Diego, Academic Press 45 79
Gallistel CR Mackintosh NJ Space and time Animal Learning and Cognition 1994 San Diego, Academic Press 221 253
Skinner BF The behavior of organisms: an experimental analysis 1938 New York, Appleton
Grice GR The relation of secondary reinforcement to delayed reward in visual discrimination learning Journal of Experimental Psychology 1948 38 1 16
Lattal KA Gleeson S Response acquisition with delayed reinforcement J Psychol Anim Behav Process 1990 16 27 39 10.1037//0097-7403.16.1.27
Dickinson A Watt A Griffiths WJH Free-operant acquisition with delayed reinforcement Quarterly Journal of Experimental Psychology, Section B - Comparative and Physiological Psychology 1992 45 241 258
Harker GS Delay of reward and performance of an instrumental response Journal of Experimental Psychology 1956 51 303 310 13319597
Perin CT A quantitative investigation of the delay-of-reinforcement gradient Journal of Experimental Psychology 1943 32 37 51
Ainslie G Specious reward: a behavioral theory of impulsiveness and impulse control Psychological Bulletin 1975 82 463 496 1099599
Ainslie G Breakdown of Will 2001 Cambridge, UK, Cambridge University Press
Evenden JL Varieties of impulsivity Psychopharmacology (Berl) 1999 146 348 361 10550486
Evenden JL Impulsivity: a discussion of clinical and experimental findings J Psychopharmacol 1999 13 180 192 10475725
Cardinal RN Winstanley CA Robbins TW Everitt BJ Limbic corticostriatal systems and delayed reinforcement Annals of the New York Academy of Sciences 2004 1021 33 50 15251872 10.1196/annals.1308.004
Dickinson A Balleine B Motivational control of goal-directed action Animal Learning & Behavior 1994 22 1 18
Cardinal RN Parkinson JA Hall J Everitt BJ Emotion and motivation: the role of the amygdala, ventral striatum, and prefrontal cortex Neurosci Biobehav Rev 2002 26 321 352 12034134 10.1016/S0149-7634(02)00007-6
Hirsh R The hippocampus and contextual retrieval of information from memory: a theory Behav Biol 1974 12 421 444 4217626 10.1016/S0091-6773(74)92231-7
Selden NR Everitt BJ Jarrard LE Robbins TW Complementary roles for the amygdala and hippocampus in aversive conditioning to explicit and contextual cues Neuroscience 1991 42 335 350 1832750 10.1016/0306-4522(91)90379-3
Kim JJ Fanselow MS Modality-specific retrograde amnesia of fear Science 1992 256 675 677 1585183
Phillips RG LeDoux JE Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning Behavioral Neuroscience 1992 106 274 285 1590953 10.1037//0735-7044.106.2.274
Kim JJ Rison RA Fanselow MS Effects of amygdala, hippocampus, and periaqueductal gray lesions on short- and long-term contextual fear Behav Neurosci 1993 107 1093 1098 8136063 10.1037//0735-7044.107.6.1093
Phillips RG LeDoux JE Lesions of the fornix but not the entorhinal or perirhinal cortex interfere with contextual fear conditioning J Neurosci 1995 15 5308 5315 7623153
Rudy JW Barrientos RM O'Reilly RC Hippocampal formation supports conditioning to memory of a context Behav Neurosci 2002 116 530 538 12148921 10.1037//0735-7044.116.4.530
Anagnostaras SG Maren S Fanselow MS Temporally graded retrograde amnesia of contextual fear after hippocampal damage in rats: within-subjects examination J Neurosci 1999 19 1106 1114 9920672
Chen C Kim JJ Thompson RF Tonegawa S Hippocampal lesions impair contextual fear conditioning in two strains of mice Behav Neurosci 1996 110 1177 1180 8919020 10.1037//0735-7044.110.5.1177
Honey RC Good M Selective hippocampal lesions abolish the contextual specificity of latent inhibition and conditioning Behav Neurosci 1993 107 23 33 8447955 10.1037//0735-7044.107.1.23
Maren S Fanselow MS Electrolytic lesions of the fimbria/fornix, dorsal hippocampus, or entorhinal cortex produce anterograde deficits in contextual fear conditioning in rats Neurobiol Learn Mem 1997 67 142 149 9075242 10.1006/nlme.1996.3752
Jarrard LE On the role of the hippocampus in learning and memory in the rat Behav Neural Biol 1993 60 9 26 8216164 10.1016/0163-1047(93)90664-4
Phillips RG LeDoux JE Lesions of the dorsal hippocampal formation interfere with background but not foreground contextual fear conditioning Learn Mem 1994 1 34 44 10467584
Good M Honey RC Conditioning and contextual retrieval in hippocampal rats Behav Neurosci 1991 105 499 509 1930720 10.1037//0735-7044.105.4.499
Holland PC Bouton ME Hippocampus and context in classical conditioning Current Opinion in Neurobiology 1999 9 195 202 10322181 10.1016/S0959-4388(99)80027-0
Good M Spatial memory and hippocampal function: where are we now? Psicológica 2002 23 109 138 1341123
Cardinal RN Pennicott DR Sugathapala CL Robbins TW Everitt BJ Impulsive choice induced in rats by lesions of the nucleus accumbens core Science 2001 292 2499 2501 11375482 10.1126/science.1060818
Cardinal RN Robbins TW Everitt BJ Heather N and Vuchinich RE Choosing delayed rewards: perspectives from learning theory, neurochemistry, and neuroanatomy Choice, Behavioral Economics and Addiction 2003 Oxford, Elsevier 183 213, 217-218
Good M Honey RC Dissociable effects of selective lesions to hippocampal subsystems on exploratory behavior, contextual learning, and spatial learning Behav Neurosci 1997 111 487 493 9189263 10.1037//0735-7044.111.5.996
McNish KA Gewirtz JC Davis M Evidence of contextual fear after lesions of the hippocampus: a disruption of freezing but not fear-potentiated startle J Neurosci 1997 17 9353 9360 9364080
Mackintosh NJ The Psychology of Animal Learning 1974 London, Academic Press
McEchron MD Bouwmeester H Tseng W Weiss C Disterhoft JF Hippocampectomy disrupts auditory trace fear conditioning and contextual fear conditioning in the rat Hippocampus 1998 8 638 646 9882021 10.1002/(SICI)1098-1063(1998)8:6<638::AID-HIPO6>3.0.CO;2-Q
McEchron MD Disterhoft JF Hippocampal encoding of non-spatial trace conditioning Hippocampus 1999 9 385 396 10495020 10.1002/(SICI)1098-1063(1999)9:4<385::AID-HIPO5>3.0.CO;2-K
Weiss C Knuttinen MG Power JM Patel RI O'Connor MS Disterhoft JF Trace eyeblink conditioning in the freely moving rat: optimizing the conditioning parameters Behav Neurosci 1999 113 1100 1105 10571493 10.1037//0735-7044.113.5.1100
Desmedt A Marighetto A Garcia R Jaffard R The effects of ibotenic hippocampal lesions on discriminative fear conditioning to context in mice: impairment or facilitation depending on the associative value of a phasic explicit cue Eur J Neurosci 2003 17 1953 1963 12752795 10.1046/j.1460-9568.2003.02615.x
Beylin AV Gandhi CC Wood GE Talk AC Matzel LD Shors TJ The role of the hippocampus in trace conditioning: temporal discontinuity or task difficulty? Neurobiol Learn Mem 2001 76 447 461 11726247 10.1006/nlme.2001.4039
Pavlov IP Conditioned Reflexes 1927 Oxford, Oxford University Press
Ellison GD Differential salivary conditioning to traces J Comp Physiol Psychol 1964 57 373 380 14155374
Schneiderman N Interstimulus interval function of the nictitating membrane response of the rabbit under delay versus trace conditioning Journal of Comparative and Physiological Psychology 1966 62 397 402
Kamin LJ Prokasy WF Temporal and intensity characteristics of the conditioned stimulus Classical Conditioning: A Symposium 1965 New York, Appleton-Century-Crofts 118 147
Odling-Smee FJ Background stimuli and the inter-stimulus interval during Pavlovian conditioning Quarterly Journal of Experimental Psychology 1975 27 387 392 1196588
Marlin NA Miller RR Associations to contextual stimuli as a determinant of long-term habituation J Exp Psychol Anim Behav Process 1981 7 313 333 7288367 10.1037//0097-7403.7.4.313
Parkinson JA Robbins TW Everitt BJ Selective excitotoxic lesions of the nucleus accumbens core and shell differentially affect aversive Pavlovian conditioning to discrete and contextual cues Psychobiology 1999 27 256 266
Marlin NA Contextual associations in trace conditioning Animal Learning and Behavior 1980 9 519 523
Kaplan PS Hearst E Bridging temporal gaps between CS and US in autoshaping: insertion of other stimuli before, during, and after CS J Exp Psychol Anim Behav Process 1982 8 187 203 7069378 10.1037//0097-7403.8.2.187
Rescorla RA Effect of a stimulus intervening between CS and US in autoshaping J Exp Psychol Anim Behav Process 1982 8 131 141 7069375 10.1037//0097-7403.8.2.131
Boughner RL Thomas BL Papini MR Effects of nonreinforced preexposure to the context on autoshaping in rats: methodological implications for demonstrations of latent inhibition International Journal of Comparative Psychology 2004 17 168 184
Wagner AR Logan FA Haberlandt K Stimulus selection in animal discrimination learning J Exp Psychol 1968 76 177 186
Dickinson A Charnock DJ Contingency effects with maintained instrumental reinforcement Quarterly Journal of Experimental Psychology Section B-Comparative and Physiological Psychology 1985 37 397 416
Colwill RC Rescorla RA Bower GH Associative structures in instrumental learning The Psychology of Learning and Motivation 1986 20 Orlando, Florida, Academic Press 55 104
Corbit LH Balleine BW The role of the hippocampus in instrumental conditioning J Neurosci 2000 20 4233 4239 10818159
Schmaltz LW Isaacson RL Effect of bilateral hippocampal destruction on the acquisition and extinction of an operant response Physiology and Behavior 1967 2 291 298 10.1016/0031-9384(67)90081-9
Corbit LH Ostlund SB Balleine BW Sensitivity to instrumental contingency degradation is mediated by the entorhinal cortex and its efferents via the dorsal hippocampus J Neurosci 2002 22 10976 10984 12486193
Balleine BW Dickinson A Goal-directed instrumental action: contingency and incentive learning and their cortical substrates Neuropharmacology 1998 37 407 419 9704982 10.1016/S0028-3908(98)00033-1
Corodimas KP LeDoux JE Disruptive effects of posttraining perirhinal cortex lesions on conditioned fear: contributions of contextual cues Behav Neurosci 1995 109 613 619 7576205 10.1037//0735-7044.109.6.1172
Sacchetti B Lorenzini CA Baldi E Tassoni G Bucherelli C Auditory thalamus, dorsal hippocampus, basolateral amygdala, and perirhinal cortex role in the consolidation of conditioned freezing to context and to acoustic conditioned stimulus in the rat J Neurosci 1999 19 9570 9578 10531459
Burwell RD Saddoris MP Bucci DJ Wiig KA Corticohippocampal contributions to spatial and contextual learning J Neurosci 2004 24 3826 3836 15084664 10.1523/JNEUROSCI.0410-04.2004
Bucci DJ Phillips RG Burwell RD Contributions of postrhinal and perirhinal cortex to contextual information processing Behav Neurosci 2000 114 882 894 11085602 10.1037//0735-7044.114.5.882
Maren S Neurotoxic or electrolytic lesions of the ventral subiculum produce deficits in the acquisition and expression of Pavlovian fear conditioning in rats Behav Neurosci 1999 113 283 290 10357453 10.1037//0735-7044.113.2.283
Bannerman DM Yee BK Lemaire M Jarrard L Iversen SD Rawlins JN Good MA Contextual fear conditioning is disrupted by lesions of the subcortical, but not entorhinal, connections to the hippocampus Exp Brain Res 2001 141 304 311 11715074 10.1007/s002210100869
Jaffard R Meunier M Role of the hippocampal formation in learning and memory Hippocampus 1993 3 Spec No 203 217 8287098
Kim JJ Baxter MG Multiple brain-memory systems: the whole does not equal the sum of its parts Trends Neurosci 2001 24 324 330 11356503 10.1016/S0166-2236(00)01818-X
Port R Curtis K Inoue C Briggs J Seybold K Hippocampal damage does not impair instrumental appetitive conditioning with delayed reinforcement Brain Res Bull 1993 30 41 44 8420633 10.1016/0361-9230(93)90037-C
Jarrard LE Meldrum BS Selective excitotoxic pathology in the rat hippocampus Neuropathol Appl Neurobiol 1993 19 381 389 8278020
Richmond J Colombo M Hippocampal lesions, contextual retrieval, and autoshaping in pigeons Brain Res 2002 928 60 68 11844472 10.1016/S0006-8993(01)03355-8
Reilly S Good M Hippocampal lesions and associative learning in the pigeon Behav Neurosci 1989 103 731 742 2765178 10.1037//0735-7044.103.4.731
Weiss C Bouwmeester H Power JM Disterhoft JF Hippocampal lesions prevent trace eyeblink conditioning in the freely moving rat Behav Brain Res 1999 99 123 132 10512579 10.1016/S0166-4328(98)00096-5
Christian KM Thompson RF Neural substrates of eyeblink conditioning: acquisition and retention Learn Mem 2003 10 427 455 14657256 10.1101/lm.59603
Shors TJ Memory traces of trace memories: neurogenesis, synaptogenesis and awareness Trends Neurosci 2004 27 250 256 15111006 10.1016/j.tins.2004.03.007
Solomon PR Vander Schaaf ER Thompson RF Weisz DJ Hippocampus and trace conditioning of the rabbit's classically conditioned nictitating membrane response Behav Neurosci 1986 100 729 744 3778636 10.1037//0735-7044.100.5.729
Schmitt WB Arianpour R Deacon RM Seeburg PH Sprengel R Rawlins JN Bannerman DM The role of hippocampal glutamate receptor-A-dependent synaptic plasticity in conditional learning: the importance of spatiotemporal discontiguity J Neurosci 2004 24 7277 7282 15317854 10.1523/JNEUROSCI.1093-04.2004
Wallenstein GV Eichenbaum H Hasselmo ME The hippocampus as an associator of discontiguous events Trends Neurosci 1998 21 317 323 9720595 10.1016/S0166-2236(97)01220-4
Savage LM Buzzetti RA Ramirez DR The effects of hippocampal lesions on learning, memory, and reward expectancies Neurobiol Learn Mem 2004 82 109 119 15341796 10.1016/j.nlm.2004.05.002
Takehara K Kawahara S Takatsuki K Kirino Y Time-limited role of the hippocampus in the memory for trace eyeblink conditioning in mice Brain Res 2002 951 183 190 12270496 10.1016/S0006-8993(02)03159-1
Holland PC Lamoureux JA Han JS Gallagher M Hippocampal lesions interfere with Pavlovian negative occasion setting Hippocampus 1999 9 143 157 10226775 10.1002/(SICI)1098-1063(1999)9:2<143::AID-HIPO6>3.0.CO;2-Z
Heldt SA Coover GD Falls WA Posttraining but not pretraining lesions of the hippocampus interfere with feature-negative discrimination of fear-potentiated startle Hippocampus 2002 12 774 786 12542229 10.1002/hipo.10033
Rawlins JN Feldon J Butt S The effects of delaying reward on choice preference in rats with hippocampal or selective septal lesions Behav Brain Res 1985 15 191 203 4005029 10.1016/0166-4328(85)90174-3
Cardinal RN Robbins TW Everitt BJ The effects of d-amphetamine, chlordiazepoxide, alpha-flupenthixol and behavioural manipulations on choice of signalled and unsignalled delayed reinforcement in rats Psychopharmacology 2000 152 362 375 11140328 10.1007/s002130000536
Ho MY Mobini S Chiang TJ Bradshaw CM Szabadi E Theory and method in the quantitative analysis of "impulsive choice" behaviour: implications for psychopharmacology Psychopharmacology (Berl) 1999 146 362 372 10550487
Kesner RP Williams JM Memory for magnitude of reinforcement: dissociation between the amygdala and hippocampus Neurobiol Learn Mem 1995 64 237 244 8564377 10.1006/nlme.1995.0006
Gilbert PE Kesner RP The amygdala but not the hippocampus is involved in pattern separation based on reward value Neurobiol Learn Mem 2002 77 338 353 11991762 10.1006/nlme.2001.4033
Gibbon J Malapani C Dale CL Gallistel C Toward a neurobiology of temporal cognition: advances and challenges Current Opinion in Neurobiology 1997 7 170 184 9142762 10.1016/S0959-4388(97)80005-0
Dietrich A Allen JD Bunnell BN Is the hippocampus involved in temporal discrimination and the memory of short intervals? Int J Neurosci 1997 90 255 269 9352431
Dietrich A Allen JD Functional dissociation of the prefrontal cortex and the hippocampus in timing behavior Behav Neurosci 1998 112 1043 1047 9829782 10.1037//0735-7044.112.5.1043
Rawlins JN Winocur G Gray JA The hippocampus, collateral behavior, and timing Behav Neurosci 1983 97 857 872 6651961 10.1037//0735-7044.97.6.857
Port RL Romano AG Steinmetz JE Mikhail AA Patterson MM Retention and acquisition of classical trace conditioned responses by rabbits with hippocampal lesions Behav Neurosci 1986 100 745 752 3778637 10.1037//0735-7044.100.5.745
Meck WH Church RM Olton DS Hippocampus, time, and memory Behav Neurosci 1984 98 3 22 6696797 10.1037//0735-7044.98.1.3
Olton DS Meck WH Church RM Separation of hippocampal and amygdaloid involvement in temporal memory dysfunctions Brain Res 1987 404 180 188 3567565 10.1016/0006-8993(87)91369-2
Meck WH Hippocampal function is required for feedback control of an internal clock's criterion Behav Neurosci 1988 102 54 60 3355658 10.1037//0735-7044.102.1.54
Hata T Okaichi H [Effects of fimbria-fornix lesion on the temporal discrimination revealed by peak interval procedure in rats] Shinrigaku Kenkyu 1998 69 304 309 10067236
Winstanley CA Theobald DE Cardinal RN Robbins TW Contrasting roles of basolateral amygdala and orbitofrontal cortex in impulsive choice J Neurosci 2004 24 4718 4722 15152031 10.1523/JNEUROSCI.5606-03.2004
Mobini S Body S Ho MY Bradshaw CM Szabadi E Deakin JF Anderson IM Effects of lesions of the orbitofrontal cortex on sensitivity to delayed and probabilistic reinforcement Psychopharmacology 2002 160 290 298 11889498 10.1007/s00213-001-0983-0
Kheramin S Body S Mobini S Ho MY Velazquez-Martinez DN Bradshaw CM Szabadi E Deakin JF Anderson IM Effects of quinolinic acid-induced lesions of the orbital prefrontal cortex on inter-temporal choice: a quantitative analysis Psychopharmacology 2002 165 9 17 12474113 10.1007/s00213-002-1228-6
Kheramin S Body S Ho M Velazquez-Martinez DN Bradshaw CM Szabadi E Deakin JF Anderson IM Role of the orbital prefrontal cortex in choice between delayed and uncertain reinforcers: a quantitative analysis Behav Processes 2003 64 239 250 14580695 10.1016/S0376-6357(03)00142-6
Cardinal RN Cheung THC Nucleus accumbens core lesions retard instrumental learning and performance with delayed reinforcement in the rat BMC Neuroscience 2005 6 9 15691387 10.1186/1471-2202-6-9
Brog JS Salyapongse A Deutch AY Zahm DS The patterns of afferent innervation of the core and shell in the "accumbens" part of the rat ventral striatum: immunohistochemical detection of retrogradely transported fluoro-gold Journal of Comparative Neurology 1993 338 255 278 8308171 10.1002/cne.903380209
Stroustrup B The C++ Programming Language 1986 Reading, Massachusetts, Addison-Wesley
Cardinal RN Aitken MRF Whisker (version 2) [ www.whiskercontrol.com ] 2001 Cambridge, UK, Cambridge University Technical Services Ltd
Evenden JL Ryan CN The pharmacology of impulsive behaviour in rats: the effects of drugs on response choice with varying delays of reinforcement Psychopharmacology 1996 128 161 170 8956377 10.1007/s002130050121
Howell DC Statistical Methods for Psychology 1997 Fourth Belmont, California, Wadsworth
Levene H Oklin I Robust tests for the equality of variance Contributions to probability and statistics 1960 Palo Alto, California, Stanford University Press
Keppel G Design and analysis: a researcher's handbook 1982 Second London, Englewood Cliffs: Prentice-Hall
Mauchly JW Significance test for sphericity of a normal n-variate distribution Annals of Mathematical Statistics 1940 11 204 209
Huynh H Feldt LS Conditions under which mean square ratios in repeated measures designs have exact F-distributions Journal of the American Statistical Association 1970 65 1582 1589
Amaral DG Witter MP Paxinos G Hippocampal formation The Rat Nervous System 1995 Second London, Academic Press 443 493
Paxinos G Watson C The Rat Brain in Stereotaxic Coordinates 1998 Fourth London, Academic Press
| 15892889 | PMC1156904 | CC BY | 2021-01-04 16:39:09 | no | BMC Neurosci. 2005 May 13; 6:36 | utf-8 | BMC Neurosci | 2,005 | 10.1186/1471-2202-6-36 | oa_comm |
==== Front
BMC OphthalmolBMC Ophthalmology1471-2415BioMed Central London 1471-2415-5-111591671210.1186/1471-2415-5-11Case ReportPrimary graft failure associated with epithelial downgrowth: a case report Aldave Anthony J [email protected] David A [email protected] Bruno [email protected] Brooks [email protected] Richard L [email protected] From the Department of Ophthalmology, The University of California, San Francisco/ Francis I. Proctor Foundation, San Francisco, California, USA2 The Jules Stein Eye Institute, The University of California, Los Angeles, Los Angeles, California, USA2005 25 5 2005 5 11 11 7 1 2005 25 5 2005 Copyright © 2005 Aldave et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Epithelial downgrowth is a rare complication of ocular surgery. While the features of epithelial downgrowth following corneal transplantation are well described, its association with primary graft failure has only been reported once previously. We report a case of primary corneal graft failure (PGF) associated with retrocorneal epithelial cell ingrowth.
Case presentation
A 59 year-old male underwent an uncomplicated penetrating keratoplasty for Fuchs' corneal dystrophy. The patient developed PGF, and a second transplant was performed 5 weeks after the initial surgery. The initial host corneal button and the failed corneal graft were examined with light microscopy. Histopathologic examination of the excised corneal button demonstrated multilaminar epithelial cells on the posterior corneal surface and absence of endothelial cells. DNA extraction and polymerase chain reaction (PCR) for herpes simplex virus (HSV) DNA was performed on the failed corneal graft. Polymerase chain reaction performed on the failed corneal graft was negative for HSV DNA, which has been implicated in selected cases of PGF. Three years following repeat penetrating keratoplasty, there was no evidence of recurrent epithelial ingrowth.
Conclusion
This is only the second report of PGF associated with epithelialization of the posterior corneal button, which most likely developed subsequent to, instead of causing, the diffuse endothelial cell loss and primary graft failure.
==== Body
Background
Epithelial downgrowth is an unusual but recognized complication of ocular surgery [1]. Characteristic features of epithelial downgrowth following corneal transplantation are well described [2], although the association of epithelial downgrowth and PGF has only been made once previously [3]. We present the second case of PGF in which the excised corneal button demonstrated epithelialization of the posterior cornea, most likely secondary to postoperative epithelial cell migration across a denuded Descemet's membrane.
Case presentation
A 59 year-old male with Fuchs' endothelial dystrophy presented with visually limiting pseudophakic corneal edema in his right eye with a best-corrected visual acuity of 20/50. Cataract extraction had been performed two years previously in the right eye through a 3.2 mm limbal incision. An uncomplicated penetrating keratoplasty was performed, with an 8.25 mm donor cornea secured into an 8 mm recipient opening with 16 interrupted 10-0 nylon sutures. On the first postoperative day, the graft demonstrated diffuse 3+–4+ stromal edema and a nearly complete epithelial defect. The anterior chamber was deep, and no aqueous leakage was noted from either the wound or suture tracks. The epithelial defect resolved in the first postoperative week, although 1+–2+ stromal edema persisted. As the graft remained edematous 4 weeks later, the patient was diagnosed with PGF, and a repeat penetrating keratoplasty was performed 5 weeks after the primary graft. Three years later, the second graft remains clear, with a central pachymetry of 495 μm, and corrected visual acuity of 20/25+.
The host corneal button removed at the initial penetrating keratoplasty and the failed donor corneal button were formalin-fixed and analyzed with light microscopy after staining with hematoxylin and eosin, periodic acid-Schiff, and Masson trichrome stains. Histopathologic examination of the host corneal button revealed stromal edema, thickening and excrescences of Descemet's membrane, and a decreased number of endothelial cells, consistent with Fuchs' corneal dystrophy. Histopathologic examination of the failed corneal graft revealed a markedly thickened stroma and near complete absence of endothelial cells. Both sides of the excised donor corneal button were covered with epithelial cells, consistent with epithelial downgrowth. (Figure 1) The corneal sections containing suture tracks were stained with the periodic acid Schiff stain to demonstrate Descemet's membrane. (Figure 2) The suture tracks passed through the true epithelial layer, Bowman's layer, and extended into the stroma, confirming that the button had not been inverted when secured into the recipient opening. Additionally, serial sectioning of the host corneal button failed to demonstrate the presence of epithelial cells on the posterior corneal surface.
To investigate the possibility of HSV infection of the donor cornea, which has been associated with one-third of cases of PGF [4,5], PCR analysis of the donor corneal button for HSV DNA was performed. Positive and negative controls confirmed adequate extraction of purified DNA from the failed donor corneal button, as well as the sensitivity and specificity of the HSV primers used. PCR analysis for HSV DNA was negative in the donor corneal button.
A complete review of the first donor's medical and eye bank records was performed. The tissue was obtained through the University of California, San Francisco tissue bank from a 56 year-old female donor who died from a myocardial infarction and cardiac tamponade. Routinely performed serologic screening tests and biomicroscopic examination of the donor button were unremarkable. The endothelial cell count was 2852 cells/mm2, and death to preservation was 12 hours, with a 51-hour preservation time in Optisol-GS.
The paired donor cornea was transplanted into another recipient several hours after this patient's original transplant. This cornea also appeared unremarkable on biomicroscopic examination, remained in Optisol-GS for 53 hours prior to transplantation, and had a cell count of 2841 cells/mm2. Five months later, after a similar 16 interrupted suture technique, the graft was clear and compact with central corneal pachymetry of 538 μm.
Conclusion
Epithelial downgrowth is an unusual complication after penetrating keratoplasty [1-3,6-17]. (Table 1) The incidence in patients undergoing aphakic penetrating keratoplasty has been estimated at 0.25% [3], and risk factors include poor wound apposition, full-thickness suture passes, and globe hypotony [1]. Characteristic signs, such as a migratory endothelial line, elevated intraocular pressure and effacement of the normal iris architecture, have been reported between 14 days and 13 years after penetrating keratoplasty [2,3].
The histopathologic finding of epithelial downgrowth on the posterior corneal surface after a diagnosis of PGF is rare. Sugar and colleagues reported a patient with PGF 1 week after penetrating keratoplasty [3]. Histopathologic examination of the failed corneal graft excised 2 weeks after the initial transplant demonstrated stratified squamous epithelium covering Descemet's membrane. Though the patient had undergone a previous uncomplicated intracapsular cataract extraction, there was no evidence of epithelial downgrowth on the patient's native cornea. The authors speculated that epithelial migration over Descemet's membrane had occurred during storage.
Identifying the origin of the epithelial cells (donor or host), as well as the timing of epithelial cell access to the posterior cornea, in our case required further investigation. Inadvertent inversion of the donor corneal button during penetrating keratoplasty was excluded by identifying suture tracks extending into the stroma from the true anterior corneal surface.
If PGF had occurred secondary to epithelial cell migration over Descemet's membrane, the migration would have had to begin while the donor cornea was in storage medium. Since epithelial migration is inhibited by an intact endothelium, epithelial extension would have required diffuse endothelial cell loss [18]. In this case, there was no evidence of endothelial loss during harvesting. Additionally, there was no evidence of HSV infection of the donor cornea.
Storage at 4°C, as in the case we report, is designed to inhibit microbial cellular metabolism, and we are unaware of any evidence that human corneal epithelial cells retain sufficient metabolic activity at 4°C to allow proliferation and migration. Furthermore, the donor button in this case was preserved for only 51 hours prior to transplantation, an insufficient amount of time for the epithelial cells to have migrated over the scleral rim and onto the posterior corneal surface [19].
If the epithelial cells on the posterior corneal surface were not of donor origin, then they must be host-derived. The possibility that these cells gained entry during the prior cataract surgery was dismissed by both the absence of epithelial cells on the original host button and the lack of evidence of epithelial downgrowth prior to penetrating keratoplasty.
The most likely explanation is that epithelial cells from the peripheral host cornea rim gained access to the posterior corneal surface after the graft was secured into the recipient opening. The complete absence of endothelial cells on the excised donor button indicates that the PGF occurred secondary to a diffuse endothelial cell injury while the corneoscleral button was in the preservative medium or at the time of penetrating keratoplasty. Interestingly, none of the identified risk factors for epithelial downgrowth, such as wound dehiscence, hypotony or a shallow anterior chamber, were observed.
Methods
DNA extraction
A 20 milligram formalin-fixed, paraffin-embedded tissue section from the failed corneal donor button was deparaffinized using xylene, and then treated with alcohol to remove the residual xylene. Protein digestion was performed with 40 μl of 20mg/ml proteinase K at 55°C for 12 hours. A silica-gel-membrane technology (Qiagen, Valencia, CA) was used to isolate 50 μl of purified DNA product.
Polymerase chain reaction
Polymerase chain reaction (PCR) for herpes simplex virus (HSV) I DNA was performed on the failed donor button, with a number of simultaneously performed positive and negative controls. Positive controls consisted of varied amounts of HSV I viral DNA (1, 10, 100 and 1000 copies), allowing quantification of the assay's sensitivity to HSV DNA present in the donor corneal button. DNA extracted from the cornea of a patient with keratoconus served as a negative control.
A previously described primer set was used for amplification of a 92bp HSV sequence: forward, 5'- CATCACCGACCCGGAGAGGGAC and reverse, 5'- GGGCCAGGCGCTTGTTGGTGTA1. These are regions of identity in the genomes of HSV types 1 and 2, and do not discriminate between these virus types [20-22]. PCR amplification was performed on a 100 μl final volume containing: 1 μl of 19.7mM of each primer (Biomolecular Research Center, University of California, San Francisco), 5 μl of the extracted DNA product, 10 μl of a 10X PCR Buffer (Sigma Chemical, Saint Louis, MO), 2 μl of 10mM deoxy nucleotide triphosphate (dNTP) (PE Biosystems, Foster City, CA), 10 μl of 25 mM of magnesium chloride, 2.5 μl of RedTaq DNA polymerase (Sigma Chemical), and 68.5 μl of water. Thermal cycling was performed in a GeneAmp® PCR System 9700 (Applied Biosystems, Foster City, CA) with the following program: initial denaturation for 2 min at 94°C, followed by 41 cycles of 94°C for 30 seconds, 60°C for 30 seconds, and 72°C for 30 seconds, followed by a final extension for 2 minutes at 72°C. The PCR amplification products (10 μl) were separated on a 4% polyacrylamide gel and visualized by ethidium bromide staining.
To confirm the presence of purified DNA extracted from the failed donor corneal button, previously published primers were used to perform PCR amplification of exon 12 of the TGFBI gene from the purified DNA product [23]. A mixture containing water in place of purified DNA served as a negative control.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Anthony J. Aldave, M.D. Assisted with care of the patient, coordinated performance of PCR for HSV DNA, wrote majority of the manuscript.
David A. Hollander, M.D., M.B.A. Assisted with manuscript preparation and revisions.
Bruno Branco, M.D. Performed PCR for HSV DNA and wrote description of methods for Material and Methods section as well as relevant portion of Results section.
Brooks Crawford, M.D. Performed histopathologic analysis and wrote description of methods for Material and Methods section as well as relevant portion of Results section.
Richard L. Abbott, M.D. Primary ophthalmologist for the patient. Performed initial and subsequent corneal transplants.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
Supported by the AOS Knapp Fellowship Fund, Cleveland, OH (Dr. Aldave); Heed Foundation Fellowship Fund, Cleveland, OH (Dr. Hollander); a grant from CNPq-Brazil (Dr. Branco); and an unrestricted grant from Research to Prevent Blindness, New York, NY.
Written consent was obtained from the patient for publication of the study.
Figures and Tables
Figure 1 Histopathologic section of excised corneal button demonstrating a markedly thickened, edematous stroma and the presence of multilayered epithelial cells on the anterior and posterior surfaces (hematoxylin and eosin stain, original magnification × 50).
Figure 2 Histopathologic section of excised corneal button demonstrating anterior portion of suture track seen as focal tissue separation on superior side of corneal button. The identification of Descemet's membrane as a dark band on the inferior side of the corneal button indicates that the donor corneal button had been correctly oriented when secured into the recipient opening (periodic acid Schiff stain, original magnification × 50).
Table 1 Summary of Reported Cases of Epithelial Downgrowth (ED) after Penetrating Keratoplasty (PK)
Author Previous ocular surgery PK Indication Time from PK to ED diagnosis Clinical features of ED Diagnosis of ED Final Anatomic/ Visual Status
Chiou et al [1] 3 PKs, 2 Trabs, 1 CPC Graft Failure 13 months Concentric peripheral retrocorneal membrane Confocal microscopy BSCVA 20/400
Feder & Krachmer [2] ICCE ABK 6 months Iritis, glaucoma, posterior corneal line Paracentesis, laser Graft decompensation, glaucoma
ICCE PBK 7 months Iritis, glaucoma Paracentesis, laser End-stage glaucoma
ICCE Fuchs' dystrophy 5 months Posterior corneal line Paracentesis, laser Evisceration
ECCE KCN 13 years Iritis Paracentesis, laser Phthisis bulbi
Mazow [7] None Endothelial dystrophy 5 months Glaucoma Histopathologic Enucleation
Karabatsas [8] 2 PKs PBK 8 months Posterior corneal line, glaucoma Clinical features Not available
Sugar [3] PK, ICCE Graft failure 7 months Retrocorneal membrane, pre-iris membrane Histopathologic Recurrent retrocorneal membrane; CF VA*
ICCE ABK;Fuchs' dystrophy 2 weeks Descemet's deposits Histopathologic BSCVA 20/200†
LK, ICCE Opacification of lamellar graft 16 months Pre-iris membrane Histopathologic Scarred regraft; VA 10/400
PK, ICCE Graft failure 6 months Retrocorneal membrane Clinical features Vascularized, edeamatous regraft, VA HM
Leibowitz [9] 2 PKs Graft failure 1 month Retrocorneal membrane; anterior chamber epithelial cyst Histopathologic Vascularized, opacified regraft
Groh & Naumann [10] None KCN 13 months Anterior chamber epithelial cyst Clinical features VA 14/20
None Lattice corneal dystrophy 47 months Anterior chamber epithelial cyst Clinical features VA 10/20
Wearne [11] ‡ None Scarring 2° to HSV keratitis 10 years Anterior chamber epithelial cyst Clinical features VA 6/20
Yamaguchi et al. [12] PK, ICCE Graft failure 12 months Retrocorneal membrane Clinical features and histopathologic Opacified regraft
Sidrys [13] ¶ ICCE PBK 3 weeks§ Retrocorneal membrane Clinical features and histopathologic VA 6/60Θ
Weiner [14]# Unknown Various Unknown Retrocorneal membrane, glaucoma Clinical features and histopathologic Unknown
Kurz & D'Amico [16] None Corneal scar 10 months None Histopathologic Unknown
None Corneal scar 8 weeks None Histopathologic Enucleation**
Bennett & D'Amico [15] None KCN 3 years Iris epithelial inclusion cyst Clinical features and histopathologic S/p sector iridectomy without recurrence
Claoue et al [17] None KCN 21 months Iris epithelial inclusion cyst Clinical features and histopathologic S/p sector iridectomy without recurrence
Daneshvar et al. [6] CE ABK 1 year Glaucoma, retrocorneal membrane Laser, ultrasound biomicroscopy No recurrence 16 months after repeat PK
BSCVA = best spectacle corrected visual acuity; ABK = aphakic bullous keratopathy; CF = counting fingers; CPC = cyclophotocoagulation; ECCE = extracapsular cataract extraction; HM = hand motions; HSV = herpes simplex virus; ICCE = intracapsular cataract extraction; KCN = keratoconus; LK = lamellar keratectomy; PBK = pseudophakic bullous keratopathy; Trab = trabeculectomy; * limited by ARMD; † h/o CME and RD s/p SB; ‡same patient reported again in; 10§ epithelial downgrowth through a cataract wound fistula; Θ limited by CME; ¶ same patient reported again in; 12 # 13 patients reported; ** performed after traumatic wound dehiscence and total retinal detachment
==== Refs
Chiou AG Kaufman SC Kaz K Beuerman RW Kaufman HE Characterization of epithelial downgrowth by confocal microscopy J Cataract Refract Surg 1999 25 1172 4 10445210 10.1016/S0886-3350(99)00133-9
Feder RS Krachmer JH The diagnosis of epithelial downgrowth after keratoplasty Am J Ophthalmol 1985 99 697 703 3893142
Sugar A Meyer RF Hood CI Epithelial downgrowth following penetrating keratoplasty in the aphake Arch Ophthalmol 1977 95 464 7 320971
Cockerham GC Bijwaard K Sheng ZM Hidayat AA Font RL McLean IW Primary graft failure : a clinicopathologic and molecular analysis Ophthalmology 2000 107 2083 90 discussion 2090-1 11054337 10.1016/S0161-6420(00)00361-4
Cockerham GC Krafft AE McLean IW Herpes simplex virus in primary graft failure Arch Ophthalmol 1997 115 586 9 9152124
Daneshvar H Brownstein S Mintsioulis G Chialant D Punja K Damji KF Epithelial ingrowth following penetrating keratoplasty: a clinical, ultrasound biomicroscopic and histopathological correlation Can J Ophthalmol 2000 35 222 4 10900520
Mazow ML Stephens RW An unusual complication after keratoplasty Surv Ophthalmol 1966 11 205 8
Karabatsas CH Hoh HB Easty DL Epithelial downgrowth following penetrating keratoplasty with a running adjustable suture J Cataract Refract Surg 1996 22 1242 4 8972378
Leibowitz HM Elliott JH Boruchoff SA Epithelization of the anterior chamber following penetrating keratoplasty Arch Ophthalmol 1967 78 613 7 4860776
Groh MJ Naumann GO Cystic epithelial growth after penetrating keratoplasty: successful curative treatment by block excision Br J Ophthalmol 2001 85 240 11225574 10.1136/bjo.85.2.238b
Wearne MJ Buckley RJ Cree IA Naumann GO Cystic epithelial ingrowth as a late complication of penetrating keratoplasty Arch Ophthalmol 1999 117 1444 5 10532467
Yamaguchi T Polack FM Valenti J Electron microscopic study of epithelial downgrowth after penetrating keratoplasty Br J Ophthalmol 1981 65 374 82 7018561
Sidrys LA Demong T Epithelial downgrowth after penetrating keratoplasty Can J Ophthalmol 1982 17 29 31 7044502
Weiner MJ Trentacoste J Pon DM Albert DM Epithelial downgrowth: a 30-year clinicopathological review Br J Ophthalmol 1989 73 6 11 2920156
Bennett T D'Amico RA Epithelial inclusion cyst of iris after keratoplasty Am J Ophthalmol 1974 77 87 9 4363106
Kurz GH D'Amico RA Histopathology of corneal graft failures Am J Ophthalmol 1968 66 184 99 4874267
Claoue C Lewkowicz-Moss S Easty D Epithelial cyst in the anterior chamber after penetrating keratoplasty: a rare complication Br J Ophthalmol 1988 72 36 40 3277665
Cameron JD Flaxman BA Yanoff M In vitro studies of corneal wound healing: epithelial-endothelial interactions Invest Ophthalmol 1974 13 575 9 4841864
Zagorski Z Holbach L Rummelt C Gossler B Naumann GO Proliferation of corneal epithelial and endothelial cells in the trabecular region of human donor corneas in organ culture Ophthalmic Res 1990 22 51 6 2342779
Cao M Xiao X Egbert B Darragh TM Yen TS Rapid detection of cutaneous herpes simplex virus infection with the polymerase chain reaction J Invest Dermatol 1989 92 391 2 2918243 10.1111/1523-1747.ep12277232
Cunningham ET JrShort GA Irvine AR Duker JS Margolis TP Acquired immunodeficiency syndrome – associated herpes simplex virus retinitis. Clinical description and use of a polymerase chain reaction – based assay as a diagnostic tool Arch Ophthalmol 1996 114 834 40 8660167
Tsurumi T Maeno K Nishiyama Y Nucleotide sequence of the DNA polymerase gene of herpes simplex virus type 2 and comparison with the type 1 counterpart Gene 1987 52 129 37 3038677 10.1016/0378-1119(87)90039-4
Munier FL Korvatska E Djemai A Le Paslier D Zografos L Pescia G Schorderet DF Kerato-epithelin mutations in four 5q31-linked corneal dystrophies Nat Genet 1997 15 247 51 9054935 10.1038/ng0397-247
| 15916712 | PMC1156905 | CC BY | 2021-01-04 16:03:50 | no | BMC Ophthalmol. 2005 May 25; 5:11 | utf-8 | BMC Ophthalmol | 2,005 | 10.1186/1471-2415-5-11 | oa_comm |
==== Front
BMC Pregnancy ChildbirthBMC Pregnancy and Childbirth1471-2393BioMed Central London 1471-2393-5-111590720710.1186/1471-2393-5-11Research ArticleSSRI'S and other antidepressant use during pregnancy and potential neonatal adverse effects: Impact of a public health advisory and subsequent reports in the news media Einarson A [email protected] AK [email protected] R [email protected] E [email protected] G [email protected] The Motherisk Program, The Hospital for Sick Children, Toronto, Canada2 Department of Pharmacoepidemiology & Pharmacotherapy, Utrecht Institute for Pharmaceutical Sciences, Utrecht, The Netherlands3 The Leslie Dan Faculty of Pharmacy, University of Toronto, Toronto, Canada2005 20 5 2005 5 11 11 11 3 2005 20 5 2005 Copyright © 2005 A et al; licensee BioMed Central Ltd.2005A et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
On Aug 9th 2004 Health Canada released an advisory, which followed a similar one from the FDA regarding the use of SSRI's and other antidepressants during pregnancy and potential adverse effects on newborns. In neither advisory was it stated that women should discontinue their antidepressant. In the seven days following the release of this advisory, The Motherisk Program received 49 calls from anxious women in response to the media reporting of this information.
Objective
To examine the impact of the advisory and subsequent reporting in the media, on the decision-making of women, currently taking an antidepressant, who called The Motherisk Program after becoming aware of this information.
Methods
We attempted to follow up all the women who had called us who were alarmed by this advisory and asked them to complete a specially designed questionnaire.
Results
We were able to complete 43/49 (88%) follow-ups of the women who contacted us. All of the callers reported that the messages in the media caused a great deal of anxiety. Seven misunderstood the advisory, ie their children were more than 1 year old, five had discontinued their antidepressant (3 abruptly (2 later restarted after speaking with Motherisk counsellors)and 2 with some form of tapering off) and(6) were considering discontinuation, but decided to continue following reassurance from Motherisk
Conclusion
Medical information regarding fetal and infant safety, disseminated in the public domain, should be transferred in a way that does not influence a pregnant woman to make decisions that may not be in the best interest of hers or her child's health.
==== Body
Background
The media plays a significant role in the dissemination of medical information to the general public, through newspapers, television, magazines and more recently through the internet. While it is important to get these messages to the public, at times it appears that the media may have a different agenda, such as selling newspapers or widening their TV audiences. They seem to prefer alarmist headlines and display a preference for stories that catch attention, rather than to inform the public of new scientific data that has both positive and negative results. For example, stories describing safe products are often unreported, or reported in small print, whereas unsafe products usually make headlines. [1] Since the thalidomide scare in the sixties, this has been especially true of stories concerning the safety/risk of exposures during pregnancy.
A recent example in another field, can be found in the study of the use of hormone replacement therapy (HRT) following the release of The Women's Health Initiative (WHI) which was widely publicized in the lay press. [2] A recent survey, completed by more than 1000 women, that examined their opinions and understanding from the lay press of this study, found that women dramatically overestimated the risks of HRT.[3] Another study found that information on the benefits of HRT was mainly provided by health care professionals, whereas information on the risks was provided mainly by the mass media. [4]
The Motherisk Program, at the Hospital for Sick Children in Toronto, is a counseling service that provides pregnant, breastfeeding women, and health care providers with evidence-based information on the safety and risks of exposures to prescription and over-the-counter (OTC) medications, natural health products, chemicals, radiation, and infectious agents. Women and their health care professionals often call us when alarming stories regarding pregnancy exposures are reported in the media. Consequently, we receive a dramatic increase in the number of calls to our service following a report in the media that involves a study that has produced results that are distressing to a pregnant woman.
On June 9th 2004 the FDA(Food and Drug Administration) instructed the manufacturers of antidepressants to issue warnings about perinatal complications associated with their products. FDA: Medwatch Drug Alert. June 03/2004). On Aug 9th 2004, Health Canada, followed suit and released on their website an advisory warning of the potential adverse effects of Selective Serotonin Reuptake Inhibitors (SSRI's) and other antidepressants on newborns. It was stated that the advisory was intended to increase awareness among mothers and physicians of the possible symptoms that may occur in the newborn, so that they can be recognized and addressed quickly. Some of the symptoms listed were from reports describing feeding and/or breathing difficulties, seizures, muscle rigidity, jitteriness and constant crying. The advisory also stated that if a woman is pregnant and taking an antidepressant she should discuss with her health care professional the potential benefits and risks of treatment options. It was also stated that it is very important that women do NOT stop these medications without consulting their doctor, however, physicians may consider slowly decreasing the dose in the third trimester of pregnancy. It must be noted, that nowhere in this advisory was it stated that women should avoid taking antidepressants during pregnancy. [5]
This advisory was reported in all forms of the media. Table 1 shows some examples, which were selected both randomly and from the women's reports.
Table 1 Samples of media reporting of advisory (headlines)
Radio: (Ottawa) "Antidepressants pose risk to unborn babies"
Internet: (Canada.com News) "Depression drugs can hurt babies"
Newspaper: (The Toronto Star)"Avoid antidepressants in pregnancy"
Television: (CTV) Antidepressants may put unborn babies at risk
Magazine: (Health and Wellness) "Pregnant women warned about antidepressants"
In the seven days following the media reports, we received 49 calls to the Motherisk line from anxious women. This number was in addition to the already substantial number of calls regarding the use of antidepressants we receive each day. We were not surprised that there was this sudden influx of calls, as we felt that this advisory was clearly ambiguous and it was understandable that the media may have misinterpreted some of the intended message.
The objective of this study was to examine the impact of the media translation of this advisory, on the decision-making of women who were currently taking an antidepressant and called Motherisk for information.
Methods
Researchers at the Motherisk Program developed a questionnaire that consisted of 4 questions, (Appendix). Potential participants were women currently treated with an antidepressant who had contacted the Motherisk Program after hearing about the Health Canada advisory in the media and becoming alarmed.
After receiving evidence based information provided by a Motherisk counselor, (table 2) each woman was followed up by phone within one week of their initial call. Upon contact, she was asked if she would be willing to participate in a telephone survey to be conducted by a student. Oral consent was obtained from each participant after the survey was fully explained over the telephone. If the interviewer was unable to reach the women on the first contact, two more attempts were made before giving up.
Table 2 Evidence-based information given by Motherisk counsellors
1. Antidepressants are an important pharmacological tool in the treatment of depression, including during pregnancy.
2. Based on epidemiologic studies they are considered an exposure that would not harm the fetus.
3. Untreated maternal depression is associated with adverse effects on both the mother and the fetus.
4. In a minority of infants there is a mostly self limited but if required, treatable discontinuation syndrome. Consequently, the baby should be observed carefully after birth for signs of withdrawal.
5. If a woman has discussed the benefits and risks of taking an antidepressant during pregnancy with her physician and if the decision is to be pharmacologic treatment, based on current epidemiologic data, there is no reason to discontinue or decrease the drug anytime during pregnancy.
The data were analyzed by descriptive statistics.
Results
A total of 49 women agreed to participate and 43 completed the survey, a response rate of 88%. Of the women who agreed to participate but were lost to follow-up; 2 had incorrect telephone numbers (either because they had given the wrong number or it had been noted incorrectly) and 4 of the women were unavailable to complete the questionnaire. All of the callers reported that the information they received from the media caused a great deal of anxiety. They all felt that this was important information to know, however would not have been so alarmed if it had been translated by the media in a less "scary" fashion.
The following results underscore how the information was translated in such a misleading fashion. This information was specifically aimed at women taking an antidepressant in late pregnancy. Thirty-two of the women were not in the third trimester at the time they called the Motherisk Program and eleven were not even pregnant. Four of them were planning a pregnancy and seven had taken an antidepressant when they were pregnant and their children were now more than one year old. Three of the women who stopped taking their antidepressant, discontinued it abruptly, although two restarted following reassurance from Motherisk. Six women were considering discontinuing their medication, but were reassured by Motherisk and decided to continue. Table 2
Discussion
To our knowledge, this is the first survey of it's kind that has been conducted to examine the impact of public health advisories and the subsequent reports in the media on the decision-making of pregnant women regarding taking antidepressants during pregnancy. We were able to interview a convenience sample of 43 women who had called our line following the dissemination of this information in the media. We did not set out to conduct a randomized controlled study, we simply wanted to document in an observational fashion, how the women who called our program felt about the information they had received. As such, we were unable to ascertain how many other women in the general population were also possibly negatively affected by rash decision-making prompted by these stories in the media. Conversely, we were also unable to examine how many women were unaffected by the media attention to this advisory, as they did not call us.
To date, based on a fairly substantial body of epidemiologic data, there is no evidence that antidepressants are not safe to take during pregnancy [6], in fact, emerging data in the literature documents evidence that not taking an antidepressant if it is warranted may be more harmful. The decision to discontinue taking an antidepressant during pregnancy can have deleterious effects on both the health of the mother and her baby as untreated depression during pregnancy carries its own risks. These include adverse maternal outcomes such as: poor compliance with prenatal care, inadequate nutrition, poor pregnancy outcomes including increased risk of preterm delivery and importantly, an increase risk of post partum depression.[7] Additionally, abrupt discontinuation of antidepressant medication during pregnancy can result in serious physical and psychological adverse effects, which may include substitution of alcohol for medication, acute onset of a major depressive episode and suicidal ideation. [8] Furthermore, studies of infants born to depressed mothers have reported that these infants exhibit "depression-like behavior, demonstrated by decreased facial expressions, inferior orientation skills, excitability and abnormal reflexes.[9] Conversely, the neonatal SSRI poor adaptation syndrome occurs in a minority of cases, is usually self-limited and to date there is no evidence to suggest that a woman should discontinue her antidepressant in late pregnancy.[10] However, the research is ongoing, so it is not possible to say definitively how often this occurs or that there are no long term adverse effects on the child whose mother was exposed to an antidepressant in late pregnancy. A woman and her physician should always discuss the evidence-based information on both the positive and negative effects of treatment with an antidepressant in pregnancy before making a decision, which will ensure the best possible outcome for the mother and child.
It is important that scientific information is disseminated to the public to empower them to take care of their health. Paternalistic models of health care delivery have been replaced by a more balanced approach, which includes both patient and provider goals Today, patients are encouraged to actively participate in the decision-making regarding management and treatment of their health and the various forms of the news media particularly the Internet are powerful tools to facilitate this process. However, the way in which this information is translated from scientific data and disseminated to the public must be carried out in a responsible and circumspect fashion. Shuchman et al. documented the four problem areas in the reporting of medical information to the public by journalists: sensationalism, biases and conflicts of interest, lack of follow-up and stories that are not covered. [11] Motherisk has previously shown that studies reporting results that showed no harmful effects, were underreported as compared to studies that reported results that showed harmful effects, for example, the use of cocaine in pregnancy.[12]
The major limitation of this study, is that we do not know how this media translation of scientific information affected the general population of pregnant women who are taking antidepressants during pregnancy for example, women who did not call Motherisk. It remains unclear how many women may have suffered from preventable adverse events secondary to abrupt discontinuation of their antidepressant as a result of these news stories, as we only have information regarding the women who did call The Motherisk Program. On the other hand some women may have been unaffected by these stories and continued to take their medication without any concerns.
In summary, this is an example of how the news media in their misguided reporting of information from a Health Canada Advisory, which in itself was rather ambiguous, influenced a number of pregnant women to make decisions that may not have been in the best interests of their health or the health of their unborn child.
Appendix
Questionnaire
1. Were you taking an antidepressant when you called Motherisk for information?
YES NO
2. If no:
1) Did you stop taking the medicine:
Abruptly Gradually (i.e. taper off)
2) After speaking with Motherisk, did you restart the antidepressant?
YES NO
3. After speaking with Motherisk, did you speak with your MD?
1) YES NO
2) If Yes:
Did your MD recommend you continue to take the antidepressant?
YES NO
3) If No, why?__________________________________________
4) If Yes:
Did you continue to take the antidepressant?
YES No
4. In your own words, "what impact do you feel that the media stories had on your decision making regarding whether you took the antidepressant or not"?
Table 3 Background of the women who called Motherisk
N = 43
Retrospective 7
Planning 4
Not in 3rd trimester 32
Discontinued drug (3 abruptly) 5
Considered discontinuing drug but did not 6
Recommendations to discontinue (MD) 5
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Larsson A Oxman AD Carling C Herrin J Medical messages in the media-barriers and solutions to improving medical journalism Health Expect 2003 6 323 31 15040794 10.1046/j.1369-7625.2003.00228.x
Neves-e-Castro M Samsioe G Doren M O Skouby S European Menopause & Andropause Society Results from WHI and HERS II – implications for women and the prescriber of HRT 42 255 8 2002 Aug 30
Levens E Williams RS Current opinions and understanding of menopausal women about hormone replacement therapy(HRT)-The University of Florida experience Am J Obstet Gynecol 2004 191 641 6 discussion 646-7 15343254 10.1016/j.ajog.2004.06.074
Ruiz I Bermejo MJ Knowledge of hormone replacement therapy among menopausal women Gac Sanit 2004 18 32 7 14980170 10.1157/13057975
Health Canada advises of potential adverse effects of SSRI's and other antidepressants on newborns Health Canada Online August 9th 2004
Kalra S Born L Sarkar M Einarson A The safety of antidepressant use in pregnancy Expert Opin Drug Saf 2005 4 273 84 15794719 10.1517/14740338.4.2.273
Bennett HA Einarson A Taddio A Koren G Einarson TR Prevalence of depression during pregnancy: systematic review Obstet Gynecol 2004 103 698 709 15051562
Einarson A Selby P Koren G Abrupt discontinuation of psychotropic drugs during pregnancy: fear of teratogenic risk and impact of counselling J Psychiatry Neurosci 2001 26 44 8 11212593
Lundy B Jones NA Field T Prenatal depression effects on neonates Infant Behav Dev 1999 22 119 129 10.1016/S0163-6383(99)80009-5
Koren G Matsui D Einarson A Steiner M Is maternal use of selective serotonin reuptake inhibitors in the third trimester of pregnancy harmful to neonates CMAJ 2005
Schuman M Wilkes MS Medical scientists and health news reporting: a case of Miscommunication Ann Intern Med 126 975 82 1997 Jun 15
Koren G Graham K Shear H Einarson T Bias against the null hypothesis: the reproductive hazards of cocaine Lancet 2 1440 2 1989 Dec 16 2574369
| 15907207 | PMC1156906 | CC BY | 2021-01-04 16:32:05 | no | BMC Pregnancy Childbirth. 2005 May 20; 5:11 | utf-8 | BMC Pregnancy Childbirth | 2,005 | 10.1186/1471-2393-5-11 | oa_comm |
==== Front
BMC Pregnancy ChildbirthBMC Pregnancy and Childbirth1471-2393BioMed Central London 1471-2393-5-91587634510.1186/1471-2393-5-9Research ArticleIncidence of stillbirth and perinatal mortality and their associated factors among women delivering at Harare Maternity Hospital, Zimbabwe: a cross-sectional retrospective analysis Feresu Shingairai A [email protected] Siobán D [email protected] Kathy [email protected] Brenda W [email protected] Department of Preventive and Societal Medicine, University of Nebraska Medical Center 984350 Nebraska Medical Center, Omaha, NE 68198-4350, USA2 Department of Epidemiology, School of Public Health, University of Michigan, 109 Observatory Rd, Ann Arbor, MI 48105, USA3 Department of Community Medicine, University of Zimbabwe, Box A178, Avondale, Harare, Zimbabwe4 Center for Statistical Consultation and Research, University of Michigan, 3554 Rackham Building, 915 E Washington St, Ann Arbor, MI 48109-1070, USA2005 5 5 2005 5 9 9 20 10 2004 5 5 2005 Copyright © 2005 Feresu et al; licensee BioMed Central Ltd.2005Feresu et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Death of an infant in utero or at birth has always been a devastating experience for the mother and of concern in clinical practice. Infant mortality remains a challenge in the care of pregnant women worldwide, but particularly for developing countries and the need to understand contributory factors is crucial for addressing appropriate perinatal health.
Methods
Using information available in obstetric records for all deliveries (17,072 births) at Harare Maternity Hospital, Zimbabwe, we conducted a cross-sectional retrospective analysis of a one-year data, (1997–1998) to assess demographic and obstetric risk factors for stillbirth and early neonatal death. We estimated risk of stillbirth and early neonatal death for each potential risk factor.
Results
The annual frequency of stillbirth was 56 per 1,000 total births. Women delivering stillbirths and early neonatal deaths were less likely to receive prenatal care (adjusted relative risk [RR] = 2.54; 95% confidence intervals [CI] 2.19–2.94 and RR = 2.52; 95% CI 1.63–3.91), which for combined stillbirths and early neonatal deaths increased with increasing gestational age (Hazard Ratio [HR] = 3.98, HR = 7.49 at 28 and 40 weeks of gestation, respectively). Rural residence was associated with risk of infant dying in utero, (RR = 1.33; 95% CI 1.12–1.59), and the risk of death increased with increasing gestational age (HR = 1.04, HR = 1.69, at 28 and 40 weeks of gestation, respectively). Older maternal age was associated with risk of death (HR = 1.50; 95% CI 1.21–1.84). Stillbirths were less likely to be delivered by Cesarean section (RR = 0.64; 95% CI 0.51–0.79), but more likely to be delivered as breech (RR = 4.65; 95% CI 3.88–5.57, as were early neonatal deaths (RR = 3.38; 95% CI 1.64–6.96).
Conclusion
The frequency of stillbirth, especially macerated, is high, 27 per 1000 total births. Early prenatal care could help reduce perinatal death linking the woman to the health care system, increasing the probability that she would seek timely emergency care that would reduce the likelihood of death of her infant in utero. Improved quality of obstetric care during labor and delivery may help reduce the number of fresh stillbirths and early neonatal deaths.
==== Body
Background
Perinatal mortality remains a challenge in the care of pregnant women worldwide, particularly in developing countries [1-3]. To address the problem of perinatal mortality, factors associated with stillbirth, a major contributor of over 50% of perinatal deaths in developing countries, [4] must be understood. Stillbirths are both common and devastating, and in developed countries, about one third has been shown to be of unknown or unexplained origin [4,5]. As is the perinatal mortality rate, the stillbirth ratio is an important indicator of the quality of antenatal and obstetric care [2,3,6], but studies have not distinctively differentiated the frequency of and risk factors for macerated versus fresh stillbirths. Understanding the distribution of fresh and macerated stillbirths and deaths within the immediate postpartum period may help identify the quality of antenatal and obstetric care available to the pregnant women and prioritize appropriate intervention strategies. Macerated stillbirths are often associated with insults that occur in utero during the antenatal period, while fresh stillbirths and early neonatal deaths or mortality (ENNM) may suggest problems with the care available during labor and at delivery [3,7,8]. Few studies from Zimbabwe [9-11], have examined frequency of perinatal mortality and how this outcome varies across important demographic subgroups. Studies from developing countries have not considered the frequency of macerated and fresh stillbirths and their relationship to preterm birth or low birth weight (LBW) [1], and no such study has been conducted in Zimbabwe.
In Zimbabwe, perinatal mortality remains unacceptably high. In Harare, the capital city, perinatal mortality declined from 83 per 1,000 live births in 1978, to 34 per 1,000 live births in1984 and has changed little since then [12,13]. In 1983, an audit of all births occurring within the Greater Harare Maternity Unit (GHMU), which comprises of Harare Maternity Hospital (HMH) and the 12 municipal clinics in Harare, estimated perinatal mortality to be 34.5 per 1,000 live births, with preterm birth being the leading cause of perinatal mortality, accounting for 19.3% of perinatal deaths [14]. By 1989, perinatal mortality had risen to 47 per 1,000 live births [12,13]. Iliff and Kenyon [12,13] estimated that an increase in the number of and mortality from preterm births accounted for about half this increase. In the same study, stillbirth ratio was estimated to be 26 per 1,000 total births. A more recent study estimated the frequency of stillbirth at HMH to be 57 per 1,000 total births [9], which, using conservative assumptions, translates to 33 per 1,000 total births for the GHMU. Prevention of perinatal deaths is critical, especially those associated with LBW and preterm birth, since intuitively, infants who are born early or small have increased risk of morbidity and mortality [2,3,9,14].
Data on the frequency and distribution of adverse birth outcomes are important for planning maternal and child health care services in developing countries, and knowledge of local patterns of morbidity and mortality is essential for improving antenatal and obstetric care. Ensuring a safe and healthy delivery for both mother and child is a priority of the Zimbabwe health care delivery system and is an essential component of safe motherhood initiatives. In this preliminary study, we assessed the contribution of socio-demographic and reproductive/obstetric risk factors to the frequency of fresh and macerated stillbirth and ENNM over a one-year period at HMH, in Zimbabwe.
Methods
The study was carried out at HMH, the largest referral hospital in Zimbabwe. The study was approved by the University of Michigan Institutional Review Board and the Medical Research Council of Zimbabwe, and permission to conduct the study was obtained from the Ministry of Health in Zimbabwe, HMH and from the Harare City Health Department.
The study methods have been described elsewhere [9]. Briefly, information on all births occurring from October 1, 1997, through September 30, 1998, at HMH was abstracted from the maternity delivery registry records. For each birth, information was abstracted on the date of birth, residential area of the woman (rural/urban), whether the mother attended prenatal care or not, maternal age and parity, estimated gestation, birth weight; sex and vital status of the baby at birth, whether the infant was a single or multiple delivery, and the type of delivery. It was not possible to link births from multiple gestations in this data set. A woman was considered to have received prenatal care when she had at least one visit for prenatal care during her pregnancy. Parity was the number of previous pregnancies ending after 20 completed weeks of gestation including stillbirth (categorized as 0, 1 to 2, and more than 2 pregnancies). Type of delivery denoted whether the infant was delivered vaginally presenting as cephalic, face to pubis or breech; vaginally with instrumental assistance; or by Cesarean section.
Eligibility criteria for this study were based on the WHO definition of viability, that is, a birth weight of at least 500 grams gestational age at least of 20 weeks [15]. Births without information on vital status were excluded. A stillbirth was defined as intrauterine death of a fetus weighing at least 500 grams after 20 completed weeks of gestation occurring before the complete expulsion or extraction from its mother. A fresh stillbirth was defined as the intrauterine death of a fetus during labor or delivery, and a macerated stillbirth was defined as the intrauterine death of a fetus sometime before the onset of labor, where the fetus showed degenerative changes [15] as reported in the obstetric records by the attending physician/midwife. An ENNM was defined as a death that occurred within the first hour of life.
Gestational age (GA) at birth was estimated by the number of days between the first day of the last menstrual period (LMP) and date of birth expressed in completed weeks after LMP and a clinical estimate as recorded in the maternity delivery record. Preterm birth was defined as a birth occurring at or before 37 completed weeks of gestation. A post-term birth was defined as a birth occurring after 44 weeks of gestation. Birth weight was defined as the first measurement of body weight, usually in the first hour of life, measured to the nearest gram. A LBW birth was defined as the birth of an infant weighing less than 2,500 grams at birth irrespective of gestational age. We also defined three LBW subgroups; term LBW birth, preterm LBW birth and very LBW birth defined as infants weighing below 1,500 grams. A high birth weight birth, based on the upper 10th percentile of our birth weight distribution, was defined as the birth of an infant weighing above 3,500 grams.
A total of 18,149 births were recorded in the 12-month study period. On average between 50 to 60 deliveries occurred daily. We excluded 32 (0.2%) births below 20 weeks of gestation, 68 (0.4%) that did not have information on vital status, 78 (0.4%), that weighed below 500 grams at birth and 795 (4.4%) that were missing information on birth weight or estimated gestation. Finally we excluded an additional 103 (0.6%) births for which birth weight/gestational age combination were implausible based on the algorithm advocated by Alexander et al. [16], leaving 17,072 births for this analysis.
Statistical analysis
The numbers of fresh, macerated, and un-typed stillbirths are presented as a proportion of all births, and ENNM are presented as a proportion of live births. To examine predictors of these outcomes, cross-tabulations by each covariate were examined using chi-square tests of homogeneity. As the complete population of births over a one-year period within the hospital was ascertained, we could directly estimate the risk of stillbirth and ENNM for each covariate. In the unadjusted analyses, relative risks (RR) and 95% confidence intervals (CI) were calculated for each demographic and reproductive risk factor using EP-INFO 2000. All 17,072 eligible births were included for stillbirth analyses. Given the differences in risk for singleton and multiple births, the subsequent multivariate analyses are limited to singleton births, therefore 16,023 singleton births were used for the stillbirths and combined stillbirth and ENNM analyses. For ENNM only analyses, 15,117 live singleton births were included.
Multivariable generalized linear regression models with a log-link function and binomial error using SAS [SAS Institute Inc., Cary, NC, USA] version 8.1, were used to model each outcome. A log (rather than logit) link function and binomial errors were used to allow estimation of relative risks (rather than odds ratios). Relative risks of stillbirth and 95% confidence intervals were calculated after adjusting for maternal age, residence, prenatal care, parity, infant sex and type of delivery. As estimates were not stable due to collinearity, we combined births delivered as face to pubis with those of normal vaginal delivery for the type of delivery analysis, adjusting for maternal age, residence, parity, and infant sex.
Recent epidemiological literature has suggested at least two alternative methods of calculating the probability of stillbirth by GA, with the differences being in the definition of the denominator [17,18]. One method calculates the GA-specific probability of stillbirth as the number of stillbirths divided by the number of births at a given GA. A second method calculates the GA-specific hazard of stillbirth as the number of stillbirths at a given GA divided by the number of fetuses at the given GA still remaining to be born [17-19]. We provide estimates using the first definition in Tables 2 and Additional files 1 and 2, and use the second definition in Tables 3 and 4 and Figures 1 and 2. Covariate effects on gestational age-specific mortality were estimated using Cox regression model with estimated gestation at birth as the time variable. Hazard ratios (HR) for stillbirth and 95% CI were calculated [17,18]. All 16,023 singleton births were used for this analysis, and stillbirths and ENNM were combined as outcomes of interest [17,18]. The interaction between each covariate and time (GA) was tested for evidence of non-proportion hazards. Covariates with no evidence of non-proportion hazards (maternal age, infant sex and parity) were refitted without the interaction term. For covariates with evidence of non-proportion hazards (residence and prenatal care) we also calculated the risk of death at 20, 28, 32, 36, 40 and 42 weeks of gestation. The probabilities of death at each stage of GA were estimated using life tables for prenatal care and for urban versus rural residents, and are presented in Figures 1 and 2.
Table 2 Distribution and risk of Deaths within First Hour of Life for 15,117 Live, Singleton Deliveriesat Harare Maternity Hospital; October 1997 to September 1998
All Live Births All Deaths Within First Hour of Life Relative Risk (95% CI) Adjusted a,b Relative Risk (95% CI)
n n %
Total 15,117 129 0.9
Maternal age
Below 20 3,214 41 1.3 1.69 (1.16 – 2.45) 1.21 (0.77 – 1.89)
20 to 35 10,860 82 0.8 Reference Reference
Above 35 969 4 0.4 0.55 (0.20 – 1.49) 0.57 (0.19 – 1.70)
Infant sex
Male 7,594 85 1.0 1.69 (1.17 – 2.45) 1.67 (1.15 – 2.43)
Female 7,076 50 0.6 Reference Reference
Residence
Urban 13,915 106 0.8 Reference Reference
Rural 2,118 22 1.0 1.27 (0.80–2.00) 1.36 (0.85 – 2.18)
Prenatal care
At least one visit 13,527 101 0.8 Reference Reference
No prenatal care 1,463 26 1.8 2.38 (1.55 – 3.65) 2.52 (1.63 – 3.91)
Parity
Para 0 6,945 77 1.1 1.81 (1.22 – 2.70) 1.52 (0.97 – 2.38)
Para 1–2 5,723 35 0.6 Reference Reference
Para above 2 2,377 15 0.6 1.03 (0.56 – 1.89) 1.25 (0.65 – 2.40)
Delivery type
Normal vaginal delivery 11,658 82 0.7 Reference Reference
Breech 386 27 7.0 9.94 (6.52 – 15.18) 10.53 (6.78 – 6.34)
Instrumental 344 9 2.6 3.72 (1.88 – 7.34) 3.38 (1.64 – 6.96)
Cesarean section 2,707 10 0.4 0.53 (0.27 – 1.01) 0.56 (0.29 – 1.09)
a For Adjusted analysis, maternal age, sex residence, prenatal care, parity were entered in one model
b Delivery type was adjusted for mother' age, residence, sex and parity
* Excludes 1,049 multiple gestations births, and 906 stillbirths
Table 3 Demographic and Obstetric Characteristics: Hazard Ratios of Combined Stillbirth and Early Neonatal Deaths for 16,023 Singleton Deliveriesat Harare Maternity Hospital; October 1997 to September 1998
Hazard Ratio (95% Confidence Intervals)
Maternal age
Below 20 0.86 (0.73 – 1.01)
20 to 35 Reference
Above 35 1.50 (1.21 – 1.84)
Infant sex
Male 1.09 (0.96 – 1.24)
Female Reference
Parity
Para 0 0.91 (0.80 – 1.05)
Para 1–2 Reference
Para above 2 1.20 (1.01 – 1.42)
*There was no evidence of time dependence of mortality risk (i.e., no evidence of non-proportion hazards) for maternal age, infant sex, or parity
Table 4 Demographic and Obstetric Characteristics for which the Hazard Ratio Changed over Gestational Age, for Combined Stillbirth and Early Neonatal Deaths for 16,023 Singleton Deliveries at Harare Maternity Hospital; October 1997 to September 1998
Hazard Ratio (95% Confidence Interval)
Gestational Age 28 32 36 40 42
Residence
Urban Reference Reference Reference Reference Reference
Rural 1.04 (0.82 – 1.32) 1.22 (1.03 – 1.45) 1.44 (1.21 – 1.70) 1.69 (1.34 – 2.12) 1.83 (1.39 – 2.40)
Prenatal care
At least one visit Reference Reference Reference Reference Reference
No prenatal care 4.76 (4.00 – 5.66) 3.32 (2.84 – 3.87) 2.32 (1.90 – 2.82) 1.62 (1.23 – 2.11) 1.35 (0.99 – 1.84)
* There was evidence of time dependence of mortality risk for residence or receiving prenatal care
Figure 1 Probability of Combined Stillbirth and Early Neonatal Death among 16,023 Singleton Deliveries by Prenatal Care at Harare Maternity Hospital; October 1997 through September 1998
Figure 2 Probability of Death among 16,023 Singleton Deliveries by Residence at Harare Maternity Hospital; October 1997 through September 1998
Results
Demographic and obstetric risk factors of stillbirth by type
Of the 959 stillbirths, 201 (21%) were fresh and 458 (48%) were macerated stillbirths; the type of stillbirth for 300 (31%) births could not ascertained from the available obstetrical records. The annual stillbirth ratio at HMH was 56 per 1,000 total births, of which 12 per 1,000 were fresh stillbirths, 27 per 1,000 were macerated stillbirths, and 17 per 1,000 were for stillbirths whose type was not indicated in the obstetric records (Table 1). Age of mothers ranged from 10 to 50 years. Most mothers were 20 to 35 years old, with a mean age at delivery of 24.6 years. Most mothers (86%) resided in urban areas. Women present late for prenatal care, GA less than 28 weeks accounted for 384 singleton births, of which 180 (47%) did not receive prenatal care. It is uncommon in Zimbabwe for women to present for prenatal care prior to 28 weeks of pregnancy [9,20], and in the rural areas women could present at a health center at the time of labor if they are having difficulty. About 45% of the women were primiparous, with parity ranging from 0 to 12. The mean age of mothers with parity of 0, was 20.7 (range 10–41 years), compared to 27.7 (range 14–50) for women with parity 1 and more. As expected, there were slightly more male than female births. A total of 17.5% of the deliveries were by Cesarean section, and an additional 2.2% required some form of instrumentation during delivery, while 5.9% of the births were either breech or face to pubis presentations. About 6% of the infants were from multiple gestations, and in crude analysis, multiple gestation births were less likely to be macerated stillbirths (RR = 0.57; 95% CI 0.35–0.93), but more likely to die in the first hour of life (RR = 1.88; 95% CI 1.12–3.14).
Table 1 Distribution of births by Vital Status at Harare Maternity Hospital; October 1997 to September 1998
Total Births All Stillbirths Fresh Stillbirths Macerated Stillbirths Un-typed Stillbirths
n n % n % n % n %
Total 17,072 959 5.6 201 1.2 458 2.7 300 1.7
Maternal age
Below 20 3,475 149 4.3 34 1.0 69 2.0 46 1.2
20 to 35 12,333 700 5.7 152 1.2 341 2.8 207 1.7
Above 35 1,177 102 8.7 15 1.3 44 3.7 43 3.7
Infant sex
Male 8,546 462 5.4 91 1.1 226 2.6 145 1.7
Female 7,978 419 5.3 93 1.2 197 2.5 129 1.6
Residence
Urban 14,576 794 5.5 170 1.2 391 2.7 233 1.6
Rural 2,402 163 6.8 31 1.3 65 2.7 67 2.8
Prenatal care
At least one visit 15,135 729 4.8 158 1.1 353 2.3 218 1.4
No prenatal care 1,796 218 12.1 41 2.3 101 5.6 76 4.2
Parity
Para 0 7,608 375 4.9 81 1.1 183 2.4 111 1.4
Para 1–2 6,563 383 5.8 76 1.1 189 2.9 118 1.8
Para above 2 2,821 195 6.9 42 1.5 85 3.0 68 2.4
Gestation type
Singleton 16,023 906 5.7 189 1.2 441 2.8 276 1.7
Multiple 1,049 53 5.1 12 1.1 17 1.6 24 2.4
Delivery type
Normal vaginal delivery 12,654 662 5.2 142 1.1 357 2.8 163 1.3
Breech 795 140 17.6 28 3.5 78 9.8 34 4.3
Instrumental 377 24 6.4 10 2.7 6 1.6 8 2.1
Cesarean section 2,990 92 3.1 9 0.3 10 0.3 73 2.5
Face to pubis 201 8 4.0 5 2.5 1 0.5 2 1.0
Additional file 1 presents the crude relative risks of stillbirth by type of demographic and reproductive characteristics of the study population. Young mothers were less likely than older women (RR = 0.73; 95% CI 0.61–0.88) to deliver a stillborn infant, with the reduction in risk particularly evident for macerated stillbirth (RR = 0.72; 95% CI 0.56–0.93). In contrast, women above 35 years had a 59% increased risk of stillbirth and 43% increase in the likelihood of delivering a macerated stillbirth. Rural women delivering at HMH had a 24% increased risk of stillbirth compared with women who resided in urban areas. Women who did not receive prenatal care consistently had over a 2.3-fold increase in the risk of stillbirth of any type. Compared to a normal vaginal delivery, breech deliveries were over 4.7 times more likely to be stillbirth, while births by Cesarean section were less likely to result in any type of stillbirth. Delivery by instrumentation was 2.2 times more likely to result in a fresh stillbirth than was normal vaginal delivery.
Additional file 2 presents adjusted relative risks for stillbirth by demographic and reproductive characteristics. Except for parity, which is correlated with maternal age (Pearson coefficient r = 0.76 p-value = 0.0001), the risks did not change in the adjusted analysis. Overall, risks for un-typed stillbirths did not differ from those for fresh and macerated stillbirths, except for residence and delivery by Cesarean section.
Demographic and obstetric risk factors of ENNM
The early neonatal mortality ratio was 9 per 1000 live births. Table 2 presents the distribution and risk for ENNM by demographic and reproductive characteristics. Mothers under 20 years old were 69% more likely than mothers 20 to 35 years of age to deliver an infant who died within the first hour of life, and similarly, risk for primiparous women was 81% compared to multiparous women. Male infants had a 69% increased risk of dying within the first hour of life. Women who did not receive prenatal care consistently had over a 2.4-fold increase in the risk of ENNM. Compared to a normal vaginal delivery, breech deliveries were 9.9 times more likely to end up as ENNM. Delivery by instrumentation was 3.7 times more likely to result in ENNM than was normal vaginal delivery. Except for maternal age and parity the risks remained the same or elevated in adjusted analysis.
Relationship of mortality to demographic and obstetric factors
Table 3 presents the crude hazard ratios for mortality (stillbirths and ENNM combined) for demographic and obstetric characteristics that had constant hazard ratios over GA. There was a 50% increased risk of death for mothers over age 35 years compared with mothers 20 to 35 years of age. There was a 20% increased risk of death for mothers parity above 2 compared with mothers with parity 1 to 2. Infant sex was not significantly associated with mortality.
Relationship of mortality to prenatal care and residence
Table 4 presents crude hazard ratios for mortality (stillbirths and ENNM combined) for demographic and obstetric characteristics for which hazard ratios change over GA. These variables included prenatal care and residence. The hazard ratio for prenatal care decreased with increasing gestation, from 4.76 at 28 weeks to 1.35 at 42 weeks. Figure 1 depicts the hazard function by GA comparing mothers who did and did not receive prenatal care. At 20 weeks, the hazard functions for mothers who did not receive prenatal care and those who did are similar. However, because of the low probability of death before week 35 among those receiving prenatal care, the relatively higher probability among those not receiving prenatal care results in a high relative risk. After week 35, the mortality (stillbirths and ENNM combined) risk in both groups increases proportionally, but for those without prenatal care remain at higher risk and the effect is attenuated at 42 weeks.
As gestation increased, the risk of mortality for rural residence increased, from 1.04 at 28 weeks to 1.69 at 40 weeks and 1.84 at 42 weeks of gestation (Table 6). Figure 2 depicts the hazard function by gestational age comparing rural and urban residence. Up to 35 weeks the hazards are very similar, beyond 35 weeks, the hazards increase proportionally for both groups, but predominantly higher for births from mothers who resided in rural areas.
Distribution of stillbirth and ennm by birth weight and ga categories
Additional file 3 shows the distribution of stillbirth and ENNM by birth weight and GA categories, using the traditional methods (birth weight-specific mortality, where LBW is analyzed as preterm LBW, term LBW) [18]. Sixteen percent of all stillbirths were LBW and 17% were preterm. Almost 3% of neonatal deaths were LBW, and 3% were preterm births.
Discussion
This paper evaluates the distribution of and risk factors for fresh and macerated stillbirth and ENNM among mothers giving birth at the largest hospital serving Harare, Zimbabwe. The proportion of macerated stillbirths in Harare is higher than in more developed countries, suggesting the presence of insults to the developing fetus and the need for timely screening and management of chronic conditions and infections. A considerable proportion of the stillbirths were fresh stillbirths, and the frequency of ENNM was high, suggesting the need for improved obstetric care and availability of emergency services during the delivery period. Lack of prenatal care was associated with increased risk of stillbirth and ENNM whether we analyzed using traditional methods (gestational age-specific mortality) [18] or with GA as a time-varying factor as argued by other experts [21-23]. Similarly, rural residence was associated with increased risk of all stillbirth and ENNM whether we analyzed using traditional methods (gestational age-specific mortality) or with GA as a time-varying factor. But most importantly, using GA as a time-varying factor clarifies where and when the risk of death is more prominent. Fresh or macerated stillbirths and ENNM were more likely to be delivered breech, but less likely to be delivered by Cesarean section. Cesarean section appears to protect against stillbirth in this population. Fresh stillbirths and ENNM were also associated with delivery by instrumentation.
The incidence of stillbirth, 56 per 1,000 total births we report for HMH, is higher than the 26 per 1,000 total births reported by Iliff and colleagues using 1989 data from HMH and 9 Harare municipal clinics [12,13], and higher than the 45 per 1,000 live births at Mpilo Maternity Hospital [24], another large referral hospital in the second largest city in Zimbabwe. Our findings differ from previous Zimbabwean studies because HMH is the largest referral center in this country, and would be expected to have higher mortality rates than other hospitals referring their most complicated cases. When we recalculate our rates based on the number of deliveries in the GHMU, 56% of which occur at HMH [9,25], and assuming no stillbirths occurred in the clinics, we estimate a population-based stillbirth ratio of 33 per 1,000 total births, a figure more comparable to that reported by Iliff and colleagues.
In this population, more stillbirths were macerated, suggesting existence of problems linked to the antenatal period, which could be related to congenital malformations [2,4]; obstetric hemorrhage; preclampsia [2,4-6,26]; infections such as syphilis [7,8,26,27]; or existing maternal chronic conditions such as hypertension, cardiac disease, and diabetes [2,4-6], none of which our study had the ability to evaluate. Smoking, which is an important factor especially in developed countries, was not a factor for this population [28]. Fresh stillbirths contribute 1.2% of all births while ENNM contribute 0.9% of live births at this institution, and both are likely to be related to fetal hypoxia [2,12,26], congenital malformations [2,4,12], quality of delivery care given to a woman during labor and delivery, and poor access to emergency obstetric care.
As would be expected, lack of prenatal care was consistently and strongly associated with stillbirths and ENNM, similar to what other studies have reported [7,29-31]. Although women in Zimbabwe usually begin prenatal care at 28 weeks of gestation a considerable number will present at hospital before that time for a problem related to their pregnancy leading to an adverse birth outcome [[9,20], and [28]]. Had the adverse birth outcome not occurred prior to 28 weeks of gestation, these women could have had an opportunity to present for prenatal care, later in their pregnancy as is common for most women in Zimbabwe. The risk of mortality (stillbirths and ENNM combined) increases with increasing GA before 35 weeks of gestation, but decreases proportionally thereafter. There was a crossover of risk of mortality by prenatal care much later in the course of pregnancy, at 42 weeks of gestation, a phenomenon reported but at earlier gestational age by other studies [17-19]. Thus, the risk of mortality was much higher for women who did not receive prenatal care compared to those who did in the earlier gestational ages, and was moderately higher proportionally after 35 weeks of gestation, and was attenuated at term. In developing countries, prenatal care, even if only attended once, remains an important factor in obstetric care, as this may be a critical linkage between the woman with maternity care services [9,28]. In contrast, research findings in middle-income countries emphasize the importance of the number of prenatal care visits and the adequacy and quality of prenatal care services. WHO recommends using prenatal care as a strategy for improved obstetric care [32]. Our data suggest that prenatal care may help ensure that interventions occur in a timely manner.
Rural residence was associated with increased risk of all stillbirths as reported by previous studies [9,28]. The risk of mortality (stillbirths and ENNM combined) increased with increasing GA, similar to other study reports [17-19]. Before 35 weeks both rural and urban women have a similar risk of mortality. Although the risk of mortality increases proportionally to term for both groups, rural women have a higher risk of having their infant dying in utero and within the first hour of life. Prior to 28 weeks a considerable number of women from rural residence did not receive prenatal care, 39 (10%). After 35 weeks rural women who end up with their infant dying in utero or within the fist hour of life might have been women referred to HMH from rural centers with a condition/complication related or leading to the adverse birth outcome. Caution should be taken when interpreting this finding, because we have two artifacts. Intuitively, after 35 weeks, women who did not receive prenatal care are primarily from rural areas and were likely not to receive prenatal care throughout their pregnancy. Secondly, we do not have the counter population (rural women who attended prenatal care) in our denominator.
For maternal age, infant sex and parity in our study, which were not time-dependent, using either traditional methods (gestational age-specific mortality) [18] or the analysis where GA is a time varying factor did not change the risk of mortality (stillbirths and ENNM combined). But, using GA as a time-varying factor for the analysis helps us to understand further the relationship between some of the maternal factors and mortality, which otherwise would have been missed. For prenatal care and residence, we were able to show how the risk of mortality was distributed at each stage of GA. We were able to separate effects at early periods versus later stages in pregnancy, which is useful for health-care planners, policy makers, and implementers, in terms of targeting resources. For example, for prenatal care, we were able to show that the risk of morality is high at early gestation. The risk of mortality persists later in pregnancy, but decreasing proportionally throughout pregnancy, being higher for women who did not receive prenatal care. This finding suggests a need to focus and emphasize on early booking and the critical role of prenatal care in developing countries. With regards to residence, knowledge that infants of rural women who get referred to urban institutions have the highest risk of mortality may suggest the need to pay more attention during the antenatal period and to improve the referral system and emergency care services.
Stillbirths, irrespective of type, and ENNM were less likely to be delivered by Cesarean section. It is conceivable that factors leading to stillbirth may cause mothers to have a Cesarean section, but our results show that Cesarean section was consistently protective of either stillbirth or ENNM. This finding may suggest that obstetricians are careful not to perform Cesarean section unless it is indicated for stillbirths, or ENNM, or that whenever a Cesarean section is performed, it saves life of the infant.
Stillbirths and ENNM were likely to be delivered breech. For fresh stillbirths and ENNM, this finding is consistent with the clinical observation that because of the nature and dynamics of this type of delivery, these infants are likely to die during or at delivery [9,28]. Similarly, infants delivered by some form of instrumentation were more likely to die within the first hour of life. For macerated stillbirths, this finding may be more related to preterm infants and would be consistent with the clinical observation that the infants turn to optimal birth presentation at about 34 weeks of gestation. About 242 (47%) of singleton births delivered as breech were preterm. Intuitively, infants that are at risk because of their small size and level of maturity are likely to face the additional risk of breech presentation.
Maternal age effects were common in stillbirths, consistent with other studies [33-35]. But the effects of maternal age were more prominent for macerated versus fresh stillbirths, again strengthening the possibility that maternal chronic disease conditions in later years of life may play a significant role. Additionally, older mothers were at greater risk for stillbirth, but lower risk for neonatal death. The crude risk for young maternal age, which was similar to crude risk for primiparity for early neonatal births, was attenuated after controlling for residence, parity, prenatal care and infant sex.
This study has some limitations. As the study was a retrospective analysis of data obtained from delivery logs, we were unable to examine risk factors such as chronic and comorbid conditions, congenital malformations, obstetric complications, and infections. Although we could not identify the stillbirth status of nearly one-third of stillbirths, risk estimates for un-typed stillbirths were similar to those for fresh and macerated stillbirths. Arguably, focusing solely on births within HMH raises concerns about selection bias. However, when we adjust our estimated rate to the base population, our rates are comparable to those previously reported [9,12-14]. Information on gestational age was limited to the clinicians' estimate and LMP information reported as weeks recorded in the obstetric log, thus some error in the classification of preterm births is likely [36,37]. Regardless, this pilot study is one of the few to characterize socio-demographic and reproductive risk factors for stillbirth and ENNM in this population [9]. Our ability to distinguish risks for macerated and fresh stillbirth has direct implications on quality of care given to pregnant women in Zimbabwe.
We were not able to show risk by combined birth weight and GA categories, which would otherwise be important for clinicians [17-19], [38,39], for to do so is impossible as birth weight varies with GA [17]. Therefore, it would be not feasible to put both variables in the same model. Our use of GA as a time-varying variable in this analysis helps conceptualize and define where risk occurs in the GA continuum.
Conclusion
Our findings suggest that earlier perinatal care could assist in early identification and treatment of risk factors for macerated stillbirth, especially those that are preventable. Zimbabwean women enter prenatal care late in pregnancy, booking at 28 weeks or later [9,29-31]. Effective programs to decrease the frequency of stillbirth may require that entry to prenatal care begin by at least 20 weeks of gestation. Increased focus on health education programs, which emphasize the benefits of prenatal care and early booking in the first trimester or by 20 weeks of pregnancy, is needed. Earlier booking for prenatal care creates a critical linkage between the woman and the health care system, which may increase the probability that the woman will seek emergency care in a timely manner. In Zimbabwe, more focus is needed on the timing and adequacy of care in maternal and child health programs, and more research is needed on barriers to early entry to prenatal care.
There is also a need to improve quality of care and access to emergency care during labour and delivery to reduce the number of fresh stillbirths and ENNM. Cesarean section should be made readily available as it improves birth outcomes. Further studies should incorporate information from women served by the entire GHMU, to improve infant mortality and morbidity in Zimbabwe.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
SF conceived of the study, participated in design of the study, carried out the data collection and all analyses and drafted the manuscript.
SD participated in conceiving and designing the study, directed the analysis of the study and reviewed the manuscript.
KW participated in the analysis of the study and formulated the probability functions used to construct the figures.
BG participated in the analysis of the study, reviewed models to calculate the relative risk, and reviewed the manuscript. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional File 1
Demographic and Obstetric Characteristics and Crude Risks of Stillbirth for 16,023 Singleton Deliveries at Harare Maternity Hospital; October 1997 to September 1998
Click here for file
Additional File 2
Adjusteda Demographic and Obstetric Characteristics and Risks of Stillbirth for 16,023 Singleton Deliveries at Harare Maternity Hospital; October 1997 to September 1998
Click here for file
Additional File 3
Frequency of Stillbirth by Birth Weight and Gestational Age Categories for 17,072 Deliveries at Harare Maternity Hospital; October 1997 to September 1998
Click here for file
Acknowledgements
Funding for the research was from the University of Michigan, University of Zimbabwe and W.K. Kellogg Foundation. We thank the Harare Maternity Hospital staff, and the Department of Community Medicine at the University of Zimbabwe for providing space and support during data collection. This project was partially supported by grant #D43-TW01276 from the Fogarty International Center and the National Institute of Child Health and Human Development. We acknowledge the work done by research assistants, Ms J Musengi, Ms D Matsika, Ms K Sithole, Mrs. F Shonhiwa and Ms T. Feresu, and S Nardie for editing this manuscript.
==== Refs
Kramer MS The Epidemiology of Adverse Pregnancy Outcomes; An Overview J Nutr 2003 133 1592S 1596S 12730473
Kumar MR Bhat BV Oumachigui A Perinatal mortality trends in a referral hospital Indian J Pediatr 1996 63 357 361 10830011
Kambarami RA Levels and risk factors for mortality in infants with birth weights between 500 and 1800 grams in a developing country: a hospital based study Cent Afr J Med 2002 48 133 136 14562599
Cnattingius S Stephansson O The Epidemiology of Stillbirth Semin Perinatol 2002 26 25 30 11876563
Gardosi J Mul T Mongelli M Fagan D Analysis of birthweight and gestational age in antepartum stillbirths Br J Obstet Gynaecol 1998 105 524 530 9637122
Wessel H Cnattingius S Bergstrom S Dupret A Reitmaier P Maternal risk factors for preterm and low birthweight in Cape Verde Acta Obstet Gynecol Scand 1996 75 360 366 8638457
Conde-Agudelo A Belizan JM Diaz-Rossello JL Epidemiology of fetal death in Latin America Acta Obstet Gynecol Scand 2000 79 371 378 10830764 10.1034/j.1600-0412.2000.079005371.x
Osman NB Challis K Cotiro M Nordahl G Bergstöm S Perinatal Outcome in an Obstetric Cohort of Mozambican Women J Trop Pediatr 2001 47 30 38 11245348 10.1093/tropej/47.1.30
Feresu SA Welch K Gillespie B Harlow SD Incidence of and Sociodemographic Risk Factors for Stillbirth, Pre-term birth and Low Birthweight in Zimbabwean Women Paediatr Perinat Epidemiol 2004 18 154 163 14996257 10.1111/j.1365-3016.2003.00539.x
Moyo SR Tswana SA Nystrom L Mahomed K Bergstrom S Ljungh A An Incident Case-Referent Study of stillbirths at Harare Maternity Hospital: Socio-Economic and Obstetric Risk Factors Gynecol Obstet Invest 1994 37 34 39 8125406
Kambarami RA Chirenje M Rusakaniko S Anabwani G Perinatal mortality rates associated socio-demographic factors in two rural districts in Zimbabwe Cent Afr J Med 1997 43 158 162 9431742
Iliff PJ Kenyon N Perinatal mortality statistics in Harare 1980–1989 Cent Afr J Med 1991 37 133 136 1790552
Iliff PJ Bonduelle MM Gupta V Changing stillbirth rates: quality of care or quality of patients? Cent Afr J Med 1996 42 89 92 8791862
De Muylder X Perinatal mortality audit in Zimbabwean district Paediatr Perinat Epidemiol 1989 3 284 293 2788880
Department of Reproductive Health Research, World Health Organization Guidelines Overview Safe Motherhood Needs Assessment (WHO Publication No. WHO/RHT/MSM/96.18 Rev.1) 2001 Geneva, Switzerland: World Health Organization
Alexander GR Himes JH Kaufman RB Mor J Kogan M A United States National Reference for Fetal Growth Obstet Gynaecol 1996 87 163 168 10.1016/0029-7844(95)00386-X
Platt RW Joseph KS Anath CV Grondines J Abrahamowicz M Kramer MS A proportional Hazards Model with Time-dependent Covariates and Time-varying Effects for Analysis of Fetal and Infant Death Am J Epidemiol 2004 160 199 206 15257989 10.1093/aje/kwh201
Joseph KS Incidence-based measures of birth, growth restriction, and death can free perinatal epidemiology from erroneous concepts of risk J Clin Epidemiol 2004 57 889 897 15504632 10.1016/j.jclinepi.2003.11.018
Yudkin PL Risk of unexplained stillbirth at different gestational ages Lancet 1987 8543 1192 1194 2883499
Munjanja S Lindmark G Nystrom L Randomised controlled trial of reduced-visits programme of antenatal care in Harare, Zimbabwe Lancet 1996 348 364 369 8709734 10.1016/S0140-6736(96)01250-0
Cheung YB On the Definition of Gestational-Age-specific Mortality Am J Epidemio 2004 160 207 210 10.1093/aje/kwh202
Klebanoff MA Schoendorf KC Invited Commentary: What's so bad about curves crossing anyway? Am J Epidemiol 2004 160 211 212 15257991 10.1093/aje/kwh203
Wilcox AJ Weinberg CR Invited Commentary: Analysis of Gestational-Age-specific Mortality – On what Biologic Foundations? Am J Epidemiol 2004 160 213 214 15257992 10.1093/aje/kwh204
Aiken CG The causes of perinatal mortality in Bulawayo, Zimbabwe Cent Afr J Med 1992 38 263 281 1477875
Tshimanga M Makunike B Wellington M An audit of maternity referrals in labour from primary health care clinics to a central hospital in Harare, Zimbabwe Cent Afr J Med 1997 43 279 283
Chalumaeau M Bouvier-Colle M Breart G the MOMA Group Can clinical risk factors for late stillbirth in West Africa be detected during antenatal care or only during labour? Int J Epidemiol 2002 31 661 668 12055171 10.1093/ije/31.3.661
Kambarami RA Manyame B Macq J Syphilis in Murewa District, Zimbabwe: an old problem that rages on Cent Afr J Med 1998 44 229 232 10101428
Feresu SA Woelk GB Harlow SD Risk factors for Prematurity at Harare Maternity Hospital, Zimbabwe Int J Epidemiol 2004 33 1194 201 15522924 10.1093/ije/dyh120
Axemo P Liljestrand J Berstrom S Gebre-Medhin Aetiology of Late Fetal Death in Maputo Gynecol Obstet Invest 1995 39 103 109 7537709
Fawcus SR Crowther CA Van Baelen P Marumahoko J Booked and unbooked mothers delivering at Harare Maternity Hospital, Zimbabwe: a comparison of maternal characteristics and fetal outcome Cent Afr J Med 1992 38 402 408 1308714
Galvin J Woelk GB Mahomed K Wagner N Mudzamari S Williams MA Prenatal care utilization and fetal outcomes at Harare maternity Hospital, Zimbabwe Cent Afr J Med 2000 47 87 92
UNDP/UNFPA/WHO/World Bank Special Programme of Research Development and Research Training in Human Reproduction Who Antenatal Care Randomized Trial: Manual for the Implementation of the New Model 2002 Geneva: World Health Organization
Fretts RC Schmittdiel J Mclean FH Usher RH Goldman MB Increased maternal age and risk of fetal death N Engl J Med 1995 333 955 957 10.1056/NEJM199510123331501
Fretts RC Usher RH Causes of Fetal Death in Advanced Maternal Age Obstet Gynaecol 1997 89 40 45 10.1016/S0029-7844(96)00427-9
Waldoer T Haidinger G Langgassner J Tuomilehto J The effect of maternal age and birth weight on the temporal trend in stillbirth rate in Austria during 1984–1993 Wien Klin Wochenschr 1996 108 643 48 8940769
Feresu SA Gillespie BW Sowers MF Johnson TRB Welch K Harlow SD Improving the assessment of gestational age in a Zimbabwean population Int J Gynaecol Obstet 2002 78 7 18 12113965 10.1016/S0020-7292(02)00094-2
Feresu SA Harlow SD Gillespie B Welch K Johnson TRB Birthweight Adjusted Dubowitz methods: Reducing Misclassification of Assessments of gestational age in a Zimbabwean population Cent Afr J Med 2003 49 47 53 15214282
Platt RW Ananth CV Kramer MS Analysis of neonatal mortality: is standardizing for relative birth weight biased? BMC Pregnancy and Childbirth 2004 4 1 10 15005809
Kulmala T Vaahtera M Ndekha M Koivisto AM Cullinan T Salin ML Ashorn P The importance of preterm births for peri- and neonatal mortality in rural Malawi Paediatr Perinat Epidemiol 2000 14 219 226 10949213 10.1046/j.1365-3016.2000.00270.x
Kaushik SL Parmar VR Grover N Kaushik R Neonatal mortality rate: relationship to birth weight and gestational age Indian J Pediatr 1998 65 429 433 10771994
| 15876345 | PMC1156907 | CC BY | 2021-01-04 16:32:05 | no | BMC Pregnancy Childbirth. 2005 May 5; 5:9 | utf-8 | BMC Pregnancy Childbirth | 2,005 | 10.1186/1471-2393-5-9 | oa_comm |
==== Front
BMC PediatrBMC Pediatrics1471-2431BioMed Central London 1471-2431-5-101590451010.1186/1471-2431-5-10Research ArticleEarly discontinuation of intravenous antimicrobial therapy in pediatric oncology patients with febrile neutropenia Hodgson-Viden Heather [email protected] Paul E [email protected] Joan L [email protected] Department of Pediatrics Stollery Children's Hospital Edmonton, Alberta, Canada2005 18 5 2005 5 10 10 14 12 2004 18 5 2005 Copyright © 2005 Hodgson-Viden et al; licensee BioMed Central Ltd.2005Hodgson-Viden et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
There are no standard criteria for when to discontinue intravenous antimicrobial therapy (IVAMT) in children with febrile neutropenia (FN), but it is now common to discontinue IVAMT and discharge patients with an absolute neutrophil count (ANC) ≤ 500 /mm3. The purpose of this study was to evaluate the outcome of a large cohort of children with FN who had IVAMT discontinued with an ANC ≤ 500 /mm3
Methods
A retrospective chart review was completed of patients in the Northern Alberta Children's Cancer Program with FN and no apparent clinical source of fever from June 1, 1997 to July 1, 2002.
Results
Out of a total of 275 patients, 127 (46%) had at least one episode of FN, with FN occurring in patients with sarcomas more commonly than in those with leukemia/ lymphoma and least in those with other solid tumors. In 59 of 276 episodes of FN (21%) patients had a microbiologically defined infection at admission. Of the 217 remaining episodes, 112 of 199 patients (56%) with known neutrophil counts had IVAMT discontinued before their absolute neutrophil count (ANC) reached 500 /mm3 at the discretion of the clinician. Fever recurred in only two of these patients after discharge, and there were no bacterial infections diagnosed after parenteral antibiotics were discontinued.
Conclusion
Even without use of standard criteria for early discharge, clinicians appear to be skilled at selecting children with FN who can safely have IVAMT discontinued with an ANC ≤ 500 /mm3.
==== Body
Background
Febrile neutropenia (FN) is a common problem in pediatric oncology patients. The majority of patients with FN do not have a microbiologically defined infection (MDI), but those who do are at risk for overwhelming sepsis. It is therefore standard practice to start empiric intravenous (IV) antibiotics if the absolute neutrophil count (ANC) is below 500–1000/mm3 [1]. However, there are no definitive guidelines on the endpoints for stopping antibiotics if a bacterial source is not found for the fever, nor is it clear if the antibiotics need to be given by IV for the entire course. Physicians must balance the potential risk of inadequate antibiotic therapy with the potential risk of antibiotic resistance from overuse of antibiotics, while taking into consideration the emotional, social, and financial costs to hospitalized children and their families.
Management of FN in the Northern Alberta Children's Cancer Program (NACCP) has traditionally consisted of broad spectrum IVAMT (usually tobramycin and piperacillin) until the patient is afebrile, blood cultures are negative at ≥ 48 hours incubation, and the ANC is > 500/mm3. However, in recent years, there has been a trend towards discontinuing IVAMT in children who do not have an ANC > 500/mm3. Clinicians have used their clinical judgment rather than standard criteria in deciding when to discontinue IVAMT. The objective of this study was to document the clinical course and outcome of FN in the face of this changing management.
Methods
Patients
Children ≤ 17 years of age receiving therapy for a primary or recurrent malignancy through the NACCP from June 1, 1997 to July 1, 2002 were enrolled in this study. Inpatient charts were reviewed for these patients at the Stollery Children's Hospital in Edmonton to document the details of all FN episodes. Outpatient charts were also reviewed to determine if the fever recurred after hospital discharge. Fever was defined as a temperature ≥ 38.0 degrees Celsius (°C) measured either at home or in hospital. Neutropenia was defined as an ANC ≤ 500/mm3. Children with leukemia who were not yet in remission were excluded from this study as they usually have concomitant complications that prolong hospitalization beyond resolution of the FN.
Data collection
For each episode of FN, data was compiled from the time of diagnosis of FN until discharge from hospital or death. If cultures done during the first 48 hours of the admission were positive for an organism that would account for the fever, no further data was collected. Fever was not attributed to beta-hemolytic streptococci isolated from a throat swab, as the carrier state is common. Fever was not attributed to toxin producing Clostridium difficile in stool, as symptomatic or asymptomatic carriage is common in pediatric oncology patients, yet fever is expected only with severe disease. Organisms isolated from blood cultures were considered contaminants if common skin organisms were isolated from a single blood culture.
For all admissions for FN where an infectious etiology was not identified from cultures done within 48 hours of admission, data was collected on the choice and duration of IVAMT (including outpatient antibiotics), use of oral antibiotics for empiric treatment of FN, use of systemic anti-fungal agents, microbiologic laboratory results, absolute phagocytic count (APC), ANC, absolute monocyte count (AMC), length of stay in hospital attributable to FN, duration of fever and recurrence of fever during or after the admission. A recurrent fever was defined as a new fever that occurred after the patient had been afebrile for ≥ 24 hours, and before they had an ANC > 500/mm3.
Statistical analysis
Data was analyzed using SAS (Cary, NC; Version 8.2). Descriptive statistics were used to summarize the patient cohort.
The Health Research Ethics Board of the University of Alberta approved this study.
Results
Patients and episodes of FN
From June 1, 1997 to July 1, 2002, the NACCP treated 275 children. Table 1 is a summary of the number of episodes of FN by diagnosis. All but a small number of children with neuroblastoma and central nervous system tumors received chemotherapy. Of the 275 patients, 148 (54%) were never admitted for FN, and the other 127 patients had 276 admissions for FN. The highest incidence of admissions for FN was in patients with osteogenic and Ewing's sarcoma, with the lowest incidence being in patients with Wilms tumor and central nervous system tumors.
Table 1 Episodes of febrile neutropenia by diagnosis in 275 pediatric oncology patients.
Diagnosis Number of patients (%) Number of patients with FN episodes (%) Mean number of episodes per patient Range of number of episodes per patient
Osteogenic sarcoma 8 (3%) 7 (87%) 3.13 0–7
Ewings sarcoma 5 (2%) 4 (80%) 2.20 0–5
Rhabdomyosarcoma 15 (5%) 11 (79%) 1.93 0–6
Hepatoblastoma 9 (3%) 5 (56%) 1.22 0–5
Acute myeloid leukemia (AML) 16 (6%) 7 (44%) 1.00 0–5
Neuroblastoma 23 (8%) 11 (48%) 0.83 0–5
Non-Hodgkin's lymphoma 24 (9%) 14 (58%) 0.75 0–4
Acute lymphoblastic leukemia (ALL) 85 (31%) 41 (48%) 0.73 0–7
Others * 22 (8%) 9 (41%) 0.41 0–3
Hodgkin's lymphoma 25 (9%) 8 (32%) 0.36 0–5
Central nervous system tumors 26 (9%) 7 (27%) 0.23 0–3
Wilms tumor 17 (6%) 3 (18%) 0.12 0–1
Total 275 127 (46%)
Children with leukemia not in remission were excluded. All but a small number of children with neuroblastoma and central nervous system tumors received chemotherapy.
n = number of patients
* adrenocortical carcinoma, epithelial tumor, juvenile myelomonocytic leukemia, Langerhans cell histiocytosis, small cell tumor, teratoma
Microbiologically documented infections
Cultures taken within 48 hours of admission detected MDI that could account for the fever in 59 episodes of FN (21%) in 45 children. Sixteen of these episodes were attributed to viral infections (9 respiratory syncytial virus, 5 parainfluenza, 1 influenza A and 1 influenza B) and 44 episodes were attributed to bacterial infections (5 urinary tract infections, 33 episodes of bacteremia, and 5 other bacterial infections). There were no systemic fungal infections diagnosed in the 276 admissions for FN. Table 2 shows the incidence of infection according to age group.
Table 2 Incidence of proven bacterial or viral infections.
Type of infection Age 0–1 year
n = 40 Age 2–4 years
n = 91 Age ≥ 5 years
n = 145
Viral infection 2 (5%) 6 (7%) 7 (5%)
Bacterial infection 9 (23%) 17 (19%) 18 (12%)
Cultures negative 29 (73%) 68 (75%) 120 (83%)
The table shows the number of proven bacterial or viral infections diagnosed on cultures done within 48 hours of admission during 276 admissions for FN in pediatric oncology patients.
n = number of episodes of FN in this age range
Cultures taken more than 48 hours after admission detected a MDI in eight episodes of FN, but the organisms isolated may not have been the cause of the original fever. These infections were documented at a median of 6.5 days (range 3 – 29 days) after admission to hospital. There were four viral infections (influenza A, parainfluenza, herpes stomatitis (2 episodes)) and four episodes of bacteremia (Salmonella enteritidis, viridans streptococci, coagulase-negative staphylococci (2 episodes)).
Duration of fever
Fever was documented only prior to the hospitalization and not after admission in 72 of the 217 episodes of FN without proven bacterial or viral infection at admission. However, for the 145 episodes where fever was documented in hospital, the median duration of the initial fever was 2 days (range 1 – 33 days). There were 39 episodes of FN in which the fever recurred prior to hospital discharge and persisted for a median of 3 days (range 2 – 19 days).
Antimicrobial therapy
For two episodes of FN, IVAMT was not initiated, but the reason for this was not clear. For the remaining 215 episodes, the most frequently prescribed IV IVAMT was piperacillin/tobramycin (75% of episodes). The median duration of IVAMT was 5 days (range 1 to 28 days), and outpatient IVAMT were not used. Oral antibiotic therapy was used in 17 episodes prior to hospital discharge and in another six episodes at discharge, but always to treat a specific infection rather than as empiric therapy for FN.
In 16 episodes of FN, patients were on fluconazole prophylaxis at the time of admission. A systemic antifungal agent was added during the hospitalization in another 19 episodes (fluconazole in 15, amphotericin B in 3 and ketoconazole in 1). The anti-fungal was added a median of 2 days after admission (range 1–10 days).
Clinical course and outcome
The median length of stay in hospital for all FN episodes was 5 days (range 1 to 33 days). Almost all patients were discharged on the day their IVAMT was discontinued.
Phagocytic cell counts at the time of discontinuation of IVAMT are shown in Table 3 (not available in 77 episodes). The median ANC at the time of discontinuation of IVAMT was 400/mm3, and in 28/199 episodes (14%), IVAMT were discontinued before the ANC rose to 100/mm3. In 13/199 episodes (7%), IVAMT was discontinued when the absolute phagocyte count (APC) was < 100/mm3.
Table 3 Hematologic parameters when antibiotics were discontinued and at discharge in pediatric oncology patients with FN with negative cultures at admission.
Median Range Number (%) < 100 Number (%) 100–500 Number (%) > 500
ANC when antibiotics discontinued (n = 199) 400 0–34900 28 (14%) 84 (42%) 87 (44%)
Monocyte count when antibiotics discontinued (n = 199) 400 0–4400 20 (10%) 117 (59%) 62 (31%)
APC when antibiotics discontinued (n = 199) 800 0–39300 13 (7%) 56 (28%) 129 (65%)
ANC at discharge (n = 194) 500 0–42200
The table shows the hematologic parameters when antibiotics were discontinued and at discharge in pediatric oncology patients with FN with negative cultures at admission. There were 217 episodes of FN analyzed in the study, but not all patients had bloodwork done when antibiotics were discontinued or at discharge.
n = number of episodes of FN
ANC – absolute neutrophil count
APC – absolute phagocyte count (ANC + monocyte count)
Two patients died during their admission for FN. One patient was receiving palliative care. The other patient died from complications of graft versus host disease. None of the patients required readmission because of clinical deterioration, suspected infection or recurrent fever prior to their neutropenia resolving.
With four other episodes of FN, a fever recurred after the patient was discharged, but prior to their next chemotherapy treatment. With two of these four episodes, the patient was not neutropenic at the time of the fever and cultures were not taken. For the other two episodes, the patients were neutropenic when the fever recurred (ANC had been 0 and 200 at discharge), but all cultures were negative and the children were not readmitted.
Discussion
This study was a review of 276 admissions for FN in 275 pediatric oncology patients over a 5-year period. Almost half the patients had at least one hospital admission for FN, with the maximum number of admissions being seven. Admissions for FN occurred most commonly in patients with sarcomas, less so in those with leukemia or lymphoma and least in those with other solid tumors. However, episodes of FN in children with leukemia who were not yet in remission were excluded since FN in this clinical setting often occurs in patients that have other complicating factors.
Management of FN is a controversial issue. It is clear that empiric treatment of FN with IVAMT (with addition of antifungal agents if fever persists) decreases mortality [1]. This is because many of these patients have bacteremia, and septic shock is a common outcome in neutropenic patients who are bacteremic with virulent organisms such as coliforms, Staphylococcus aureus, and even some typically non-virulent organisms such as viridans streptococci. However, cultures are never 100% sensitive, not all body sites are accessible to culture, and it is possible that some infections in immunocompromised patients are due to organisms that are cell-wall deficient so are not detected in routine cultures [2]. A study using high-resolution computed tomography of the chest in adults with FN revealed a high incidence of pulmonary infiltrates that were not evident on plain films [3], so it is possible that at least some patients with FN have a bacterial pneumonia that is not diagnosed. Therefore, IV or oral antibiotics are often continued even if all cultures at presentation are negative. However, it is now clear that the majority of adults and children with FN, negative bacterial cultures, and without clinical evidence of bacterial infection do not have a partially-treated or occult bacterial infection, and IVAMT can be discontinued prior to the ANC reaching 500/mm3 [1].
In the current study, an MDI was identified by cultures on admission in 21% of episodes (16% bacterial, 6% viral), and based on cultures done more than 48 hours after admission in 2.8% of episodes (1.4% bacterial, 1.4% viral). There was no mortality and minimal morbidity attributable to infection in the other 208 hospital admissions for FN. Almost all episodes were managed with IVAMT initially, but in 56% of episodes, IVAMT was discontinued when the ANC was ≤ 500/mm3. Early discontinuation of IVAMT was based on the judgment of the clinician rather than on standard criteria. Fever recurred in only 4/112 (4%) of these episodes, and the fever was not due to bacterial infection and did not result in readmission in any of these four episodes. This suggests that the clinicians were able to select patients who could safely have IVAMT discontinued and be discharged when the ANC was ≤ 500/mm3.
The main potential risk of limiting the duration of IVAMT is recurrence of a partially treated bacterial infection. Benefits of limiting the duration of antibiotic therapy include decreasing the duration of hospitalization, decreasing the risk of antibiotic-related toxicity, and possibly decreasing the risk of secondary infection with organisms that are resistant to the antibiotics that were used for empiric treatment of FN. In the current study, there was no evidence of recurrence of undiagnosed partially treated bacterial infections after discontinuation of antibiotics. With regard to secondary bacterial infections, a bacterial infection was diagnosed based on cultures done more than 48 hours after admission in only 4/276 (1.4%) of admissions, and all four patients were still on IVAMT. In the only published pediatric study where 73 patients were randomized to continue or discontinue antibiotics before the ANC reached 500/mm3, the rate of documented or probable secondary bacterial infections was 8% in the group who continued vs. 6% in the group who discontinued antibiotics [4]. Clearly it would require a very large study to demonstrate if early discontinuation of antibiotics in selected patients can decrease the risk of secondary bacterial infection.
There were 21 pediatric studies up to 2002 looking at the outcomes of children with FN who had all antibiotics discontinued or were changed to oral antibiotics with ANC ≤ 500 /mm3 [5]. In general, outcomes were good if patients had evidence of marrow recovery at the time of discontinuation of IVAMT. However, there is a need for specific criteria in determining when to stop empiric IVAMT in patients with FN. Risk factors for bacteremia or other serious bacterial infections that have been validated and would be identifiable soon after admission include:
1) recent chemotherapy with a platelet count <50,000/mm3 [6]
2) relapsed leukemia [6]
3) hypotension [6]
4) CRP ≥ 90 mg/L [6]
5) absolute monocyte count (AMC) < 100 [7-9]
6) peak temperature ≥ 39°C [8,9]
Even these risk factors are not consistent between studies, which could be because the definitions of significant bacterial infections and the populations studied varied [8]. Other factors that have been reported to increase the risk of bacterial infection such as young age [5], high number of previous FN episodes [10], marrow involvement [11], presence of a central venous catheter [11], anticipated neutropenia > 7 days [5,12], presence of co-morbidities [12] or recent stem-cell transplant [5] require validation.
Risk factors for bacterial infections after discontinuation of IVAMT in children with FN are even less well defined. Most studies to date had insufficient power to look at the occurrence of bacterial infections after discontinuation of IVAMT, and used recurrence of fever as an endpoint [5]. As mentioned previously, there is a correlation between the absence of marrow recovery and adverse outcome after discontinuation of IVAMT [1,5], but it is not clear if the APC, the ANC, the AMC, the platelet count, or some combination of these parameters should be analyzed to determine that marrow recovery is occurring. In our study, none of the patients had bacterial infections after IVAMT was discontinued and only four patients with recurrent fever, so we were not able to analyze risk factors. Our study suggests that children selected for early discontinuation of IVAMT by experienced clinicians are likely to have an excellent outcome. However, further prospective studies are needed to determine whether using objective criteria (such as an AMC > 100) in addition to clinical judgment increases the number of children who can safely have their IVAMT discontinued while they have a low ANC.
If IVAMT is discontinued with an ANC ≤ 500/mm3, there is no consensus as to whether oral antibiotics should be substituted until the ANC reaches 500/mm3. All antibiotics were discontinued with an ANC ≤ 500/mm3in at least some patients in 10 of 21 previous pediatric studies, and oral antibiotics were substituted in the other 11 studies [5]. The outcomes appear to be comparable with these two strategies. The only randomized controlled trial of oral antibiotics versus placebo found no advantages in using antibiotics [4]. This would suggest that undiagnosed partially treated bacterial infection is not common in these patients.
Conclusion
About half of the pediatric oncology patients in our program had at least one admission for FN. Intravenous antibiotics were discontinued with an ANC ≤ 500/mm3 in about half of episodes of FN, with oral antibiotics then being prescribed only for patients with proven bacterial infections. The few secondary bacterial infections occurred prior to discontinuation of IVAMT, and no patients required readmission because of suspected or proven infection. Based on this and previous studies, it is not necessary to routinely prescribe oral antibiotics in children who are discharged with an ANC ≤ 500/mm3. A prospective randomized study could determine if use of objective criteria in addition to clinical judgment could increase the number of children who can safely have IVAMT discontinued while still neutropenic.
List of abbreviations
AMC – absolute monocyte count
ANC – absolute neutrophil count
APC – absolute phagocyte count
°C – degrees Celsius
FN – febrile neutropenia
IVAMT – intravenous antimicrobial therapy
MDI – microbiologically defined infection
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
HH collected all data and wrote part of the manuscript. PEG helped with study design and reviewed the manuscript. JLR helped with study design and wrote part of the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors would like to thank **EPICORE (Epidemiology Coordinating and Research) Centre at the University of Alberta for their valuable assistance with data entry and statistical analysis. Heather Hodgson was funded by the ACADRE network (Aboriginal Capacity and Development Research Environments).
==== Refs
Hughes WT Armstrong D Bodey GP Bow EJ Brown AE Calandra T Feld R Pizzo PA Rolston KV Shenep JL Young LS 2002 guidelines for the use of antimicrobial agents in neutropenic patients with cancer Clin Infect Dis 2002 34 730 751 11850858 10.1086/339215
Woo PC Wong SS Lum PN Hui W Yuen K Cell-wall-deficient bacteria and culture-negative febrile episodes in bone-marrow-transplant recipients Lancet 2001 357 675 679 11247551 10.1016/S0140-6736(00)04131-3
Heussel CP Kauczor HU Heussel GE Fischer B Begrich M Mildenberger P Thelen M Pneumonia in febrile neutropenic patients and in bone marrow and blood stem cell transplant recipients: use of high-resolution computer tomography J Clin Oncol 1999 17 796 805 10071269
Klaassen RJ Allen U Doyle JJ Randomized placebo-controlled trial of oral antibiotics in pediatric oncology patients at low-risk with fever and neutropenia J Pediatr Hematol Oncol 2000 22 405 411 11037850 10.1097/00043426-200009000-00004
Orudjev E Lange BJ Evolving concepts of management of febrile neutropenia in children with cancer Med Pediatr Oncol 2002 39 77 85 12116054 10.1002/mpo.10073
Santolaya ME Alvarez AM Aviles CL Becker A Cofre J Enriquez N O'Ryan M Paya E Salgado C Silva P Tordecilla J Varas M Villarroel M Viviani T Zubieta M Prospective evaluation of a model of pediction of invasive bacterial infection risk among children with cancer, fever, and neutropenia Clin Infect Dis 2002 35 678 683 12203164 10.1086/342064
Klaassen RJ Goodman TR Pham B Doyle JJ "Low-risk" prediction rule for pediatric oncology patients presenting with fever and neutropenia J Clin Oncol 2000 18 1012 1019 10694551
Rackoff W Breitfeld P Risk factors in children with fever and neutropenia J Clin Oncol 2001 19 4270 4271 11709573
Madsen K Rosenman M Hui S Breitfeld PP Value of electronic data for model validation and refinement: bacteremia risk in children with fever and neutropenia J Pediatr Hematol Oncol 2002 24 256 262 11972092 10.1097/00043426-200205000-00008
Ammann RA Hirt A Luthy AR Aebi C Predicting bacteremia in children with fever and chemotherapy-induced neutropenia Pediatr Infect Dis J 2004 23 61 65 14743049
Ammann RA Hirt A Luthy AR Aebi C Identification of children presenting with fever in chemotherapy-induced neutropenia at low risk for severe bacterial infection Med Pediatr Oncol 2003 41 436 443 14515382 10.1002/mpo.10320
Alexander SW Wade KC Hibberd PL Parsons SK Evaluation of risk prediction criteria for episodes of febrile neutropenia in children with cancer J Pediatr Hematol Oncol 2002 24 38 42 11902738 10.1097/00043426-200201000-00011
| 15904510 | PMC1156908 | CC BY | 2021-01-04 16:31:06 | no | BMC Pediatr. 2005 May 18; 5:10 | utf-8 | BMC Pediatr | 2,005 | 10.1186/1471-2431-5-10 | oa_comm |
==== Front
BMC PediatrBMC Pediatrics1471-2431BioMed Central London 1471-2431-5-141591670110.1186/1471-2431-5-14Research ArticleBehaviour sequelae following acute Kawasaki disease Carlton-Conway Daniel [email protected] Raju [email protected] Lucy [email protected] Colin [email protected] Louise [email protected] Robert [email protected] Department of Paediatric Cardiology, Royal Hospital for Children, Upper Maudlin Street, Bristol, BS2 8BJ, UK2 Department of Paediatric Cardiology, Guy's and St Thomas' Hospital Trust, Guy's Hospital, St Thomas' Street, London, SE1 9RT, UK3 Department of Paediatrics, Ealing Hospital, Uxbridge Road, Southall, Middlesex, UB1 3EW, UK4 Institute of Psychiatry, de Crespigny Park, London, SE5 8AF, UK2005 25 5 2005 5 14 14 5 11 2004 25 5 2005 Copyright © 2005 Carlton-Conway et al; licensee BioMed Central Ltd.2005Carlton-Conway et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Kawasaki disease is a systemic vasculitis and may affect cerebral function acutely. The aim of the present study was to measure a number of behaviour and social parameters within a cohort of Kawasaki disease patients.
Methods
Parents of children with past diagnosis of Kawasaki disease were recruited to complete several behaviour screening questionnaires. Sixty five sets of questionnaires relating to the patient cohort received were eligible for inclusion. Two control groups were used, a hospital (HC) control and a sibling control (SC) group.
Results
40% of the Kawasaki disease group showed elevated internalising scores in the clinical or borderline-clinical range. This compared with 18% of hospital controls and 13% of sibling controls. Additionally, the Kawasaki disease (KD) group were shown to be experiencing greater overall total difficulties when compared with the controls (KD 13.7, HC 8.6, SC 8.9). The KD group attained higher behavioural scores within the internalising sub-categories of somatic problems (KD 61, HC 57, SC 54) and withdrawn traits (KD 56, HC 53, SC 51). The KD group were also shown to be suffering more thought problems (KD 57, HC 53, SC 50) compared with the controls. Further difficulties relating to conduct (KD 3.3, HC 1.4) and social interactions (KD 6.7, HC 8.3) are also highlighted for the KD group compared with hospital controls. Positron emission tomograms were performed on nine patients to investigate severe behavioural problems. Three showed minor changes, possibly a resolving cerebral vasculopathy.
Conclusion
Kawasaki disease can be associated with significant behavioural sequelae. This is an important consideration in the long-term follow up and referral to a clinical psychologist may be necessary in selected patients.
Kawasaki DiseaseCerebral VasculitisPsychological DifficultiesLong Term Management
==== Body
Background
In 1967, Tomisaku Kawasaki first described an illness in children called 'acute febrile mucocutaneous lymph node syndrome' [1]. It has now become more commonly known as Kawasaki disease. The diagnostic criteria according to the Centres for Disease Control (Atlanta) are laid out below. Fever of 5 days plus 4 out of the 5 remaining criteria [2], or fever plus coronary artery aneurysms at echocardiography and 3 additional criteria are generally required for a diagnosis of Kawasaki disease to be made [3] (see table 1).
Table 1 Criteria necessary to confirm diagnosis of Kawasaki disease [2]
Criteria Description
Fever 5 days or more
AND
1. Conjunctivitis Bilateral
2. Mucous membrane changes of oropharynx Red cracked lips; strawberry tongue; or diffuse erythema
3. Changes of peripheral extremities Initially: erythema and oedema of palms and soles Convalescent stage: desquamation
4. Rash Polymorphous, non vesicular
5. Lymphadenopathy Cervical
Cardiac complications can occur which include coronary artery aneurysms, which have been documented as developing during the illness in 25% of patients treated with aspirin alone (determined at echocardiography) [3,4]. Treatment with intravenous gammaglobulin and aspirin has been shown to reduce this incidence considerably [5,6].
Pathologically, Kawasaki disease is a multisystem vasculitis affecting small and medium sized arteries. Inflammation is progressive, initially mild affecting only the subendothelium and progressing to a severe panarteritis of the coronary arteries. Injured arteries are weakened and thus aneurysms may form as a result [3].
Neurological and psychological complications associated with Kawasaki disease have also been noted, although reports in this area are limited. Neurological complications include aseptic meningitis occurring in 26–50% of cases as well as facial nerve palsy, sensorineural hearing loss, hemiplegia, cerebral infarction and severe lethargy, which have all been reported in various case studies [3-6]. Amano and Hazama have examined histological specimens from 30 Kawasaki disease patients and found evidence of the occurrence of endoarteritis and periarteritis within the brain [3] and it is likely that many of the CNS complications are secondary to such a process of cerebral vasculitis [8].
Extreme irritability is a well documented observation seen in children during the acute phase of the illness, which generally resolves after treatment. However, a recent study performed by King and workers have reported that behavioural problems may develop or persist once the acute phase of the illness has passed and then continue for months or years after the initial illness [7]. The study consisted of 32 patients with Kawasaki disease in the USA and used siblings as controls. They demonstrated that post acute Kawasaki disease patients had a significantly greater tendency to internalising and attentional behavioural problems, although no firm aetiology was uncovered [7]. However, this study was somewhat limited as it was carried out on a relatively small cohort and did not include a control group of patients that had been hospitalised for a similar time period.
It is these behavioural sequelae that we are interested in as we have documented many cases in clinic of reported behavioural problems persisting for months and years after the initial episode of Kawasaki disease.
We initially set out to screen for potential behavioural problems within a British patient cohort. 123 children with a past diagnosis of Kawasaki disease reviewed in a long term Kawasaki disease follow up clinic were included. Parents were asked at time of their visit to answer a series of questions relating to behaviour and certain aspects of their child's disease. Additionally, 142 questionnaires used by a national Kawasaki disease parent support group were reviewed retrospectively. All the data was collected between 1996–1999, with 265 children being reviewed in total. The questions in clinic and within the help group questionnaire were derived from Rutter's B questionnaire, with the aim to identify significant changes in physical and psychological health as perceived by the child's carer [3]. The results demonstrated that 90 of the children (34%) suffered from behavioural changes, lasting for more than one year following their recovery from the acute stages of Kawasaki disease. These behavioural problems were significant enough to lead to consultation with an educational psychologist, clinical psychologist or a GP. The behavioural problems recognised in the data included hyperactivity, decreased concentration, increased aggression and emotional lability.
Based on these initial results, we therefore set out to investigate these behavioural difficulties in a more detailed and quantitative fashion within a British Kawasaki disease patient cohort.
There is nothing in the current management protocol that is aimed at addressing behavioural difficulties, nor in the advice given to parents. Finding a significant correlation between Kawasaki disease and behavioural sequelae would provide the basis for an important added dimension in the long-term management of the disease. We would hope to provide a greater insight and awareness into behavioural sequelae, which would lead to earlier referrals to a child psychologist where necessary, more specific advice to the parents on how best to manage their child as well as providing reassurance that the child's behavioural difficulties may be within the normal sequelae of Kawasaki disease.
Methods
A retrospective cohort study was selected as the most appropriate method in view of the uncommon frequency of Kawasaki disease within the UK population. Parents of children who had been seen in clinic between 1995 – 2001 were contacted by telephone and informed verbal consent gained. When parents could not be contacted by phone, a detailed covering letter explained the nature of the research was sent out and parents were invited to respond. The study was approved by the Local Ethics Committee.
The Kawasaki disease patient group (KD group)
Inclusion criteria for this stage of the research were: (a) all patients aged under 18 years at time of diagnosis and at the time of the study were between 3 to 18 years old; (b) A diagnosis of KD formally having been reached according to the criteria in table 1; (c) sufficient level of English spoken by parent/carer to understand questionnaires; and (d) no history of CNS dysfunction unrelated to KD, or any other history of severe illness.
112 sets of questionnaires were sent out to parents. 77 questionnaires were received and 65 were suitable for inclusion according to our predetermined criteria.
Controls
Two control groups with different characteristics were used. The first consisted of previous hospital inpatients who had stayed in hospital for a short period and undergone a cardiac catheterisation, with follow-up at regular intervals. This group was chosen in order to control for the effects of a hospital stay on child behaviour. Patients who had subsequently progressed to surgery, or suffered from an unrelated CNS disorder or other chronic illness were not deemed suitable for inclusion. These controls were matched to the KD group, as best as possible, according to age and sex. 47 sets of questionnaires were sent out and 23 were returned completed.
The second control group consisted of siblings of the KD group matched for age and sex, as best as possible. This control group was selected in order to control for the influence of genetics and environment on child behaviour. 31 sets of questionnaires were sent out and 17 were returned.
Details of the groups are included in Table 1.
Data collection instruments
Mothers of patients were sent out a series of psychological screening questionnaires for behavioural difficulties, which they were invited to complete. The questionnaire set comprised: the Child Behaviour Check List – CBCL/4–18 (Achenbach 1991) [13]; Strengths & Difficulties Questionnaire (Goodman 1997) [14], and the Parenting Stress Index – PSI, (Abidin 1983) [15]. A general information sheet was also devised for inclusion with the questionnaires.
The Child Behaviour Check List (CBCL) tests the parental reports of a child's competencies and quantifies behavioural and emotional problems along predetermined categories. It has a one week test-retest reliability of r = 0.93 [13]. The CBCL was used to obtain standardised reports from parents on behavioural and emotional problems for their children. The CBCL contains 120 problems items and measures in two broad categories: Internalising problems (consisting of the scales: anxious/depressed, somatic complaints and withdrawn behaviour) and Externalising complaints (aggressive, and delinquent behaviour). Additional scales of social problems, thought problems and attention problems are also included.
The parenting stress index (PSI) short form is designed to measure stresses occurring within the parent-child relationship and provides insight as to whether any difficulties originate from the parent, the child or their interaction. It uses the following categories; Total stress, Parental distress, Parent child dysfunctional interaction, and Difficult child. Also, scores for defensive responding were recorded.
The Strengths and Difficulties Questionnaire (SDQ), looks at behavioural, emotional and/or relationship difficulties in children as well as a positive category that rates social strengths and interactions. The SDQ provides scores across 5 subscales; conduct, hyperactivity, emotional, peer, and prosocial problems. It also generates a total difficulty score.
Statistical analysis
One-way analyses of variance (ANOVA) were conducted on each measure, with Bonferoni post hoc testing to examine the differences between the groups. Fisher's Exact Test was also employed to test the strength of a trend within the groups. Means for each measure and standard errors are given in Table 2.
Table 2 Comparison between patient cohort and controls
KD patient Group Hospital Controls Sibling Controls
Number in group 65 23 17
Male : Female 1.6:1 0.9:1 0.9:1
Age on completion (mean, range) 7.38 (3–18) 8.95 (3–17) 8.68 (5–14)
Patient age at presentation with Kawasaki disease (median, range) 3.60 (0–14) N/A N/A
Number with coronary artery dilatations or aneurysms secondary to Kawasaki disease 13 (20%) N/A N/A
Reported difficulties (%) 56.3% 35.3% 46.2%
Results
When asked by a single question, 56% of the Kawasaki disease patient cohort were reported to be experiencing minor to severe difficulties: emotionally, behaviourally or socially. This compared with 35% of hospital controls and 46% of sibling controls. Looking at the scores on the Child Behaviour Check List (CBCL), the KD group consistently achieved higher mean behavioural scores than the two control groups within all categories. (Table 2).
Significantly higher behavioural scores were attained by the KD group within the withdrawn (P = 0.045) and somatic (P = 0.030) categories, when compared with the sibling controls (Table 3). This suggests a greater tendency towards internalising behaviour within the KD group. Indeed, 40% of the KD group fell within the clinical or borderline range for internalising problems, compared with 18.2% of the hospital controls and 12.5% of the sibling controls (FE = 0.040). No significant differences between the groups were found for externalising problems. The KD group was found to have significantly higher behavioural scores on the thought problem category of the CBCL compared with the sibling control group (P = 0.006). Thought problems reported included obsessions, compulsions as well as strange behaviour. Results from the social and schooling categories of the CBCL did not suggest any significant differences between the three groups.
Table 3 Selected results, including all that were significant, from the scores to the questionnaires. FE = Fisher's exact test; SEM = Standard Error of Mean; † indicates P values less than 0.05 * indicates which control groups differed significantly (P < 0.05) from the KD group (see text for p values)
Category (mean, SEM) KD Group Hospital Controls Sibling controls Significance (ANOVA unless indicated)
CBCL/4–18
Internalising score in clinical or borderline range 40.0% 18.2% 12.5% 0.04 † (FE)
Externalising score in clinical or borderline range 26.2% 9.1% 25% 0.27 (FE)
Thought Problems 57.0 (1.06) 52.8 (1.19) 50.2 (1.74) * 0.03 †
Withdrawn 55.9 (0.96) 52.8 (1.19) 51.2 (0.82) * 0.02 †
Somatic 61.0 (1.27) 57.3 (1.63) 54.1 (1.81) * 0.02 †
Sex Problems 54.0 (1.10) 52.9 (1.57) 47.3 (1.28) * 0.01 †
Aggressive 56.5 (1.19) 52.0 (0.71) 52.6 (2.07) 0.05
Internal T score 53.4 (2.0) 48.5 (2.38) 44.1 (2.81) 0.05
Total T score 53.4 (1.73) 48.4 (2.45) 45.6 (3.36) 0.07
Strengths + Difficulties
Prosocial 6.7 (0.31) 8.3 (0.48) * 7.4 (0.63) 0.04 †
Conduct 3.3 (0.38) 1.4 (0.34) * 2.8 (0.73) 0.04 †
Total Difficulties 13.7 (1.08) 8.6 (1.44) * 8.9 (1.93) 0.02 †
Parenting Stress Index
Difficult child (raw) 30.3 (1.52) 24.2 (1.47) 26.5 (2.80) 0.07
Scores on the Strengths and Difficulties Questionnaire showed the KD group achieving higher mean scores than both control groups within the four negative categories of conduct, peer, emotional and hyperactivity. Along the positive prosocial category (rating social strengths and interactions) the KD group achieved a lower mean score than both the control groups. The KD group were significantly less helpful and considerate (prosocial category) than the hospital control group (P = 0.045). The KD group also experienced significantly more problems associated with their conduct such as temper tantrums, disobedience and argumentative behaviour when compared with the hospital control group (P = 0.030). Greater total behavioural difficulties experienced within the KD group when compared to the hospital controls were also demonstrated (P = 0.048).
On the Parental Stress Index, the KD group achieved greater mean scores compared to the control groups. However, the differences fell short of significance. There may be a trend here towards a 'difficult' child being the main source of stress within the parent-child relationship (P = 0.084), but further research is required to confirm this.
No significant differences were shown between the hospital control and the sibling control groups within any of the questionnaires.
13 patients within the KD group were documented as suffering or having suffered from coronary artery dilatation or aneurysms secondary to their Kawasaki disease. No significant differences or trends to significance were shown when comparing the mean behavioural scores of the KD group with dilatations/aneurysms and the KD group without dilatations/aneurysms.
Furthermore, no significant differences or values approaching significance were demonstrated when comparing age of onset of Kawasaki disease with mean behavioural scores.
Discussion
Our initial indications that a significant number of children experience behavioural sequelae in the short to medium-term post Kawasaki disease was confirmed by more detailed investigation. The KD group showed higher scores on measures of behavioural and emotional difficulties when compared with both control groups, with the emergence of significant differences demonstrated between the KD group and the controls across various categories of the questionnaires. The CBCL showed these significant differences to exist between the KD group and the sibling control group, whereas the SDQ showed significant differences between the KD and the hospital control group.
Like King et al. (2000) [7], we have shown a strong preponderance towards problems of an internalising nature within the KD group. 40% of the KD group were shown to have internalising scores in the clinical or borderline-clinical range, representing an incidence greater than twice that of the control groups. In line with this, significantly higher scores were attained by the KD group compared to sibling controls within the internalising sub-categories of somatic problems and withdrawn traits. However, unlike King et al., our results also highlight further difficulties relating to thought and conduct existing within the KD group.
Importantly for parents, no significant differences were demonstrated between the groups with regard to participation in activities and hobbies, social interaction or performance at school. This, hopefully could form the basis of some reassuring parental advice.
However, despite the lack of any proven link between adverse effects on the child's overall social and academic performance, the high incidence of clinical and borderline-clinical internalising behaviour should make the paediatrician aware of the potential need for referral to a clinical psychologist during long term management. On account of the fact that internalising behaviour can be harder to detect than overt externalising traits, some simple screening questions into the child's behaviour should be posed to the parents during long-term follow up of Kawasaki disease.
This study has a few potential sources of bias. The relatively low return rate of the questionnaires raises the possibility of self selection bias. Bias originating from the parent answers to the questionnaires may also exist. Kawasaki disease is a relative rare disease with the potential for serious cardiac complications. The associated fears and anxieties likely to be experienced by the parents may serve to exaggerate any observed behavioural difficulties in the minds of the parents. This would be reflected in higher than expected behavioural scores. However, the CBCL has been argued to mitigate this bias and moreover, Green et al (1998) suggest that the CBCL is in fact intrinsically conservative in discriminating mild emotional and behavioural problems [16]. The relatively small size of the groups, especially the controls, may have limited the extent of the significant behavioural and emotional problems seen, and we would therefore expect to see more significant results emerge should the size of the groups be expanded.
Pathological associations
Kawasaki disease is a systemic vasculitis which seems to have a predilection to coronary arteries, although it has also been shown to affect the cerebral arteries. Evidence suggesting the nature of the consequences arising from the cerebral vasculopathy has come from a single-photon emission computed-tomography study carried out on 21 children in Japan by Ichiyama and Nishikawa. They were able to demonstrate localised cerebral hypoperfusion occurring in 29% of the patients during the acute phase of Kawasaki disease [17].
Systemic lupus erythematous is also a multi-organ systemic vasculitis, which is known to affect the cerebral nervous system (CNS), causing various neuropsychiatric symptoms [18,19]. Multiple areas of hypoperfusion have been shown on SPECT studies of lupus patients positive for neuropsychiatric symptoms [20]. Interestingly, one study of 35 patients by Colamussi et al. (1995) found that in the absence of neuropsychiatric symptoms, uptake was normal in nearly all of the patients, whereas hypoperfusion was demonstrated in most patients with severe neuropsychiatric sequelae [21].
It is suggested that these areas of hypoperfusion documented within lupus and Kawasaki disease may be caused by a process of cerebral vasculitis resulting in patchy ischaemic areas. The damage to the CNS which is likely to arise may then alter neurological function for some time after the acute phase of the illness, and this would certainly provide one reasonable explanation for the behavioural difficulties that have been shown to persist in the medium to long-term.
Indeed, supporting evidence comes from Suzuki et al, (2000) who suggested that the remodelling of coronary aneurysms is a process that continues for many years after the onset of the disease [22]. One might surmise that continued behavioural problems may be the result of a parallel remodelling process occurring in the brain.
Thus, the evidence of CNS pathology adds further weight to our belief that the behavioural changes arise secondary to a cerebral vasculopathy, and are not merely due to the psychological complications of an acute severe illness.
Further experimental evidence
In order to further investigate any identifiable CNS pathology affecting the brain, active or otherwise, we have performed PET (glucose and ammonium) and MRI scans on 7 post Kawasaki disease patients with severe behavioural difficulties (Table 4). In all seven cases these were carried out at least 6 months after their last episode of Kawasaki disease. Three of these patients were reported to have slightly reduced uptake of tracer, occurring within the occipital lobes, although other areas were also affected in two of the cases. Although these minimal changes are of uncertain significance, they could represent resolving areas of cerebral hypoperfusion secondary to cerebral vasculitis that occurred during the acute phase of the disease. An ethically approved study with a larger patient base would be the next step to investigate these changes further.
Table 4 Cerebral PET scan results following acute Kawasaki disease
No. Year of Kawasaki disease PET scan result Date of PET scan
1 1997 Diffusely reduced uptake in occipital lobes and cerebellum 24/6/99
2 1999 Normal 20/9/00
3 1999 Asymmetry in occipital region lower on R side 20/3/00
4 1995 Normal 27/5/99
5 1997 Normal 20/5/99
6 1999 Slight reduction in posterior parietal and occipital lobes symmetrically. 28/10/99
7 1997 Normal 7/7/99
Summary
In conclusion, we have shown significant behavioural sequelae can occur following acute Kawasaki disease. Therefore, the paediatrician should bear in mind the potential need for long term follow up of Kawasaki disease patients in clinic even in the absence of coronary artery aneurysms. Parents can be told that behavioural difficulties experienced may be within the normal sequelae of the disease process and they can be reassured that they do not appear to affect school performance to any significant degree. However, a few probing questions should be asked to flesh out other possible causes before any difficulties are attributed to Kawasaki disease. The paediatrician should consider referral to a clinical psychologist where necessary. This may indeed prove to be a common requirement in long-term management, when it is considered that 40% of the KD group in this study fell within the clinical or borderline-clinical range for internalising problems.
Future research should examine the behavioural difficulties in greater detail, employing formal neuropsychiatric testing. Additionally, further research in this area is required to rule out other possible causes of behavioural sequelae. In particular, standard treatment with aspirin and intravenous gammaglobulin should be eliminated as potential aetiological factors.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We are very grateful to the Kawasaki disease support group for their assistance with the funding of this study. We would also like to warmly thank the families of patients who have helped with this study, as well as their clinicians for referring patients to the Kawasaki Clinic.
==== Refs
Kawasaki T Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desquamation of the fingers and toes in children (Japanese) Japanese Journal of Allergy 1967 16 178 222 6062087
Centres for Disease Control Kawasaki Disease - New York MMWR 1980 29 61 3
Dhillon R Newton L Rudd PT Hall SM Management of Kawasaki disease in the British Isles Arch Dis Child 1993 69 631 638 8285773
Kato H Sugimura T Akagi T Sato N Hashino K Maeno Y Kazue T Long-term Consequences of Kawasaki Disease. A 10-to 21-Year Follow-up Study of 594 Patients Circulation 1996 94 1379 1385 8822996
Levin M Tizard E Dillon M Kawasaki disease: recent advances Arch Dis Child 1991 66 1369 74 1776877
Rowley AH Gonzalez-Crussi J Shulman ST Kawasaki Syndrome Rev of Infectious Diseases 1998 10 1 15 3281216
King W Schlieper A Birdi N The Effect of Kawasaki Disease on Cognition and Behaviour Archives of Pediatric and Adolescent Medicine 2000 154 463 468
Poon L Lun K Ng Y Facial Nerve Palsy and Kawasaki Disease HKMJ 2000 6 224 226 10895149
McDonald D Buttery J Pike M Neurological Complications of Kawasaki Disease Arch Dis Child 1998 79 200 9797614
Bushara K Wilson A Rust R Facial Palsy in Kawasaki Syndrome Pediatric Neurology 1997 17 362 364 9436804 10.1016/S0887-8994(97)00103-3
Amano S Hazama F Neural involvement in Kawasaki disease Acta Pathol Jpn 1980 30 365 73 7395511
Michie CA Tulloh R Mills G Kawasaki syndrome may be associated with behavioural and learning disabilitie Ped Res 1998 43 14A
Achenbach TM Manual for Child Behaviour Check List/4–18 and the 1991 profile Burlington, University of Vermont, Dept of Psychiatry 1991
Goodman R The strengths and difficulties questionnaire J Child Psychol Psychiatry 1997 38 581 6 9255702
Abidin RR Parenting stress and the utilization of pediatric services Child Health care 1983 11 70 3 10262147
Green M Foster AF Morris MK Parent assessment of psychological functioning following paediatric acquired brain injury Journal of Paediatric Psychology 1998 23 289 99
Ichiyama T Nishikawa M Hayashi T Cerebral Hypoperfusion During Acute Kawasaki Disease Stroke 1998 29 1320 1321 9660380
Steinlin M Blaser S Gilday D Neurological manifestations of paediatric systemic lupus erythematous Pediatric Neurology 1995 13 191 7 8554655 10.1016/0887-8994(95)00110-2
Iverson G Anderson K The etiology of psychiatric symptoms in patients with systemic lupus erythematous Scand J Rheumatol 1994 23 277 82 7973483
Rubbart A Marienhagen J Pirner K Single-photon-emission computed tomography analysis of cerebral blood flow in the evaluation of central nervous system involvement in patients with systemic lupus erythematous Arthritis and Rheumatism 1993 36 1193 5 8216412
Colamussi P Giganti M Cittanti C Brain single-photon emission tomography with 99mTc-HMPAO in neuropsychiatric systemic lupus erythematous: relations with EEG and MRI findings and clinical manifestations European Journal of Nuclear Medicine 1995 22 17 24 7698150 10.1007/BF00997243
Suzuki A Miyagawa-Tomita S Komatsu K Active Remodelling of the coronary arterial lesions in the late phase of Kawasaki disease Circulation 2000 101 2935 41 10869266
| 15916701 | PMC1156909 | CC BY | 2021-01-04 16:31:06 | no | BMC Pediatr. 2005 May 25; 5:14 | utf-8 | BMC Pediatr | 2,005 | 10.1186/1471-2431-5-14 | oa_comm |
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-441588246510.1186/1471-2458-5-44Research ArticleMental health literacy in an educational elite – an online survey among university students Lauber Christoph [email protected] Vladeta [email protected] Nadja [email protected] Niklaus [email protected]össler Wulf [email protected] Psychiatric University Hospital, Zurich, Switzerland2005 9 5 2005 5 44 44 7 12 2004 9 5 2005 Copyright © 2005 Lauber et al; licensee BioMed Central Ltd.2005Lauber et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Mental health literacy is a prerequisite for early recognition and intervention in mental disorders. The aims of this paper are to determine whether a sample of university students recognise different symptoms of depression and schizophrenia and to reveal factors influencing correct recognition.
Methods
Bivariate and correspondence analyses of the results from an online survey among university students (n = 225).
Results
Most participants recognised the specific symptoms of depression. The symptoms of schizophrenia were acknowledged to a lower extent. Delusions of control and hallucinations of taste were not identified as symptoms of schizophrenia. Repeated revival of a trauma for depression and split personality for schizophrenia were frequently mistaken as symptoms of the respective disorders. Bivariate analyses demonstrated that previous interest in and a side job related to mental disorders, as well as previous personal treatment experience had a positive influence on symptom recognition. The correspondence analysis showed that male students of natural science, economics and philosophy are illiterate in recognising the symptoms depression and schizophrenia.
Conclusion
Among the educational elite, a wide variability in mental health literacy was found. Therefore, it's important for public mental health interventions to focus on the different recognition rates in depression and schizophrenia. Possibilities for contact must be arranged according to interest and activity (e.g., at work). In order to improve mental health literacy, finally, education and/or internship should be integrated in high school or apprenticeship curricula. Special emphasis must be given towards the effects of gender and stereotypes held about mental illnesses.
==== Body
Background
Mental health literacy is defined as the ability to gain access to, understand and use information in ways that promote and maintain good mental health. It refers to knowledge and beliefs about mental disorders that aid their recognition, management or prevention. It also includes the ability to recognise specific disorders, know risk factors and causes, know self-treatments and available professional help, and it is an attitude that promotes recognition and appropriate help-seeking [1]. The concept of mental health literacy implies that it is crucial to increase the public's knowledge about mental health and mental diseases since it is a prerequisite for early recognition and intervention in mental disorders.
The lifetime risk of developing any mental disorder is nearly 50% [2]. Thus, during a lifetime almost everybody has direct contact with an affected person. Recognition of mental disorders is essential as it influences a person's attitude and behaviour towards those affected [3]. Moreover, mental health literacy is an important determinant of help-seeking behaviour. Less knowledge about mental illness and its symptoms and possible treatment approaches are negatively associated with health care use [4]. In addition, mental disorders account for a considerable economic and emotional burden. Five out of ten leading illnesses associated with disease burden are psychiatric disorders. WHO estimates that in 2020 major depression will become the second most leading cause of disease burden at all [5]. Mental disorders are generally life shortening [6]. As a consequence, mental health literacy has gained increased attention within the last few years [7-20].
Recent studies have shown that mental health literacy is low, regardless of the population considered [21-25]. Although people distinguish abnormal from normal behaviour at a relatively satisfactory level, the recognition of a particular diagnosis is poor. However, in everyday life, people complain about specific symptoms. So far, studies have assessed mental health literacy exclusively by means of vignettes, i.e. short stories depicting a person with clinical symptoms of a particular diagnosis, mostly depression or schizophrenia [10,24,26-28]. Although this method allows one to draw conclusions about the recognition of a disorder in general, no statement about the recognition of the specific symptoms is possible. Thus, we investigated in a sample of university students the recognition of different symptoms of depression and schizophrenia. The following questions were used to assess the students' recognition:
1. What is the recognition rate of different symptoms of depression and schizophrenia in a sample of university students?
2. What are determining factors for recognition? Does faculty affiliation have an impact?
3. What conclusions can be drawn from these results for mental health policy?
Methods
Sampling and sample
This study is based on an online questionnaire completed by students of the University of Zurich/Switzerland. The baseline population included 2014 students in their 9th or 10th University semester (cut-off date: November 21st, 2003). The study sample was taken from a sub-population of 1228 students (61%) who had given permission that their e-mail address can be given out for research purposes. Students were then invited to participate by e-mail. Approval was granted by the University Legal Department. The anonymity of the participants was guaranteed.
The personal form of address was used. One of the authors (N.F.) introduced herself and gave preliminary information about the anonymity of the questionnaire. The test persons were assured they would receive feedback after the study terminated. Additionally, each session was introduced by questions from previous log-ins to control for repeated completion.
A pre-test with 34 students was preformed. Afterwards, the questionnaire was slightly modified with respect to the level of knowledge required and the wording of the questions. There was no criticism regarding the layout or the length of the survey. None of the test persons complained about the format of the questionnaire or the time it took to complete the study.
241 students completed the online-questionnaire (19.6% of the e-mail sub-population), three quarters within the first two days after receiving the invitation. 225 questionnaires were included (18.3% of the e-mail sub-population; 16 questionnaires were excluded due to missing data). The sample showed no statistically significant differences to the baseline population and the e-mail subpopulation regarding sex, age, faculty and main subject (Table 1). However, it cannot be excluded that knowledge differences exist between the two groups, for example their experience with mental disorders and the respective treatments.
Table 1 Demographic and study subject characteristics of the sample (N = 222–225)1
Semester population e-mail sample Study sample
Sex 2012 1294 222
Women 1038 (51.6%) 651 (50.3%) 105 (52.7%)
Men 974 (48.4%) 643 (49.7%) 117 (47.3%)
Age 2012 1294 224
Mean (years) 27.14 (± 5.19) 27.01 (± 5.10) 26.41 (± 5.07)
Faculty 2012 1294 225
Law 377 (18.7%) 261 (20.6%) 37 (16.4%)
Economics 339 (16.8%) 254 (20.0%) 46 (20.4%)
Medical school 224 (11.1%) 137 (10.8%) 15 (6.7%)
Natural sciences2 182 (9.0%) 105 (8.3%) 18 (8.0%)
Philosophy / arts3 731 (36.3%) 429 (34.1%) 109 (36.0%)
Psychology 159 (7.9%) 108 (8.3%) 28 (12.4%)
1 Varying sample sizes due to missing values
2 6 students in veterinary medicine included
3 1 student in theology included
The questionnaire
The self-constructed questionnaire is comprised of the following parts:
1. Presentation of the survey aims and instructions for further proceedings.
2. Assessment of demographic factors (age, sex, faculty affiliation), personal experience with mental disorders and previous contact to people with mental illnesses (5 questions).
3. Assessment of knowledge about depression and schizophrenia:
• Prompting of 10 symptoms where 5 are part of the diagnostic criteria of the respective disease (according to the ICD-10 diagnostic criteria) and 5 are not (see Tables 2 and 3). Answer categories for each symptom according to the ICD-10 categorisation: main symptoms, additional symptoms and no symptoms of the respective disorder.
Table 2 Distribution of the interviewees who considered a symptom presented to be a main, an additional, or no symptom of depression (N = 221–225); only the first 5 symptoms are part of the diagnostic criteria of depression
Main Sy.1 (%) Additional Sy.2 (%) False Sy.3 (%)
Depressed mood 209 (93.3%) 13 (5.8%) 2 (0.9%)
Reduced energy 200 (89.3%) 22 (9.8%) 2 (0.9%)
Bleak and pessimistic views of the future 192 (85.3%) 33 (14.7%) 0 (0%)
Disturbed sleep 112 (50.0%) 101 (45.1%) 11 (4.9%)
Considerable distress or agitation 84 (37.3%) 125 (55.6%) 16 (7.1%)
Disorientation for the own person 1 (0.4%) 22 (9.8%) 203 (89.7%)
Compulsion to wash 0 (0%) 28 (12.3%) 119 (57.6%)
Disturbed perception 7 (3.1%) 62 (27.6%) 158 (69.3%)
Vague thinking and distorted speaking 7 (3.2%) 67 (30.3%) 148 (66.5%)
Repeated revival of a trauma 15 (6.8%) 138 (62.2%) 69 (31.1%)
1 Main symptoms: respondents who considered the variables to be a main symptom of depression
2 Additional symptoms: respondents who considered the variables to be an additional symptom of depression
3 False symptoms: respondents who considered the variables not to be a symptom of depression
Table 3 Distribution of the interviewees who considered a symptom presented to be a main, an additional, or no symptom of schizophrenia (N = 222–225); only the first 5 symptoms are part of the diagnostic criteria of schizophrenia
Main Sy.1 (%) Additional Sy.2 (%) False Sy.3 (%)
Auditory hallucinations 158 (70.5%) 58 (25.9%) 8 (3.6%)
Feelings or actions experience as made or influenced by external agents4 111 (49.6%) 93 (41.5%) 20 (8.9%)
Delusions 107 (47.8%) 95 (42.4%) 22 (9.8%)
Delusions of control 53 (23.8%) 104 (46.6%) 66 (29.6%)
Hallucinations of taste 42 (18.8%) 110 (49.1%) 72 (32.1%)
Increased prevalence of allergies 2 (0.9%) 33 (14.7%) 189 (84.4%)
Agoraphobia with panic attacks 12 (5.3%) 74 (32.9%) 139 (61.8%)
Recklessly money spending in combination with grandiosity 21 (9.5%) 91 (41.0%) 110 (49.5%)
Both sex have an increased readiness for violence during and outside of illness episodes 17 (7.7%) 127 (57.2%) 78 (35.1%)
Split personality 144 (64.3%) 53 (23.7%) 27 (12.1%)
1 Main symptoms: respondents who considered the variables to be a main symptom of schizophrenia
2 Additional symptoms: respondents who considered the variables to be an additional symptom of schizophrenia
3 False symptoms: respondents who considered the variables not to be a symptom of schizophrenia
4 The so-called "Gefühl des Gemachten"
To avoid a bias due to the use of technical terms, we described the symptoms according to ICD-10. E.g., 'delusions' were explained as 'the feeling of being chased or threatened by an organisation like, e.g., the mafia without any external reason.' 'Depressed mood' was illustrated as 'feeling of continuous low mood that is rarely changing from day to day and unresponsive to and out of keeping with current circumstances.' However, to facilitate the understanding of the paper we consequently used the technical terms.
• One question each to assess the knowledge about lifetime prevalence (percentage of the general population), sex distribution (answers: more males, more females, no gender difference) and the age of the first episode of the respective disorder (age of onset). However, these data will be reported in a separate paper.
Literacy scores
Three literacy sum scores were calculated. The first score, called true symptoms, was calculated by summing up each correct positive assignment of the items presented. Two points were given when an item correctly referred to the category main symptoms and one point for an item referring to the category additional symptoms. The maximum score that could be achieved was 10. The maximum score for false symptoms was five. The false symptoms score ('false symptoms') is representing the correct recognition of the items that are not part of the diagnostic criteria of the respective disorder (maximum score: 5). The overall score was calculated using the sum of 'true symptoms' and 'false symptoms'. Since the first and the second score had shown different results in bivariate and multivariate analysis the overall score was dropped from further analyses.
Statistical analyses
The statistical analyses were carried out using SPSS for Macintosh (Version 6 and Version 11). The usual descriptive and bivariate analyses χ2-test, T-test and F-test were applied. Associations between faculty affiliation and depression/schizophrenia literacy were analysed by means of correspondence analysis [29]. Faculty affiliations were law, economics, medicine, natural and social sciences. Psychology was discriminated from social sciences. Literacy was analysed using a correspondence analysis of two 'interactive variables'. The first is matching gender and faculty affiliation and the second high/low literacy levels of depression and schizophrenia. The cut-off values with respect to the latter variable distinguished scores of ≤ 8 vs. >8 in depression (46.8% vs. 53.2% of the sample) and ≤ 6 vs. >6 in schizophrenia (51.4% vs. 48.6% of the sample) resulting in four categories (low/low, low/high, high/low and high/high literacy).
Results
The recognition of depressive symptoms, whether main or additional, was consistent over 90% (Table 2). Depressed mood was well recognised as a symptom of depression whereas sleep problems, distress and agitation were only recognised as additional disorder. Repeated revival of a trauma was not well recognised as a symptom of depression. Schizophrenia compared to depression symptoms were not well recognised (Table 3). Split personality and increased readiness for violence were often mistaken for symptoms of schizophrenia.
Table 4 is shows the bivariate analyses of explanatory variables and the two knowledge scores for depression and schizophrenia symptoms, respectively. The mean sum scores are clearly higher for depression compared to schizophrenia indicating that symptoms of depression are better recognised. The probabilities of the T-tests and F-tests, respectively, indicate that true symptoms scores and false symptoms scores are related to different explanatory variables. This is despite the fact that all variables are more or less closely associated with the faculty affiliation. Medical and psychology students have the highest true symptoms scores. Their advantage is more noticeable with respect to schizophrenia than to depression. However, medical students do not differ from other faculties regarding false symptoms scores.
Table 4 Bivariate analyses (t-test or F-test) of explanatory variables and knowledge mean sum scores regarding depression and schizophrenia (N = 222–225)
Sample Depression Schizophrenia
N1 (%) True symptoms2 False symptoms3 True symptoms2 False symptoms3
Sex Men 105 (47.3%) 8.2 3.4 5.9 2.6
Women 117 (52.7%) 8.6 3.4 6.5 2.3
p .007 .903 .059 .122
Age-group -24 147 (65.9%) 8.4 3.4 6.1 2.3
25–29 57 (25.6%) 8.5 3.5 6.5 2.6
30+ 19 (8.5%) 8.4 3.9 7.0 2.7
p .917 .183 .163 .312
Faculty Law 37 (16.4%) 8.7 3.3 6.0 2.5
Economics 46 (20.4%) 8.2 3.2 5.7 2.0
Medical school 15 (6.7%) 9.0 3.1 8.2 1.7
Natural sciences4 18 (8.0%) 8.3 3.3 6.0 2.4
Philosophy / arts5 109 (48.4%) 8.3 3.4 5.7 2.4
Psychology 28 (12.4%) 8.6 4.5 8.4 3.4
p .081 .000 .000 .000
Previous interest in this subject yes 105 (46.7%) 8.6 3.6 7.2 2.6
no 122 (53.3%) 8.3 3.4 5.5 2.3
p .026 .221 .000 .029
Side job related to mental disorders yes 33 (14.7%) 9.0 3.5 8.1 2.7
no 192 (84.6%) 8.3 3.4 5.9 2.4
p .003 .631 .000 .198
Self-experience of mental illness yes 133 (58.6%) 8.5 3.6 6.5 2.4
no 92 (40.5%) 8.3 3.3 5.9 2.5
p .386 .078 .070 .675
Own treatment experience due to mental problems yes 44 (19.4%) 8.6 3.8 7.3 2.7
no 192 (79.7%) 8.4 3.4 6.0 2.4
p .254 .042 .001 .069
Knowledge about mental disorders yes 122 (53.7%) 8.4 3.7 6.8 2.6
no 103 (45.4%) 8.4 3.1 5.7 2.2
p .957 .000 .001 .037
1 Varying sample sizes due to missing values
2 True symptoms: mean sum score about the recognition of the items that are part of the diagnostic criteria of depression or schizophrenia, respectively
3 False symptoms: mean sum score about the recognition of the items that are not part of the diagnostic criteria of depression or schizophrenia, respectively. A high false negative symptoms score reflects greater mental health literacy.
4 6 students in veterinary medicine included
5 1 student in theology included
The mean true symptoms score for depression was 8.42 (± 1.17) and for schizophrenia 6.27 (± 2.32), respectively. The mean false symptoms score for depression was 3.45 (± 1.24) and for schizophrenia 2.43 (± 1.15), respectively. In both depression and schizophrenia symptoms score, the correlations between the positive and negative symptoms scores were not significantly different from 0.
Correspondence analysis
The correspondence analysis of the interactive variables gender*faculty (e.g., m_law) vs. depression*schizophrenia (e.g., LOWLOW) literacy yielded a total inertia of 0.319 (inertia represents the percent of variance explained by each dimension). The inertia of the first dimension was 0.203 (64% of total inertia) and of the second 0.098 (31%). The contribution of the distinct categories to the inertia of each dimension is depicted in Table 5.
Table 5 Contribution of each dimension to the inertia*
Dimension
Variable 1 2
depression*schizophrenia literacy 11 LOWLOW .360 .358
12 LOWHIGH .053 .055
21 HIGHLOW .104 .520
22 HIGHHIGH .483 .067
gender*faculty affiliation 12 m_law .000 .246
13 m_econ .181 .143
14 m_med .114 .035
16 m_scie .027 .113
17 m_arts .092 .037
18 m_psyc .083 .025
22 f_law .042 .089
23 f_econ .008 .012
24 f_med .220 .036
26 f_scie .002 .181
27 f_arts .041 .049
28 f_psyc .190 .034
* Inertia: represents the percent of variance explained by each dimension
Figure 1 shows how gender and faculty diverge among the categories of literacy. There are three different groups. Medical and psychology students had an overall high level of literacy. Male students of natural sciences, economics and philosophy were especially illiterate.
Discussion
Mental health literacy has been recognised as a crucial prerequisite for early recognition and intervention in mental disorders. In this online-based questionnaire we presented symptoms of depression and schizophrenia to a sample of university students. The vast majority of the participants recognised the specific symptoms of depression. The symptoms of schizophrenia, however, were recognised to a lesser extent. Delusions of control and hallucinations of taste were not identified as symptoms of schizophrenia. In contrast, repeated revival of a trauma for depression and split personality for schizophrenia were mistaken as symptoms of the respective disorder. Bivariate analyses demonstrated that faculty affiliation, not gender, age or personal experience of mental illness, had an influence on symptom recognition in either disorders. Previous interest in and a side job related to mental disorders as well as treatment experience had a positive influence on symptom recognition. The correspondence analysis showed that male students in natural sciences, economics and philosophy are especially illiterate in recognising the two disorders. Gender became an important factor when controlling for faculty affiliation.
Thus, our questions can be answered as follows: the recognition rate of diverse symptoms of depression and schizophrenia among university students varies according to gender, symptoms and the faculty affiliation. It is also dependent on previous experience with mental disorders, be it at work, by interest or due to own treatment experience. Gender and faculty affiliation play a deciding role in recognising mental disorders.
Methodological considerations
As far as it is known, this is the first online survey in the field of mental health literacy. The Internet is likely to facilitate access to people and information. Thus, interest in online surveys are growing despite some shortcomings, e.g., sampling limitations as a result of omitting those without an email address/access as well as limits on response alternatives and interviewer observation [30]. Interpretation of responses is difficult because we do not know whether the answers were honest, i.e. whether the interviewees responded according to their own knowledge or if they made use of additional information about mental disorders – although they were instructed not to do so. There are limitations concerning public opinion surveys: they may be criticised for underestimating antipathy as those unaffected or disinterested may refuse to participate in a survey and many give socially desirable responses [31]; medical researchers tend to ask closed questions and obtain positive answers while sociologists ask open question, thus, uncovering negative stigmatising answers [32]; and attitudes should not be mistaken for actual interpersonal behaviour, but should be considered as a 'proxy' measure of social behaviour [33]. Furthermore, one could wonder whether the response rate is not higher than 18.3% of the eligible sample. However, our response rate is higher than previous online surveys at the University of Zurich (Zurich University administration, personal communication, 2003) and it must be mentioned that no incentives for participating were given. The schizophrenia symptoms did not include negative symptoms. We believe that the distinction between negative symptoms of a patient with schizophrenia and some symptoms of depression are too specific to be asked in an online survey among lay people, and could be a source of bias.
Comparison to the literature
There are few publications about how the general population recognises mental disorders. Four in ten adults who are symptomatic, but undiagnosed, have never heard of clinical depression and 84% had never heard of anxiety disorders [34]. In an Australian survey, nearly 38% of those questioned did not recognise depression [35]. Magliano et al. [36] found in a survey conducted in Italy that 21% of the general public identified schizophrenia in the case vignette. In a comparable opinion poll in Switzerland, the depression vignette was correctly recognised by 39.8% of the respondents whereas the remaining 60.2% considered the person depicted as having a 'life crisis' [37]. Schizophrenia, however, was recognised by 73.6% of the interviewees. Recognition of mental disorders was closely related to positive attitudes of psychiatry in general, but not to previous experience with treatment [38].
Our results show low recognition of distinct symptoms of mental disorders, which differs from the findings when a vignette was presented. In the latter, symptoms of schizophrenia were recognised as an illness and specific symptoms were recognised at a low rate. The symptoms of depression showed exactly the opposite results: Specific symptoms were recognised as illness-related whereas in the vignette the symptoms were classified as a 'life crisis'. For mental health policy recommendations and awareness campaigns to be useful special emphasis must be given on the differences between depression and schizophrenia.
Implications for mental health policy
Based on both the literature and the results presented in this study recommendations for mental health policy can be assessed. Correspondence analysis yields two substantial effects: education and gender-related effects. The major effect of education is very important, for example the study of medicine or psychology. But one must not forget that to study on the university is a privilege and, therefore, the tested student population is not representative of the population in general. The differences in mental health literacy would probably be even more obvious. Thus, one conclusion is that education can be very important when changing prejudice [39].
When comparing male and female students regardless of faculty, except medicine and psychology, females had a higher mental health literacy. This gender-specific effect is in line with research findings of gender differences concerning knowledge about and attitudes towards people with mental illness. Women tend to react more sympathetic towards people with mental illness and are more likely to volunteer in psychiatry wards or hospitals [40]. Therefore, gender-specific interventions should be considered.
What should be part of these lessons? The knowledge of the symptoms of depression is considerably high which could be a result of the high prevalence of affective disorders [41]. This seemingly contrasts with the findings that depression is hardly considered as a mental illness [42]. In general, depressive states are not viewed as illnesses but rather as normal psychological phenomena commonly called life crises [43]. This could also explain why a 'repeated revival of a trauma' is often mistaken as a symptom of depression. The consequences of public unawareness is that medical treatment for depression is not regarded as being necessary while non-medical interventions are thought to be helpful [44]. Thus, for mental health policy the emphasis on categorisation of depression as a mental illness (and not as a 'life crisis') and to a lower extent the recognition of distinct symptoms is very important.
Further suggestions for improving health literacy concerning schizophrenia are based on different findings: the low recognition rate of symptoms of schizophrenia compared with depression and two stereotypical attitudes about schizophrenia, namely, that it is associated with more violence and that split personality is one of its main symptoms. Firstly, the recognition is influenced by interest and job experience. Thus, previous practical or theoretical contacts to the topic improve the recognition of the disorder and its respective symptoms. This supports the findings that in the general population contact with mentally ill people is increasing along with the recognition of their disorder [45]. Additionally, contact has been shown to positively influence various parameters concerning people with mental illness, e.g., social distance or restrictions towards the mentally ill [39,46-48].
Stereotypes such as 'people with schizophrenia areviolent' or 'have a split personality' must be addressed. Whenever people with schizophrenia are more violent than the general public, it involves a combination of factors, for example male gender, comorbid substance use, less treatment compliance or severe psychopathology [49]. These constantly disseminated stereotypes can only be fought by persistent and differentiated information and clarification. This proposal is in line with previous findings that information improves mental health literacy [39]. In this regard, the mass media may play a crucial role. Instead of demonising people with mental illnesses and, thus, increasing prejudices against the affected, they should be motivated to engage in the fight against discrimination and stigmatisation of mental illness.
Conclusion
There is a lack of knowledge among university students, especially among those in natural sciences, philosophy and economics. Differences between distinct groups among the general population are expected to be even more extreme. Poor mental health literacy was significant regarding symptoms of schizophrenia. The following conclusions can be drawn to improve mental health literacy:
• Contact to mental disorders, either in a theoretical way (e.g., by interest) or by practical activity (e.g., at work), improves mental health literacy. Thus, possibilities for contact must be available.
• As it is true for health in general [50], education regarding mental health is needed. The time before university is advantageous, either in high school or during the apprenticeship.
• Education should emphasise that depression is a mental illness
• Stereotypes of schizophrenia should be addressed.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
C.L., V.A.-G. and W.R. designed the study and drafted the paper. N.F. collected the data and gave significant contributions to the content of the paper. N.S. gave significant contributions to the content of the paper and did the final editorial revision. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors would like to thank the students who participated in this online study.
==== Refs
Jorm AF Korten AE Jacomb PA Christensen H Rodgers B Pollitt P 'Mental health literacy': a survey of the public's ability to recognise mental disorders and their beliefs about the effectiveness of treatment Med J Aust 1997 166 182 186 9066546
Kessler RC McGonagle KA Zhao S Nelson CB Hughes M Eshleman S Wittchen HU Kendler KS Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey Arch Gen Psychiatry 1994 51 8 19 8279933
Jorm AF Mental health literacy. Public knowledge and beliefs about mental disorders Br J Psychiatry 2000 177 396 401 11059991 10.1192/bjp.177.5.396
Jorm AF Mental health literacy. Public knowledge and beliefs about mental disorders Br J Psychiatry 2000 177 396 401 11059991 10.1192/bjp.177.5.396
Murray CJ Lopez AD Global Burden of Disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020 1996 Harvard: Harvard University Press
Hannerz H Borga P Borritz M Life expectancies for individuals with psychiatric diagnoses Public Health 2001 115 328 337 11593442 10.1016/S0033-3506(01)00471-1
Chen H Parker G Kua J Jorm A Loh J Mental health literacy in Singapore: a comparative survey of psychiatrists and primary health professionals Ann Acad Med Singapore 2000 29 467 473 11056777
Christensen H Griffiths K The Internet and mental health literacy Aust N Z J Psychiatry 2000 34 975 979 11127628 10.1046/j.1440-1614.2000.00830.x
Goldney RD Fisher LJ Wilson DH Mental health literacy: an impediment to the optimum treatment of major depression in the community J Affect Disord 2001 64 277 284 11313096 10.1016/S0165-0327(00)00227-5
Goldney RD Fisher LJ Wilson DH Cheok F Mental health literacy of those with major depression and suicidal ideation: an impediment to help seeking Suicide Life Threat Behav 2002 32 394 403 12501964 10.1521/suli.32.4.394.22343
Jorm AF Korten AE Jacomb PA Christensen H Rodgers B Pollitt P Public beliefs about causes and risk factors for depression and schizophrenia Soc Psychiatry Psychiatr Epidemiol 1997 32 143 148 9130866
Jorm AF Korten AE Jacomb PA Christensen H Rodgers B Pollitt P 'Mental health literacy': a survey of the public's ability to recognise mental disorders and their beliefs about the effectiveness of treatment Med J Aust 1997 166 182 186 9066546
Jorm AF Mental health literacy. Public knowledge and beliefs about mental disorders Br J Psychiatry 2000 177 396 401 11059991 10.1192/bjp.177.5.396
Kitchener BA Jorm AF Mental health first aid training for the public: evaluation of effects on knowledge, attitudes and helping behavior BMC Psychiatry 2002 2 10 12359045 10.1186/1471-244X-2-10
Lauber C Nordt C Falcato L Rössler W Do people recognise mental illness? Factors influencing mental health literacy Eur Arch Psychiatry Clin Neurosci 2003 253 248 251 14504994 10.1007/s00406-003-0439-0
Mubbashar MH Farooq S Mental health literacy in developing countries Br J Psychiatry 2001 179 75 11435276 10.1192/bjp.179.1.75-a
Yeo SG Parker G Yap HL Mahendran R Mental health literacy beliefs. A comparison of psychiatric trained nurses and enrolled nurses in Singapore J Psychosoc Nurs Ment Health Serv 2003 41 34 41 12643081
Magliano L De Rosa C Fiorillo A Malangone C Guarneri M Marasco C Maj M the Working Group of the Italian National Study Beliefs of psychiatric nurses about schizophrenia: a comparison with patients' relatives and psychiatrists Int J Soc Psychiatry 2004 50 319 330 15648745 10.1177/0020764004046073
Magliano L Fiorillo A De Rosa C Malangone C Maj M Beliefs about schizophrenia in Italy: a comparative nationwide survey of the general public, mental health professionals, and patients' relatives Can J Psychiatry 2004 49 322 330 15198469
Magliano L De Rosa C Fiorillo A Malangone C Maj M Perception of patients' unpredictability and beliefs on the causes and consequences of schizophrenia- a community survey Soc Psychiatry Psychiatr Epidemiol 2004 39 410 416 15133599 10.1007/s00127-004-0764-2
Chen H Parker G Kua J Jorm A Loh J Mental health literacy in Singapore: a comparative survey of psychiatrists and primary health professionals Ann Acad Med Singapore 2000 29 467 473 11056777
Jorm AF Mental health literacy. Public knowledge and beliefs about mental disorders Br J Psychiatry 2000 177 396 401 11059991 10.1192/bjp.177.5.396
Lauber C Nordt C Falcato L Rössler W Do people recognise mental illness? Factors influencing mental health literacy Eur Arch Psychiatry Clin Neurosci 2003 253 248 251 14504994 10.1007/s00406-003-0439-0
Nordt C Falcato L Lauber C Rössler W Assessment of depression and schizophrenia as mental illness and its influence on the help-seeking recommendations Neuropsychiatrie 2002 16 103 107
Scott TL Gazmararian JA Williams MV Baker DW Health literacy and preventive health care use among Medicare enrollees in a managed care organization Med Care 2002 40 395 404 11961474 10.1097/00005650-200205000-00005
Goldney RD Fisher LJ Wilson DH Mental health literacy: an impediment to the optimum treatment of major depression in the community J Affect Disord 2001 64 277 284 11313096 10.1016/S0165-0327(00)00227-5
Jorm AF Korten AE Jacomb PA Christensen H Rodgers B Pollitt P 'Mental health literacy': a survey of the public's ability to recognise mental disorders and their beliefs about the effectiveness of treatment Med J Aust 1997 166 182 186 9066546
Lauber C Nordt C Falcato L Rössler W Public attitude to compulsory admission of mentally ill people Acta Psychiatr Scand 2002 105 385 389 11942946 10.1034/j.1600-0447.2002.1o267.x
Weller SC Kimball Romney A Metric scaling: correspondence analysis 1990 Newbury Park, CA: Sage Publications
Rosenvinge JH Laugerud S Hjortdahl P Trust in health Websites: a survey among Norwegian Internet users J Telemed Telecare 2003 9 161 166 12877778 10.1258/135763303767149979
Byrne P Psychiatric stigma Br J Psychiatry 2001 178 281 284 10.1192/bjp.178.3.281
Hayward P Bright JA Stigma and mental illness: A review and critique J Mental Health 1997 6 345 354 10.1080/09638239718671
Penn DL Corrigan PW The effects of stereotype suppression on psychiatric stigma Schizophr Res 2002 55 269 276 12048150 10.1016/S0920-9964(01)00207-9
America's Mental Health Survey 2001 2001 Roper Starch Worldwide Inc
Lepine JP Gastpar M Mendlewicz J Tylee A Depression in the community: the first pan-European study DEPRES (Depression Research in European Society) Int Clin Psychopharmacol 1997 12 19 29 9179630
Magliano L Fiorillo A De Rosa C Malangone C Maj M Beliefs about schizophrenia in Italy: a comparative nationwide survey of the general public, mental health professionals, and patients' relatives Can J Psychiatry 2004 49 322 330 15198469
Lauber C Nordt C Falcato L Rössler W Do people recognise mental illness? Factors influencing mental health literacy Eur Arch Psychiatry Clin Neurosci 2003 253 248 251 14504994 10.1007/s00406-003-0439-0
Lauber C Nordt C Falcato L Rössler W Do people recognise mental illness? Factors influencing mental health literacy Eur Arch Psychiatry Clin Neurosci 2003 253 248 251 14504994 10.1007/s00406-003-0439-0
Corrigan PW River LP Lundin RK Penn DL Uphoff-Wasowski K Campion J Mathisen J Gagnon C Bergman M Goldstein H Kubiak MA Three strategies for changing attributions about severe mental illness Schizophr Bull 2001 27 187 195 11354586
Lauber C Nordt C Falcato L Rössler W Determinants of attitude to volunteering in psychiatry: results of a public opinion survey in Switzerland Int J Soc Psychiatry 2002 48 209 219 12413249 10.1177/002076402128783253
Kessler RC McGonagle KA Zhao S Nelson CB Hughes M Eshleman S Wittchen HU Kendler KS Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey Arch Gen Psychiatry 1994 51 8 19 8279933
Lauber C Nordt C Falcato L Rössler W Do people recognise mental illness? Factors influencing mental health literacy Eur Arch Psychiatry Clin Neurosci 2003 253 248 251 14504994 10.1007/s00406-003-0439-0
Lauber C Nordt C Falcato L Rössler W Lay recommendations on how to treat mental disorders Soc Psychiatry Psychiatr Epidemiol 2001 36 553 556 11824850 10.1007/s001270170006
Lauber C Nordt C Falcato L Rössler W Lay recommendations on how to treat mental disorders Soc Psychiatry Psychiatr Epidemiol 2001 36 553 556 11824850 10.1007/s001270170006
Lauber C Nordt C Falcato L Rössler W Do people recognise mental illness? Factors influencing mental health literacy Eur Arch Psychiatry Clin Neurosci 2003 253 248 251 14504994 10.1007/s00406-003-0439-0
Lauber C Nordt C Sartorius N Falcato L Rössler W Public acceptance of restrictions on mentally ill people Acta Psychiatr Scand Suppl 2000 102 26 32 11261636 10.1034/j.1600-0447.2000.00005.x
Lauber C Nordt C Falcato L Rössler W Public attitude to compulsory admission of mentally ill people Acta Psychiatr Scand 2002 105 385 389 11942946 10.1034/j.1600-0447.2002.1o267.x
Lauber C Nordt C Falcato L Rössler W Factors influencing social distance toward people with mental illness Community Ment Health J 2004 40 265 274 15259631 10.1023/B:COMH.0000026999.87728.2d
Soyka M Substance misuse, psychiatric disorder and violent and disturbed behaviour Br J Psychiatry 2000 176 345 350 10827882 10.1192/bjp.176.4.345
Kemm J Health education: a case for resuscitation Public Health 2003 117 106 111 12802976 10.1016/S0033-3506(02)00013-6
| 15882465 | PMC1156910 | CC BY | 2021-01-04 16:28:57 | no | BMC Public Health. 2005 May 9; 5:44 | utf-8 | BMC Public Health | 2,005 | 10.1186/1471-2458-5-44 | oa_comm |
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-491590451410.1186/1471-2458-5-49Research ArticleA mass vaccination campaign targeting adults and children to prevent typhoid fever in Hechi; Expanding the use of Vi polysaccharide vaccine in Southeast China: A cluster-randomized trial Yang Jin [email protected] Camilo J [email protected] Guo-ai [email protected] Jun [email protected] Cui-yun [email protected] Da-bin [email protected] R Leon [email protected] Anne-Laure [email protected] M Carolina [email protected] Jie [email protected] Bao-de [email protected] He-zhuang [email protected] Ming-liu [email protected] Dong-mei [email protected] Zhen-zhu [email protected] Jian [email protected] Jin-Kyung [email protected] Mohammad [email protected] Bernard [email protected] Gui-chen [email protected] Hong-hui [email protected] Tikki [email protected] Zhi-yi [email protected] Allan [email protected] Claudia M [email protected] Bai-qing [email protected] John D [email protected] Division of Bacterial Diseases, Guangxi Center for Disease Control, Nanning, Guangxi, China 80 Taoyuan Road, Nanning 530021, Guangxi, China2 International Vaccine Institute, Kwanak P.O. Box 14, Seoul, Korea 151-6003 Hechi Anti-epidemic Center, Hechi, Guangxi, China 2 Xinjian Road, Jinchengjiang 54700, Guangxi, China4 Vaccines and Other Biologicals, World Health Organization, Geneva, Switzerland5 Research Policy and Cooperation, World Health Organization, Geneva, Switzerland6 University of Western Ontario, Canada2005 18 5 2005 5 49 49 28 12 2004 18 5 2005 Copyright © 2005 Yang et al; licensee BioMed Central Ltd.2005Yang et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
One of the goals of this study was to learn the coverage, safety and logistics of a mass vaccination campaign against typhoid fever in children and adults using locally produced typhoid Vi polysaccharide (PS) and group A meningococcal PS vaccines in southern China.
Methods
The vaccination campaign targeted 118,588 persons in Hechi, Guangxi Province, aged between 5 to 60 years, in 2003. The study area was divided into 107 geographic clusters, which were randomly allocated to receive one of the single-dose parenteral vaccines. All aspects regarding vaccination logistics, feasibility and safety were documented and systematically recorded. Results of the logistics, feasibility and safety are reported.
Results
The campaign lasted 5 weeks and the overall vaccination coverage was 78%. On average, the 30 vaccine teams gave immunizations on 23 days. Vaccine rates were higher in those aged ≤ 15 years (90%) than in adolescents and young adults (70%). Planned mop-up activities increased the coverage by 17%. The overall vaccine wastage was 11%. The cold chain was maintained and documented. 66 individuals reported of adverse events out of all vaccinees, where fever (21%), malaise (19%) and local redness (19%) were the major symptoms; no life-threatening event occurred. Three needle-sharp events were reported.
Conclusion
The mass immunization proved feasible and safe, and vaccine coverage was high. Emphasis should be placed on: injection safety measures, community involvement and incorporation of mop-up strategies into any vaccination campaign. School-based and all-age Vi mass immunizations programs are potentially important public health strategies for prevention of typhoid fever in high-risk populations in southern China.
==== Body
Background
The People's Republic of China has led the use of Vi vaccine as a public health tool to contain typhoid fever in some provinces in which the disease is endemic. With Vi vaccine proven to be efficacious in clinical trials [1,2], in 1989 the Lanzhou Institute of Biological Products began planning for the production of Vi vaccine in cooperation with the original developers [3]. The production followed the specifications published by the World Health Organization (WHO) [4]. In 1995, 2 large-scale licensing randomized, placebo-controlled trials in China of this locally produced vaccine demonstrated its safety and protective efficacy (71% and 69% protection 12 and 19 months after vaccination, respectively) [5], [6]. During an outbreak in 1999, the vaccine was found to be 71% effective [7], similar to the efficacy results reported in phase III trials.
Vi vaccine is currently produced by 6 vaccine institutes in China. Some provincial and district governments have encouraged its use through school-based campaigns financed with reasonable user fees (e.g., US $0.30 to $0.60 per dose). In 1996, the Guangxi Zhuang Autonomous Region (Guangxi Province) in southern China introduced Vi vaccine as a public health tool for school-aged children and for use during outbreaks. Although the reduction in typhoid fever cases from government surveillance has been reported, to date, there has been no formal assessment of the effectiveness of the Vi vaccine use as a public health tool.
Guangxi Province plans to expand the Vi vaccination strategy beyond the school-aged children. An open cluster randomized controlled Vi vaccine effectiveness evaluation was designed and launched in 2002 with the goal of generating policy-relevant data to enable expanded use of Vi vaccine in public health program [8]. Herein, we report initial results on vaccine coverage, safety and logistics of the Vi vaccine when delivered through a mass vaccination campaign that targeted a population aged 5–60 years in a city in Guangxi Province.
Methods
Study site and design
The vaccination campaign was carried out in Hechi Prefecture (figure 1), Guangxi Province, between April 8 and May 12, 2003. The study area (referred to in the text as Hechi) is located 420 km northwest of the provincial capital, Nanning, and includes Jin Cheng Jiang (urban) and Don Jiang (rural), the two most populous areas in the prefecture. Agriculture is the major source of income in this subtropical area. The average household annual income is 1,700 RMB (1$ US = 8.3 RMB). Some 80% of the population is of the Zhuang ethnic group. Birth and death rates were 8% and 6%, respectively, in 2001.
Figure 1 Hechi in Guangxi Province, PR China. shows the location of Hechi and Guangxi in China.
In recent years national child immunization program coverage has been over 90% in Hechi, which has a variety of health facilities (85 in total) including hospitals, health centers, factory clinics and private clinics. The Hechi Center for Disease Control (CDC) is responsible for the collection of data on notifiable infectious diseases. In this area, in 1995–1999, typhoid fever incidence rates were 27–153 per 100,000 residents [9] and typhoid fever affected both school-aged children and adults.
The target population for the Vi vaccination campaign was 118,588 persons aged 5–60 years who were registered in the project census conducted in January 2003. The study area was divided into 107 geopolitical clusters, each resembling a unit to be targeted in a public health program. The mean cluster size was 1,103 persons (range: 319–2,610). Of the clusters, 77 with 85,815 persons were in urban areas and the remaining 30, with 32,256 persons, in rural areas. Of households within the clusters, the majority (41%) lived in factory compounds, 21% in private homes, 6% in boarding schools, 6% in government housing and 1% in health institutions.
Clusters were randomly allocated to receive either Vi vaccine or a group A meningococcal (menA) vaccine, the active control. Details of the rationale and design of this open-cluster randomized-effectiveness trial are provided elsewhere [10], [11]. All 107 clusters were stratified according to population size (large ≥ 1,000 or small) and setting (urban or rural) before randomization. The exclusion criteria were pregnancy, breastfeeding, fever >37.5°C at the time of vaccination and presence of severe chronic diseases. The last exclusion criterion is applied to all vaccines in China.
Both vaccines used for this study are licensed and commercially available in China; however, due to the nature of research aspects in determining the effectiveness of the Vi vaccine, ethical and research clearance was obtained from the Guangxi ethical review board, the International Vaccine Institute's institutional review board and the ethical committee of the World Health Organization (WHO). Trial information was disseminated at the community level (meetings and media) and consent was recorded individually on the day of immunization.
Mass vaccination campaign
Promotional activities were initiated 2 months prior to immunization and included meetings with community leaders and campaign advertisements via local television, loudspeakers, posters and information flyers. The mass vaccination campaign was carried out in 4 stages. Stage I (i.e., the pilot phase) vaccinated a few clusters in order to test and fine-tune the system and to provide final training for the other teams. The next stage was the large-scale vaccination at pre-defined vaccination posts in each cluster according to a pre-set schedule. In stage III, the initial mop-up, the vaccination team re-visited its corresponding vaccination post at the cluster for 1 day. The final mop-up (stage IV) was at a centralized vaccination center (Hechi CDC) after the study population was again invited to be vaccinated.
Both the Vi and the menA vaccines are licensed. They are produced by Lanzhou Institute of Biological Products and administered in 1-dose regimes. A Vi vaccine vial contains two 0.5-ml doses (each with 30 μg of the purified Vi of Salmonella enterica subspecies enterica serovar Typhi) for both children and adults. This vaccine is administered intramuscularly. The menA vaccine vial contains five freeze-dried 0.2-ml doses (each is 30 μg of the purified group A PS of Neisseria meningitidis). Each dose must be reconstituted with 1 ml of PBS (diluent provided with the vaccine) before subcutaneous injection.
Guangxi and Hechi CDC personnel planned, coordinated and launched the mass campaign. Training in good clinical practices (GCPs) was given to all vaccination team members and supervisors 1 month prior to the campaign. The vaccines were delivered by methods that simulated public health program conditions. One site was selected in each cluster (school, health facility, factory or locations such as intersections and squares) to become the vaccination posts. Table 1 depicts the type of personnel involved, responsibilities and numbers needed during the campaign.
Table 1 Human resources involved in the Hechi, China, typhoid fever vaccine campaign 2003
Personnel Function Number
Vaccine team (30 teams)*
Leaders (physicians) Overall responsible for cluster vaccination; daily collection of vaccines and supplies; treat SAE 30
Nurses Vaccinator 43
Other health workers Recorder 24
Non-health workers Recorder 9
Community helpers Facilitated immunization process; liaison between community and Hechi Center for Disease Control 78
Other
Storage room Maintained cold chain 4
Data management Data entry of vaccine records 6
Field health workers SAE 3-day home visit 3
Drivers 3
Supervisors Assured adherence to standard operating procedures 7
NOTE: The research-component of the project is not included
* Each team had a leader, vaccinator, recorder and community helper.
Each team was expected to vaccinate about 200 persons each day and to cover 3 or 4 clusters in 20 days. Identification cards similar to those used in the local expanded program of immunization (EPI) were distributed to each household 1 month before vaccination. The cards had identifying information and a study number that corresponded to the household's date and time of vaccination. In order to avoid errors in vaccine allocation, each vaccine team delivered only one vaccine. Those registered in the project census were vaccinated and recorded. However, persons not registered in the project census were offered menA vaccine at the end of the campaign.
Each vaccination team was equipped with: 1 or 2 cold boxes, injecting material, emergency kits, 1 safety disposal box and stationary. All eligible participants were listed in a vaccination record book available at the vaccine posts. Those who came for immunization and gave informed consent were assessed for eligibility by the vaccination team and vaccination status, regardless of received or not received, was recorded in the vaccination record book, which was later entered into a database. Vaccination coverage was calculated from this database.
All aspects of safety for each of the study vaccines were emphasized and the vaccination teams received intensive training. Two particular aspects were monitored: safe injection practices and adverse events (AE). Proper vaccine administration (intramuscular for typhoid Vi and subcutaneous for menA), aseptic injection technique and injection safety (not recapping and disposing of used needles into safety boxes) were closely monitored by supervisors and external observers. Team members were instructed to report all accidents involving sharps and needles to their supervisor. The complete incineration of used safety boxes was monitored by the Hechi CDC staff.
Safety of the vaccines was monitored systematically. All vaccinees were asked to remain in the vaccination posts for at least 15 minutes after vaccination to be monitored for any immediate serious adverse event (SAE) by the physician of the vaccination team. 535 randomly chosen vaccinees were visited for 3 consecutive days after vaccination by the project personnel for solicited adverse event surveillance. Unsolicited (passive) adverse event surveillance was carried out for one month following vaccination by having all healthcare facilities in the study area to report signs and symptoms of patients with history of receiving the vaccine, or hospitalization cases of the vaccinees to Hechi CDC. Physicians attending the patients were asked to fill in a form for all the cases with causality specified. Inadvertent vaccination of pregnant women was to be reported passively during the vaccination campaign; these cases would be followed until delivery. Deaths of study participants are being ascertained through established mortality surveillance by Hechi CDC, which gathers information from death certificates and cremation and hospital records for the 2-year follow-up period. Project personnel reviewed all incoming forms to determine the causality and when unclear, a trial clinical monitor assisted in reviewing the forms.
For both vaccines the manufacturer recommends storage temperatures of 2–8°C. All supplies, including the vaccines, were stored at the logistic hub at Hechi CDC: a 3 × 12 m2 room equipped with 8 refrigerators and 1 freezer. The number of vaccine vials and supplies distributed to each team was recorded daily to calculate vaccine usage and wastage. The maintenance of the cold chain was verified at all stages from the time the vaccine left the manufacturer until it reached the vaccination posts. Maximum/minimum (max/min) thermometers and battery-automated temperature recorders were used to monitor the vaccine temperature during transportation and storage. Regular thermometers in cold boxes were used at the vaccine posts. Temperatures were documented daily in temperature charts – twice for stored vaccines and 3 times for the cold boxes.
Results
The vaccination campaign lasted 31 workdays, during which 53 clusters received typhoid Vi vaccine and 54 the menA PS vaccine. Thirty teams gave immunizations on an average of 23 days, covering a target population of 118,588 in 107 clusters. Overall, 96,504 people came to the vaccination posts and 92,476 were immunized, yielding an overall vaccine coverage of 78%. In all, 25,605 persons were not vaccinated: 80% (20,472) did not appear for immunization. Table 2 summarizes reasons for not being vaccinated. 80% of the non-immunized did not show-up and 13% were excluded at time of vaccination (serious illness, lactation, pregnancy or fever).
Table 2 Reasons for not being vaccinated during the vaccine campaign
Reason Pregnancy Lactation Fever Serious disease Refusal* Did not show-up Vaccinated previously Total
No. person 533 714 143 1,851 788 20,472 1,104 25,605
(%) 2 3 1 7 3 80 4 100
* at the time of seeking consent
The vaccine coverage was 77% and 80% for Vi vaccine and menA vaccine, respectively. Figure 2 shows household locations (mapped by global positioning system technology) and magnitude of coverage by urban and rural area (and by cluster). As shown in figure 3, the highest vaccination coverage (≈90%) was achieved in children less than 15 years of age and was lowest in adolescents (15–19 years) and young adults (20–29 years), reaching 70%. Coverage by gender was similar: 79% in males and 82% in females. High vaccine coverage rates (92.5%) occurred in clusters that corresponded to schools. Figure 4 illustrates vaccine coverage by program stage: 74% (range: 56% – 92%) of the overall coverage was achieved in the first two phases of the vaccination campaign. During mop-up (phases III and IV), 5,269 persons (17% of the total remaining target population) were vaccinated, achieving final coverage of 78% (range: 65% – 93%). Stages I and II lasted 4 days and 5 days (range: 2–9) per cluster, respectively. The initial mop-up stage (stage III) required 4 days and the final mop-up (stage IV) was performed in 3 days. Both of these stages were preceded by 1-day with re-invitation activities.
Figure 2 Vaccine coverage in Hechi. shows the vaccine coverage by cluster and household locations in Hechi. Darker the color, higher the coverage is. Boundary of the city is also shown.
Figure 3 Vaccine coverage by age group. shows the coverage of each vaccine by age group.
Figure 4 Cluster vaccine coverage by campaign stages. shows the cluster vaccine coverage by campaign stages.
Three needle-stick injuries were reported during the mass vaccination campaign. All cases happened to the vaccinators of the team. One case occurred during crowded hours at the vaccination post, while two cases occurred due to mishandling of the sharp disposal box. The 3 persons exposed to blood were tested for hepatitis B surface antigen (HBsAg) antibodies and immediately given 1 dose of hepatitis B vaccine according to the local guidelines. No further doses were necessary as all 3 were anti-HBsAg positive.
No life-threatening immediate AE were observed. The surveillance system did identify 1 serious AE; however, this auto-limited febrile case which required hospitalization was not severe and the patient recovered in 12 hours. Of the 413 persons interviewed for solicited AE, 408 were surveyed on all 3 days, and no SAE were reported. In total 129 (38 Vi and 91 menA) AEs determined to be caused by the vaccine were reported from 66 individuals (26 Vi and 40 menA), consisting mainly in fever (21% Vi and 19% menA; ranging between 37.5°C to 39.5°C), malaise (19% Vi and 10% menA), and/or minor rash (19% Vi and 5% menA) around the site of injection. In all, 3,430 hospitalizations were reported 1 month after the immunizations. Of these persons, 300 had been immunized during the campaign. The trial clinical monitor determined that none of the hospitalizations (the majority were scheduled surgeries) were related to vaccine injections.
The Lanzhou Institute initially shipped 60,000 doses of Vi vaccine and 80,000 doses of menA vaccine to the study site by refrigerated (2–4°C) truck monitored by max/min thermometer. Once the temperature reached 14°C for approximately 3 hours. The vaccines were stored in cold rooms at the Guangxi CDC for 6 days where temperature ranged from 2°C to 8°C. On March 29, 2003, the vaccines were sent to the Hechi CDC in a refrigerated truck. This 420-km journey took 6 hours. The cold chain (monitored by battery-powered temperature recorder) was maintained between 4°C and 9°C. At the Hechi CDC, all 140,000 vaccine doses had temperature readings of 2°C to 11°C. The mean temperature recorded in the field was 4.2°C (range: 1.8°C – 12°C). In all, 26 vials of Vi vaccine were discarded due to freezing.
Wastage based on breakage, missing at inventory and unused opened vials was higher as expected for the 5-dose vial of the menA vaccine vial than for the Vi vaccine (12.8% vs. 9.2%., respectively). In conformity with the product information sheet, vials of menA vaccine were discarded within 1 hour of opening if not used. At the end of the campaign the unused doses of both vaccines were shipped to the Hechi CDC headquarters so they could be used elsewhere.
The number of personnel required during the campaign is shown in table 1. The majority were volunteer community members. Transportation from the Hechi CDC to the vaccination posts required a variety of transport: foot, bicycles and motorized vehicles (motorcycles, cars, buses or ambulances).
Conclusion
The results verify that a mass vaccination campaign with the 1-dose parenteral Vi vaccine or the menA vaccine, which targeted both school-aged children and adults in Hechi, south China, was logistically feasible and safe. An important coverage rate was attained with no major disruption to the normal EPI or other health activities.
In general, there are three main concerns when launching population-based large-scale vaccination campaigns: (1) resources, (2) safety and (3) cost. Especially in developing countries, there is the need to identify resources outside the well-established local vaccine delivery system (usually the EPI program). The Hechi campaign illustrates that an existing healthcare system can deliver vaccines to large populations in a fairly short period of time. Community members can play a key role in the success of promotional activities and delivery of vaccines when an entire city is targeted.
The Hechi campaign monitored and recorded safety systematically and in depth. No life-threatening serious AE were associated with either vaccine. No reported AE (solicited or not) were serious and all patients recovered quickly. Use of safety boxes was introduced for the first time in Hechi during this mass vaccination campaign. Despite instruction on safe handling of sharps, 3 needle-stick injuries were recorded. For future mass vaccination campaigns, including routine EPI activities, safe injection practices must be introduced and emphasized to prevent unnecessary accidents and spread of blood-borne diseases.
A mop-up strategy increased the coverage by 9.6% on average in the lower-coverage clusters and by 0.7% increase in the higher-coverage clusters. Because mop-up strategies imply the need for additional time and costs, this strategy should be planned in advance. Adult coverage was slightly lower than that of children, but a 70% coverage indicates the ability to reach adults during a vaccination campaign. The comprehensive vaccine safety surveillance system, implemented within the existing Hechi health facilities, was useful. Post-licensure monitoring can address many issues regarding safety, such as: uncommon AE, incorrect storage or administration procedures and dissemination of information to the public to maintain confidence in the scheme [12].
The cost-effectiveness of the use of Vi vaccine will be analyzed at the end of the study. However, because the vaccine is locally produced, it seems reasonable to anticipate that unlike EPI vaccines, cost savings can be achieved by reduced transportation costs and in-country production of the vaccine. In an urban area of Vietnam, a locally produced oral cholera vaccine distributed through mass immunization was found to be affordable if introduced in a public health program [13]. In countries, with decentralized health systems, such as China, public health sector introduction or expansion of the use of Vi vaccine can be initiated and financed by local governments.
Mass immunization is considered the most logical control strategy, outside the EPI program, in settings that typically manifest high incidences of disease or in populations exposed to predictable outbreaks [14]. Besides school-based immunization programs, all-age (5–60 years) Vi mass vaccinations are potentially an important public health strategy to prevent typhoid fever in high-risk populations in southern China. Follow-up of the randomized double-blind placebo-controlled trials of Vi vaccine indicate that vaccine protection of 50% lasts for 3 years (manuscript in preparation). The usefulness of Vi vaccine may be further advanced with the development of Vi conjugate vaccine [15]. We anticipate that the vaccine cost results, together with cost-effectiveness analysis of the Hechi program, will complement the above findings and provide objective information to assist policymakers who are increasingly confronted with juggling decisions regarding the use of a growing list of available vaccines while determining health priorities, often under severe budgetary constraints.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
YJ, CJA, MA, YH, TP, XZY, AD, DB, JDC have been in part of the conception and design of study, YJ, CJA, SG, ZengJ, LC, LD, RLO, AP, MCD, ZhangJ, ZB, LH, WM, TD, TZ, GJ, JKP, MA, BI, LG, YH, TP, XZY, AD, CMG, DB, JDC compiled, analyzed, and/or interpreted the data, and YJ, CJA, SG, ZengJ, LC, LD, RLO, AP, MCD, ZhangJ, ZB, LH, WM, TD, TZ, GJ, JKP, MA, BI, LG, YH, TP, XZY, AD, CMG, DB, JDC played roles in drafting and/or revising the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This work was supported by the Diseases of the Most Impoverished Program, funded by the Bill and Melinda Gates Foundation.
We thank the following persons: John M Albert, Jeremy Farrar, Eun Young Kim, Sue Kyoung Jo, John Wain, Amanda Walsh and the staff that worked for the following groups: Guangxi CDC; Hechi CDC; Guangxi and Hechi Health Bureau; private doctors of Hechi city; Oxford University-Wellcome Trust, Tropical Unit, Ho Chi Minh City, Vietnam; University of Western Ontario, London, Canada; WHO, Geneva, Switzerland and IVI, Seoul, Korea.
==== Refs
Acharya IL Lowe CU Thapa R Gurubacharya VL Shrestha MB Bryla DA Cramton T Trollfors B Cadoz M Schulz D Armand J Schneerson R Robbins JB. Prevention of typhoid fever in Nepal with the Vi capsular polysaccharide of Salmonella typhi: A preliminary report one year after immunization N Engl J Med 1987 317 1101 1104 3657877
Klugman KP Gilbertson IT Koornhof HJ Robbins JB Schneerson R Schulz D Cadoz M Armand J Vaccine Advisory Committee Protective activity of Vi capsular polysaccharide vaccine against typhoid fever Lancet 1987 2 1165 1169 2890805 10.1016/S0140-6736(87)91316-X
Robbins JD Robbins JB Reexamination of the protective role of the capsular polysaccharide (Vi antigen) of Salmonella typhi J Infect Dis 1984 150 436 449 6207249
World Health Organization Annex 1: Requirements for Vi polysaccharide typhoid vaccine (Requirements for Biological Substances No 48) WHO Technical Report Series No 840 1994 15 29
Wang ZG Zhou WZ Shi J [Efficacy and side effects following immunization with Salmonella typhi Vi capsular polysaccharide vaccine] (in Chinese) Zhonghua Liu Xing Bing Xue Za Zhi 1997 18 26 29 9812477
Yang HH Wu CG Xie GZ Gu QW Wang BR Wang LY Wang HF Ding ZS Yang Y Tan WS Wang WY Wang XC Qin M Wang JH Tang HA Jiang XM Li YH Wang ML Zhang SL Li GL Efficacy trial of Vi polysaccharide vaccine against typhoid fever in south-western China Bull World Health Organ 2001 79 625 631 11477965
Yang HH Kilgore PE Yang LH Park JK Pan YF Kim Y Lee YJ Xu ZY Clemens JD An outbreak of typhoid fever, Xing-An County, People's Republic of China, 1999: estimation of the field effectiveness of Vi polysaccharide typhoid vaccine J Infect Dis 2001 183 1775 1780 11372030 10.1086/320729
DeRoeck D Clemens JD Nyamete A Mahoney RT Policymakers' views regarding the introduction of new-generation vaccines against typhoid fever, shigellosis and cholera in Asia Vaccine 2005 23 2762 2774 2005 Apr 15 15780724 10.1016/j.vaccine.2004.11.044
Guangxi CDC Health statistical records Guangxi 1999
Acosta CJ Galindo CM Ochiai RL Danovaro-Holliday MC Page AL Thiem VD Park JK Koo HW Park E Wang XY Abu-Elyazeed R Ali M Ivanoff B Pang T Xu ZY Clemens JD The role of epidemiology in the introduction of typhoid fever Vi polysaccharide vaccines in Asia J Health Popul Nutr 2004 22 240 245 15609776
Acosta CJ Galindo CM Ali M Abu-Elyazeed R Ochiai RL Danovaro-Holliday MC Page AL Vu DT Yang J Park JK Lee H Puri MK Ivanoff B Agtini MD Soeharno R Simanjuntak CH Punjabi N Canh DG Sur D Nizami Q Manna B Dong B Dang DA Yang H Bhattacharya SK Bhutta ZA Trach DD Xu ZY Pang T Donner A Clemens JD A Multi-Country Cluster Randomized Controlled Effectiveness Evaluation to Accelerate the Introduction of Vi Polysaccharide Typhoid Vaccine in Developing Countries in Asia: Rationale and Design J Trop Med Int Health
Tozzi AE Field evaluation of vaccine safety Vaccine 2004 22 2091 2095 15121330 10.1016/j.vaccine.2004.01.013
Vu DT Hossain MM Nguyen DS Nguyen TH Rao MR Do GC Naficy A Nguyen TK Acosta CJ Deen JL Clemens JD Dang DT Coverage and costs of mass immunization of an oral cholera vaccine in Vietnam J Health Popul Nutr 2003 21 304 308 15043004
World Health Organization Background document: The diagnosis, treatment and prevention of typhoid fever 2003 WHO/V&B/03.07
Lin F-YC Ho VA Khiem HB Trach DD Bay PV Thanh TC Kossaczka Z Bryla DA Shiloach J Robbins JB Schneerson R Szu SC The efficacy of a Salmonella typhi Vi conjugate vaccine in two-to-five-year-old children N Engl J Med 2001 344 1263 1269 11320385 10.1056/NEJM200104263441701
| 15904514 | PMC1156911 | CC BY | 2021-01-04 16:28:57 | no | BMC Public Health. 2005 May 18; 5:49 | utf-8 | BMC Public Health | 2,005 | 10.1186/1471-2458-5-49 | oa_comm |
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-501590453710.1186/1471-2458-5-50Research ArticleEpidemiology of episodic adenolymphangitis: a longitudinal prospective surveillance among a rural community endemic for bancroftian filariasis in coastal Orissa, India Babu Bontha V [email protected] Abhay N [email protected] Kalpataru [email protected] Division of Epidemiology, Regional Medical Research Centre, Indian Council of Medical Research, Bhubaneswar – 751 023, India2005 19 5 2005 5 50 50 9 10 2004 19 5 2005 Copyright © 2005 Babu et al; licensee BioMed Central Ltd.2005Babu et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The epidemiological knowledge on acute condition of lymphatic filariasis is essential to understand the burden and issues on management of the disease.
Methods
A one year long longitudinal prospective surveillance of acute adenolymphangitis (ADL) was carried out in rural population of Orissa, India.
Results
The annual incidence of ADL per 1000 individuals is 85.0, and is slightly higher (P > 0.05) in male (92.0) than in female (77.6). A steady rise in the incidence of ADL episodes along with the age is recorded. The distribution indicates that persons with chronic disease are more prone to ADL attacks. The average number of episodes per year is 1.57 (1.15 SD) per affected person, and is gender dependent. Duration of the episode varies from 1 to 11 days with mean duration of 3.93 (1.94 SD) days. The chronic disease is the significant predictor for the duration of the episode. The data show that fever and swelling at inguinal regions are most common symptoms.
Conclusion
The incidence, frequency and duration of ADL episodes in this community are similar to that of other endemic areas. As the loss due to these ADL episodes is substantial, it should be considered while further estimating the burden due to lymphatic filariasis. The disability and loss caused by chronic forms of filariasis is higher, and the additional incapacity caused by the ADL episode, majority of which occur among chronic filariasis patients, further poses the burden on individuals and their families. Hence, morbidity management measures to prevent ADL episodes among endemic communities are to be implemented.
==== Body
Background
Lymphatic filariasis (LF) is associated with a wide range of clinical signs, symptoms and sequelae, which are influenced by a variety of factors related to host and parasite. Acute episodes of adenolymphangitis (ADL) is one of the symptoms and this acute clinical manifestation is characterised by recurrent attacks of fever associated with inflammation of the lymph nodes and or lymph vessels [1]. The importance of acute clinical manifestations, i.e. ADL, in natural progression of the disease, particularly the development of chronic disease has been recognised by filarialogists [2-4]. Though the need of systematic epidemiological studies on acute LF or ADL is recognised, a few studies have been undertaken in different endemic areas [5-8]. The epidemiological information is useful to estimate the burden of the disease and to develop morbidity management strategies. The present paper, based on a longitudinal prospective fortnightly surveillance, reports some epidemiological aspects of ADL episodes among rural communities of Orissa, India, which is endemic for LF caused by Wuchereria bancrofti.
Methods
Study area
This one year longitudinal prospective surveillance for acute ADL episodes was carried out in two villages in Khurda district of Orissa, India. These two villages are close together by less than two kilometres and the geographical coordinates of these villages are 20° 11' N and 85° 40' E, and 20° 10' N and 85° 38' E. The study area is known for its endemicity for LF caused by Wuchereria bancrofti, which is transmitted by Culex quinquefasciatus. The microfilaria rate and density are 9.4% and 769 mf/ml respectively [9]. The total disease and chronic disease rates are 12.5% and 7% respectively [10]. The prevalence of infection based on circulating filarial antigenaemia is 39.9% [11]. One round of mass drug administration of diethylcarbamazine was given in the district in 1997 and recorded around 30% of treatment compliance (BV Babu, unpublished data). However, no data specific to these two villages are available. The area is rural and its inhabitants are mostly small farmers and daily wage labourers.
Data collection
A population of 1329 (685 males and 644 females) of two villages were monitored for one year, during March 2000 – February 2001. Initially, the census of the villages was conducted for demographic information and identities of all individuals. The investigators visited all the households for every fortnight to detect individuals who were suffering or had suffered from an acute attack during that fortnight. In the present study, an acute ADL episode was defined as the presence of local signs and symptoms such as pain, tenderness, local swelling and warmth in the groin, with or without associated constitutional symptoms such as fever, nausea and vomiting [12]. The investigators, along with paramedical staff explained in the local language the symptoms of acute episodes, which are common in this endemic population. As there are no other diagnostic tools to identify ADL episodes, the present method of symptomatic diagnosis was used. This method of diagnosis through local terminologies was found to be highly specific (specificity = 0.980) and sensitive (sensitivity = 0.978) for diagnosing ADL [5]. For individuals identified as affected with acute attacks during last fortnight, the details including clinical symptoms, duration, etc. were recorded, by recall method and in case of ongoing episodes during visits, they were monitored. Majority of these cases are examined during the episode by physicians of local health institution. The available medical records/prescriptions were used as adjunct. A few episodes are clustered in members of the same household. The affected individuals were tested for microfilaraemia. A finger-prick thick blood film was prepared, using blood collected after 10 o'clock night, stained with Leishman's stain and entire film was examined for microfilariae.
Data analysis
The data were analysed through SPSS V.8 for Windows. The variation in incidence, frequency, number and duration of ADL episodes were assessed by employing the following statistical tests. Standardised normal Z-tests were performed to assess the variation in the incidence between different groups. Differences in number of episodes and duration of episodes were assessed by t-tests and analysis of variance (ANOVA). Multivariate ANOVA by regression was used to examine the effect of gender, age and pathological condition on the number and duration of episodes in affected individuals. The seasonal variation in the frequency of ADL episodes was assessed by χ2 test.
Results
Incidence and distribution of acute episodes
Number of fortnightly rounds completed during one year of study is 25 and 113 episodes were encountered among 72 patients in both villages. The annual incidence per 1000 population is 92.0 and 77.6 among men and women respectively, and the overall incidence is 85.0 per 1000 population. The difference between male and female is not significant (P > 0.05). However, the annual incidence is significantly high (P < 0.001) among individuals with chronic disease (957.0) than those without any chronic manifestations (19.4). The age-wise incidence of ADL episodes indicates a steady rise in the incidence of episodes along with age among males and females, except in the age group of 51–60 years (Fig. 1).
Figure 1 Relation between annual incidence of ADL episodes and different age groups among male, female and total population.
A total of 72 (52 males and 20 females) individuals of the study population (5.42%) were affected with acute ADLs. The age of the affected individuals varies from 7 years to 75 years with mean age of 36.23 ± 16.46 (SD) in case of male and 45.30 ± 12.15 (SD) in case of female, with a significant difference between two genders (P < 0.05). Among the total 72 patients suffered from acute episodes, overt chronic symptoms were found among 53 patients (73.61%). Out of total 113 episodes, 47 (41.59%) episodes are associated with lymphoedema, 38 (33.63%) with hydrocele, 4 (3.54%) with both lymphoedema and hydrocele and 24 episodes (21.24%) in individuals without chronic manifestations (Table 1). The mean age of patients with chronic manifestations (40.11 ± 15.22 SD) is slightly higher than individuals without any chronic manifestations (34.95 ± 17.68 SD), but the difference is not significant (P > 0.05).
Table 1 Association of gender, pathology group and age group with mean number of episodes and duration of episodes and details of multivariate ANOVA.
Independent Variable Total number of patients
in each category Total number of episodes
in each category Mean number of episodes Mean duration (days) of episode
Mean ± SD Coefficient (P)a Mean ± SD Coefficient (P)b
Gender
Male 52 63 1.21 ± 0.57 1.169 (0.000) 3.95 ± 2.05 0.508 (0.221)
Female 20 50 2.50 ± 1.67 3.90 ± 1.79
Pathology group
Hydrocele 31 38 1.23 ± 0.67 0.276 (0.071) 4.56 ± 2.34 0.505 (0.034)
Lymphoedema 19 47 2.47 ± 1.71 3.74 ± 1.76
Both 3 4 1.33 ± 0.58 3.67 ± 1.15
None 19 24 1.26 ± 0.56 3.52 ± 1.62
Age group
7–20 years 10 16 1.60 ± 1.08 0.002 (0.768) 4.38 ± 1.82 0.023 (0.0557)
21–40 years 30 40 1.33 ± 0.84 4.24 ± 2.41
41–60 years 24 42 1.75 ± 1.45 3.60 ± 1.51
61–75 years 8 15 1.88 ± 1.25 3.64 ± 1.80
Total 72 113 1.57 ± 1.15 3.93 ± 1.94
aFor number of episodes: F = 9.317 (P < 0.001), Adjusted R2 = 0.260
bFor duration of episodes: F = 2.607 (P < 0.05), Adjusted R2 = 0.041
Frequency and duration of acute episodes
As 113 episodes were recorded among 72 patients, 20 (27.8%) patients experienced more than one episode. Majority of patients (52; 72.2%) experienced acute episode only once in the year. It is followed by 12.5% of patients with two episodes, 6.9% of patients with 3 episodes, 5.6% of patients with 4 episodes, and 1.4% of patients each with 5 and 7 episodes per year. Among the individuals with multiple episodes, majority (16 out of 20) are with chronic filarial symptoms (P < 0.001). The average number of episodes per year is 1.57 (1.15 SD) per affected individual. The multivariate ANOVA employed to assess the influence of age, gender and pathology on number of episodes per affected persons indicated that only gender is identified as significant predictor to the number of episodes per affected individual (P < 0.001) (Table 1). Similar analysis is carried out to assess the effect of age among hydrocele patients, and age and gender among lymphoedema patients. These variables have no impact on frequency of episodes among hydrocele and lymphoedema patients (P > 0.05). There is significant variation in the frequency of ADL episodes with seasons of the year (P < 0.001). The average numbers of episodes per month during summer (March to June), rainy season (July to October) and winter (November to February) are 11.00, 10.75 and 6.50 respectively.
Duration of the episode varies from 1 day to 11 days and the mean duration is 3.93 days (1.94 SD). In majority of cases, the episodes persisted for 3 days (26.4%), followed by 4 days (22.2%), 2 days (15.3%), 8 days (9.7%), 5 days (8.3%), 1 day (5.6%), 6 days (5.6%), 7 days (5.6%) and 11 days (1.4%). When individuals' age, gender and pathology were considered simultaneously in multivariate ANOVA, the individuals' pathological condition emerged as a significant predictor to the mean duration of episode (P < 0.05) (Table 1). When the hydrocele and lymphoedema patients are considered separately, age in hydrocele patients and, age and gender in lymphoedema patients have no significant impact on the duration of episode (P > 0.05).
Clinical symptoms associated with ADL episodes
Distribution of various clinical symptoms among these acute cases is presented in Table 2. Occurrence of fever (77.87%) and swelling of inguinal regions (63.72%) are most common symptoms of acute cases. In the local language, swelling of lymphatic nodes is known as Bagi or Pichhuli, and is commonly perceived as the initial symptom of LF among the study population. Pain and tenderness in different body parts are seen in majority of cases. Presence of local swelling, anorexia, nausea and vomiting are associated with ADL among majority of cases. It is attempted to examine the differences in clinical presentation across gender and pathological groups. There are no significant variations in occurrence of these symptoms except in occurrence of pain and tenderness. Pain and tenderness in genitals is common in hydrocele patients, and hence more frequent in male patients. In lymphoedema patients, pain and tenderness are noticed in lower limbs. No association was found between occurrence of these symptoms and age groups. The microfilarae data indicated that only 8, out of 72 acute patients (11.11%) are microfilaraemic.
Table 2 Distribution of associated symptoms among the ADL cases.
Symptom Prevalence among episodes Percentage
Pain in
Upper arms 12 10.62
Legs 83 73.45
Genitals 29 20.00
Breasts 1 0.88
Tenderness in
Upper arms 10 8.85
Legs 13 11.50
Genitals 24 21.24
Lymphangitis in
Left inguinal region 34 30.09
Right inguinal region 38 33.63
Left papillary 3 2.65
Right papillary 3 2.65
Swelling 92 81.42
Fever
High 61 53.98
Low 27 23.89
Anorexia 72 63.72
Nausea 61 53.98
Vomiting 32 28.32
Microfilaraemiae 13 11.50a
apercentage is to total number of patients
Discussion
To develop strategies for relieving and preventing the suffering of filarial patients, it is essential to understand the epidemiology of various forms of morbidity. The information on acute form of LF, particularly on ADL is sporadic from a few endemic regions. The present longitudinal prospective surveillance showed the annual incidence of 85.03 of ADL episodes per 1000 population in this rural Eastern Indian community. Similarly, a South Indian rural community, which is endemic for bancroftian filariasis recorded the annual incidence of 96.3 per 1000 population [6]. The incidence rates are available from a few other endemic countries such as Ghana (96 per 1000 population) [5] and Tanzania (33 per 1000 population) [8]. The proportion of people affected with ADL episodes in the present study (5.4%) is similar to that reported from another bancroftian filarial endemic community from South India (5.3%) [6]. The age wise distribution indicates that male recorded higher incidence in all age groups except in the age groups of 41–50 years and above 60 years of age. This age-wise incidence follows the pattern with the prevalence of microfilaraemia and chronic disease [13]. The variation could be due to differences in prevalence of chronic disease and differential susceptibility to ADL episodes for individuals with lymphoedema and hydrocele. In the present population, men recorded higher prevalence of chronic filarial disease than women [10]. The mean age of affected individuals is varying between male and female. The lower mean age of male patients may be due to occurrence of hydrocele even at lower age. In the present study population, majority of affected individuals has only one episode per year and only 27.8% of affected individuals have experience more than once. The data showed that these multiple episodes are more common in patients with chronic disease. The mean number of episodes per year is significantly higher among lymphoedema patients than even hydrocele patients. A study from South India reported a direct relationship between the number of acute attacks and the grade of lymphoedema [14]. It is also known that the frequency of these attacks is generally higher in bancroftian filariasis as compared to brugian filariasis [14,15]. Regarding the seasonal variation in the frequency of ADL episodes, it is lower in winter than is summer and rainy season. Some of the earlier studies reported higher frequency in rainy season [16,17]. In Tanzania, the higher incidence of ADL episodes during rainy season is related to increased transmission by infective mosquito bites [8]. It is consistent with the hypothesis that ADL episodes may be associated with allergic responses to massive parasite antigen release [18,19]. There is uniformity in the associated symptoms of ADL episodes amongst various endemic communities, though the etiology of the ADL is not clear. The majority of the ADL patients in this study are amicrofilaraemic, and this finding is in conformity with earlier observations [5,19,20]. The limitation of this study is that an insensitive method of microfilaria detection was used, i.e., finger prick method as opposed to Nucleopore filtration method.
The incidence and duration of the ADL episode has greater economic implication on individuals, their families and community. There will be substantial loss of work during these episodes and subsequent economic loss [21-23]. In the present study population, on average each episode persists for 4 days, which is similar to the other endemic areas [6,8,16,24,25]. The magnitude of loss due to ADL episodes is substantial and it will be constituted as a major proportion to the total burden of LF. While estimating the global burden of LF, only the chronic forms of disease were considered [26], perhaps due to lack of data at that time. As the data on incidence and duration of ADL episodes are available at least from some endemic areas, they should be considered during further estimation of disease burden due to LF. In the present study community, it is reported that the chronic condition is posing considering economic burden due to loss of work and treatment costs [27]. It is well known that the disability and economic loss caused by chronic filariasis is life long and much higher. And additional incapacity caused by the ADL episodes, majority of which occur in these chronic filarial patients, further poses the burden on the individuals and their families. It is clear that there are two types of ADL episodes, ADL secondary to bacterial or fungal infections and ADL caused directly by the parasite infection itself [1]. For the episodes among chronic lymphoedema cases, secondary bacterial infection may be plausible explanation [28-30]. It is also evident from recent findings that simple foot hygiene and prevention of secondary bacterial infection lower the incidence of ADL episodes [28,31-33]. Shenoy et al. [34] demonstrated that well designed programme of foot care significantly decreases the frequency of ADL episodes. In this programme, meticulous hygiene including use of foot wear, regular washing of affected limbs, etc. to prevent injuries and infections needs to be incorporated. Thus, it is essential to develop and promote simple, cost-effective and user-friendly measures to minimise the burden of acute disease of LF.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BVB conceived and designed the study; performed analysis and interpretation of data; and drafted the paper. ANN collected and computerised the data; KD collected the data. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This study has been carried out with intra-mural funding of Regional Medical Research Centre (Indian Council of Medical Research), Bhubaneswar, India. Authors are grateful to Dr. S.K. Kar, Director, Regional Medical Research Centre, for his encouragement for research and publications. Authors are also grateful to Prof. J.W. Kazura, Dr. K.D. Ramaiah and Dr. P.K. Das for their comments as reviewers for refinement of the manuscript.
==== Refs
Kumaraswami V Nutman TB The clinical manifestations of lymphatic filariasis Lymphatic Filariasis 2000 London: Imperial College Press 103 126
Turner LH Studies on filariasis in Malaya: the clinical features of filariasis due to Wuchereria malayi Trans Roy Soc Trop Med Hyg 1959 53 154 169 13647621 10.1016/0035-9203(59)90066-5
Edeson JFB Wilson T The epidemiology of filariasis due to Wuchereria bancrofti and Brugia malayi Ann Rev Entomol 1964 9 245 268 10.1146/annurev.en.09.010164.001333
Raghavan NGS Clinical manifestations and associated epidemiological factors of filariasis J Commun Dis 1969 1 75 102
Gyapong JO Gyapong M Adjei S The Epidemiology of acute adenolymphangitis due to lymphatic Filariasis in Northern Ghana Am J Trop Med Hyg 1996 54 591 595 8686777
Ramaiah KD Ramu K Vijay Kumar KN Guyatt H Epidemiology of acute filarial episodes caused by Wuchereria bancrofti infection in two rural villages in Tamil Nadu, South India Trans Roy Soc Trop Med Hyg 1996 90 639 643 9015500 10.1016/S0035-9203(96)90415-0
Alexander ND Perry RT Dimber ZB Hyun PJ Alpers MP Kazura JW Acute disease episodes in a Wuchereria bancrofti-endemic area of Papua New Guinea Am J Trop Med Hyg 1999 61 319 324 10463687
Gasarasi DB Premji ZG Mujinja PGM Mpembeni R Acute adenolymphangitis due to bancroftian filariasis in Rufiji district, South East Tanzania Acta Trop 2000 75 19 28 10708003 10.1016/S0001-706X(99)00090-X
Dash AP Mohapatra N Hazra RK Acharya AS Transmission dynamics of filariasis in Khurda district of Orissa, India Southeast Asian J Trop Med Pub Health 1998 29 137 140 9740288
Babu BV Acharya AS Mallick G Jangid PK Nayak AN Satyanarayana K Lymphatic filariasis in Khurda district of Orissa, India: an epidemiological study Southeast Asian J Trop Med Pub Health 2001 32 707 710 12041543
Sahoo PK Geddam JJB Satapathy AK Mohanty MC Ravindran B Bancroftian filariasis: prevalence of antigenaemia and endemic normals in Orissa, India Trans Roy Soc Trop Med Hyg 2000 94 515 517 11132379 10.1016/S0035-9203(00)90070-1
WHO Lymphatic Filariasis: Disease and its Control. Fifth Report of the WHO Expert Committee on Filariasis Technical report series No 821 1992 Geneva: World Health Organization
Brabin L Sex differentials in susceptibility to lymphatic filariasis and implications for maternal/child immunity Epidemiol Inf 1990 105 335 353
Pani SP Yuvaraj J Vanamail P Dhanda V Michael E Grenfell BT Bundy DAP Episodic adenolymphangitis and lymphoedema in patients with bancroftian filariasis Trans Roy Soc Trop Med Hyg 1995 89 72 74 7747314 10.1016/0035-9203(95)90666-5
Pani SP Krishnamoorthy K Rao AS Prathiba J Clinical manifestation in malayan filariasis with special reference to lymphoedema grading Ind J Med Res 1990 81 200 207
Rao CK Chandrasekharan A Cherian G Frequency and duration of acute filarial attacks in persons in Brugia malayi endemic community Ind J Med Res 1982 75 813 815
Ravindranathan TC Sethumadhaya KVP Roychoudhary SP Babu CS Kanoujia KH Narasimham MVVL Rao CK Seasonal variation in microfilaraemia (bancrofti) density and filarial disease J Commun Dis 1980 12 73 77 7031121
Kar SK Mania J Kar PK Humoral immune response during filarial fever in bancroftian filariasis Trans Roy Soc Trop Med Hyg 1993 87 230 233 8337737 10.1016/0035-9203(93)90505-K
Addiss DG Eberhard ML Lammie JP Filarial adenolymphangitis without filarial infection Lancet 1994 343 597 7906341 10.1016/S0140-6736(94)91547-4
Michael E Grenfell BT Bundy DAP The association between microfilaraemia and disease in lymphatic filariasis Proc Roy Soc London B 1994 256 33 40
Sabesan S Krishnamoorthy K Pani SP Panicker KN Man-days lost due to repeated acute attacks of lymphatic filariasis Trends in Life Sciences (India) 1992 7 5 7
Ramaiah KD Ramu K Das PK Kumar KN Pani SP Direct and indirect costs of the acute form of lymphatic filariasis to households in rural areas of Tamil Nadu, South India Trop Med Int Health 1998 3 108 115 9537272 10.1046/j.1365-3156.1998.00208.x
Babu BV Nayak AN Treatment costs and work time loss due to episodic adenolymphangitis in lymphatic filariasis patients in rural communities of Orissa, India Trop Med Int Health 2003 8 1102 1109 14641845 10.1046/j.1360-2276.2003.01146.x
Kimura E Penaia L Spears GFS Epidemiology of sub-periodic bancroftian filariasis in Samoa 8 years after control by mass treatment with diethylcarbamazine Bull World Health Organ 1985 63 869 878 3912078
Partono F The spectrum of disease in lymphatic filariasis Filariasis, Ciba Foundation Symposium No 127 1987 15 31
Michael E Bundy DAP Ottesen E Ramachandran CP Global burden of disease: prevalence of disease in lymphatic filariasis Parasitology 1996 112 409 428 8935952
Babu BV Nayak AN Dhal K Acharya AS Jangid PK Mallick G The economic loss due to treatment costs and work loss to individuals with chronic lymphatic filariasis in rural communities of Orissa, India Acta Trop 2002 82 31 38 11904101 10.1016/S0001-706X(02)00030-X
Shenoy RK Sandhya K Suma TK Kumaraswami V A preliminary study of filarial related acute adenolymphangitis with special reference to precipitating factors and treatment modalities Southeast Asian J Trop Med Pub Health 1995 26 301 305 8629065
Olszewski WL Jamal S Skin bacterial factor in progression of filarial lymphoedema Lymphology 1994 27 148 149 7807989
Olszewski WL Jamalm S Manokaran G Pani S Kumaraswami V Kubicka U Lukomska B Dworczynski A Swoboda E Meisel-Mikolajczyk F Bacteriologic studies of skin, tissue fluid, lymph and lymph nodes in patients with filarial lymphodema Am J Trop Med Hyg 1997 57 7 15 9242310
Shenoy RK Suma TK Rajan K Kumaraswami V Prevention of acute adenolymphangitis in brugian filariasis: comparison of the efficacy of invermectin and diethylcarbamazine, each combined with local treatment of the affected limb Ann Trop Med Parasitol 1998 92 587 594 9797832 10.1080/00034989859285
Burri H Loutan L Kumaraswami V Vijayasekaran V Skin changes in chronic lymphatic filariasis Trans Roy Soc Trop Med Hyg 1996 90 671 674 9015512 10.1016/S0035-9203(96)90428-9
Dreyer G Medeiros Z Netto MJ Leal NC de Castro LG Piessens WF Acute attacks in the extremities of persons living in an area endemic for bancroftian filariasis: differentiation of two syndromes Trans Roy Soc Trop Med Hyg 1999 93 413 417 10674092 10.1016/S0035-9203(99)90140-2
Shenoy RK Kumarswami V Suma TK Rajan K Radhakuttyamma G A double blind placebo controlled study of the efficacy of oral pencillin, diethyl-carbamazine or local treatment of the affected limb in preventing acute adenolymphangitis in lymphodema caused by brugian filariasis Ann Trop Med Parasitol 1999 93 367 377 10656038 10.1080/00034989958366
| 15904537 | PMC1156912 | CC BY | 2021-01-04 16:28:57 | no | BMC Public Health. 2005 May 19; 5:50 | utf-8 | BMC Public Health | 2,005 | 10.1186/1471-2458-5-50 | oa_comm |
==== Front
BMC PsychiatryBMC Psychiatry1471-244XBioMed Central London 1471-244X-5-211585049610.1186/1471-244X-5-21Research ArticlePosttraumatic stress disorder: An exploratory study examining rates of trauma and PTSD and its effect on client outcomes in community mental health Howgego Irene M [email protected] Cathy [email protected] Lenore [email protected] Peter [email protected] Frances [email protected] Ruth [email protected] Academic Unit of Psychological Medicine, Australian National University The Canberra Hospital, Australian Capital Territory 2605, Australia2 Medical Education Unit, Australian National University, Australian Capital Territory, 0200, Australia3 PO Box 198 Kenilworth Queensland, 4574, Australia4 Centre for Health Teaching, University of California, Davis, CA 95616, United States of America5 West End Mental Health Service Brisbane, Queensland, 4101, Australia6 Center for Mental Health Research, Australian National University, Australian Capital Territory, 0200, Australia7 Formerly at University of Queensland ST LUCIA Queensland 4072 Australia2005 26 4 2005 5 21 21 29 11 2004 26 4 2005 Copyright © 2005 Howgego et al; licensee BioMed Central Ltd.2005Howgego et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Rates of trauma and Posttraumatic Stress Disorder (PTSD) were examined in order to compare the profile in clients of an Australian Public Mental Health Service with that reported in the international literature for clients with major mental illness and to explore the effect of this on client health outcomes. Potential factors contributing to increased levels of trauma/PTSD in this group of clients and the issue of causality between PTSD and subsequent mental illness was also explored.
Methods
A convenience sample of 29 clients was screened for trauma and PTSD using the Posttraumatic Stress Diagnostic Scale™ (PDS) and selected outcome measures. Paired and independent samples t-test and ANOVA were applied to the data.
Results
High levels of undocumented trauma and PTSD were found. Twenty clients, (74%) reported exposure to multiple traumatic events; 33.3% (9) met DSM IV diagnostic criteria for PTSD. Significant difference was found for PTSD symptomatology, severity and impairment and for client and clinician-rated scores of Quality of Life (QOL) outcomes in the PTSD group. No effect for PTSD symptomatology on the Working Alliance (WA) was found. Factors that may influence higher rates of PTSD in this group were identified and included issues associated with the population studied, the predominance of assaultive violence found, and vulnerability and risks factors associated with re-traumatisation within the social and treating environments.
Conclusion
A similar trauma and PTSD profile to that reported in the international literature, including greater levels of trauma and PTSD and a poorer QOL, was found in this small sample of clients. It is postulated that the increased levels of trauma/PTSD as reported for persons with major mental illness, including those found in the current study, are primarily related to the characteristics of the population that access public mainstream psychiatric services and that these factors have specific implications for service delivery, and raise issues of efficiency and effectiveness of resource use in achieving successful outcomes in public mental health services for clients with co-morbid PTSD. Further research with a more rigorous design is needed to test these preliminary findings within Australian Community Mental Health Services.
==== Body
Background
PTSD is emerging as a major public health problem worldwide [1]. Recent epidemiological studies within Australia [2] and America [3] identified rates of PTSD within the general population as 1.3% (DSM IV criteria) 3.3% (ICD 10 criteria) and 7.8% (DSM-III-R criteria) respectively. Whilst this growing recognition of the prevalence of PTSD is stimulating research, activity in the broad spectrum of psychiatry is still limited. Despite the knowledge that high rates of trauma are associated with persons with mental illness few studies have examined trauma and PTSD in this population [4-6], and the majority of these stem from the United States of America [7]. PTSD research in mainstream psychiatry clearly highlights the complexity of trauma/PTSD in persons with a co-existing psychiatric illness [6,8]. The emerging profile showed a group of respondents with high levels of previously undetected trauma (51% – 98%) [6,9] and PTSD (22.2% – 66%) [9,10] who exhibited the phenomena of multiple traumatisation [5,6,10-14]. The primary type of trauma identified was interpersonal in nature (physical and sexual assault) and included both childhood and adulthood victimisation. The general lack of recognition and documentation of trauma and PTSD evidenced in a number of studies [4,6,9,11,12,14,15] coupled with the high rates of trauma and PTSD found, reflects the general consensus in the psychiatric literature that the problem is under diagnosed and potentially untreated in this population.
Whilst existing studies provide valuable data and insight into PTSD in persons with another mental illness, more needs to be done to determine the applicability of these findings outside of the American context. Australian PTSD research in conventional psychiatry is in its infancy with only one study to date reporting on PTSD in a psychiatric in-patient unit [4]. This showed clients with an undocumented trauma rate of 61% and PTSD rate of 28% based on DSM-III-R criteria [4]. PTSD was also found to be the incident disorder in 50% of clients and preceded major depression in 83% of these clients (n = 141).
Despite these findings, routine assessment of trauma (and therefore, diagnosis of PTSD) in persons presenting to Community Mental Health Services (CMHS) is often "overlooked" in the absence of PTSD symptomatology as the presenting complaint [15]. This has critical implications for clinical management as clients generally will not volunteer this information either from a reluctance to re-visit the trauma, fear of clinician response or simply not recognising the relevance of any prior trauma to their current problem [16]. Consequently, nationally and internationally, the recognition, diagnosis and treatment of PTSD in clients with comorbid psychiatric diagnoses is at best sporadic and poorly understood by the majority of clinicians [1,17-21]. Compounding this is the fact that conventional psychiatry and mental health service delivery in Australia has been slow to embrace concepts from the field of traumatic stress [17]. Leo Sher, [21] in correspondence published in an Australian psychiatric journal, concluded that "there is a pressing need to improve recognition and treatment of PTSD".
Implications for mental health services and client outcomes
These findings from the literature have important implications for successful outcomes in Mental Health for the individual, the clinician, the service provider and society in general. On the individual level PTSD is known (and acknowledged) to be co-morbid with a variety of other psychiatric disorders particularly mood and anxiety based problems [3,4,7,15,16,19]. These disorders represent the 'core business' of CMHS. Therefore, the potential exists for clients being treated in such services for any of these disorders, to have underlying PTSD symptomatology.
Untreated co-morbid PTSD in persons with another major mental illness is associated with important negative effects such as increased symptom severity for both diagnoses, increased hospitalisation, prolonged treatment and poorer overall health outcomes for the individual [4,6,8,16,19,22]. Additionally, the therapeutic alliance between client and clinician may also be compromised, further eroding the likelihood of achieving positive client outcomes [6]. This latter factor is of key clinical importance as the role of the relationship in achieving positive outcomes is well documented [23-26]. For the service provider, all of the above result in greater overall treatment and management costs and sub-optimal resource use. Ultimately, society bears both the social and financial cost of PTSD. This burden to both individuals and society is acknowledged by contemporary researchers [7,16] leading to the conclusion that PTSD is one of the most serious and disabling psychiatric disorders [1].
There are no clear causal links between PTSD and other mental illness. The contemporary literature discusses the interplay of risk factors such as personal and family psychiatric history, gender, ethnicity, and type of trauma [22,27] and hypothesises on causal pathways and co-morbidity with other Axis I diagnoses including psychotic-based illnesses [16]. However, the exact relationship between trauma/PTSD and mental illness, how this effect is mediated, and whether it differs diagnostically across the spectrum of mental disorders is poorly understood and requires greater research [8,28].
A model of Trauma and Mental Illness (Figure 1) proposed by Mueser et al. [8] demonstrates the potential for explanation and examination of these complex interactions. Within the context of this model, it is theorised that PTSD symptomatology plays a central role in the severity and course of mental illness through two mechanisms. First, directly from the 3 symptomatology clusters of Re-experiencing, Avoidance, and Hyper-arousal either collectively or individually (represented by the solid lines). Second, indirectly (represented by broken lines) through the effects of correlates of PTSD such as substance abuse and re-traumatisation, which further compromises the course and severity of symptoms. Additionally, the potential for a poor working alliance with clinicians resulting from all these factors, may lead to the patient receiving fewer preventative illness management services such as case management and/or medication management, thereby further compromising their status [8].
Figure 1 Interactive model of Trauma, PTSD and Severe Mental Illness [8] used with permission.
The following research questions arising from the literature review were explored in this study:
1. What was the trauma and PTSD profile for clients in an Australian CMHS?
2. How did this profile compare to that reported in the international literature for similar groups of clients?
3. Did the treating health professional know of this profile?
4. What impact did PTSD symptomatology have on
a. The Working Alliance between clients and their case managers
b. Client health outcomes?
5. What factors may be contributing to the higher levels of trauma and PTSD reported in the literature for this population?
6. What are the possible links between PTSD and other mental illness?
This paper discusses the exploratory results from a small (n = 27) sample of clients with predominantly Axis I diagnoses who were screened for trauma/PTSD at baseline (T1). All clients were receiving case management services for their primary disorder within a public-funded CMHS.
Methods
Design
The primary study used a prospective time-series design. Data was collected at baseline and planned for five time points (baseline screening and selected outcome measures at 1 month, WA baseline and the remaining outcome measures at 3 months with repeat of all measures at 6 months). However, this was reduced to three time-points due to client retention problems. The study was funded by the National Health and Medical Research Council (NHMRC) and was a joint project between the University of Queensland and Mental Health ACT (MHACT). Ethical approval was obtained from both institutions.
Setting
The research initially commenced in Brisbane, Australia and was re-located to the Australian Capitol Territory (ACT) in the early phase of implementation. MHACT serves a population of approximately 320,000 residents, 21% of whom experienced a mental health disorder compared to the national average of 18% [29].
The ACT research site is an integrated Public Mental Health Service. As such, the service structure included comprehensive community and hospital-based elements that provided preventative, acute and ongoing care, case management and rehabilitation services integrated within a single agency (including a shared filing system) in order to facilitate continuity of care and ease of access for clients to the various service elements. Other elements included a GP Liaison and Court Liaison service. Specialised PTSD services did not exist within the service. Four geographically diverse adult CMHS located throughout the ACT community were used as study sites.
Case management services included the provision and/or co-ordination of a wide range of agency and external services that targeted the client's primary psychiatric diagnosis and were tailored to individual client needs. Services included medical management, medication monitoring, rehabilitation, counselling services, group therapy, such as depression, anxiety, anger management, leisure/social activities, accommodation support, drug and alcohol services, financial management, legal services/representation and respite care and hospitalisation as needed.
Sample
Eligibility Criteria
Adult clients who were entering a 'new' therapeutic relationship with a case manager (CM) and were expected to receive case management service for 12–18 months in one of the four adult community mental health teams were eligible to participate. A 'new therapeutic relationship' was defined as that of a client who was either entirely new to the service or one who was being re-assigned to a new CM. No diagnostic restriction was placed on participants. However, the ability to respond to written and spoken English was essential; interpreter services were considered, however this was rejected as a threat to validity. Only one potential client was excluded on the grounds of being non-English-speaking.
Sample selection and procedures
Pre-Recruitment
Prior to the recruitment phase a series of information sessions detailing the aims and purpose of the research together with the recruitment and selection procedures were given for the CM's at each of the study sites.
Recruitment
A researcher (IH) attended weekly team allocation meetings (assignment of new clients to CM's) in an effort to aid recruitment, address any concerns regarding the study and generally facilitate the process. Recruitment occurred over a 9-month period.
Selection
Potential participants were selected consecutively from a random start date using the referral list to the teams. The CM initially approached clients about the study. Eighty-four clients from a pool of 211 met the selection criteria, 27 of these consented to participate in the study, a recruitment rate of 34.5%. Although case managers nominated a total of 211 client names, 74 of these did not meet the selection criteria, primarily the criterion relating to the term of case management. Clinical concerns expressed by the CM 's re the client's mental state excluded another 13 clients (CM's considered that approaching the client about the research would have an adverse impact on either the client's current mental state or the developing therapeutic relationship). As the relationship was the primary independent variable examined in the larger study, it was essential to have both CM and clients engaged in the study, therefore CM clinical judgments regarding the clients well-being were respected. Thirteen clients failed to engage with the service. Time from the initial contact (1 month) was critical to baseline data collection (relative to the interaction with WA) and 27 clients had not been approached by their CM regarding the research in this time period and were, therefore, lost' to the study.
Procedures
An information leaflet was given to interested clients (by their CM) and an appointment made for a researcher to follow-up on the written material. Explanations of the purpose of the study, issues of confidentiality and consent procedures were discussed with clients. Written consent was obtained from participants. Interviews were conducted by a research officer (I. H), and clients were paid $10 for each completed interview. The final sample size was twenty-nine patient-case manager dyads.
Instruments
Trauma/PTSD
Posttraumatic Stress Diagnostic Scale™ (PDS) [30]; a 49 item self-report scale based on the DSM IV[31] PTSD diagnostic criteria A-F.
The PDS assesses current PTSD symptomatology as experienced by the client 1 month prior to interview. The possible number of endorsed symptoms is 0–17 and the Symptom Severity score 0–51. Impairment is measured over 9 Functional Life Areas and defined as 0= No Impairment, 1–2 areas = Mild Impairment, 3–6 areas = Moderate Impairment, and 7–9 areas = Severe Impairment.
Although the scale was designed for self-completion, it was administered via interview to maintain internal validity by ensuring clients were responding to items measuring PTSD symptomatology and not potential overlap symptoms from their primary diagnosis. The interview was re- focused on the reported trauma as required.
Clients who identified multiple traumas were asked to nominate the 'one that bothered them the most' consistent with the administration requirements of the instrument. The PDS has demonstrated validity and reliability and is recommended as particularly useful when used for screening and assessing PTSD, especially in 'at risk' populations in clinical and research settings [30].
The CM's knowledge of the client's trauma profile was also sought and they were asked to complete Parts 1 and 4 of the PDS. Part 1 listed the traumatic event. CM's were asked to respond to this and indicate any event/s that they knew the client had experienced. Part 4 lists nine life areas that may be impaired as a result of the trauma. If CM's answered 'yes' to any trauma in Part 1, they were asked to respond to this and indicate if they knew whether the client's life had been effected by the trauma in any or all of the areas listed. Additionally, CM's were asked, if to their knowledge, the client was receiving any treatment for the trauma.
The Working Alliance Inventory-Client Form (WAI-C) [32]; a 36-item questionnaire that assesses the client's thoughts and feelings about their relationship with their clinician through the three principal components of the Working Alliance: Tasks, Goals and Bonds. The form is available in both self-completion and interview format, the latter was used in this study.
The Working Alliance Inventory – Case Manager Form (WAI-CM) [32]; a parallel measure of the client form that assesses the clinician's thoughts and feelings regarding their relationship with the client. Both these measures have demonstrated validity and reliability [32]. Permission was obtained from the scale designer to modify some wording to better reflect the clinical context of the research, that is, clinical case management as opposed to psychotherapy/psychology.
Quality of life, symptomatology and general functioning: these key outcomes were measured using the following:
The Australian World Health Organization Quality of Life-BREF (WHOQOL-BREF) [33]. Based on subjective measures of quality of life only, it consists of 24 items in 4 domains – Physical Health, Psychological Health, Social Relationships and Environment. Two items also measure overall quality of life and general health. Validity and reliability have been established [33]. CM's completed a modified version that reflected their perception regarding the client's Quality of Life (QOL). CM's were not asked to rate this from the client's perspective but as a clinical judgment of the client's QOL.
Health of the Nation Outcome Scale – Version 4 [34]; a clinician-rated measure of client health outcomes and useful for measuring improvement in client status over time; it is a 12-item scale with four sub-scales representing different problem areas of Behaviour, Impairment, Social Functioning and Symptom domains. Reasonable reliability and validity has been established within Australia and the UK [35,36]. Clients also completed the 'Health Questionnaire', the client equivalent of the HoNOS. Although this instrument is not in the formal suite of HoNOS measures, (as it did not progress past preliminary field trials) after consultation with the HoNOS UK centre, it was decided to use it in this study and attempt validation of the measure. The results of this validation will be presented elsewhere.
The Abbreviated Life Skills Profile (LSP-16) [37,38]; a 16-item scale developed from the original Life Skills Profile. Reliability and validity have been established.
This measure together with the HoNOS is recommended for use by service providers as part of outcome measurement initiatives in the Australian National Mental Health Strategy [38].
Administration
A major aim of the larger study was to gain both client and CM perspectives; therefore, with the exception of the LSP-16 all measures had both client and CM versions. Where these did not already exist, permission was sought from the scale designer for any modifications undertaken.
Client measures
All client measures except the Health Questionnaire were by face-to-face interview that was conducted by a single researcher (IH). The Health Questionnaire was originally designed for self-completion and was administered in this mode as part of the validation process.
CM measures
these were presented as a suite of questionnaires in booklet form and provided to the CM prior to the client interview. CM's were notified of the date of the interview in writing and were requested to complete the questionnaires within a maximum 5 day time frame either side of the client interview, that is, 10 days in total, in order to improve reliability and validity of clinician and client score comparisons.
Preparation
A series of familiarity sessions, which included a sample suite of client and clinician measures, and a 'hot-line' contact for any subsequent queries, was provided for each team. The HoNOS was the only instrument that required specific training prior to use; as this is part of the validation process (methods section) the details are not included here.
Data analysis
Data was collected from June through December 2001. The Statistical Package for Social Sciences (SPSS) – versions 10 – 11.5 software package was used to design the database and conduct statistical analysis; data were entered on a personal computer.
Statistical analysis
Paired and independent samples t-tests were first conducted to compare mean symptom counts for those with and without a PTSD diagnosis. In the sample studied there was overlap of some symptom counts for those with and without a PTSD diagnosis. In addition, for both groups, PTSD symptom counts were highly skewed. In skewed distributions measures of the mean can be relatively uninformative when used as a summary measure for comparing groups. Therefore, respondents were grouped into four subgroups: those without PTSD diagnosis whose symptom counts were in the lower 50% or upper 50% for that subset and those with PTSD diagnosis whose symptom counts were in the lower 50% or upper 50% for that subgroup. These analyses gave a more informative picture of the overall symptom presentation of those with and without a PTSD diagnosis. ANOVA were applied to this data. One sample-t-test was used for comparison with other population data. Eta-squared was used to calculate effect size in the ANOVA's.
Data relevant to outcome is based on the subjects who were still in the study at Time 2 (n = 17). Intention to treat analysis was not used as the study did not use randomised controlled methods and was an observational study only.
Results and discussion
Attrition
The attrition rate for the sample was 37%. Ten client/CM dyads had exited the study by the 12-month data collection point; the primary reasons for these exits were clinical and systems-based. Clinically, seven client's no longer needed case management services, therefore, were 'closed', that is, were no longer 'active' clients of the service. A further three clients changed case managers as a result of staff movements (either that of leaving the service or changing role functions) after a critical point in the research period and, therefore, exited the study as the effect for WA on outcome was not able to be measured for these clients.
Response rate
The low responses rate of 34% was primarily a function of the large number of clients that were nominated by case managers that did not meet selection criteria. As noted in the methods section, the criterion that presented the most difficulty was that of the duration of case management service required. For the study, this was a mid- to long-term period to allow for the tracking of the therapeutic relationship (the primary variable being measured in the larger study). As clients were selected on entry to the service, CM's were required to make a judgment as to the likely term of service required. This was not always easy given the acuity of the psychopathology involved. Furthermore, in the reality of the clinical setting, the selection criteria were not uppermost in the minds of the clinicians referring the clients to the study. Attempts to address this were undertaken during the recruitment phase, for example, a large poster listing the selection criteria was designed for each team, with the request that it be prominently displayed at team meetings in which clients were allocated to case managers to aid referral. This was in addition to the planned strategies discussed in the methods section to facilitate the recruitment process. Systemic issues also impacted on the issue of eligibility as some clients referred to the study had been assigned to an interim CM only; this again made them ineligible for the purpose of measuring the therapeutic relationship as they would be changing CM's within 3 months or less.
Data
Data on trauma/PTSD is reported for a sample group of 27 client/CM dyads (two clients withdrew their consent during baseline data collection; these data are omitted). Where applicable, data are provided for the total sample followed by the discrete results for subgroups with PTSD (n = 9) and those reporting trauma that did not meet diagnostic threshold (n = 11) total n = 20.
Demographics: client
Presented in Table 1. Note that data for CM contacts and times is at 12 months and therefore only includes data for the 17 clients remaining at this point.
Table 1 Client Demographic details for the whole sample in sub-groupings for trauma/PTSD (n = 27)
Variable PTSD n = 9 Trauma/no PTSD n = 11 No Lifetime Trauma n = 7
Gender M/F 5/4 5/6 4/3
Mean Age 34(SD = 12.2) 39(SD = 11) 37(SD = 14)
Country of Birth
Australia 5 9 6
Indigenous 0 0 0
Overseas 4 2 1
Highest educational level
Completed Secondary -Junior Level 5 4 4
Completed Secondary- Senior Level 3 6 1
Tertiary 1 1 2
Marital status
Single never married 7 7 4
Married 1 2 1
Divorced/Separated 1 2 1
Widowed 0 0 1
Employment Status
Employed 3 0 1
Unemployed 5 11 4
Student 1 0 0
Homemaker 0 0 2
Main Source of Income
Government Payment 8 11 6
Private 1 0 1
Primary Diagnostic Profile
Schizophrenia – various types 4 7 5
Borderline Personality Disorder 4 1 0
Depression 1 1 1
Bipolar Affective Disorder 0 2 1
Mean age of onset for primary diagnosis 25(SD = 12.5) 25(SD = 11.2) 25 (SD= 7.9)
Dual Diagnosis (co-occurrence substance misuse) 7 5 1
Mean number years with Tx service 5 (SD= 3.4) 8 (SD= 3.2) 8 (SD= 2.6)
Persons subject to involuntary mental health order at time of study 0 1 3
Mean number recorded contacts with CM over 12 months 29 (SD = 7) n = 5 21 (SD7.5) n = 7 33 (9) n = 5
Mean time in hours of contacts 17.5(SD= 6.5) 13(SD = 6.3) 14.6(SD = 3.2)
Table 2 Categories of traumatic events from the PDS screen as reported by clients (n = 20)
TRAUMATIC EVENT PTSD N = 9 NO PTSD N = 11
No. % No. %
Serious Accident, fire or explosion 7 78% 4 36%
Natural disaster 1 11% 0
Non-sexual assault – family member or someone known 6 67% 4 36%
Non-sexual assault – stranger 3 33% 3 27%
Sexual assault – family member or someone known 5 55.5% 3 27%
Sexual assault – stranger 3 33% 2 18%
Military combat/war zone 1 11% 1 9%
Sexual contact under 18 years with someone 5 or more years older 7 78% 4 36%
Imprisonment 3 33% 2 18%
Torture 1 11% 0
Life-threatening Illness 4 44% 2 18%
Other type of event 4 44% 7 63%
Co-Morbidity levels
Co-morbidity was defined as at least one other diagnosis (excluding PTSD). The rate of co-morbidity in the whole sample (n = 27) was 30% (n = 7) with the majority having one other diagnosis; PTSD group n = 9: 44% (n = 4) all of whom had at least two other diagnoses; non-PTSD group n = 11: 17% (n = 3) all of whom had only one other diagnosis.
Demographics: case manager (n = 17)
Data is adjusted for CM's with multiple clients in the study. The majority of CM's were male (71% n = 12). The mean age was 42 years (SD= 10) range 35. The respondents had a mean of 4 years (SD = 1.2) clinical experience plus a mean of 2.5 years in mental health (SD = 1.4). Nurses comprised the largest professional discipline (71% n = 12). The remainder were Clinical Psychologists (4) and one Social Worker. The majority of CM'S had an undergraduate qualification only (59%), four had a specialised qualification in psychiatric nursing and three CM's had a higher degree at the masters' level. Fifty-three percent of respondents had post-graduate training in a related field (CBT, Counselling, Psychoeducation).
Discussion: client demographics
Levels of employment
These are high at 70% in this sample, despite the high level of respondents who completed high school. The unemployment rate in the whole of the adult service of the ACT CMHS is recorded as 34%, however, this is not adjusted for acuity or diagnosis, therefore, as the sample has a high percentage of clients with a psychotic-based illness, direct comparison cannot be made.
The high educational level reported may reflect the general profile for the ACT, which has a higher level of retention of students in secondary school with 89% of year 7 students remaining in school at year 12 compared to the national rate of 73% [29]. However, place of schooling was not a demographic that was included in the screen, therefore, this is not a definitive explanation of reported educational levels in the sample.
Australia has six state and two territory governments and a federal level government, with differing levels of political responsibility for services. That of health service delivery rests at the state and territory level, and there is currently no central data collection point for employment status for persons with a mental illness in a treatment context. However, two national surveys [39,40] provide some insight into this, each reporting higher rates on unemployment (approximately 30%) across genders for persons with a mental illness. A further survey on employment in persons with psychosis [41] reported that 85% (n = 980) had their main source of income from government payments. The general unemployment rate for the ACT is 5% compared to that of 6% nationally [29].
The issue of unemployment in person with a mental illness in Australia is complex, and influenced by several factors such as access to employment-related services, stigma amongst employers and society generally, and systemic issues such as availability of vocational and rehabilitation services in mental health service delivery systems. Underpinning all of these factors is the influence of a federally-funded social welfare system that supports unemployed persons to varying degrees [29,39-41].
Primary psychiatric diagnosis
This was the major differentiating demographic factor between the two groups, with a greater number of clients in the non-PTSD groups diagnosed with schizophrenia. Whilst Mueser [6] found that diagnosis was the only demographic variable associated with a diagnosis of PTSD, the current sample is too small to draw any conclusions. Given the small body of research in this setting, it is difficult to be definitive about associations between comorbid psychiatric diagnoses and PTSD. As discussed below, there are many complex issues involved in any attempts to infer causality between trauma/PTSD and the development and course of other mental illnesses [8,16] and the area requires greater research.
Data: trauma/PTSD profile
Rates of trauma and PTSD n = 27
Twenty clients (74%) reported exposure to at least one traumatic event; seven clients (26%) reported no experience of a traumatic event in their lifetime. Sixty-seven percent of respondents identified multiple traumatic events; two reported exposure to a single event only. The mean number of events reported was four (SD = 2.3). Nine clients from the total sample (33%) met diagnostic criteria for PTSD. Eleven clients (41%) reported trauma symptomatology that did not meet diagnostic threshold for current PTSD. Only one patient had a formal diagnosis of PTSD in their medical record.
Please note
The data pertinent to trauma symptomatology detailed below relates to the 20 persons who reported experiencing a traumatic event in their lifetime, this includes those meeting diagnostic criteria for PTSD (n = 9), and those reporting trauma but who did not meet diagnostic criteria for PTSD (n = 11). The 7 clients who did not report any lifetime trauma are excluded from this section, but are included in the subsequent outcome section.
Number of events (n = 20)
The total number of traumatic events reported by clients was 77; the CM reported a total of 21. Paired samples t-test showed significant difference for the mean number of reported events between CM M = 1.0, SD = 1.12 and Clients M = 3.8, SD = 2.4 t (19), p< .0005.
PTSD group (n = 9)
clients reported a total of 45 traumatic events (M = 5, SD = 2.95); the CM reported a total of 14 (M = 1.55, SD = 1.13) or 31 % of the events reported by the patient. The majority of CM's (78% (7)) – knew about the patient's exposure to trauma; two CM's (22%) reported no knowledge of the patient's trauma. Only one CM reported that the client was receiving treatment for trauma/PTSD.
Non- PTSD group (n = 11)
clients reported a total of 32 events (M = 3, SD = 1.30); CM's reported a total of 6 (M = .54, SD = .93) events or 19% of those reported by the client. The majority of CM's 64% (7) had no knowledge of the client's reported exposure to trauma.
Type of event
Table 2 shows the reported events. The three most frequently were Serious Accident, Physical Assault and Sexual Assault, however, the combination of the different types of sexual assault (childhood and adulthood) makes this the most frequently reported type of trauma across the groups. Fifty-five percent (11) of clients reported sexual assault before the age of eighteen, of these 58% (7) also reported adult sexual assault.
Table 3 Itemised PTSD symptoms for Criteria B, C, D, reported at 2–5 times per week
SYMPTOM CLUSTER PTSD GROUP N = 9 NON-PTSD GROUP N = 11
B: Re-experiencing
Upsetting thoughts or images 55.5% (5) 9.1 %(1)
Bad dreams or nightmares 55.5% (5) 0
Reliving the traumatic event 44.4% (4) 18.2% (2)
Feeling emotionally upset when reminded of the event 66.6% (6) 36.4% (4)
Experiencing physical reactions when reminded of the event 33.3% (3) 27.3% (3)
C: Avoidance
Trying not to think, talk, or have feelings about the event 66.6% (6) 9.1 %(1)
Trying to avoid activities, places or people that recall the event 55.5% (5) 0
Unable to remember important part of the event 11.1%(1) 9.1 %(1)
Having less interest in important activities 66.6% (6) 9.1 %(1)
Feeling distant or cut-off from people 77.8% (7) 36.4% (4)
Feeling emotionally numb 44.4% (4) 27.3% (3)
Feeling that future plans will not come true 44.40/0 (4) 27.3% (3)
D: Arousal
Having trouble sleeping 33.3% (3) 18.2% (2)
Feeling irritable 66.6% (6) 27.3% (3)
Having trouble concentrating 55.5% (5) 27.3% (3)
Being overly alert 55.5% (5) 27.3% (3)
Being jumpy or easily startled 44.40/0 (4) 9.1 %(1)
Table 4 Number and severity of symptoms for the 3 clusters – Median-split
CLUSTER ITEM NUMBER IN MEAN SD P*
Lower 50% Upper 50%
PTSD
Re-experiencing (B)
Number of symptoms Md: ≤ 3 Yes 2 7 3.67 1.41
No 6 5 2.18 1.94 0.072
Severity score Md: 4.5 Yes 3 6 7.11 4.43
No 7 4 3.36 3.04 0.038
Avoidance (C)
Number of symptoms Md: 3.5 Yes 2 7 5.11 1.36
No 8 3 3.00 2.10 0.018
Severity score Md: 6.5 Yes 3 6 10.78 5.43
No 7 4 4.82 4.19 0.013
Hyper-arousal (D)
Number of symptoms Md: ≤ 3 Yes 5 4 3.78 1.20
No 8 3 2.27 1.74 0.041
Severity score Md: 4.0 Yes 4 5 7.78 4.66
No 6 5 4.60 3.66 0.115
Total number of symptoms endorsed Md: <14.5 Yes 2 7 12.67 3.39
No 8 3 7.18 5.72 0.021
Total symptom severity score Md: 9.5 Yes 3 6 25.67 13.77
No 7 4 12.36 9.60 0.020
* Derived from ANOVA analysis
PTSD group n = 9
78% (7) of the respondents reported childhood sexual assault; 53% (5) also reported adult sexual assault (3 females, 2 males). Non- PTSD group n = 11: 36% (4) reported childhood sexual assault; of these, 50% (2) also reported adult sexual assault (1 female, 1 male).
Trauma with the most effect
Interpersonal assault, physical and sexual, was the type of trauma nominated by 45% (9) of respondents as the type of trauma that 'bothered them the most'. The second highest rating trauma nominated was that of a 'serious accident or explosion' the remaining categories were unequally distributed amongst the remaining types of trauma. Eleven clients (55% n = 20) reported that the event nominated by them as the one that 'bothered them the most' happened more than 5 years ago.
Symptom profile
Symptom details for the B C D criteria are shown in tables 3 and 4 .
Symptomatology specifiers (E criterion)
All clients in both groups met the chronic symptom duration criteria (all had experienced the symptoms for more than 3 months). PTSD group n = 9: The majority of clients (89% (8)) in this group had 'acute' onset of symptoms, with only one client showing 'delayed' onset. Non- PTSD group n = 11: In contrast, these clients were almost equally divided between the acute and delayed onset categories.
Impairment of functioning (F criterion)
PTSD group n = 9
The majority of respondents – 67% (6) – met the 'severe' impairment criteria (7–9 functional areas of life effected). Two respondents met the criterion for 'moderate' impairment (3–6 areas effected); and one demonstrated 'mild' impairment (1–2 areas effected).
Non- PTSD group n = 11
Forty-five percent of respondents (5) showed no impairment; thirty-six percent (4) met the 'severe impairment' criteria, with one respondent in each of the remaining categories. Independent samples t-test showed significant difference in mean levels of impairment between the two groups. PTSD (M = 8, SD = 2.3), non-PTSD (M = 4, SD = 4), t (16) 2.8, p.014). Eta squared = 0.30 indicating a large effect size (Cohen).
Discussion: trauma/PTSD profile
Findings from this profile have particular importance for CMHS and underscore the importance of trauma assessment as a routine part of entry assessment to the service as indicated below.
Chronicity of trauma
Despite the lengthy service contact and the historical nature of the trauma, PTSD symptomatology was still current and largely unknown to treating clinicians and was, therefore, chronic in nature. Chronicity of PTSD is associated with co-morbidity; the longer the history of PTSD, the greater the chance of an individual developing a comorbid disorder [16]. This finding highlights the importance of trauma assessment in clients with major mental illness and the hidden impact that undiagnosed and untreated trauma/PTSD may have on the course and treatment of comorbid psychiatric illness [16]
Multi-traumatisation
This phenomenon (another feature of PTSD) is also evident in the current study, with most clients in the PTSD group demonstrating double the trauma exposure to those of the non-PTSD Group. This finding is particularly relevant to clinicians in CMHS as multiplicity of trauma exposure is also predictive of PTSD within general and psychiatric populations [3,8]. Ipso facto, clients with multiple trauma experiences are more likely to have PTSD; therefore, screening for trauma is important in alerting clinicians to the possibility that PTSD symptomatology may be present in clients being treated for another primary psychiatric diagnosis.
Types of traumatic events
The pattern of events reported, although consistent with that of other studies, is quite different to that of the Australian population as reported in the findings from the NSMHW [28]; the top three categories of events in that study were 'witnessing someone being killed', 'being involved in a life-threatening accident' and 'being involved in a natural disaster'. Sexual assault (defined as rape or sexual molestation in the above survey) was comparatively small – approximately 12% (n = 10 641). These findings distinguish the 'uniqueness' of the trauma profiles in the different populations and underscore the potential reasons for the increased levels of PTSD seen in treatment populations in CMHS as discussed in a subsequent section of this article.
Symptom profile
The currency of the symptomatology and impairment reported as arising from the traumatic event is of interest given the chronicity of the trauma experienced. The data in table 3 showed that in the PTSD group in particular, the reports of symptomatology are clearly not an aberration or a 'one-off' experience, but a persistent experience of negative feelings and emotions associated with the trauma. Again, this is an important finding given that the trauma was largely unknown to health professionals and therefore, untreated.
Also of interest is the symptom cluster showing the greatest effect in the PTSD group -that of the 'Avoidance/Numbing' criterion. Breslau [42] notes that this has previously been the least met criterion in the PTSD symptomatology clusters and, therefore, the most critical to the diagnosis.
This cluster is also of particular interest for its interaction with other psychiatric symptomatology. As noted earlier, the predominant type of trauma experienced by clients with mental illness is inter-personal in nature, therefore, the avoidance of social interactions and feelings of detachment, feature large in this criterion, (as evidenced by the data), consequently, they have a high potential to lead to social isolation and reduced social networks. Social isolation and poor social networks are also a feature of several types of other mental illnesses and are a known predictor of relapse and hospitalisation [43,44]. Ipso, facto, co-morbid PTSD symptomatology, particularly, that of the avoidance cluster, may increase this effect and lead to a worsening of symptoms and functioning and, ultimately, relapse and hospitalisation. The hypothesised pathway by which this occurs is illustrated in the model discussed earlier [8].
Rates of PTSD
This finding is of major importance as it demonstrates a rate 26-times greater than that found in the National Survey of Mental Health and Wellbeing (NSMHWB) [28] (using the most conservative results for rates of PTSD) and it is also consistent with findings of other studies in persons with a major mental illness in treatment settings as discussed above. If this finding was representative of CMHS Australia wide, one third of all clients in a service at a given time may have current PTSD symptomatology. There is no reason to suspect that these findings are unique to the study site, as it does not differ greatly in structure, services or client base to other services nationwide.
Potential contributory factors to the higher rates of trauma and PTSD found in clients of mainstream psychiatric services
The explanation for the finding of higher rates of trauma and PTSD in persons with major mental illness is unclear, however, it is evident from the discussion in the literature that this phenomena is not unique to the current study. Several potential factors (in addition to the role of chronicity discussed earlier) including victimisation, associations between age and type of trauma experienced, gender issues and vulnerability, substance use and the psychiatric setting have been proposed in the literature as possible explanations for the increased rates of trauma and PTSD in this population and these are briefly outlined below. In general these factors cannot be directly examined in relation to the results of the current study due to methodological limitations, but they provide a basis for exploration of this variable in future research studies in this population.
Age and type of trauma
The age at which victimisation occurs may influence later victimisation; several studies have noted the link between childhood sexual abuse and sexual and physical abuse in adulthood [3,6,45,46].
The additional relevance of this finding to general psychiatry is that sexual victimisation in childhood is also associated with the development of psychiatric disorders (other than PTSD) in adulthood [8,45,46]. The current study showed a large percentage of respondents reporting sexual abuse in both childhood and adulthood; victims of such trauma are more likely to develop other psychiatric disorders and be treated for such in the mental health system, therefore, treating health professionals need to be alert to this potential link to an unknown trauma history.
Gender effects
It has been suggested that women who have been subjected to sexual victimisation may be less aware of inherent dangers and have poorer risk recognition, therefore, are slower to remove themselves from sexually dangerous situations than are women with no history of sexual victimisation [8,42,47], hence increasing their vulnerability to further traumatisation. Whilst this effect is greater in some women with a history of sexual victimisation and PTSD symptomatology, this influence may be concomitant with the severity of PTSD symptoms. Better risk-awareness was reported in women with greater symptom severity, particularly those of the hyper-arousal cluster, so that PTSD symptomatology may, in some instances, act as a 'buffer' for women in sexually dangerous situations as a result of increased sensitivity to cues [47]. However, the opposite may be true for women with major mental illness, particularly those with a diagnosis of schizophrenia, who have experienced sexual victimisation. Vulnerability to re-victimisation may be exacerbated for these women as a result of the negative effect of their illness that may further compromise their social competence and decrease their ability to act positively to avert the risk or remove themselves from dangerous situations [8,48]. Sample size prohibits analysis of this factor in the current sample.
The Psychiatric setting
Vulnerability issues related to assaultive violence may also be a factor inherent in the psychiatric setting [20,49]; in-patient units in particular, have been the subject of a broad range of studies in relation to patient violence and the use of restraint and seclusion. However, the focus of these studies has primarily been staff and patient safety issues, the aetiology of violence, measurement issues, staff training and legislation. Even though patient to patient assault may occur, particularly in mixed gender units, few studies have examined the psychological impact of this on the individual [20]. It has been suggested that the procedures and processes of in-patient units, particularly those related to restraint and seclusion, may also re-traumatise victims [20,49]. Additionally, the experience of the mental illness itself, particularly if it involves psychosis, can result in PTSD symptomatology and/or exacerbation of previous trauma [6,49,50]. Several clients in the current study nominated this latter stressor (being diagnosed with a mental illness, particularly psychotic-based) as a 'traumatic event'. However, as it did not meet DSM IV criteria for a traumatic event, further assessment was not undertaken.
Substance use and misuse
This is a known clinical correlate of both PTSD and major mental illness, and may also play a role in the higher levels of trauma evidenced in this population [5,6,51]. The decreased inhibitory effects and resultant risk taking associated with substance abuse may place the person in increasingly unsafe social and physical environments exposing them to greater risks of interpersonal violence [6,51]. Levels of substance misuse are high in the PTSD and trauma groups of the current study, but again, the sample size prohibits any definitive analysis.
Implications of these contributing factors for CMHS and mainstream psychiatry
The variety of issues contributing to increased rates of trauma and PTSD identified in the preceding discussion may be distilled into three main factors all of which have important implications for public-sector service providers in mainstream psychiatry, particularly CMHS. These factors relate to the discrete sub-group of persons accessing mainstream psychiatric services, the inherent risk factors for PTSD within this group and the trauma characteristics demonstrated by this group. Increased awareness of these factors by service providers and clinicians is the first step in responding to the identified need for service provision for clients with co-morbid PTSD.
First, the nature of the population; all reported studies of PTSD in persons with a mental illness are from a treatment population of persons receiving current intervention for another mental health disorder. Given the known co-morbidity associated with PTSD, it is not unreasonable that higher rates would be found in this group of persons as they are suffering non-diagnosed PTSD, of chronic duration and, therefore, the likelihood of another mental disorder developing, for which the individual seeks treatment is increased. However, awareness of this factor by service providers not only gives a contextual awareness for interpretation of research findings in the field, but also provides them with insight into the potential service needs of their customer base and the need to include trauma/PTSD screening as a routine element of entry assessment protocols in this consumer group.
Second, the trauma profile of persons with a major mental illness (as reported in the literature and the results of this study) is dominated by interpersonal violence, particularly sexual victimisation and is potentially a major contributing factor to the higher rates found. Whilst women in particular are at greater risk of assaultive violence [42], the type of the trauma experienced and the individual's perception of the trauma as 'upsetting' is known to influence the development of PTSD. This cognitive/emotional response is a known feature of sexual victimisation; for example, rape is one such event that is perceived as 'upsetting' across genders and in treatment and non-treatment populations and, as such, is strongly linked to the subsequent development of PTSD [3,6,28]. Whilst this underscores the need for trauma screening, it also highlights the need for a sensitive and supportive environment that allows the traumatised individual to verbalise the nature of the abuse.
Finally, whilst increased risk of PTSD following exposure to trauma in persons with a major mental illness is reported in both treatment and population surveys [6,15,52], this risk may be incremental depending on the type and severity of mental illness experienced; in a treatment population, the nature of the illness is likely to be more acute and/or severe, therefore, individuals may be subject to greater vulnerability to the risk factors discussed earlier, including re-traumatisation and subsequent development of PTSD.
This has particular implications for the processes and procedures associated with issues of admission and management practices in hospital units, involuntary processes, the experience of mental illness itself, particular psychotic-based illness, and the impact this has on the psychological integrity of the individual. It also highlights the need for inclusive and collaborative management of substance abuse issues and those associated with residential and environmental safety in community based services.
The interaction of PTSD and other mental illness
Co-morbidity and PTSD
PTSD is strongly co-morbid with other psychiatric disorders as demonstrated in the two community surveys in America [3] and Australia [2]. The NSMHWB [2] found a 12-month prevalence of co-morbidity for PTSD with at least one other Axis 1 diagnosis in 85.2% of males and 79.7% of females, whilst the National Co-morbidity survey showed 88.3% males and 79% females with at least one other psychiatric disorder [3]. This high rate of co-morbidity was also demonstrated in the current study particularly in the PTSD group, who accounted for most of the co-morbidity in the sample, however, the sample was too small to examine gender differences.
Notwithstanding the potential influence of the above on rates of PTSD in persons with a co-morbid psychiatric diagnosis, no definitive causality chain between PTSD and other mental disorders can be identified. Although some researchers discuss the seemingly intuitive link between traumatic experiences, PTSD and the course of other mental illnesses (based on the widespread occurrence of the phenomenon), there is no definitive answer to the question at this time and the issue has not been widely studied or reported.
The possible link between the two broad types of disorders may be more to do with shared risk factors for PTSD and other mental health problems, such as the mood disorders, and the interplay between differential effects of specific traumas, individual risk factors and personal coping mechanisms [16,28]. The full explanation may lie in a complex matrix of all of these factors differing across individuals, communities and diagnoses, but is unlikely to lie in the potential for symptom overlap between PTSD and common co-morbid diagnoses such as depression, anxiety, and dysthymia. Rather, the literature suggests that this factor, far from overstating the case for PTSD, may result in misdiagnosis if detailed trauma histories are not sought and potential PTSD diagnosis excluded [10,15,16].
The potential influence of this factor cannot be ruled out of the current study given the level of documentation of respondents' trauma/PTSD profile found. Although this local finding was not unexpected, as there was no formal or standardised assessment of trauma done in the service it does, however, demonstrate the potential for diagnostic ambiguity in clients presenting with symptoms of depression, anxiety, psychosis and/or substance abuse, if trauma histories are not routinely sought.
Whilst the inter-relationship between trauma, PTSD and other mental illness is complex and largely unexamined, thus compounding attempts to explain the higher rates found in persons with other mental disorders, contemporary findings indicate that multiple traumatisation is a strong predictor of PTSD in treatment and non treatment populations regardless of the ultimate relationship [3,6]. The single definitive answer that can be gleaned from the current findings and discussions on the subject, is that much more rigorous and longitudinal research is needed from an epidemiological and targeted perspective in order to achieve optimal outcomes for the commonly treated psychiatric disorders [1,3,7,8,16,22,28,45,53,54].
Data: effect of PTSD diagnosis on client outcome
As the study sought to explore the potential effect of untreated PTSD symptomatology on client health outcomes, this section compares data for clients with PTSD to those without PTSD, and therefore, includes those clients who reported no experience of trauma either current or lifetime. The sample size was 17 (5 clients with PTSD, 12 without), the number remaining enrolled in the study at the T2 data point, the 6-month period following engagement with the service and CM. Only two of the four outcome measures used showed any significant difference for clients with PTSD – the HoNOS and the WHOQOL. Only one of these, WHOQOL, is reported in this paper for reasons discussed earlier. Data are presented in table 5/6. Two population 'norms' were used for comparison, the first was from a similar population of clients with a major mental disorder, primarily Axis 1 diagnoses, 70% of whom had a diagnosis of Schizophrenia [55]. That study was conducted in an Australian Community Mental Health setting and used the WHOQOL-BREF, client and CM format and, therefore, provides a suitable comparison in view of the lack of studies in clients with major mental illness and PTSD. The second comparison is that of the Australian population 'norms' for the WHOQOL [33].
Table 5 QOL PTSD/No PTSD diagnosis (n = 17): Client data and comparisons with community samples.
Domain Study Group: Client Data Population 'Norms' Psychosis n = 173 Australian Population n = 396
M SD P Effect size (Eta2) M SD P* M SD P*
Physiological Health 60.7 15.4 80 17.1
PTSD n = 5 37.14 8.60 .034 0.26 .004 .0005
No PTSD n = 12 59.22 20.17 .805 .004
Psychological Health 56.8 17.4 72.6 14.2
PTSD 32.50 9.50 .010 0.36 .005 .001
No PTSD 58.33 18.37 .778 .021
Social Relationships 51.3 20.3 72.2 18.5
PTSD 38.33 26.74 .527 .339 .047
No PTSD 45.13 16.46 .221 .0005
Environment 61.1 13.8 74.8 13.7
PTSD 50.62 10.22 .015 0.33 .084 .006
No PTSD 69.72 13.57 .061 .186
Derived from one sample t-test.
Table 6 QOL PTSD/No PTSD diagnosis: CM data and comparisons with 'psychosis" sample
Domain Study Group: CM Data Psychosis 'Norms'
P*
M SD P Effect size (Eta2) M SD
Physiological Health 57 12.5 .110
PTSD n = 5 38.75 20.10 .001
No PTSD n = 12 70.23 11.82
Psychological Health 51 13.0 .032
PTSD 36.66 9.94 .007
NO PTSD 54.16 10.87
Social Relationships 43.4 18.8 .026
PTSD 26.66 10.86 .066
No PTSD 40.97 14.41
Environment 55.4 13.5 .178
PTSD 46.87 11.69 .006
No PTSD 65.62 10.74
There were no significant correlations between client and CM ratings on any of the outcome measures used in this study.
Discussion: effect of PTSD on client outcomes
The findings on outcome are severely constrained by the short follow-up time and no definitive conclusions can be drawn from the lack of effect for PTSD Diagnosis on the other outcomes selected.
Whilst QOL is considered the most important outcome in mental health research and is central to outcomes management [56], there is an obvious lack of literature on PTSD and quality of life in the study population with which to compare the research findings [56,57]. Contemporary QOL research in PTSD stems primarily from a veteran's perspective [56,57]; research of civilian trauma has focused on specific trauma-related perspectives, for example, female victims of violence [58], victims of major trauma [59], or persons with specific medical conditions [60,61] and those involved in drug trials [57]. Although there remains the issue of different QOL measurements used in the various studies to date, the emerging trend in anxiety research suggests that PTSD in particular, has a major negative effect on QOL [56,57].
The results of the current study support this growing body of research, with clients and CM's both reporting data that was significantly different for those clients with PTSD than those without, in three of the four domains of QOL measured.
The impact of co-morbid PTSD on QOL in persons with another major mental illness is further evidenced in the study group when compared with a similar population with major mental illness but without PTSD. Clients with PTSD showed significantly greater impairment in physiological and psychological health than did those in the comparison group who had a psychotic illness only (considered to be one of the most disabling disorders with a lower quality of life than that reported in physical illness and the general population). In contrast, those study clients without PTSD had scores very similar to that reported for the comparison group. Additionally, the study demonstrated that for clients with and without PTSD, reported QOL was significantly worse in all domains (with the exception of that of the Environmental Domain in the non-PTSD group) when compared with that of the general Australian population.
Limitations
The primary limitations of this study were the small sample size and the self-report nature of the trauma/PTSD data collected from the client. Although an attempt to minimise this latter point was taken by conducting the PDS as an interview, clients still needed to engage in re-call, as the trauma was an historical event for most of them. Furthermore, it also required discrimination of similar symptomatology from that of their primary psychiatric diagnosis. Both of these limitations are addressed in the recommendations.
Conclusion
The trauma/PTSD profile for this small sample of Australian CMHS clients with major mental illness was consistent with findings from other reported studies of similar populations on all key elements and was largely unknown by treating clinicians.
Standardised assessment of trauma was not a routine element of service entry at the study site or generally within Australian Community Mental Health Services, but there is growing evidence within the literature and from the current study to support the introduction of such a measure.
This study also identified poorer outcomes in clients with PTSD, albeit for a single outcome only, that of QOL, which was shown to be significantly compromised both within the study group and in comparison to that reported for an external population of clients with psychotic-based illness (no PTSD) in another CMHS within Australia in which the same measure of QOL was used. No effect was found for other key outcome measures with the exception of the HoNOS, which will be reported elsewhere.
Whilst no single explanation for the findings of increased trauma/PTSD in persons with another major mental illness was evidenced either in the current study or the literature, the authors propose that three interconnected factors, primarily related to the characteristics of the population treated in CMHS, may figure large in any such explanation. Each of these has important implications for service delivery across all elements of an integrated mental health service.
Recommendations
Finally, within the context of these findings it is recommended that further research be undertaken with a larger sample to determine the relevance of these findings to the broader population of clients in Public Community Mental Health Services. In addition to the use of a screening measure, a follow-up clinical interview, such as the Clinician Administered Posttraumatic Stress Scale (CAPS) [62] is recommended in order to provide more robust data that clearly distinguishes PTSD symptomatology from that of the primary psychiatric disorder.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
IH participated in study design, data collection and analysis and drafted the manuscript. CO coordinated and facilitated the study at the clinical interface in ACT, participated in study design, and preparation of manuscript. PY conceived the study and participated in study design and implementation in Brisbane, and acted as overall consultant. LM participated in the study design and implementation in Brisbane and provided specialist advice relating to Posttraumatic Stress Disorder measures. FD participated in study design for the Brisbane elements of implementation and facilitated the study at the clinical interface. RP provided statistical advice and analysis and participated in the preparation of the manuscript. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This study was undertaken with funding from the National Health and Medical Research Council (NHMRC) Grant Number 98/NHMRC7001G.
==== Refs
Davidson J New Strategies for the Treatment of Posttraumatic Stress Journal of Clinical Psychiatry 2000 61 3 51
Andrews G Hall W Teesson M Henderson S The Mental Health of Australians: Survey of National Mental Health and Wellbeing 1999 Canberra, ACT: Commonwealth Department of Health and Aged Care
Kessler R Sonnega A Bromet E Hughes M Nelson CB Posttraumatic Stress Disorder in the National Comorbidity Survey Archives of General Psychiatry 1995 52 1048 1060 7492257
McFarlane A Bookless C Air T Posttraumatic Stress Disorder in a General Psychiatric Inpatient Population Journal of Traumatic Stress Studies 2001 14 633 645 10.1023/A:1013077702520
Neria Y Bromet S Seivers S Trauma exposure and posttraumatic stress disorder in psychosis: Findings from a first-admission cohort Journal of Consulting and Clinical Psychology 2002 7 246 251 11860051 10.1037//0022-006X.70.1.246
Mueser K Trumbetta SL Rosenberg S Trauma and posttraumatic stress disorder in severe mental illness Journal of Consulting and Clinical Psychology 1998 66 493 499 9642887 10.1037//0022-006X.66.3.493
Kessler RC Posttraumatic Stress Disorder: The burden to the individual and society Clin Psychiatry 2000 61 4 12
Mueser K Rosenberg S Goodman LA Trumbetta SL Trauma, PTSD and the course of severe mental illness: and interactive model Schizophrenia Research 2002 53 123 143 11728845 10.1016/S0920-9964(01)00173-6
Craine S Henson CE Colliver JA MacLean DG Prevalence of a history of sexual abuse among female psychiatric patients in a state hospital system Hosp Community Psychiatry 1988 39 300 304 3356438
Davidson J Smith R Traumatic experiences in psychiatric outpatients Journal of Traumatic Stress Studies 1990 3 459 475 10.1007/BF00974785
Jacobson A Richardson B Assault experiences of 100 psychiatric inpatients: Evidence of the need for routine inquiry Am J Psychiatry 1987 144 908 913 3605402
Lipscihitz D Kaplan ML Sorkenn JB Prevalence and characteristics of physical and sexual abuse among psychiatric outpatients Psychiatric Services 1996 47 189 191 8825258
Muenzenmaier K Meyer I Struening E Ferber J Childhood abuse and neglect among women outpatients with chronic mental illness Hosp Community Psychiatry 1993 44 666 670 8192738
Rose SM Peabody CG Stratigeas B Undetected abuse among intensive case management clients Hospital and Community Psychiatry 1991 42 449 503
Zimmerman M Mattia Jl Is posttraumatic stress disorder under-diagnosed in routine clinical settings? Journal of Nervous and Mental Diseases 1999 187 420 428 10.1097/00005053-199907000-00005
Brady KT Killeen TK Brewerton T Lucerini S Comorbidity of Psychiatric Disorders and Posttraumatic Stress Disorder J Clin Psychiatry 2000 61 22 33 10795606
McFarlane A Traumatic Stress in the 21 st Century Australian and New Zealand Journal of Psychiatry 2000 34 896 902 11127620 10.1046/j.1440-1614.2000.00826.x
McFarlane A Yehuda R Clinical Treatment of posttraumatic stress disorder: conceptual challenges raised by recent research Australian and New Zealand Journal of Psychiatry 2000 39 940 953 11127624 10.1046/j.1440-1614.2000.00829.x
Switzer G Dew MA Thompson K Goycoolea JM Derricott T Mullins SD Posttraumatic stress disorder and service utilization among urban mental health center clients Journal of Traumatic Stress 1999 12 25 39 10027140 10.1023/A:1024738114428
Rosenberg S Mueser KT Friedman MT Developing effective treatments for posttraumatic stress disorder among people with severe mental illness Psychiatric Services 2001 52 1453 1461 11684740 10.1176/appi.ps.52.11.1453
Sher L Posttraumatic stress disorder: the importance of recognition Australian and New Zealand Journal of Psychiatry 2004 38 84 84 14731199 10.1111/j.1440-1614.2004.01303.x
Hidalgo RB Davidson JR Posttraumatic Stress Disorder: Epidemiology and Health-Related Considerations J Clin Psychiatry 2000 61 5 13 10795604
Howgego I Yellowlees P Owen C Meldrum L Dark F The Therapeutic Alliance: the key to effective patient outcome? A descriptive review of the evidence in community mental health case management Australian and New Zealand Journal of Psychiatry 2003 37 169 183 12656956 10.1046/j.1440-1614.2003.01131.x
Horvath A Symonds B Relation between working alliance and outcome in psychotherapy: a meta-analysis Journal of Counselling Psychology 1991 38 139 149 10.1037//0022-0167.38.2.139
Neal M Rosenheck RA Therapeutic alliance and outcome in a VA intensive case management program Psychiatric Services 1995 46 719 721 7552565
Solomon P Draine J Delaney MA The working alliance and consumer case management Journal of Mental Health Administration 1995 22 126 134 10142126
Brewin C Andrews B Valentine JD Meta-Analysis of Risk Factors for Posttraumatic Stress Disorder in Trauma-Exposed Adults Journal of Consulting and Clinical Psychology 2000 68 748 766 11068961 10.1037//0022-006X.68.5.748
Creamer M Burgess P McFarlane AC Post-traumatic stress disorder: Findings from the Australian National Survey of Mental Health and Well- being Psychological Medicine 2001 31 1237 1247 11681550 10.1017/S0033291701004287
ACT Chief Health Officer's Report 2000–2003
Foa E Posttraumatic Stress Diagnostic Scale (PDS) Manual 1995 Minneapolis: National Computer Systems Inc
APA Diagnostic and Statistical Manual of Mental Disorders – DSM1V 1994 Fourth Washington DC: American Psychiatric Association
Horvath A Greenberg L Development and Validation of the Working Alliance Inventory Journal of Counselling Psychology 1989 36 223 233 10.1037//0022-0167.36.2.223
Murphy B Herrman H Hawthorne G Pinzone T Evert H Australian WHOQOL Instruments: User's manual and interpretation guide 2000 Melbourne, Australia: Australian WHOQOL Field Study Centre, University of Melbourne, Department of Psychiatry
Wing J Curtis R Beevor A Health of the Nation Outcome Scales: Report on Research 1996 London: College Research Unit, Royal College of Psychiatrists
Trauer T Callaley T Hantz P Little J Sheilds R Smith J Health of the Nation Outcome Scales: Results of the Victorian field trial British Journal of Psychiatry 1999 174 380 388 10616601
Orrell M Yard P Handysides J Schapira R Validity and Reliability of the Health of the Nation Outcome Scales in psychiatric patients in the community British Journal of Psychiatry 1999 174 409 412 10616606
Care DoHaA Mental Health Information Development: National Information Priorities under the Second National Mental Health Plan 1998–2003 1998 Canberra: Commonwealth Department of Health and Family Services
Buckingham W Burgess P Solomon S Developing a Casemix Classification for Mental Health Services, Volume 2: Resource Material 1998 Canberra: Commonwealth Department of Health and Family Services
Trewin D National Health Survey: Mental Health 2001 Canberra: Australian Bureau of Statistics
McLennan W Mental health and Wellbeing: Profile of Adults, Australia 1997 Canberra: Australian Bureau of Statistics
Frost B Carr V Halpin S Employment and Psychosis 2002 Canberra: Commonwealth Department of Health and Ageing
Breslau N Chilcoat HD Kessler RC Vulnerability to assaultive violence: further specification of sex difference in post-traumatic stress disorder Psychological Medicine 1999 29 813 821 10473308 10.1017/S0033291799008612
Kent S Yellowlees P Psychiatric and social reasons for frequent rehospitalization Hosp Community Psychiatry 1994 45 347 350 8020919
Kent S Fogarty M Yellowlees P A review of studies of heavy service users of psychiatric services Psychiatric Services 1995 46 1247 1253 8590109
Jacobson A Herald C The relevance of childhood sexual abuse to adult psychiatric inpatient care Hosp Community Psychiatry 1990 41 154 158 2303215
Briere J Woo R McRAE BM Foltz J Sitzman R Lifetime History, Demographics, and Clinical Status in Female Psychiatric Emergency Room Patients Journal of Nervous and Mental Diseases 1997 185 95 101 10.1097/00005053-199702000-00005
Wilson A Calhoun KS Bernat JA Risk Recognition and Trauma-Related Symptoms Among Sexually Revictimized Women Journal of Consulting and Clinical Psychology 1999 67 705 710 10535237 10.1037//0022-006X.67.5.705
Gearon J Bellack AS Women with Schizophrenia and Co-Occurring Substance Use Disorders: An Increased Risk for Violent Victimization and HIV Community Mental Health Journal 1999 35 401 419 10547116 10.1023/A:1018778310859
Frueh C Cusack KJ Heirs TG Monogan S Cousins VC Cavenaugh SD Improving public mental health services for trauma victims in South Carolina Psychiatric Services 2001 52 812 814 11376230 10.1176/appi.ps.52.6.812
McGorry PD Chanen A McCarthy E Van Riel R McKenzie D Singh BS Posttraumatic Stress Disorder following recent-onset psychosis. An unrecognised syndrome J Nerv Ment Dis 1991 179 253 258 1781819
Hiday VA Swartz MS Swanson JW Borum R Wagner HR Criminal Victimisation of Persons with Severe Mental Illness Psychiatr Serv 1999 59 62 68 9890581
Cascardi M Mueser KT DeGiralomo J Murrin M Physical aggression against psychiatric inpatients by family members and partners Psychiatric Services 1996 47 531 533 8740498
Creamer M O'Donnell M Post-Traumatic Stress Disorder Current Opinion in Psychiatry 2002 15 163 167 10.1097/00001504-200203000-00007
McFarlane A Traumatic Stress in the 21st Century Australian and New Zealand Journal of Psychiatry 2000 34 896 902 11127620 10.1046/j.1440-1614.2000.00826.x
Herrman H Hawthorne G Thomas R Quality of life assessment in people living with psychosis Social Psychiatry & Psychiatric Epidemiology 2002 37 510 518 12395140 10.1007/s00127-002-0587-y
Mendlowicz M Stein MB Quality of Life in Individuals with Anxiety Disorders American Journal of Psychiatry 2000 157 669 682 10784456 10.1176/appi.ajp.157.5.669
Hansson L Quality of Life in Depression and Anxiety International Review of Psychiatry 2002 14 185 189 10.1080/09540260220144966
Laffaye C Kennedy C Stein MB Post-traumatic stress disorder and health-related quality of life in female victims of intimate partner violence Violence Vict 2003 18 227 238 12816406
Holbrook T Hoyt DB Stein MB Sieber WJ Gender differences in long-term posttraumatic stress disorder outcomes after major trauma: women are at higher risk of adverse outcomes than men Journal of Trauma-Injury Infection and Critical Care 2002 53 882 8 10.1097/00005373-200211000-00012
Amir M Ramati A Post-traumatic symptoms, emotional distress and quality of life in long-term survivors of breast cancer; preliminary research Journal of Anxiety Disorders 2002 16 195 206 12194544
Meeske K Ruccione K Globe DR Stuber ML Posttraumatic stress, quality of life, and psychological distress in young adult survivors of childhood cancer Oncology Nursing Forum 2001 28 481 9 11338757
Blake D Weathers FW Nagy LM Kaloupek DG Gusman FD Charney DS Keane TM The development of a clinician-administered PTSD scale Journal of Traumatic Stress 1995 8 75 91 7712061
| 15850496 | PMC1156913 | CC BY | 2021-01-04 16:33:02 | no | BMC Psychiatry. 2005 Apr 26; 5:21 | utf-8 | BMC Psychiatry | 2,005 | 10.1186/1471-244X-5-21 | oa_comm |
==== Front
BMC PsychiatryBMC Psychiatry1471-244XBioMed Central London 1471-244X-5-261592150810.1186/1471-244X-5-26Research ArticleAntipsychotic monotherapy and polypharmacy in the naturalistic treatment of schizophrenia with atypical antipsychotics Faries Douglas [email protected] Haya [email protected] Baojin [email protected] Christoph [email protected] John [email protected] Outcomes Research, Eli Lilly and Company, Indianapolis, Indiana, USA2 Department of Psychiatry, The Zucker Hillside Hospital, Glen Oaks, New York, USA2005 27 5 2005 5 26 26 13 1 2005 27 5 2005 Copyright © 2005 Faries et al; licensee BioMed Central Ltd.2005Faries et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Antipsychotic monotherapy is recognized as the treatment of choice for patients with schizophrenia. Simultaneous treatment with multiple antipsychotics (polypharmacy) is suggested by some expert consensus guidelines as the last resort after exhausting monotherapy alternatives. This study assessed the annual rate and duration of antipsychotic monotherapy and its inverse, antipsychotic polypharmacy, among schizophrenia patients initiated on commonly used atypical antipsychotic medications.
Methods
Data were drawn from a large prospective naturalistic study of patients treated for schizophrenia-spectrum disorders, conducted 7/1997–9/2003. Analyses focused on patients (N = 796) who were initiated during the study on olanzapine (N = 405), quetiapine (N = 115), or risperidone (N = 276). The percentage of patients with monotherapy on the index antipsychotic over the 1-year post initiation, and the cumulative number of days on monotherapy were calculated for all patients and for each of the 3 atypical antipsychotic treatment groups. Analyses employed repeated measures generalized linear models and non-parametric bootstrap re-sampling, controlling for patient characteristics.
Results
During the 1-year period, only a third (35.7%) of the patients were treated predominately with monotherapy (>300 days). Most patients (57.7%) had at least one prolonged period of antipsychotic polypharmacy (>60 consecutive days). Patients averaged 195.5 days on monotherapy, 155.7 days on polypharmacy, and 13.9 days without antipsychotic therapy. Olanzapine-initiated patients were significantly more likely to be on monotherapy with the initiating antipsychotic during the 1-year post initiation compared to risperidone (p = .043) or quetiapine (p = .002). The number of monotherapy days was significantly greater for olanzapine than quetiapine (p < .001), but not for olanzapine versus risperidone, or for risperidone versus quetiapine-initiated patients.
Conclusion
Despite guidelines recommending the use of polypharmacy only as a last resort, the use of antipsychotic polypharmacy for prolonged periods is very common during the treatment of schizophrenia patients in usual care settings. In addition, in this non-randomized naturalistic observational study, the most commonly used atypical antipsychotics significantly differed on the rate and duration of antipsychotic monotherapy. Reasons for and the impact of the predominant use of polypharmacy will require further study.
==== Body
Background
Guidelines for treating patients with schizophrenia [1-5] have long recognized antipsychotics as the core treatment modality and have consistently recommended antipsychotic monotherapy as the treatment of choice. Although expert consensus guidelines do not advocate antipsychotic polypharmacy, some [4] suggest antipsychotic polypharmacy as the last resort after having exhausted prior monotherapy alternatives. Only one of the consensus guidelines [5] offers guidance on the duration of antipsychotic polypharmacy, which is recommended when switching from one antipsychotic to another (cross titration or overlap and taper) and for not longer than 60 days.
Monotherapy is recognized as the preferred mode of treatment because it allows clinicians to accurately evaluate the patient's response to a new course of treatment [6]. Monotherapy permits documenting patient's response to an adequate trial of each medication, helping to reduce the complexity of the medication regimen, reducing the risk of adverse events, and making it easier to assess and manage future symptom exacerbations [7].
Despite consistent recommendations of antipsychotic monotherapy, polypharmacy is widespread in the treatment of schizophrenia [7-15]. The proliferation of antipsychotic polypharmacy is likely driven by increased availability of pharmacologically diverse atypical antipsychotics that augment an extensive armamentarium of typical antipsychotics. Generally, the concurrent use of more than one antipsychotic, particularly of typical and atypical agents, was reported to vary from 13% to 60%, depending on the population studied, the year when the study has been conducted, the study method, the type of treatment site, and the duration of the study period [8,9,14,16-19].
The body of evidence supporting the benefits of antipsychotic polypharmacy is limited [7,8,20,21] and is in contrast to the extensive and compelling body of evidence supporting monotherapy with atypical antipsychotics [6,7,9,11-13,22-24].
Antipsychotic polypharmacy was reported to increase the risk of medication-related adverse events and of drug-drug interactions, to increase the need for additional medications to treat emerging side-effects, to decrease adherence with medication due to increased treatment complexity, confounding clinicians' ability to discern helpful from unhelpful medications, and to increase cost of care [7,10,25]. A recent case-control study of psychiatric inpatients demonstrated that short-term treatment with antipsychotic polytherapy was associated with major increases in drug exposure, in adverse events, and in hospitalized duration but with no apparent clinical benefit [16].
Antipsychotic polypharmacy appears to be used for various reasons [7,11], with the one cited most often being the wish to bolster medication effectiveness in treating patients with refractory psychotic symptoms, mood symptoms, or behavioral problems [9]. It is, however, unclear if antipsychotic polypharmacy is associated with specific atypical antipsychotics more often than with others.
The body of research on differential monotherapy or polypharmacy among atypical agents is sparse [8,26-30]. It is unclear if risperidone and quetiapine-treated patients differ on monotherapy parameters. Most previous studies reported higher monotherapy or lower polypharmacy rates for olanzapine compared to risperidone [18,19,28,29], and compared to quetiapine-treated patients [25,27,30], whereas olanzapine and risperidone-treated patients were not found to significantly differ on polypharmacy rates in three studies [25-27]. The reasons for the inconsistent findings are unclear but may stem from methodological issues, including lack of complete information about the use of depot antipsychotics, about use of medications during psychiatric hospitalizations, and inadequate control for potential selection effects, which arise from clinicians' tendency to tailor treatment regimens to patients' illness profiles and previous treatment patterns.
This study expanded on prior research by using comprehensive medication data from a large prospective multi-site naturalistic study of patients treated for schizophrenia-spectrum disorders in the United States to assess the annual rate and duration of monotherapy, and its inverse, antipsychotic polypharmacy. We also focused on patients initiated on commonly used atypical antipsychotics – olanzapine, quetiapine, or risperidone – and compared their rates and duration of monotherapy during the year following initiation on the antipsychotic medication.
Methods
Data source
This study used data of the U.S. Schizophrenia Care and Assessment Program (US-SCAP), a large (N = 2327) non-randomized, naturalistic, 3-year prospective multi-site study conducted between 7/1997 and 9/2003. The goal of US-SCAP was to understand the treatment of patients with schizophrenia in usual care settings. Approximately 400 patients at each of the study's six regional sites were enrolled. All participants were diagnosed with schizophrenia, schizoaffective, or schizophreniform disorders based on DSM-IV criteria, and were at least 18 years of age. Patients were excluded if they were unable to provide informed consent or had participated in a clinical drug trial within 30 days prior to enrollment.
In order to reduce selection bias, outpatients were randomly selected from site medical information rosters of active clients. Inpatients were sequential admissions. Of 3332 patients who met inclusion criteria, 765 (23.0%) refused, and 240 (7.2%) were not enrolled for other reasons. Most enrollees competed 1 year of follow-up (78.1%), with fewer participants completing 2 years (69.6%), and 3 years (65.2%). Of the 2327 enrollees, 21.2% were hospitalized at enrollment or during the 6 months prior to enrollment, and 36.2% were hospitalized in the year prior to enrollment.
Participants were enrolled from six states (California, Colorado, Connecticut, Florida, Maryland, and North Carolina) and represented treatment in diverse systems of care including community mental health centers, university health care systems, the Department of Veterans Affairs Health Services (VA), and community and state hospitals. Institutional Review Board (IRB) approval was received at each regional site and informed consent was received from all participants.
At enrollment, almost all participants (94.7%) were treated with at least one antipsychotic medication, including oral typical (36.7%), oral atypical (58.1%), and depot typical antipsychotics (19.6%). Medication changes during the study period, including initiations and discontinuations, if any, were based on physicians' decisions as they occur in usual care. Further details about US-SCAP are available elsewhere [31,32].
The current study included data of US-SCAP participants who were initiated on one of three commonly used atypical antipsychotics – olanzapine, quetiapine, or risperidone, in regular oral formulation. Participants were defined as initiators if they were not prescribed the index antipsychotic for at least 60 days prior to initiation, had at least 2 months of treatment data available prior to initiation of the index antipsychotic, and had at least 12 months of follow-up information in the study post initiation. Inclusion of data for the 2 months prior to initiation allowed for identification of variables on which the treatment groups differed at the time of initiation. Further, because this study aimed to assess use of antipsychotics over a 1-year period, availability of a 12-month follow-up period was necessary. Note that patients were not excluded from the analysis if they stopped their antipsychotic therapy – only if they discontinued or completed the study within less than 12 months post initiation. Some patients may have met inclusion criteria for being an initiator of more than one medication throughout the 3-year period. Such patients were considered to be initiators of the first antipsychotic medication they initiated in the study.
Outcome measures
Medical records provided information about prescribed psychiatric medications and were systematically abstracted for the 6 months prior to enrollment and for each 6-month interval thereafter. Patients were queried about the use of medications and other mental health resources outside those received at their regular treatment site. When this occurred, systematic efforts were made to abstract out-of-site medical records.
On a daily basis, monotherapy (polypharmacy) was defined as the occurrence of one (more than one) ongoing antipsychotic medication prescription. Predominant use of monotherapy (polypharmacy) was defined as the use of monotherapy (polypharmacy) for > 300 days out of the year. Substantial monotherapy (polypharmacy) use was defined as the use of monotherapy (polypharmacy) for > 60 to ≤ 300 days out of the year. Consistent with expert consensus guidelines [5], prolonged polypharmacy was defined as a period of more than 60 consecutive days of polypharmacy.
Two parameters of antipsychotic monotherapy were used as the outcome measures for initiating treatment group comparisons (olanzapine, quetiapine, or risperidone): (a) the percentage of patients on antipsychotic monotherapy with the index atypical antipsychotic over the 1-year post initiation, and (b) the cumulative number of days of antipsychotic monotherapy.
Participants were assessed with standard psychiatric measures at enrollment and at 12-month intervals thereafter. These measures were not administered, however, at the time of initiation or discontinuation of any medication. Consequently, these variables were not used in this study. Furthermore, the study did not assess reasons for medication initiation or discontinuation, thus eliminating the ability to evaluate the reasons for any medication changes.
Statistical methods
Summary statistics were used to quantify monotherapy/polypharmacy use at initiation on the medication and across the 1-year period post initiation for all patients (N = 796). This included summarizing the distribution of durations on monotherapy/polypharmacy as well as classification of patients based on predominant, substantial monotherapy/polypharmacy definitions and the prolonged use of polypharmacy.
To assess differences in the use of monotherapy between the treatment groups, two statistical approaches were utilized. First, treatment group differences in the percentage of patients on monotherapy each day over the 1-year following initiation were assessed using a repeated measures general linear model (GEE, Generalized Estimating Equations) [33] with an exchangeable correlation matrix. GEE models are widely used for analyzing longitudinal binary data as they provide consistent parameter estimates and account for the correlation of responses within patients over time. The model consisted of terms for treatment, time, treatment by time interaction, and a set of a priori determined covariates discussed below. Second, treatment group differences in the mean number of days of monotherapy over the 1-year following initiation were assessed. As the distribution of the number of days was non-normal – with a high percentage of patients with zero days, a non-parametric propensity score adjusted bootstrap re-sampling approach was utilized [34].
In all treatment comparisons, differences were adjusted for a set of available covariates selected a priori. These covariates were selected based on the expectations that they may be associated with monotherapy use. Some clinical variables that were collected at 1-year intervals in this study were not used as covariates because they were not collected at the point of initiation or discontinuation of any medication. Thus, the covariates used for adjustment in this study included available socio-demographic and medication history variables: age, gender, ethnicity, illness duration, monotherapy status at initiation (yes/no), investigational site, time, time by treatment interaction, and the following variables based on the 2 months prior to initiation of the index antipsychotic: any psychiatric hospitalization, number of days of antipsychotic monotherapy, and any use of: typical antipsychotics, atypical antipsychotics, antiparkinsonian agents, antidepressants, anti-anxiety medications, sleep agents, mood stabilizers, and typical antipsychotics in depot formulation.
No adjustment for multiple comparisons was performed as we considered the repeated measures GEE model the primary analysis and the monotherapy duration analysis as a secondary analysis to assess the robustness of the primary results.
Results
Patient characteristics
Table 1 summarizes the socio-demographic characteristics and medication history for the patient population. The sample (N = 796) included patients who were initiated during the study on olanzapine (N = 405), quetiapine (N = 115), or risperidone (N = 276). While similar in many aspects, the treatment groups differed on several characteristics – most notably the percentage of patients on antipsychotic polypharmacy at initiation, which was highest among quetiapine-treated patients and lowest for the risperidone treatment group. During the 2 months prior to initiation on the index antipsychotic, a greater percentage of quetiapine-treated patients were treated with atypical antipsychotics and mood stabilizers; a greater percentage of olanzapine-treated patients were treated with typical antipsychotics in both oral and depot formulations; risperidone patients had fewer antipsychotic monotherapy days. Further, the quetiapine treatment group had least males, the lowest illness duration, and fewest patients with substance use disorders.
Table 1 Baseline characteristics of patients initiated on olanzapine, risperidone, and quetiapine
Variable Olanzapine N = 405 Quetiapine N = 115 Risperidone N = 276
Age, mean (S.D.) 41.8 (10.5) 39.6 (10.9) 40.4 (12.1)
Illness duration (yrs), mean (S.D.) * a,b 22.1 (11.3) 18.5 (11.2) 19.9 (12.4)
Male gender* b 61.7% 47.8% 54.7%
Ethnicity
White 46.7% 53.9% 45.3%
Black 41.2% 33.9% 36.6%
Other 12.1% 12.2% 18.1%
Schizoaffective disorder 32.6% 40.0% 33.3%
Substance-use disorder* b,c 30.9% 19.3% 29.4%
No insurance 6.9% 8.9% 9.2%
Polypharmacy at initiation * a,c 67.2% 76.5% 59.8%
Prior treatment pattern †
Prior Antipsychotic monotherapy days, mean (S.D.)* a 42.4 (26.0) 39.2 (27.7) 36.8 (28.0)
Prior psychiatric hospitalization* a 21.0% 21.7% 29.4%
Prior atypical antipsychotic* b,c 23.5% 69.6% 28.6%
Prior typical oral antipsychotic* b,c 63.5% 39.1% 58.7%
Prior typical depot antipsychotic* a,b 23.2% 13.9% 15.9%
Prior antidepressants 35.6% 45.2% 40.9%
Prior anti-anxiety agents 10.4% 14.8% 12.0%
Prior antiparkinsonians agents 46.9% 36.5% 47.5%
Prior mood stabilizers* a,c 30.1% 39.1% 23.2%
Prior sleep agents 0.7% 3.5% 2.2%
* Significant groups difference at p < 0.05
asignificant pairwise comparisons olanzapine vs. risperidone
b significant pairwise comparisons olanzapine vs. quetiapine
c significant pairwise comparisons risperidone vs. quetiapine
† Treatment pattern during the 2 months prior to initiation on the index drug
Medication dose
The dosing of each antipsychotic during the year post initiation, at initiation and at endpoint, and by monotherapy versus polypharmacy status at initiation is presented in Table 2. Mean doses were higher at endpoint than at initiation and were within the package insert guidelines for the vast majority of the patients. In addition, the mean doses were higher for those on monotherapy than polypharmacy at initiation, but this was not always found at endpoint.
Table 2 Daily Dose (mg/day) of Index Atypical Antipsychotic: During the 1-Year Post Initiation, at Initiation, and at Endpoint
Group/Time Mean (SD) Median Min/Max Mean Mono/Poly*
Olanzapine
1-Year 13.9 (7.4) 10.0 2.5/40 15.1/13.5
Initiation 10.0 (6.1) 10.0 2.5 / 40 10.7 / 9.5
Endpoint 14.2 (8.4) 10.0 2.5 / 50 13.9 / 14.5
Quetiapine
1-Year 330 (214) 295 25/850 305/334
Initiation 164.2 (151) 100 12.5 / 800 223 / 138
Endpoint 341 (277) 300 25 / 1300 369 / 330
Risperidone
1-Year 4.2 (2.3) 3.9 0.28/12.9 4.7/3.9
Initiation 2.7 (2.0) 2.0 0.25 / 12 3.2 / 2.4
Endpoint 4.3 (2.7) 4.0 0.25 / 16 4.5 / 4.1
* Mean Mono/Poly indicates the mean dose (mg/day) for patients while on antipsychotic monotherapy, and while on antipsychotic polypharmacy
Overall rates and duration of antipsychotic monotherapy/polypharmacy
Approximately one-third (34%) of the 796 patients were initiated on antipsychotic monotherapy with the index antipsychotic. Conversely, most patients (66%) received another antipsychotic, or polypharmacy, at the time of initiation. Only 30% of the polypharmacy-initiated patients were deemed to be in the process of medication change because their polypharmacy ceased within 60 days following initiation on the atypical agent.
The distribution of monotherapy and polypharmacy treatment categories during the 1-year period (Figure 1) indicates that about a third (35.7%) of the patients were treated predominately with monotherapy (>300 days), 26.9% were treated predominately with polypharmacy (>300 days), 30.2% had a mix of both substantial monotherapy and polypharmacy treatment periods (61–300 days of each), and 0.6% were not treated with any antipsychotic for more than 300 days. A small proportion of the patients (6.6%) had periods without antipsychotic medications along with substantial treatment periods with either monotherapy or polypharmacy (61–300 days).
Figure 1 Percent of patients in each monotherapy/polypharmacy treatment category. Abbreviations: Mono, monotherapy; Poly, polytherapy; No AP, no antipsychotic treatment. Definitions: Monotherapy: Predominant antipsychotic monotherapy (> 300 days); Polypharmacy: Predominant antipsychotic polypharmacy (> 300 days); Mix: Mono/Poly: Substantial periods of monotherapy (61 to 300 days) and substantial periods of polypharmacy (61 to 300 days), but no substantial periods of no antipsychotic treatment (< 60 days); No Antipsychotic: Patients predominately without prescribed antipsychotics (> 300 days); Mix: No AP/Mono: Patients predominately without prescribed antipsychotics (> 300 days) and substantial periods of monotherapy (61 to 300 days); Mix: No AP/Poly: Patients predominately without prescribed antipsychotics (> 300 days) and substantial periods of polypharmacy (61 to 300 days).
Overall, patients averaged 195.5 days on monotherapy (54% of the year), 155.7 days on polypharmacy (43% of the year), and 13.9 days without antipsychotic therapy (3% of the year). The distribution of the average number of polypharmacy days for each patient over the 1-year period (Figure 2) indicates that 40.6% had polypharmacy for less than 60 days. In fact, most patients (57.7%) had at least one prolonged period of antipsychotic polypharmacy (longer than 60 consecutive days).
Figure 2 Percent of patients with antipsychotic polypharmacy during the 1-year period by duration category.
Specific rates and duration of monotherapy/polypharmacy
A significantly greater percentage of olanzapine-treated patients were on antipsychotic monotherapy during the 1-year period compared to risperidone and to quetiapine. Olanzapine-treated patients were 2.08 times more likely to be on monotherapy than quetiapine (Odds Ratio (OR) = 2.08, 95% Confidence Interval (CI), 1.30–3.31, p = .002), and 1.36 times more likely to be on monotherapy than risperidone (OR = 1.36, 95%CI, 1.01–1.84, p = .043). Differences between risperidone and quetiapine initiators showed a trend toward significance (OR = 1.53, 95% CI, 0.94–2.47, p = .085).
The percentages of patients on antipsychotic monotherapy during every week in the 1-year following initiation for each treatment group are provided in Figure 3. Due to the strong influence of monotherapy status at initiation on later monotherapy status, and the treatment differences in monotherapy status at initiation, results are also shown stratified by monotherapy status at initiation in Figure 4.
Figure 3 Percent of patients on monotherapy during the 1-year following initiation on the atypical antipsychotic medication. Using a GEE (generalized estimating equations) repeated measures binary data model, pair-wise comparisons found a significantly higher rate of monotherapy for olanzapine compared to quetiapine (p = .002) and risperidone (p = .043). The difference between risperidone and quetiapine approached statistical significance (p = .085).
Figure 4 Percent of patients on monotherapy after initiation on antipsychotic polypharmacy or antipsychotic monotherapy. Abbreviations: OLZ, olanzapine; Poly, polypharmacy at initiation; QUE, quetiapine; RIS, risperidone; Mono, monotherapy at initiation
The observed mean number of monotherapy days on the initiating antipsychotic for olanzapine, quetiapine, and risperidone are presented in Table 3 for patients initiating on monotherapy and for patients initiating on polypharmacy. For monotherapy or polypharmacy initiators, patients on olanzapine had a greater number of days of monotherapy on the initiating medication while quetiapine initiators had the fewest. Using a propensity score bootstrap analysis, pair-wise comparisons found a significantly longer duration on monotherapy for olanzapine compared to quetiapine (p < .001) and a trend toward longer duration for risperidone initiators as compared to quetiapine (p = .052). The difference between olanzapine and risperidone was not statistically significant.
Table 3 Number of monotherapy days on the initiating antipsychotic during the 1-year post initiation
Status at Initiation
Treatment Monotherapy Polypharmacy Overall Mean
N Mean (SD) N Mean (SD) (SD)
Olanzapine 133 252.1 (132.8) 272 90.2 (126.8) 143.4 (149.5)
Quetiapine 27 159.8 (142.4) 88 53.6 (102.1) 78.5 (120.9)
Risperidone 111 230.6 (143.8) 165 78.0 (121.2) 139.3 (150.5)
Abbreviation: SD, standard deviation.
Discussion
In this large prospective naturalistic study of patients with schizophrenia, antipsychotic polypharmacy was found to be highly prevalent and to be of prolonged duration. Overall, most patients (57.7%) had at least one period of antipsychotic polypharmacy longer than 60 consecutive days, and only a third (35.7%) of the patients were treated predominately with monotherapy. More specifically, two thirds (66%) of the patients were treated with another antipsychotic at the time of initiation on the atypical agent, a practice that could have signaled a medication change process. However, most of these patients continued on antipsychotic polypharmacy for a substantial duration and only a small proportion of those patients (30%) were deemed to have gone through medication changes as polypharmacy ceased within the first 60 days after medication initiation. Findings suggest that for the majority of patients, polypharmacy is a prolonged and deliberate treatment choice rather than an interim, brief, or unintentional practice.
A third (34%) of the patients were not receiving another antipsychotic at the time of initiation on the studied atypical medications (olanzapine, quetiapine, and risperidone). These monotherapy-initiated patients continued on monotherapy for only about a third of the year post initiation, and at the end of that year, more than 50% of them were no longer receiving monotherapy with the initiating antipsychotic. Further attesting to the pervasive practice of antipsychotic polypharmacy is the finding that over 40% of all patients had no days of monotherapy with the initiating atypical antipsychotics during the 1-year treatment period. On the average, patients were treated with monotherapy for 54% of the year, with polypharmacy for 43% of the year, and without antipsychotic therapy for 3% of the year.
While polypharmacy was found to be prevalent in this patient population, the question of generalizability remains. However, a goal in designing the US-SCAP study was to generate a sample of patients representative of those treated in usual care. Participants in this large prospective naturalistic study were treated for schizophrenia at large public health care delivery systems in the United States, and were enrolled from multiple sites across six states. The patients were randomly identified from active client rosters at each site and only then approached about enrollment in the study. In addition, there were few exclusion criteria that would restrict the patient population.
This study provided a 1-year longitudinal perspective on the rate and duration of polypharmacy following initiation on the index antipsychotic whereas most previous studies assessed polypharmacy during shorter time periods, such as a 2-month window [15], or during inpatient hospitalization [16,19,25,26]. Generally, the larger is the studied time window, the higher is the likelihood of finding polypharmacy. Another major difference is the age of the data because polypharmacy has increased over the years. Studies using data from the earlier years after the introduction of the atypical antipsychotics [10,14] tend to report lower prevalence rates of antipsychotic polypharmacy, whereas studies using more recent data reported higher polypharmacy rates [8,15]. The complexity involved in comparing findings across studies is further compounded by variations in the definition of polypharmacy. Although most studies defined polypharmacy as any time with more than one antipsychotic [15], others have set specific time requirements, such as at least 14 days of concurrent antipsychotic use [8]. Other differences in study methods can generate different results. While the current study followed patients after their initiation on certain antipsychotic medications, other studies used a cross sectional method, assessing the prevalence of polypharmacy at a given time window, without using time of initiation on the studied antipsychotic medications as the reference point. While cross sectional designs provide important information about prevailing practices at a given time window, the cross sectional method is not well suited for comparisons between antipsychotic treatment groups. When the date of initiation on the antipsychotic is not used as a starting point, treatment group differences may be obscured by differential duration on the medication prior to the studied time window, data that are not included in cross sectional designs.
This study also found significant differences in monotherapy and polypharmacy between the most commonly prescribed atypical antipsychotic medications. Patients initiated on olanzapine were significantly more likely to be on monotherapy compared to quetiapine (rate and duration) or risperidone-initiated patients (rate only) during the 1-year post treatment initiation. The current findings appear consistent with several previous studies in which olanzapine-treated patients were found to have significantly higher rate of monotherapy compared to risperidone [18,19,28,29], and compared to quetiapine-treated patients [25,27,30], with the quetiapine treatment group being the least likely to be treated monotherapy [8,27]. Findings are particularly congruent with those reported in a study of medication patterns in the Michigan Medicaid database [28], in which the proportion of schizophrenia patients treated with olanzapine monotherapy remained relatively steady over the first 3 months of treatment while the proportion of patients receiving risperidone monotherapy decreased over time.
The present study helps demonstrate the dynamic and complex nature of medication management of patients with schizophrenia in usual clinical practice [14]. In addition to providing new information on the rate of monotherapy during treatment with various atypical agents, this study further contributes to the literature by providing a longitudinal perspective on the duration of monotherapy/polypharmacy following treatment initiation. The strengths of this study appear to lie in its large representative and diverse sample, the ability to provide comparative data on a number of commonly used atypical antipsychotics, the availability of comprehensive medication information about use of antipsychotics in depot formulation and about antipsychotics used during hospitalizations (types of data often absent in claims databases), and notably, the ability to generalize the findings to patients treated at large public systems of health care across the United States.
This study also has its limitations. First is the use of naturalistic observational data to compare antipsychotic treatment groups, because observed group differences in rates or duration of monotherapy could result from pre-existing differences between the treatment groups rather than differences in medication choice. Physicians tend to select treatments for different types of patients and illness profiles, a practice that may lead to lack of comparability between the treatment groups at initiation. Analyses that do not appropriately control for such differences can be biased, and one can never be certain that all important group differences have been controlled for statistically. Indeed, in this study differences were observed between groups at initiation. Although differences in several available pre-existing patient and treatment characteristics were controlled for in the analysis, it is possible that other pre-existing group differences were present. For instance, the design of this study did not include assessment of symptom severity or substance use at the time of initiation on the index antipsychotic. Thus, differences in symptom severity or substance use at initiation could not be controlled for.
Another study limitation is lack of information about reasons for initiating antipsychotic medication changes or the specific treatment indications. This type of information was not assessed in US-SCAP but could be valuable in discerning whether there is a link between treatment effectiveness and antipsychotic monotherapy (or polypharmacy). Treatment indication is also important because some antipsychotics, particularly quetiapine in low doses, may have been used to treat insomnia rather than core symptoms of schizophrenia. Although some have speculated that higher doses of quetiapine produce better outcome, the available data [35] do not support this. Davis and Chen [35] observed, however, "the sponsor of quetiapine has 2 studies of dose response, one nearing completion, and the other just beginning, so more information will be available in the future". They also noted "the fact that individuals do clinically improve at higher doses in open studies does not prove that the higher dose was responsible, as such improvement might reflect the passage of time, augmenting drugs, or other confounding factors."
Further, this study did not control for potential "sponsorship" bias that could have influenced clinicians' prescription rates and duration of monotherapy in favor of the sponsor's antipsychotic medication (in this case, olanzapine). Although we cannot completely rule out this possibility, sponsor bias is unlikely to have played a role in this study because the current findings are highly consistent with findings of other studies conducted by independent researchers who used non-industry-sponsored data such as Medicaid claims database [18,19,28,29]. Additional support for the validity of our findings and their freedom from sponsorship bias comes from a recent double-blind randomized study conducted by the sponsors of risperidone [35], comparing risperidone and quetiapine on the rate of antipsychotic polypharmacy. That short term trial found the relative risk (quetiapine vs. risperidone) of using antipsychotic polypharmacy to be higher for quetiapine compared to risperidone (odds ratio 1.90, 95% CI, 1.29–2.80, p = 0.001), a finding similar to that in our observational study, in which the odds ratio for antipsychotic polypharmacy of quetiapine vs. risperidone was 1.53 (95% CI 0.94–2.47, p = 0.085).
This study also did not present information about concomitant use of psychotropic medications other than antipsychotics, thus providing a partial understanding of the full range of psychotropic polypharmacy that takes place in usual practice. This is a complex phenomenone that will benefit from a separate and detailed study. We, however, cognizant of the potential role of prior concomitant psychotropic medications and used five types of psychotropics (antidepressants, anti-anxiety, antiparkinsonian, mood stabilizers, and sleep agents) in the analyses. Because prior use of concomitant psychotropics is highly correlated with continued use of these concomitant medications post initiation of the new antipsychotic regimen, we opted to include patients' prior concomitant medication as covariates in the analyses. The correlations between concomitant psychotropic medication use prior to initiation on the index drug and its concomitant use post initiation on the antipsychotic were consistently high, ranging from .65 to .81 (r = .81 for antidepressants, r = .80 for anti-anxiety agents, r = .73 for antiparkinsonians, r = .81 for mood stabilizers, and r = .65 for sleep agents). These correlations suggest that the bigger factor in the use of concomitant psychotropic medications post antipsychotic initiation is likely patients' prior medication history and not the antipsychotic monotherapy/polypharmacy status. Generally, the mean daily total number of concomitant psychotropic medications was higher for patients on antipsychotic polypharmacy (1.56) than for patients on antipsychotic monotherapy (1.24). Lastly, this study did not address the potential impact of polypharmacy on the total cost of antipsychotic treatment, or on patients' clinical outcomes, an area of investigation that will require further study.
Conclusion
This large prospective study of treatment in usual care settings demonstrated that antipsychotic medication management of schizophrenia patients is a complex process characterized by prevalent and prolonged polypharmacy. Current findings also highlight differences between the most commonly used atypical antipsychotics on the rate and duration of antipsychotic monotherapy, and its inverse, antipsychotic polypharmacy. Future research is needed to clarify the reasons for the observed treatment group differences, and to investigate the potential impact of antipsychotic polypharmacy on the costs of treatment and other important treatment outcomes in the long-term medication management of patients with schizophrenia.
Competing interests
The Schizophrenia Care and Assessment Program (SCAP-US) was funded by Eli Lilly and Company. The article-processing charge for this article was paid by Eli Lilly and Company. Dr. Kane serves as a consultant and/or lecturer for Eli Lilly and Company, Abbot, Bristol-Meyers Squibb, Pfizer and Janssen. Dr. Correll serves as a consultant and/or lecturer for Astra Zeneca, Bristol-Meyers Squibb, Elli Lilly and Company, and Janssen. Drs. Ascher-Svanum, Faries and Zhu are employees of Eli Lilly and Company. Drs. Ascher-Svanum, Faries and Zhu are minor stock holders in Eli Lilly and Company. Drs. Kane and Correll do not hold any stocks or shares in an organization that may in any way gain or lose financially from the publication of this manuscript, either now or in the future. The authors do not hold and are not applying for any patents relating to the content of the manuscript, and have not received reimbursements, fees, funding, or salary from an organization that holds or has applied for patents relating to the content of the manuscript. The authors have no other financial competing interests, and have no non-financial competing interests.
Authors' contributions
DF conceived of the study, designed the analytical plan, performed the statistical analyses, interpreted the results, and drafted the manuscript. HAS participated in study design, the analytical plan, interpretation of the results, and drafted the manuscript. BZ participated in study design, the analytical plan, statistical analyses, interpretation of the results, and the drafting of the manuscript. JK participated in the design of the analytical plan, interpretation of the results, and the drafting of the manuscript. CC participated in the design of the analytical plan, interpretation of the results, and the drafting of the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
American Psychiatric Association Practice Guidelines for the Treatment of Patients with Schizophrenia Am J Psychiatry 2004 161 2 American Psychiatric Association 1 56 10.1176/appi.ajp.161.1.1
Lehman AF Kreyenbuhl J Buchanan RW Dickerson FB Dixon LB Goldberg R Green-Paden LD Tenhula WN Boerescu D Tek C Sandson N Steinwachs DM The Schizophrenia Patient Outcomes Research Team (PORT): updated treatment recommendations 2003 Schizophr Bull 2004 30 193 217 15279040
National Institute for Clinical Excellence Guidance on the use of newer (atypical) antipsychotic drugs for the treatment of schizophrenia 2002 Technology Appraisal Guidance London
Miller AL Chiles JA Chiles JK Crismon ML Rush AJ Shon SP The Texas Medication Algorithm Project (TMAP) schizophrenia algorithms J Clin Psychiatry 1999 60 649 657 10549680
The expert consensus guideline series Treatment of schizophrenia J Clin Psychiatry 1999 3 80
Weiden PJ Casey DE "Polypharmacy": Combining antipsychotic medications in the treatment of schizophrenia J Prac Psych and Behav Hlth 1999 7 229 233
Miller AL Craig CS Combination antipsychotics: pros, cons, and questions Schizophr Bull 2002 28 105 109 12047009
Ganguly R Kotzan JA Miller LS Kennedy K Martin BC Prevalence, Trends, and Factors Associated With Antipsychotic Polypharmacy Among Medicaid-Eligible Schizophrenia Patients, 1998–2000 J Clin Psychiatry 2004 65 1377 1388 15491242
Tapp A Wood AE Secrest L Erdmann J Cubberley L Kilzieh N Combination antipsychotic therapy in clinical practice Psychiatr Serv 2003 54 55 59 12509667 10.1176/appi.ps.54.1.55
Loosbrock DL Zhao Z Antipsychotic medication use patterns and associated costs of care for individuals with schizophrenia Journal of Mental Health Policy Economics 2003 6 67 75
Schumacher JE Makela EH Griffin HR Multiple antipsychotic medication prescribing patterns Annals of Pharmacotherapy 2003 37 951 955 12841799 10.1345/aph.1C420
McCue RE Waheed R Urcuyo L Polypharmacy in patients with schizophrenia Journal of Clinical Psychiatry 2003 64 984 989 14628972
Stahl SM Antipsychotic polypharmacy: Squandering precious resources? J Clin Psychiatry 2002 63 93 94 11874226
Covell NH Jackson CT Evans AC Essock SM Antipsychotic prescribing practices in Connecticut's Public Mental Health System: Rates of changing medications and prescribing styles Schizophr Bull 2002 28 17 29 12047017
Weissman EM Antipsychotic prescribing practices in the Veterans Healthcare Administartion – New York Metropolitan Region Schizoph Bull 2002 28 31 42 12047020
Centorrino F Goren JL Hennen J Salavtore P Kelleher JP Baldessarini RJ Multiple versus single antipsychotic agents for hospitalized psychiatric patients: Case-control study of risks versus benefits Am J Psychiatry 2004 161 700 706 15056517 10.1176/appi.ajp.161.4.700
Clark RE Bartels SJ Mellman TA Peacock WJ Recent trends in antipsychotic combination therapy of schizophrenia and schizoaffective disorder: implications for state mental health policy Schizophr Bull 2002 28 75 84 12047024
Jerrell JM Cost-effectiveness of risperidone, olanzapine, and conventional antipsychotic medications Schizophr Bull 2002 28 589 605 12795493
Procyshyn RM Kennedy NB Tse G Thompson B Antipsychotic polypharmacy: a survey of discharge prescriptions from a tertiary care psychiatric institution Can J Psychiatry 2001 46 334 339 11387789
Yuzda MSK Combination antipsychotics: what is the evidence? J Inform Pharamcother 2000 2 300 305
Freudenreich O Goff DC Antipsychotic combination therapy in schizophrenia. A review of efficacy and risks of current combinations Acta Psychiatr Scand 2002 106 323 330 12366465 10.1034/j.1600-0447.2002.01331.x
Meltzer HY Kostakoglu A Combining antipsychotics: is there evidence for efficacy? Psychiatric Times 2000 17 25 32
Chiles JA Miller AL Crismon LM Rush AJ Krasnoff AS Shon SS The Texas medication algorithm project: development and implementation of the schizophrenia algorithm Psychiatr Serv 1999 50 69 74 9890582
Koshino Y Algorithm for treatment-refractory schizophrenia Psychiatry Clin Neurosci 1999 53 S9 S13 10560891
Jaffe AB Levine J Antipsychotic medication coprescribing in a large state hospital system Pharmacoepidemiol Drug Saf 2003 12 41 48 12616846 10.1002/pds.783
Advokat C Dixon D Schnedier J Comaty JE Comparison of risperidone and olanzapine as used under "real world" conditions in a state psychiatric hospital Progress in Neuro-Psychopharmacology & Biological Psychiatry 2004 28 484 495
Correll CU Kane JM O'Shea D Razi K Malhotra AK Antipsychotic polypharmacy inn the treatment of schizophrenia Schizophr Res 2003 37 12765741 10.1016/S0920-9964(03)80107-X
Gibson PJ Damler R Jackson EA Wilder T Ramsey J The impact of olanzapine, risperidone, or haloperidol on the cost of schizophrenia care in a Medicaid population Value in Health 2004 7 22 35 14720128 10.1111/j.1524-4733.2004.71272.x
Zhao Z Tunis SL Lage M Medication treatment patterns following initiation on olanzapine versus risperidone Clin Drug Invest 2002 22 741 749
Wang PF Zhao Z Comparison of olanzapine versus quetiapine in the treatment of hospitalized patients with schizophrenia Value in Health 2003 6 355
Ascher-Svanum H Zhu B Faries D Ernst FR A comparison of olanzapine and risperidone on the risk of psychiatric hospitalization in the naturalistic treatment of patients with schizophrenia Ann Gen Hosp Psychiatry 2004 3 Epub June 02, 2004
Swanson JW Swartz MS Elbogen EB Effectiveness of Atypical Antipsychotic Medications in Reducing Violent Behavior Among Persons with Schizophrenia in Community-Based Treatment Schizophr Bull 2004 30 3 20 15176758
Liang KY Zeger SL Longitudinal Data Analysis Using Generalized Linear Models Biometrika 1986 73 13 22
Chernick MR Bootstrap Methods: A Practicioner's Guide 1999 John Wiley & Sons, Inc. New York, NY
Davis J Chen N Dose response and dose equivalence of antipsychotics J Clin Psychopharmacol 2004 24 192 208 15206667 10.1097/01.jcp.0000117422.05703.ae
Rupnow M Stahl S Greenspan A Kosik-Gonzalez C Gharabawi G Use and cost of polypharmacy in schizophrenia: Data from a randomized double-blind study of risperidone and quetiapine Schizophr Bull 2005 31 550
| 15921508 | PMC1156914 | CC BY | 2021-01-04 16:33:02 | no | BMC Psychiatry. 2005 May 27; 5:26 | utf-8 | BMC Psychiatry | 2,005 | 10.1186/1471-244X-5-26 | oa_comm |
==== Front
BMC SurgBMC Surgery1471-2482BioMed Central London 1471-2482-5-111594637910.1186/1471-2482-5-11Research ArticleIs the routine drainage after surgery for thyroid necessary? - A prospective randomized clinical study [ISRCTN63623153] Khanna Jotinder [email protected] RS [email protected] [email protected] Dinesh [email protected] MK [email protected] M [email protected] Magan [email protected] Department of Surgery, Institute of Nuclear Medicine and Allied Sciences, Delhi, India2 Department of Cytopathology, Institute of Nuclear Medicine and Allied Sciences, Delhi, India3 Vardhman Mahavir Medical College Safdarjang Hospital, New Delhi-India4 Department of Radiology, Institute of Nuclear Medicine and Allied Sciences, Delhi, India2005 19 5 2005 5 11 11 13 11 2004 19 5 2005 Copyright © 2005 Khanna et al; licensee BioMed Central Ltd.2005Khanna et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Drains are usually left after thyroid surgery to prevent formation of hematoma and seroma in the thyroid bed. This is done to reduce complications and hospital stay. Objective evaluation of the amount collected in the thyroid bed by ultrasonography (USG) can help in assessing the role of drains.
Methods
A randomized prospective control study was conducted on 94 patients undergoing 102 thyroid surgeries, over a period of fifteen months. Patients included in the study were randomly allocated to drain and non-drain group on the basis of computer generated random number table. The surgeon was informed of the group just before the closure of the wound Postoperatively USG neck was done on first and seventh postoperative day by the same ultrasonologist each time. Any swelling, change in voice, tetany and tingling sensation were also recorded. The data was analyzed using two-sample t-test for calculating unequal variance.
Results
Both groups were evenly balanced according to age, sex, and size of tumor, type of procedure performed and histopathological diagnosis. There was no significant difference in collection of thyroid bed assessed by USG on D1 & D7 in the two groups (p = 0.313) but the hospital stay was significantly reduced in the non-drain group (p = 0.007). One patient in the drain group required needle aspiration for collection in thyroid bed. No patient in either group required re-operation for bleeding or haematoma.
Conclusion
Routine drainage of thyroid bed following thyroid surgery may not be necessary. Not draining the wound results in lesser morbidity and decreased hospital stay.
==== Body
Background
Most surgeons give into tradition of leaving a drain following thyroid surgery with the hope that this will obliterate the dead space and evacuate collected blood and serum. This belief is further reinforced by the fact that postoperative drains usually yield fluid. The need for drainage has been questioned after various types of surgeries with much larger potential dead spaces like cholecystectomy and colonic anastamosis [1,2]. These procedures are now routinely not drained. Blood and serum that they are supposed to drain usually block drains. They add to discomfort, give extra scar and increase hospital stay. We carried out a prospective randomized study to study the role of drains after thyroid surgery and monitor the fluid collection in thyroid bed objectively by USG.
Methods
The approval was taken from the reviewer board and ethical committee of the hospital before initiating the study and informed consent was taken from the patients regarding the use of drains. The study was carried out on 94 patients who underwent 102 thyroid surgeries in a single unit between January 2001 and March 2002. This discrepancy was due to 8 patients undergoing completion thyroidectomies for well differentiated thyroid carcinomas confirmed by histopathology.. These 8 patients were randomized as fresh cases with no consideration given to previous surgery. Patients with cervical lymph nodes metastases requiring neck dissection and those with clinical or laboratory indicators of coagulation disorders were excluded from the study. No patient was excluded on the basis of size of the gland, difficulty in surgery, surgery involving both lobes and re-operation in the neck. Alll patients underwent routine preoperative and postoperative laryngoscopy(indirect/direct) as a part of the protocol at our center. The patients were randomly allocated to drain and non-drain group on the basis of computer generated random number table. The operating surgeon was informed of the group just before the closure of the wound. In the drain group a closed suction drain with negative pressure (Romovac®) was brought out through a separate wound. Ultra sound of the neck using B mode with linear frequency of 7.5 MHz was performed in both groups between 24 – 48 hours of surgery and seventh postoperative day each time by the same radiologist or under his supervision. Volume of fluid collection in the operative bed was calculated by measuring the maximum diameter in three dimensions. The volume of fluid collected in the suction drain was measured separately. The drains were removed in all the patients after 48 hours. Before the present study was contemplated, a pilot study was carried out on 20 patients(excluded from the present study)to ascertain the duration of drainage and the drains were removed after the drainage reduced to less than 30 ml in 24 hours following which the patients were discharged. It was observed that the average time taken for the drains to be removed was 4 days while the ultrasound assessment did not reveal any collection in the thyroid bed after 48 hours. The review committee of the hospital therefore recommended keeping of the drains for 48 hours in order to compare the morbidity in the two groups. Therefore a cut off period of 48 hours for removal of drains was considered. All patients were observed for any postoperative respiratory distress, change in voice, wound collection, tingling sensation and tetany. The specimens were subjected to histopathological examination for final confirmation of diagnosis. Using two sample "t" test all the data was statistically analyzed for any significant difference in the two groups for a) fluid collection in thyroid bed on day one and day seven, b) size of nodule and amount of fluid collection and c) complication rate.
Results
The average age of patients in the present study was 34.56 years (range 8–60 years). Male to female ratio was 1: 6.84, with equal distribution in both the groups based on the type of surgery and size of nodule. The amount of fluid collection in thyroid bed as assessed by USG for both the groups on day one and day seven is shown in Table 1. Two-sample T- test was applied for detecting any difference in means of fluid collected between the two groups. There was no statistically significant difference in the volume of fluid collection on day one (p = 0.371) and day seven (p = 0.577) between the two groups. On the other hand the amount of fluid collected in the suction drain was noted for 48 hours. The average collection was 167.14 ml (range 30 – 120 ml/ day).
Table 1 Volume of fluid collection in the two groups as assessed by USG on D 1 & D7
Amount of fluid Drain group Non drain group
D1 D7 D1 D7
Mean 3.258 ml 1.819 ml 2.345 ml 1.366 ml
Minimum 0 ml 0 ml 0 ml 0 ml
Maximum 40 ml 35 ml 19 ml 14.2 ml
Total number of patients 51 51 51 51
Two sample T test for detection of difference in means of fluid collected on D1 & DD7 respectively
Unequal variance T DF P
D1 0.9 81.5 0.371
D7 0.58 72.9 0.577
Average size of thyroid nodule was 4.03 cm & 4.12 cm in the drain and non-drain groups respectively. Kruskal – Wallis one way non parametric analysis showed no significant statistical difference in the amount of collection according to size of nodule arbitrarily taken as < 4 cm & > 4 cm on D1 & D7 respectively (Table 2 &3).
Table 2 Chi Squared approximation for difference in volume of collection based on size on D1
Amount of fluid in ml Drain group Non drain group
<4 cm >4 cm < 4 cm >4 cm
Mean D1 2.30 ml 5.935 ml 2.452 ml 2.011
Total number of patients 37 14 34 17
P = 0.313 (ns)
Table 3 Chi Squared approximation for difference in volume of collection based on size on D7
Amount of fluid Drain group Non drain group
<4 cm >4 cm < 4 cm >4 cm
Mean 0.956 ml 1.956 ml 1.388 ml 1.244 ml
Total number of patients 37 14 34 17
P = 0.0712 (ns)
Average duration of hospital stay was 3.715 days for the entire group; 4.35 days for drain group and 3.07 days for non-drain group. This was statistically significant when analysed using two-sample T test (P = 0.0072).
Complications that were observed in the present study are shown in Table 4. Of the six patients who developed collection, four had undergone bilateral subtotal thyroidectomy with an average amount 13.40 ml. Only one patient from the drain group required single aspiration. None of the patients had respiratory distress. Three patients had tingling sensation and two patients developed transient tetany. Five patients developed transient change in voice four of these belonged to drain group, the difference being non significant (p= 0.36). One patient in each group developed wound infection. The presence or absence of drains, expectedly did not contribute significantly to the postoperative complications.
Table 4 Distribution of complications into drain and non-drain group
Complication Drain group Non drain group
Swelling 3 3
Tetany 0 2
Tingling 1 2
Infection 1 1
Discussion
Drains have been traditionally used in most of the surgical procedures with limited evidence to suggest any benefit [1-4]. The present prospective randomized study has failed to show any advantage of routinely using drain after uncomplicated thyroid surgery. Except for patients undergoing simultaneous neck dissection or coagulation disorder no other exception was made based on factors mentioned above.
Other authors in randomized studies reported similar results with sample size varying from 100 to 200 patients (5 -10). Two large non-randomized studies of 250 and 400 patients have also documented no benefit of using drains after thyroid surgery [11,12]. Objective assessment of fluid collection in thyroid bed by USG has been done very infrequently in randomized settings. Debry et al have used it selectively in two patients developing postoperative haematoma while Schwartz used it to compare two types of drains [5,13]. In the present study absence of fluid in the thyroid bed on USG but its presence in the suction drain could therefore be due to the drain itself [8]. The drains by virtue of the inflammation induced due to their presence may actually increase the drainage. The vaccum created by the negative suction of the drain may prevent the lymphatics from sealing off and thus cause increase in the seroma formation and drainage [5-9].
Conclusion
The present randomized study highlights that placement of drains after routine thyroid surgery may induce rather than prevent fluid collection, is not related to the type of surgery or size of nodule, has no influence on complications, leads to an extra scar and may increase the hospital stay (if the patients can not be discharged with drains in situ). Meticulous haemostasis and attention to finer details during surgery are more important. Routine use of drains after thyroid surgery may therefore not be necessary.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
DB was the chief surgeon in charge of the unit, the department of surgery also responsible for the designing of the study, RSM, CM were the senior operating surgeons, JK, MM were the senior residents in charge of the cases did the statistical analysis, MK did the ultrasonographic assessment and MS was responsible for the pathological analysis. All authors read and approved the manuscript.
Acknowledgements
We would like to acknowledge with gratitude the contribution of Dr Rajvir Singh PhD (Biostatistics) Statistician All India Institute Of Medical sciences for the statistical analysis.
Pre-publication history
The pre-publication history for this paper can be accessed here:
==== Refs
Lewis RT Goodall RG Marien B Park M Llyod-Smith W Weigand FM Simple elective cholecystectomy; to drain or not Am J Surg 1990 159 242 245
Hoffmann J Lorentzen M Drainage after cholecystectomy Br J Surg 1985 72 423 427 3893617
Hoffmann J Shokouh-Amiri M Damm P Jensen R A prospective controlled study of prophylactic drainage after colonic anastomoses Dis colon rectum 1987 30 449 452 3595364
Johnson CD Lamont PM Orr N Lennox M Is drain necessary after colonic anastomosis? JR Soc Med 1989 82 661 664
Debry C Renou G Fingerhut A Drainage after thyroid surgery: a prospective randomised study The Journal of Laryngology and Otology 1999 113 49 51 10341919
Peix JL Teboul F Feldman H Massard JL Drainage after thyroidectomy: a randomized clinical trial Int surg 1992 77 122 124 1644539
Ayyash K Khammash M Tibblin MS Drain Vs no drain in primary thyroid and para thyroid surgery European Journal of Surgery 1991 157 113 114 1676302
Wihlborg O Bergljung L Matensson H To drain or not to drain in thyroid surgery. a controlled clinical study. a controlled clinical study Archives of surgery 1988 123 40 41 3276296
Kristofferson A Sandzen B Jarhult J Drainage in uncomplicated thyroid and para thyroid surgery Br J Surg 1986 73 121 122 3947901
Schoretsanitis G Melisas J Sanidas E Christodoulakis M Vlachonikolis JG Does draining the neck affect morbidity following thyroid surgery? Am surgeon 1998 64 778 780 9697913
Shaha AR Jaffe BM Selective use of drain in thyroid surgery Journal of Surgical Oncology 1993 52 241 243 8468987
Arriyanayagam DC Narayan Singh V Busby D Sieunarine K Raju G Thyroid surgery without drainage; 15 years of clinical experience Journal of Royal College of Surgery Edinburgh 1993 38 69 70
Schwartz W Willy Ndzee C Gerugross H Gravity or suction drainage in thyroid surgery? Control of efficacy with ultrasound determination of residual haematoma Archives of Surgery 1996 381 337 342
| 15946379 | PMC1156915 | CC BY | 2021-01-04 16:28:04 | no | BMC Surg. 2005 May 19; 5:11 | utf-8 | BMC Surg | 2,005 | 10.1186/1471-2482-5-11 | oa_comm |
==== Front
BMC SurgBMC Surgery1471-2482BioMed Central London 1471-2482-5-121592152710.1186/1471-2482-5-12Study ProtocolPostoperative Delirium after elective and emergency surgery: analysis and checking of risk factors. A study protocol Agnoletti Vanni [email protected] Luca [email protected] Fausto [email protected] Rabbih [email protected] Cataldis Angelo [email protected] Nino Gianfranco [email protected] Claudio [email protected] Stefano [email protected] Rita Maria [email protected] Antonella [email protected] Mario [email protected] Department of Anaesthesiology, "S. Orsola-Malpighi" Hospital, University of Bologna, Italy2 Department of Emergency Surgery, S. Orsola-MalpighiHospital, University of Bologna, Italy3 Department of Psychology, University of Bologna, Italy4 Department of Experimental Pathology, University of Bologna, Italy2005 28 5 2005 5 12 12 4 3 2005 28 5 2005 Copyright © 2005 Agnoletti et al; licensee BioMed Central Ltd.2005Agnoletti et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Delirum is common in hospitalized elderly patients and may be associated with increased morbidity, length of stay and patient care costs. Delirium (acute confusional state) is defined as an acute disorder of attention and cognition. In elderly patients, delirium is often an early indicator of patho-physiological disturbances. Despite landmark studies dating back to the 1940s, the pathogenesis of Delirium remains poorly understood. Early investigators noted that Delirium was characterized by global cortical dysfunction that was associated predominantly with specific electroencephalographic changes. It's important to understand the risk factors and incidence of Delirium. Some of the risk factors are already identified in literature and can be summarized in the word "VINDICATE" which stands for: Vascular, Infections, Nutrition, Drugs, Injury, Cardiac, Autoimmune, Tumors, Endocrine. Aims of this study are: to re-evaluate the above mentioned clinical risk factors, adding some others selected from literature, and to test, as risk factors, a pattern of some genes associated to cognitive dysfunction and inflammation possibly related to postoperative Delirium.
Design
All patients admitted to our Emergency Unit who are meet our inclusion/exclusion criteria will be recruited. The arising of postoperative Delirium will select incidentally two groups (Delirium/non Delirium) and the forward analysis of correlate risk factors will be performed. As in a typical observational case/control study we will consider all the exposure factors to which our population are submitted towards the outcome (presence of Delirium). Our exposures are the following: ASA, Pain (SVS; VAS), Blood gas analysis (pH; Hb; pO2; pCO2), Residence pharmacological therapy (BDZ; hypnotics; narcotic drugs; alcohol; nitrous derivates), Body temperature, Arterial pressure, Heart frequency, Breath frequency, Na, K, Creatinin, Glicemia, Albumin, Hct, White blood cells, Glasgow Coma Scale (GCS), Cognitive state (SPMSQ), Functional state (ADL and IADL), Psychological Distress (HADS), Cumulative Illness Rating Scale (CIRS), Hypotension (classified in: light; moderate and severe and duration), Blood loss (classified in: < 2 lt and > 2 lt), Blood transfusions (< 2 lt and > 2 lt), Quantity of red cells and plasma transfusions, Visual VAS / SVS (timing: I-II-III post-operative day), Red cells and Plasma transfusions, Blood count evaluation and Saturation (O2%), Postoperative analgesia (Emilia-Romagna protocol), Presence of malignant tumoral disease, APACHE Score II. Moreover the presence of some relevant genetic polymorphisms will be studied in different genes such as IL-6, IL-10, TNF-alpha, and IL-1 cluster.
==== Body
Background
Delirum is common in hospitalized elderly patients and may be associated with increased morbidity[1], length of stay and patient care costs[2]. The classic manifestations of this syndrome are impaired cognition and decreased ability to maintain attention[3]. Efforts to understand this syndrome require a thorough understanding of its causes and the ability to predict who is at risk. Each year Delirium complicates hospital stays for more than 2.3 milion older people, involves more than 17.5 million inpatients days, and accounts for more than $4 billion of Medicare expenditure[4]. Substantial additional costs accrue after discharge from the hospital, because of the increased need for institutionalazisation, rehabilitation, and home care[5].
Delirium (acute confusional state) is defined as an acute disorder of attention and cognition[6].
It represents a syndrome of disruption of one's state of consciousness, concentration, perception, memory, cognition, orientation and psychomotor behavior. The most prominent symptoms among these are the inability to concentrate and changes in the state of alertness[7].
The reported incidence of Delirium in acutely ill elderly patients during hospitalization ranges from 7 to 61.3% in the US depending on the population studied and the criteria used for diagnosis[8]. In elderly patients, delirium is often an early indicator of patho-physiological disturbances.
It's important to understand the risk factors and incidence of Delirium, because we agree with the Inouye's model of the cumulative effects of baseline vulnerability factors and precipitating factors for Delirium. Baseline vulnerability factors are defined as predisposing factors for Delirium present upon the admission. Precipitating factors are defined as noxious insults or hospitalization-related factors that contribute to Delirium[9].
Despite landmark studies dating back to the 1940s, the pathogenesis of Delirium remains poorly understood. Early investigators noted that Delirium was characterized by global cortical dysfunction that was associated predominantly with specific electroencephalographic changes. These findings suggest an abnormality on a biochemical and electrophysiological level.
Although Delirium can develop at any time during hospitalization, it typically presents early in the postoperative period. A good preoperative evaluation should include a formal cognitive assessment in patients at risk of developing Delirium. Delirium usually persists for hours to days and can fluctuate throughout the course of the day. Although several formal cognitive test are useful in identifying Delirium, including the Mini-Mental-Status-Exam (MMSE) [10], the Confusion-Assessment-Method (CAM) and the Delirium Writing, as the MMSE can't distinguish Delirium from dementia, the CAM remains the most suitable test to check the presence of Delirium.
Some of the risk factors are already identified in literature and can be summarized in the word "VINDICATE" which stands for: Vascular, Infections, Nutrition, Drugs, Injury, Cardiac, Autoimmune, Tumors, Endocrine. During last year in the Emergency Surgery Unit of S. Orsola-Malpighi Hospital (Bologna, Italy) a pilot study was performed with the aim of checking and analyzing some of the most important risk factors for postoperative Delirium by using an observational case/control study. A series of 100 over 65 years old patients, admitted for either emergency or elective surgery, were included and submitted to a psychometric, clinical and biochemical assessment. Among all risk factors examined the following were statistically significantly related to postoperative Delirium:
• Origin from home or other medical garrisons;
• Cardiovascular pathology;
• Metabolic diseases (hypo-hyper-tyroidism; glycemic disturbances)
• ECG abnormalities (ECG);
• Intraoperative hypotension;
• Preoperative low level of hemoglobin;
• Preoperative low level fo hematocritus;
• Postoperative hypoglicemia;
• Use of nitrous derivate during anesthesia;
• Use of anticolynergic drugs during surgery;
• Blood and plasma transfusion during surgery;
• Mini Mental Status (MMS);
• ADL (Activity Day Living).
The aims of the present study are the followings:
• to re-evaluate the above mentioned clinical risk factors, adding some others selected from the literature.
• to test, as risk factors, a pattern of some genes associated to cognitive dysfunction and inflammation possibly related to postoperative Delirium.
Methods
Design
Observational analytical case/control study (outcome-exposure).
All patients admitted to our Emergency Unit who meet our inclusion/exclusion criteria will be recruited. The arising of postoperative Delirium will select incidentally two groups (Delirium/non Delirium) and the forward analysis of correlate risk factors will be performed.
As in a typical observational case/control study we will consider all the exposure factors to which our population are submitted towards the outcome (presence of Delirium).
Our exposures are the following:
1) Preoperative
• ASA
• Pain (SVS; VAS)
• Blood gas analysis (pH; Hb; pO2; pCO2)
• Residence pharmacological therapy (BDZ; hypnotics; narcotic drugs; alcohol; nitrous derivates)
• Body temperature
• Arterial pressure
• Heart frequency
• Breath frequency
• Na; K; Creatinin; Glicemia; Albumin
• Hct
• White blood cells
• Glasgow Coma Scale (GCS)
• Cognitive state (SPMSQ)
• Functional state (ADL and IADL)
• Psychological Distress (HADS)
• Cumulative Illness Rating Scale (CIRS) (classified in 4 different steps):
- heart;
- hypertension;
- cardiovascular diseases;
- respiratory tract diseases;
- othorinolaryngoiatry diseases;
- digestive tract diseases (divided in to two tracts: proximal and distal);
- liver diseases;
- kidney diseases;
- urogenital diseases;
- bone and muscle diseases;
- neurological diseases;
- endocrino-metabolic diseases;
- mind status and behavior;
2) Intraoperative
• Hypotension (classified in: light; moderate and severe and duration);
• Blood loss (classified in: < 2 lt and > 2 lt);
• Blood transfusions (< 2 lt and > 2 lt);
• Quantity of red cells and plasma transfusions.
3) Postoperative
• VAS / SVS (timing: I-II-III post-operative day);
• Red cells and Plasma transfusions;
• Blood count evaluation and Saturation (O2%);
• Postoperative analgesia (Emilia-Romagna protocol);
• Presence of malignant tumoral disease;
• APACHE Score II
4) Genetic
The presence of some relevant genetic polymorphisms will be studied in different genes such as IL-6, IL-10, TNF-alpha, and IL-1 cluster. The polymorphisms are the following :
1. -174 (C/G) at 5'-upstream of IL-6[15];
2. -1082 (G/A) at IL-10 promoter sequence[13],
3. -308 (G/A) at TNF-alpha promoter sequence[14],
4. -889 (C/T) in IL-1 alpha; -511 (C/T) in IL-1 beta; IL-1 Receptor Antagonist (86 bp repeated sequenze in introne 2)[11].
All these polymorphisms have an important role for immune response/inflammation pathways[12] and in this project their frequency will be correlated to postoperative Delirium insurgence. The products of these genes will be also measured in the plasma, together with cortisol levels, because the last is known to have an important role in Immune-Neuroendocrine System and in Delirium insurgence[16].
Blood samples (15–20 ml) from the enrolled patients at S. Orsola-Malpighi Hospital, at Department of Emergency Surgery, will be collected in tripotassium EDTA sterile tubes. These samples will be transported to the Laboratory of Immunology, directed by Prof Claudio Franceschi (Dipartimento di Patologia, Sperimentale, via S. Giacomo 12, Bologna). Plasma will be harvested as soon as possible and stored at -80°C, while the cell pellet will be stored separately at -80° for further DNA extraction. Genomic DNA extraction will be carried out by classical phenol/chloroform techniques according to standard procedures and then stored at -20°C. The genetic polymorphisms, will be studied according to the detailed protocol reported in above mentioned papers.
Briefly, genotyping analysis will be performed by means of PCR methods and gel electrophoresis analysis to test all genes, but IL-10 polymorphisms that will require a slight different method, named ARMS(Amplification Refractory Mutation System)-PCR, which is applied for single base biallelic polymorphism/mutation[13]. In particular, the study of single base biallelic polymorphism -1082 (C/A) at IL-10 promoter sequence will be analysed using the protocol here described. Two different 5' primers respectively specific for IL-10 -1082C and -1082A are separately mixed with a 3' primer in the following conditions: 200 ng of template DNA are mixed with a final concentration of 2 U of TaqGold-DNA polymerase, 200 mmol/L each of deoxynucleotyde and 1X reaction buffer and 0.5 mmol/L of each specific primer. Cycling is performed at 98°C for 12 min and 35 cycles at 96°C for 30, 54°C for 30, and 72°C for 30, followed by a final extension of 10 min at 72°C. PCR products are detected by electrophoresis on 2% agarose.
The plasma levels of gene products, i.e. the IL-6, IL-10, IL-1, TNF-alpha cytokines will be evaluated by Enzyme Linked ImmunoSorbent Assay (ELISA) standard procedures. Cortisol will be evalueted by usual biochemical-haematological assays.
DNA samples will be stored in the Department of Experimental Pathology, Section of Immunology, in the respect of the Italian current law related to privacy protection.
In case that some patients will decide to leave the study, DNA samples will be destroyed by hydrochloric acid pre-treatment and then disposed off in the biological refusals.
The presence of Delirium among patients enrolled in our study will be evaluated through CAM test interviews at the 1st, 2nd, 3rd and 6th postoperative day. To assess the severity of Delirium (previously diagnosed through CAM) the Delirium Rating Scale (DRS) will be used.
The follow-up of patients enrolled in this study will be completed by two others checks (at 3 and 6 months after the operation) in order to establish a possible correlation between the arising of postoperative Delirium and development of permanent cognitive dysfunctions and Dementia.
Sample Size
The exact sample size will be 300 patients with a probable distribution of 1/4,26 in the two groups (Delirium/non Delirium), that has been calculated through our above mentioned pilot study. This study showed 19 Delirium patients over 100 included: for this reason we are expecting, among our population of 300 patients, 57 with postoperative Delirium.
Power calculation
Sample size has been calculated to reach a confidence level of 95% with a power of 80% according to the results of the above mentioned pilot observational case/control study on 100 patients previously admitted to our Operative Unit. In this study 19 patients on 100 presented postoperative Delirium with a ratio, between two groups (81 not Delirium vs. 19 Delirium), of 4.26. Furthermore we selected MMSE with a cut off value of 7 as a variable considered as a risk factor in developing postoperative Delirium. Then we calculated the ratios of patients in the two groups with a value of MMSE below 7. This ratio resulted 0.26 for those who presented postoperative Delirium and 0.09 for the other group. So we calculated our sample size by a Proportion Difference Power/Sample Size Calculation using an online statistical calculator[17]. The calculated sample size was of 257 patients (49 Delirium and 208 Not Delirium). We expect to enrol 300 patients.
Inclusion and exclusion criteria
Inclusion criteria are:
• Age > 65 years
Exclusion criteria
• Any condition preventing a correct evaluation of cognitive functions such as speech and sensory impairment or psychiatric disorders (language difficulties or psychiatric organic dysfunctions) which make difficult the administration of psychometric tests.
Informed consent
In the informed consent form, patients will receive all the information about the study protocol, the confidential nature of personal data and will fill up a questionnaire before signing for acceptance or refusal, including the authorization for taking a sample of blood. All the medical information obtained from the patients will be kept confidentially among the research scientists conducting the study. In case of refusal to participate, there will be not any inconvenience to the patients. Patients will be free to withdraw from the study whenever they want without any obligation.
Ethical approval
Approved by Ethical Committee of S. Orsola-Malpighi Hospital on March 1st, 2005
Stopping rules
In our study we don't include the experimental use of any drug or surgical procedures, so we don't expect any inconvenience or complication.
Primary endpoint
The evaluation of risk factors above mentioned and their power of correlation with incidence of Delirium.
The evaluation of Delirium presence as a possible risk factor in developing successive Dementia.
Interim analysis
A preliminary statistical analysis will be performed after enrolling the first 150 patients.
Data management – type of analysis – statistics test stated
All our data will be reported on a form on purpose created and subsequently collected in an appropriate computer database hold in the Emergency Surgery Unit.
After a preliminary descriptive analysis, univariate analysis will be performed by using Odds Ratio for categorical variables and T-test or U-Mann-Withney test for continuous variables depending on their distribution (Normal/Not normal) as appropriate. Multivariate analysis will be performed by using Logistic Regression (forward-conditional) introducing only variables estimated statistically significant in the univariate analysis. All statistical analysis will be performed using the statistical software SPSS 8.0 for Windows.
Indemnities specified
No incentives are planned for the patients regarding the operation or the follow-up.
Is the study clinically necessary?
Postoperative Delirium is one of the factors which influences morbidity and mortality after elective and emergency surgery and even engraves on economical budget of the Surgical Units.
Check and knowing the most important risk factors could help in establish prevention measures in order to reduce postoperative Delirium's incidence and could let us to manage it appropriately.
Time frame – finishing date – reporting date
We expect to start the study next June. According to the number of patients admitted monthly in our Unit, the duration of the study can be approximately estimated of about one year.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
VA, LA, RC, AD, CF, AP, edited the manuscript
RMM, GD, MT drafted the manuscript
AD, SG performed the statistical analysis
All Authors participated in the design of the study.
All Authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
No sponsor supports this study. All the funds belong to our Institutions.
==== Refs
American Psychiatric Association Diagnostic and statistical Manual of Mental Disorders, Washington, DC 1994 4 American Psychiatric Association: Washington DC
Williams MA Campbell EB Raynor WJ Reducing acute confusional states in elderly patients with hip fractures Res Nurs Health 1985 8 329 337 3853245
Lipowsky ZJ Delirium (acute confusional states) JAMA 1987 258 1789 1792 3625989 10.1001/jama.258.13.1789
The Administration on Aging A profile of older Americans Washington DC: American Association of Retired Persons 1995
Inouye SK Rushing JT Foreman MD Palmer RM Pompei P Does delirium contribute to poor hospital outcomes? A three site epidemiologic study J Gen Intern Med 1998 13 234 242 9565386 10.1046/j.1525-1497.1998.00073.x
Inouye SK van Dick CH Alessi CA Clarifyng confusion: the confusion assessment method – a new method for detection of delirium Ann Intern Med 1990 113 941 948 2240918
Mesulam M Attention, confusional state and neglect; in Mesulam M (ed): Principles of Behavioral Neurology Philadelphia, Saunders 1985 125 168
Foreman MD Confusion in the hospitalised elderly: incidence, onset and associated factors Res Nurs Health 1989 12 21 29 2922488
Inouye SK Charpentier PA Precipitating factors for delirium in hospitalized elderly patients: predictive model and interrelationship with baseline vulnerability JAMA 1996 275 852 857 8596223 10.1001/jama.275.11.852
Dyer CB Ashton CM Teasdale TA Postperative Delirium Arch Intern Med 1995 155 461 465 7864702 10.1001/archinte.155.5.461
Cavallone L Bonafe M Olivieri F Cardelli M Marchegiani F Giovagnetti S Di Stasio G Giampieri C Mugianesi E Stecconi R Sciacca F Grimaldi LM De Benedictis G Lio D Caruso C Franceschi C The role of IL-1 gene cluster in longevity: a study in Italian population Mech Ageing Dev 2003 124 533 8 12714264 10.1016/S0047-6374(03)00033-2
Franceschi C Olivieri F Marchegiani F Cardelli M Cavallone L Capri M Salvioli S Valensin S De Benedictis G Di Iorio A Caruso C Paolisso G Monti D Genes involved in immune response/inflammation, IGF1/insulin pathway and response to oxidative stress play a major role in the genetics of human longevity: the lesson of centenarians Mech Ageing Dev 2005 126 351 61 15621218 10.1016/j.mad.2004.08.028
Lio D Scola L Crivello A Colonna-Romano G Candore G Bonafe M Cavallone L Franceschi C Caruso C Gender-specific association between -1082 IL-10 promoter polymorphism and longevity Genes Immun 2002 3 30 3 11857058 10.1038/sj.gene.6363827
Lio D Scola L Crivello A Colonna-Romano G Candore G Bonafe M Cavallone L Marchegiani F Olivieri F Franceschi C Caruso C Inflammation, genetics, and longevity: further studies on the protective effects in men of IL-10 -1082 promoter SNP and its interaction with TNF-alpha -308 promoter SNP J Med Genet 2003 40 296 9 12676903 10.1136/jmg.40.4.296
Olivieri F Bonafe M Giovagnetti S Stecconi R Cardelli M Cavallone L Spazzafumo L Marchegiani F Carrieri G Mugianesi E Giampieri C Centurelli M Moresi R Tesei S Lisa R Viticchi C Falsetti L Salvioli S Franceschi C In vitro IL-6 production by EBV-immortalized B lymphocytes from young and elderly people genotyped for -174 C/G polymorphism in IL-6 gene: a model to study the genetic basis of inflamm-aging Mech Ageing Dev 2003 124 549 53 12714266 10.1016/S0047-6374(03)00035-6
Olsson T Activity in the hypothalamic-pituitary-adrenal axis and delirium Dement Geriatr Cogn Disord 1999 10 345 9 10473937 10.1159/000017168
| 15921527 | PMC1156916 | CC BY | 2021-01-04 16:28:04 | no | BMC Surg. 2005 May 28; 5:12 | utf-8 | BMC Surg | 2,005 | 10.1186/1471-2482-5-12 | oa_comm |
==== Front
Biomed Eng OnlineBioMedical Engineering OnLine1475-925XBioMed Central London 1475-925X-4-331590452210.1186/1475-925X-4-33Book ReviewReview of "Advances in Silicon Carbide Processing and Applications" 1st Edition by Steven E. Saddow and Anant Agarwal Horsfall Alton [email protected] School of Electrical, electronic and Computer Engineering, University of Newcastle, Newcastle, Tyne and Wear, NE1 7RU, UK2005 17 5 2005 4 33 33 Advances in silicon carbide processing and applications .
Saddow , Steven E , and Anant Agarwal . Artech House, Norwood, MA 02062 . ISBN 1580537405 xiv+212 pages . 11 5 2005 17 5 2005 Copyright © 2005 Horsfall; licensee BioMed Central Ltd.2005Horsfall; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
==== Body
"Advances in Silicon Carbide Processing and Applications" specifically targets the technology of two key application areas, propulsion systems in electronic vehicles and sensors for deployment in extreme environments. Edited by Steven Saddow & Anant Agarwal, two highly respected researchers in the field of SiC, with contributions from a range of eminent scientists, the book offers a real insight to these fascinating areas. The technology lead approach on these two areas sets the book apart from other recent offerings, such as 'Process Technology for SiC Devices', by Zetterling [1] and 'Silicon Carbide, Recent Major Advances', by Choyke, Matsunami and Pensl [2].
After the introductory chapter one the growth and potential market of silicon carbide wafers, is suitable for those people who are new to the field of SiC and offers a strong background to support the remainder of the material presented in the later chapters. The indication that the market forces for SiC power electronics will be in the 300V plus market however, goes against the significant inroads made by both Infineon [3] and Cree [4] into the switch mode power supply market for domestic customers. This is followed by a chapter which focuses on the technological requirements of implantation and annealing for the production of the selectively doped regions required for device fabrication. With the crystal lattice of SiC being particularly dense, the damage caused by ion implantation is severe and significantly degrades the carrier mobilities in these regions. The chapter provides a thorough review of the techniques being used to investigate the extent of this damage and the measures being used to minimise it in device structures.
The chapters at the rear of the book focus on the development of power transistors (both MOSFETS and bipolars) and their suitability for replacing silicon based power devices in power electronic circuitry, as would be suitable for propulsion systems. Because SiC devices offer the possibility of operation with junction temperatures substantially above the 175°C limit associated with Si power devices, they can be operated without the need of heavy and cumbersome heat sinks. This, along with their higher efficiency makes them particularly suitable for mobile propulsion systems. The material properties of SiC also make it suitable for the fabrication of high power radio frequency devices, such as IMPATT diodes and bipolar transistors and the final section shows the recent research in this area which offers a potentially lucrative market for SiC devices as they offer superior properties in comparison to silicon.
The remaining chapter focuses on the development sensor structures for deployment in environments beyond those possible with current technology, such as the detection of gas species in exhaust systems for both transportation and domestic systems. The chapter offers a review of the latest results from research groups in both Sweden and the US, who are arguable at the forefront of research in this field. This includes an introduction to the operation of the devices and the importance of the dissociation of gas species with a catalytic metal, before demonstrating the influence of a range of gas species (such as hydrogen, nitrous oxides and ammonia) to the behaviour of these devices. Results in the chapter illustrate the response of these sensors to gas species whilst being deployed in situ, such as a car exhaust, where a mis-firing cylinder can be easily identified from the change in gas mixture in the manifold and give details about the practical nature of deploying these sensors, such as mounting the SiC device and electrical connections. The application of this technology to biomedical applications is an area ripe for development, as silicon carbide material properties are far superior to those observed in traditional semiconductors, such as silicon. To date only a single reference exists showing development in this field [5]. The content of this chapter alone is sufficient to warrant the purchase of this book and given the popularity of this volume with our postgraduate students, I would recommend it to anyone wishing to gain more knowledge in this exciting field of work.
==== Refs
Zetterling CM Process Technology for Silicon Carbide Devices 2002 Stevenage: Institute of Electrical Engineers
Choyke WJ Matsunami H Pensl G Silicon Carbide: Recent Major Advances 2003 Berlin : Springer Verlag
Infineon web page
Cree Inc web page
Godigon P SiC materials and technologies for sensors development Materials Science Forum 2005 483–485 1009 1014
| 0 | PMC1156917 | CC BY | 2021-01-04 16:37:37 | no | Biomed Eng Online. 2005 May 17; 4:33 | utf-8 | Biomed Eng Online | 2,005 | 10.1186/1475-925X-4-33 | oa_comm |
==== Front
Cancer Cell IntCancer Cell International1475-2867BioMed Central London 1475-2867-5-141590451910.1186/1475-2867-5-14Primary ResearchLate stage inhibition of hematogenous melanoma metastasis by cystatin C over-expression Ervin Heather [email protected] James L [email protected] Department of Biochemistry, Kirksville College of Osteopathic Medicine, 800 West Jefferson, Kirksville, Missouri 63501-1497 USA2005 17 5 2005 5 14 14 17 5 2004 17 5 2005 Copyright © 2005 Ervin and Cox; licensee BioMed Central Ltd.2005Ervin and Cox; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Tumor metastasis is a frequent cause of treatment failure for cancer patients. A key feature of metastatic cancer cells is their invasive ability. Cysteine proteases contribute to invasive properties of many cancer cell types. To analyze the contribution of cysteine proteases to metastasis we have over-expressed in B16 melanoma cells the natural cysteine protease inhibitor, cystatin C. We measured in vitro invasion of cystatin over-expression clones with Boyden chamber type assays. Tail-vein injections of cells were used to compare lung tumor colonization. Subcutaneous tumor growth and tumor cell metastasis from primary tumors were also analyzed. Apoptosis of tumor cells was measured in lung tissues following melanoma cell injection.
Results
Results show the in vitro invasion of cystatin C over-expressing cells was dramatically inhibited. Lung tumor colonization was also reduced. Increased tumor cell apoptosis was found to be an important factor and may be related to the reduced tumor burden noted in this system of melanoma metastasis.
Conclusion
Cysteine proteases therefore, may be a target for future anti-metastatic therapies.
==== Body
Background
Tumor metastasis is a major reason for treatment failure in cancer patients. Invasion of host tissues is a hallmark feature of metastasis and requires alterations in tumor cell adhesion, cell migration, and proteolytic degradation of tissue matrices [1,2]. A critical initial step of metastasis is that tumor cells enter blood vessels or the lymphatic system by invasion of host basement membranes. After being dispersed through vascular spaces, metastatic tumor cells lodge in distant capillaries. Late stages of metastasis involve further steps including invasion from vascular spaces into foreign tissues and growth of secondary metastases in permissive environments (seed and soil hypothesis) [2]. Growth of metastases in tissues also requires invasion of host endothelial cells into tumor masses to support further the metabolic needs of tumor growth [3].
Members of the papain cysteine protease family of enzymes (primarily cathepsins B, H, and L) are elevated in cancer cells and contribute to invasion of many tumor types [4,5]. This high cysteine protease activity influences the activity of other proteolytic enzymes during acquisition of an invasive phenotype [6]. Cysteine proteases help create an invasive phenotype at one or more steps in tumor progression to metastasis [7]. Details are lacking in the exact roles of cysteine proteases in tumor progression, but it is clear that highly metastatic variants of various tumor types express an excess of protease activity of the papain cysteine protease type [8]. Because elevated cysteine protease activity correlates well with highly invasive cancers, inhibition of this activity may be a target for anti-metastatic therapies.
Cystatins are potent inhibitors of cysteine proteases that are found in all tissues and most biological fluids. In general, a decrease in cysteine protease inhibitor levels is also found in metastatic tumors, contributing to higher cysteine protease levels [9]. Cystatin C is a type II cysteine protease inhibitor that is normally secreted from cells [10]. Previously we have investigated the role of the natural cysteine protease inhibitor, cystatin C, and found inhibition of lung metastasis [11]. To help establish mechanism of cystatin C action in metastasis we have fused cystatin C to the N-terminus of green fluorescent protein. In this study we show over-expression of cystatin C in melanoma cells is associated with reduced metastasis and increased apoptosis in lung tissues.
Results
Expression of cystatin C-GFP fusion
The expression of cystatin C-GFP fusion was observed in B16F10 melanoma cells following transient transfection of plasmid construct DNA. Confocal microscopy was used in an attempt to localize cystatin C-GFP protein expression in cells. A cytoplasmic expression was noted for GFP alone and cystatin C-GFP fusion when transiently expressed in B16 melanoma cells (Figure 1). This pattern of expression was not anticipated since cystatin C is primarily a secreted protein that should have produced a vesicular localization of fluorescence within the cell. Figure 2 shows expression of the cystatin C-GFP fusion protein. Cystatin C fused to GFP should be about 44 kD (30 kD GFP plus 14 kD cystatin) and has also formed multimers under denaturing conditions. Recently Paraoan et al. have described cytoplasmic distribution of human cystatin C mutated within the signal peptide [16]. Similarly, sequence analysis of RT-PCR product revealed a signal peptide amino acid difference (-4 Ala→Gly) between Balb c (inserted cDNA) and B16 melanoma cystatin C messenger RNA (endogenous) (data not shown). We recloned the cystatin C-GFP fusion and re-transfected the DNA construct and obtained identical results. We believe cystatin C over-expression effects are primarily due to intracellular expression in our B16 melanoma cell model.
Figure 1 GFP fluorescence. B16 melanoma cells transfected with cystatin C-GFP fusion construct A, B) transiently transfected B16F10C) stably transfected B16F10 cells.
Figure 2 Western blot analysis of cystatin C-GFP fusion. Lane 1, human cystatin C, 1 ug; lane 2, c23 fusion clone (40 ug/lane total cell extract). Probed with anti-human cystatin C antibody. Fusion is 14 kD cystatin C plus 30 kD GFP protein ~44 kD. Multimers form with denatured cystatin C.
Cysteine protease activity
The total cellular cysteine protease activity was measured with a sensitive fluorometric assay. This assay indirectly measures protease inhibitor levels in the cell as cystatins are often found at low levels. Table 1 shows decreased cysteine protease activity in cystatin C over-expression clone c23. Most of the cellular cysteine protease activity is localized to lysosomes that would be unaffected by altered cystatin levels.
Table 1 Cellular cysteine protease activity in vitro
Clone Activity* Relative activity %
G1 20 +/- 0.8 100
c23 15.6 +/- 0.7 78
*Arbitrary fluorescence units. Experiment carried out 3 times. Students t-test, p < .05.
In vitro invasion of B16 melanoma is inhibited by cystatin C over-expression
One important characteristic of metastasis is the invasive ability of tumor cells. We used an in vitro Boyden chamber assay to quantify the invasive potential of our clonal lines. [17]. Over-expression of cystatin C-GFP (c23, c28) decreased the invasion of B16F10 melanoma cells through Matrigel coated filters by about 90% relative to control G1 cells (GFP transfected) (Fig 3). Cystatin-GFP fusion clones behave similarly to cystatin over-expressing clones without the GFP fusion. Growth rate is the same as control; adhesion is the same as control; and cell migration is reduced in comparison to control ([12] and personal observations)
Figure 3 Invasion assay. Boyden chamber type assays comparing cell line B16 F10, clone G1 (GFP-expressing), c23 and c28 (cystatin C-GFP fusion) invasion through Matrigel. Experiment was run in triplicate with three filters for each clone per experiment. Data is expressed in cells / filter ± s.d. Student's t test P < .001.
Cystatin C over-expression reduces lung colonization and increases survival in mice
We examined the effect of cystatin C over-expression on lung tumor growth and overall survival of mice injected via tail vein. Tail vein injections are measures of experimental or late stage metastasis only because the initial invasion and intravasation are bypassed. We hypothesized that the large decrease measured for in vitro invasion might lead to similar decreases in lung colonization in mice. Although we measured a decrease in lung tumor burden following tail vein injection in the cystatin C over-expressing clone, this clone still resulted in tumor colonization (Table 2). We next examined the effect of cystatin C over-expression on survival of mice injected with B16F10 melanoma cells and found about a 10% increase in median survival time (Figure 4). Our conclusion from this data is that over-expression of cystatin C, while decreasing tumor burden, causes only modest increase in the survival of animals following tail vein injection of melanoma cells. We show that a large decrease in melanoma invasive ability with cystatin over-expression does not by itself prevent subsequent growth of metastases.
Table 2 Lung tumor colonization by B16 F10 transfected cell lines
Tumors / Lung
Clone* Average. Range
G1 25 ± 12 12–45
C23 14 ± 5 5–23
* n = 6 mice for each clone
Figure 4 Survival curve. Comparison of survival of mice injected via tail-vein with either clone G1 or c23 cells (7 mice per group).
Subcutaneous growth of B16 melanoma with cystatin C over-expression is similar to control
We compared the subcutaneous growth of B16 melanoma control and cystatin C over-expressing clones in syngeneic C57 BL6 mice. Following subcutaneous, scapular injection of B16 melanoma tumor cells, tumors become measurable by calipers after about ten days. Figure 5 shows the relative growth of subcutaneous tumors in mice. A slight decrease in subcutaneous tumor growth was noted for a cystatin C over-expression clone. In a separate experiment we compared the relative lung metastasis of melanoma cells originating from subcutaneous tumors. These secondary lung metastases were quantitated at 14 days following subcutaneous injection of either clone G1 or clone c23. After sacrifice of the mice, tumor burden was estimated by S100 staining of frozen lung tissue cross-sections. We found a significant reduction in lung metastasis from primary subcutaneous tumors upon cystatin C over-expression (Table 3).
Figure 5 Subcutaneous tumor growth. Comparison of tumor volume of clone G1(control) or c23 (cystatin over-expressor). (4 mice per group) Paired Wilcoxon test P < 0.01.
Table 3 Melanoma metastasis to lung from subcutaneous primary tumors
Clone Average area ± s.e.m. (%_total)*
G1 26.0 ± 4.9
C23 4.3 ± 1.4
* Imaged S100 antibody positive staining for 4 tumors with 3 sections per tumor and three fields per section. (P = .005 for Student's t test).
Apoptosis of micrometastases is increased with cystatin C over-expression
We examined the distribution of both GFP labeled cells and cystatin C-GFP labeled cells in lung tissue at 24 hours following tail vein injection. The cystatin C-GFP fusion product that we introduced into B16 F10 melanoma allowed us to follow the distribution of metastatic cells in animals. The melanoma cell density in the lung and melanoma distributions were similar for both GFP labeled and cystatin C-GFP clones (data not shown). This observation suggests delivery of tumor cells for each clone to lung capillaries is equivalent and cannot account for differences in subsequent lung tumor burden. We instead focused on tumor growth in late stage metastasis. An examination of melanoma cell apoptosis at early times after lung tissue seeding of melanoma cells provides insight into tumor outgrowth at later stages. One week after melanoma cell tail-vein injection, we excised lung tissue and examined the levels of apoptotic cells. We used the Dead-End TUNEL Assay to quantitate apoptotic cells in frozen tissue sections. Figure 6 shows typical tissue sections stained for apoptotic cells at one week post-injection. Quantitation of apoptosis showed an average two-fold increase in apoptotic melanoma cells within lung tissue of animals injected with cystatin C over-expressing cells (Table 4). Thus, increased apoptosis of lung metastasized B16 melanoma cells correlates with decreased tumor burden upon cystatin C over-expression.
Figure 6 Apoptosis. Representative lung tissue sections stained for apoptosis with TUNEL reagents. A. Control (G1) cell injected mouse. B. c23 (cystatin C over-expressor) cell injected mouse.
Table 4 Apoptosis in lung tissue sections
Apoptotic cells/field
Mouse Clone G1 Clone c23 G1/C23 ratio
Expt 1 1 36 ± 3 78 ± 9 2.2
2 8.6 ± 1.5 44 ± 7 5.1
3 11.6 ± 1.3 25 ± 2.0 2.1
4 13.7 ± 2.0 26 ± 4.3 1.9
2.8 average
Mouse Clone G1 Clone c23 G1/C23 ratio
Expt 2 1 10.5 ± 1.6 35.8 ± 2.5 3.4
2 8.6 ± 0.8 20 ± 4.3 2.3
3 6.5 ± 0.8 12.2 ± 1.6 1.9
2.5 average
Discussion
Blockade of the metastatic spread of cancer could provide dramatic clinical improvements for the treatment of cancer patients. The therapeutic blockade of metastasis is a formidable challenge due to many factors including redundant invasion mechanisms of metastatic cells, acquisition of chemotherapeutic resistance, and evasion of host immune responses. In spite of these problems, inhibition of metastatic cell invasion could be a possible adjunctive treatment for metastatic disease. We have previously found cystatin C over-expression inhibits both melanoma cell migration and metastasis [12]. In this work, expression of cystatin C fused to GFP also showed a dramatic inhibition of in vitro invasion of B16 melanoma cells. The number of lung metastases following tail-vein injection were decreased, however growth of subcutaneous melanoma tumors over-expressing cystatin C was only slightly inhibited. The expression of cystatin C fused to GFP we obtained exhibited primarily cytosolic localization. Another group has reported cytosolic localization of GFP-labeled cystatin C due to a signal peptide mutation [16]. We have noted an amino acid change near the signal peptide cleavage site and suggest abnormal cellular distribution may result. The point is, our results on melanoma metastasis may only pertain to intracellular over-expression of cystatin C.
We have demonstrated cystatin over-expression in melanoma cells appears to result in an increase in apoptosis in lung micrometastases in vivo. The growth of secondary metastases is considered to be rate limiting in the metastatic cascade of tumor spread [18,19]. Over-expression of the metalloprotease inhibitor TIMP-1 does not block B16 melanoma cell extravasation into lung tissue, but does alter subsequent tumor growth within lung tissue [20]. Our work supports the idea that inhibition of cell invasion alone is not sufficient to prevent lung tissue colonization by metastatic cells [21]. In this work we demonstrate that apoptosis is increased in lung micrometastases when melanoma cell invasion is inhibited by cystatin C over-expression. Increased tumor cell apoptosis may be the result of alterations in the tumor cell microenvironment in vivo, for example, alterations in tumor cell adhesion or local release of growth factors required for tumor cell survival. Inhibition of angiogenesis is probably not involved because after at one-week analysis, virtually all micrometastases were only a few cells in diameter. Generally, angiogenesis begins only as tumor size approaches 1 mm [22].
Additional effects cystatin on melanoma cell metastasis might have been expected if our melanoma cystatin over-expressing clones had secreted cystatin C. Multiple roles extracellular cathepsin B have been suggested, including matrix degradation [23]. A synthetic inhibitor of cysteine proteases has been found to decrease metastasis in a colon cancer model, however the point of metastatic block was not defined [24]. Recently, Konduri et al. [25] have shown over-expression of cystatin C in glioblastoma cells blocks tumor cell invasion and tumor growth in vivo, however in their system cystatin C was secreted.
Conclusion
Metastatic melanoma is a rapidly spreading and growing cancer type that thwarts virtually all attempts at growth arrest. Blockade of metastasis would improve current treatments of melanoma that are primarily targeted at tumor cell proliferation. We have found that cystatin C over-expression dramatically reduces melanoma cell invasion. Late stage metastasis is influenced for cystatin C over-expressing cells by an apparent increased apoptosis in the target tissue (in this case, lung). Our results indicate that inhibition of cell invasion alone is insufficient to dramatically improve the survival of mice at late stage melanoma metastasis. More favorable results may occur at earlier steps of metastasis intervention with anti-invasive agents or with combined therapies that target tumor growth and /or angiogenesis.
Methods
Cell culture
The B16-F10 melanoma cell line was kindly provided by Dr. I. Fidler (MD Anderson Cancer Center, Houston, TX). Cells were grown in culture on plastic dishes as monolayers in RPMI-1640 media containing 10% fetal bovine serum (v/v) and supplemented with antibiotics (100 units penicillin, 0.1 mg streptomycin, 0.25 μg amphotericin B per milliliter) (Sigma). Cells were cultured in a 5% CO2 humidified atmosphere at 37°C until near confluence.
DNA constructs and cell transfections
Murine cystatin C cDNA was described in a previous paper [12]. Fusion of cystatin C cDNA to the N-terminus of green fluorescent protein (GFP) (Invitrogen) was carried out with a fusion fragment comprised of two annealed oligos (5'-CTGCAAAAATGCCA-3', 5'-CGTGGCATTTTTGCAG3') (Integrated DNA Technologies, Corallville, Iowa). Ligated DNA was transformed into DH5a E. coli cells by electroporation with a Biorad Gene Pulser following the manufacturer's instructions. Fusion clones were confirmed by restriction analysis and lysis-purified DNA was used for transfection of B16 melanoma cells. Transfection of plasmid DNA into B16 F10 melanoma cells was carried out by the calcium phosphate method [13]. Transiently transfected cells were imaged by confocal microscopy one or two days after transfection. Stable transformants were selected in media containing G418 (geneticin)(Sigma) at 1 mg/ml. Three clonal lines were chosen for study after three weeks selection: a GFP-control clone (designated G1) and two cystatin over-expressing clones (designated c23 and c28).
Western blot analysis
B16 melanoma F10 clone c23 was grown to near confluence and lysed with buffer (0.1% Triton X-100 in PBS with 1% protease inhibitors (Sigma)) through three freeze thaw cycles and centrifuged at 10 krpm for 5 minutes to pellet cell debris. Cell extracts (40 ug) were run on 10% SDS PAGE gels and electro-blotted to PVDF membranes for antibody detection. Human cystatin C (1 ug) (Calbiochem) was run as a control. Rabbit anti-human cystatin C (Upstate Biotechnology) was used at 1:1000 dilution. Secondary antibody, goat anti-rabbit HRP (Upstate Biotechnology) was used at 1:1000 and detection was by ECL luminescence system. (Amersham). Protein was determined by the Bradford assay.
Assay of cysteine protease activity
We used a fluorometric assay to measure total cellular cysteine protease activity [14]. This assay indirectly measures cysteine protease inhibitor levels that are difficult to measure directly due to low levels. Either G1 (GFP) or c23 (cystatin C-GFP fusion) cells were seeded at 5 × 104 cells per well in a 24 well cell culture plate. After 16 hours growth the cells were incubated with PAB buffer (Hank's buffered saline plus 0.6 mM CaCl2, 0.6 mM MgCl2, 2 mM cysteine, and 25 mM PIPES pH7.0) for 30 minutes at 37°C. The buffer was then replaced with 0.2 ml of the same buffer containing 0.1% Triton X-100 and 100 uM Z-Phe-Arg AMC (Enzyme Systems Products, Livermore, CA). After 20 minutes further incubation at 37°C, fluorescence was read in a BioTek flx800 fluorescent plate reader at 360 excitation/460 emission and the results were expressed in relative fluorescence units. The assay was found to be linear for 30 minutes under the conditions used.
Cell invasion
Sub-confluent B16 F10 cells or permanently transfected B16 melanoma clones were detached with trypsin/EDTA solution (0.25%, 5 minutes) (Sigma). Following inactivation of trypsin by addition of an equal volume of serum-containing media, the cells were counted with a hemocytometer. Cells (2 × 104) were placed into top wells of Boyden chambers (BD Biosciences), the filters (8 um pore size, Osmonics, Inc.) pre-coated with Matrigel (65 ug/filter) [12]. Cell invasion was allowed to occur for 24 hours, after which time cells were fixed and stained [methanol, 10 minutes; 0.1% Triton X-100, 2 minutes; Harris hematoxylin (Sigma), 10 minutes]. After 24 hours, cells on the bottom of the filter were counted with an inverted Olympus IMT-2 microscope at 400× magnification. Cell numbers for 18 fields were summed for each filter.
Survival curve
An experiment was conducted where mice (seven mice per group) were tail vein injected (1 × 105 cells per mouse) with either control GFP (G1) or cystatin C-GFP fusion clone (c23). The mice were followed until time of death or were moribund in which case mice were sacrificed.
Lung metastasis-assay
Clones G1 and c23 were grown to near confluence and detached with trypsin solution. Trypsin was neutralized with an equal volume of cell culture media and washed twice with phosphate buffered saline (PBS). Cells were re-suspended in PBS and 1 × 105 were injected into tail veins of C57BL6 female mice (seven per group). After 30 days mice were sacrificed and lung tissues were excised. Lung surface tumors were counted under a low-power dissecting microscope.
Subcutaneous tumor growth
Cells of clones to be tested (G1 and c23) were grown to near confluence, detached with brief trypsin treatment, and washed twice with PBS. After re-suspension in PBS, the cells (3 × 105) were injected subcutaneously into scapular regions of C57BL6 mice (six mice per group). When tumors began to appear (after about ten days), tumor diameter was measured on successive days in two orthogonal directions with calipers. Tumor volumes were calculated by the formula V = (a2 × b)/2, were a = width and b = length, in millimeters [15].
Lung metastasis assay from subcutaneous tumors
C57BL6 mice (6 mice per group) were injected subcutaneously with either clone G1 or clone c23 cells at 2 × 105 cells per mouse. After 3 weeks mice were sacrificed and lungs were removed and frozen in OCT media for sectioning. Frozen tissue sections (10 uM) from each lung tissue sample were stained with S100 antibody (DAKO, 1:1000 dilution) and immunologically detected with a secondary horseradish peroxidase antibody (Sigma, 1:1000). Microscopic images (100×) of stained tissue sections were collected from 4 mice in each group. Quantitation of stained area per section was carried out with Image J (NIH software).
Apoptosis of lung tumor cells
Mice (C57BL6, 3–4 animals per group) were injected with B16 F10 clones G1 (GFP), or c23 (cystatin C-GFP) at 1 × 105 cells via tail vein. After one week mice were sacrificed and lung tissue was removed. Lung tissues were stored in OCT embedding compound (Miles Inc.) overnight at 4°C and then frozen at -80°C. Frozen tissue sections (10 um) were prepared and stained for TUNEL reactivity with the Dead-End TUNEL kit (Promega) directly on glass coverslips. Manufacturers instructions were followed with side-by-side staining of G1 and c23. Stained cells were summed for at least five 200× microscopic fields per tissue section and six independent sections per animal. The average number of apoptotic cells per high powered field were recorded ± s.d.
List of abbreviations
GFP, Green fluorescent protein; AMC, 7-amino-4-methylcoumarin; PBS, Phosphate buffered saline; ECL, enhanced chemical luminescence.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
HE carried out the apoptosis assays, western blot analysis, and cell maintenance. JC conducted the remaining procedures and drafted the manuscript.
Acknowledgements
The authors thank Tom Moninger of the University of Iowa for assistance with the confocal imaging.
==== Refs
Engers R Gabbert HE Mechanisms of tumor metastasis: cell biological aspects and clinical implications J Cancer Res Clin Oncol 2000 126 682 92 11153140 10.1007/s004320000148
Fidler IJ The pathogenesis of cancer metastasis: the 'seed and soil' hypothesis revisited Nat Rev Cancer 2003 3 453 8 12778135 10.1038/nrc1098
Rosen LS Angiogenesis inhibition in solid tumors Cancer J 2001 7 S120 8 11779082
Friedrich B Jung K Lein M Turk I Rudolph B Hampel G Schnorr D Loening SA Cathepsins B, H, L and cysteine protease inhibitors in malignant prostate cell lines, primary cultured prostatic cells and prostatic tissue Eur J Cancer 1999 35 138 44 10211102 10.1016/S0959-8049(98)00273-1
Bervar A Zajc I Sever N Katunuma N Sloane BF Lah TT Invasiveness of transformed human breast epithelial cell lines is related to cathepsin B and inhibited by cysteine proteinase inhibitors Biol Chem 2003 384 447 55 12715895 10.1515/BC.2003.050
DeClerck YA Imren S Montgomery AM Mueller BM Reisfeld RA Laug WE Proteases and protease inhibitors in tumor progression Adv Exp Med Biol 1997 425 89 97 9433492
Podgorski I Sloane BF Cathepsin B and its role(s) in cancer progression Biochem Soc Symp 2003 263 76 14587299
Hazen LG Bleeker FE Lauritzen B Bahns S Song J Jonker A Van Diel BE Lyon H Hansen U Kohler A Van Noorden CJ Comparative localization of cathepsin B protein and activity in colorectal cancer J Histochem Cytochem 2000 48 1421 30 10990495
Chambers AF Colella R Denhardt DT Wilson SM Increased expression of cathepsins L and B and decreased activity of their inhibitors in metastatic, ras-transformed NIH 3T3 cells Mol Carcinog 1992 5 238 45 1586450
Turk B Turk V Turk D Structural and functional aspects of papain-like cysteine proteinases and their protein inhibitors Biol Chem 1997 378 141 50 9165064
Cox JL Sexton PS Green TJ Darmani NA Inhibition of B16 melanoma metastasis by overexpression of the cysteine proteinase inhibitor cystatin C Melanoma Res 1999 9 369 74 10504055
Sexton PS Cox JL Inhibition of motility and invasion of B16 melanoma by the overexpression of cystatin C Melanoma Res 1997 7 97 101 9167174
Sambrook J Fritsch EF Maniatis T Molecular Cloning: A Laboratory Manual 1989 Cold Spring Harbor, NY: Cold Spring Harbor Press
Anastasi A Brown MA Kembhavi AA Nicklin MJ Sayers CA Sunter DC Barrett AJ Cystatin, a protein inhibitor of cysteine proteinases. Improved purification from egg white, characterization, and detection in chicken serum Biochem J 1983 211 129 38 6409085
Montgomery AM Mueller BM Reisfeld RA Taylor SM DeClerck YA Effect of tissue inhibitor of the matrix metalloproteinases-2 expression on the growth and spontaneous metastasis of a human melanoma cell line Cancer Res 1994 54 5467 73 7923181
Paraoan L Ratnayaka A Spiller DG Hiscott P White MR Grierson I Unexpected intracellular localization of the AMD-associated cystatin C variant Traffic 2004 5 884 95 15479453 10.1111/j.1600-0854.2004.00230.x
Albini A A rapid in vitro assay for quantitating the invasive potential of tumor cells Cancer research 1987 47 3239 3245 2438036
Chambers AF Schmidt EE MacDonald IC Morris VL Groom AC Early steps in hematogenous metastasis of B16F1 melanoma cells in chick embryos studied by high-resolution intravital videomicroscopy J Natl Cancer Inst 1992 84 797 803 1573668
Morris VL Tuck AB Wilson SM Percy D Chambers AF Tumor progression and metastasis in murine D2 hyperplastic alveolar nodule mammary tumor cell lines Clin Exp Metastasis 1993 11 103 12 8422701 10.1007/BF00880071
Koop S Khokha R Schmidt EE MacDonald IC Morris UL Chambers AF Groom AC Overexpression of metalloproteinase inhibitor in B16F10 cells does not affect extravasation but reduces tumor growth Cancer Res 1994 54 4791 7 8062280
Van Noorden CJ Jonges TG Meade-Tollin LC Smith RE Koehler A In vivo inhibition of cysteine proteinases delays the onset of growth of human pancreatic cancer explants Br J Cancer 2000 82 931 6 10732768 10.1054/bjoc.1999.1021
Folkman J Role of angiogenesis in tumor growth and metastasis Semin Oncol 2002 29 15 8 12516034 10.1053/sonc.2002.37263
Erdel M Trefz G Spiess E Habermaas S Spring H Lah T Ebert W Localization of cathepsin B in two human lung cancer cell lines J Histochem Cytochem 1990 38 1313 21 2201737
Van Noorden CJ Jonges TG Van Marle J Bissell ER Griffini P Jans M Snel J Smith RE Heterogeneous suppression of experimentally induced colon cancer metastasis in rat liver lobes by inhibition of extracellular cathepsin B Clin Exp Metastasis 1998 16 159 67 9514097 10.1023/A:1006524321335
Konduri SD Yanamandra N Siddique K Joseph A Dinh DH Olivero WC Gujrate M Kouraklis G Swaroop A Kyritsis AP Rao JS Modulation of cystatin C expression impairs the invasive and tumorigenic potential of human glioblastoma cells Oncogene 2002 21 8705 12 12483523 10.1038/sj.onc.1205949
| 15904519 | PMC1156918 | CC BY | 2021-01-04 16:24:16 | no | Cancer Cell Int. 2005 May 17; 5:14 | utf-8 | Cancer Cell Int | 2,005 | 10.1186/1475-2867-5-14 | oa_comm |
==== Front
Cancer Cell IntCancer Cell International1475-2867BioMed Central London 1475-2867-5-161591890810.1186/1475-2867-5-16Primary ResearchProton-Induced X-ray Emission (PIXE) Analysis and DNA-chain Break study in rat hepatocarcinogenesis: A possible chemopreventive role by combined supplementation of vanadium and beta-carotene Chattopadhyay Mitali Basu [email protected] Sutapa [email protected] Indira [email protected] V [email protected] Manika [email protected] NB [email protected] Malay [email protected] Division of Biochemistry, Department of Pharmaceutical Technology Jadavpur University, Kolkata 700 032, India2 Institute of Physics, Sachivalaya Marg, Bhubaneswar – 751 005, Orissa, India3 All Indian Institute of Hygiene and Public Health, Kolkata 700 072, India2005 26 5 2005 5 16 16 7 8 2003 26 5 2005 Copyright © 2005 Chattopadhyay et al; licensee BioMed Central Ltd.2005Chattopadhyay et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Combined effect of vanadium and beta-carotene on rat liver DNA-chain break and Proton induced X-ray emission (PIXE) analysis was studied during a necrogenic dose (200 mg/kg of body weight) of Diethyl Nitrosamine (DENA) induced rat liver carcinogenesis. Morphological and histopathological changes were observed as an end point biomarker. Supplementation of vanadium (0.5 ppm ad libitum) in drinking water and beta-carotene in the basal diet (120 mg/Kg of body weight) were performed four weeks before DENA treatment and continued till the end of the experiment (16 weeks). PIXE analysis revealed the restoration of near normal value of zinc, copper, and iron, which were substantially altered when compared to carcinogen treated groups. Supplementation of both vanadium and beta-carotene four weeks before DENA injection was found to offer significant (64.73%, P < 0.001) protection against generation of single-strand breaks when compared with the carcinogen control counter parts. A significant stabilization of hepatic architecture of the cells was observed as compared to carcinogen control in vanadium plus beta-carotene treated group. This study thus suggests that vanadium, a prooxidant but potential therapeutic agent yield safe and effective pharmacological formulation with beta-carotene, an antioxidant, in the inhibition of experimental rat hepatocarcinogenesis.
DENAvanadiumPIXEbeta-caroteneDNA-chain break
==== Body
Background
Trace element studies on cancer and normal tissues have interested groups of researchers for many years [1]. Accumulating evidence demonstrated that many essential/trace elements play an important role in a number of biological processes by activating or inhibiting enzymatic reactions, by competing the permeability of cell membranes, maintaining genomic stability and/or by other mechanisms. It is therefore, reasonable to assume that this element would exert action, directly or indirectly on the carcinogenic process [1,2]. The technique of PIXE (Proton induced X-ray emission), since its introduction has been successfully used by various groups throughout the world for trace elemental analysis of biomedical samples due to high sensitivity, multielemental capability and its non-destructive nature [3].
Vanadium (V), which is an endogenous constituent of all or most mammalian tissues, is believed to have a regulatory role in biological system [4]. Previous studies from our laboratory [5] and other laboratories [6,7] have established V as a significant biological regulator in assessing the physiological and biochemical state of the animals in a dose related manner [6]. It has been shown that V at 0.5 ppm in drinking water was effective in inhibiting the chemical induced two stage rat hepatocarcinogenesis without any toxic manifestation [8]. The data from several animal experiments established the role for beta-carotene (BC) as an antioxidant and even as an anticancer agent. [9]. BC has been found to be prophylactic for UV and carcinogen-induced cancers of the skin, mammary gland, buccal pouch epithelia and colon in animal models [10,11]. This current work is directed to elucidate the efficacy of both the trace element V and the antioxidant BC when given together can effectively prevent the transformation of neoplastic cells of liver treated with a single necrogenic dose of diethylnitrosamine (DENA).
The use of single injection of DENA allows observation of the chemopreventive potential of V and BC in combination in the initiation phase of hepatocarcinogenesis by the DNA chain break analysis. DNA strand can be detected with great sensitivity by methods that utilize observation of the role of the unwinding of the two DNA strand [12]. In this study, we have used a technique of fluorimetric analysis of DNA unwinding (FADU) using a fluorescent dye for monitoring DNA unwinding, according to our method of Sarkar et al., 1997 [13]. In addition, a 16-weeks study was further conducted first, to demonstrate the protective response of combined supplementation of V and BC in modulating morphological, histopathological changes and second to examine the role of micronutrient concentrations in the carcinogenic process.
Information on the combined effect of vanadium and beta-carotene and their antineoplastic potential is still meager. In this communication, we report for the first time that the trace element, vanadium in combination with antioxidant, beta-carotene can modulate the activity/or status/ or concentration of other elements at the particular dose and tend to stabilizes the genome to some extent in chemical carcinogenesis.
Materials and methods
Animals and diet
Male Sprague Dawley rats weighing 80–100 g at the beginning were obtained from the Indian Institute of Chemical Biology (CSIR), Kolkata 32, India and maintained on standard basal diet. We followed the recommendations of the NIH guide for the Care and Use of Laboratory Animals for the maintenance, treatment and sacrifice of the animals used in this study.
Experimental design
Two separate sets of rats were used for the study one for PIXE analysis and histopathological study (Set-I) another for the DNA-Strand assay (Set-II). In Set-I, the rats were divided into eight (8) groups (Fig. 1) with ten rats in each. In group B, C, D and E, hepatocarcinogenesis was initiated by a single intraperitoneal (i.p) injection of DENA at a dose of 200 mg/Kg body weight in 0.9% sodium chloride solution [8,15] at week four. Normal vehicle control rats (Group A) received only sodium chloride solution till the end of the experiment (16 weeks). V (as ammonium monovanadate) was supplemented at a dose of 0.5 ppm in drinking water ad libitum [8] and BC was given in the basal diet at a regimen of 120 mg/Kg of body weight [13] six days a week for the entire period of study. Group B rats were the DENA controls, group C were supplemented with V only, group D rats were treated with BC only and group E were treated with both V and BC. Group c, d and e served as the respective V, BC and V plus BC control groups. For DNA-strand breaks analysis (Set-II) rats were supplemented at the same dose regimen of PIXE analysis four weeks prior to DENA injection. All the rats were killed 18–20 h after DENA injection, livers are promptly excised and DNA was isolated.
Figure 1 The Basic Experimental Protocol.
Experimental set-up for PIXE analysis
About 5 ml of whole blood samples from different control and experimental groups were collected and lyophilized at the Regional Medical Research Center (RMRC), Bhubaneswar. Then samples were converted into fine powder, homogenized and finally pressed into pellets. Conversion of the sample into fine powder before pelletising is necessary to avoid particle size effects in the analysis. These effects are particularly important for the lighter elements. To obtain consistent pellets which remain stable during bombardment, an equal amount of binder such as graphite are added to the powdered sample prior to pelletisation so that bombardment in vacuum does not lead to charging of specimen [14]. The targets are normally thin layers of the sample deposited uniformly over Mylar films usually of 3 μm to 6 μm thickness. Yttrium is added as internal standard. During each run a set of IAEA (international atomic energy agency) animal blood reference standard A13, are prepared and analyzed along with other biomedical samples under similar experimental conditions.
Morphology of liver
For histological studies, representative longitudinally cut blocks were taken from the left, right-median and right-anterior lobes and the samples were fixed immediately in 10% buffered formalin to prevent deformation. Five contiguous sections were made from each liver slice for routine histopathological evaluation by hematoxylin and eosin (H & E) staining. Specific hepatocellular lesions observed in H & E staining were recognized by light microscopy (Adcon, Model No. 5591) according to published criteria [15].
DNA Strand break assay
DNA was isolated from the rat liver by a modification of the published criteria [16]. After isolation of DNA from each experimental and control group, DNA was assayed for extent of DNA unwinding by the FADU technique [13]. Data were analysed statistically or differences between the means using Student's t-test.
Statistical analysis
Significance of difference between mean values of different groups of rats was assessed by Student's t-test values of P < 0.05 were taken to imply statistical significance.
Result
Effect of Vanadium and Beta-carotene on essential/Trace elements
Thick target PIXE analysis was performed using GUPIX-95 software, which provides non-linear least squares fitting of the spectrum, together with subsequent conversion of the X-ray peak intensities (not shown) into elemental concentrations via a defined standardization technique involving fundamental parameters and user defined instrument constant. Full account was taken matrix effects and secondary fluorescence contributions in both the spectrum fitting portion and the calculation of concentrations.
Mean concentrations of 8-essential/trace elements were measured by PIXE analysis of blood samples of different groups are shown in Table 1. Results from this study show that the content of elements calcium (Ca), copper (Cu) and zinc (Zn) were lower in the DENA control Group B than the certified values (Table I). On the other hand, the element potassium (K), iron (Fe), manganese (Mn), selenium (Se) showed elevated concentrations in Group B (DENA control). In contrast, elemental analysis of blood samples in group C, D and E treatment with V or BC or both in combination exhibited marked closeness to normal value (Table 1). Data for the control groups of animals (group c, d and e) did not show any marked differences from the normal vehicle control (Group A) counterparts and hence were not included in the study.
Table 1 The Elemental concentrations in μg/ml of five different groups. The Values given in column 2 are the certified concentrations (reference standard A13) of various elements.
Elements Detected (in μg/ml) Certified Value A (Normal) B (DENA Control) C (DENA+V) D (DENA+BC) E (DENA+V+BC)
K 2500 ± 350 3714 ± 0.03a 5440.6 ± 0.02b 3810.5 ± 0.04c 3805.0 ± 0.04c 3782.5 ± 0.01d
Ca 286 ± 53 269.7 ± 0.07 119.1 ± 0.03b 242.5 ± 0.02c 247.6 ± 0.02c 261.9 ± 0.022d
Mn 1.4 ± 1 1.6 ± 0.06 2.0 ± 0.05b 1.5 ± 0.023c 1.8 ± 0.01c 1.6 ± 0.01d
Fe 356 ± 150 370.5 ± 0.04 577.5 ± 0.04b 392.7 ± 0.03c 387.0 ± 0.06c 379.8 ± 0.02d
Cu 2.3 ± 0.5 1.5 ± 0.06 0.89 ± 0.05b 0.98 ± 0.06c 0.96 ± 0.02c 1.2 ± 0.01d
Zn 13.0 ± 10 18.9 ± 0.01 9.0 ± 0.06e 17.9 ± 0.02f 17.3 ± 0.04f 18.4 ± 0.04g
Se 0.24 ± 0.5 0.18 ± 0.04 0.19 ± 0.03b 0.10 ± 0.07c 0.16 ± 0.05c 0.18 ± 0.07d
Pb 0.18 ± 0.01 0.21 ± 0.03 0.25 ± 0.04m 0.22 ± 0.02n 0.24 ± 0.03n 0.22 ± 0.03l
aValues are mean ± S.E. of five different run using different sets of targets.
bp < 0.01 as compared with group A,
cp < 0.01 as compared with group B,
dp < 0.001 as compared with group B.
ep < 0.005 as compared with group A,
fp < 0.01 as compared with group B,
gp < 0.005 as compared with group B.
mp < 0.001 as compared with group A,
np < 0.05 as compared with group B,
lp < 0.005 as compared with group B.
Effect of vanadium and beta-carotene on hepatic Histopathology
Histopathological examination in liver sections from normal untreated group, i.e. Group 'A' (Figure 2) as well as other control groups, i.e. Groups 'c', 'd' (Figures not shown) and 'e' (Figure 5) which received V or BC alone or both in combination respectively revealed normal liver parenchymal cells with granulated cytoplasm and small uniform nuclei radially arranged around the central vein. On the contrary, gross structural alterations were seen in Group B rats (DENA control) with predominantly basophilic, eosinophilic and clear cell foci (Figure 3). Extensive vacuolation was observed in the cytoplasm encircling the nucleus with masses of acidophilic materials. Some nuclei in the cells were large and hyperchromatic with prominent and centrally located nucleoli (Figure 3). Phenotypically altered hepatocyte populations in the form of altered liver cell foci and nodules in varying extent were noticeable throughout the hepatic parenchyma of all groups treated with DENA. The cellular architecture of hepatic lobules in Group E rats (Figure 4) those received both V and BC in combination for 16 weeks seemed to be almost comparable to normal. This elicited the maximum protection of both V and BC (Group E) for the entire period of study against DENA-induced hepatocarcinogenesis, which was reflected in the almost normal hepatocellular architecture. Liver cells from this group were found to contain compact cytoplasmic material with only clear cell foci (Figure 4). The nucleocytoplasmic ratio was decreased considerably as compared to Group B (DENA control). However, a moderate improvement in hepatic cellular architecture was observed in Groups C (Figure not shown) and D (Figure not shown) rats receiving V and BC alone respectively. A considerable vacuolation was still observed in the cytoplasm and the compactness of the hepatocytes was not like normal untreated control group (Group A). The liver sections from these groups presented a predominance of clear cell foci rather than eosinophilic and/or basophilic foci.
Figure 2 Thin section of the normal untreated rat liver (Group A) showing normal hepatocellular architecture [H&E × 400].
Figure 3 Section of the rat liver (Group B) showing abnormal architecture after initiation with DENA (200 mg/ Kg b. wt.) for 16 weeks [H&E × 400].
Figure 4 Section of the DENA treated rat liver (Group E) showing almost normal architecture after continuous supplementation of vanadium (0.5 ppm) and beta-carotene (120 mg/kg of basal diet) for 16 weeks. [H&E × 400]. cv, central vein;
Figure 5 Thin section of the rat liver of vanadium and beta-carotene control (Group e) showing normal hepatocellular architecture, CV, Central Vein [H&E × 400].
Effects of Vanadium and Beta-carotene on DEN-induced hepatic DNA chain break
A significant rise in total percentage of hepatic DNA single-strand break could be observed in group B rats 20 h following DENA injection (Fig. 6) when compared with normal control, i.e., Group A. The percentage of native double-stranded DNA of group B (DENA control) rats was found to be three-fold less (P < 0.001) whereas the total aberrant single-strand regions were found to be more than ten fold (P < 0.001) higher than that of normal vehicle control (Group A) [Fig. 6]. This is actually shows the direct DNA damaging efficacy of the hepatocarcinogen DENA. Treatment with V or BC or both in combination in group C, D and E rats respectively strictly abated the generation of single-stranded DNA 20 h following DEN injection. Result shows that maximum protective effect against generation of single-stranded region was found in Group E rats that received both V and BC. Moreover, the native double-stranded DNA in group E rats was almost two fold higher than in DENA control rats (Group B) [Fig. 6]. Table 2 shows the number of single-strand breaks/ DNA 20 h following DENA injection in the presence of V and BC either alone or in combination. In group B, a significant (P < 0.01) increase in the number of single-strand breaks/DNA could be observed. Treatment with V or BC alone showed decrement at the level of 47.23% and 52.87% respectively (P < 0.01) in the generation of single-strand DNA over and above DENA control but the maximum protective effect (64.73%, P < 0.005) was found in Group E that received both V and BC in combination.
Figure 6 Inhibitory effect of vanadium and beta-carotene (in combination) on the generation of DNA-strand breaks initiate with DENA. *P < 0.01 as compared to Group A *fP < 0.01 as compare with Group B. **fP < 0.005 as compared with Group B.
Table 2 Effect of vanadium and beta-Carotene on the number of single-strand breaks/DNA fragment in rat liver 18–20 h after a single injection of DEN
Group Treatment Percentage of Double-strand DNA Number of single-strand breaks/DNA fragment Inhibition (%)
A Normal 93.56 0.07 ± 0.02a
B DENA Control 31.68 1.32 ± 0.01 b
C DENA+V 47.23 0.58 ± 0.01 c 56.06
D DENA+BC 52.87 0.54 ± 0.04 c 59.09
E DENA+V+BC 64.73 0.48 ± 0.023 d 63.6
The mean values of the number of the single-strand breaks/DNA fragment for Vanadium control (group c) was 0.07 ± 0.01; for β-carotene control (group d), 0.08 ± 0.01 and the Vanadium and beta-carotene control (group e), 0.10 ± 0.01. No statistical significance could be observed between these drug controls when compared with normal vehicle control (Group A).
aValues are means ± S.E of 5 rats (each rat liver genomic was tested in triplicate)
bp < 0.01 as compared with group A,
cp < 0.01 as compared with group B,
dp < 0.005 as compared with group B.
Discussions
The V plus BC supplemented group E animals registered a 64.73% inhibition in the number of single strand-breaks per DNA (Fig. 6). Such strand breaks are generally assumed to be lethal to cells. It is well known that DNA strand breaks responsible for chromosomal alterations are generated from DNA base lesions induced by most chemical mutagens. The substantial decrement of the single strand breaks can explain one possible mechanism of the anticlastogenic potential of V and BC. One hypothesis is that the in vivo protective effect of V and BC when give together as observed here, may be due to excision repair activity. DNA double-strand breaks (DDBs) are generated from mutagen induced DNA lesions in the S phase of the cell cycle and repaired by post replication repair in the G2 phase and that unrepaired DDBs result in breakage type chromosomal aberrations [17]. In this context, the suppression of break type aberrations by the combined action of V and BC may be due to a modification of the capability of the post-replication repair of DDBs.
Several reports suggest that antioxidants suppress the clastogenic action of tumor promoters and carcinogens [18]. The inhibitory effect of different antioxidants like vitamin E, beta-carotene and selenium in the production of SCEs in cultured mammalian cells by stimulated human phagocytes, which occurs by a mechanism involving oxygen metabolites, has been successfully demonstrated by Weitberg et al., 1985 [19]. This is in line with the findings of Sarkar et al [13] that confirm the anticlastogenic effect of beta-carotene. Based upon the present study and those reported previously from our laboratory [20] as well as from other laboratories [21] it has been confirmed that V at low concentration is not clastogenic in experimental animals. Information on the combined effect of a trace element (V) and an antioxidant (BC) in alleviating chemical carcinogenesis in animal model is meager, particularly at molecular level. However, a recent study from our laboratory [22] demonstrate an inhibitory role of concomitant treatment with micronutrient vanadium and VD3 on hepatic chromosomal aberrations as well as DNA strand break during rat liver carcinogenesis induced by DENA. In this study we have shown that the occurrence of DDBs are greatly reduced by exploring the inhibitory effect of two antineoplastic agents. V and BC thus may be effective in offering protection by possible mechanism by preventing the breaks at DNA by acting on the antiproliferative and differentiation inducing genes.
Histology was taken as end point biomarkers. Histological observations clearly show that DENA in carcinogen control group animals clearly damages the normal architecture of hepatic tissue. A typical vacuolation was observed in the cytoplasm encircling the nucleus with masses of acidophilic materials. Some nuclei in the cells were large and hyperchromatic with prominent and centrally located nucleoli (Figure 3). This condition is known as preneoplastic condition. V plus BC supplementation seemed to diminish these changes (Figure 4) and brought back to normal tissue architecture. Mild preneoplastic condition was observed proving that V plus BC synergistically offers chemoprevention and protects against carcinogen-induced damage.
The distribution of trace element levels in various organs indicates a significant differentiation between normal and malignant tissue, which is in good agreement with several previous findings [23]. It is well that malignant cells differ biochemically in many ways from normal cells. Oberley et al., 1984 [24] reported that cancer cells have an altered pattern of super oxide dismutase (SOD) activity from that seen in normal cells. They added that cancer cells usually have lowered levels of Cu or Zn-SOD when compared with an appropriate control. Within cells SOD (Mn, Cu, Zn), glutathione peroxidase (Se) and catalase (Fe) remove 02- and H2O2 before they approach available promoters of Fenton chemistry [25]. H2O2 is reduced by glutathione (GSH) [50]. Despite these protective enzymes, some 02- and H2O2 may escape and in the presence of decompartmentalized Fe, are catalyzed to more reaction reactive oxygen species (ROS) [25]. This condition is aggravated during carcinogenesis as SOD and GSH concentration reduced during carcinogenic process [26]. In the present study, raised level of Cu or Zn in blood samples of group C, D and especially in group E (which received V, BC and V plus BC respectively) could be explained by the ability of liver cells to form SOD-model compounds in vivo. These complexes scavenge super oxide radicals and account for the antitumor activity of vanadium and beta-carotene in rat hepatic neoplasia presumably by stabilizing the genome.
Several studies have shown augmentation of tumor growth in rodent models with Fe supplementation [27,28]. In vitro and in vivo data suggest that Fe may be capable of mutagenic effects mediated through free radical generation or tumor promotion through nutritional mechanisms [29]. As a transition metal Fe is possessed of loosely bound electrons that are capable of participating in lipid peroxidation (LPO) reactions. Such reactions are thought to lead to DNA damage and ultimately neoplasia [30]. This is in accordance with observations made in the present study. In this regard, marked lowering in Fe concentration by combined supplementation of V and BC to DEN-induced animals in Group E (Table 1) could be explained by the suggested mechanism(s). Treatment with V or BC alone or in combination showed decrease in the level of K, Mn, Se concentrations when compare with carcinogenesis control group (group B). At present it is doubtful that minor changes in mineral levels induced by V and BC combination played a major role in tumorigenesis in this experiment, but the possibility cannot be excluded.
Conclusion
These studies suggest that trace element concentration in vivo could be used as an early biomarker for carcinogenic process. The results documented indicate that vanadium plus beta-carotene at a given dose can modulate other trace element concentration in serum by some unknown mechanism that is yet to be elucidated and in so doing preserve the genetic stability. The study remains noteworthy in the aspect that it records V and BC both-mediated suppression of DNA-strand break at the initiation stage in addition to essential / trace elemental concentrations alteration(s) in the liver and in the maintenance of the normal hepatic architecture during the preneoplastic stage of liver carcinogenesis.
Authors' contributions
Author 1 MBC carried out the rat hepatocarcinogenesis, morphology, morphometry and histopathological studies. Author 1 MBC and author 3 IK made DNA-chain break studies. Proton induced X-ray emission analysis and assessment was made by author 2 SM and author 4 VV and author 5 MD. Author 6 NBK, a senior medical specialist has done the expert consultant for the pathological aspects of the work. The planning, design and coordination of the experimental work was made by author 7 MC, the principal investigator.
Acknowledgements
M. B Chattopadhyay is indebted to Indian Council of Medical Research, Govt. of India (Grant No. 45/2/2001/BMS/TRM) for the financial support to carry out this work.. We are also grateful to Dr. P.K. Sen, Senior Reader, Department of Mathematics, Jadavpur University, Calcutta, India for his excellent guidance in statistical interpretation of the data. The authors are grateful to Dr. Amitabha Chatterjee, Calcutta University College of Medicine, Calcutta for his help in interpretation of the histopathological study. The authors thanks Dr. Ranjan Basak and Dr. Barun Kanti Saha of Jadavpur University for technical advise and support.
==== Refs
Schwartz MK Introduction to session on biological markers in diagnosis and therapy Cancer 1975 35 2325 2326 1212648
Sky-Peck HH Trace metals and neoplasia Clin Physiol Biochem 1986 4 99 111 3514058
Tong YP Zang YPYL Determination of eight elements in six human cancer cell lines and two humans normal cell lines by proton-induced x-ray emission Biological Trace Elemental Research 1991 29 31 44
Crans DC Bunch RJ Theisen LA Interaction of trace levels of vanadium (IV) and vanadium (V) in biological systems J Am Chem Soc 1989 11 7597 7607 10.1021/ja00201a049
Bishayee A Chatterjee M Inhibition of altered liver cell foci and persistent nodule growth by vanadium during diethylnitrosamine-induced hepatocarcinogenesis in rats Anticancer Res 1995 15 455 462 7763021
Djordjevic C Antitumor activity of vanadium compounds Metal Ions in Biol Sys 1995 31 595 616
Sakurai H Tamura H Okatawi K Mechanism for a new antitumor vanadium complex: hydroxyl radical dependent DNA cleavage by 1,10-phenanthroline-vandyl complexes in the presence of hydrogen peroxide Biochem Biophys Res Commun 1995 206 113 137 10.1006/bbrc.1995.1019
Bishayee A Chatterjee M Inhibitory effect of vanadium on rat liver carcinogenesis initiated with diethylnitrosamine and promoted by Phenobarbital British J Cancer 1995 71 1214 1220
Potter JD Food and cancer prevention II: summary of the meeting Cancer Letters 1997 114 337 338 9103326 10.1016/S0304-3835(97)04697-1
Gerster H Anticarcinogenic effect of common carotenoids Int J Vit Nutr Res 1993 63 93 12
Tanaka T Makita H Ohnishi M Hirose Y Wang A Mori H Satoh K Hara A Ogawa H Chemoprevention of 4-nitroquinoline 1-oxide-induced oral carcinogenesis by dietary curcumin and hesperidin: comparison with the protective effect of beta-carotene Cancer Res 1994 54 4653 4659 8062259
Sheriden RB IIIHuang PC Single strand breakage and repair of eukaryotic DNA as assayed by SI nuclease Nucleic Acids Res 1977 4 299 318 840643
Sarkar A Basak R Bishayee A Basak J Chatterjee M β-carotene inhibits rat liver chromosomal aberrations and DNA chain breaks a single injection of diethylnitrosamine British J Cancer 1997 76 855 861
Maenhaut W Malmquist KG Rene Van Grienken Hand Book of X-ray Spectrometry 1992 11 Andrej a Markowicz New York 517
Stewart HL Williams GM Keysser CH Lombard Ls Montail RJ Histological typing of liver tumours of the rat J Nat Cancer Inst 1980 64 117 207
Gupta RC Nonrandom binding of the carcinogen N-hydroxy-2 acetylaminofluorene to repetitive sequences of rat liver in vivo Proc Natl Acad Sci USA 1984 81 6943 6947 6594673
Kihlman BA Hansson K Andersson HC The effects of post-treatments with caffeine during S-G2 on the frequencies of chromosomal aberrations induced by the thiotepa in roots tips of Vicia fava Environ Exp Bot 1982 20 271 286 10.1016/0098-8472(80)90024-6
Cerutti PA Prooxidant status and tumour promotion Science 1985 227 375 381 2981433
Weitberg AB Weitzman SA Clark EP TP Stossel Effects of antioxidants on oxidant-induced sister chromatid exchange formation J Clin Invest 1985 75 1835 1841 3924956
Bishayee A Banik S Marimuthu P Chatterjee M Vanadium-mediated suppression of diethylnitrosamine-induced chromosomal aberrations in rat hepatocytes and its correlation with induction of hepatic glutathione and glutathione-S-transferase Int J oncology 1997 102 413 423
Bird RP Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen: preliminary findings Cancer Lett 1987 37 147 151 3677050 10.1016/0304-3835(87)90157-1
Basak R Saha BK Chatterjee M Inhibition of diethylnitrosamine-induced rat liver chromosomal aberrations and DNA-strand breaks by synergistic supplementation of vanadium and 1α, 25-dihydroxy vitamin D3 Biocheim Biophys Acta 2000 1502 273 282
Shenberg C Feidstein H Cornelis R Mees L Versieck J Vanhaltenberghe L Cafmeyer J Maenhout W Br, Rb, Zn, Fe, Se and K in blood of colorectal patients by INAA and PIXE J Trace Elem Med Biol 1995 9 193 199 8808190
Oberley L Buettner G Role of superoxide dismutase in cancer: a review Cancer Res 1979 39 1141 1149 217531
Halliwell B Oxidants and human disease: some new concepts Fed Am Soc Exp Biol J 1987 1 358
Sato K Glutathione transferase as markers of preneoplasia and neoplasia Adv Cancer Res 1990 52 205 255 2662713
Siegers CP Bumann D Baretten G Younes M Dietary iron enhances the tumour rate in dimethylhydrazine-induced colon carcinogenesis in mice Cancer Lett 1988 41 251 256 3409203 10.1016/0304-3835(88)90285-6
Nelson RL Yoo SJ Tanure JC Andriannopoulus G Misumi A The effect of iron on experimental colorectal carcinogenesis Anticancer Res 1989 9 1477 82 2560618
Singh M Lu J Briggs SP McGinley JN Haegele AD Thompson HJ Effect of excess dietary iron on the promotion stage of 1-methyl-1-nitrosourea-induced mammary carcinogenesis: pathogenetic characteristics and distribution of iron Carcinogenesis 1994 15 1567 1570 8055634
de Kok TM ten Vaarwerk F Zwingman I van Maanen JM Kleinjans JC Peroxidation of linoleic, arachidonic and oleic acid in relation to the induction of oxidative damage and cytogenetic effects Carcinogenesis 1994 15 1399 1404 8033317
| 15918908 | PMC1156919 | CC BY | 2021-01-04 16:24:16 | no | Cancer Cell Int. 2005 May 26; 5:16 | utf-8 | Cancer Cell Int | 2,005 | 10.1186/1475-2867-5-16 | oa_comm |
==== Front
Cancer Cell IntCancer Cell International1475-2867BioMed Central London 1475-2867-5-171591889510.1186/1475-2867-5-17Primary ResearchEndothelial cell pseudopods and angiogenesis of breast cancer tumors Cameron Ivan L [email protected] Nicholas [email protected] LuZhe [email protected] W Elaine [email protected] Department of Cellular and Structural Biology, University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78229, USA2 Pennington Biomedical Research Center, Louisiana State University, 6400 Perkins Road, Baton Rouge, LA 70808, USA2005 26 5 2005 5 17 17 14 7 2004 26 5 2005 Copyright © 2005 Cameron et al; licensee BioMed Central Ltd.2005Cameron et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
A neoplastic tumor cannot grow beyond a millimeter or so in diameter without recruitment of endothelial cells and new blood vessels to supply nutrition and oxygen for tumor cell survival. This study was designed to investigate formation of new blood vessels within a human growing breast cancer tumor model (MDA MB231 in mammary fat pad of nude female mouse). Once the tumor grew to 35 mm3, it developed a well-vascularized capsule. Histological sections of tumors greater than 35 mm3 were stained with PAS, with CD-31 antibody (an endothelial cell maker), or with hypoxia inducible factor 1α antibody (HIF). The extent of blood vessel and endothelial cell pseudopod volume density was measured by ocular grid intercept counting in the PAS stained slides.
Results
The tumor area within 100–150 μm of the well-vascularized capsule had few blood vessels and only occasional endothelial cell pseudopods, whereas the area greater than 150 μm from the capsule had more blood vessels, capillaries, and a three-fold increase in volume density of pseudopods sprouting from the capillary endothelial cells. This subcortical region, rich in pseudopods, some of which were observed to have vacuoles/lumens, was strongly positive for presence of HIF. In some larger tumors, pseudopods were observed to insinuate for mm distances through hypoxic regions of the tumor.
Conclusion
The positive correlation between presence of HIF and the increased extent of pseudopods suggests volume density measure of the latter as a quantifiable marker of tumor hypoxia. Apparently, hypoxic regions of the tumor produce HIF leading to production of vascular endothelial growth factors that stimulate sprouting of capillary endothelial cells and formation of endothelial cell pseudopods.
endothelial cell sproutingbreast cancerangiogenesis
==== Body
Background
Most tissues in the non-growing adult body have an adequate vascular supply with no need for formation of new blood vessels. Growth in tissue or tumor mass however requires recruitment of endothelial cells to form new blood vessels. In an early report, new blood capillaries were directly observed to form by sprouting of pseudopods from existing endothelial cells [1]. Speidel reported endothelial cell sprouting in the regenerating tail fin of the frog (Rana clamitans) 6 day after amputation. He saw capillary sprouts that progressed for 32 μm through the tissue over a 50-minute period, leaving a solid capillary sprout behind. Up to seven side sprouts or pseudopods also arose along the length of this main capillary sprout during this 50-minute period. He also observed a sprout that acquired a vacuole that fused with the capillary lumen and allowed entry of an erythrocyte from the capillary lumen. This report of morphological events involved in capillary cell sprouting and pseudopod formation remains the most definite description of endothelial cell sprouting and pseudopod formation.
The recruitment of endothelial cells by a tumor is a key early step in tumor angiogenesis [2]. This tumor angiogenic process has become an important target in cancer therapy [2-4]. There is now strong molecular and genetic support for targeting tumor angiogenesis [2-4]. However, there is much more to be learned about tumor angiogenesis. In our own investigation of angiogenesis of the human breast cancer tumors (MDA MB231) growing in the female nude mouse xenograft, we discovered that staining 8 μm thick histological mid-cross sections of the tumor with the periodic acid-Schiff (PAS) technique for glycoproteins [5] revealed an unexpected abundance of PAS positive endothelial cell pseudopods. The distribution of endothelial cell pseudopods was observed not to be random throughout the tumor. This is the report of a study done to determine the spatial pattern and the possible explanation for this non-random pattern of endothelial cell pseudopods during tumor angiogenesis.
Results
A total of 10 tumors greater than 35 mm3 were fixed, embedded in paraffin wax and midsections were cut at 8 μm thickness for staining and histological study. Histological examination of midsections of PAS stained tumors was done to assess the tumor vascularization patterns. Tumors with a volume >35 mm3 demonstrated a connective tissue capsule with blood vessels (Fig. 1A). At higher magnification the cortical area within about 100 μm of the capsule revealed few blood vessels while the area greater than 100 μm from the capsule showed considerable evidence of blood vessels and of capillaries with many endothelial pseudopods extending away from the capillaries (Fig. 1B&C). The general direction of the pseudopods was parallel to the tumor capsule surface. Immunohistochemical localization of CD-31, used as a specific marker of blood vessel endothelium, demonstrated a positive reaction of pseudopods (Fig. 1D). Areas of tumor necrosis were observed below the subcortical area (Fig. 1E&F). Immunohistochemical localization of hypoxia inducible factor 1-α (HIF) reveals the subcortical area of the tumor to contain HIF positive cells while the tumor capsule, cortex and necrotic areas of the tumor demonstrate no evidence of HIF. Thus, the areas found to be HIF positive were enriched in endothelial pseudopods.
Figure 1 Photomicrographs illustrating the pattern of the tumor vascularization. A, The tumor capsule (left) of this PAS stained section reveals blood vessels. The cortex under the capsule reveals no blood vessels and few endothelial pseudopods while the subcortical area to the top right has more pseudopods. B, at higher magnification the subcortical area of the tumor reveals a blood capillary with multiple endothelial pseudopods protruding at right angles into the tumor mass. C, Endothelial pseudopods are seen to branch and occasionally have a vacuole/lumen (arrow). D, The endothelial pseudopods react immunohistochemically positive for the CD-31 specific endothelial cell marker, using the avidin-biotin peroxidase complex method. E, Viable cell area can be seen beneath tumor capsule (left) while necrotic area can be seen to the right. F, Enlarged subcortical area from E. In E, the HIF-α positive area between the viable and the necrotic tissue is stained brown.
Ocular grid intercept counting was used to quantify tumor vascularization in 8 μm thick PAS stained histological sections of the tumors. The numbers of ocular grid intercepts were scored over: 1.) blood vessels and capillaries, 2.) endothelial pseudopods and 3.) areas with no indication of these structures. This method has been shown to be a usable measure of volume density occupied by recognizable structures. The results of the scoring of blood vessels and of endothelial pseudopods in the subcortical regions of the tumors are summarized in Fig. 2. As illustrated in Fig. 2, there was significantly less pseudopod volume density in the cortical versus the subcortical region of the tumor.
Figure 2 Quantification of volume density differences of endothelial cell pseudopods in the tumor cortex, within 100 to 150 μm of the well-vascularized tumor capsule, and in the subcortical region at a distance of greater than 150 μm beneath the tumor capsule. The mean ± SEM of the pseudopod volume density of 10 tumors in the two different regions of the tumors indicate a significant (p < 0.001) three fold higher value of pseudopod volume density in the subcortical region compared to the cortical region of the tumor.
Can the observation of PAS positive endothelial cell pseudopods be generalized to human breast cancer surgical biopsy specimens? To address this question, specimens from 15 breast cancer patients, obtained from the University of Texas Health Science Center at San Antonio Department of Pathology, were examined. Of these 15 patients, nine were diagnosed with ductal confinement intraductal carcinoma, or "DCIC." Although five showed multilayering of cancer cells within the ducts, none showed evidence of endothelial pseudopods within this cancer cell population. Seven of the biopsy specimens were diagnosed as invasive ductal carcinoma, or "IDS." Six of the seven IDS specimens showed evidence of endothelial pseudopods but to a lesser extent than in the MDA MB231 human breast cancer tumor used in the current study. Two IDS tumor specimens had visible regions of necrosis. The region of the tumor adjacent to these necrosis regions had more pseudopods than seen elsewhere in the tumor. Thus, PAS positive endothelial cell pseudopods were commonly present in the human IDS specimens.
Discussion
The PAS stained histological sections of the tumors were useful for identification of the spatial pattern of tumor vascularization. Not only could blood vessels and capillaries be identified in the tumor but numerous pseudopods were also observed sprouting from capillaries. The positive reaction to the CD-31 endothelial marker confirmed that the PAS stained structures were blood vessels, capillaries and pseudopods of endothelial cell origin. Figure 3 is a drawing that summarizes the observed distribution of blood vessels and endothelial cell pseudopods. Tumors greater then 35 mm3 revealed distinct areas of viable and necrotic tissue. An area of viable tumor cells was observed beneath the vascularized tumor capsule. The area of viable tumor cells had two distinct subdivisions. The cortical areas of the tumor within 100 to 150 microns of the tumor capsule was not well vascularized and had few endothelial pseudopods whereas the subcortical area further than 100 to 150 microns from the tumor capsule had blood vessels and had many more endothelial pseudopods. Even further beneath the tumor capsule, in the areas of tumor necrosis, there was no evidence of blood vessels, capillaries or endothelial pseudopods.
Figure 3 Model of vascularization as observed in the human MDA MB231 breast cancer cell tumor grown as a xenograph in a female nude mouse. The well vascularized tumor capsule overlies the tumor cortex. The cortex extends between 100 and 150 μm beneath the capsule. The HIF-α positive region is located in the subcortical region, at distances greater than 150 μm of the capsule. More endothelial pseudopods are located in this subcortical region than are located in the cortical region of the tumor. Even further away from the vascularized capsule is the region of tumor necrosis which contains no blood vessels or endothelial cell pseudopods. This model is consistent with the conclusion that the hypoxic regions of the tumor produce HIF leading to production of vascular endothelial growth factors and sprouting of endothelial cell pseudopods.
The observed localization of HIF in the tumors by immunohistochemistry helped explain the vascular pattern of the encapsulated tumors. The HIF positive reactivity in the tumor was localized in the subcortical areas of the tumor in the same area found to have the most endothelial cell pseudopods (Fig. 2). These findings suggest that this subcortical area of the tumor is hypoxic and is producing angiogenesis growth factors, which in turn acts on the host endothelial cells to sprout pseudopods. Sprouting of cultured endothelial cells in response to vascular endothelial growth factors has been reported [6]. In the same report, Feraud et al. demonstrated that several angiostatic agents could inhibit this endothelial cell sprouting. The presence of pseudopodial like structures in viable areas of a tumor immediately adjacent to necrotic areas of the tumor has been previously reported [7]. The lack of HIF reactivity beneath the well vascularized tumor capsule in the cortical area suggests this cortical area is not hypoxic and not in need of an extensive vasculature. Apparently, the PAS staining of endothelial cell pseudopods in tumors gives quantifiable spatial information on where in the tumor that HIF and vascular endothelial growth factors are located.
Materials and Methods
A detailed account of the materials and methods used in this study has been published elsewhere [8]. Thus, only a summary is given here. Fifteen 6-week-old female athymic nude mice where fed AIN-76 semipurified diet altered to contain 10% corn oil. Two millions human breast cancer cells were inoculated into the inguinal mammary fat pad of each mouse. Once the tumors had grown to greater than 35 mm3 the mice were euthanized by injection of a ketamine/rompun anesthesia supplied by the University of Texas Health Science Center at San Antonio Laboratory Animal veterinarian, then cervically dislocated and exsanguinated by cardiac puncture. The tumors were then removed and fixed in Omni Fix II (Mt. Vernon, NY), dehydrated and embedded in paraffin, sectioned at 4 or 8 μm thickness, mounted on slides, deparaffinized, and stained with H&E or with periodic acid-Schiff (PAS). Additional slides were stained for the immunohistochemical markers for endothelial cells, CD-31 (PECAM-1, PharMingen) or for hypoxia-inducible factor 1-alpha (#OSA-601, Stressgen). Quantification of the volume density of endothelial cell pseudopods was done on the 8 μm thick PAS stained histological sections of the tumors using counts of the number of ocular grid line intercepts.
Archived surgical biopsy specimens that had been fixed in 10% neutralized formalin were dehydrated and embedded in paraffin. Histological sections of the specimens were stained with H&E for routine histopathology examination or with PAS to assess endothelial cell pseudopod distribution.
Acknowledgements
This work was supported by LSU fund 44096, NIH grant CA7553, and EMF Therapeutics, Inc.
==== Refs
Speidel CC Studies of living nerves II. Activities of ameboid growth cones, sheath cells, and myelin segments, as revealed by prolonged observation of individual nerve fibers in frog tadpoles Am J Anat 1933 52 29 79 10.1002/aja.1000520102
Folkman J Browder T Palmblad J Angiogenesis research: Guidelines for translation to clinical application Thromb Haemost 2001 86 23 33 11487011
Carmeliet P Mechanisms of angiogenesis and arteriogenesis Nat Med 2000 6 389 395 10742145 10.1038/74651
Longo R Sarmiento R Fanelli M Capaccetti B Gattuso D Gasparini G Anti-angiogenic therapy: Rationale, challenges, and clinical studies Angiogenesis 2002 5 237 256 12906317 10.1023/A:1024532022166
Rambourg A Neutra M Leblond CP Presence of a 'cell coat' rich in carbohydrate at the surface of cells in the rat Anat Rec 1966 154 41 72 4162458 10.1002/ar.1091540105
Feraud O Cao Y Vittet D Embryonic stem cell-derived embryoid bodies development in collagen gels recapitulates sprouting angiogenesis Lab Invest 2001 81 1669 1681 11742037
Yang M Li L Jiang P Moossam AR Penman S Hoffman RM Dual-color fluorescence imaging distinguishes tumor cells from induced host angiogenic vessels and stromal cells Proc Natl Acad Sci USA 2003 100 14259 14262 14614130 10.1073/pnas.2436101100
Cameron IL Sun LZ Short N Hardman WE Williams CD Therapeutic electromagnetic field (TEMF) and gamma irradiation on human breast cancer xenograph growth, angiogenesis and metastasis Pending in Cancer Cell International 2004
| 15918895 | PMC1156920 | CC BY | 2021-01-04 16:24:16 | no | Cancer Cell Int. 2005 May 26; 5:17 | utf-8 | Cancer Cell Int | 2,005 | 10.1186/1475-2867-5-17 | oa_comm |
==== Front
Cardiovasc DiabetolCardiovascular Diabetology1475-2840BioMed Central London 1475-2840-4-81594147110.1186/1475-2840-4-8EditorialStatins research unfinished saga: desirability versus feasibility Fisman Enrique Z [email protected] Yehuda [email protected] Alexander [email protected] The Sackler Faculty of Medicine, Tel-Aviv University, 69978 Ramat-Aviv, Tel-Aviv, Israel2 Cardiac Rehabilitation Institute, Sheba Medical Center, 52621 Tel-Hashomer, Israel2005 7 6 2005 4 8 8 3 6 2005 7 6 2005 Copyright © 2005 Fisman et al; licensee BioMed Central Ltd.2005Fisman et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Drugs in the same class are generally thought to be therapeutically equivalent because of similar mechanisms of action (the so-called "class effect"). However, statins differ in multiple characteristics, including liver and renal metabolism, half-life, effects on several serum lipid components, bioavailability and potency. Some are fungal derivatives, and others are synthetic compounds. The percentage absorption of an oral dose, amount of protein binding, degree of renal excretion, hydrophilicity, and potency on a weight basis is variable. These differences may be even greater in diabetic patients, who may present diabetes-induced abnormalities in P450 isoforms and altered hepatic metabolic pathways. Thus, it is obvious that head-to-head comparisons between different statins are preferable than trial-to-trial comparisons. Such assessments are of utmost importance, especially in cases in which specific populations with a distinct lipid profile and altered metabolic pathways, like diabetics, are studied. It should be specially pinpointed that patients with metabolic syndrome and diabetes constitute also a special population regarding their atherogenic dyslipidemia, which is usually associated with low HDL-cholesterol, hypertriglyceridemia and predominance of small dense LDL-cholesterol. Therefore, these patients may benefit from fibrates or combined statin/fibrate treatment. This policy is not accomplished since in the real world things are more complex. Trials would require very large sample sizes and long-term follow-up to detect significant differences in myocardial infarction or death between two different statins. Moreover, the fact that new compounds are under several phases of research and development represents an additional drawback for performing the trials. Ideally, head-to-head trials regarding clinically important outcomes should be conducted for all drugs. Nonetheless, the desirability of performing such trials, which epitomize modern evidence-based medicine, is frequently superseded by the feasibility dictated by pragmatic and economic circumstances. In the latter case, in absence of solid systematic documentation of comparable health benefits and long-term safety, both researchers and practicing physicians should allude to the weight of scientific endorsement behind the arguments and seek for the possible strengths and weaknesses intrinsic to each specific study. In any case, conclusions based on surrogate endpoints cannot completely substitute head-to-head comparisons regarding patients' outcome.
coronary artery diseasediabeteshyperlipidemiastatinstrials
==== Body
During the last fifteen years large-scale clinical trials were conducted aiming to determine the antihyperlipidemic activity of several 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors (statins). These studies included the Scandinavian Simvastatin Survival Study (4S) [1], the West of Scotland Coronary Prevention Study (WOSCOPS) [2], the Expanded Clinical Evaluation of Lovastatin (EXCEL) study [3], among others. Several important concepts emerged from these studies. It was shown that lipid-lowering pharmacotherapy is justified for patients with hypercholesterolemia who are at risk of coronary artery disease (CAD). Lipid-lowering medication produced significant reductions in death from CAD and non-fatal myocardial infarction (MI), all cardiovascular death and all-cause mortality, clearly supporting the use of statins in such patients. In addition, statins induced modest increases in high-density lipoprotein cholesterol, and decreases in triglycerides [1-3]. No direct comparisons between these drugs were available at this stage. Thus, questions regarding the relative superiority and/or safety of a given compound remained unanswered. These queries have need of an appropriate response.
Head-to-head comparisons
Following the pioneering initial studies, more recent head-to-head comparative parallel-group trials unveiled additional features of this type of drugs. In 2001, the Reversal of Atherosclerosis with Aggressive Lipid Lowering (REVERSAL) trial demonstrated that atorvastatin 80 mg/day partially reversed carotid intima-media thickness in patients with familial hypercholesterolemia, as compared with simvastatin 40 mg/day [4].
Furthermore, the Atorvastatin versus Simvastatin on Atherosclerosis Progression (ASAP) trial study showed that high-dose atorvastatin delayed progression of coronary atherosclerosis, as assessed by intravascular ultrasound, while 40 mg/day pravastatin did not [5]. These trials demonstrated that aggressive statin therapy is more effective in terms of atherosclerosis progression. Importantly, statins have also demonstrated to reduce endothelial dysfunction, inflammation, noxious cytokines concentration and blood thrombogenicity, which all seem to be co-responsible for plaque thrombosis [6,7].
In the Pravastatin or Atorvastatin Evaluation and Infection Therapy (PROVE-IT) trial [8] 4162 patients who had suffered an acute coronary syndrome in the preceding 10 days were randomized to receive pravastatin 40 mg/day vs atorvastatin 80 mg/day in addition to gatifloxacin vs placebo. The data of the lipid-lowering therapy showed mean low-density lipoprotein (LDL) levels of 95 mg/dl in the pravastatin group, in compliance with present guidelines of clinical practice, and 62 mg/dl in the atorvastatin group. After a follow-up of 2 years, the intensive lipid-lowering group showed a significant reduction in the incidence of the combined end-point of death, MI, unstable angina requiring hospitalization, revascularization and stroke. The PROVE-IT findings were strengthened by the Aggressive Lipid-Lowering Initiation Abates New Cardiac Events (ALLIANCE) study, which confirmed the hypothesis of further benefits with aggressive cholesterol-lowering [9]. In this trial, 2442 patients with coronary artery disease were randomized to achieve LDL levels <80 mg/dl, receiving atorvastatin titrated at doses ≤ 80 mg/day, or standard therapy with the statins and doses chosen by the treating physicians. Intensive lipid-lowering therapy demonstrated a significant reduction in the primary end-point of combined incidence of cardiac death, myocardial infarction, coronary revascularization and unstable angina requiring hospitalization.
However, it should be pinpointed that lipid reduction, decrease of markers of hemostasis and inflammation or reduction in number of atherosclerotic plaques should be viewed as surrogate end points of statins activity. The extent to which these results can be extrapolated to clinically relevant outcomes remains to be established [10]. It is yet unclear whether all statins are equally effective in preventing recurrent MI and death at a long-term follow-up. Drugs in the same class are generally thought to be therapeutically equivalent because of similar mechanisms of action (the so-called "class effect" [11]). In this context, however, statins differ in multiple characteristics, including liver and renal metabolism, half-life, effects on several serum lipid components, bioavailability and potency. In addition, some are fungal derivatives, and others are synthetic compounds. The percentage absorption of an oral dose, amount of protein binding, degree of renal excretion, hydrophilicity, and potency on a weight basis varies among the individual agents [12,13]. Despite these differences, a just published Canadian retrospective cohort study comparing atorvastatin, pravastatin, simvastatin, lovastatin and fluvastatin in patients aged >65 years after their first MI claims that all these statins are equally effective [10]. The study included above 18500 patients who began statin treatment within 90 days after discharge, and the primary end point was the combined outcome of recurrent MI or death from any cause.
Trial-versus-trial comparisons
These comparisons present the hazard of being potentially misleading because of fundamental "apples-to-oranges" comparisons secondary to differences between trials in patient populations, inclusion criteria, management algorithms, and end-point definitions [14]. Although statins share common main actions, they may have clinically important differences in terms of efficacy and safety. In this context, it should be remembered that as a result of 31 reported cases of fatal rhabdomyolysis, cerivastatin was withdrawn from the market, underscoring the risk of therapeutic interchangeability [15]. Modern pharmacological decision-making and guidelines development often rely on meta-analyses in order to handle the vast amount of clinical information available from clinical trials in cardiovascular medicine. Despite the establishment of standards for reporting meta-analyses, there are also limitations that need to be acknowledged. Statistical testing for heterogeneity of treatment effects across drugs in a class is not a particularly sensitive analysis. Thus, without access to original source data, reliably identifying adverse events can be very difficult [14].
An event illustrating this problem comes from New Zealand. After a government decision to replace simvastatin with fluvastatin as the approved statin for reimbursement, Thomas and Mann [16] investigated the effect of the switch in 126 patients; the lipid concentrations went up in 115 (94%) of the patients after the switch to fluvastatin. In addition, during the 6 months before the switch, there were nine arterial thrombotic events compared with 27 during the same time after the switch (p < 0.05). This study spells out the danger of allowing budgetary arguments about class effects to influence drug selection.
Statins in diabetic patients
Part of the patients included in studies evaluating statins are diabetics. Both the PROVE-IT [8] and the ALLIANCE [9] trials were performed in populations comprising both diabetic and nondiabetic patients, albeit the percentage of diabetic patients was relatively small, averaging about 20% of the examinees.
Various trials demonstrated improved prognosis and reduction of cardiovascular events in patients with diabetes or impaired fasting glucose taking several statins [17-20]. However, none of these was a head-to-head comparison of two statins regarding their effectiveness in a diabetic population. The recently published study by Berne and Siewert-Delle [21] addresses precisely this point. Rosuvastatin was compared with atorvastatin in type 2 diabetes mellitus for the reduction of LDL-cholesterol. They found that at 4 weeks, 65% of rosuvastatin patients had reached their 2003 European LDL- cholesterol goal (<2.5 mmol/L), compared with 33% of atorvastatin patients (p < 0.0001). Both treatments were similarly well tolerated with no unexpected safety concerns.
The mechanism behind these differences has not yet been elucidated. A possible explanation is the dissimilarity in their hepatic metabolization. Atorvastatin is a relatively lipophilic compound; lipophilic statins are more susceptible to metabolism by the cytochrome P450 system. On the contrary, rosuvastatin is relatively hydrophilic and not significantly metabolized by cytochrome P450 enzymes [22]. Diabetes-induced abnormalities in P450 isoforms were described in both experimental and clinical studies [22-24]. This could explain the better compliance of diabetic patients to the latter drug.
Lipid-modifying strategy in metabolic syndrome and diabetes: beyond the LDL-cholesterol
Whereas statins remain the useful drug for patients who need to achieve the LDL-cholesterol goal, other lipid-lowering compounds – fibrates – may represent the alternative intervention for subjects with atherogenic dyslipidemia typical for metabolic syndrome and an LDL-cholesterol already close to goal values. An atherogenic dyslipidemia is characterized by elevated levels of triglycerides, reduced levels of high-density lipoprotein cholesterol and a preponderance of small dense LDL-cholesterol particles.
The Adult Treatment Panel III guidelines suggest that because elevated trigycerides are an independent CAD risk factor, some trigyceride-rich lipoproteins, commonly called remnant lipoproteins, must be atherogenic. A novel method for the isolation and quantification of plasma remnants was developed by Nakajima and Nakamura [25]. Specific immunoaffinity-based gel is incubated with plasma, which results in binding of high-density lipoproteins (HDL), LDL, and the majority of very low-density lipoproteins (VLDL) particles to the gel. Unbound lipoproteins are quantified on the basis of their cholesterol content; this is termed the remnant-like particles cholesterol (RLP-C) concentration. It is established that elevated plasma RLP-C levels are associated with endothelial dysfunction, a marker for atherosclerotic disease. Patients with established coronary heart disease present elevated plasma levels of RLP-C. Elevated levels of plasma RLP-C were predictive of future coronary events in patients with CAD independently of other risk factors [26].
Extensive evidence from intervention trials has been presented to demonstrate that "old" statin treatment results in reduced cardiovascular morbidity and mortality. By contrast, statin treatment has not been associated consistently with reduction in plasma RLP-C levels. In general, however, it can be concluded that RLP-C lowering by statins depends on their ability to reduce triglyceride levels. Therefore, among all currently clinically available statins only atorvastatin and rosuvastatin have potential to reduce RLP-C levels [27].
In addition, the concomitant use of fibrate and statin seems to be attractive in patients whose LDL-cholesterol is controlled by statin but whose HDL-cholesterol and/or triglyceride levels are still inappropriate [28]. A combination statin/fibrate may be necessary to control all lipid abnormalities in patients with metabolic syndrome and diabetes. Safety concerns about some fibrates such as gemfibrozil may lead to exaggerate precautions regarding fibrate administration and therefore diminish the use of these agents. However, other fibrates- such as bezafibrate and fenofibrate – appear to be safer and better tolerated [29-32].
The combination of fibrate with statin which are not mainly metabolized within the liver by the cytochrome P450 system (like pravastatin or fluvastatin) may be even less hazardous. We believe that a proper coadministration of statins and fibrates in some cases, selected on basis of their safety, could be more effective in achieving a comprehensive lipid control as compared with monotherapy. An alternative sort of future combination may hypothetically be represented by ezetimibe and fibrate. Ezetimibe represents the newest class of lipid-modifying agents. It exerts its cholesterol-lowering effect by inhibiting the absorption of both the dietary cholesterol and biliary cholesterol at the brush border of the small intestine. Because of its distinct mechanism of action, ezetimibe appears to be most useful as part of combination therapy with other lipid-modifying agents rather than as monotherapy. Ezetimibe is approved currently for use alone or with statins, and initial study with fibrates are promising [32-34].
Desirability and feasibility
Prospective randomized parallel-group head-to-head comparisons between different drugs were not performed during the early stages of statins human use. The main reason was the unwillingness of industry sponsors to take on the risk of failing to show their drug is noninferior to another agent in the same class [14]. Moreover, such a study may potentially became a menacing 'boomerang" if it demonstrates that the competitor's drug is preferable, as it actually happened in the PROVE-IT trial [8]. While statistical techniques for estimating the comparative therapeutic efficacy of "competing" compounds using indirect methods have been proposed, the soundness of these adjusted indirect comparisons is limited, and depends on the accuracy and similarity of the included trials [35].
Thus, it seems even needless to state that head-to-head comparisons are preferable. It would be preferable that every drug (and indeed every dose and every formulation [36]) be evaluated in randomized comparative trials with active comparators from the same class for their effects on clinically important outcomes. This is not accomplished since. Regretfully, in the real world things are more complex.
Pharmaceutical companies spend millions of dollars to convince consumers and doctors that their products are superior to competitor brands; however, these companies cannot be compelled to design and finance trials in which a future monetary profit is dubious.
In addition, trials would require very large sample sizes and long-term follow-up to detect significant differences in myocardial infarction or death between two different statins [35,36]. Moreover, the fact that new compounds, like the presumably potent pitavastatin [37], are under several phases of research and development represents an additional drawback for performing the trials.
The assertion that there really does not appear to be much difference across statins in terms of their effectiveness and they appear to be similar one to another (in a study which did not include rosuvastatin) [10] is seriously defied by head-to-head comparisons performed so far [5-9,21,38]. Such assessments are of utmost importance, especially in cases in which specific populations with a distinct lipid profile and altered metabolic pathways [22-24], like diabetics, are studied.
Ideally, head-to-head trials regarding clinically importantoutcomes should be conducted for all drugs. Nonetheless, the desirability of performing such trials, which epitomize modern evidence-based medicine, is frequently superseded by the feasibility dictated by pragmatic and economic circumstances. In absence of solid systematic documentation of comparable health benefits and long-term safety, both researchers and practicing physicians should allude to the weight of scientific endorsement behind the arguments and seek for the possible strengths and weaknesses intrinsic to each specific study. In addition, surrogate end points, like LDL-cholesterol level, allow conducting shorter and smaller trials with acceptable level of credibility regarding clinical significance of the new compound. A surrogate outcome will be reliable only if there is a validated causal connection between change in surrogate and change in the clinically important outcomes (like morbidity and mortality), and if the surrogate reflects most of the effects of treatment on a specific outcome. The recent paper of Berne and Siewert-Delle [21] denotes a good example for an accurate approach using a relatively reliable (LDL-cholesterol level) surrogate end-point. However, conclusions based on surrogate endpoints can not completely substitute head-to-head trials, particularly in relation to the magnitude of potential benefits in clinical practice.
List of abbreviations used
ALLIANCE – Aggressive Lipid-Lowering Initiation Abates New Cardiac Events
ASAP – Atorvastatin versus Simvastatin on Atherosclerosis Progression
CAD – coronary artery disease
EXCEL – Expanded Clinical Evaluation of Lovastatin
HDL – high-density lipoproteins
LDL low-density lipoprotein
MI – myocardial infarction
PROVE-IT – Pravastatin or Atorvastatin Evaluation and Infection Therapy
REVERSAL – Reversal of Atherosclerosis with Aggressive Lipid Lowering
RLP-C – remnant-like particles cholesterol
VLDL – very low-density lipoproteins
WOSCOPS – West of Scotland Coronary Prevention Study
4S – Scandinavian Simvastatin Survival Study
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
All authors contributed equally in the conception and drafting of the manuscript.
Acknowledgements
This work was supported in part by the Cardiovascular Diabetology Research Foundation (RA 58-040-684-1), Holon, Israel, and the Research Authority of Tel-Aviv University (Dr. Ziternick and Haia Silva Ziternick Fund, grants 01250238 and 01250239).
==== Refs
Scandinavian Simvastatin Survival Study (4S): Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S) Lancet 1994 344 1383 1389 7968073 10.1016/S0140-6736(94)92521-6
Shepherd J Cobbe SM Ford I Isles CG Lorimer AR MacFarlane PW McKillop JH Packard CJ Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. West of Scotland Coronary Prevention Study Group N Engl J Med 1995 333 1301 1307 7566020 10.1056/NEJM199511163332001
Bradford RH Shear CL Chremos AN Dujovne CA Franklin FA Grillo RB Higgins J Langendorfer A Nash DT Pool JL Expanded Clinical Evaluation of Lovastatin (EXCEL) study results: two-year efficacy and safety follow-up Am J Cardiol 1994 74 667 673 7942524 10.1016/0002-9149(94)90307-7
Nissen SE Tuzcu EM Schoenhagen P Brown BG Ganz P Vogel RA Crowe T Howard G Cooper CJ Brodie B Grines CL DeMaria AN REVERSAL Investigators Effect of intensive compared with moderate lipid-lowering therapy on progression of coronary atherosclerosis: a randomized controlled trial JAMA 2004 291 1071 1080 14996776 10.1001/jama.291.9.1071
Smilde TJ van Wissen S Wollersheim H Trip MD Kastelein JJ Stalenhoef AF of aggressive versus Effect conventional lipid lowering on atherosclerosis progression in familial hypercholesterolaemia (ASAP): a prospective, randomised, double-blind trial Lancet 2001 357 577 581 11558482 10.1016/S0140-6736(00)04053-8
Blanco-Colio LM Blanco-Colio LM Munoz-Garcia B Gomez-Hernandez A Arribas A Ortega L Tunon J Egido J NF-kappaB activation and Fas ligand overexpression in blood and plaques of patients with carotid atherosclerosis: potential implication in plaque instability Stroke 2004 35 458 463 14739407 10.1161/01.STR.0000114876.51656.7A
Tunon J Blanco-Colio LM Blanco-Colio LM Egido J Intensive treatment with statins and the progression of cardiovascular diseases: the beginning of a new era? Nephrol Dial Transplant 2004 19 2696 2699 15340095 10.1093/ndt/gfh472
Cannon CP Braunwald E McCabe CH Rader DJ Rouleau JL Belder R Joyal SV Hill KA Pfeffer MA Skene AM Pravastatin or Atorvastatin Evaluation and Infection Therapy-Thrombolysis in Myocardial Infarction 22 Investigators Comparison of intensive and moderate lipid lowering with statins after acute coronary syndromes N Engl J Med 2004 350 1495 1504 15007110 10.1056/NEJMoa040583
Koren MJ Hunninghake DB ALLIANCE Investigators Clinical outcomes in managed-care patients with coronary heart disease treated aggressively in lipid-lowering disease management clinics: the alliance study J Am Coll Cardiol 2004 44 1772 1779 15519006 10.1016/j.jacc.2004.07.053
Zhou Z Rahme E Abrahamowicz M Tu JV Eisenberg MJ Humphries K Austin PC Pilote L Effectiveness of statins for secondary prevention in elderly patients after acute myocardial infarction: an evaluation of class effect CMAJ 2005 172 1187 1194 15851712
Furberg CD Herrington DM Psaty BM Are drugs within a class interchangeable? Lancet 1999 354 1202 1204 10513728 10.1016/S0140-6736(99)03190-6
Vaughan CJ Gotto AM JrBasson CT The evolving role of statins in the management of atherosclerosis J Am Coll Cardiol 2000 35 1 10 10636252 10.1016/S0735-1097(99)00525-2
Hsu I Spinler SA Johnson NE Comparative evaluation of the safety and efficacy of HMG-CoA reductase inhibitor monotherapy in the treatment of primary hypercholesterolemia Ann Pharmacother 1995 29 743 759 8520093
Antman EM Ferguson JJ Should evidence-based proof of efficacy as defined for a specific therapeutic agent be extrapolated to encompass a therapeutic class of agents? Circulation 2003 108 2604 2607 14638523 10.1161/01.CIR.0000090571.83005.CE
Chong PH Lack of therapeutic interchangeability of HMG-CoA reductase inhibitors Ann Pharmacother 2002 36 1907 1917 12452755 10.1345/aph.1C116
Thomas M Mann J Increased thrombotic vascular events after change of statin Lancet 1998 352 1830 1831 9851392 10.1016/S0140-6736(05)79893-7
Pyörälä K Pedersen TR Kjekshus J Faergeman O Olsson AG Thorgeirsson G Cholesterol lowering with simvastatin improves prognosis of diabetic patients with coronary heart disease. A subgroup analysis of the Scandinavian Simvastatin Survival Study (4S) Diabetes Care 1997 20 614 620 9096989
Goldberg RB Mellies MJ Sacks FM Moye LA Howard BV Howard WJ Davis BR Cole TG Pfeffer MA Braunwald E Cardiovascular events and their reduction with pravastatin in diabetic and glucose-intolerant myocardial infarction survivors with average cholesterol levels: subgroup analyses in the Cholesterol And Recurrent Events (CARE) trial Circulation 1998 98 2513 2519 9843456
Keech A Colquhoun D Best J Kirby A Simes RJ Hunt D Hague W Beller E Arulchelvam M Baker J Tonkin A LIPID Study Group Secondary prevention of cardiovascular events with long-term pravastatin in patients with diabetes or impaired fasting glucose: results from the LIPID trial Diabetes Care 2003 26 2713 2721 14514569
Collins R Armitage J Parish S Sleigh P Peto R Heart Protection Study Collaborative Group MRC/BHF Heart Protection Study of cholesterol-lowering with simvastatin in 5963 people with diabetes: a randomised placebo-controlled trial Lancet 2003 361 2005 2016 12814710 10.1016/S0140-6736(03)12475-0
Berne C Siewert-Delle A on behalf of the URANUS study investigators Comparison of rosuvastatin and atorvastatin for lipid lowering in patients with type 2 diabetes mellitus: results from the URANUS study Cardiovasc Diabetol 2005 4 7 15935095 10.1186/1475-2840-4-7
Schachter M Chemical, pharmacokinetic and pharmacodynamic properties of statins: an update Fundam Clin Pharmacol 2005 19 117 125 15660968 10.1111/j.1472-8206.2004.00299.x
Enriquez A Leclercq I Farrell GC Robertson G Altered expression of hepatic CYP2E1 and CYP4A in obese, diabetic ob/ob mice, and fa/fa Zucker rats Biochem Biophys Res Commun 1999 255 300 306 10049703 10.1006/bbrc.1999.0202
Wang Z Hall SD Maya JF Li L Asghar A Gorski JC Diabetes mellitus increases the in vivo activity of cytochrome P450 2E1 in humans Br J Clin Pharmacol 2003 55 77 85 12534643 10.1046/j.1365-2125.2003.01731.x
Nakajima K Saito T Tamura A Suzuki M Nakano T Adachi M Tanaka A Tada N Nakamura H Campos E Cholesterol in remnant-like lipoproteins in human serum using monoclonal anti apo B-100 and anti apo A-I immunoaffinity mixed gels Clin Chim Acta 1993 223 53 71 8143370 10.1016/0009-8981(93)90062-9
Kugiyama K Doi H Takazoe K Kawano H Soejima H Mizuno Y Tsunoda R Sakamoto T Nakano T Nakajima K Ogawa H Sugiyama S Yoshimura M Yasue H Remnant lipoprotein levels in fasting serum predict coronary events in patients with coronary artery disease Circulation 1999 99 2858 2860 10359727
Lamendola C Abbasi F Chu JW Hutchinson H Cain V Leary E McLaughlin T Stein E Reaven G Comparative effects of rosuvastatin and gemfibrozil on glucose, insulin, and lipid metabolism in insulin-resistant, nondiabetic patients with combined dyslipidemia Am J Cardiol 2005 95 189 193 15642550 10.1016/j.amjcard.2004.09.005
Tenenbaum A Fisman EZ Which is the best lipid-modifying strategy in metabolic syndrome and diabetes: fibrates, statins or both? Cardiovasc Diabetol 2004 3 10 15574199 10.1186/1475-2840-3-10
Tenenbaum A Fisman EZ Motro M Rhabdomyolysis and lipid-lowering drugs JAMA 2005 293 1448 15784863 10.1001/jama.293.12.1448-a
Kyrklund C Backman JT Kivisto KT Neuvonen M Laitila J Neuvonen PJ Plasma concentrations of active lovastatin acid are markedly increased by gemfibrozil but not by bezafibrate Clin Pharmacol Ther 2001 69 340 345 11372002 10.1067/mcp.2001.115542
Beggs PW Clark DW Williams SM Coulter DM A comparison of the use, effectiveness and safety of bezafibrate, gemfibrozil and simvastatin in normal clinical practice using the New Zealand Intensive Medicines Monitoring Programme (IMMP) Br J Clin Pharmacol 1999 47 99 104 10073746 10.1046/j.1365-2125.1999.00846.x
Farnier M Salko T Isaacsohn JL Troendle AJ Dejager S Gonasun L Effects of baseline level of triglycerides on changes in lipid levels from combined fluvastatin + fibrate (bezafibrate, fenofibrate, or gemfibrozil) Am J Cardiol 2003 92 794 797 14516878 10.1016/S0002-9149(03)00885-3
Ballantyne CM Abate N Yuan Z King TR Palmisano J Dose-comparison study of the combination of ezetimibe and simvastatin (Vytorin) versus atorvastatin in patients with hypercholesterolemia: the Vytorin Versus Atorvastatin (VYVA) study Am Heart J 2005 149 464 473 15864235 10.1016/j.ahj.2004.11.023
Keech AC the FIELD Study Investigators The need for a large-scale trial of fibrate therapy in diabetes: the rationale and design of the Fenofibrate Intervention and Event Lowering in Diabetes (FIELD) study. ISRCTN64783481 Cardiovasc Diabetol 2004 3 9 15571637 10.1186/1475-2840-3-9
Song F Altman DG Glenny AM Deeks JJ Validity of indirect comparison for estimating efficacy of competing interventions: empirical evidence from published meta-analyses BMJ 2003 326 472 12609941 10.1136/bmj.326.7387.472
Tu JV Naylor CD Choosing among drugs of different price for similar indications Can J Cardiol 1998 14 349 351 9551028
Mukhtar RY Reid J Reckless JP Pitavastatin Int J Clin Pract 2005 59 239 252 15854203 10.1111/j.1742-1241.2005.00461.x
Jones PH Davidson MH Stein EA Bays HE McKenney JM Miller E Cain VA Blasetto JW for the STELLAR Study Group Comparison of the efficacy and safety of rosuvastatin versus atorvastatin, simvastatin and pravastatin across doses (STELLAR) trial Am J Cardiol 2003 92 152 160 12860216 10.1016/S0002-9149(03)00530-7
| 15941471 | PMC1156921 | CC BY | 2021-01-04 16:25:03 | no | Cardiovasc Diabetol. 2005 Jun 7; 4:8 | utf-8 | Cardiovasc Diabetol | 2,005 | 10.1186/1475-2840-4-8 | oa_comm |
==== Front
Cardiovasc UltrasoundCardiovascular Ultrasound1476-7120BioMed Central London 1476-7120-3-141591889310.1186/1476-7120-3-14Case ReportLeft ventricular apical thrombus after systemic thrombolysis with recombinant tissue plasminogen activator in a patient with acute ischemic stroke Doepp Florian [email protected] Wasiem [email protected] Stephan J [email protected] Gert [email protected] Adrian C [email protected] Department of Neurology, University Hospital Charité, Berlin, Germany2 Department of Cardiology, University Hospital Charité, Berlin, Germany2005 26 5 2005 3 14 14 7 4 2005 26 5 2005 Copyright © 2005 Doepp et al; licensee BioMed Central Ltd.2005Doepp et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Thrombolysis with recombinant tissue plasminogen activator (rtPA) is an established treatment in acute stroke. To prevent rethrombosis after rtPA therapy, secondary anticoagulation with heparin is commonly performed. However, the recommended time-point and extent of heparin treatment vary and are not well investigated.
Case presentation
We report a 61-year-old man who developed an acute global aphasia and right-sided hemiparesis. Cranial CT was normal and systemic thrombolytic therapy with tPA was started 120 minutes after symptom onset. Low-dose subcutaneous heparin treatment was initiated 24 hours later. Transthoracic echocardiography (TTE) 12 hours after admission showed slightly reduced left ventricular ejection fraction (LVEF) but was otherwise normal. 48 hours later the patient suddenly deteriorated with clinical signs of dyspnea and tachycardia. TTE revelead a large left ventricular apical thrombus as well as a reduction of LVEF to 20 %. Serial further TTE investigations demonstrated a complete resolution of the thrombus and normalisation of LVEF within two days.
Conclusion
Our case demonstrates an intracardiac thrombus formation following rtPA treatment of acute stroke, probably caused by secondary hypercoagulability. Rethrombosis or new thrombus formation might be an underestimated complication of rtPA therapy and potentially explain cases of secondary stroke progression.
thrombolysishypercoagulabilityischemic strokecardiac thrombus
==== Body
Background
Recombinant tissue plasminogen activator (rtPA) is an approved treatment option for acute ischemic stroke within three hours of symptoms onset. It is well known that reperfusion of ischaemic brain tissue, when performed in a timely manner, improves clinical outcome [1,2]. However, there is evidence, that rtPA induces a secondary activation of coagulation cascades and inhibition of endogenous fibrinolysis [3,4]. These hemostatic changes might contribute to severe complications of rtPA treatment such as vessel reocclusion or recurrent stroke. Therefore, an adjuvant anticoagulation with heparin is performed to prevent reocclusion in myocardial infarction [5] and after intra-arterial thrombolysis of the basilar artery [6]. However, no studies or consensus exist regarding early anticoagulation after systemic thrombolysis in acute stroke.
We present a patient with acute ischaemic stroke who developed a transient left ventricular apical thrombus after systemic thrombolysis.
Case presentation
A 61-year-old man was admitted to the Emergency Department immediately after he experienced global aphasia and right-sided hemiparesis. The clinical diagnosis of acute cerebral ischaemia in the middle cerebral artery (MCA) territory was made by a neurologist. A cranial CT-scan showed no early signs of cerebral infarct and no intracranial bleeding. Atrial fibrillation was recognised on the initial ECG and an cerebral infarct of cardiac embolic origin was assumed. The patient had no history of stroke but suffered from coronary heart disease and sustained an anterior cardiac wall infarction in the past. Further known vascular risk factors were hypertension and smoking. Even so, the patient took no medication at the time of admission.
Thrombolysis with 70 mg rt-PA (0.9 mg/kg) was started two hours after the onset of symptoms. The neurological status during and after rt-PA treatment remained unchanged. Transthoracic echocardiography (TTE) performed 12 hours after initiation of thrombolytic therapy revealed a moderately reduced left ventricular systolic function (ejection fraction: 41%), left ventricular regional dyssynergy, and no intracardiac thrombus.
The D-Dimer antigen plasma concentration as a marker of coagulation activation was raised 24 hours after rt-PA treatment and showed a tendency to normalisation during the following days.
Extracranial Duplexsonography revealed moderate arteriosclerotic plaque formation in the carotid bulb on both sides but no hemodynamic stenosis of the carotid and vertebral arteries. Using transcranial duplexsonography a severe stenosis was found in the proximal segment of the left middle cerebral artery. No diabetes or hyperlipidaemia were diagnosed. Cerebral MRI, twenty-four hours after lysis revealed a cortical infarct with a clinically asymptomatic haemorrhagic transformation in the left-sided MCA territory. Antithrombotic treatment with subcutaneous low-dose heparin (Dalteparin 2/d) was initiated 24 hours after thrombolysis. The fluid balance was continuously monitored and kept net 0 ml/24 hours to avoid fluid overload.
On the 3rd day after admission, the patient developed acute dyspnea, tachypnea, tachycardia, and an elevated blood pressure. Acute myocardial infarction was excluded on the basis of ECG and blood tests (CK, CK-MB, Troponin were normal). A CT-scan detected no pulmonary embolism, but massive bilateral pleural effusion. TTE revealed an apical thrombus measuring 2.0 × 2.0 cm in the left ventricle (figure 1) with left ventricular ejection fraction reduced to 20%. The thrombus, was fixed to the hypokinetic anteroseptal wall. Serial TTE performed two days later revealed complete resolution of the thrombus and hypokinesia; with a left ventricular ejection fraction of 40%. In the mean time the patient recovered from the acute cardiac decompensation. The neurological clinical picture remained stable during the acute phase with a slight improvement of aphasia and hemiparesis within the following 5 weeks.
Figure 1 Cardiac duplex ultrasound image (4-chamber view), obtained 2 days after thrombolysis showing the thrombus attached to left ventricular apex (arrow).
Discussion
Systemic rt-PA thrombolysis in acute stroke has been implemented into daily clinical practice during the last decade. The treatment within three hours of stroke appears to be effective in reducing the neurological deficit [1,7]. However, the use of rtPA therapy remains limited due to the narrow time-to-treatment windows and the potential complications of intracranial haemorrhage [7]. The optimal time-point of anticoagulation after systemic thrombolysis is unclear due to a lack of evidence. In accordance with the NINDS-protocol anticoagulation is interdicted within 24 hours after thrombolysis [1]. Also, latest guidelines for early stroke management recommend no antithrombotic therapy within 24 hours after application of rt-PA although there is increasing evidence for vessel reocclusion in about one third of the patients [8]. However, in clinical practice the use of anticoagulative agents seems to be much more heterogeneous [9].
Some authors have reported a significantly higher incidence of parenchymal haematomas if thrombolysis was immediately followed by intravenous or subcutaneous heparin administration [10]. However, recent analyses have shown that full-dose intravenous anticoagulant treatment within 24 hours does not increase the incidence of parenchymal hemorrhage [9].
rt-PA has a short biological half-life of 8–12 minutes but alters the physiological balance between coagulation and anticoagulation for a longer time period [11]. After discontinuation of rtPA infusion several mechanisms, potentially leading to a secondary hypercoagulability status have been discussed. For instance, the activation of plasminogen activator inhibitor-1, resulting in suppression of endogenous fibrinolysis [4], the increase of thrombin generation and activity [3], the increase of thrombin-antithrombin-III-complex levels [12] or the induction of hypoperfusion in ischaemic brain tissue [11]. The procoagulant response to rt-PA has been shown to persist up to 72 hours [3].
Otherwise, clinical studies of fibrinolytic therapy in myocardial infarction show, that early heparin treatment starting immediately after thrombolysis significantly decreases the risk of cardiac vessel reocclusion [5]. The capacity to attenuate the increased coagulation activity seems to be similar, regardless if low-molecular-weight or unfractionated heparin is used [13]. In stroke patients with an occlusion in the posterior cerebral circulation undergoing intra-arterial thrombolysis, heparin treatment is also known to reduce the rate of early reocclusion [6].
In our patient, treated with low dose heparin, an apical thrombus developed within 72 hours after thrombolytic therapy, strongly suggesting a causal relationship of secondary hypercoagulability and thrombus formation. The patient's history and risk profile suggests an increased risk for cardiac events [14]. We postulate, that he developed an acute coronary syndrome (dyspnea and tachycardia) on the basis of the pre-existing cardiac disease with kinetic disturbance which subsequently enabled the formation of a ventricular thrombus – promoted by the risk factors hypercoagulability, atrial fibrillation and previous myocardial infarction [15]. The transient increase of D-Dimer probably indicates the state of hypercoagulability after rt-PA treatment [3]. The low-dose heparin therapy in our patient was not sufficient to prevent thrombus development although a number of studies have shown that sufficient systemic anticoagulation with heparin is able to do so [16].
Coronary angiography was not performed, considering the cerebral state of the patient and the observation that repeated ECGs were normal and cardiac enzymes not elevated. The clinical status stabilised spontaneously and the cardiac thrombus resolved within two days without initiation of a specific therapy, further supporting the hypothesis of a temporary status of hypercoagulation after thrombolysis.
Another hypothesis for the temporary cardiac detoriation in our patient could be an acute neurogenic stunned myocardium. This phenomenon is described as sudden, reversible left ventricular dysfunction with abnormal left ventricular wall motion and reduced ejection fraction. Levels of creatine kinase MB and troponin may be elevated and ECG alterations as depression or elevation of the ST segment or T wave inversion can be observed. However, acute neurogenic myocardial stunning has so far only been reported after subarachnoid hemorrhage [17-20] and in isolated cases of subdural hematoma [21] and Guillain-Barré syndrome [22]. We found no evidence in the literature for acute neurogenic myocardial stunning after stroke. In addition, our patient suffered from coronary artery disease whereas neurogenic myocardial stunning has been reported in patients without cardiac disease. Finally, the cardiac failure in our patient occurred on the 3rd day and not directy after stroke which makes the diagnosis of neurogenic stunned myocardium further unlikely [17,20,21].
In summary, our case demonstrates an intracardiac thrombus formation following rtPA treatment of acute stroke, probably caused by secondary hypercoagulability. Rethrombosis or new thrombus formation might be an underestimated complication of rtPA therapy and potentially explain cases of secondary stroke progression. Early systemic anticoagulation with heparin might reduce the risk of rethrombosis but also increase the risk of a bleeding complication. Systematic studies concerning the incidence of thrombus formation after rtPA therapy and the effects of different post-thrombolysis anticoagulation strategies are required to assess the clinical relevance of the discussed secondary hypercoagulability. Closed echocardiographic monitoring in stroke patients treated with systemic thrombolysis might be useful for early detection of the described potential cardiac complications especially because repeated measurement of ECG and cardiac enzymes alone might fail.
==== Refs
The National Institute of Neurological Diseases and Stroke rt-PA Stroke Study Group Tissue plasminogen activator for ischaemic stroke N Engl J Med 1995 333 1581 1587 7477192 10.1056/NEJM199512143332401
Hacke W Kaste M Fieschi C Toni D Lesaffre E von Kummer R Boysen G Bluhmki E Hoxter G Mahagne MH for the ECASS Group Intravenous thrombolysis with recombinant tissue plasminogen activator for acute hemispheric stroke: The European Cooperative Acute Stroke Study JAMA 1995 274 1017 1025 7563451 10.1001/jama.274.13.1017
Fassbender K Dempfle CE Mielke O Schwartz A Daffertshofer M Eschenfelder C Dollman M Hennerici M Changes in Coagulation and fibrinolysis markers in acute ischemic stroke treated with recombinant plasminogen activator Stroke 1999 30 2101 2104 10512913
Paganelli F Alessi MC Morange P Maixent JM Levy S Vague IJ Relationship of plasminogen activator inhibitor-1 levels following thrombolytic therapy with rt-PA as compared to streptokinase and patency of infarct related coronary artery Thromb Haemost 1999 82 104 108 10456462
Frostfeldt G Gustavsson G Lindahl B Nygren A Wallentin L Influence on coagulation activity by subcutaneous LMW heparin as an adjuvant treatment to fibrinolysis in acute myocardial infarction Thrombosis Research 2002 105 193 199 11927123 10.1016/S0049-3848(02)00017-8
Kuker W Friese S Vogel W Schmidt F Weller M Incomplete resolution of basilar artery occlusion after intra-arterial thrombolysis: abciximab and heparin prevent early rethrombosis Cerebrovasc Dis 2000 10 484 486 11070384 10.1159/000016115
Wardlaw JM Zoppo G Yamaguchi T Berge E Thrombolysis for acute ischaemic stroke Cochrane Databaes Syst Rev 2003 CDOOO213
Adams H Adams R Del Zoppo G Goldstein LB Guidelines for the Early Management of Patients With Ischemic Stroke: 2005 Guidelines Update: A Scientific Statement From the Stroke Council of the American Heart Association/American Stroke Association Stroke 2005 36 916 921 15800252 10.1161/01.STR.0000163257.66207.2d
Schenkel J Weimar C Knoll T Haberl RL Busse O Hamann GF Koennecke HC Diener HC R1-Systemic Thrombolysis in German Stroke Units. The experience from the German Stroke Data Bank J Neurol 2003 250 320 324 12638023 10.1007/s00415-003-1004-5
Trouillas P Nighoghossian N Derex L Adeleine P Honnorat J Neuschwander P Riche G Getenet JC Li W Froment JC Turjman F Malicier D Fournier G Gabry AL Ledoux X Berthezene Y Ffrench P Dechavanne M Thrombolysis with intravenous rtPA in a series of 100 cases of acute carotid territory stroke. Determination of etiological, topographic, and radiological outcome factors Stroke 1998 29 2529 2540 9836764
Kilic E Bähr M Hermann DM Effect of recombinant tissue plasminogen activator after intraluminal thread occlusion in mice. Role of hemodynamic alterations Stroke 2001 32 2641 2647 11692029
Maki A Shirato C Ohguchi M Aoki N Ochiai I Ishikawa K Mechanism of re-thrombosis after thrombolytic therapy: angiographic findings and investigation of the coagulation system in dogs Angiology 1998 49 447 453 9631890
Nilsen DW Goransson L Larsen AI Hetland O Kierulf P Systemic thrombin generation and activity resistant to low molecular weight heparin administered prior to streptokinase in patients with acute myocardial infarction Thromb Haemost 1997 77 57 61 9031450
Adams RJ Chimowitz MI Alpert JS Awad IA Cerqueria MD Fayad P Taubert KA Coronary risk evaluation in patients with transient ischemic attack and ischemic stroke: a scientific statement for healthcare professionals from the Stroke Council and the Council on Clinical Cardiology of the American Heart Association/American Stroke Association Circulation 2003 108 1278 1290 12963684 10.1161/01.CIR.0000090444.87006.CF
Greaves SC Zhi G Lee RT Solomon SD MacFadyen J Rapaport E Menapace FJ Rouleau JL Pfeffer MA Incidence and natural history of left ventricular thrombus following anterior wall acute myocardial infarction Am J Cardiol 1997 80 442 448 9285655 10.1016/S0002-9149(97)00392-5
Vaitkus PT Barnathan ES Embolic potential, prevention and management of mural thrombus complicating anterior myocardial infarction: a meta-analysis J Am Coll Cardiol 1993 22 1004 1009 8409034
Kono Left ventricular wall motion abnormalities in patients with subarachnoid hemorrhage: neurogenic stunned myocardium J Am Coll Cardiol 1994 24 636 640 8077532
Robert Evidence of stunned myocardium in humans: a 2001 update Coron Artery Dis 2001 12 349 356 11491199 10.1097/00019501-200108000-00003
Bulsara Use of the peak troponin value to differentiate myocardial infarction from reversible neurogenic ventricular dysfunction associated with aneurysmal subarachnoid hemorrhage J Neurosurg 2003 98 524 528 12650423
Jain Management of patients with stunned myocardium associated with subarachnoid hemorrhage Am J Neuroradiol 2004 25 126 129 14729541
Ohtsuka Neurogenic stunned myocardium Circulation 2000 101 2122 2124 10790357
Bernstein R Mayer SA Magnano A Neurogenic stunned myocardium in Guillaine-Barre syndrome Neurology 2000 54 759 762 10680822
| 15918893 | PMC1156922 | CC BY | 2021-01-04 16:38:31 | no | Cardiovasc Ultrasound. 2005 May 26; 3:14 | utf-8 | Cardiovasc Ultrasound | 2,005 | 10.1186/1476-7120-3-14 | oa_comm |
==== Front
Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health : CP & EMH1745-0179BioMed Central 1745-0179-1-61596705610.1186/1745-0179-1-6ResearchHeterozygous β-thalassaemia as a susceptibility factor in mood disorders: excessive prevalence in bipolar patients Bocchetta Alberto [email protected] Bernard B. Brodie Department of Neurosciences, University of Cagliari, Italy2005 1 6 2005 1 6 6 14 4 2005 1 6 2005 Copyright ©2005 Bocchetta; licensee BioMed Central Ltd.2005Bocchetta; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Background
Previous preliminary reports have suggested potential interactions between microcytic anaemia and mood disorders. In particular, heterozygous β-thalassaemia has been implicated in the bipolar spectrum. This study surveyed relevant haematological parameters in a large sample of psychiatric outpatients with the aim of clarifying previous observations.
Methods
Mean Corpuscular Volume (MCV) was analysed in 1014 consecutive patients diagnosed according to modified Research Diagnostic Criteria (RDC). Haemoglobin electrophoresis and/or chromatography was performed in blood samples from 143 patients with reduced MCV. Prevalence of heterozygous β-thalassaemia was estimated based on the rates of patients with reduced MCV and increased haemoglobin A2 concentration.
Results
MCV lower than 80 μ3 was found in greater proportions among bipolar compared with the remaining RDC subgroups (183/732 = 25.0% versus 51/282 = 18.1%; p = 0.02; relative risk = 1.38; Fisher's exact test). This difference can mainly be attributed to heterozygous β-thalassaemia, the estimated prevalence of which was 16.4% among bipolar and 9.9% among non-bipolar subgroups (p = 0.01; relative risk = 1.65).
Conclusion
The results are consistent with the hypothesis that heterozygous β-thalassaemia might play a role as a susceptibility factor in bipolar spectrum disorders in specific populations.
==== Body
Background
Our interest in potential interactions between mood disorders of the bipolar spectrum and blood conditions typical of the Mediterranean area, such as glucose-6-phosphate dehydrogenase (G6PD) deficiency and heterozygous β-thalassaemia, was first aroused in the 1980 s. At that time, chromosomes Xq28 and 11p15, where the G6PD gene and the β-globin gene cluster are located, became very popular in psychiatric genetics following the reports of positive linkage with manic-depression [1-3].
We subsequently surveyed G6PD deficiency and heterozygous β-thalassaemia as phenotypic markers to be used in linkage studies [4]. Unexpectedly, we found excessive proportions of the two markers among bipolar patients, especially those with psychotic symptoms [5-7]. Among potential mechanisms underlying the observed associations, we considered linkage disequilibrium between the markers and mood disorder genes located in Xq28 and 11p15. This hypothesis, which at first was most intriguing, weakened considerably when subsequent studies reported diminished evidence of Xq28 or 11p15 linkage [8,9]. Over the following years, the use of phenotypic markers in psychiatric genetics vanished due to the advent of molecular DNA markers. Nevertheless, with regard to heterozygous β-thalassaemia, our interest was periodically revived by the appearance of relevant reports. On the one hand, co-occurrence of the thalassaemic trait and bipolar disorder was reported in several cases from various countries, including Japan, where thalassaemia is very rare [10-13]. On the other hand, a psychiatric population surveyed in Argentina revealed excessive prevalence of microcytic anaemia among patients suffering from recurrent mood disorder [14]. Given the latter series of intriguing reports, an extension of our previous surveys was warranted: this is an updated account of haematological parameters regarding 1014 patients.
Methods
Outpatients consecutively admitted to the Department of Neuroscience, University of Cagliari, between 1980 and 2000 were studied. The department is one of the reference centres for the management of lithium and related treatments of mood disorders in southern Sardinia. On admission, detailed information was obtained from the patient and from other available sources (such as relatives and referring clinicians) concerning demographic characteristics and lifetime medical and psychiatric history. Moreover, all available records regarding previous prescriptions or hospitalisations were examined. Psychiatric diagnosis was made according to modified Research Diagnostic Criteria (RDC) [15,7]. Routine laboratory tests, including blood count and serum iron concentration were available for 1014 patients. All subjects gave their informed consent to participate in the study.
Additional procedures were used aiming at the identification of heterozygous β-thalassaemia, which is ordinarily diagnosed based on reduced MCV, increased haemoglobin A2 (HbA2), and normal serum iron concentration. In an initial survey [5], haemoglobin electrophoresis was systematically performed in samples from 180 patients who had attended the centre between April 3rd and May 3rd, 1989. Thereafter, MCV lower than 80 μ3 was chosen as the cut-off for the performing of haemoglobin electrophoresis and/or chromatography (98% of cases with abnormally high HbA2 concentration from the initial survey had MCV<80 μ3). Overall, electrophoresis and/or chromatography were obtained for 143 patients with MCV<80 μ3. Of the latter, 62% had increased HbA2 (heterozygous β-thalassaemia); 22% had normal HbA2 and normal serum iron (in Sardinia, this pattern has been characterised molecularly as heterozygous α-thalassaemia in the large majority of cases [16]). In the remaining 16%, HbA2 concentration was normal but serum iron was reduced, therefore heterozygous β-thalassaemia could not be ruled out (iron-deficiency may maintain HbA2 within the normal range).
Results
Overall, MCV<80 μ3 was found in 23.1% of patients (Table 1). Greater proportions were found among subgroups with bipolar course compared with the remaining subgroups (p = 0.02 at the Fisher's exact test; 183/732 = 25.0% versus 51/282 = 18.1%; relative risk = 1.38). The highest proportion was found in Manic Schizoaffective disorder, predominantly affective subtype (74/219 = 33.8%).
Table 1 Reduced Mean Corpuscular Volume (MCV) and Heterozygous β-Thalassaemia in 1014 Psychiatric Outpatients
RDC Diagnosis Number of Patients MCV<80 μ3 N (%) Heterozygous β-Thalassaemiaa
Definite Probable Total (%)
Manic Schizoaffective 288 88 (30.6) 42 16 58 (20.1)
Bipolar with Mania 269 59 (21.9) 17 21 38 (14.1)
Bipolar with Hypomania 175 36 (20.6) 16 8 24 (13.7)
Depressive Schizoaffective 96 19 (19.8) 6 4 10 (10.4)
Recurrent Major Depression 124 24 (19.4) 7 7 14 (11.3)
Other (Schizophrenia, Minor Depression, etc) 62 8 (12.9) 1 3 4 (6.5)
TOTAL 1014 234 (23.1) 89 59 148 (14.6)
aRates of heterozygous β-thalassaemia are estimated based on the number of definite cases (MCV <80 μ3 and increased HbA2) plus the number of probable cases (62 % of patients with MCV <80 μ3 but unknown HbA2).
Heterozygous β-thalassaemia (increased HbA2 concentration) accounted for the majority of patients with MCV<80 μ3. Rates did not vary between the initial survey (27/42 = 64%) and subsequent cases who underwent HbA2 analysis (62/101 = 62%). If similar rates were assumed regarding the remaining 91 patients with MCV<80 μ3 but unknown HbA2 concentration, prevalence of heterozygous β-thalassaemia in the overall sample could be estimated as at least 16.4% among bipolar and 9.9% among non-bipolar subgroups. A simulation of the Fisher's exact test provided a bipolar/non-bipolar relative risk of 1.65 with a p value of 0.01. Estimated rates of heterozygous β-thalassaemia across psychiatric diagnoses are shown in Table 1. The subgroup of patients with Manic Schizoaffective disorder, predominantly affective subtype (not shown), had the highest estimated rate of heterozygous β-thalassaemia (22%).
The distribution of patients with MCV<80 μ3 and normal HbA2 concentration did not vary across psychiatric diagnoses (not shown).
Discussion
Reduced Mean Corpuscular Volume and Mood Disorder
In this survey, an impressive 24% of the 952 patients suffering from major mood disorders had an MCV lower than 80 μ3. Rates of reduced MCV were lower (13%) in the small subgroup of patients (N = 62) with other psychiatric diagnoses. The direction of results strikingly parallels that from an Argentinean survey [14], which reported a 17.7 % prevalence of microcytic anaemia among 79 patients with recurrent mood disorders compared with 8.0 % prevalence among 25 patients with other psychiatric diagnoses (schizophrenia, anxiety, etc). The lack of similar studies in the literature is surprising given the number of symptoms shared between anaemia and depression (fatigue, asthenia, somatic complaints, etc). On the contrary, many studies have investigated macrocytic anaemia and the effects of vitamin B12 deficiency in the CNS.
Prevalence of reduced MCV appeared to follow a hierarchy across the spectrum of mood disorders, according to the presence of psychotic and manic features. Since severity of symptoms may reflect an increased underlying biological/genetic load, the hypothesis of microcytic anaemia as a susceptibility factor is warranted.
Heterozygous β-Thalassaemia and Mood Disorder
In this survey, variation in prevalence of reduced MCV by psychiatric diagnosis was mostly attributable to heterozygous β-thalassaemia. Data substantially confirmed those from our initial 1989 survey [5], in which haemoglobin electrophoresis was systematically performed in blood samples from 180 patients. The highest rates of heterozygous β-thalassaemia had been found in Manic Schizoaffective Disorder and an apparent bipolar/non-bipolar distinction had already been signalled. In this extended sample, the estimated rates were 16.4 % in bipolar and 9.9 % in non-bipolar subgroups. Results paralleled those from the above mentioned Argentinean survey [14], where heterozygous β-thalassaemia was diagnosed in four of the 51 bipolar patients (7.8 %) and in only one of the 53 remaining cases (1.9 %). Prevalence in the general population in Argentina has been estimated at approximately 5% [14], even if geographical variation is assumed reflecting non-homogeneous ethnic distribution. In Sardinia, 12.6 % is the rate found in the general population [17], but a wide variation has been reported following once malarial morbidity [18].
Potential geographical stratification is a relevant point in the interpretation of our findings: in fact, if mood disorder with psychotic and/or manic features is more prevalent for unknown reasons in once malarial areas in Sardinia, a spurious association with heterozygous β-thalassaemia might have derived. The latter possibility is however challenged by the similarity in results between two independent studies (ours and the Argentinean one). Another approach capable of ruling out a potential geographical bias is the study of segregation between marker and disease within families, given the shared origin of relatives. Indeed, we found a higher than expected concordance between mood disorder and the blood condition within families of our patients with heterozygous β-thalassaemia (to be reported in a different account). A tendency towards familial co-segregation was also inferable by pooling data from published case reports of bipolar disorder and heterozygous β-thalassaemia ascertained in Canada [10], Eire [11], Japan [12], and Australia [13].
In the presence of a true genetic association, different mechanisms may be hypothesised, including linkage disequilibrium between the globin gene and a mood disorder gene closely located in chromosome 11p15 or a direct effect of heterozygous β-thalassaemia in clinical presentation or severity of mood disorder.
Whatever the mechanism, the relevance of heterozygous β-thalassaemia as a susceptibility factor in mood disorder can be inferred from two recently published studies from India [19] and Greece [20]. Since the context was that of studies of psychosocial adjustment of parents of (or at risk of bearing) children with Cooley's anaemia, attention was mostly directed to current depression and anxiety (rather than to history of bipolar or schizoaffective disorder, if any). Moreover, the potential role of the mild or silent blood condition in parents was not taken into account. In the Indian study [19], depression was diagnosed in 70% of parents (19 mothers and 11 fathers) accompanying their thalassaemic child for blood transfusion. Unfortunately no group of parents of children suffering from other severe chronic diseases was included for comparison.
The Greek study [20] reported clinically relevant depression in 10.1% of 159 women with heterozygous β-thalassaemia undergoing chorionic villus sampling due to increased possibility of carrying an embryo with Cooley's anaemia (spouses were β-thalassaemic carriers too). As comparison groups, 150 women undergoing karyotyping for risk of trisomy and 309 undergoing a routine first trimester scan were studied. Depression was found in 4.7 % and 0.3 %, respectively.
In another Greek study [21], more psychiatric disorders, including a broad range of DSM-III-R diagnoses, were found among 71 siblings of subjects with Cooley's anaemia compared with 71 control subjects. In the age group from 10 to 19 years, the difference was significant (61% compared with 37%). Even if the latter data are intriguing, no conclusion can be drawn regarding the role of heterozygous β-thalassaemia in this sample, because haematological data of participants were not envisaged. However, since the expected proportion of thalassaemic carriers among siblings of subjects with Cooley's anaemia is 2/3, similar samples would be very informative. Inclusion of older age groups, when mood disorders are better characterised, would be preferable.
Conclusion
The results from this survey and the series of intriguing literature reports are consistent with the hypothesis that heterozygous β-thalassaemia might play a role as a susceptibility factor in mood disorders, particularly of the bipolar spectrum. Since the evidence cannot be considered established, further studies are warranted. Potential approaches include: a) replication surveys of microcytic anaemia and the thalassaemic trait in psychiatric populations from high prevalence areas; b) surveys of bipolar spectrum disorders among subjects with an already established diagnosis of heterozygous β-thalassaemia (such as relatives of patients with Cooley's anaemia or participants in specific screening programs); c) family studies of patients with bipolar disorder and an established thalassaemic trait.
Acknowledgements
The Author is grateful to Professor Antonio Cao for his advice regarding the haematological diagnoses and Ms Anne Farmer for language editing of the manuscript.
==== Refs
Mendlewicz J Linkowski P Willmotte J Linkage between glucose-6-phosphate dehydrogenase deficiency and manic-depressive psychosis Br J Psychiatry 1980 137 337 342 7448473
Baron M Risch N Hamburger R Mandel B Kushner S Newman M Drumer D Belmaker RH Genetic linkage between X-chromosome markers and bipolar affective illness Nature 1987 326 289 292 10.1038/326289a0 3493438
Egeland JA Gerhard DS Pauls DL Sussex JN Kidd KK Allen CR Hostetter AM Housman DE Bipolar affective disorder linked to DNA markers on chromosome 11 Nature 1987 325 783 787 10.1038/325783a0 2881209
Del Zompo M Bocchetta A Goldin LR Corsini GU Linkage between X-chromosome markers and manic-depressive illness Acta Psychiatr Scand 1984 70 282 287 6333780
Bocchetta A Del Zompo M Bipolar affective disorder and heterozygous β-thalassaemia Am J Psychiatry 1990 147 1094 2375451
Bocchetta A Severino G Bernardi F Del Zompo M Associations between heterozygous β-thalassaemia, glucose-6-phosphate dehydrogenase deficiency, and affective disorders of the bipolar spectrum It J Psychiat Behav Sci 1993 3 107 110
Bocchetta A Piccardi MP Del Zompo M Is bipolar disorder linked to Xq28? Nat Genet 1994 6 224 10.1038/ng0394-224 8012380
Baron M Freimer NF Risch N Lerer B Alexander JR Straub RE Asokan S Das K Peterson A Amos J Diminished support for linkage between manic-depressive illness and X-chromosome markers in three Israeli pedigrees Nat Genet 1993 3 49 55 10.1038/ng0193-49 8490654
Kelsoe JR Ginns EI Egeland JA Gerhard DS Goldstein AM Bale SJ Pauls DL Long RT Kidd KK Conte G Housman DE Paul SM Re-evaluation of the linkage relationship between chromosome 11p loci and the gene for bipolar affective disorder in the Old Order Amish Nature 1989 342 238 243 10.1038/342238a0 2682265
Joffe RT Horvath Z Tarvydas I Bipolar affective disorder and thalassemia minor Am J Psychiatry 1986 143 933 3717440
Singh AN Maguire J Bipolar affective disorder and thalassaemia minor. A genetic linkage? Br J Psychiatry 1988 152 581 3167431
Harada H Nakajima T Inazawa J Abe T Bipolar affective disorder associated with β-thalassaemia minor Biol Psychiatry 1995 37 477 480 10.1016/0006-3223(94)00172-Y 7786963
Brett A Dunn P Bipolar affective disorder and thalassaemia minor Aust N Z J Psychiatry 1998 32 141 9565201
Ciprian-Ollivier J Albin J de Lopez Mato AM Talasemia y enfermedad bipolar Alcmeon 1991 4 547 555
Spitzer RL Endicott J Robins E Research Diagnostic Criteria for a selected group of functional disorders, 3rd edition 1977 New York: New York State Psychiatric Institute
Galanello R Sollaino C Paglietti E Barella S Perra C Doneddu I Pirroni MG Maccioni L Cao A Alpha-thalassamia carrier identification by DNA analysis in the screening for thalassemia Am J Hematol 1998 59 273 278 10.1002/(SICI)1096-8652(199812)59:4<273::AID-AJH2>3.0.CO;2-3 9840907
Cao A Galanello R Furbetta M Muroni PP Garbato L Rosatelli C Scalas MT Addis M Ruggeri R Maccioni L Melis MA Thalassaemia types and their incidence in Sardinia J Med Genet 1978 15 443 447 745215
Siniscalco M Bernini L Filippi G Latte B Meera Khan P Piomelli S Rattazzi M Population genetics of haemoglobin variants, thalassaemia and glucose-6-phosphate dehydrogenase deficiency, with particular reference to the malaria hypothesis Bull World Health Organ 1966 34 379 393 5296398
Rao P Pradhan PV Shah H Psychopathology and coping in parents of chronically ill children Indian J Pediatr 2004 71 695 699 15345869
Lykeridou K Daskalakis G Papadatou D Vaslamatzis G Dafni O Antsaklis A Dysthymic reactions of women undergoing chorionic villus sampling for prenatal diagnosis of hemoglobinopathies or karyotyping Fetal Diagn Ther 2004 19 149 154 10.1159/000075140 14764960
Labropoulou S Beratis S Psychosocial adjustment of thalassaemic children's siblings J Psychosom Res 1995 39 911 919 10.1016/0022-3999(95)00061-4 8636923
| 15967056 | PMC1156923 | CC BY | 2021-01-04 18:01:02 | no | Clin Pract Epidemiol Ment Health. 2005 Jun 1; 1:6 | utf-8 | Clin Pract Epidemiol Ment Health | 2,005 | 10.1186/1745-0179-1-6 | oa_comm |
==== Front
Comp HepatolComparative Hepatology1476-5926BioMed Central London 1476-5926-4-41587173410.1186/1476-5926-4-4ResearchEffects of redox cycling compounds on DT diaphorase activity in the liver of rainbow trout (Oncorhynchus mykiss) Sturve Joachim [email protected] Eiríkur [email protected]örlin Lars [email protected] Department of Zoology, Zoophysiology, Göteborg University, Box 463, 405 30, Göteborg, Sweden2005 4 5 2005 4 4 4 22 12 2004 4 5 2005 Copyright © 2005 Sturve et al; licensee BioMed Central Ltd.2005Sturve et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
DT diaphorase (DTD; NAD(P)H:quinone oxidoreductase; EC 1.6.99.2) catalyses the two electron reduction of quinones, thus preventing redox cycling and consequently quinone dependent production of reactive oxygen species. In rat and mouse, a wide range of chemicals including polyaromatic hydrocarbons, azo dyes and quinones induces DTD. Bifunctional compounds, such as β-naphthoflavone (β-NF) and benzo(a)pyrene (B(a)P), induce DTD together with CYP1A and phase II enzymes by a mechanism involving the aryl hydrocarbon receptor (AHR). Monofunctional induction of DTD is mediated through the antioxidant response element and does not lead to the induction of AHR dependent enzymes, such as CYP1A. The aim of this study was to investigate the effects of prooxidants (both bifunctional and monofunctional) on the activity of hepatic DTD in rainbow trout (Oncorhynchus mykiss) in order to evaluate DTD suitability as a biomarker. We also investigated the effect of β-NF on hepatic DTD activity in perch (Perca fluviatilis), shorthorn sculpin (Myoxocephalus scorpius), eelpout (Zoarces viviparus), brown trout (Salmo trutta) and carp (Cyprinus carpio). In addition, the effect of short term exposure to prooxidants on catalase activity was investigated.
Results
In rainbow trout, hepatic DTD activity is induced by the bifunctional AHR agonists β-NF and B(a)P and the monofunctional inducers naphthazarin, menadione and paraquat. Although exposure to both B(a)P and β-NF led to a strong 7-ethoxyresorufin-O-deethylase (EROD) induction, none of the monofunctional compounds affected the rainbow trout EROD activity. DTD was not induced by β-NF in any of the other fish species. Much higher DTD activities were observed in rainbow trout compared to the other fish species. Catalase activity was less responsive to short term exposure to prooxidants compared to DTD.
Conclusion
Since rainbow trout hepatic DTD activity is inducible by both monofunctional and bifunctional inducers, it is suggested that rainbow trout DTD may be regulated by the same mechanisms, as in mammals. The fact that DTD is inducible in rainbow trout suggests that the enzyme may be suitable as a part of a biomarker battery when rainbow trout is used in environmental studies. It appears as if DTD activity in rainbow trout is higher and inducible compared to the other fish species studied.
==== Body
Background
The aquatic environment is exposed to a great number of pollutants. Effluents from industries and sewage treatment plants as well as drainage from urban and agricultural areas contain pollutants that may damage aquatic life. A large part of these compounds exert their toxic effect by generating reactive oxygen species (ROS), causing oxidative stress [1]. Compounds such as quinones, certain polycyclic aromatic hydrocarbons (PAH) metabolites and bipyridils generate ROS through their ability to redox cycle, a process where an enzymatic one electron reduction of the parent compound is followed by an autooxidation in the presence of molecular oxygen [2]. In this reaction, the oxygen will be reduced to a superoxide ion that can lead to the formation of other ROS such as hydrogen peroxide (H2O2) and hydroxyl radicals [3]. ROS causes cell injury by oxidizing lipids, proteins and DNA leading to membrane damage, enzyme malfunction and/or tumor formation. The cell has evolved an antioxidant defense system consisting of antioxidant enzymes and molecules as a defense against oxidative damage. Failure of the antioxidant system to counteract ROS mediated damage, either due to an increased ROS production or a malfunctioning antioxidant defense, will lead to a state of oxidative stress with concomitant oxidative damage [3].
Responses to xenobiotics, including molecular or biochemical changes or cellular damage are used as biomarkers of exposure or injurious effects [4]. Several field studies show changes in antioxidant enzyme activities and levels of antioxidant molecules, as well as oxidative damage, in fish from areas supposedly exposed to prooxidants [1,5,6]. Both short term and heritable tolerance of killifish (Fundulus heteroclitus) to toxic sediments in the Elisabeth river (VA, USA) was suggested to be partly due to an upregulation of the antioxidant defence system [7].
The antioxidant defense system is generally less responsive to xenobiotics compared to other biomarkers such as the cytochrome P4501A (CYP1A) mediated 7-ethoxyresorufin-O-deethylase (EROD) activity [4]. Despite this fact, and since oxidative stress is an important mechanism in the pathology of fish, it would be of interest to establish new oxidative stress biomarkers in aquatic organisms. Elevated rates of idiopathic lesions and neoplasia in fish from polluted sites were suggested to be related to pollutant induced oxidative stress [2]. DT diaphorase (DTD; NAD(P)H:quinone oxidoreductase; NQO1; EC 1.6.99.2) was proposed as a biomarker for quinones in the aquatic environment [8]. Quinones are of interest due to their widespread occurrence in the environment, both naturally as metabolites in plants as well as environmental contaminants, such as quinonoid metabolites derived from benzene and polyaromatic hydrocarbons [9-11].
DTD is a mainly cytosolic flavoenzyme that catalyzes the two-electron reduction of quinones into hydroquinones, thus counteracting redox cycling of prooxidants. Hydroquinones are more stable and less likely to undergo autooxidation [12-15]. DTD is characterized by its ability to utilize both NADH and NADPH as electron donors and to be inhibited by the anticoagulant dicoumarol [16]. In mammals, DTD can be induced by monofunctional and bifunctional inducers together with other phase II enzymes, such as glutathione S-transferases (GST) [17,18]. In mouse, the two regulatory elements ARE (antioxidant responsive element) and XRE (xenobiotic responsive element) have been found on the 5' promoter region of the NQO1 gene, and the XRE shows significant homology to the CYP1A1 XRE [19]. Monofunctional DTD induction is mediated through the ARE, whereas bifunctional compounds can induce DTD activity by two different mechanisms: (i) directly via the aryl hydrocarbon receptor (AHR) acting on the XRE on the NQO1 gene; or (ii) through electrophilic metabolites from the CYP1A1 phase 1 reaction acting on the ARE [18]. Bifunctional inducers consist of large planar aromatic compounds, such as PAHs, whereas monofunctional inducers are a diverse group of chemicals with a majority containing, or acquiring by metabolism, a Michael reaction acceptor structure [20]. Although DTD has been proposed as a biomarker for quinones and redox cycling compounds in fish [8], few studies address the effects of prooxidants on DTD in fish.
The aim of the present study was to study the effect of the monofunctional inducers paraquat (PQ), menadione (MN) and naphtazarin (5,8-dihydroxy-1,4-naphthoquinone; DHNQ), and of the bifunctional inducers β-naphthoflavone (β-NF) and benzo(a)pyrene (B(a)P), on the catalytic activity of DTD in rainbow trout (Oncorhynchus mykiss) liver. The aim was also to investigate the suitability of rainbow trout DTD as a biomarker for redox cycling compounds. β-NF, which proved to be a potent inducer of DTD activity in rainbow trout, was also chosen to study the inducibility of DTD in other fish species used as sentinel species in monitoring studies. These fishes were the perch (Perca fluviatilis), shorthorn sculpin (Myoxocephalus scorpius), eelpout (Zoarces viviparus), brown trout (Salmo trutta), and carp (Cyprinus carpio). In addition, we studied the effects of prooxidants on the activities of catalase and EROD. Catalase metabolizes H2O2 into molecular oxygen and water [21] and EROD reflects the catalytic activity of the phase I detoxifying enzyme CYP1A [22].
Results
DT diaphorase activity
All studied compounds, both bifunctional (β-NF and B(a)P) and monofunctional (MN, PQ and DHNQ), caused significant increases in hepatic DTD activity in rainbow trout (Table 1). DTD activity increased in all β-NF exposed groups, although the increase was only significant in the group exposed to the high dose (15 mg Kg-1) and after 5 days (Table 1). Treatment with a high B(a)P dose (15 mg Kg-1) also caused increased DTD activity after 5 days (Table 1).
Table 1 Hepatic DT-diaphorase, EROD and catalase activities in rainbow trout exposed to benzo(a)pyrene (B(a)P), β-naphthoflavone (β-NF), naphthazarin (DHNQ), menadione (MN), or paraquat (PQ), for 2 and 5 days.
Dose (mg Kg-1) DT-diaphorase nmol/(min × mg protein) EROD pmol/(min × mg protein) Catalase mmol/(min × mg protein)
2 days 5 days 2 days 5 days 2 days 5 days
B(a)P 0 49.7 (9.4) 51.4 (20.3) 7.7 (3.7) 4.9 (1.5) 278 (52) 327 (33)
5 58.3 (12.6) 59.4 (11.1) 159.3 (76.1)a 237.6 (146)a 334 (128) 398 (126)
15 38.8 (16.1) 74.4 (15.8)ab 323.4 (280)a 491.8 (220)a 376 (96) 357 (114)
β-NF 0 50.5 (11.9) 54.0 (9.0) 8.5 (3.3) 7.7 (6.7) 297 (86) 247 (69)
5 69.2 (16.3) 77.4 (19.8) 896.5 (156)a 943.3 (327)a 268 (93) 182 (75)
15 72.1 (28.1) 101.7 (42.1)a 832.5 (399)a 877.3 (340)a 334 (131) 226 (95)
DHNQ 0 38.2 (5.3) 30.2 (7.6) 5.1 (3.2) 3.2 (1.8) 256 (103) 261 (171)
1 45.3 (14.6) 40.4 (5.7)a 3.6 (1.9) 4.0 (3.9) 360 (132) 311 (152)
3 41.7 (8.7) 48.3 (8.0)a 3.2 (1.6) 0.8 (0.5)ab 261 (69) 320 (64)
MN 0 40.7 (3.4) 47.0 (7.6) 7.3 (8.8) 10.0 (7.4) 144 (65) 204 (97)
5 54.0 (9.4)a 65.0 (13.2)a n.m. 6.6 (6.2) 136 (37) 215 (47)b
15 35.7 (6.7) 51.8 (6.4) 5.5 (2.1) 5.2 (2.5) 127 (38) 158 (65)
PQ 0 47.2 (11.7) 46.9 (7.6) 5.5 (3.9) 2.2 (1.1) 245 (73) 248 (97)
3 53.0 (8.3) 71.0 (6.0)a 4.3 (2.4) 1.4 (0.8)b 229 (47) 348 (55)ab
10 43.7 (6.0) 61.8 (20.4)a 3.9 (1.1) 2.7 (1.3) 155 (23)a 171 (37)a
Note: Values (n = 7) are presented as: mean (standard deviation); n.m. = not measured. aSignificantly different from control – p < 0.05; bSignificantly different from day 2 – p < 0.05.
Among the monofunctional inducers studied MN was the only one to cause a significant increase in DTD activity after 2 days (Table 1). After 2 days exposure to the low dose (5 mg Kg-1) of MN, rainbow trout displayed a significant increase in DTD activity, whereas no change was observed in the group exposed to a high dose (15 mg Kg-1) of MN. The same pattern was observed after 5 days exposure with a significant increase in DTD activity in the low dose (5 mg Kg-1) group and no increase in DTD activity in the group exposed to a high dose (15 mg Kg-1) of MN. The fish exposed to PQ showed the same trend as the MN exposed fish, with lower DTD activity in the high dose groups compared to the low dose groups. After 2 days exposure, DTD activity increased in the group treated with a low dose of PQ (3 mg Kg-1) but the difference was not statistically significant. After 5 days treatment, both the low and the high dose groups (3 and 10 mg Kg-1) displayed significantly higher DTD activity compared to the control group, even though the high dose group displayed lower DTD activity than the low dose group (Table 1). DHNQ treatment led to an increase in DTD activity in both the low dose (1 mg Kg-1) and the high dose (3 mg Kg-1) groups, after 5 days exposure (Table 1).
EROD activity
EROD activity was significantly and strongly increased in all groups treated with β-NF and B(a)P (both low and high dose and after 2 and 5 days) (Table 1). The monofunctional inducers PQ and MN treatment did not affect EROD activity, whereas the group exposed to the high dose of DHNQ (15 mg Kg-1) displayed a significant reduction in EROD activity after 5 days exposure (Table 1).
Catalase activity
PQ caused a significant decrease in catalase activity in the high dose groups (after both 2 and 5 days exposure) and a significant increase in the low dose group after 5 days exposure (Tab. 1). B(a)P, β-NF, MN and DHNQ caused no statistically significant effects on the catalase activity (Table 1).
Comparison of DTD activity in different fish species
In contrast to rainbow trout, DTD activity did not increase in perch, carp, brown trout, eelpout or shorthorn sculpin treated with a high dose (15 mg Kg-1) of β-NF for 5 days. The DTD activity in these fishes was considerably lower compared to DTD activity in rainbow trout (Table 2).
Table 2 DT diaphorase activity in rainbow trout, brown trout, perch, carp, eelpout and shorthorn sculpin exposed to 15 mg Kg-1 of β-naphthoflavone (β-NF) for 5 days.
DT diaphorase nmol/(min × mg protein)
Control β-NF
Rainbow trout 54.0 (9.0) 101.7 (42.1)a
Brown trout 14.9 (4.7) 11.1 (5.5)
Perch 4.7 (2.7) 6.3 (2.8)
Carp 5.2 (2.4) 4.0 (1.3)
Eelpout 3.2 (1.4) 2.8 (2.0)
Shorthorn sculpin 6.0 (2.8) 4.5 (2.8)
Note: Values are presented as mean (standard deviation); n = 7 for rainbow trout, and n = 6 for the additional fish species studied.
aSignificantly different from control; p < 0.05.
Discussion
The effects of β-NF and B(a)P on cellular defense systems have been extensively studied in fish, including rainbow trout. Both compounds have proved to be potent AHR agonists inducing enzymes in the CYP system, especially CYP1A [22]. Most studies in fish address effects of these compounds on the CYP system and relatively few have investigated effects on oxidative stress parameters. Stephensen et al. [23] demonstrated, in a short term study in rainbow trout liver, that treatment with 15 mg Kg-1 of β-NF caused a moderate increase in cytosolic glutathione reductase (GR) activity and a decrease in the cytosolic GST activity. Longer exposure to a higher dose of β-NF (50 mg Kg-1 for 14 days) increases GST activity in rainbow trout [24]. Regoli et al. [25] showed that both β-NF and B(a)P increased GR activity in European eel (Anguilla anguilla) liver, whereas both compounds suppressed GST and catalase activities, indicating an increase in oxidative stress caused by these two AHR agonists. Some AHR agonists can also induce DTD activity in rainbow trout. Förlin et al. [26] observed in a long term study of rainbow trout that polychlorinated biphenyls (PCB) and 3-methylcholanthrene (3-MC) caused an increase in DTD activity. It has previously been demonstrated that also β-NF induces DTD activity in rainbow trout [27]. In the present study, both β-NF and B(a)P caused an increase in hepatic DTD activity in rainbow trout. β-NF and B(a)P also caused a strong increase in the CYP1A mediated EROD activity in all exposed groups indicating strong AHR activation. The fact that exposure to the bifunctional inducers β-NF and B(a)P induced both DTD and CYP1A activities implies that the DTD induction can be mediated through activation of the AHR. In mammals, it has been demonstrated that both β-NF and B(a)P induce DTD activity [28,29]. Both compounds are classified as bifunctional since they induce DTD activity (via a mechanism involving the AHR) and have electrophilic metabolites that induce DTD activity (via the ARE) [18]. Therefore, the results obtained in this study imply that rainbow trout DTD activity can be induced by a mechanism involving the AHR, as in mammals. It is not yet known whether the promoter region of the rainbow trout DTD gene contains functional XRE or ARE. Thus, it is not clear whether the observed induction of DTD activity is mediated directly by the AHR, through the XRE, or it is caused by metabolites acting as monofunctional inducers through the ARE, or even whether both mechanism coexist. This should be investigated in future studies.
Relatively few studies address the effects of monofunctional inducers on oxidative stress parameters in fish. Stephensen et al. [23] investigated the effects on glutathione and glutathione dependent enzymes in rainbow trout exposed to the monofunctional inducers PQ, MN and DHNQ. All three compounds induced the catalytic activities of GR and GST, PQ being the most potent inducer. DHNQ also induced the activities of glutathione peroxidase and γ-glutamylcysteine synthetase, the rate-limiting enzyme in the glutathione synthesis. The same study also showed that glutathione levels were increased after exposure to PQ and MN. In the present study, the monofunctional inducers DHNQ, MN and PQ elevated hepatic DTD activity in rainbow trout, but none of those compounds induced EROD activity. This suggests that DTD activity in rainbow trout may be induced through an oxidant responsive element resembling the mammalian AREs. A study by Samson et al. [30] suggests that H2O2 dependent metallothionein induction in rainbow trout was mediated through ARE-like sequences on the metallothionein gene.
Our results show that high doses of prooxidants can lead to a decrease in DTD activity. Exposure to high doses of both PQ and MN resulted in lower DTD activities compared to exposure to low doses. Previous studies have shown that PQ is a potent redox cycler, strongly inducing GR and G6PDH activities [23,31]. The effect on G6PDH activity suggests that PQ exposure lead to the depletion of NADPH, a molecule crucial to the redox state in the cell and a cofactor in the catalytic activity of DTD and CYP1A enzymes. However, since DTD can also utilize NADH as an electron donor, the NADPH depletion should not affect DTD activity. The lower DTD activities in the rainbow trout exposed to high doses of MN and PQ, when compared to activities from animals exposed to lower doses, could be due to an overproduction of ROS; causing an oxidation dependent malfunction of the DTD enzyme. Our results also show that catalase activity was decreased after the exposure to a high dose of PQ. Catalase can be partially inhibited by the ROS superoxide [3] and it is possible that also DTD is affected by a similar mechanism. In contrast, exposure to a high dose of DHNQ led to a decrease in EROD activity and an increase in DTD activity in rainbow trout. As previously reported [32,33], the decrease in EROD activity could be due to ROS mediated inactivation of the CYP1A enzyme.
A great number of evidence shows that the main function of mammalian DTD is to protect the cell against harmful ROS production caused by redox cycling compounds [34]. Exposure to such compounds causes an increase in DTD activities in rodents [29]. Studies also demonstrate that a lack of DTD activity, either due to knockout of the DTD gene or by dicoumarol enzyme inhibition, causes an increase in quinone dependent toxicity [12,35]. The cells inability to break the redox cycling of compounds, such as quinones, increases the ROS generation that leads to increases in oxidative cellular damage, which may result in tissue malfunction. The monofunctional compounds included in this study exert their toxic effect through redox cycling, which causes ROS production [23]. Metabolites from the bifunctional compounds may also cause ROS production through redox cycling [23]. The increase in hepatic DTD activity in rainbow trout by these compounds implies that DTD has a protective role also in this species. The increase of hepatic DTD activity in response to prooxidant exposure in rainbow trout may also suggest DTD activity as part of an oxidative stress biomarker battery. Cultivated rainbow trout are often used in fresh water caging studies for environmental biomonitoring. Biomarkers can be defined as specific biological changes resulting from exposure to chemicals. In order to qualify as an enzymatic biomarker, the enzyme needs to be induced or changed in a measurable way as a result to an exposure. This study shows that hepatic DTD activity in rainbow trout is inducible by prooxidants in short term exposure studies, suggesting that hepatic DTD may serve as a biomarker for prooxidants in rainbow trout. However, rainbow trout were exposed to high doses of the test compounds and the elevation of DTD activity was moderate (2-fold). Long-term studies with ecologically relevant doses would be necessary to fully evaluate the suitability of DTD as a biomarker.
Hepatic catalase activity is frequently used as a biomarker to assess oxidative stress in biomonitoring programs in the aquatic environment [36-38]. Catalase is mainly a peroxisomal enzyme, and it is thus possible that an elevation of catalase activity reflects peroxisomal proliferation rather than antioxidant defense. It has been shown that catalase activity positively correlates the number of peroxisomes in mice [39]. Though elevation in catalase activity is observed in field studies, few laboratory studies have reported increased catalase activities in fish exposed to prooxidants [4]. In the present study, only exposure to PQ and not to β(a)P, β-NF, DHNQ or MN affected catalase activity. This suggests that catalase is only slightly responsive to acute prooxidant exposure.
Among the fish species included in this study, rainbow trout was the only species that displayed inducible DTD activities. Treatment with 15 mg Kg-1 of β-NF for five days doubled the DTD activity in rainbow trout liver cytosol, whereas no increase in DTD activity was observed in the other fish species receiving the same treatment, including the brown trout (Salmo trutta) that is more phylogenetically related to the rainbow trout. β-NF is believed to be rapidly metabolized by CYP1A enzymes and a dose of 15 mg Kg-1 causes a high but not maximal induction of CYP1A activities in rainbow trout. Since different species differ in their sensitivity towards AHR agonists, 15 mg Kg-1 was chosen as a moderate dose to initially screen the effects of a possible DTD inducer on fish species commonly used as sentinel ones in monitoring studies. This dose was chosen to avoid a too high dose capable of inhibiting DTD activity but also high enough for inducing DTD activity. Nevertheless, further studies should be conducted to investigate effects of increased number of doses and duration of exposures. On the other hand, studies in other fish species report low levels of DTD activity and inability of AHR agonists to induce DTD activity [27,40]. For example, wild sea bass (Dicentrarachus labrax) and dab (Limanda limanda) exposed to 3-MC displayed low DTD activities (ca. 4 and 3 nmol min-1 mg-1, respectively, in control fish) and no DTD induction [27]. Pretti et al. [40] reported low DTD activities in gilthead seabream (Sparus auratus) (0.7 nmol min-1 mg-1 in control fish) and observed no elevation in DTD activity after exposing the fish to a high dose (50 mg Kg-1) of β-NF. Previous studies in rainbow trout exposed to the AHR agonist 3-MC and β-NF showed increased hepatic DTD activities [26,27].
Rainbow trout also displayed much higher DTD activities compared to the other fish species included in this study. When comparing the activities of different detoxification enzymes in different fish species, Förlin et al. [41] described a similar pattern with higher DTD activities in rainbow trout compared to other fish species. It was suggested that the observed species difference could be due to the presence of DTD inducing compounds in the commercial fish food, being rainbow trout the only cultivated fish species included in the study. This suggestion, however, is not fully supported by the fact that DTD activity was not induced in the other fish species treated with β-NF. Instead, it seems that the rainbow trout has higher natural basal levels of DTD activity compared to other fish species. It is also possible that rainbow trout express an inducible isoform of the enzyme, different from that found in other investigated fish species. The reason for this apparent species difference in relation to DTD is not known. It would be of interest to investigate whether the rainbow trout specific feature of DTD is natural or acquired. Stocks of rainbow trout have been cultivated for up to one hundred years and the species may have adapted to the special conditions in a fish farm. Higher basal levels of a more effective DTD enzyme could be one such adaptive response to prooxidants in commercial fish foods.
Conclusion
Since rainbow trout DTD activity is inducible by both monofunctional and bifunctional inducers, we suggest that rainbow trout DTD gene expression may be regulated by the same mechanisms, as in mammals. The inducibility of rainbow trout hepatic DTD may also suggest that rainbow trout DTD have a protective role. The rainbow trout has a higher basal DTD activity compared to other fish species studied, and its activity is inducible. The fact that DTD activity is inducible in rainbow trout suggests that the enzyme may be suitable as a part of a biomarker battery; for example, in biomonitoring caging studies where cultivated rainbow trout are being used. It appears that DTD activity is not suitable as a biomarker in the other fish species investigated in this study, and that further studies are needed to elucidate the DTD response to prooxidants in these fish species.
Methods
Chemicals
β-naphthoflavone (β-NF) was obtained from Jansson Chimica and menadione (MN), paraquat (PQ), 5,8-Dihydroxy-1,4-naphthoquinone (naphthazarin; DHNQ) and benzo-a-pyrene (B(a)P) were obtained from Sigma (St. Louis, USA). 7-ethoxyresorufin, reduced β-nicotinamide adenine dinucleotide 2'-phosphate (NADPH), reduced β-nicotinamide adenine dinucleotide (NADH), bovine serum albumine (BSA) and 2,6-dichlorophenol indophenol (DCPIP) were obtained from Sigma (St. Louis, USA). Rhodamine B and H2O2 were obtained from MERCK (Darmstadt, Germany) and Dicoumarol from Aldrich (Milwaukee, USA). All other chemicals were of analytical grade.
Fish
Cultured juvenile rainbow trout of both sexes, weighting 108 (SD = 24) g, were obtained from Antens fiskodling AB, a hatchery close to Göteborg. At least a week prior to the experiments, the fish were acclimatized to the laboratory conditions in 500 l tanks with filtered, aerated fresh water at 10°C in a flow-through system (5 l water/min). During the experiment, fish were kept in 50 l glass aquaria (7 fish per aquaria) with filtered, aerated water at 10°C, also in a flow-through system (0.5 l water/min). The fish were not fed and were exposed to a 12:12 h light/dark cycle.
Perch (Perca fluviatilis) weighing 6 (SD = 0.8) g, shorthorn sculpin (Myoxocephalus scorpius) weighing 153 (SD = 41) g, eelpout (Zoarces viviparus) weighing 35 (SD = 13) g, and brown trout (Salmo trutta) weighing 102 (SD = 23) g, were provided by local fishermen, and carp (Cyprinus carpio) weighing 155 (SD = 42) g was obtained from a local hatchery. Perch, brown trout and carp were treated as described for rainbow trout. Eelpout and shorthorn sculpin were kept in salt water (30‰), also under similar conditions as described for rainbow trout.
Experimental design
At the start of the experiment, fish were injected intraperitoneally with the test compound dissolved in its respective carrier solution (peanut oil for β-NF, B(a)P, MN and DHNQ and 0.15 M KCl for PQ) or the carrier solution alone (0.5 ml/100 g fish). Rainbow trout were divided into three groups: one control group that received the carrier solution alone; one group that was exposed to a low dose of a single test compound (B(a)P, β-NF or MN 5 mg Kg-1, K DHNQ: 1 mg Kg-1, K PQ: 3 mg Kg-1); and, finally, one group that was exposed to a high dose of a single test compound (B(a)P, β-NF or MN 15 mg Kg-1, DHNQ, 3 mg Kg-1, PQ, 10 mg Kg-1). Doses were chosen relative to PQ toxicity data (manufacturers label, mammalian data) with the high PQ dose 10 times lower than recorded LD50 values. Since B(a)P, β-NF and MN are less toxic than PQ, these compounds were administered in higher doses. Due to DHNQ's higher toxicity we used lower doses of this compound. The exposure times were 2 and 5 days, and 7 fish were sampled per exposure group and time point.
Fish were killed with a sharp blow to the head and the weight and length recorded. Each fish was cut open and the liver excised, weighed and homogenized in Na+/K+-phosphate buffer (pH 7.4) containing 0.15 M KCl. The homogenate was centrifuged for 10 min at 700 g, and the supernatant recentrifuged for 20 min at 10,000 g. An aliquot of the supernatant (S9 fraction) was stored at -80°C for EROD measurements, and the rest of the supernatant centrifuged for 60 min at 105,000 g. Aliquots of the supernatant (cytosolic fraction) were stored at -80°C until analyzed. All subcellular preparation steps were performed at 4°C.
The additional fish species studied were treated with a high dose (15 mg Kg-1) for 5 days. The fish were exposed and sampled and the samples treated as described for rainbow trout. Six fish were sampled per exposure group.
Biochemical assays
DT-diaphorase activity was measured in liver cytosolic fraction according to Ernster [13], modified and adapted to a microplate reader [42]. The reaction mixture contained 50 mM Tris-HCl (pH 7.6), 40 μM DCPIP and 0.3 mM NADH in a final volume of 210 μl. Ten μl of the sample were pipetted into microplate wells and the reaction was started by the addition of 200 μl of reaction mixture. Samples were measured with or without the addition of 0.1 mM dicoumarol, dissolved in 0.15 % NaOH. DTD activity was defined as dicoumarol inhibitable DCPIP reduction. Change in absorbance was monitored at 600 nm, and DTD activity calculated using the extinction coefficient for DCPIP (ε = 21 mM-1cm-1).
EROD activity was measured in liver S9 fraction according to the method described by Förlin et al. [43] using rhodamine as standard. The reaction mixture contained 0.1 M Na-phosphate buffer (pH 8.0), 0.5 μM ethoxyresorufin and 25–50 μl of sample in a final volume of 2 ml. The reaction was started with the addition of 10 μl 10 mM NADPH. The increase in fluorescence was monitored at 530 nm (excitation) and 585 (emission).
Catalase activity was measured in liver cytosol according to the method described by Aebi [21]. The reaction mixture contained 50 mM K-phosphate buffer (pH 6.5) and 50 mM H2O2 diluted in 80 mM K-phosphate buffer (pH 6.5). The mixture was incubated at 25°C, the baseline recorded, and the decrease in absorbance further recorded at 240 nm, after the addition of the sample. Catalase activity was calculated using the extinction coefficient for H2O2 (ε = 40 M-1cm-1).
Protein content was determined using the BCA Protein Assay Kit (Pierce, USA), with BSA as standard.
Statistical analysis
Data were analyzed with one-way analysis of variance (ANOVA) and, following significant differences, with Newman-Keuls post hoc test. Levene's test was used to test for homogeneity of variances. Data displaying heterogeneity of variances were instead analysed using Kruskal-Wallis H test, followed by Dunnett's C test. When comparing only two groups, the Mann-Whitney U test was used. All statistical analyses were calculated using SPSS® for Windows. The significance level (α) was set at 0.05. Data are presented as: mean (standard deviation).
Authors' contributions
JS participated in the experimental design, fish exposure, sampling, performed most of the analysis and writing of the manuscript. ES participated in the experimental design, fish exposure and the sampling. LF rose funding and participated in the experimental design and writing of the manuscript. All authors read and approved the final manuscript.
Acknowledgements
The authors would like to thank Helge Ax:son Johnsons Stiftelse, Kungliga Vetenskaps-och Vitterhets-Samhället i Göteborg, MISTRA and the EU-BEEP project for financial support. We thank Aina Stenborg for technical assistance. We would also like to thank Dr. Malin Celander for valuable comments on the manuscript.
==== Refs
Livingstone DR Contaminant-stimulated reactive oxygen species production and oxidative damage in aquatic organisms Mar Pollut Bull 2001 42 656 666 11525283 10.1016/S0025-326X(01)00060-1
Winston GW Di Giulio RT Prooxidant and antioxidant mechanisms in aquatic organisms Aquat Toxicol 1991 19 137 161 10.1016/0166-445X(91)90033-6
Halliwell B Gutteridge JMC Free radicals in biology and medicine 1999 third Oxford: Oxford University Press
van der Oost R Beyer J Vermeulen NPE Fish bioaccumulation and biomarkers in environmental risk assessment: a review Environ Toxicol Pharmacol 2003 13 57 149 10.1016/S1382-6689(02)00126-6
Di Guilio RT Benson WH Sanders BM Van Veld PA Rand G Biochemical mechanisms:metabolism, adaptation and toxicity Fundamentals of Aquatic Toxicology Effects, Environmental Fate and Risk Assessment 1995 2 London: Taylor and Francis Ltd
Oakes KD Van Der Kraak GJ Utility of the TBARS assay in detecting oxidative stress in white sucker (Catostomus commersoni) populations exposed to pulp mill effluent Aquat Toxicol 2003 63 447 463 12758008 10.1016/S0166-445X(02)00204-7
Meyer JN Smith JD Winston GW Di Giulio RT Antioxidant defenses in killifish (Fundulus heteroclitus) exposed to contaminated sediments and model prooxidants: short-term and heritable responses Aquat Toxicol 2003 65 377 395 14568353
Bagnasco M Camoirano A De Flora S Melodia F Arillo A Enhanced liver metabolism of mutagens and carcinogens in fish living in polluted seawater Mutat Res 1991 262 129 137 1705659 10.1016/0165-7992(91)90119-O
Bolton JL Trush MA Penning TM Dryhurst G Monks TJ Role of quinones in toxicology Chem Res Toxicol 2000 13 135 160 10725110 10.1021/tx9902082
Willson RL Quinones, semiquinone free radicals and one-electron transfer reactions: a walk in the literature from Peru to S.O.D Free Radic Res Commun 1990 8 201 217 2191902
Nebert DW Petersen DD Fornace AJ Jr Cellular responses to oxidative stress: the [Ah] gene battery as a paradigm Environ Health Perspect 1990 88 13 25 2272308
Ernster L DT-Diaphorase – A historical review Chemica Scripta 1987 27A 1 13
Ernster L DT diaphorase Methods Enzymol 1967 10 309 317
Gutierrez PL The role of NAD(P)H oxidoreductase (DT-Diaphorase) in the bioactivation of quinone-containing antitumor agents: a review Free Radical Biol Med 2000 29 263 275 11035255 10.1016/S0891-5849(00)00314-2
Cadenas E Antioxidant and prooxidant functions of DT-Diaphorase in quinone metabolism Biochem Pharmacol 1995 49 127 140 7530954 10.1016/S0006-2952(94)00333-5
Lind C Cadenas E Hochstein P Ernster L DT-Diaphorase: purification, properties and function Methods Enzymol 1990 186 287 301 2233301
Daniel V Glutathione S-transferases: gene structure and regulation of expression Crit Rev Biochem Mol Biol 1993 28 173 207 8325038
Prochaska HJ Talalay P Regulatory mechanisms of monofunctional and bifunctional anticarcinogenic enzyme inducers in murine liver Cancer Res 1988 48 4776 4782 3409219
Ross D Kepa JK Winski SL Beall HD Anwar A Siegel D NAD(P)H:quinone oxidoreductase 1 (NQO1): chemoprotection, bioactivation, gene regulation and genetic polymorphisms Chem Biol Interact 2000 129 77 97 11154736 10.1016/S0009-2797(00)00199-X
Dinkova-Kostova AT Fahey JW Talalay P Chemical structures of inducers of nicotinamide quinone oxidoreductase 1 (NQO1) Methods Enzymol 2004 382 423 448 15047115
Aebi H Bergmeyer, HU Catalase Methods of enzymatic analysis 1985 New York: Academic press 671 684
Goksøyr A Förlin L The cytochrome P450 system in fish, aquatic toxicology and environmental monitoring Aquat Toxicol 1992 22 287 312 10.1016/0166-445X(92)90046-P
Stephensen E Sturve J Förlin L Effects of redox cycling compounds on glutathione content and activity of glutathione-related enzymes in rainbow trout liver Comp Biochem Physiol C 2002 133 435 442
Zhang YS Andersson T Förlin L Induction of hepatic xenobiotic biotransformation enzymes in rainbow trout by beta naphthoflavone. Time-course studies Comp Biochem Physiol B 1990 95 247 254 10.1016/0305-0491(90)90072-2
Regoli F Winston GW Gorbi S Frenzilli G Nigro M Corsi I Focardi S Integrating enzymatic responses to organic chemical exposure with total oxyradical absorbing capacity and DNA damage in the European eel Anguilla anguilla Environ Toxicol Chem 2003 22 2120 2129 12959540 10.1897/02-378
Förlin L Blom S Celander M Sturve J Effects on UDP glucuronosyl transferase, glutathione transferase, DTdiaphorase and glutathione reductase activities in rainbow trout liver after long-term exposure to PCB Mar Environ Res 1996 42 213 216 10.1016/0141-1136(95)00061-5
Lemaire P Forlin L Livingstone DR Responses of hepatic biotransformation and antioxidant enzymes to CYP1A-inducers (3-methylcholanthrene, [beta]-naphthoflavone) in sea bass (Dicentrarchus labrax), dab (Limanda limanda) and rainbow trout (Oncorhynchus mykiss) Aquat Toxicol 1996 36 141 160 10.1016/S0166-445X(96)00819-3
Joseph P Jaiswal A NAD(P)H:quinone oxidoreductase 1 reduces the mutagenicity of DNA caused by NADPH:P450 reductase-activated metabolites of benzo(a)pyrene quinones Brit J Cancer 1998 77 709 719 9514048
Talalay P Prochaska H Mechanisms of induction of of NAD(P)H-Quinone reductase Chemica scripta 1987 27A 61 66
Samson SL-A Paramchuk WJ Gedamu L The rainbow trout metallothionein-B gene promoter: contributions of distal promoter elements to metal and oxidant regulation Biochim Biophys Acta 2001 1517 202 211 11342100
Akerman G Amcoff P Tjarnlund U Fogelberg K Torrissen O Balk L Paraquat and menadione exposure of rainbow trout (Oncorhynchus mykiss)-studies of effects on the pentose-phosphate shunt and thiamine levels in liver and kidney Chem Biol Interact 2003 142 269 283 12453665 10.1016/S0009-2797(02)00122-9
Schlezinger JJ White RD Stegeman JJ Oxidative inactivation of cytochrome P-450 1A (CYP1A) stimulated by 3,3',4,4'-tetrachlorobiphenyl: production of reactive oxygen by vertebrate CYP1As Mol Pharmacol 1999 56 588 597 10462547
Schlezinger JJ Stegeman JJ Induction and suppression of cytochrome P450 1A by 3,3',4,4',5-pentachlorobiphenyl and its relationship to oxidative stress in the marine fish scup (Stenotomus chrysops) Aquat Toxicol 2001 52 101 115 11164533 10.1016/S0166-445X(00)00141-7
Dinkova-Kostova AT Talalay P Persuasive evidence that quinone reductase type 1 (DT diaphorase) protects cells against the toxicity of electrophiles and reactive forms of oxygen Free Radical Biol Med 2000 29 231 240 11035251 10.1016/S0891-5849(00)00300-2
Radjendirane V Joseph P Lee YH Kimura S Klein-Szanto AJ Gonzalez FJ Jaiswal A Disruption of the DT diaphorase (NQO1) gene in mice leads to increased menadione toxicity J Biol Chem 1998 273 7382 7389 9516435 10.1074/jbc.273.13.7382
Tagliari KC Cecchini R Rocha JAV Vargas VMF Mutagenicity of sediment and biomarkers of oxidative stress in fish from aquatic environments under the influence of tanneries Mutat Res 2004 561 101 117 15238235
Lionetto MG Caricato R Giordano ME Pascariello MF Marinosci L Schettino T Integrated use of biomarkers (acetylcholinesterase and antioxidant enzymes activities) in Mytilus galloprovincialis and Mullus barbatusin an Italian coastal marine area Mar Pollut Bull 2003 46 324 330 12604066 10.1016/S0025-326X(02)00403-4
Regoli F Pellegrini D Winston GW Gorbi S Giuliani S Virno-Lamberti C Bompadre S Application of biomarkers for assessing the biological impact of dredged materials in the Mediterranean: the relationship between antioxidant responses and susceptibility to oxidative stress in the red mullet (Mullus barbatus Mar Pollut Bull 2002 44 912 922 12405216 10.1016/S0025-326X(02)00120-0
De Craemer D Vamecq J Roels F Vallee L Pauwels M Van Den Branden C Peroxisomes in liver, heart, and kidney of mice fed a commercial fish oil preparation: Original data and review on peroxisomal changes induced by high-fat diets J Lipid Res 1994 35 1241 1250 7964185
Pretti C Salvetti A Longo V Giorgi M Gervasi PG Effects of beta-naphthoflavone on the cytochrome P450 system, and phase II enzymes in gilthead seabream (Sparus aurata) Comp Biochem Physiol C 2001 130 133 144
Förlin L Lemaire P Livingstone DR Comparative studies of hepatic xenobiotic metabolizing and antioxidant enzymes in different fish species Mar Environ Res 1995 39 201 204 10.1016/0141-1136(94)00030-S
Stephensen E Svavarsson J Sturve J Ericson G Adolfsson-Erici M Förlin L Biochemical indicators of pollution exposure in shorthorn sculpin (Myoxocephalus scorpius), caught in four harbours on the southwest coast of Iceland Aquat Toxicol 2000 48 431 442 10794829 10.1016/S0166-445X(99)00062-4
Förlin L Goksøyr A Husöy AM Kramer KJM Cytochrome P450 monooxygenase as indicator of PCB/dioxin like compounds in fish Biomonitoring of coastal waters and estuaries 1994 Boca Raton, Florida: CRC Press 135 150
| 15871734 | PMC1156924 | CC BY | 2021-01-04 16:38:24 | no | Comp Hepatol. 2005 May 4; 4:4 | utf-8 | Comp Hepatol | 2,005 | 10.1186/1476-5926-4-4 | oa_comm |
==== Front
Curr Control Trials Cardiovasc MedCurrent Controlled Trials in Cardiovascular Medicine1468-67081468-6694BioMed Central 1468-6708-6-81590720110.1186/1468-6708-6-8ResearchEffect of increased convective clearance by on-line hemodiafiltration on all cause and cardiovascular mortality in chronic hemodialysis patients – the Dutch CONvective TRAnsport STudy (CONTRAST): rationale and design of a randomised controlled trial [ISRCTN38365125] Penne E Lars [email protected] Peter J [email protected] Michiel L [email protected] den Dorpel Marinus A [email protected] Muriel P [email protected]é Menso J [email protected] der Tweel Ingeborg [email protected] Wee Piet M [email protected] CONTRAST study group 1 Department of Nephrology, University Medical Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands2 Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Universiteitsweg 100, 3584 CG Utrecht, The Netherlands3 Department of Internal Medicine, Rijnmond-Zuid Medical Center, Clara Location, Olympiaweg 350, 3078 HT Rotterdam, The Netherlands4 Department of Nephrology, VU Medical Center, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands5 Department of Internal Medicine, Medical Center Alkmaar, Wilhelminalaan 12, 1815 JD Alkmaar, The Netherlands6 Center for Biostatistics, Utrecht University, Padualaan 14, 3584 CH Utrecht, The Netherlands2005 20 5 2005 6 1 8 8 3 5 2005 20 5 2005 Copyright © 2005 Penne et al; licensee BioMed Central Ltd.2005Penne et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The high incidence of cardiovascular disease in patients with end stage renal disease (ESRD) is related to the accumulation of uremic toxins in the middle and large-middle molecular weight range. As online hemodiafiltration (HDF) removes these molecules more effectively than standard hemodialysis (HD), it has been suggested that online HDF improves survival and cardiovascular outcome. Thus far, no conclusive data of HDF on target organ damage and cardiovascular morbidity and mortality are available. Therefore, the CONvective TRAnsport STudy (CONTRAST) has been initiated.
Methods
CONTRAST is a Dutch multi-center randomised controlled trial. In this trial, approximately 800 chronic hemodialysis patients will be randomised between online HDF and low-flux HD, and followed for three years. The primary endpoint is all cause mortality. The main secondary outcome variables are fatal and non-fatal cardiovascular events.
Conclusion
The study is designed to provide conclusive evidence whether online HDF leads to a lower mortality and less cardiovascular events as compared to standard HD.
End stage renal diseasehemodialysishemodiafiltrationconvective transportmiddle moleculesmortalitycardiovascular diseaseoutcome
==== Body
Background and rationale
Atherosclerotic cardiovascular disease (CVD) is common among hemodialysis (HD) patients. In fact, approximately 50% of the deaths is attributed to cardiovascular causes, which is much higher than in the general population [1]. In addition, chronic HD patients suffer from atherosclerotic complications at a relatively younger age and die younger from ischemic heart disease [2]. The origin of CVD in chronic HD patients is most probably multi-factorial, as the extremely high prevalence in this patient group is not easily explained by traditional risk factors, either alone or in combination [3]. In recent years, other contributing factors have emerged, including the accumulation of uremic toxins, disturbances in the immuno-inflammatory system, as reflected by a chronic micro-inflammatory state, increased oxidative stress, and endothelial dysfunction [4-6]. In particular the retention of larger uremic toxins, the so-called middle molecules (MM, molecular weight [MW] 0.5 – 50 kDa), may play an important role in the pathogenesis of CVD [7,8]. Therefore, it is conceivable that dialysis modalities with superior MM removal reduce CVD and improve survival.
In contrast to diffusive dialysis strategies, which mainly remove small MW solutes, such as urea and creatinine, convective dialysis strategies are particularly effective in the removal of larger molecules. In hemodiafiltration (HDF), diffusive and convective transport are combined, providing an optimal removal of both small and larger MW substances up to the range of 30 – 40 kDa. Clinical studies have shown that beta-2-microglobulin (β2 m), which is a typical MM with a MW of 11.8 kDa and therefore incapable of passage through the membrane of low flux devices, is effectively removed during HDF leading to lower pre-dialysis levels in the long term [9,10]. Similarly, the removal of other MM such as advanced glycation end-products (AGEs), leptin, and complement factor D is enhanced by convective transport [11-13]. Apart from the increased MM clearance, it has been suggested that HDF improves the removal of smaller molecules that are highly protein bound, due to a better elimination of the unbound fraction [14]. With respect to homocysteine, which is >90% protein bound, the observed decrease may also be explained by an improved removal of uremic substances with inhibitory effects on its metabolism [15]. Finally, it has been shown that treatment with online HDF leads to lower plasma phosphate concentrations, as compared to standard HD [16,17].
At present, it is unclear whether HDF has a favourable effect on the micro-inflammatory state in dialysis patients. Although a reduction of pro-inflammatory proteins has been shown during HDF [18], anti-inflammatory cytokines may also be removed. Of note, besides solute removal, other factors may influence the inflammatory-state as well, such as the bio-incompatibility of the dialyser membrane and the microbiological quality of the dialysate [19,20]. Considering oxidative stress and endothelial dysfunction, data on the effects of HDF on these parameters are limited.
In summary, compared to standard HD, HDF improves the uremic state by an increased clearance of MM and other, mainly protein bound, uremic toxins. Circumstantial evidence implies that these effects result in less vascular damage and ultimately in decreased cardiovascular morbidity and mortality (figure 1). Although observational studies suggest that online HDF improves cardiovascular outcome in chronic HD patients [21,22], two small randomised studies failed to show any differences between online HDF and standard HD [23,24]. However, the latter analysis lacked adequate power to detect differences in clinical endpoints.
Figure 1 Hypothesis
Based on the above-mentioned theoretical considerations, the scarcity of reliable clinical data, and the growing interest in convective techniques under nephrologists, the CONvective TRAnsport STudy (CONTRAST) has been initiated. CONTRAST is a randomised controlled trial investigating the effects of online HDF on clinical endpoints, compared to low-flux HD. If online HDF indeed leads to an improvement in CV morbidity and mortality, this finding will imply a breakthrough in the treatment of chronic HD patients.
Methods
Objectives
The primary objective of CONTRAST is to assess the effect of on-line HDF on all cause mortality, when compared to standard low-flux HD. The main secondary outcomes are fatal and non-fatal cardiovascular events. Other secondary outcome measures include differences between treatment regimens on the progression of left ventricular mass index (LVMi), as assessed by echocardiography, the progression of atherosclerosis as assessed by measurement of carotid intima-media thickness (CIMT) and the progression of arterial stiffness, as assessed by measurement of aortic pulse wave velocity (PWV). Furthermore, several laboratory markers of endothelial function, inflammatory state, and oxidative stress will be assessed over time and compared between the two treatment groups. In addition, subjective global assessment (SGA) is performed in the study patients as a measure of nutritional state, and a questionnaire is used to investigate the effects of on-line HDF on quality of life.
Study design
In this randomised controlled trial, participants are randomised centrally into a 1:1 ratio for treatment with online HDF or treatment with low-flux HD. Randomisation is stratified by the participating centres and occurs in blocks. The follow-up period is three years. At present, 24 dialysis centres have agreed to recruit the required number of patients. The study is conducted according to good clinical practice (GCP) guidelines.
Patients
The in- and exclusion criteria are given in table 1. Since the study results may be of importance for chronic HD patients of all ages, no upper age limit has been set. Severe incompliance is defined as non-adherence to the dialysis prescription, especially the frequency and duration of dialysis treatment. Permission for participation in other (e.g. observational) studies will be discussed with and decided by the executive committee.
Table 1 Inclusion and exclusion criteria
Inclusion criteria
patients treated by HD 2 or 3 times a week, for at least 2 months.
patients able to understand the study procedures.
patients willing to provide written informed consent.
Exclusion criteria
current age < 18 years
treatment by HDF or high flux HD in the preceding 6 months
severe incompliance
life expectancy < 3 months due to non renal disease
participation to other clinical intervention trials evaluating cardiovascular outcome
Stabilisation period
Before randomisation, patients will be dialysed 3 times (or 2 times) per week with low-flux synthetic dialysers (UF-coefficient < 20 ml/mmHg/h) for at least 6 months in case of a prevalent dialysis patient and at least 2 months in case of a new dialysis patient.
Blood flow will be maintained at 250–400 ml/min. Anticoagulation is performed with low molecular weight heparin (LMWH) before HD. Patients on coumarins receive 50% of the LMWH dose. Treatment times will be adapted to a target dialysis spKt/V urea of ≥ 1.2 per treatment. Ultra pure water is used for preparation of dialysis fluid. Bicarbonate is provided from powder cartridges to avoid the risk of a bacterial load from bicarbonate concentrates. For instance, the biBAGR system (Fresenius) and BiCartR system (Gambro) will be used. The dialysate flow is 500 ml/min and the temperature of the dialysate is 36°C.
Routine patient care
Metabolic control will be performed according to the guidelines of the Quality of Care Committee of the Dutch Federation of Nephrology. Anti-hypertensive medication, lipid lowering therapy, platelet aggregation inhibitors and medication to treat renal anemia and renal osteodystrophy will also be prescribed according to these guidelines, and, if not available, according to usual care.
Randomisation
The patients will be randomised as soon as they are considered to be stable. When a patient has been randomised for low-flux HD, the treatment as performed in the stabilisation period will be continued. Treatment times will be adjusted only if dialysis spKt/V urea < 1.2 per treatment or if ultrafiltration goals can not be achieved, according to the attending nephrologist. When randomised for online HDF, patients will be treated with a target post-dilution dose of 6 l/h (~100 ml/min) and a high-flux synthetic dialyser (UF-coefficient > 20 ml/mmHg/h). Blood flow will be set at >300 ml/min, if possible, in order to achieve a substitution volume of 100 ml/min. If the blood flow is less than 300 ml/min, the post-dilution volume will be decreased accordingly (filtration and post-dilution <25–33% of blood flow). If necessary, the dose of LMWH will be increased and given in two separate doses. Treatment times will be fixed according to the prescription in the stabilisation period and adjusted only when spKt/V urea is < 1.2 / treatment. Metabolic control and medication is similar to the low-flux group, as described above.
Dialyser membranes
Dialysers with comparable biocompatible membrane material and surface area will be used in both treatment groups, to ascertain that differences in clearance result from differences in convective transport, rather than differences in dialyser characteristics. Only if the target dose of 6 l/h post-dilution is not achieved in the online HDF patients, it is allowed to prescribe a membrane with a larger surface area. The membranes advised by the study group are summarised in table 2. As many low-flux membranes with a membrane surface > 1.5 m2 have a UF coefficient 10–20 ml/mmHg/h, in this study low-flux is defined as a UF coefficient of < 20 ml/mmHg/h.
Table 2 Dialyser characteristics for both treatment arms
Low-flux HD Online HDF
Company Gambro Fresenius Gambro Fresenius
Dialyser Polyflux 17L F8HPS Polyflux 170H FX80
Membrane material polyamide polysulfone polyamide polysulfone (helixone)
Sterilisation method heat heat heat heat
Surface area (m2 ) 1.7 1.8 1.7 1.8
Membrane thickness (μm) 50 40 50 35
UF coefficient (ml/mmHg/h) 13 18 65 59
In vitro clearance:#
Urea 260 251 268 276
Phosphate 198 193 229 239
Vit B12 111 118 158 175
# (QB = 300 ml/min, QD = 500 ml/min)
Online HDF technique
During hemodiafiltration, the removal of larger solutes is increased by excess ultrafiltration, leading to solute removal by convection. As fluid removal exceeds the desired weight loss of the patient, fluid balance is maintained by the infusion of a pyrogen-free substitution solution. In addition, dialysate is used to create a concentration gradient for solute removal by diffusion, as in standard HD. At the introduction of HDF more than 20 years ago, the substitution fluid was supplied in bags. The infusion volumes were limited due to the high costs and laborious procedure, limiting the efficiency of HDF.
In recent years, however, technical advances have made it possible to prepare the substitution solution online from ultra pure water and dialysate concentrates. As a result, the volume of the substitution fluid could be increased considerably, without the disadvantages of inconvenient bags. Hence, the UF rate can be increased up to 50L per treatment in the pre-dilution mode and 25L in the post-dilution mode [25].
Online dialysate and substitution fluid preparation
Ultra-pure water is used for the preparation of bicarbonate-containing dialysis fluid, which undergoes one step of ultrafiltration converting it into ultra pure dialysis fluid. Dialysis fluid is produced at a rate of 600–800 ml/min of which approximately 100 ml/min is diverted for further processing into substitution fluid. The electrolyte composition of the dialysis fluid is: Na+ 138–140 mmol/l; K+ 1.0–3.0 mmol/l; HCO3 - 30–35 mmol/l; Ca++ 1.0–1.7 mmol/l; Mg++ 0.5 mmol/l; Cl- 108–109.5 mmol/l; glucose 0–5.6 mmol/l; acetate 3 mmol/l.
The substitution fluid is prepared from the dialysis fluid by one additional step of controlled ultrafiltration, before it is infused post-filter into the blood. The electrolyte composition of the substitution fluid is the same as the composition of the dialysis fluid. Ultrafiltration procedures will be performed according to the manufacturers' instructions, as described below.
- The on-line system, ONLINEplus™ (Fresenius Medical Care, Bad Homburg, Germany) is integrated into the dialysis machine (4008 series; Fresenius Medical Care) and consists of two ultrafilters (DIASAFE® plus), an infusate pump module, and disposable infusate lines. Infusate is prepared continuously by double-stage ultrafiltration. Both filters are subjected to automated membrane integrity tests before dialysis, and are replaced after 100 treatments or 12 weeks of use, whichever comes first. Dialysis fluid downstream from the first filter stage enters the dialyser; part of the stream is subjected to cross-flow filtration in the second filter in order to produce infusate. The infusate stream is connected with the venous bubble catcher for post-dilution HDF [25,26].
- The AK 100/200 ULTRA dialysis machine (Gambro AB, Lund, Sweden) prepares ultra pure water and ultra pure dialysis fluid by stepwise ultrafiltration of water and bicarbonate -containing dialysis fluid (BiCart) using two polyamide ultrafilters (U8000 S). When used for HDF, sterile non-pyrogenic solution is prepared on-line from the ultra pure dialysis fluid by an additional step of ultrafiltration using a sterile polyamide ultrafilter (U2000) integrated in a sterile line set (Steriset). The hygiene of the fluid pathway, including the U8000S ultrafilters, will be assured by heat disinfections after each treatment. The U8000S filters are changed bimonthly. The final ultrafilter (U2000) is employed on a single-use basis [26,25].
Data collection
Baseline and follow-up data registration
At baseline, all relevant information will be documented: i.e. demographical data, information on cardiovascular risk factors, time on dialysis, cause of renal insufficiency, and medication. A follow-up visit will be scheduled every three months. During this visit, the occurrence of CV events, death, and hospitalisation will be documented. In addition, blood pressure, body weight and the achieved filtration and substitution dose per treatment will be registered. Case record forms are provided using the TeleForm system (version 8.1.1, Cardiff Software Inc, Vista, CA, USA). As all completed forms are scanned, no data entry by typing is needed. Registration of all data will be performed in each centre by the attending nephrologists and research nurses.
Recording outcome events
CV events include fatal or non-fatal myocardial infarction, stroke, therapeutic coronary procedure (PTCA or stenting), therapeutic carotid procedure (endarterectomy or stenting), and PTA and vascular intervention (revascularisation, PTA or stenting). Congestive heart failure is excluded as a CV event, since the discrimination with fluid overload is often hard to make in chronic HD patients. Furthermore, hospitalisations, duration of the hospitalisations and main diagnosis (including the occurrence of infections) will be recorded during the study period.
An independent event committee will evaluate all causes of death, cardiovascular events, and infections. The primary investigators will collect sufficient information of the events for the event committee. The event committee is blinded for information on the received treatment and consists of physicians with different specialisations: neurologists, vascular surgeons, nephrologists, internists, and cardiologists. Events will be coded as fatal and non-fatal, definite, probable and possible and not codeable (i.e. insufficient information). Only definite and probable events will be used in the final analysis. This procedure is successfully applied in a number of studies coordinated by the Julius Center, e.g. in the SMART study [27].
Left ventricular hypertrophy
Using transthoracic M-mode echocardiography from the parasternal long axis position, left ventricular end-diastolic diameter (LVEDD), end-systolic diameter (LVESD) as well as posterior and septal wall thickness will be determined at baseline, after 6 months, after 12 months and annually afterwards, on a midweek non-dialysis day according to a central uniform protocol. From these parameters left ventricular ejection fraction (LVEF) will be determined as LVEDD – LVESD/LVEDD, while the left ventricular mass index (LVMi) will be calculated using the formula of Devereux and Reichek [28], modified in accordance with the recommendations of the American Society of Echocardiography [29]. The ultrasound investigations will be recorded on a compact disc and analysed off-line by experienced cardiologists in a core laboratory.
Vessel wall measurements
With respect to carotid intima-media thickness (CIMT), the outcome is the change in mean common CIMT, defined as the average of the intima-media thickness measurements performed circumferentially at pre-defined angles for the near and far wall of 10 mm segments of the right and left distal common carotid arteries [30]. A limited number of centres will be involved in the CIMT measurements in this study. Centres will be trained according to a central uniform carotid ultrasound protocol. Before actually starting the study, sonographers need to be certified as outlined in the CIMT ultrasound protocol.
Measurements will be performed at baseline and then annually on a midweek non-dialysis day. The ultrasound scan is being recorded on videotape and analysed off line by a core laboratory. Quality Assurance and Quality Control procedures as existing and applied in several (inter)national trials will be implemented [31,32].
Pulse wave velocity (PWV) is determined to provide additional information on functional changes of the arterial wall [33]. The outcome measurement is the change in aortic PWV. A limited number of centres will be involved in the PWV measurements in this study. Centres will be trained according to a central uniform PWV protocol. Data are checked regularly on quality control aspects as defined in the PWV protocol as described earlier [34].
Measurements will be performed at baseline and then annually on a midweek non-dialysis day.
Nutritional state
At base-line, after 1, 2 years and at the end of the study, nutritional state is assessed by subjective global assessment (SGA), pre-albumin and dry weight [35].
Quality of life
Patient well-being will be estimated at baseline, and once a year by the Kidney Disease Quality of Life Short Form (KDQOL-SF). This version is validated in American and Dutch dialysis patients [36,37].
Laboratory assessments
Three monthly, blood samples will be drawn for routine laboratory assessments. In addition, blood samples will be taken at baseline, and after 6, 12, 18, 24, and 36 months for determinations of oxidative stress, inflammatory and endothelial function markers. Finally, a whole blood sample will be stored for future research on the effect of genetics on the response to HDF, after specific permission of the patients in the informed consent form.
Statistical methods
The results of the study will be analysed according to the 'intention to treat' principle.
Primary outcome
The primary outcome variable is the time until the occurrence of an event defined as 'all cause mortality'. Results will be presented as Kaplan-Meier curves for the two treatment arms and the difference between the two treatments will be analysed using a log-rank test. The log-rank test will be adjusted for the effect of cumulative data analyses (see below).
Secondary outcomes
CV events are considered as secondary outcome variables. They will be analysed and presented as described for the primary outcome variable.
The primary analysis of CIMT progression will employ a linear random coefficient (Laird-Ware) model using real visit days, treatment and clinical center as independent variables. for each participant, the intercept and slope of CIMT change over time is assumed to be a normally distributed random variable with different means for the two treatment groups. The mean slope for the HDF treatment group will be compared to that for the low flux group using linear contrasts and a 5% significance level. Additional exploratory analyses will evaluate the impact of including baseline CIMT, lumen diameter, and ultrasound reader as additional co-variates.
The data analytic approach to arrive at the PWV outcome variable and the LVMi outcome variable is similar to that of the CIMT outcome. Adjustments that will be taken into account in the estimates are changes in MAP and changes in heart rate, since both are closely related to PWV.
Sample size considerations
The sample size of the present study is based on the following event rates: the 3-year all cause mortality rate among subjects with ESRD is 44% based on data from the Dutch renal replacement registry (RENINE) [38]. CV mortality constitutes 40–60% of the total group of deaths, leading to a 3 year CV mortality rate of 22% in HD patients. Assuming that the incidence of non-fatal CVD is equal to the CV mortality rate (22%), the three-year incidence of fatal and non-fatal CVD is 44%. In addition, based on experience ± 8% of the ESRD patients will undergo renal transplantation yearly and as such is being censored in the trial.
Assuming that HDF will reduce all cause mortality with 20%, it has been estimated that with a two-sided alpha of 0.05 and a power of 80%, about 772 patients need to be enrolled and followed for three years. In these patients about 250 events are expected to come to a decision. Note that the total number of patients to be included cannot be specified in advance because of the planned sequential interim analyses, as described below.
Interim analysis
In this study, group sequential interim analyses will be performed to evaluate the primary outcome variable. The reason for this approach is that, on average, fewer patients are needed in the study when the expected difference in the primary outcome variable is real or when no difference of the hypothesised magnitude can be expected anymore.
Sequential analysis is a statistical approach where one conducts significance tests over time as the data are collected. Sequential analysis and its application in clinical trials have been described extensively by Whitehead [39]. Sequential design and analysis is implemented in the computer program PEST version 4 [40].
The general approach is as follows. A null hypothesis H0 and an alternative hypothesis H1 are formulated for a suitable measure θ of treatment difference. For this study with a survival type outcome variable, θ is equal to the negative of the logarithm of the hazard ratio (HR). The HR is defined as the ratio of the logarithm of the (expected) cumulative survival under HDF (= 0.648) and the logarithm of the (expected) cumulative survival under HD (= 0.56). H0 is formulated as "no difference in the occurrence of the primary endpoint between the two trial arms" or θ = 0 (i.e. HR = 1). The alternative hypothesis H1 is formulated as |θ| ≥ -log(0.75) = 0.29. Two test statistics, Z and V, can be derived depending on the type of response variable. Z is a measure of the treatment difference; for survival data Z is the observed number of events in the control group minus the expected number of events given treatment equivalence. V reflects the amount of information about θ contained in Z; for survival data V is approximately equal to a quarter of the number of events observed. The sequential analysis requires critical boundaries to be specified in advance. These boundaries depend on θ, the type I error α and the power 1-β. For each new group of patients, values of Z and V are calculated and presented graphically by plotting Z against V (see Fig. 2 for an illustration of a double-sided sequential test). Based on the path of cumulative (Z,V)-points, one of the following three decisions is made:
Figure 2 Sequential analysis. Boundaries for a double sequential triangular test with α = 0.05, power 0.80 and hazard ratio 0.75. Z is the observed number of events in the control group minus the expected number of events given treatment equivalence. V is approximately equal to a quarter of the number of events observed.
1) the upper or the lower (continuous) boundary is crossed: stop the data collection and reject the null hypothesis;
2) one of the inner wedge-shaped (dashed) boundaries is crossed: stop the data collection and accept the null hypothesis;
3) continue the data collection: the cumulative data are inadequate to draw a conclusion yet.
An independent Data Safety and Monitoring Board (DSMB) will evaluate the results of the sequential interim analyses. The DSMB consists of a biostatistician (chair), a nephrologist, an internist, and a clinical epidemiologist. The biostatistician will perform the sequential analyses. The executive committee will provide the DSMB every 2 months with the relevant database to perform the unblinded analyses. The main task of the DSMB is to decide whether the analyses provide enough evidence of either efficacy or no efficacy with respect to the primary outcome and formulates recommendations for the executive committee on the continuation of the trial. The DSMB may also offer unsolicited recommendations on the continuation of the trial, for example after publication of results of similar trials.
Conclusion
Online HDF is gaining popularity, as recent technical advances have made it possible to safely replace considerable amounts of fluid at reasonable cost. In addition, accumulating evidence indicates that the correction of the uremic state is improved by online HDF, if compared to standard HD. However, at present it is unclear whether long-term treatment with HDF ultimately results in an improved clinical outcome. Therefore, CONTRAST is initiated, a randomised controlled trial of sufficient sample size to detect differences in survival and cardiovascular events. Patients will be randomised between low-flux HD and online HDF and followed for 3 years. Over 20 Dutch dialysis centers participate in this study and approximately 800 incident and prevalent HD patients will be recruited. By April 2005, more than 150 patients were included.
Appendix
Steering committee
The steering committee consists of the primary investigators (nephrologists) of the participating centers. The members and institutions in the Netherlands are:
W. Bax, Medical Center Alkmaar, Alkmaar;
W.H. Boer, University Medical Center Utrecht, Utrecht;
H. Boom, Reinier de Graaf Hospital, Delft;
G.J. Bruinings, Slingeland Hospital, Doetinchem;
M. van Buren, Leyenburg Hospital, The Hague;
G.W. Feith, Gelderse Vallei Hospital, Ede;
A.B. Geers, St Antonius Hospital, Nieuwegein;
J.O. Groeneveld, Onze Lieve Vrouwe Gasthuis, Amsterdam;
H.W. van Hamersvelt, University Medical Center St Radboud, Nijmegen;
F. de Heer, Maasland Hospital, Sittard;
B.C. van Jaarsveld, Dianet Dialysis Centers, Utrecht;
M.G. Koopman, Academic Medical Center, Amsterdam;
A.T. Lavrijssen, Oosterschelde Hospital, Goes;
C.J. Konings, Catharina Hospital, Eindhoven;
M.I. Koolen, Jeroen Bosch Hospital, 's Hertogenbosch;
T.K. Kremer Hovinga, Martini Hospital, Groningen;
W.H. van Kuijk, VieCuri Medical Center, Venlo;
J.J. Offerman, Isala Clinics, Zwolle;
L.J. Reichert, Rijnstate Hospital, Arnhem;
P.L. Rensma, St Elisabeth Hospital, Tilburg;
C.E. Siegert, Sint Lucas Andreas Hospital, Amsterdam;
P.J. van de Ven, Rijnmond-Zuid Medical Center, Rotterdam;
M.G. Vervloet, VU Medical Center, Amsterdam.
In Norway:
H.J. Kloke, Haukeland Hospital, Bergen.
Executive committee
P.J. Blankestijn, nephrologist, University Medical Center Utrecht, Utrecht (chair);
M.L. Bots, epidemiologist, University Medical Center Utrecht, Utrecht;
M.A. van den Dorpel, nephrologist, Rijnmond-Zuid Medical Center, Rotterdam;
M.P. Grooteman, nephrologist, VU Medical Center, Amsterdam;
M.J. Nubé, nephrologist, Medical Center Alkmaar, Alkmaar;
E.L. Penne, research physician, University Medical Center Utrecht, Utrecht;
P.M. ter Wee, nephrologist, VU Medical Center, Amsterdam (chair).
Data Safety and Monitoring Board
D.E. Grobbee, epidemiologist, University Medical Center Utrecht, Utrecht;
A.J. Rabelink, nephrologist, Leiden University Medical Center, Leiden;
C.D. Stehouwer, internist, Academic Hospital Maastricht, Maastricht;
I. van der Tweel, biostatistician, Center for Biostatistics, Utrecht University, Utrecht.
Event committee
P.A. Doevendans, cardiologist, University Medical Center Utrecht, Utrecht;
L.J. Kapelle, neurologist, University Medical Center Utrecht, Utrecht;
G. Ligtenberg, nephrologist, Utrecht;
F. Stam, internist, VU Medical Center, Amsterdam;
G. Veen, cardiologist, VU Medical Center, Amsterdam;
M.C. Visser, neurologist, VU Medical Center, Amsterdam;
F.L. Visseren, internist, University Medical Center Utrecht, Utrecht;
W. Wisselink, vascular surgeon, VU Medical Center, Amsterdam.
Advisory committee
P. Boer, biochemist, University Medical Center Utrecht, Utrecht;
M.J. Cramer, cardiologist, University Medical Center Utrecht, Utrecht;
O. Kamp, cardiologist, VU Medical Center, Amsterdam;
B. van Rossum, cardiologist, VU Medical Center, Amsterdam;
C. Schalkwijk, biochemist, VU Medical Center, Amsterdam;
C.D. Stehouwer, internist, Academic Hospital Maastricht, Maastricht;
T. Teerlink, biochemist, VU Medical Center, Amsterdam.
Trial coordinating centre
M.L. Bots, epidemiologist, University Medical Center Utrecht, Utrecht;
L. ten Brinke, CRA, Hans Mak Institute, Naarden;
M.P. Grooteman, nephrologist, VU Medical Center, Amsterdam;
E.A. Ram, project manager, University Medical Center Utrecht, Utrecht.
Project management and data management is provided by the Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
ELP has drafted the manuscript and has contributed to the design. PJB, MLB, MAD, MPG, MJN and PMW have designed the trial and have been involved in revising the article. IT has made contributions to the study design and has drafted the statistical methods section of this manuscript. All authors have approved the final manuscript.
Acknowledgements
CONTRAST has been made possible by grants from the Dutch Kidney Foundation (Nierstichting Nederland grant C02.2019), Fresenius Medical Care Netherlands, and Gambro Corporation. Additional support has been received from the Dr E.E. Twiss Fund, Roche Netherlands, and the International Society of Nephrology/Baxter Extramural Grant Program.
==== Refs
Parfrey PS Foley RN The clinical epidemiology of cardiac disease in chronic renal failure J Am Soc Nephrol 1999 10 1606 1615 10405218
Goodman WG Goldin J Kuizon BD Yoon C Gales B Sider D Wang Y Chung J Emerick A Greaser L Elashoff RM Salusky IB Coronary-artery calcification in young adults with end-stage renal disease who are undergoing dialysis N Engl J Med 2000 342 1478 1483 10816185 10.1056/NEJM200005183422003
Cheung AK Sarnak MJ Yan G Dwyer JT Heyka RJ Rocco MV Teehan BP Levey AS Atherosclerotic cardiovascular disease risks in chronic hemodialysis patients Kidney Int 2000 58 353 362 10886582 10.1046/j.1523-1755.2000.00173.x
Himmelfarb J Stenvinkel P Ikizler TA Hakim RM The elephant in uremia: oxidant stress as a unifying concept of cardiovascular disease in uremia Kidney Int 2002 62 1524 1538 12371953 10.1046/j.1523-1755.2002.00600.x
Vanholder R Argiles A Baurmeister U Brunet P Clark W Cohen G De Deyn PP Deppisch R Descamps-Latscha B Henle T Jorres A Massy ZA Rodriguez M Stegmayr B Stenvinkel P Wratten ML Uremic toxicity: present state of the art Int J Artif Organs 2001 24 695 725 11817319
Zimmermann J Herrlinger S Pruy A Metzger T Wanner C Inflammation enhances cardiovascular risk and mortality in hemodialysis patients Kidney Int 1999 55 648 658 9987089 10.1046/j.1523-1755.1999.00273.x
Vanholder R Glorieux G Lameire N Uraemic toxins and cardiovascular disease Nephrol Dial Transplant 2003 18 463 466 12584262 10.1093/ndt/18.3.463
Beerenhout CH Kooman JP Luik AJ Jeuken-Mertens SG Van Der Sande FM Leunissen KM Optimizing renal replacement therapy--a case for online filtration therapies? Nephrol Dial Transplant 2002 17 2065 2070 12454212 10.1093/ndt/17.12.2065
Lornoy W Becaus I Billiouw JM Sierens L Van Malderen P D'Haenens P On-line haemodiafiltration. Remarkable removal of beta2-microglobulin. Long-term clinical observations Nephrol Dial Transplant 2000 15 Suppl 1 49 54 10737167
Lin CL Yang CW Chiang CC Chang CT Huang CC Long-term on-line hemodiafiltration reduces predialysis beta-2-microglobulin levels in chronic hemodialysis patients Blood Purif 2001 19 301 307 11244190 10.1159/000046958
Lin CL Huang CC Yu CC Yang HY Chuang FR Yang CW Reduction of advanced glycation end product levels by on-line hemodiafiltration in long-term hemodialysis patients Am J Kidney Dis 2003 42 524 531 12955680 10.1016/S0272-6386(03)00747-9
Van Tellingen A Grooteman MP Schoorl M ter Wee PM Bartels PC Schoorl M Van Der PT Nube MJ Enhanced long-term reduction of plasma leptin concentrations by super-flux polysulfone dialysers Nephrol Dial Transplant 2004 19 1198 1203 14993491 10.1093/ndt/gfh122
Ward RA Schmidt B Hullin J Hillebrand GF Samtleben W A comparison of on-line hemodiafiltration and high-flux hemodialysis: a prospective clinical study J Am Soc Nephrol 2000 11 2344 2350 11095657
Bammens B Evenepoel P Verbeke K Vanrenterghem Y Removal of the protein-bound solute p-cresol by convective transport: a randomized crossover study Am J Kidney Dis 2004 44 278 285 15264186
Van Tellingen A Grooteman MP Bartels PC Van Limbeek J Van Guldener C Wee PM Nube MJ Long-term reduction of plasma homocysteine levels by super-flux dialyzers in hemodialysis patients Kidney Int 2001 59 342 347 11135089 10.1046/j.1523-1755.2001.00496.x
Minutolo R Bellizzi V Cioffi M Iodice C Giannattasio P Andreucci M Terracciano V Di Iorio BR Conte G De Nicola L Postdialytic rebound of serum phosphorus: pathogenetic and clinical insights J Am Soc Nephrol 2002 13 1046 1054 11912265
Zehnder C Gutzwiller JP Renggli K Hemodiafiltration--a new treatment option for hyperphosphatemia in hemodialysis patients Clin Nephrol 1999 52 152 159 10499310
Gil C Lucas C Possante C Jorge C Gomes F Candeias M Lages H Arranhado E Ferreira A On-line haemodiafiltration decreases serum TNFalpha levels in haemodialysis patients Nephrol Dial Transplant 2003 18 447 448 12543911 10.1093/ndt/18.2.447
Ward RA Ultrapure dialysate: a desirable and achievable goal for routine hemodialysis Semin Dial 2000 13 378 380 11130260 10.1046/j.1525-139x.2000.00103.x
Horl WH Hemodialysis membranes: interleukins, biocompatibility, and middle molecules J Am Soc Nephrol 2002 13 Suppl 1 S62 S71 11792764
Canaud B Bragg-Gresham JL Marshall MR Desmeules S Gillespie BW Depner TA Klassen P Port FK Patients recieving hemodiafiltration or hemofiltration have lower mortality risk than patients recieving hemodialysis without replacement fluid (HD) in Europe: The Dialysis Outcomes and Practice Patterns Study (DOPPS) [abstract] J Am Soc Nephrol 2003 14 31A
Locatelli F Marcelli D Conte F Limido A Malberti F Spotti D Comparison of mortality in ESRD patients on convective and diffusive extracorporeal treatments. The Registro Lombardo Dialisi E Trapianto Kidney Int 1999 55 286 293 9893138 10.1046/j.1523-1755.1999.00236.x
Wizemann V Lotz C Techert F Uthoff S On-line haemodiafiltration versus low-flux haemodialysis. A prospective randomized study Nephrol Dial Transplant 2000 15 Suppl 1 43 48 10737166
Locatelli F Mastrangelo F Redaelli B Ronco C Marcelli D La Greca G Orlandini G Effects of different membranes and dialysis technologies on patient treatment tolerance and nutritional parameters. The Italian Cooperative Dialysis Study Group Kidney Int 1996 50 1293 1302 8887291
Ledebo I On-line hemodiafiltration: technique and therapy Adv Ren Replace Ther 1999 6 195 208 10230890
Canaud B Bosc JY Leray H Stec F Microbiological purity of dialysate for on-line substitution fluid preparation Nephrol Dial Transplant 2000 15 Suppl 2 21 30 11051034 10.1093/ndt/15.suppl_1.21
Simons PC Algra A Bots ML Grobbee DE van der GY Common carotid intima-media thickness and arterial stiffness: indicators of cardiovascular risk in high-risk patients. The SMART Study (Second Manifestations of ARTerial disease) Circulation 1999 100 951 957 10468526
Devereux RB Reichek N Echocardiographic determination of left ventricular mass in man. Anatomic validation of the method Circulation 1977 55 613 618 138494
Schiller NB Shah PM Crawford M DeMaria A Devereux R Feigenbaum H Gutgesell H Reichek N Sahn D Schnittger I . Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. American Society of Echocardiography Committee on Standards, Subcommittee on Quantitation of Two-Dimensional Echocardiograms J Am Soc Echocardiogr 1989 2 358 367 2698218
Oren A Vos LE Uiterwaal CS Grobbee DE Bots ML Cardiovascular risk factors and increased carotid intima-media thickness in healthy young adults: the Atherosclerosis Risk in Young Adults (ARYA) Study Arch Intern Med 2003 163 1787 1792 12912713 10.1001/archinte.163.15.1787
Crouse JRIII Grobbee DE O'Leary DH Bots ML Evans GW Palmer MK Riley WA Raichlen JS Measuring Effects on intima media Thickness: an Evaluation Of Rosuvastatin in subclinical atherosclerosis--the rationale and methodology of the METEOR study Cardiovasc Drugs Ther 2004 18 231 238 15229392 10.1023/B:CARD.0000033645.55138.3d
Bots ML Evans GW Riley W Meijer R McBride KH Paskett ED Helmond FA Grobbee DE The Osteoporosis Prevention and Arterial effects of tiboLone (OPAL) study: design and baseline characteristics Control Clin Trials 2003 24 752 775 14662281 10.1016/S0197-2456(03)00096-5
Blacher J Guerin AP Pannier B Marchais SJ Safar ME London GM Impact of aortic stiffness on survival in end-stage renal disease Circulation 1999 99 2434 2439 10318666
Lebrun CE van der Schouw YT Bak AA de Jong FH Pols HA Grobbee DE Lamberts SW Bots ML Arterial stiffness in postmenopausal women: determinants of pulse wave velocity J Hypertens 2002 20 2165 2172 12409954 10.1097/00004872-200211000-00015
Locatelli F Fouque D Heimburger O Drueke TB Cannata-Andia JB Horl WH Ritz E Nutritional status in dialysis patients: a European consensus Nephrol Dial Transplant 2002 17 563 572 11917047 10.1093/ndt/17.4.563
Hays RD Kallich JD Mapes DL Coons SJ Carter WB Development of the kidney disease quality of life (KDQOL) instrument Qual Life Res 1994 3 329 338 7841967 10.1007/BF00451725
Korevaar JC Merkus MP Jansen MA Dekker FW Boeschoten EW Krediet RT Validation of the KDQOL-SF: a dialysis-targeted health measure Qual Life Res 2002 11 437 447 12113391 10.1023/A:1015631411960
Registratie Nierfunctievervanging Nederland (RENINE) Statistisch Jaarverslag 2002. 2002 Rotterdam
Whitehead J The design and analysis of sequential clinical trials 1997 Revised 2nd edition Chichester, John Wiley & Sons
MPS Research Unit Pest 4: Operating Manual. 2000 Reading (UK): University of Reading
| 15907201 | PMC1156925 | CC BY | 2021-01-04 16:47:32 | no | Curr Control Trials Cardiovasc Med. 2005 May 20; 6(1):8 | utf-8 | Curr Control Trials Cardiovasc Med | 2,005 | 10.1186/1468-6708-6-8 | oa_comm |
==== Front
Global HealthGlobalization and Health1744-8603BioMed Central London 1744-8603-1-81595339710.1186/1744-8603-1-8Short ReportThe World Summit on Sustainable Development: reaffirming the centrality of health von Schirnding Yasmin [email protected] Sustainable Development and Healthy Environments Cluster World Health Organisation Geneva Switzerland2005 10 5 2005 1 8 8 8 1 2005 10 5 2005 Copyright © 2005 von Schirnding; licensee BioMed Central Ltd.2005von Schirnding; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The World Summit on Sustainable Development (WSSD) was held in Johannesburg in 2002 to review progress since the Rio conference in 1992, and to agree a new global deal on sustainable development. Unlike its predecessor, it was primarily concerned with implementation rather than with new treaties and targets, although a number of new targets were agreed, for example one on sanitation. Failure to agree a target on renewable energy was regarded as a major disappointment of the conference. While relatively modest in its achievements, and with difficulties in achieving consensus in key areas such as energy, trade, finance and globalisation, WSSD nevertheless succeeded in placing sustainable development back on the political agenda, giving new impetus, in particular to the environment and development needs of Africa, with a strong focus on local issues like household energy, water and sanitation. Health was singled out as one of five priority areas, along with water, energy, agriculture and biodiversity, and was devoted a separate chapter in the resulting Plan of Implementation, which highlighted a range of environmental health issues as well as issues relating to health services, communicable and non-communicable diseases. A number of new partnerships were formed at WSSD, including the Healthy Environments for Children Alliance (HECA) launched by WHO, which will form an important platform for implementation. The Commission on Sustainable Development has been designated main responsibility for monitoring and follow up, with its programme of work reorganised to focus on thematic clusters of issues. From the perspective of health, WSSD must be seen as a reaffirmation of the central place of health on the sustainable development agenda, and in the broader context of a process which began in Rio and was given added impetus with the Monterrey Financing for Development conference and the World Trade Organisation meeting held in Doha. Translating policies into action at all levels- global to local – remains the single biggest challenge in the years that lie ahead.
==== Body
The world summit on sustainable development: reaffirming the centrality of health
It will be years before we can assess the true impact of the world's largest UN Summit, which brought together governments, the private sector and civil society to agree on a new global deal on sustainable development. Over one hundred Heads of State and Governments addressed the World Summit on Sustainable Development (WSSD) held in August 2002 in Johannesburg, South Africa [1], and over twenty two thousand people participated, including more than ten thousand delegates and eight thousand representatives of NGOs and civil society. It had as its key aims to review progress since the UN Conference on Environment and Development held in Rio in 1992 and to recommend measures to strengthen implementation of Agenda 21 (the global programme of action on sustainable development) and other outcomes of the Rio conference.
The term sustainable development, as originally defined by the 1987 World Commission on Environment and Development [2] (the "Brundtland Commission") was meant to entail "development that meets the needs of the present without compromising the ability of future generations to meet their own needs". It was coined as part of an effort to bring environmental issues into the mainstream of development, recognizing that in order to address the escalating problems related to the environment, the root causes which lay in the broader development process and the global economic system needed to be addressed. As originally articulated "sustainable" captured the environmental issues (assumed to centre on the needs of future generations) while "development" captured the economic/poverty issues (assumed to centre on the needs of the present generation). The concept has since been broadened, in recognition of the non-environmental aspects of sustainability, and the non-economic aspects of development [3].
In some ways we are only now beginning to judge the success of the Rio conference held ten years earlier. In comparison with the WSSD, which was largely concerned with implementation rather than with new visions, treaties and agreements, Rio led to a dramatic new paradigm shift in thinking on sustainable development, and to new legally binding agreements such as those concerned with biodiversity, and with climate change. Over the ensuing years however, there has occurred a dwindling in high-level political interest and engagement with sustainable development issues.
While WSSD in comparison with Rio was more modest in its immediate achievements, it nevertheless resulted in placing sustainable development back on the political and public agenda. New impetus was given to global action to protect the environment and fight poverty, with the development needs of Africa identified for special attention and support by the international community. A significant departure from Rio was a greater concern with social and economic issues-perhaps not surprising given that the conference was hosted by South Africa. There was also a stronger emphasis on local, as opposed to global, issues – for example with issues such as household energy, water and sanitation, rather than with the global problems associated with climate change which received so much attention in Rio.
The major outcomes of WSSD included a negotiated Plan of Implementation, a Political Declaration and a number of implementation partnerships and initiatives [1]. New targets and agreements were negotiated in a number of important areas, for example in sanitation. Previous agreements such as those relating to the achievement of the Millenium Development Goals [4], were also reaffirmed, making the Johannesburg Plan of Implementation a somewhat eclectic mix of new and past agreements and affirmations, albeit with many important implications for health.
One of the difficulties however was that so many of the key global issues – AIDS, biodiversity, climate change, trade – had their own conference processes, treaty and other mechanisms in place, such that WSSD could only serve to reaffirm these rather than cut new ground. The preparatory meetings preceding Johannesburg were characterized by difficult negotiations and attempts to achieve consensus on key aspects of the plan, particularly on energy, trade, finance and globalization. While consensus was eventually achieved in Johannesburg, contentious issues were often solved by falling back on previously agreed positions.
A separate chapter was included on globalization in the Plan of Implementation (strongly pushed for by the G77/China), which emphasized the need for successful completion of the Doha round of trade negotiations and the implementation of the Monterrey Consensus negotiated at the International Conference on Financing for Development held in Monterrey, Mexico. In addition, the final chapter dealing with means of implementation addressed also health-related commitments related to trade. Issues of contention included the mobilization of financial resources, and the EUs agricultural subsidies. Critics characterized the failure to go beyond Doha to reduce trade-distorting energy and agricultural subsidies in the rich countries as a shortcoming of the summit.
The issue of setting a time-bound target to reverse the trend in natural resource degradation caused disagreement among countries, as did references to the precautionary principle and the ecosystem approach, and the principle of "Common but Differentiated Responsibilities" agreed to in Rio. Other issues which were controversial included debate about the need for stronger governance at the international and national levels (the former emphasized by developing countries and the latter by developed countries), the role of the UN in follow up to WSSD, partnerships and their possible modalities, and the relationship between human rights and environmental protection.
Centrality of Health
A significant departure from Rio was that health issues featured centrally in WSSD, reflecting increased recognition of health as a resource for, and as an indicator of, sustainable development. Already in 1992 the Rio Declaration stated that "Human beings are at the centre of concerns for sustainable development. They are entitled to a healthy and productive life in harmony with nature." This stressed the important interlinkages between the social, economic and environmental pillars of sustainable development, all of which are underpinned by good health. Further, Chapter 6 of Agenda 21 emphasized the fundamental commitment within sustainable development to "protecting and promoting human health" [5].
At the WSSD there was a greater emphasis on development sectors, and health was singled out by the UN Secretary General as one of five priority issues, in what became known as the "WEHAB" initiative, with emphasis on water, energy, health, agriculture and biodiversity [6,7]. Health was devoted a separate chapter in the negotiated Plan of Implementation [8], and health issues permeated the text throughout. A key message in the wide ranging health agenda at WSSD was that sustainable development cannot be achieved where there is a high prevalence of debilitating illnesses, and the health of the population cannot be maintained without a healthy environment.
Particular emphasis was placed on health issues in relation to environment and poverty concerns. At least a quarter of the global burden of disease can be attributed to environmental factors, many of them poverty-related. The Johannesburg agenda reflected a major shift in recent thinking which has occurred regarding the role of health in poverty reduction and development. Health is far more central to poverty reduction than previously thought, and that realization is now beginning to shape national and global policies [9].
While there were no major breakthroughs in health nor dramatic new agreements reached, a key aspect of significance was recognition of the importance of health in the context of environment, water, energy, agriculture, biodiversity and other issues. Indeed the conference called for a stronger emphasis on health and environment linkages – a move initiated by the Canadians with strong backing from WHO and UNEP.
Environmental Health Issues
Improving access to safe water, sanitation, clean air, improved waste management and sound management of chemicals were among the key issues which received special attention in Johannesburg. Of particular note was the call for increasing access to sanitation to improve human health and reduce infant and child mortality, prioritising water and sanitation in national sustainable development strategies and poverty reduction strategies where they exist. A new target was eventually agreed, namely to halve by the year 2015 the proportion of people who do not have access to basic sanitation. This new target complements the Millenium Development Goal on access to safe drinking water, and was the subject of much acrimonious debate.
Another important agreement was to aim, by 2020, to use and produce chemicals in ways that lead to the minimisation of significant adverse effects on human health and the environment...taking into account the precautionary approach. Also called for was the promotion of reduction of risks posed by heavy metals that are harmful to human health and the environment.
An agreement to diversify energy supply and substantially increase the global share of renewable energy sources with the objective of increasing its contribution to total energy supply was significant (especially for its health implications), even though the Summit failed to reach agreement on a specific time-bound target. There was also an agreement to improve access to reliable and affordable energy services for sustainable development and resources sufficient to achieve the Millennium Development Goals. Many countries however, viewed the failure to set targets to increase the percentage of the world's power generated by renewable energy sources as the Summit's most significant missed opportunity. The EU reacted by announcing its intention to develop renewable energy sources according to a set timetable with like-minded countries.
Other aspects of note were the call to enhance health education (with the objective of achieving improved health literacy on a global basis by 2010), and an emphasis on capacity building to better assess health and environment linkages. The need to strengthen occupational health programmes was highlighted, as was the necessity of reducing air pollution exposures and related health impacts (including through use of cleaner fuels, modern pollution control techniques and reducing dependance on traditional fuel sources for cooking and heating), as well as controlling lead exposure through eliminating lead in petrol, paints and other sources of human exposure.
Communicable and Non-Communicable Diseases
Reflecting the knowledge that has accumulated about how ill-health creates and perpetuates poverty, triggering a vicious cycle which hampers economic and social development, the WSSD was also concerned with addressing the main causes of avoidable death in low-income countries. These include HIV/AIDS, malaria, tuberculosis (TB), childhood infectious diseases, maternal and perinatal conditions, nutritional deficiencies and tobacco-related illnesses. There was a strong reaffirmation of targets and goals previously agreed, with emphasis on vulnerable groups such as women and children. While AIDS received due attention, it failed in many respects to be fully recognised as a key development issue at WSSD. This was somewhat surprising given previous concerns that AIDS might dominate the entire WSSD agenda.
A call was made to strengthen the capacity of health care systems to deliver basic health services to all, aimed at improving access to essential drugs, immunisation services, vaccines and medical technology, improving maternal and obstetric care and reproductive and sexual health. These commitments were made with an emphasis on meeting the Millenium Development Goals related to health, including reducing maternal and child mortality. One aspect of the health chapter that proved to be among the most contentious of the Summit and which was of the last to receive agreement on, related to references to "health care and services". There was concern among some that this could be construed to include abortion services. Specific measures to combat and treat HIV/AIDS, malaria, TB and other diseases were called for, with special emphasis placed on the need to mobilise financial resources and to support the Global Fund to Fight AIDS, TB and Malaria.
While many countries continue to see their development efforts hampered by the burden of communicable diseases, at the same time they are faced with the rising incidence of non-communicable diseases (NCDs). The rapid rise of NCDs is also threatening economic and social development as well as the lives and health of millions of people. They represent a major health challenge to global development in the coming century. Low- and middle-income countries suffer the greatest impact, and the rapid increase in these diseases disproportionately affects poor and disadvantaged populations; contributing to widening health gaps between and within countries. In this regard Johannesburg called for programmes to combat non-communicable diseases, mental health, injuries and violence and associated risk factors such as tobacco, alcohol, unhealthy diets and lack of physical activity. In addition, a ten-year framework of programmes in support of regional and national initiatives to accelerate the shift towards sustainable consumption and production (aimed at promoting social and economic development within the carrying capacity of ecosystems) was agreed, with industrialised countries taking the lead.
The Call for Integrated Strategies and Partnerships
Other aspects emphasised in the Plan of Implementation included the need to integrate health concerns into strategies, policies and programmes for poverty eradication and sustainable development, implementation of the WHO Health for All strategy [10], and creating more effective national and regional policy responses to environmental threats to human health, as well as encouraging health promoting production and consumption policies.
Addressing the underlying determinants of health through intersectoral efforts is key to ensuring sustained health improvements and ecologically sustainable development [11]. In this regard Johannesburg called for increased action from the international community, NGOs, the private sector and local communities to implement sustainable development objectives through partnerships and alliances at all levels- global to local. This represented a major departure from previous UN conferences and was strongly pushed by the US amid initial opposition from many in the G77 bloc, who feared that this would result in an abdication of responsibility away from governments in favour of the private sector and donor interests. Over 220 partnerships (including 16 in health) with 235 million dollars in resources were identified in advance of the summit and around 60 partnerships were announced during the Summit by a variety of countries, with many more announced outside of the formal proceedings.
With health unquestionably recognised as an intersectoral issue the health sector will have to seriously deliberate on its changing role in this complex international landscape. In many parts of the world intersectoral approaches and partnerships have been successfully developed to tackle particular diseases, both communicable (infectious) and non-communicable [3]. Much progress has been made in forging closer links between national health care and other sectors, particularly through local and national intersectoral health and development planning; increased use of planning tools such as health impact assessment procedures; integrated monitoring and surveillance systems; and improved health information systems and indicators [3].
Many countries have instituted new policy and planning frameworks over the past decade, and have developed tools to make health and environment concerns an integral part of the planning process. For example at the national level, National Environmental Health Action Plans have been developed and at the local level, Local Agenda 21 and related activities such as the WHO Healthy Cities Movement, and UN Habitat and UNEPs Sustainable Cities Movement [12] have been important developments.
Effective health, environment and sustainable development policies and programmes depend however on convenient access to information about a large variety of hazards, ranging from biological hazards in food and water, to chemical hazards such as pesticides, to various physical and social factors. This is necessary if health authorities are to effectively discharge their responsibility to protect public health. But it also serves to clarify the extent to which health hazards are attributable to environmental conditions and/or to the activities of sectors other than health.
In general, knowledge of environment and health risks is segmented, and incomplete. Mechanisms to ensure coordination at national, regional and local levels regarding health effects assessment and the development of adequate reporting systems, are commonly lacking. Equally, mechanisms are frequently not in place to ensure that such information, once obtained, is transmitted to the various relevant sectors for action. Integrated databases on development hazards, environmental exposures and health, are urgently required. Well-developed health-and-environment information systems, based on relevant data sets, are essential if scientific monitoring information is to be provided in support of policy and decision-making, planning and evaluation [13]. This is one of the key challenges to the health sector in taking forward the sustainable development agenda.
The Way Forward
The Summit called on all countries to take immediate steps to formulate national sustainable development strategies and to begin implementation efforts by 2005, with international cooperation supporting the special needs of developing countries. It recommended that Governments immediately enact and enforce clear and effective laws that support sustainable development, develop and strengthen the necessary infrastructure and promote public participation in implementation. Most implementation efforts will take place at the local, national and regional levels, with Governments bearing the primary responsibility.
The main responsibility for monitoring and reviewing progress in carrying out the Johannesburg decisions falls however with the UN Commission on Sustainable Development, which is a functional commission of the UNs Economic and Social Council (ECOSOC). It was set up in 1992 to ensure effective follow-up of the UN Conference on Environment and Development held in Rio. The WSSD called for a strengthened Commission to play a larger role in promoting implementation, including by facilitating partnership initiatives and the sharing of best practices. The Commission has also been charged with developing indicators that will help to determine the state of conditions around the world and will be the basis for discussions on how to overcome obstacles to implementation [14].
The Commission on Sustainable Development has consequently been revitalized and its programme of work reorganised along a multi-year programme of work divided into seven two-year cycles (the first year devoted to a review of progress and the second to policy recommendations), with each cycle focusing on selected thematic clusters of issues. The thematic clusters of issues will be addressed in an integrated manner, taking account of the economic, social and environmental dimensions of sustainable development. Issues being addressed during the first cycle are water, sanitation and human settlements, while in 2006/2007 issues to be addressed are energy, industrial development, air pollution/atmosphere and climate change.
Already there has been much work underway at country level and the various partnerships created have been actively pursuing their respective agendas. This includes the Healthy Environments for Children Alliance (HECA) which was launched by WHO in conjunction with a number of other UN agencies, NGOs and governments [15]. While the Commission has responsibility for tracking partnerships, the partnerships themselves are voluntary and the Commission cannot hold them accountable through the same formal process to monitor government action. The UN has also created new mechanisms to help coordinate efforts in the area of water and energy, namely "UN Water" and "UN Energy" which will help to improve consistency, coherence and cooperation within the UN system.
From the perspective of health, importantly, Johannesburg must be seen as a reaffirmation of the central place of health on the sustainable development agenda [16]. It must also be seen in the broader context of a process which began in Rio and which was given added impetus with the Monterrey Financing for Development Conference and the World Trade Organization meeting held in Doha. The health sector and its allies in other sectors must now muster the necessary resources and commitment to follow through on the sustainable development agenda, and translate policy into action which will ensure a more sustainable path of development for future generations to come.
==== Refs
United Nations World Summit on Sustainable Development Johannesburg 2002
World Commission on Environment and Development Our Common Future 1987 Oxford University Press, Oxford
von Schirnding Y Mulholland C Health in the Context of Sustainable Development. Background Document for WHO Meeting: Making Health Central to Sustainable Development: Planning the Health Agenda for the WSSD. Oslo, Norway WHO, Genca; 2002 29 November–1 December 2001. WHO/HDE/HID02.6
United Nations Millenium Assembly 55th Session New York 2000
United Nations Agenda 21: Programme of Action for Sustainable Development. UN, New York; 1993
Annan K Towards a Sustainable Future New York: the American Museum of Natural History's Annual Environmental Lecture 2002
United Nations A Framework for Action on Health and the Environment WEHAB Working Group UN, New York; 2002
United Nations World Summit on Sustainable Development The Johannesburg Plan of Implementation UN, New York; 2002
World Health Organisation Macroeconomics and Health: Investing in Health for Development – Report of the Commission on Macroeconomics and Health 2001 WHO, Geneva
World Health Organisation Health-for-All in the 21st Century 1998 WHO, Geneva
von Schirnding YER Intersectoral Action for Health: Addressing Health and Environment Concerns in Sustainable Development 1997 WHO, Geneva
von Schirnding YER Addressing Health and Environment Concerns in Sustainable Development, with Special Reference to Participatory Planning Initiatives such as Healthy Cities Ecosystem Health 1997 3 220 228
von Schirnding YER Health in Sustainable Development Planning: the Role of Indicators 2002 WHO, Geneva WHO/HDE/HID/02.11
United Nations Department of Economic and Social Affairs The Road from Johannesburg: What was Achieved and the Way Forward UNDESA, New York; 2003
World Health Organisation Healthy Environments for Children Alliance. Report of the Secretariat 2004 WHO, Geneva
von Schirnding YER Health and Sustainable Development: Can we rise to the Challenge? Lancet 2002 360 632 7 12241950 10.1016/S0140-6736(02)09777-5
| 15953397 | PMC1156926 | CC BY | 2021-01-04 16:39:02 | no | Global Health. 2005 May 10; 1:8 | utf-8 | Global Health | 2,005 | 10.1186/1744-8603-1-8 | oa_comm |
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-331587782410.1186/1477-7525-3-33ResearchInternal consistency reliability is a poor predictor of responsiveness Puhan Milo A [email protected] Dianne [email protected] Gordon H [email protected] Diane [email protected]ünemann Holger J [email protected] Horten Centre, University Hospital, Postfach Nord, 8091 Zurich, Switzerland2 Department of Clinical Epidemiology and Biostatistics, McMaster University, 1200 Main St. W., Health Sciences Centre, Room 2C12, Hamilton, Ontario, L8N 3Z5 Canada3 Department of Medicine, McMaster University, Health Sciences Centre, Hamilton, Ontario, L8N 3Z5 Canada2005 9 5 2005 3 33 33 7 4 2005 9 5 2005 Copyright © 2005 Puhan et al; licensee BioMed Central Ltd.2005Puhan et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Whether responsiveness represents a measurement property of health-related quality of life (HRQL) instruments that is distinct from reliability and validity is an issue of debate. We addressed the claims of a recent study, which suggested that investigators could rely on internal consistency to reflect instrument responsiveness.
Methods
516 patients with chronic obstructive pulmonary disease or knee injury participating in four longitudinal studies completed generic and disease-specific HRQL questionnaires before and after an intervention that impacted on HRQL. We used Pearson correlation coefficients and linear regression to assess the relationship between internal consistency reliability (expressed as Cronbach's alpha), instrument type (generic and disease-specific) and responsiveness (expressed as the standardised response mean, SRM).
Results
Mean Cronbach's alpha was 0.83 (SD 0.08) and mean SRM was 0.59 (SD 0.33). The correlation between Cronbach's alpha and SRMs was 0.10 (95% CI -0.12 to 0.32) across all studies. Cronbach's alpha alone did not explain variability in SRMs (p = 0.59, r2 = 0.01) whereas the type of instrument was a strong predictor of the SRM (p = 0.012, r2 = 0.37). In multivariable models applied to individual studies Cronbach's alpha consistently failed to predict SRMs (regression coefficients between -0.45 and 1.58, p-values between 0.15 and 0.98) whereas the type of instrument did predict SRMs (regression coefficients between -0.25 to -0.59, p-values between <0.01 and 0.05).
Conclusion
Investigators must look to data other than internal consistency reliability to select a responsive instrument for use as an outcome in clinical trials.
==== Body
Background
Health-related quality of life (HRQL) instruments should demonstrate adequate test-retest reliability, cross-sectional and longitudinal validity before investigators use them to assess outcomes in research studies. Whether responsiveness, the ability of an instrument to detect change in HRQL when change occurs, is a measurement property distinct from reliability and validity remains, however, controversial [1-4].
Lindeboom et al. purportedly tested the assumption that responsiveness is not a distinct measurement property, but is embodied in internal consistency reliability [5]. To investigate their hypothesis, the authors removed the item contributing most to internal consistency (as determined using Cronbach's alpha) in a step-wise fashion from the physical component of the Sickness Impact profile, the Barthel activities of daily living scale and the psychosocial domain of the Graves' ophthalmology quality of life instrument using data from three previous studies. Following each step-wise removal, they recalculated Cronbach's alpha and the standardised response means (SRM, change score divided by standard deviation of change score) of the remaining items. They then assessed the correlation of these new Cronbach's alphas with the new SRMs and observed strong associations (Spearman rank correlation coefficients between 0.90 and 1.00). They concluded that internal consistency reliability adequately reflects an instrument's responsiveness and that investigators can use the two entities interchangeably.
The first conceptual problem with the approach Lindeboom et al. chose is that they looked at the correlation of internal consistency reliability and responsiveness within single studies and instruments only. However, this approach does not take into account that responsiveness depends on the type of an intervention while internal consistency reliability does not. Most HRQL measures may be very reliable, but internal consistency reliability has nothing to do with the therapy that is producing the change. In contrast, if an intervention targets aspects of HRQL that are specifically covered by a disease-specific instrument, for example, responsiveness is likely to be high. If the effect of another intervention targeting aspects other than those covered by the instrument, responsiveness will be lower. Thus the within study approach does not take into account that responsiveness is not a fixed measurement property.
Another important issue to consider is the influence of other determinants of an instrument's responsiveness such as the type of instrument, generic or disease-specific. There is ample evidence that responsiveness depends on the type of instrument. [6-9] Lindeboom's within instrument approach does not take into account this issue.
Finally, if the within instrument approach with step-wise deconstruction of domains is used, one would expect step-wise decreases of internal consistency reliability, responsiveness and other measurement properties such as cross-sectional validity for the following reasons. Internal consistency reliability is reduced when the items contributing most to internal consistency reliability are removed because the error term in the denominator increases. For the same reason, responsiveness deteriorates if the number of items is decreased[10]. Thus it is likely to see a parallel decline of internal consistency reliability and responsiveness even if there is no relationship between these two measurement properties. Indeed using Lindeboom's approach one would expect high correlations between internal consistency reliability and other measurement properties such as cross-sectional validity and could consequently conclude that they are all embodied in internal consistency reliability. The assessment of the relationship between internal consistency reliability and responsiveness should include entire domains, as they were developed, validated and used in research.
Having considered the methodological challenges and constraints above, we analysed the relationship between internal consistency reliability and responsiveness of entire domains across different instruments and studies using data from several of our previous studies.
Methods
Studies
A priori we defined the following eligibility criteria to ensure an unbiased selection of datasets as possible and to ensure that it was theoretically possible to detect a correlation between internal consistency reliability and responsiveness if one existed. We applied the following criteria:
1. Studies must have longitudinal follow-up with a baseline assessment and at least one follow-up assessment completed by the CLARITY research group (McMaster University, Hamilton, Ontario, Canada) within the last five years.
2. Studies must have investigated an intervention of established effectiveness that induces changes in HRQL.
3. Studies must include ≥ 2 multi-item HRQL instruments that allow calculation of Cronbach's alpha and instruments within a study must have different degrees of responsiveness (e.g. generic versus disease-specific) to ensure variability in responsiveness. We expected variability in Cronbach's alpha to be limited to values ≥ 0.60 because only those are generally accepted to represent sufficient internal consistency reliability [3].
Statistical analysis
We calculated Cronbach's alpha using baseline scores for each domain of each HRQL instrument or for the total instrument if domains did not exist. Similarly, for each domain or for a total score we calculated SRMs (change score divided by standard deviation of change score).
We calculated the correlation between Cronbach's alpha and the corresponding SRM using Pearson correlation coefficients across all studies and for each study separately. We then built linear regression models with the SRM as the dependent variable and Cronbach's alpha as the independent variable. Since the type of instrument (generic or disease-specific) affects the SRM [6-9], we introduced the type of instrument as a covariate into the regression models. For all regression models, we adjusted for possible clustering for data originating from the same group of patients (for example, patients from one study providing data for eight domains of the Short-Form Survey 36) by using the cluster function of STATA. We performed all statistical analysis with STATA for Windows version 8.2 (StataCorp, College Station, Texas, USA).
Results
Eligible Studies
The following four studies met the eligibility criteria:
Study 1 [11]
This prospective study measured HRQL in 85 patients with chronic obstructive pulmonary disease (COPD) before and after participation in Canadian inpatient respiratory rehabilitation programs similar to many inpatient programs worldwide [12]. All patients completed the interviewer-administered Chronic Respiratory Questionnaire (CRQ) including individualised and standardised dyspnea questions. In addition, patients completed the St. Georges Respiratory Questionnaire (SGRQ) and the Short-From Survey 36 (SF-36) [13] at the beginning and end of the rehabilitation program.
Study 2 [14]
This was a prospective randomised study of 177 patients with COPD before and after respiratory rehabilitation in Canada and the United States. We randomised patients to complete either the interviewer or self-administered CRQ [11,15]. All patients answered the individualised and standardised dyspnea questions of the CRQ. Patients also completed the SGRQ and the SF-36 at the beginning and end of the rehabilitation program.
Study 3 [16,17]
This prospective study enrolled 71 patients with COPD following a respiratory rehabilitation program at four cites in Switzerland, Germany and Austria. We also randomised patients to complete either the interviewer or self-administered CRQ as in study 2 [11,15] and all patients answered the individualised and standardised dyspnea questions of the CRQ. Patients also completed the SF-36 [18] at the beginning and end of the rehabilitation program.
Study 4 [19]
This prospective study enrolled patients undergoing anterior crucial ligament reconstruction (study 4a, n = 66) and knee arthroscopy (study 4b, n = 117) to determine their ability to recall pre-operative quality of life and functional status. Patients completed the disease-specific Anterior Crucial Ligament Quality Of Life questionnaire (ACL-QOL) [20] (study 4a) or the Western Ontario Meniscal Evaluation Tool (WOMET) [21] (study 4b) as well as the International Knee Documentation Committee (IKDC) Subjective Form [22], the Knee Injury and Osteoarthritis Outcome Score (KOOS) [23] and the SF-36 pre- and one year post-operatively.
Relationship between internal consistency reliability and responsiveness
Tables 1 and 2 show the reliability coefficients and standardised response mean for each study and instrument. The mean Cronbach's alpha across all studies was 0.83 (SD 0.08, range 0.61 to 0.97) and the mean standardised response mean was 0.59 (SD 0.33, range -0.08 to 1.45).
Table 1 Internal consistency reliability and responsiveness. Studies 1 and 2
Study Instrument and domain Internal consistency reliability¶ Standardised response mean#
Study 1 CRQ-IA
Dyspnea individualised 0.78 1.16
Dyspnea standardised 0.82 0.81
Fatigue 0.86 0.85
Emotional function 0.88 0.76
Mastery 0.75 0.95
SGRQ
Symptoms 0.63 0.27
Activities 0.75 0.49
Impact 0.73 0.69
Total 0.84 0.66
SF-36
Physical Functioning 0.80 0.83
Role Physical 0.77 0.54
Bodily Pain 0.93 0.29
General Health 0.64 0.13
Vitality 0.80 0.65
Social Functioning 0.74 0.63
Role Emotional 0.84 0.50
Mental Health 0.84 0.41
Study 2 CRQ-IA
Dyspnea individualised 0.77 0.66
Dyspnea standardised 0.86 0.50
Fatigue 0.80 0.25
Emotional function 0.90 0.24
Mastery 0.92 0.38
CRQ-SA
Dyspnea individualised 0.83 0.84
Dyspnea standardised 0.86 0.69
Fatigue 0.84 0.60
Emotional function 0.89 0.56
Mastery 0.87 0.70
SGRQ
Symptoms 0.72 0.34
Activities 0.81 0.33
Impact 0.80 0.46
Total 0.88 0.51
SF-36
Physical Functioning 0.87 0.31
Role Physical 0.91 0.19
Bodily Pain 0.83 0.22
General Health 0.76 0.22
Vitality 0.86 0.32
Social Functioning 0.90 0.21
Role Emotional 0.92 0.07
Mental Health 0.87 0.15
¶Cronbach's alpha; # Change score (follow-up minus baseline) / standard deviation of change score; * Pearson correlation coefficient; CRQ-IA interviewer-administered CRQ; CRQ-SA self-administered CRQ
Table 2 Internal consistency reliability and responsiveness. Studies 3 and 4
Study Instrument and domain Internal consistency reliability¶ Standardised response mean#
Study 3 CRQ-IA
Dyspnea 0.73 0.94
Fatigue 0.81 0.96
Emotional function 0.77 1.35
Mastery 0.76 1.09
CRQ-SA
Dyspnea 0.78 0.86
Fatigue 0.83 1.38
Emotional function 0.89 1.00
Mastery 0.86 0.86
SF-36
Physical composite score 0.61 0.59
Mental composite score 0.70 0.48
Study 4a ACL-QOL 0.96 1.45
IKDC 0.86 1.17
KOOS
Symptoms 0.74 0.64
Pain 0.89 0.74
Function 0.96 0.59
Sports 0.92 0.83
QOL 0.88 1.15
SF-36
Physical Function 0.90 0.64
Role Physical 0.79 0.67
Bodily Pain 0.80 0.52
General Health 0.72 0.11
Vitality 0.87 0.44
Social Functioning 0.81 0.50
Role Emotional 0.82 0.33
Mental Health 0.76 0.34
Study 4b WOMET 0.97 0.88
IKDC 0.90 0.85
KOOS
Symptoms 0.82 0.85
Pain 0.93 0.95
Function 0.97 0.82
Sports 0.94 0.66
QOL 0.90 0.71
SF-36
Physical Function 0.93 0.59
Role Physical 0.93 0.40
Bodily Pain 0.88 0.60
General Health 0.85 -0.08
Vitality 0.86 0.08
Social Functioning 0.84 0.25
Role Emotional 0.88 0.12
Mental Health 0.79 0.21
¶Cronbach's alpha
# Change score (follow-up minus baseline) / standard deviation of change score
* Pearson correlation coefficient
Figure 1 shows the relationship between Cronbach's alpha and SRM across all studies. The correlation coefficient was 0.10 (95% CI -0.12 to 0.32). When we analysed each study separately, correlation coefficients ranged from -0.17 to 0.62 (Figure 2).
Figure 1 Relationship between internal consistency reliability and responsiveness, all studies Relationship between Cronbach's alpha and standardised response mean for 79 domains or total scores of health-related quality of life instruments and symptoms scales. The data come from four studies including 333 patients with chronic obstructive pulmonary disease following a pulmonary rehabilitation and 183 patients with knee injury undergoing anterior crucial ligament reconstruction or knee arthroscopy.
Figure 2 Relationship between internal consistency reliability and responsiveness, per study.
Table 3 shows the regression equations to predict the SRM from Cronbach's alpha. In an analysis of all studies including internal consistency reliability as the sole independent variable did not predict responsiveness (p = 0.59, r2 = 0.01). In contrast, an analysis that included the type of instrument showed that the generic versus specific categorisation predicted responsiveness (p = 0.01, r2 = 0.37). Analysing the studies separately showed similar results (Figure 2). Only in study 4 was Cronbach's alpha a significant predictor in unadjusted analyses. Even in this case, when we introduced the type of instrument into the model, Cronbach's alpha was no longer a significant predictor.
Table 3 Prediction of responsiveness from internal consistency reliability
Study Dependent variable Independent variables Constant Regression coefficient (p-value) R2
All studies SRM Cronbach's alpha 0.25 0.41 (0.59) 0.01
SRM Cronbach's alpha 0.66 0.13 (0.83) 0.37
Type of instrument* -0.40 (0.01)
Study 1 SRM Cronbach's alpha -0.09 0.90 (0.30) 0.07
SRM Cronbach's alpha -0.07 1.04 (0.18) 0.32
Type of instrument* -0.25 (0.04)
Study 2 SRM Cronbach's alpha 0.94 -0.64 (0.46) 0.03
SRM Cronbach's alpha 0.52 -0.02 (0.98) 0.48
Type of instrument* -0.29 (<0.01)
Study 3 SRM Cronbach's alpha -0.51 1.88 (0.11) 0.29
SRM Cronbach's alpha 1.42 -0.45 (0.74) 0.60
Type of instrument* -0.59 (0.05)
Study 4a SRM Cronbach's alpha -1.81 2.94 (0.013) 0.39
SRM Cronbach's alpha -0.46 1.58 (0.15) 0.60
Type of instrument* -0.37 (0.03)
Study 4b SRM Cronbach's alpha -2.61 3.52 (0.03) 0.32
SRM Cronbach's alpha -0.43 1.36(0.22) 0.74
Type of instrument* -0.48 (<0.01)
SRM: standardised response mean
* Disease-specific or generic
Discussion
We assessed the relationship between internal consistency reliability and responsiveness and found no evidence to support the claim that investigators can use them interchangeably. In general, internal consistency reliability is a poor predictor of responsiveness. Consistent with previous findings [6], we showed that in contrast to Cronbach's alpha, a significant predictor of responsiveness is whether the instrument is a generic or a disease-specific HRQL instrument.
Our findings contradict those presented by Lindeboom et al. We suspect that these differences are largely due to differences in conceptual and, thus, statistical approaches. In particular, Lindeboom's within instrument and within study approach fails to take into account that responsiveness depends on the type of instrument and on the intervention that produces change in HRQL. In our analyses, we evaluated the relationship between internal consistency reliability and responsiveness across instruments and studies.
One might argue that our failure to demonstrate a relationship between Cronbach's alpha and the SRM results from the limited variability in Cronbach's alpha across the instruments and their domains. Indeed, this limited variability in part explains the lack of relationship. Nevertheless, when choosing instruments for clinical trials, investigators will face Cronbach's alpha coefficients such as those shown in Table 1 and 2. If they rely on these results to predict responsiveness, they will be misled. In particular, some domains with very high Cronbach's alpha coefficients (SF-36 bodily pain, 0.93; CRQ IA emotional function 0.90) had low responsiveness (SRMs of 0.29 and 0.24, respectively).
Strengths of our study include the definition of a priori criteria to ensure an unbiased selection of studies that ensure large variability responsiveness creating the greatest potential to detect a relationship if one existed. Furthermore, the inclusion of very different patient populations (chronic lung disease and knee pathology) and the consistency of results across these studies and populations enhances the generalizability of our study. Replication in other populations would further strengthen our conclusions.
Conclusion
Our study demonstrates that internal consistency reliability is a poor predictor of responsiveness and that both conceptual and statistical evidence exists to support the argument that they are distinct measurement properties of evaluative instruments.
Authors' contributions
MAP, DB, GHG and HJS designed the study and wrote the study protocol. MAP, GHG, HJS and DB collected the data. MAP, DB and DHA performed the statistical analysis. MAP drafted and DB, GHG, DHA and HJS critically revised the manuscript. All authors read and approved the final manuscript.
==== Refs
Guyatt GH Feeny DH Patrick DL Measuring health-related quality of life Ann Intern Med 1993 118 622 629 8452328
Hays RD Hadorn D Responsiveness to change: an aspect of validity, not a separate dimension Qual Life Res 1992 1 73 75 1301117 10.1007/BF00435438
Streiner D Norman G Health measurement scales Oxford Medical Publications 2003 Third Oxford, Oxford University Press
Terwee CB Dekker FW Wiersinga WM Prummel MF Bossuyt PM On assessing responsiveness of health-related quality of life instruments: guidelines for instrument evaluation Qual Life Res 2003 12 349 362 12797708 10.1023/A:1023499322593
Lindeboom R Sprangers MA Zwinderman AH Responsiveness: a reinvention of the wheel? Health Qual Life Outcomes 2005 3 8 15691385 10.1186/1477-7525-3-8
Wiebe S Guyatt G Weaver B Matijevic S Sidwell C Comparative responsiveness of generic and specific quality-of-life instruments J Clin Epidemiol 2003 56 52 60 12589870 10.1016/S0895-4356(02)00537-1
de Torres JP Pinto-Plata V Ingenito E Bagley P Gray A Berger R Celli B Power of Outcome Measurements to Detect Clinically Significant Changes in Pulmonary Rehabilitation of Patients With COPD(*) Chest 2002 121 1092 1098 11948037 10.1378/chest.121.4.1092
Guyatt GH King DR Feeny DH Stubbing D Goldstein RS Generic and specific measurement of health-related quality of life in a clinical trial of respiratory rehabilitation J Clin Epidemiol 1999 52 187 192 10210235 10.1016/S0895-4356(98)00157-7
Singh SJ Sodergren SC Hyland ME Williams J Morgan MD A comparison of three disease-specific and two generic health-status measures to evaluate the outcome of pulmonary rehabilitation in COPD Respir Med 2001 95 71 77 11207021 10.1053/rmed.2000.0976
Moran LA Guyatt GH Norman GR Establishing the minimal number of items for a responsive, valid, health-related quality of life instrument J Clin Epidemiol 2001 54 571 579 11377117 10.1016/S0895-4356(00)00342-5
Schunemann HJ Griffith L Jaeschke R Goldstein R Stubbing D Austin P Guyatt GH A Comparison of the Original Chronic Respiratory Questionnaire With a Standardized Version Chest 2003 124 1421 1429 14555575 10.1378/chest.124.4.1421
Lacasse Y Brosseau L Milne S Martin S Wong E Guyatt GH Goldstein RS Pulmonary rehabilitation for chronic obstructive pulmonary disease Cochrane Database Syst Rev 2004 CD003793
Ware JEJ Sherbourne CD The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection
1 Med Care 1992 30 473 483 1593914
Schunemann HJ R G Mador MJ D. MK E S M.A. P L G B. G P A R. C G.H. G A Randomized Trial to Evaluate the Self-Administered Standardized CRQ Europ Respir J 2005 25 31 40 10.1183/09031936.04.00029704
Guyatt GH Berman LB Townsend M Pugsley SO Chambers LW A measure of quality of life for clinical trials in chronic lung disease Thorax 1987 42 773 778 3321537
Puhan MA Behnke M Frey M Grueter T Brandli O Lichtenschopf A Guyatt GH Schunemann HJ Self-administration and interviewer-administration of the German Chronic Respiratory Questionnaire: instrument development and assessment of validity and reliability in two randomised studies
1 Health Qual Life Outcomes 2004 2 1 14713317 10.1186/1477-7525-2-1
Puhan MA Behnke M Laschke M Lilhtenschopf A Brandli O Guyatt GH Schunemann HJ Self-administration and standardisation of the chronic respiratory questionnaire: A randomised trial in three German-speaking countries Respiratory Medicine 2004 98 342 350 15072175 10.1016/j.rmed.2003.10.013
Bullinger M German translation and psychometric testing of the SF-36 Health Survey: preliminary results from the IQOLA Project. International Quality of Life Assessment Soc Sci Med 1995 41 1359 1366 8560303 10.1016/0277-9536(95)00115-N
Bryant D Norman G Stratford P Marx R Walter S al KA Can patients undergoing knee surgery provide an accurate rating of pre-operative quality of life, functional status and general health at 2 weeks post-operatively? A matter of efficiency Manuscript in progress 2005
Mohtadi N Development and validation of the quality of life outcome measure (questionnaire) for chronic anterior cruciate ligament deficiency Am J Sports Med 1998 26 350 359 9617395
Griffin S Huffman H Bryant D Kirkley A The development and validation of a quality of life measurement tool for patients with meniscal pathology: the Western Ontario Meniscal Evaluation Tool (WOMET) Manuscript in progress 2005
Irrgang JJ Anderson AF Boland AL Harner CD Kurosaka M Neyret P Richmond JC Shelborne KD Development and validation of the international knee documentation committee subjective knee form Am J Sports Med 2001 29 600 613 11573919
Roos EM Roos HP Lohmander LS Ekdahl C Beynnon BD Knee Injury and Osteoarthritis Outcome Score (KOOS)--development of a self-administered outcome measure J Orthop Sports Phys Ther 1998 28 88 96 9699158
| 15877824 | PMC1156927 | CC BY | 2021-01-04 16:38:14 | no | Health Qual Life Outcomes. 2005 May 9; 3:33 | utf-8 | Health Qual Life Outcomes | 2,005 | 10.1186/1477-7525-3-33 | oa_comm |
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-341590452710.1186/1477-7525-3-34ReviewHealth-related quality of life measurement in pediatric clinical practice: An appraisal and precept for future research and application Varni James W [email protected] Tasha M [email protected] Mariella M [email protected] Department of Pediatrics, College of Medicine, Department of Landscape Architecture and Urban Planning, College of Architecture, Texas A&M University, 3137 TAMU, College Station, TX 77843-3137 USA2 Department of Anesthesiology, University of Washington, 1945 NE Pacific Street, Seattle, WA 98195 USA3 Department of Psychology, College of Liberal Arts, Texas A&M University, College Station, TX 77843 USA2005 16 5 2005 3 34 34 5 5 2005 16 5 2005 Copyright © 2005 Varni et al; licensee BioMed Central Ltd.2005Varni et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Health-related quality of life (HRQOL) measurement has emerged as an important health outcome in clinical trials, clinical practice improvement strategies, and healthcare services research and evaluation. HRQOL measures are also increasingly proposed for use in clinical practice settings to inform treatment decisions. In settings where HRQOL measures have been utilized with adults, physicians report such measures as useful, some physicians alter their treatment based on patient reports on such instruments, and patients themselves generally feel the instruments to be helpful. However, there is a dearth of studies evaluating the clinical utility of HRQOL measurement in pediatric clinical practice. This paper provides an updated review of the literature and proposes a precept governing the application of pediatric HRQOL measurement in pediatric clinical practice. Utilizing HRQOL measurement in pediatric healthcare settings can facilitate patient-physician communication, improve patient/parent satisfaction, identify hidden morbidities, and assist in clinical decision-making. Demonstrating the utility of pediatric HRQOL measurement in identifying children with the greatest needs, while simultaneously demonstrating the cost advantages of providing timely, targeted interventions to address those needs, may ultimately provide the driving force for incorporating HRQOL measurement in pediatric clinical practice.
health-related quality of lifeclinical practicepediatricschildrenhealthcare services
==== Body
Review
The last decade has evidenced a dramatic increase in the development and utilization of health-related quality of life (HRQOL) measures in an effort to improve patient health and determine the value of healthcare services [1]. Most research efforts on HRQOL in medical decision-making thus far have focused on measuring patient-reported outcomes (PROs), healthcare utilization, and the incorporation of HRQOL assessments in clinical trials with the goal of identifying group discrimination between treatments. Unfortunately, a dearth of empirical peer-reviewed studies exists that have evaluated the clinical utility of HRQOL measurement in pediatric clinical practice.
The use of HRQOL measurement in clinical practice has focused primarily on adult patients. In settings where HRQOL instruments have been employed with outpatient adults, physicians report that such measures are useful, and that they use PROs to alter treatment. Patients generally perceive the instruments to be helpful in communicating their healthcare needs to their physicians [2,3]. Further, a number of position papers have suggested that the routine assessment of HRQOL in clinical practice may facilitate best practices [4,5], with the caveat that the routine use of HRQOL assessment should be considered within the overall framework of evidence-based practice [6].
This paper serves to provide an updated review of the literature and to build a rationale for the incorporation of HRQOL measurement into pediatric clinical practice by evaluating both conceptual arguments for its benefit and existing empirical evidence of outcomes for its use in clinical practice. Since a paucity of data exists on these issues in pediatric clinical care, the adult literature is reviewed and implications for pediatric settings are addressed. Logistical issues related to incorporating such assessment procedures into clinical settings are then examined, and recommendations are made for future research and practice.
Rationale and evidence for the incorporation of HRQOL measurement in clinical practice
Facilitation of patient-physician communication
One potential benefit of incorporating HRQOL assessment into clinical practice is the impact on patient-physician communication [7]. The adult literature supports the idea that both patients and physicians believe HRQOL feedback to physicians assists communication [3,8-11]. For example, patients are often willing to discuss HRQOL with their physicians even when they are not experiencing significant impairments [8,12]. Unfortunately, although patients desire to share HRQOL information, physicians are not likely to ask [8,12-14]. In fact, research evidence suggests that physician behaviors (e.g., neglecting to ask about HRQOL issues) can impede a patient's willingness to introduce HRQOL concerns, that physicians vary greatly in their ability to elicit important information from patients, and that patients vary in their ability or desire to articulate it [15,16]. Patients are, however, generally willing to complete a HRQOL questionnaire at outpatient visits [8,12]. In this way, HRQOL measures can be a catalyst for communication in that they can objectively identify areas of concern that can then be discussed at the point of healthcare service.
Unfortunately, there are no randomized controlled studies in the published literature that have examined the impact of pediatric patient self-report or parent proxy-report HRQOL measurement on patient-physician communication. In pediatric practice, communication can be a key issue, as young children may not have the language skills necessary to accurately communicate their symptoms or feelings verbally. For these children, child self-report HRQOL measures that are sensitive to children's language and cognitive development may indicate problem areas, even those unrelated to the patient's chief complaint, that need to be further explored. In this way, one benefit of standardized HRQOL measurement in pediatric practice is the communication of healthcare needs by the child who may not be able to otherwise accurately report symptoms or other problems with daily functioning.
Improvement in patient satisfaction
The adult literature suggests that patient satisfaction with medical care has been related to physician discussion of psychosocial topics and active physician elicitation of patient opinion [17,18]. Street and colleagues [13] reported that physicians who asked about pain, health perceptions, vitality, and role limitations due to physical problems were evaluated more favorably than physicians who did not. Another study by Brody et al.[19] reported that patients of physicians who received mental health feedback reported more satisfaction with their physician than patients in a control group. In the pediatric literature, no research has examined patient satisfaction and HRQOL assessment in pediatric practice. While it is unclear whether children would respond similarly, parents of pediatric patients are likely to report satisfaction or dissatisfaction with medical care based on their perceptions of their child's HRQOL outcomes as a result of treatment.
Identification of hidden morbidities in pediatric clinical care
The adult literature has indicated a disparity in physician and patient ratings of health status [20-23]. Physicians frequently underestimate or fail to detect psychosocial and functional disabilities in patients, and often even high levels of psychological distress are undetected and untreated [22,24-28]. Studies with adult patients, however, indicate that physicians who receive HRQOL data have significantly higher detection rates for psychological problems [29-32], and for functional and HRQOL impairments [10,33]. With this information a physician may be better able to make appropriate referrals and tailor interventions to the specific needs of his/her patients.
Research in pediatric healthcare has also demonstrated the continuing under-identification of psychosocial problems, termed the "new hidden morbidity" [34], in routine pediatric practice [35,36]. Given the marked under-detection of these problems by physicians, HRQOL measures can serve as standardized screening instruments for identifying physical and psychosocial health concerns from the perspectives of both the child and parent at the point of service that pediatricians might otherwise overlook [37]. However, there are no existing randomized controlled pediatric trials that address this issue. Nevertheless, it seems reasonable to assume that pediatricians are likely to be assisted by standardized HRQOL measurement in the identification of hidden morbidities that might not otherwise be detected.
Impact on clinical decision-making
In addition to traditional morbidity and mortality outcomes, evidence-based clinical decision-making requires validated patient-reported outcomes, particularly for chronic health conditions [38]. The impact of a HRQOL measurement instrument on clinical decision-making can be tested under the working hypothesis that HRQOL measurement and feedback to physicians must occur at the point of service in order to improve health outcomes [37]. Clinical decision-making may also benefit from a serial screening approach, which can assist with tracking changes in HRQOL over time. This longitudinal monitoring of HRQOL can help physicians evaluate the effectiveness of treatment or monitor disease status and associated symptoms such as pain, fatigue, coughing, wheezing, and the like.
In the adult literature, it has been estimated that up to 52% of physicians make management decisions (e.g., medication changes, ordering lab tests) on the basis of patient-reported HRQOL [6,8,39]. One study found that functional status feedback provided new information for approximately 25% of patients, resulting in specific clinical management actions by physicians in approximately 40% of those patients [3]. HRQOL measurement may also assist physicians with the referral process, in that they may be more likely to refer patients to mental health specialists or other health professionals as a result of this information [8,31].
Although these studies with adult patients suggest that improving detection through the use of HRQOL measurement does affect clinical care in various ways, the data regarding changes to clinical practice, after receiving HRQOL feedback, is inconclusive. For example, some adult studies have reported a lack of statistically significant differences between groups of physicians who received HRQOL feedback versus those who did not with regard to changes in prescription of medications, ordering of tests, referrals, or resource utilization [2,6,10,39-41].
The available evidence suggests that simply providing primary care physicians with HRQOL screening information without specific resource and management suggestions may not be sufficient in changing healthcare provider behavior [2,6]. Some physicians even express concern that HRQOL instruments measure some problems they are not prepared to treat; that is, they question the extent to which they could intervene, practically and financially, when certain problems are detected beyond their time and resources to address adequately [42,43]. The failure to produce changes in treatment may also be related to physician's limited experience with HRQOL measures or difficulty in translating HRQOL data into specific interventions to improve functioning.
The clinical utility of HRQOL measurement in clinical practice is likely to be enhanced by the availability of measures that are easy to use and interpret, physician education regarding the purpose and utility of HRQOL measurement, and specific management guidelines for deficits in HRQOL [6,44,45]. In fact, some investigators have concluded that the provision of targeted recommendations in conjunction with HRQOL scores are needed to yield beneficial effects of HRQOL measurement information on outcomes of care [8,40,46,47]. Providing practitioners with easy to use linkage tools (e.g., to appropriate referrals, resources, and evidence-based interventions) may greatly improve the integration of HRQOL data into clinical practice. However, linking HRQOL data to appropriate treatment recommendations, with subsequent improvements in HRQOL outcomes, has not yet been demonstrated empirically in pediatrics.
Thus, while the above studies provide valuable information about the impact of HRQOL on clinical decision making in adult patients, virtually no pediatric studies exist. One study with the Pediatric Quality of Life Inventory™ (PedsQL™) Generic Core Scales and Rheumatology Module in pediatric rheumatology [48], however, found that for those patients for whom the pediatrician reported that HRQOL measurement at the point of service influenced her clinical decision-making, subsequent HRQOL outcomes demonstrated significant improvement in comparison to patients in which HRQOL assessment was not utilized [37]. Nevertheless, this single empirical study only highlights the need for further research.
Improvement in patient outcomes over time
A major hypothesized benefit to incorporating HRQOL measurement into routine clinical practice is the potential for identifying symptoms and problems that may result in improved patient outcomes over time. The findings are nevertheless mixed, with some physician-feedback intervention studies in the adult literature demonstrating no significant differences in patient outcomes such as functional status or health status between intervention and control groups [2,40,44,48,49], while other studies has yielded some support for improvement in mental health (e.g., depression) [10,48] and role functioning [10] following feedback interventions.
Overall, the results from the adult literature suggest that routine implementation of standardized HRQOL screening may be a necessary but not sufficient condition for enhancing patients' HRQOL. For example, Rubenstein and colleagues [47] reported that providing physicians with patient functional status information in conjunction with management and community resource suggestions resulted in better mental health and social activity scores. Thus, incorporating specific resource management suggestions, such as appropriate referrals and tailored treatments, may enhance the potential potency of HRQOL measurement by providing physicians with variable options to identified problems.
In sum, the adult research on the integration of HRQOL measurement into clinical practice has demonstrated that the provision of HRQOL data to physicians is more likely to impact detection and diagnosis, but is generally not sufficient in changing physicians' management behaviors unless the HRQOL findings are linked to the available resources necessary to improve HRQOL outcomes. The general failure to produce changes in treatment may be related to physician's limited experience with such measures or the difficulty of healthcare professionals in translating HRQOL data into specific interventions to improve functioning. Finally, there may be significant barriers to the implementation of HRQOL measures into clinical practice, as discussed next.
Barriers to incorporating standardized HRQOL measurement in clinical practice
Barriers such as perceived lack of time, money, and human resources needed to collect, analyze, and interpret HRQOL data, as well as the lack of ongoing computer support for storing and retrieving data, can greatly impede implementation of routine HRQOL measurement in clinical practice [6,50]. Thus, while HRQOL measurement logically appears to have potential utility in pediatric clinical practice, there are a number of perceived barriers to use that must be addressed, including that: (1) measures may be perceived as too long, impractical, and difficult to score for a clinic with multiple time constraints; (2) physicians may not be convinced that HRQOL is related to clinical signs and symptoms, which traditionally have been the focus of clinical intervention; (3) implementation of HRQOL measurement may be perceived too costly, requiring additional resources such as staff time and computer scoring systems; (4) implementation of HRQOL measurement is perceived as potentially interfering with clinic operation, and (5) the information provide by HRQOL instruments is perceived to be already available through conventional testing methods (e.g., semistructured questions about health and symptoms).
Concerns about resources
Based on their semi-structured interviews of physicians concerning practice-based functional health assessment, McHorney and Bricker [42] found that physicians desired a point-of-service data collection process that did not interrupt the flow of care; perceived barriers included insufficient staff and computing resources, as well as the need for summary output to be able to be immediately implemented. A study in an outpatient oncology clinic addressed several of these issues, reporting that completion, scoring, and printing of patient-reported HRQOL data could be conducted during waiting room time [9]. Additionally, summaries provided to physicians did not lengthen average consultation time, and anecdotally, some physicians expressed the belief that the summary increased their efficiency [9]. Thus, intervention studies comparing duration of physician consultation have found that having access to HRQOL data does not necessarily increase the amount of time spent with patients [9,10].
The application of computer-assisted assessment technology to the measurement of patient self-report and parent proxy-report in pediatric clinical care may reduce a number of the perceived barriers associated with the administration, completion, and scoring of standardized HRQOL instruments, and consequently represents one method for potentially overcoming several barriers to the use of these measures in pediatric practice. Computer-assisted HRQOL measurement as a screening methodology may facilitate the identification of areas of concern at the point of service, eliminate a lengthy interview process, and assist with targeting interventions [51-54]. For instance, a one page electronically-generated report can provide a brief summary of symptoms and problem areas for the individual patient on a real time basis, which may influence clinical decision-making at the point of service. This summary report could also generate targeted referral suggestions to facilitate the clinical decision-making process and make suggestions for targeted evidence-based healthcare based on the HRQOL findings.
Attitudinal barriers
One attitudinal barrier to the incorporation of HRQOL assessment in clinical care is clinicians' views of the relevance and value of this information in patient care. Physicians vary greatly in their degree of willingness to discuss HRQOL issues with patients, and this willingness impacts patients' behavior (e.g., patients do not discuss their HRQOL concerns with their physician unless prompted to do so) [12,16,55].
Measuring the HRQOL of children in clinical practice may also meet with some resistance by clinicians who feel that the completion of these measures represents too great a burden for children and their parents, particularly children with serious illnesses. Oftentimes, a more qualitative approach is sought, with the belief that such methods are less burdensome and intrusive than standardized quantitative methods [56]. While an understandable concern, similar concerns have been seen in the past in regards to pain measurement in children [57] and HRQOL measurement in children with newly diagnosed cancer [58]. These concerns were met with brief instruments (to reduce respondent burden), developed with focus groups and cognitive interviews (to hear the voices of children and their parents), and by careful attention to the methodological details involved in establishing the reliability and validity of these instruments [59]. Subsequent to the development, field-testing, and documentation of the measurement properties of these standardized quantitative instruments, clinical trials have been conducted targeting pain management and enhancing HRQOL through symptom reduction using these quantitative instruments as outcome measures. Thus, standardized HRQOL instruments have the potential to improve the lives of children by providing the systematic documentation of the efficacy of treatment interventions designed for symptom relief [60]. In palliative care, for instance, standardized measures have the potential to increase the accountability and the quality of the care provided by comparing healthcare institutions or practitioners, and utilizing that information to inform consumers and aid quality improvement efforts [61].
A related attitudinal barrier to use in pediatric clinical practice is the perception by some physicians that HRQOL measures are neither sufficiently associated with nor predictive of subtle changes in physiological parameters [51,62]. For example, small-to-medium correlation effect sizes were found between perceived HRQOL and Hba1c in children and adolescents with Type 1 diabetes [63], a finding consistent with the broader literature across diseases [64]. However, as succinctly summarized by McHorney [64]: "QOL scores correlate modestly at best with clinical outcomes. This finding suggests that clinical and human function are relatively independent. It does not imply that one or the other is inherently superior or correct. They simply measure different things, and using both will likely yield more information than any set alone" (p III-58). In fact, the use of physiological parameters to assess disease severity may be problematic, as many health conditions lack physiological markers, and such markers do not often apply across disease categories [65]. Furthermore, a patients' HRQOL may change in the absence of a measured biological change, or vice versa [62]. From this perspective, then, the incorporation of HRQOL measures into clinical practice should not be viewed as proxies for physiological parameters, but rather as a means of providing a more comprehensive evaluation of patient functioning across multiple life domains, particularly when meaningful physiological parameters are not available. This can be the hallmark for providing pediatric comprehensive care; multidimensional assessment leading to targeted interventions based on patient perceived needs.
As medicine shifts to a more patient-centered paradigm [38], willingness to incorporate HRQOL measurement is likely to increase. Physician education can be used to target the aforementioned attitudinal barriers, but Golden [45] notes the inadequacy of continuing medical education in changing physician behaviors and underscores the need for physician leaders who will advocate for and model the use of HRQOL measurement in clinical practice.
Finally, positive experiences with HRQOL measurement may reduce the aforementioned attitudinal barriers. Physicians in intervention studies who have used HRQOL measures report positive attitudes toward HRQOL assessment, expressing beliefs that the HRQOL data were accurate, were useful in clinical practice, could improve patient health status and satisfaction with care, provided new patient information, and facilitated physician-patient communication [2,8-10,29,39,40,44,46]. These positive experiences are a necessary but not sufficient condition for change, as detailed later.
Selection of HRQOL instruments for use in pediatric clinical practice
For HRQOL instruments to be utilized in pediatric clinical practice, they must demonstrate their utility in the clinic setting by addressing the following criteria: (1) they must be brief, yet maintain reliability and validity and provide novel, valuable information; (2) they should be well-designed both for ease of use by patients/parents and for quick, easy scoring and interpretation; and (3) they must be responsive to meaningful patient change. While a comprehensive review of specific measures and their measurement properties is beyond the scope of this paper, interested readers may find reviews by Eiser & Morse [67] and Matza [68] helpful. Below we provide a briefly summary of the relative merits of generic and disease-specific measures, child and parent reports, and the measurement properties to consider in selecting an appropriate HRQOL measure for pediatric clinical practice.
Generic and disease-specific HRQOL measures in clinical practice
There are potential advantages to an integrated modular approach combining the relative merits of generic and disease-specific HRQOL instruments [59,69,70]. A generic HRQOL measurement instrument allows for screening in healthy populations, enables standardized comparisons across pediatric chronic health conditions, and facilitates benchmarking with healthy populations [71,72]. While providing useful benchmarking data for comparisons with healthy children and across different disease groups, generic HRQOL measures may not, however, be as responsive to changes in disease-specific symptoms [73], and typically do not measure specific disease symptoms and treatment side effects of relevance to particular disease groups. In contrast, disease-specific measures (e.g., for asthma, cancer, diabetes, and arthritis) may be more sensitive to specific clinical changes in patients with a particular illness than a generic measure and may be particularly informative for pediatric chronic disease management at the individual patient level. Nevertheless, disease-specific instruments are unable to provide comparisons across diseases, including benchmarking with healthy population norms. Thus, there is an emerging perspective that for pediatric chronic health conditions, both generic and disease-specific HRQOL measures should be administered so as to gain as a more comprehensive evaluation of the patient's HRQOL.
Child self-report versus parent proxy-report in pediatric practice
Imperfect agreement between self and proxy report, termed cross-informant variance [74], has been consistently documented in the HRQOL assessment of children with chronic health conditions and healthy children [75]. The existence of cross-informant variance and the subjective nature of HRQOL indicate an essential need for reliable and valid child self-report instruments for the broadest age range possible. However, while self-report is considered the standard for measuring perceived HRQOL, there may be circumstances when the child is too young, or too ill or fatigued to complete a HRQOL instrument, and parent proxy-report may be needed in such cases. Further, it is typically parents' perceptions of their children's HRQOL that influences healthcare utilization [35,76]. Thus, an instrument should be selected that meets the need to measure the perspectives of both the child and parent since these perspectives may be independently related to healthcare utilization, risk factors, and quality of care.
Measurement properties requisite for individual patient interpretation
Key instrument properties that are important when HRQOL measurement is employed at the individual patient level include high internal consistency reliability (e.g., alpha ~ .90) and documented responsiveness to change over time as a result of treatment [77]. The establishment of clinically meaningful cut-points to identify at-risk patient status, and the documentation of the magnitude of score changes representative of clinically meaningful change, are essential measurement properties when using HRQOL instruments in clinical decision-making.
Summary
The dramatic increase in use of HRQOL measures to evaluate patient outcomes and treatment alternatives reflects an emerging shift in health care toward valuing the patient perspective. However, only a few studies, primarily with adults, have evaluated the utility of HRQOL measurement in clinical practice. This research has demonstrated some support for the benefits of HRQOL measurement in facilitating patient-physician communication, improving patient satisfaction, increasing detection of problems with functioning, informing and altering clinical management, and improving patient outcomes. However, consistency in results is generally lacking, and similar data are virtually nonexistent for pediatric patients. Further, while the provision of HRQOL measurement feedback to physicians is likely to impact detection and diagnosis, it is often insufficient to change physicians' management behaviors (and subsequently improve patient outcomes) unless HRQOL findings are linked to suggestions for specific interventions, including appropriate referral sources.
Routine use of standardized HRQOL instruments in pediatric clinical practice may be impeded by physician attitudinal barriers or concern that completion of these measures represents too great a burden. Further, physicians desire information about the effectiveness of practice-based functional health assessment and are reluctant to routinely use HRQOL measurement if evidence is lacking that it improves patient clinical outcomes. Intervention studies are needed to provide evidence-based demonstration supporting the setting aside of valuable resources needed for the integration of HRQOL measurement in pediatric clinical practice. Intervention studies that provide physicians with targeted recommendations linked to HRQOL data are especially needed.
A precept for future research and practice
What is the overarching precept or principle that should guide future research in determining the utility of HRQOL measurement in pediatric clinical practice? We propose that while doing the right thing for children is always the foremost principle, the realities of today's healthcare industry mandates that any change to current standard practice demonstrate that the change either does not cost more to the organization, or even more ideally, saves the organization costs in the future.
While cost savings are rarely the sole driving force behind change in pediatric healthcare systems (the corollary to the above is typically true, i.e., cost savings should have no negative impact on care, and ideally improve clinical care through greater efficiencies and reduced variances in the provision of care), convincing senior management to implement a change that has upfront costs is challenging at best. Children's Hospitals and pediatric healthcare systems often find themselves struggling to survive during economic downturns. Without clear economic value, as perceived by senior management, or regulatory or legislative mandate, changes in healthcare systems are rarely endorsed.
Thus, we suggest that the overarching precept for determining the likely implementation of HRQOL measurement in pediatric clinical practice requires a win-win circumstance, that is, in doing the right thing for children, the organization actually saves healthcare cost in the future, or at the very least, finds the change "revenue neutral" (i.e., the change pays for itself somehow).
What are the conditions under which such a set of circumstances could be achieved? We believe that nuggets of evidence exist in the literature that portend the set of circumstances above, although only suggestive thus far. Specifically, prospective cohort studies that have demonstrated the utility of HRQOL measurement in predicting future healthcare utilization and costs suggest the necessary but not sufficient condition for the overarching precept above. For example, in the adult literature, lower asthma-specific HRQOL was associated with higher risk of asthma-related emergency department visits or hospitalizations and higher healthcare costs during the subsequent 12 months [78]. In pediatric patients, generic HRQOL (PedsQL™) prospectively accounted for significant variance in healthcare costs at 6, 12, and 24 months [79]. Adjusted regression models that included both HRQOL scores and chronic health condition status accounted for 10.1%, 14.4%, and 21.2% of the variance in healthcare costs at 6, 12, and 24 months. HRQOL and chronic health condition status together defined a 'high risk' group, constituting 8.7% of the sample and accounting for 37.4%, 59.2%, and 62% of healthcare costs at 6, 12, and 24 months. The high risk group's per member per month healthcare costs were, on average, 12 times that of other enrollees' at 24 months.
Pediatric risk prediction is of increasing importance for healthcare providers, purchasers, payers, and policy makers. Predicting resource utilization is key to managing defined populations in a prospective payment system and for proactively case-managing those at greatest risk of poor health. If providers knew in advance, for example, at health plan enrollment which children were most likely to become ill, health care resources could be proactively targeted to those children in order to minimize or prevent that morbidity and associated costs. Being able to demonstrate the utility of pediatric HRQOL measurement in identifying children with the greatest needs, while simultaneously demonstrating the cost advantages of providing timely, targeted interventions to address those needs, may ultimately provide the driving force for incorporating HRQOL measurement in pediatric clinical practice.
Thus, the findings above suggest that pediatric HRQOL data can be used to prospectively predict healthcare costs. When combined with chronic health condition status, HRQOL data can identify an at risk group of children as candidates for proactive care coordination. The win-win proposition entails doing the right thing for children when their needs are identified at the point of healthcare contact, with the potential for healthcare cost savings in the future for the healthcare industry. It seems like a precept worth testing in the service of children and society.
List of abbreviations
HRQOL Health-Related Quality of Life
PedsQL™ Pediatric Quality of Life Inventory™
Authors' contributions
All authors contributed to the conceptualization and drafting of the manuscript. All authors read and approved the final manuscript.
==== Refs
Fayers PM Machin D Quality of life: Assessment, analysis and interpretation 2000 New York: Wiley
Rubenstein LV Calkins DR Young RT Cleary PD Fink A Kosecoff J Jette AM Davies AR Delbanco TL Brook RH Improving patient function: A randomized trial of functional disability screening Ann Intern Med 1989 111 836 842 2683917
Nelson EC Landgraf JM Hays RD Wasson JH Kirk JW The functional status of patients: How can it be measured in physicians' offices? Med Care 1990 28 1111 1126 2250496
Jenkinson C Quality of life measurement: Does it have a place in routine clinical assessment J Psychosom Res 1994 38 377 381 7965926 10.1016/0022-3999(94)90098-1
Albrecht G Using subjective health assessments in practice and policy-making Health Care Anal 1996 4 284 292 10163999 10.1002/(SICI)1099-1042(199611)4:4<284::AID-HCA194>3.0.CO;2-W
Greehnalgh J Meadows K The effectiveness of the use of patient-based measures of health in routine practice in improving the process and outcomes of patient care: A literature review J Eval Clin Pract 1999 5 401 416 10579704 10.1046/j.1365-2753.1999.00209.x
Theirac SO Forces motivating the use of health status assessment measures in clinical settings and related research Med Care 1992 30 MS15 MS22 1583929
Wagner AK Ehrenberg BL Tran TA Bungay KM Cynn DJ Rogers WH Patient-based health status measurement in clinical practice: A study of its impact on epilepsy patients' care Qual Life Res 1997 6 329 341 9248315 10.1023/A:1018479209369
Detmar SB Aaronson NK Quality of life assessment in daily clinical oncology practice: A feasibility study Eur J Cancer 1998 34 1181 1186 9849476 10.1016/S0959-8049(98)00018-5
Detmar SB Muller MJ Schornagel JH Wever LD Aaronson NK Health-related quality of life assessments and patient-physician communication: a randomized controlled trial JAMA 2002 288 3027 3034 12479768 10.1001/jama.288.23.3027
Jacobsen PB Davis K Cella D Assessing quality of life in research and clinical practice Oncology (Huntingt) 2002 16 133 139 12380963
Detmar SB Aaronson NK Wever LDV Muller M Schornagel JH How are you feeling? Who wants to know? Patients' and oncologists' preferences for discussing health-related quality of life issues J Clin Oncol 2000 18 3295 3301 10986063
Street RL Gold WR McDowell T Using health status surveys in medical consultations Med Care 1994 32 732 744 8028407
Detmar SB Muller MJ Wever LDV Schornagel JH Aaronson NK Patient-physician communication during outpatient palliative treatment visits: An observational study JAMA 2001 285 1351 1357 11255393 10.1001/jama.285.10.1351
Maguire P Faulkner A Booth K Elliott C Hillier V Helping cancer patients disclose their concerns Eur J Cancer 1996 32A 78 81 8695247 10.1016/0959-8049(95)00527-7
Beckman HB Frankel RM The effect of physician behavior on the collection of data Ann Intern Med 1984 101 692 696 6486600
Stewart MA What is a successful doctor-patient interview? A study of interactions and outcomes Soc Sci Med 1984 19 167 175 6474233 10.1016/0277-9536(84)90284-3
Bertakis KD Roter D Putnam SM The relationship of physician medical interview style to patient satisfaction J Fam Pract 1991 32 175 181 1990046
Brody DS Lerman CE Wolfson HG Caputo GC Improvement in physicians' counseling of patients with mental health problems Arch Intern Med 1990 150 993 998 2331204 10.1001/archinte.150.5.993
Wartman SA Morlock LL Malitz FE Palm E Impact of divergent evaluations by physicians and patients of patients' complaints Public Health Rep 1983 98 141 145 6602349
Slevin ML Plant H Lynch D Drinkwater J Gregory WM Who should measure quality of life, the doctor or the patient? Br J Cancer 1988 57 109 112 3348942
Calkins DR Rubenstein LV Cleary PD Davies AR Jette AM Fink A Kosecoff J Young RT Brook RH Delbanco TL Failure of physicians to recognize functional disability in ambulatory patients Ann Intern Med 1991 114 451 454 1825267
Nerenz DR Repasky DP Whitehouse FW Kahkonen DM Ongoing assessment of health status in patients with diabetes mellitus Med Care 1992 30 MS112 MS124 1583926
Brody DS Physician recognition of behavioral, psychological, and social aspects of medical care Arch Intern Med 1980 130 1286 1289 7425763 10.1001/archinte.140.10.1286
Steward MA Buck CW Physicians' knowledge of and response to patients' problems Med Care 1977 15 578 585 875502
Nelson E Conger B Douglass R Gephart D Kirk J Page R Clark A Johnson K Stone K Wasson J Zubkoff M Functional health status levels of primary care patients JAMA 1983 249 3331 3338 6602226 10.1001/jama.249.24.3331
Sprangers MA Aaronson NK The role of healthcare providers and significant others in evaluating the quality of life of patients with chronic disease: A review J Clinic Epidemiol 1992 45 743 760 10.1016/0895-4356(92)90052-O
Paykel ES Priest RG Recognition and management of depression in general practice: Consensus statement BMJ 1992 305 1198 1202 1467723
Dowrick C Does testing for depression influence diagnosis or management by general practitioners? Fam Pract 1995 12 461 465 8826066
Magruder-Habib K Zung WWK Feussner JR Improving physicians' recognition and treatment of depression in general medical care Med Care 1990 28 239 250 2179650
Mazonson PD Mathias SD Fifer SK Buesching DP Malek P Patrick DL The Mental Health Patient Profile: Does it change primary care physicians' practice patterns? J Am Board Fam Pract 1996 9 336 345 8884672
Gold I Baraff LJ Psychiatric screening in the emergency department: Its effects on physician behavior Ann Emerg Med 1989 18 875 880 2757286
Rubenstein LV McCoy JM Cope DW Barrett PA Hirsch SH Messer KS Young RT Improving patient quality of life with feedback to physicians about functional status J Gen Intern Med 1995 10 607 614 8583263
Costello EJ Edelbrock C Costello AJ Dulkan MK Burns BJ Brent D Psychopathology in pediatric primary care: The new hidden morbidity Pediatrics 1988 82 415 424 3405677
Varni JW Setoguchi Y Screening for behavioral and emotional problems in children and adolescents with congenital or acquired limb deficiencies Am J Dis Child 1992 146 103 107 1531284
Wolfe J Grier HE Klar N Levin SB Ellenbogen JM Salem-Schatz S Emanuel EJ Weeks JC Symptoms and suffering at the end of life in children with cancer N Engl J Med 2000 342 236 333 10.1056/NEJM200002033420506
Varni JW Seid M Knight TS Uzark K Szer IS The PedsQL™ 4.0 Generic Core Scales: Sensitivity, responsiveness, and impact on clinical decision-making J Behav Med 2002 25 175 193 11977437 10.1023/A:1014836921812
Sullivan M The new subjective medicine: Taking the patient's point of view on health care and health Soc Sci Med 2003 56 1595 1604 12614708 10.1016/S0277-9536(02)00159-4
Espallargues M Vaalderas HM Alonso J Provision of feedback on perceived health status to health care professionals Med Care 2000 38 175 186 10659691 10.1097/00005650-200002000-00007
Kazis LE Callahan LF Meenan RF Pincus T Health status reports in the care of patients with rheumatoid arthritis J Clin Epidemiol 1990 43 1243 1253 2243259 10.1016/0895-4356(90)90025-K
Gilbody SM House AO Sheldon T Routine administration of health related quality of life (HRQoL) and needs assessment instruments to improve psychological outcome- a systematic review Psychol Med 2002 32 1345 1356 12455933 10.1017/S0033291702006001
McHorney CA Bricker DE A qualitative study of patients' and physicians' views about practice-based functional health assessment Med Care 2002 40 1113 1125 12409856 10.1097/00005650-200211000-00012
Meadows KA Twidale F Rogers D Action research- a model for introducing standardized health assessment in general practice: An exploratory study J Eval Clin Pract 1998 4 225 239 9744710 10.1046/j.1365-2753.1998.00005.x
Wasson J Hays R Rubenstein L Nelson E Leaning J Johnson D Keller A Landgraf J Rosenkrans C The short term effect of patient health status assessment in a health maintenanace organization Qual Life Res 1992 1 99 106 1301124 10.1007/BF00439717
Golden WE Health status measurement: Implementation strategies Med Care 1992 30 MS187 MS195 1583933
Gilbody SM House AO Sheldon TA Routinely administered questionnaires for depression and anxiety: A systematic review BMJ 2001 322 406 409 11179161 10.1136/bmj.322.7283.406
Rubenstein LV McCoy JM Cope DW Improving patient functional status: A randomized trial of computer-generated resource and management suggestions Clin Res 1989 37 801A
Varni JW Seid M Knight TS Burwinkle TM Brown J Szer IS The PedsQL™ in pediatric rheumatology: Reliability, validity, and responsiveness of the Pediatric Quality of Life Inventory™ Generic Core Scales and Rheumatology Module Arthritis Rheum 2002 46 714 725 11920407 10.1002/art.10095
Calkins DR Rubenstein LV Cleary PD Davies AR Jette AM Fink A Kosecoff J Young RT Brook RH Delbanco TL Functional disability screening of ambulatory patients: A randomized controlled trial in a hospital-based group practice J Gen Intern Med 1994 9 590 592 7823232
Russak SM Croft JD Furst DE Hohlbauch A Liang MH Moreland L Ofman JJ Paulus H Simon LS Weisman M Tugwell P Evidence-Based Medicine Working Groups in Rheumatology The use of rheumatoid arthritis health-related quality of life patient questionnaires in clinical practice: Lessons learned Arthritis Rheum 2003 49 574 584 12910566 10.1002/art.11208
Carlson L Speca M Hagen N Taenzer P Computerized quality of life screening in a cancer pain clinic J Palliat Care 2001 17 46 52 11324185
O'Reilly JM Hubbard ML Lessler JT Biemer PP Turner CF Audio and video computer assisted self-interviewing: Preliminary tests of new technologies for data collection J Official Stat 1994 10 197 214
Turner CF Rogers SM Lindberg LD Pleck JH Sonenstein FL Adolescent sexual behavior, drug use, and violence: Increased reporting with computer survey technology Science 1998 280 867 873 9572724 10.1126/science.280.5365.867
Couper MP Rowe B Evaluation of a computer-assisted self-interview component in a computer-assisted personal interview survey Pub Opin Quarterly 1996 60 89 105 10.1086/297740
Levinson W Roter D Physicians' psychosocial beliefs correlate with their patient communication skills J Gen Intern Med 1995 10 375 379 7472685
Taylor KM MacDonald KA Bezjak A Ng P DePetrillo AD Physicians perspective on quality of life: An exploratory study of oncologists Qual Life Res 1996 5 5 14 8901361 10.1007/BF00435963
Varni JW Thompson KL Hanson V The Varni/Thompson Pediatric Pain Questionnaire: I. Chronic musculoskeletal pain in juvenile rheumatoid arthritis Pain 1987 28 27 38 3822493 10.1016/0304-3959(87)91056-6
Varni JW Katz ER Seid M Quiggins DJL Friedman-Bender A The Pediatric Cancer Quality of Life Invenory-32 (PCQL-32): I. Reliability and Validity Cancer 1998 82 1184 1196 9506367 10.1002/(SICI)1097-0142(19980315)82:6<1184::AID-CNCR25>3.0.CO;2-1
Varni JW Seid M Rode CA The PedsQL™: Measurement model for the Pediatric Quality of Life Inventory Med Care 1999 37 126 139 10024117 10.1097/00005650-199902000-00003
Varni JW Jacobs JR Seid M Drotar D Treatment adherence as a predictor of health-related quality of life Promoting adherence to medical treatment in chronic childhood illness: Concepts, methods, and interventions 2000 Mahwah, NJ: Erlbaum 287 305
Teno J Byock I Field M Research agenda for developing measures to examine quality of care and quality of life of patients diagnosed with life-limiting illness J Pain Symptom Manag 1999 17 75 82 10.1016/S0885-3924(98)00134-1
Frost MH Bonomi AE Ferrans CE Wong GY Hays RD Clinical Significance Consensus Meeting Group Patient, clinician, and population perspectives on determining the clinical significance of quality-of-life scores Mayo Clin Proc 2002 77 488 494 12004999
Varni JW Burwinkle TM Jacobs JR Gottschalk M Kaufman F Jones KL The PedsQL™ in Type 1 and Type 2 Diabetes: reliability and validity of the Pediatric Quality of Life Inventory™ Generic Core Scales and Type 1 Diabetes Module Diabetes Care 2003 26 631 637 12610013
McHorney CA The potential clinical value of quality of life information response to Martin Med Care 2002 40 III56 62 12064759
Stein REK Gortmaker SL Perrin EC Perrin JM Pless IB Walker DK Weitzman M Severity of illness: Concepts and measurements Lancet 1987 2 1506 1509 2892059 10.1016/S0140-6736(87)92633-X
Moore AA Siu AL Partridge JM Hays RD Adams J A randomized trial of office-based screening for common problems in older persons Am J Med 1997 102 371 378 9217619 10.1016/S0002-9343(97)00080-6
Eiser C Morse R The measurement of quality of life in children: past and future perspectives Dev Beh Peds 2001 22 248 256
Matza LS Swensen AR Flood EM Secnik K Leidy NK Assessment of health-related life in children: a review of conceptual, methodological and regulatory issues Val Health 2004 79 92 10.1111/j.1524-4733.2004.71273.x
Patrick DL Deyo RA Generic and disease-specific measures in assessing health status and quality of life Med Care 1989 27 S217 S233 2646490
Sprangers MAG Cull A Bjordal K Groenvold M Aaronson NK The European Organization for Research and Treatment of Cancer approach to quality of life assessment: Guidelines for developing questionnaire modules Qual Life Res 1993 2 287 295 8220363 10.1007/BF00434800
Varni JW Burwinkle TM Seid M Skarr D The PedsQL™ 4.0 as a pediatric population health measure: Feasibility, reliability, and validity Ambul Pediatr 2003 3 329 341 14616041 10.1367/1539-4409(2003)003<0329:TPAAPP>2.0.CO;2
Varni JW Burwinkle TM Katz ER Meeske K Dickinson P The PedsQL™ in pediatric cancer: Reliability and validity of the Pediatric Quality of Life Inventory™ Generic Core Scales, Multidimensional Fatigue Scale, and Cancer Module Cancer 2002 94 2090 2106 11932914 10.1002/cncr.10428
Guyatt GH Feeny DH Patrick DL Measuring health-related quality of life: Basic Science Review Ann Int Med 1993 70 225 230
Varni JW Katz ER Colegrove R Dolgin M Adjustment of children with newly diagnosed cancer: Cross-informant variance J Psychosoc Oncol 1995 13 23 38
Koot HM Wallander JL Eds Quality of life in child and adolescent illness: Concepts, methods and findings 2001 East Sussex, UK: Brunner-Routledge
Janicke DM Finney JW Riley AW Children's health care use: A prospective investigation of factors related to care-seeking Med Care 2001 39 990 1001 11502956 10.1097/00005650-200109000-00009
Nunally JC Bernstein IR Psychometric Theory 1994 3 New York: McGraw-Hill
Eisner MD Ackerson LM Chi F Kalkbrenner A Buchner D Mendoza G Lieu T Health-related quality of life and future health care utilization for asthma Ann Asthma Allergy Immunol 2002 89 46 55
Seid M Varni JW Segall D Kurtin PS Health-related quality of life as a predictor of pediatric healthcare costs: A two-year prospective cohort analysis Health Qual Life Outcomes 2004 2 48 15361252 10.1186/1477-7525-2-48
| 15904527 | PMC1156928 | CC BY | 2021-01-04 16:38:14 | no | Health Qual Life Outcomes. 2005 May 16; 3:34 | utf-8 | Health Qual Life Outcomes | 2,005 | 10.1186/1477-7525-3-34 | oa_comm |
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-351590452810.1186/1477-7525-3-35ResearchEffects of a self-care program on quality of life of cirrhotic patients referring to Tehran Hepatitis Center Zandi Mitra [email protected] Mohsen [email protected] Robabeh [email protected] Anooshiravan Kazem [email protected] Seyed Moayed [email protected] Kashan University of Medical Sciences and Health Services, kashan, Iran2 Tarbiat Modarres University, Tehran, Iran3 Baghiyatallah University of Medical Sciences, Tehran Hepatitis Center, Tehran, Iran2005 18 5 2005 3 35 35 11 2 2005 18 5 2005 Copyright © 2005 Zandi et al; licensee BioMed Central Ltd.2005Zandi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Chronic liver disease especially liver cirrhosis is one of the medical problems that substantially reduces the quality of life of its victims. Because of the chronic and irreversible nature of the disease, it needs self-care programs to be developed according to client's needs and to maintain their independence and sense of well-being. The purpose of this study was to determine the effects of a self-care educational program on Quality of Life (QoL) of a sample of Iranian cirrhotic patients.
Methods
A quasi-experimental study was conducted on 44 cirrhotic patients in Tehran Hepatitis Center. Longitudinal case registry and random allocation technique were used to divide the sample into experimental (n = 21) and control (n = 23) groups. Chronic liver disease questionnaire (CLDQ) was used for measuring the quality of life. The experimental group was given a questionnaire to assess their educational needs. A self-care educational program was conducted and the patients were followed for 3 months. Then the quality of life of both groups was compared using descriptive and analytical statistics.
Results
The experimental and control groups were the same concerning the effective factors on the quality of life, such as age, sex, etc (P > 0.05). There was no significant difference between QOL mean score of both groups before the intervention, however the QoL significantly improved in the experimental group after the intervention (P= 0.001), while the QoL decreased in control group.
Conclusion
The result of the present study confirmed the positive effects of the educational and self care programs on the QoL of cirrhotic patients. Extensive educational and self-care programs along with long-term follow up such as the program conducted in this study are suggested.
==== Body
Background
Cirrhosis represents a late stage of progressive hepatic fibrosis characterized by distortion of the hepatic architecture and the formation of regenerative nodules. It is generally considered to be irreversible in its advanced stages [1,2].
While alcohol abuse is the most common cause of cirrhosis in the western world, hepatitis B is the primary cause in the third world [1]. The only relatively acceptable remedy for this condition is liver transplantation, currently performed with many limitations in Iran because of complex operation technique and high expenses [3].
Cirrhosis of the liver is the third leading cause of death in people between the ages of 25 and 65 years, exceeded only by cardiovascular disease and cancer. Cirrhosis and chronic liver diseases accounted for more than 25,000 death and 373,000 hospital discharges annually in the adult in the United States. The cost of cirrhosis in terms of human suffering, financial burden, and loss of productive life is devastating [2,4,5].
Complications such as encephalopathy, ascites, bacterial peritonitis, and frequent bleeding from variceal veins dramatically alter the well being of cirrhotic patients [6] as well as their Quality of life. Studies have shown the negative effects of the disease on the patients' activities, social functioning, and emotional status [7-10]. Because of irreversible nature of the disease, and because the current therapies are not yet good and available enough, patients quality of life has moved to the forefront of patient concern.
Several studies have documented impaired health-related quality of life (HRQoL) associated with chronic hepatitis [11,12]. In 1998, Foster et al compared the Health Related QoL (HRQoL) of liver patients with viral hepatitis B and C and reported that social functioning, energy and fatigue and role limitations due to physical problems were significantly more impaired in hepatitis C patients [14]. Younossi et al found an increasing impairment of generic HRQoL with increasing disease severity; while Marchesini et al found that the most relevant determinants of impaired health status were severity of disease and muscle cramps [7,8,15].
These studies contributed substantially to our knowledge of the physical, social and mental problems of chronic liver patients. However, little is published related to the QoL of these chronic liver patients in Iran. Moreover, none of these studies included the effect of self-care programs on the QoL of these patients, while Quality of life research as well as self-care researches could contribute to fulfill the quest for providing a better living for these chronic patients. According to Orem, self-care is a learned behavior, which can satisfy many needs of patients, provide growth and development, and prevent deviation from health [16]. So, The purpose of this study was to determine the effects of a self-care program on QoL of a sample of Iranian cirrhotic patients.
Methods
This quasi-experimental study was conducted with two experiment (n = 21) and control (n = 23) groups in Tehran Hepatitis Center (THC) in 2002. Using a longitudinal case registry method, 44 cirrhotic patients were selected over a four months period. Age between 20 and 65, not having a coexisting chronic debilitating illness [i.e. diabetes, chronic renal failure (CRF), stroke, inflammatory bowel disease, epilepsy and malignancy] and preferably living in Tehran) were considered as inclusion criteria. Three hundred cirrhotic patients were under the THC at the time of study. The patients had a routine visit each six months that would change to a monthly visit if any problem occurred. A total of 49 cirrhotic patients were referred for their routine (monthly or six-monthly) visit during the (first 4 months of the) study period. Of these, 44 patients had the including criteria and 5 patients were excluded because they had a coexisting chronic illness (4 because of diabetes and 1 for CRF). Random allocation technique was used to divide the sample into the experimental and control groups. To do this the patients were allocated into the groups as every other one.
Instruments
Four instruments were used including a demographic data questionnaire (consisting of questions related to the age, sex, level of education, martial status, number of children, occupation, frequency of hospitalizations, etiology, the duration and severity of the disease), a need assessment questionnaire, a self-assessment questionnaire, and the CLDQ questionnaire.
The need assessment questionnaire
Need assessment form was consistent of a list of 20 pre written questions related to the common problems such as fatigue, itching, dry mouth, muscular cramps flatulence, and also common problems related to dietary regiment and drug therapy. Each patient wanted to respond to the questions as "yes" or "no". At the end of need assessment forms the patients were asked to determine their own preferred time for attending the educational programs.
The self-report questionnaire
Six self-care checklists were designed in accordance with the six common problems and the content of educational programs (i.e. nutrition, controlling worry and depression, mouth dryness, pruritus, fatigue, muscular cramps). Each checklist was tabled for 30 days and included a list of self-care activities related to a common problem. The patients were asked to do a daily review on the checklists and mark self-care activities they followed. All the checklists were given to all members of experimental group and they were instructed how to complete them. They were also instructed to return the checklists to the researcher at the end of each month and new self-care checklists were given to them for the new month (a sample checklist is presented as additional file [see additional file 1]).
QOL instrument
CLDQ: The CLDQ is the first liver specific instrument for measuring QoL in chronic liver disease (CLD) developed by Younossi et al. [6]. The CLDQ includes 29 items in the following domains: abdominal symptoms, fatigue, systemic symptoms, activity, emotional function and worry. It has 7 likert scale type of answers ranging from "all of the time" to "none of the time" (possible range 29–203 from worst to best QoL) [6]. The construct validity of the CLDQ was supported by a strong correlation with patients' global rating scores (r = 0.84; p = 0.02) [6,8]. A particular strength of the CLDQ is that all phases of the validation process included patients with both hepatocellular and cholestatic liver disease of varying severity (no cirrhosis to Child's C cirrhosis). This should allow the CLDQ wide application in hepatology research.
Translation of CLDQ
The original version of CLDQ was translated into Farsi according to the standardized guidelines proposed by Guillemin et al. [17]. A native English speaker living in Iran who understood Farsi language quite well and did not have knowledge about QoL carried out back translation. The final version derived from reconciliation of the original and back translation and tested on 5 patients with chronic liver diseases. The content validity of translated CLDQ was approved by 10 faculty members in Tehran medical university. Translated CLDQ also was repeated in 5 patients in 2 weeks apart for test-retest analysis. Reliability was determined from Cronbach's alpha and test-retest. Cronbach's alpha was higher than 0.91 for domains and it was 0.93 for overall scores. Spearman's rank correlation was also 0.89 for CLDQ and it was higher than 0.73 for CLDQ domains.
The procedure
At the first stage, all subjects were administered CLDQ and demographic questionnaire. The experimental group was additionally given the need assessment questionnaire to determine their own educational needs as well as their own preferred time for attending the educational program. The severity of the illness was diagnosed and documented by a physician in patients' records.
Then the patients' demographic data and their educational needs were analyzed. Although the educational needs were somehow different for different patients, however, all of them were expressed their interest for intending the full educational program.
The content of educational program was designed [in six domains] based on the content of need assessment questionnaire and the domains of CLDQ questionnaire. [These were consisted of the nature of disease, coping strategies in systemic symptoms, coping strategies in worry and depression, relaxation techniques, diet and nutrition, and medical therapies, its side effects and relieving factors].
The experimental group was divided into four small groups with regard to their levels of education and their free time. The content of educational program were similar in all groups except for its simplicity [based on the subjects educational level]. Four educational sessions were held for each group. Each session lasted for 45 minutes and 5–6 patients with 3–4 relatives took part in each session [the relatives played a supportive role to maintain and restore clients' independence as much as possible].
The educational sessions consisted of a talk delivered by the main researcher. Also there was scope for questions and discussions. Posters, slides and manikin were also used to facilitate the participants learning process. Educational pamphlets or handouts (totally, 5 pamphlets and 2 handouts) covering all the contents and a related self-care checklist were given to the patients at the end of each session. The content of the four educational sessions are summarized in Table 2.
Table 2 The content of four educational sessions
Session 1 Nature of the disease Etiology, transmission route, clinical manifestations, diagnosis, management and complications
Session 2 Coping strategies in systemic symptoms Fatigue, dry mouth and, pruritus
Session 3 Coping strategies in worry, nutrition Anxiety, relaxation techniques, diet and nutrition,
Session 4 Muscular cramps, Medical therapies Medical regimen and its side effects, and relieving factors
The self-assessment checklists were given to the patients after the fourth educational session. The patients wanted to do a daily review on the checklists and mark each self-care activity if they followed them. Checklists were similar for all patients and did not modified during the study. The patients followed the program for 3 months and recorded the interventions in the self-report checklists. The main researcher phoned patients every two weeks and checked them for their compliance for the educations and reinforced them for completing the self-report forms. The completed checklists were returned to the researcher in a monthly visit at the end of each month and new self-care checklists were given to them for the upcoming month. After follow-up for three months, all patients (including the controls and experimental groups) completed the CLDQ once again. Then the CLDQ was analyzed and the two groups were compared. This study received ethics approval from the ethic committee of Tarbiat Modarres University. All subjects provided written consent before participation. The control group was also given the educational materials at the end of the third month to observe ethics. Data analysis was performed by SPSS using descriptive and analytical statistics.
Results
One patient in the experimental group and three in the control group died during the study to change the number of subjects in each group into 20. Demographic data of both groups are presented and compared in Table 1. No significant difference was found between the two groups.
Table 1 Demographic data of cirrhotic patients in the control and experimental groups
Items Control Experimental P
Mean age and SD 46 + 12.5 40 + 12.5 0.18
Sex (No.), %
Male (14), 70 (10), 50 0.19
Female (6), 30 (10), 50
Education
Illiterate (2), 10 (6), 30 0.27
Primary or secondary school (6), 30 (4), 20
Diploma or higher (12), 60 (10), 50
Marital status
Single (2), 10 (2), 10 0.38
Married (16), 80 (17), 85
Divorced (2), 10 (1), 5
No. of Children
Less than or equal to 3 (14), 70 (11), 55 0.32
More than 3 (6), 30 (9), 45
Occupation
Laborer (6), 30 (8), 40 0.32
Employee (4), 20 (6), 30
Housewife (4), 20 (4), 20
Student (2), 10 (2), 10
Retired (4), 20 -
Duration of illness (Years)
1–3 (12), 60 (14), 70 0.32
3–6 (6), 30 (4), 20
More than 6 (2), 10 (2), 10
No. of hospitalizations
None (12), 60 (8), 40 0.41
1 (2), 10 (4), 20
3 (6), 30 (8), 40
Etiologic factor
Hepatitis B (10), 50 (9), 45 0.83
Hepatitis C (2), 15 (4), 20
Autoimmune disease (4), 20 (4), 20
Cryptogenic (4), 20 (4), 20
Child of cirrhosis
A (4), 20 (4), 20 0.77
B (8), 40 (10), 50
C (8), 40 (6), 30
In general, 14 relatives took part in educational sessions with a range between 30 to 50 years of age and mostly between 40 and 50. Thirteen relatives were female, most of them had high school diploma (6 persons) and the majority were the spouses of patients (9 persons).
Patients' educational needs were used as a basis for planning educational sessions. The most common educational needs included: controlling or reducing abdominal distention (the greatest need, %70); curative ways in cirrhosis (being treatable or not treatable); ways of controlling fatigue (%65); principles of care and proper medications (%60); worry (%55); controlling pruritus and fatigue (%50); ways to decrease muscular cramps, dry mouth, and dyspnea; patterns of activity, rest, and sleep; routes of transmission as well as diagnostic tests (%45); and diagnostic procedures (%40).
In general, the items that the patients marked in the checklists were appraised over 3 months. All patients marked the nutritional items, of which %95 were observed. Items for fatigue, anxiety, and depression were ticked in %60 of cases and the rest reported no problem in this regard. Items for muscular cramps, dry mouth, and pruritus were ticked in %40, %30, and %20 of cases respectively and the rest reported no related problem.
Figure 1 shows the total score of CLDQ in both groups before and after the intervention. Mean score of CLDQ in the control group before the intervention was 137 and changed to 112.2 after 3 months, which showed a marked decline and a significant difference by Wilcoxon test (P = 0.001). In the experimental group, the mean score was 139 and changed to 171.6, which showed a rise and a significant difference by the same test (P = 0.001). There was no significant difference between QoL scores of both groups before the intervention, which was verified by Mann-Whitney test (P = 0.75). In fact, there was no significant difference before the intervention in total score of QoL between the two groups, but the same test showed a significant difference (P = 0.001), which indicated an improvement in QoL of the experimental group after 3 months.
Figure 1 Mean score of CLDQ in the control and experimental groups before and after intervention
Each domain in the CLDQ was compared before and after intervention in both groups. As presented by the Figure 2, there was a significant difference between mean scores of activity (P = 0.001), worry (P = 0.001), and emotional (P = 0.001) domains in the control group before and after intervention with a decline after 3 months. In other words, subjects in the control group experienced more emotional, anxiety, and activity problems, but no significant difference was shown in systemic (P = 0.59) and abdominal (P = 0.39) symptoms as well as fatigue (P = 0.9) after 3 months (Figure 2).
Figure 2 Mean score of CLDQ domains in cirrhotic patients in the control group before and after intervention
On the other hands, a significant difference noted between mean scores of the abdominal symptoms (P = 0.001), fatigue (P = 0.001), systemic symptoms (P = 0.001), activity (P = 0.01), worry (P = 0.001) and emotional (P = 0.001) domains in the experimental group before and after the intervention with an overall increase than before (Figure 3). Accordingly, subjects in this group experienced less abdominal, systemic, emotional, fatigue, and worry problems as well as improved activity.
Figure 3 Mean score of CLDQ domains in cirrhotic patients in the experimental group before and after the intervention
There was a significant difference between the experimental and control groups in all CLDQ domains after the intervention, as confirmed by Mann-Whitney test (P = 0.001). In other words, subjects in the experimental group experienced less systemic, abdominal, emotional, and fatigue problems than the control group. Thus, research hypothesis, "QoL of cirrhotic patients will improve following the self-care program," was verified.
However, the Wilcoxon test showed no significant difference in the severity of the disease before and after the intervention in control (P = 0.057) and experimental (P = 0.66) groups. In other word, the intervention had no effect on the severity of the disease. Mann-Whitney test also showed no significant difference in both groups before and after the intervention (P = 0.73).
Discussion
According to our knowledge this was the first study that evaluated the effects of an educational/ self-care program on the QoL of cirrhotic patients. The results of this study confirmed the beneficial effects of educational and self-care programs on the health related quality of life and also supported the findings of previous studies that reported improvement in patients QoL after self-care and educational programs [18]. This is an important finding since cirrhotic patients showed a significantly worse disease-specific and generic HRQoL than non-cirrhotic patients or healthy controls [19].
Most patients in our study were males with age between 20–50. The most common etiologic factors for their cirrhosis were also hepatitis B, cryptogenic cirrhosis, autoimmune hepatitis, and hepatitis C. These findings were similar to the previous study conducted by Alavian et al [3].
Medical practitioners have traditionally focused on organic diseases and their treatment. Patients, however, are concerned with their symptoms, regardless of the presence of organic or non-organic findings. On the other word, the quality of life is the forefront of patient concern [20,21]. Since current therapies are not yet good enough to eradicate the disease, the patients' care should considerably focus on their quality of life [21].
Coping strategies and the ways for controlling or reducing symptoms like abdominal distention; fatigue and pruritus were among the most common educational needs among our patients. These findings indicate that our health care providers including the medical practitioners and nursing staff do not exert enough effort for the patient education. This was also evident in the low level of QoL of both experimental and control groups before the intervention that was lower than the scores of QoL in healthy people reported by the other researchers [7,8,22]. So more attention should be paid to this important issue.
The severity of the disease did not decrease after the intervention, as it was not intended. However, the positive changes occurred in all aspects of QoL of experimental group. This was especially evident in the worry and emotional domains of QoL. It seems that the self-care program has made positive behavioral and cognitive changes in the experimental group. Such behavioral changes were predicted by Johnston and Orem [23,16]. Health-related quality of life generally refers to the patients' perceptions of their physical functioning, social functioning, role functioning, mental health, vitality, pain, and cognitive functioning. In many cases improvements in health-related quality of life are a natural result of improved clinical outcomes. However, patients' perception of their quality of life is also improved when they are empowered by well-designed educational programs. Empowered patients tend to feel more personally capable of positively impacting their outcomes. For patients with chronic conditions, health-related quality of life can improve significantly when they are trained in self-management techniques and empowered with education. Therefore, it could be said that the educational and self-care programs like the programs conducted in the present study could satisfy many needs of these chronic patients and will empower them to improve their quality of life.
Study limitations
The intervention has two components; the educational sessions, and the subsequent self-care program. However, we do not know exactly to which component the changes in QoL might be attributed. Was it the educational program alone, altering subjects perceptions and understanding, or was it the educational program plus self-care that mediated the change?
Although the active and continual participation of the patients in the management of their own problems may have an important role in the improved QoL in the experimental group, however, how the intervention might be altering QoL as recorded by CLDQ, is the issue that is not completely clear. This is an interesting question for future researches.
Although we tried to use a randomized sample in the experimental and control groups, however, given the small sample size, randomization may not assure a balance between arms.
The CDLQ is the first disease specific HRQL instrument for patients with Chronic Liver Disease. Although the inputs from the patients with a variety of types and stages of liver disease have been used for the development of the scale. However, the number of patients with Childs C were relatively low in the original validation studies [5], so the instrument may have limited validity in patients with Childs C. Yet, we used only the CLDQ for measuring the QoL, however, using other objective scales, such as functional status, etc, could add the reliability of the results in the future studies.
Conclusion
The result of the present study confirmed the positive effects of the educational and self care programs on the QoL of cirrhotic patients. The excessive aspiration of patients for educational programs was confirmed. Unfortunately no systematic and organized educational program is now existed for these chronic patients. Extensive educational and self-care programs along with long-term follow up such as the program conducted in this study are suggested. There was also some ambiguity related to the component that actually affected the QoL in the present study. So, further study is suggested to know to which component the change in QoL might be attributed.
Competing interests
The author(s) declare that they no competing interests.
Authors' contributions
MZ: Initiation and design of the research, collection and analysis of the data and writing the draft paper, MAH: co-analysis and editorial revision of draft papers, MR: supervisor, AKN: supervisor, SMA: supervisor, co-analysis and translation of draft papers,
Figure 4 Mean score of CLDQ domains in cirrhotic patients in the experimental and control groups after the intervention
Supplementary Material
Additional File 1
Self-care in pruritus
Click here for file
Acknowledgements
The authors would like to acknowledge all patients who participated in this research.
==== Refs
Andreoli T Cecil essential of the medicine 1997 4 U.S.A., Saunders Company
Goldberg E Chopra S Overview of the complications, prognosis and management of cirrhosis
Alavian SM Azimi K Sarafi M Alavi M Malek Zadeh R To determine etiologic factors of liver cirrhosis in hospitalized patients in shariaty hospital Iranian Journal of Gastroenteria 2002 38 19 26
Younossi ZM Boparai N Price LL Kiwi ML McCormick M Guyatt G Health related quality of life in chronic liver disease: the impact of type and severity of disease Am J Gastroenterol 2001 96 2199 2205 11467653 10.1111/j.1572-0241.2001.03956.x
Mihas AA Cirrhosis of the liver Postgraduate Medicine 2001 109 2
Younossi ZM Guyatt G Kiwi M Boparai N King D Development of a disease specific questionnaire to measure health related quality of life in patients with chronic liver disease Gut 1999 45 295 300 10403745
Younossi ZM Boparai N Price LL Kiwi ML McCormick M Guyatt G Health-related quality of life in chronic liver disease: the impact of type and severity of disease Am J Gastroenterol 2001 96 2199 2205 11467653 10.1111/j.1572-0241.2001.03956.x
Marchesini G Bianchi G Amodio P Salerno F Merli M Panella C Loguercio C Apolone G Niero M Abbiati R Factors associated with poor health-related quality of life of patients with cirrhosis Gastroenterology 2001 120 170 178 11208726
De Bona M Ponton P Ermani M Iemmolo RM Feltrin A Boccagni P Gerunda G Naccarato R Rupolo G Burra P The impact of liver disease and medical complications on quality of life and psychological distress before and after liver transplantation J Hepatol 2000 33 609 615 11059865 10.1034/j.1600-0641.2000.033004609.x
Borgaonkar MR Irvine EJ Quality of life measurement in gastrointestinal and liver disorders Gut 2001 47 444 454 10940286 10.1136/gut.47.3.444
Davis GL Balart MD Schiff ER Lindsay K Bodenheimer HC Perrillo RP Carey W Jacobson IM Payne J Dienstag JL Assessing health-related quality of life in chronic hepatitis C using the sickness impact profile Clin Ther 1994 16 334 343 8062327
Ware JE Bayliss MS Mannocchia M Davis GL the International Hepatitis Interventional Therapy Group Health-related quality of life in chronic hepatitis C: Impact of disease and treatment response Hepatology 1999 30 550 555 10421667 10.1002/hep.510300203
Bonkovsky HL Woolley JM the Consensus Interferon Study Group. Reduction of health-related quality of life in HCV and improvement with interferon therapy Hepatology 1999 29 264 270 9862876 10.1002/hep.510290124
Foster GR Goldin RD Thomas HC Chronic hepatitis C virus infection causes a significant reduction in quality of life in the absence of cirrhosis Hepatology 1998 27 209 212 9425939 10.1002/hep.510270132
Younossi ZM Kiwi ML Boparai N Price LL Guyatt G Cholestatic liver diseases and health-related quality of life Am J Gastroenterol 2000 95 497 502 10685757 10.1111/j.1572-0241.2000.01774.x
Craddock RB Adams PF Usui WM Mitchell L An intervention to increase use and effectiveness of self-care measures for breast cancer chemotherapy patients Cancer Nurs 1999 22 312 319 10452209 10.1097/00002820-199908000-00009
Guillemin F Bombardier C Beaton D Cross-cultural adaptation of health-related quality of life measures: literaturereview and proposed guidelines J Clin Epidemiol 1993 46 1417 1432 8263569 10.1016/0895-4356(93)90142-N
Ries AL Kaplan RM Limberg TM Prewitt LM Effects of Pulmonary Rehabilitation on Physiologic and Psychosocial Outcomes in Patients with Chronic Obstructive Pulmonary Disease Ann Intern Med 1995 122 823 832 7741366
Van der Plas SM Hansen BE de Boer JB Stijnen T Passchier J de Man RA Schalm SW Generic and disease-specific health related quality of life in non-cirrhotic, cirrhotic and transplanted liver patients: a cross-sectional study BMC Gastroenterol 2003 3 33 14617381 10.1186/1471-230X-3-33
Glis H Wiklund I Health-related quality of life and gastrointestinal disease Journal of Gastroenterology and Hepatology 2002 17 S72 S84 12000595 10.1046/j.1440-1746.17.s1.6.x
Bernstein D Health Related Quality of Life and Hepatitis C
Gralnek IM Hays RD Kilbourne A Rosen HR Keeffe EB Artinian L Kim S Lazarovici D Jensen DM Busuttil RW Martin P Development and evaluation of the Liver Disease Quality of Life instrument in persons with advanced, chronic liver disease – the LDQOL 1.0 Am J Gastroenterol 2000 95 3552 3565 11151892
Johnston M Mood in chronic disease: Questioning the answer Curr Psychol 1999 18 71 89
| 15904528 | PMC1156929 | CC BY | 2021-01-04 16:38:14 | no | Health Qual Life Outcomes. 2005 May 18; 3:35 | utf-8 | Health Qual Life Outcomes | 2,005 | 10.1186/1477-7525-3-35 | oa_comm |
==== Front
Int J Health GeogrInternational Journal of Health Geographics1476-072XBioMed Central London 1476-072X-4-121592708210.1186/1476-072X-4-12MethodologyNeighborhood size and local geographic variation of health and social determinants Ali Mohammad [email protected] Jin-Kyung [email protected] Vu Dinh [email protected] Do Gia [email protected] Michael [email protected] John D [email protected] International Vaccine Institute, SNU Research Park, San 4–8 Bongcheon-7 dong, Kwanak-gu, Seoul, Korea2 National Institute of Health and Epidemiology, No. 1 Yersin Street, Hanoi, Vietnam3 Robert Wood Johnson Foundation Health & Society Scholar, Columbia University, USA2005 1 6 2005 4 12 12 19 4 2005 1 6 2005 Copyright © 2005 Ali et al; licensee BioMed Central Ltd.2005Ali et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Spatial filtering using a geographic information system (GIS) is often used to smooth health and ecological data. Smoothing disease data can help us understand local (neighborhood) geographic variation and ecological risk of diseases. Analyses that use small neighborhood sizes yield individualistic patterns and large sizes reveal the global structure of data where local variation is obscured. Therefore, choosing an optimal neighborhood size is important for understanding ecological associations with diseases. This paper uses Hartley's test of homogeneity of variance (Fmax) as a methodological solution for selecting optimal neighborhood sizes. The data from a study area in Vietnam are used to test the suitability of this method.
Results
The Hartley's Fmax test was applied to spatial variables for two enteric diseases and two socioeconomic determinants. Various neighbourhood sizes were tested by using a two step process to implement the Fmaxtest. First the variance of each neighborhood was compared to the highest neighborhood variance (upper, Fmax1) and then they were compared with the lowest neighborhood variance (lower, Fmax2). A significant value of Fmax1 indicates that the neighborhood does not reveal the global structure of data, and in contrast, a significant value in Fmax2 implies that the neighborhood data are not individualistic. The neighborhoods that are between the lower and the upper limits are the optimal neighbourhood sizes.
Conclusion
The results of tests provide different neighbourhood sizes for different variables suggesting that optimal neighbourhood size is data dependent. In ecology, it is well known that observation scales may influence ecological inference. Therefore, selecting optimal neigborhood size is essential for understanding disease ecologies. The optimal neighbourhood selection method that is tested in this paper can be useful in health and ecological studies.
==== Body
Introduction
Spatial filtering can be used to create smoothed maps of health and ecological patterns [1-4]. Since population distributions are highly heterogeneous in space, an ordinary point plot of all cases is not useful. Smoothing data by adjusting for the population at risk is necessary to identify areas with higher disease rates [5]. Smoothing disease data can provide the true relative risk of a disease across a study area [6]. There are other reasons to filter health and ecological data. Field survey data gathering systems usually generate errors. Filtering removes random noise caused by inaccurate records or mislocated cases [1,7,8]. There are many intervening factors at the individual level that may influence spatial processes of disease phenomena. For instance, an individual's biological or socioeconomic status may influence their health status. Neighbors usually have similar risk, particularly for environmentally related diseases, unless the spatial process of the disease is exclusively affected by individual-level characteristics. Also, some risk factors of diseases genuinely operate at the population level [9].
People do not live in isolation; they live in groups (neighborhood) that may influence their life style, health, and health seeking behavior. Thus, a neighborhood level study is sometimes essential to identify important public health problems and to generate hypotheses about their potential causes [9]. Twigg et al. showed that the behavioral practices of an individual are influenced by neighbors [10]. Some variables do not make sense at the individual scale and should be modeled as ecological variables. For instance, a household with a good sanitation system can be exposed to bad sanitation from neighbors. Ecological factors are more meaningful if the data are measured by neighborhood. Spatial filtering can be used to model such neighborhood level phenomena.
Ecological variables can be measured at different geographic scales from local to global. In ecology, it is well known that observation scales influence ecological inference [11-13]. Determining the neighborhood size (or the area) over which densities of the phenomena are estimated is important. A large neighborhood makes the data flat over the entire study area whereby important local level variation is obscured that could point to ecological associations. In contrast, a small neighborhood may reveal individualistic patterns [1], and that may not be useful for identifying ecological relationships with health outcomes. Defining an optimal neighborhood size is difficult [14,15]. Bailey and Gatrell [16] suggested exploring different sizes and looking at the variation at those scales to come up with an optimal neighborhood size. However, literature that describes methodologies for selecting the optimal size of neighborhood is scarce. Thus, one often chooses the scale arbitrarily, and the use of an arbitrary scale may yield spurious outcomes. This paper introduces a methodological approach for selecting the optimal neighborhood size that can be used to measure ecological variables and to investigate ecological links with local variation of diseases.
Methodology
The Study area
Health and socioeconomic data of a study area in Khanh Hoa Province, on the coast of central Vietnam, were used to test the proposed method. The size of the study area is 740 square kilometers consisting of 33 communes in two districts: Nha Trang (151 square kilometres) and Ninh Hoa (589 square kilometres). A dynamic population database is maintained for the study area which is updated on yearly basis. In 2002, the population of the study area was 329,596, of which 54% of the population is from Nha Trang. We created a household-level geographic information system (GIS) database that includes a point for each active household (described below) in 2002. The household settlements are clustered, leaving a large portion non-inhabited land within the administrative boundaries, which led us to define household-based working study area by creating 500 meters radius buffer around each household point and dissolving boundaries between buffers (Figure 1). This resulted in a 394 square kilometres working study area for the entire population and 79 square kilometres for Nha Trang specifically.
Figure 1 The geographic features of the study area in Vietnam. Map showing the geographic characteristics of the study area along with the geographic positing of the study area in Vietnam
The GIS data
In 2003, we conducted a global positioning system (GPS) survey using handheld receivers to identify the geographic locations of every household in the study area. A base map of commune boundaries and other geographic features (e.g., rivers, roads, railways, lakes) was first acquired in digital format. The household GPS survey data were projected in the same geographic referencing system (i.e., Transverse Mercator) so that the household points could be accurately integrated with the study area base map. When several households shared a single structure or closely connected structures a single point was plotted. We received a list of 72,152 households from the census 2002. A total of 3,587 households could not be included in the GIS for a variety of reasons (e.g., migration out, living on a military base with no access, and household confirmation was not possible due to the absence of household members). Thus, the GIS was created for the 68,565 households, which were referenced spatially by 32,542 points. Several checks were made for missing households, misallocations, and wrong identification numbers, and the erroneous data were corrected. Finally, the household data were mapped in groups (the smallest geographic entity), and their positions were verified by ground-truthing.
The attribute data
We used two social variables, religion and literacy (i.e., years of schooling), and two enteric disease incidence variables (i.e., shigellosis and Vibrio parahaemolyticus). The disease incidence data were obtained from a population-based passive surveillance system that was begun in January 1997 [17]. The socioeconomic data were obtained from a 2002 population survey. The survey data shows approximately 85% of the population is secular, 10% Buddhist, and 5% Christian. Only 11% had not attended school, 37% had received a primary education, and the rest (52%) completed secondary or higher education.
The Vibrio parahaemolyticus data were derived from the disease surveillance of the study area from 1997 to 1999. V. parahaemolyticus, a gram-negative, halophilic bacterium, inhabits marine and estuarine environments. The microorganism was first identified as a cause of food borne illness in Japan in 1950 [18]. A polymerase chain reaction (PCR) based method to detect the toxR sequence specific to Vibro parahaemolyticus was used to identify cases as reported elsewhere [19]. The shigellosis study was carried out for three years (2001–2003) in Nha Trang in collaboration with the Diseases of the Most Impoverished Program [20]. All shigellosis cases were detected through microbiological test of stool samples.
Data categorization and manipulation
We categorized the two variables, literacy and religion, to define the social status of the study population. A person having six years of schooling or above was considered to be literate, and the religion was classified by Buddhist and non-Buddhist. Since the data were obtained at the individual level, the data were aggregated by spatially referenced household points. Neighborhood level data were then obtained for each of the spatially referenced points of household (32,542 points for the 329,596 individuals in 68,565 households) by aggregating household level data of surrounding neighbors using circular windows of various sizes. The neighborhood level social variables were estimated by the percentage of people living within neighborhoods, and the disease incidence variables were expressed in rates per 10,000 people. Our aim was to create a local-level neighborhood variable for these phenomena. Therefore, based on the working size of the study area and the spatial distribution of the population we set the minimum size to a 100-meter radius neighborhood and increased the size stepwise by 100 meters until a 2000 meter size was reached. This resulted in 20 different neighborhood sizes from which to select an optimal neighborhood size.
Statistical analysis
Since the data from smaller neighborhoods are individualistic in nature, a high variance value is expected. In contrast, a low variance value is expected for larger neighborhoods. A high variance value means that data are local and low variance means that they are global. To select an optimal size of neighborhood that can ensure that the ecological data are neither local nor global, we used Hartley's test of homogeneity of variance [21] that evaluates variation in variances across neighborhoods. The Hartley's test statistic, FMAX, is calculated by
where
= maximum value of the variances among groups
= minimum value of the variances among groups
Under the null hypothesis, the test assumes that the variances are equal. The critical value (CV) was calculated under the F-distribution with (k, nMAX - 1) degrees of freedom at α = 0.05. Here, k is the number of groups and nMAX is the maximum sample size among groups.
The Fmax test involved two steps. First the variance of each neighborhood was compared to the highest neighborhood variance (upper, Fmax1) and then they were compared with the lowest neighborhood variance (lower, Fmax2). A significant value (means the value does not fall within the CV) of Fmax1 indicates that the neighborhood does not have a global structure of data, and in contrast, a significant value in Fmax2 implies that the neighborhood data are not individualistic. The neighborhoods that are between the lower and the upper limits are the optimal neighbourhood sizes.
Results
There were 131 cases of Vibro parahaemolyticus in 127 household points in the entire study area for the three years of study (1997–1999), and 308 cases of shigellosis were observed in 295 household points for the year 2001 through 2003 in Nha Trang. Out of the total 329,596 population, 31,924 (9.7%) were Buddhists who were identified in 3,681 household points of the total study area. And, a total of 168,699 (51.2%) literate persons were observed in 30,069 household points.
The data variances for the Vibro parahaemolyticus incidence rates under various neighborhood sizes show a declining trend with an increase in neigborhood size (Figure 2). The rapid decline observed at smaller scales virtually disappears with larger neighborhood sizes. The pattern is similar for shigellosis as well as for both socioeconomic variables (figures not shown).
Figure 2 The data variance by neighborhood size. Graphical presentation of the data variances for the Vibro parahaemolyticus incidence in Khanh Hoa, Vietnam under various sizes of neighborhood.
The test results for homogeneity variance of Vibro parahaemolyticus incidence rates under various neighborhood sizes are listed in Table 1. The Fmax1 test statistic at the level α = 0.05 shows a neighborhood size above 900 meters would reveal the global structure of the data, and the Fmax2 statistic shows that any neighborhoods below 200 meters would make the data too individualistic. Thus, the choice of optimal neighborhood lies between 200 and 900 meters. Considering the values of several parameters such as minimum population size, skewness and kurtosis of the incidence rate, we argue that a 500-meter neighborhood is optimal size for modeling the local variation of the disease incidence.
Table 1 Descriptive statistics and results of variance ratio (Fmax) test for the Vibrio parahaemolyticus incidence under various neighborhoods, Khanh Hoa Province, Vietnam, 1997–99. (n = 29,211)
r † Population size Incidence Rate/10000 Population Upper Fmax Test Lower Fmax Test
Min Max Mean Min Max Mean Variance Fmax1 DF1 * CV1 ** Fmax2 DF2 * CV2 **
100 1 1779 158 .00 1429.00 4.612 646.023 49.116 20 1.571 1.000 1 3.842
200 1 5143 492 .00 370.40 4.451 195.453 14.860 19 1.587 3.305 2 2.996
300 1 7372 914 .00 208.30 4.538 108.599 8.257 18 1.604 5.949 3 2.605
400 1 9252 1416 .00 161.30 4.533 73.509 5.589 17 1.623 8.788 4 2.372
500 3 12265 1971 .00 144.90 4.494 52.889 4.021 16 1.644 12.215 5 2.214
600 4 15784 2571 .00 227.30 4.486 42.481 3.230 15 1.667 15.207 6 2.099
700 4 19178 3236 .00 92.59 4.441 33.666 2.560 14 1.692 19.189 7 2.010
800 4 21949 3953 .00 52.91 4.434 28.671 2.180 13 1.720 22.532 8 1.939
900 4 24982 4711 .00 38.46 4.446 25.298 1.923 12 1.753 25.537 9 1.880
1000 4 26772 5508 .00 35.71 4.463 22.733 1.728 11 1.789 28.418 10 1.831
1100 6 28821 6324 .00 36.50 4.4939 20.647 1.570 10 1.831 31.289 11 1.789
1200 25 31691 7160 .00 36.50 4.5101 18.831 1.432 9 1.880 34.306 12 1.753
1300 50 34877 8029 .00 46.51 4.5099 17.453 1.327 8 1.939 37.015 13 1.720
1400 63 36311 8921 .00 37.17 4.5184 16.444 1.250 7 2.010 39.286 14 1.692
1500 63 37334 9832 .00 31.65 4.5266 15.655 1.190 6 2.099 41.266 15 1.667
1600 63 38471 10760 .00 28.33 4.5429 15.116 1.149 5 2.214 42.738 16 1.644
1700 63 39259 11693 .00 28.17 4.5510 14.593 1.109 4 2.372 44.269 17 1.623
1800 63 40278 12651 .00 27.70 4.5727 14.114 1.073 3 2.605 45.772 18 1.604
1900 63 41457 13657 .00 26.32 4.5990 13.639 1.037 2 2.996 47.366 19 1.587
2000 63 42492 14703 .00 26.32 4.6152 13.153 1.000 1 3.842 49.116 20 1.571
† r = size of neighborhood in meter radius
* DF = degrees of freedom
**CV1 and CV2 = critical values at 95% confidence level for Upper Fmax and Lower Fmax respectively Bold figures in Fmax1 and Fmax2 are the upper and lower limit of optimal neighborhoods, and the bold figure in "r" column is the choice of optimal neighborhood size.
When looking at literacy, the Fmax1 test statistic shows a neighborhood above 600 meters would reveal the global pattern, and the Fmax2 test statistic demonstrates any neighborhoods below 700 meters would make the data individualist (Table 2). In this case, we believe that 700 meters is the optimal size because the difference between Fmax1 and CV1 is smaller than the difference between Fmax2 and CV2 of a 600-meter neighborhood. The summary statistics and test results of religious status under various neighborhood sizes are shown in Table 3. For religion, a 700-meter size neighborhood is also appropriate.
Table 2 Descriptive statistics and results of variance ratio (Fmax) test for the literacy status under various neighborhoods, Khanh Hoa Province, Vietnam, 2002. (n = 32,542)
r † Population size Incidence Rate/10000 Population Upper Fmax Test Lower Fmax Test
Min Max Mean Min Max Mean Variance Fmax1 DF1 * CV1 ** Fmax2 DF2 * CV2 **
100 1 1859 195 .00 100.00 50.226 221.986 3.262 20 1.571 1.000 1 3.842
200 1 5681 611 .00 100.00 50.191 167.826 2.466 19 1.587 1.323 2 2.996
300 1 8362 1143 .00 88.89 50.215 146.722 2.156 18 1.604 1.513 3 2.605
400 2 11276 1785 .00 83.33 50.229 134.220 1.972 17 1.623 1.654 4 2.372
500 2 15425 2504 .00 83.33 50.250 124.628 1.831 16 1.644 1.781 5 2.214
600 2 20047 3282 .00 80.50 50.241 117.075 1.720 15 1.667 1.896 6 2.099
700 2 23875 4146 .00 80.52 50.270 110.223 1.620 14 1.692 2.014 7 2.010
800 2 28601 5075 .00 80.46 50.296 104.018 1.528 13 1.720 2.134 8 1.939
900 2 33159 6055 .00 80.14 50.318 98.808 1.452 12 1.752 2.247 9 1.880
1000 4 36182 7081 15.70 75.96 50.333 94.737 1.392 11 1.789 2.343 10 1.831
1100 9 39576 8129 11.11 74.55 50.344 91.235 1.341 10 1.831 2.433 11 1.789
1200 25 43628 9194 15.97 73.14 50.355 88.216 1.296 9 1.880 2.516 12 1.752
1300 25 47769 10297 16.07 72.95 50.373 85.460 1.256 8 1.939 2.598 13 1.720
1400 25 49747 11427 16.52 72.44 50.386 82.739 1.216 7 2.010 2.683 14 1.692
1500 25 51108 12580 16.54 71.57 50.390 80.180 1.178 6 2.099 2.769 15 1.667
1600 66 52454 13750 16.54 70.89 50.394 77.521 1.139 5 2.214 2.864 16 1.644
1700 128 53725 14925 16.61 70.67 50.392 74.995 1.102 4 2.372 2.960 17 1.623
1800 138 55422 16133 16.61 69.93 50.393 72.569 1.066 3 2.605 3.059 18 1.604
1900 138 57913 17398 17.36 69.52 50.394 70.210 1.032 2 2.996 3.162 19 1.587
2000 148 59428 18708 17.75 68.81 50.412 68.054 1.000 1 3.842 3.262 20 1.571
† r = size of neighborhood in meter radius
* DF = degrees of freedom
**CV1 and CV2 = critical values at 95% confidence level for Upper Fmax and Lower Fmax respectively Bold figures in Fmax1 and Fmax2 are the upper and lower limit of optimal neighborhoods, and the bold figure in "r" column is the choice of optimal neighborhood size.
Table 3 Descriptive statistics and results of variance ratio (Fmax) test for the ethnicity status under various neighborhoods, Khanh Hoa Province, Vietnam, 2002. (n = 32,542)
r † Population size Incidence Rate/10000 Population Upper Fmax Test Lower Fmax Test
Min Max Mean Min Max Mean Variance Fmax1 DF1 * CV1 ** Fmax2 DF2 * CV2 **
100 1 1859 195 .00 100.00 7.283 243.304 4.026 20 1.571 1.000 1 3.842
200 1 5681 611 .00 100.00 7.344 183.209 3.032 19 1.587 1.328 2 2.996
300 1 8362 1143 .00 100.00 7.383 153.024 2.532 18 1.604 1.590 3 2.605
400 2 11276 1785 .00 100.00 7.402 133.603 2.211 17 1.623 1.821 4 2.372
500 2 15425 2504 .00 100.00 7.467 121.327 2.008 16 1.644 2.005 5 2.214
600 2 20047 3282 .00 88.57 7.534 112.520 1.862 15 1.667 2.162 6 2.099
700 2 23875 4146 .00 88.57 7.609 105.353 1.744 14 1.692 2.309 7 2.010
800 2 28601 5075 .00 82.50 7.682 99.320 1.644 13 1.720 2.450 8 1.939
900 2 33159 6055 .00 75.00 7.719 93.729 1.551 12 1.752 2.596 9 1.880
1000 4 36182 7081 .00 69.49 7.745 88.828 1.470 11 1.789 2.739 10 1.831
1100 9 39576 8129 .00 63.73 7.761 84.867 1.404 10 1.831 2.867 11 1.789
1200 25 43628 9194 .00 62.40 7.776 81.459 1.348 9 1.880 2.987 12 1.752
1300 25 47769 10297 .00 62.40 7.793 78.140 1.293 8 1.939 3.114 13 1.720
1400 25 49747 11427 .00 61.91 7.801 74.610 1.235 7 2.010 3.261 14 1.692
1500 25 51108 12580 .00 60.84 7.820 71.456 1.183 6 2.099 3.405 15 1.667
1600 66 52454 13750 .00 58.06 7.851 69.027 1.142 5 2.214 3.525 16 1.644
1700 128 53725 14925 .00 50.55 7.873 66.823 1.106 4 2.372 3.641 17 1.623
1800 138 55422 16133 .00 46.79 7.891 64.531 1.068 3 2.605 3.770 18 1.604
1900 138 57913 17398 .00 44.54 7.907 62.339 1.032 2 2.996 3.903 19 1.587
2000 148 59428 18708 .00 39.71 7.915 60.426 1.000 1 3.842 4.026 20 1.571
† r = size of neighborhood in meter radius
* DF = degrees of freedom
**CV1 and CV2 = critical values at 95% confidence level for Upper Fmax and Lower Fmax respectively Bold figures in Fmax1 and Fmax2 are the upper and lower limit of optimal neighborhoods, and the bold figure in "r" column is the choice of optimal neighborhood size.
The test results for the choice of optimal neighborhood size for shigellosis obtained from the Nha Trang subpopulation are shown in Table 4. The Fmax1test statistic reveals that a neighborhood size over 800 meters would produce a global pattern. On the other hand, the Fmax2 test statistic illustrates that a neighborhood below 300 meter would yield an individualistic pattern. Out of the choices between 400 and 800 meters, we suggest a 400 meter neighborhood size based on the criteria mentioned above for Vibro parahaemolyticus.
Table 4 Descriptive statistics and results of variance ratio (Fmax) test for shigella incidence under various neighborhoods, Nha Trang, Vietnam, 1999–2001. (n = 13565)
r † Population size Incidence Rate/10000 Population Upper Fmax Test Lower Fmax Test
Min Max Mean Min Max Mean Variance Fmax1 DF1 * CV1 ** Fmax2 DF2 * CV2 **
100 3 5692 1015 .00 333.30 6.041 197.436 25.397 20 1.571 1.000 1 3.842
200 3 17440 3223 .00 333.30 6.155 74.927 9.638 19 1.587 2.635 2 2.996
300 3 25689 6026 .00 57.14 6.112 38.808 4.992 18 1.605 5.088 3 2.606
400 10 34531 9388 .00 41.67 6.116 28.304 3.641 17 1.624 6.976 4 2.373
500 30 47539 13112 .00 57.47 6.102 22.808 2.934 16 1.644 8.656 5 2.215
600 33 61692 17105 .00 34.36 6.121 18.871 2.427 15 1.667 10.462 6 2.099
700 33 73490 21480 .00 35.46 6.129 16.566 2.131 14 1.692 11.918 7 2.010
800 87 88015 26126 .00 27.47 6.140 14.514 1.867 13 1.721 13.603 8 1.939
900 87 102106 30938 .00 22.87 6.148 13.374 1.720 12 1.753 14.763 9 1.881
1000 87 111244 35834 .00 31.15 6.158 12.651 1.627 11 1.789 15.606 10 1.831
1100 87 121667 40711 .00 19.67 6.137 11.774 1.515 10 1.831 16.769 11 1.789
1200 150 134222 45547 .00 18.28 6.116 11.057 1.422 9 1.881 17.856 12 1.753
1300 425 146883 50410 .00 17.64 6.119 10.536 1.355 8 1.939 18.739 13 1.721
1400 628 152901 55270 .00 16.40 6.135 10.079 1.297 7 2.010 19.589 14 1.692
1500 628 156963 60090 1.21 15.92 6.135 9.588 1.233 6 2.099 20.592 15 1.667
1600 628 161121 64829 1.77 15.92 6.123 9.104 1.171 5 2.215 21.687 16 1.644
1700 628 165020 69457 1.21 15.92 6.129 8.751 1.126 4 2.373 22.562 17 1.624
1800 628 170190 74078 2.73 15.92 6.135 8.411 1.082 3 2.606 23.474 18 1.605
1900 628 177965 78890 2.68 15.92 6.153 8.099 1.042 2 2.996 24.378 19 1.587
2000 628 182551 83821 2.98 15.92 6.171 7.774 1.000 1 3.842 25.397 20 1.571
† r = size of neighborhood in meter radius
* DF = degrees of freedom
**CV1 and CV2 = critical values at 95% confidence level for Upper Fmax and Lower Fmax respectively Bold figures in Fmax1 and Fmax2 are the upper and lower limit of optimal neighborhoods, and the bold figure in "r" column is the choice of optimal neighborhood size.
To get an understanding of local geographic variation of the disease and social variables, we created isopleth maps with the spatially smoothed data by using the optimal neighborhood sizes. Spatially smoothed data are more appropriate for disease and ecological mapping than the raw data [22]. A widely used geostatistical interpolation method called kriging [23,24] was used to create those maps. The maps ware produced as a quintile distribution for the respective phenomenon. Figure 3 shows the local geographic pattern of the Vibro parahaemolyticus incidence rate in Khanh Hoa province, Figure 4 shows the geographic pattern of literacy status in Khanh Hoa province, Figure 5 shows the pattern of religion in Khanh Hoa province, and Figure 6 shows the pattern of shigellosis incidence in Nha Trang. All of the maps show clear local geographic variation of the phenomena.
Figure 3 Local geographic pattern of Vibro parahaemolyticus incidence rate in Khanh Hoa province, Vietnam. The map was created based on the household point locations, thus the upper part of the study where no households are located have been omitted during the creation of the surface map. The lighter tones indicate lower Vibro parahaemolyticus incidence rate and the darker tones indicate higher Vibro parahaemolyticus incidence rate.
Figure 4 Local geographic pattern of literacy status in Khanh Hoa province, Vietnam. The map was created based on the household point locations, thus the upper part of the study where no households are located have been omitted during the creation of the surface map. The lighter tones indicate lower literacy status and the darker tones indicate higher literacy status.
Figure 5 Local geographic pattern of ethnicity status in Khanh Hoa province, Vietnam. The map was created based on the household point locations, thus the upper part of the study where no households are located have been omitted during the creation of the surface map. The lighter tones indicate lower proportion of ethnically minority group and the darker tones indicate higher proportion of the ethnically minority group.
Figure 6 Local geographic pattern of shigella incidence rate in Nha Trang, Vietnam. The map was created based on the household point locations, thus the upper part of the study where no households are located have been omitted during the creation of the surface map. The lighter tones indicate lower shigella incidence rate and the darker tones indicate higher shigella incidence rate.
Discussion and conclusion
The Hartley's Fmax test of homogeneity provides a solution for determining the optimal neighborhood size for modeling the local variation of health and social determinants. The methodological approach illustrates that the choice of optimal neighborhood is data dependent. Vibrio parahaemolyticus incidence requires a scale from 200 and 900 meters, and we argued that a 500 meter neighborhood is most appropriate based on the values of other parameters. The choice of neighborhood size for social variables (i.e., literacy and religion) ranged from 600 to 700 meters, and we suggested 700 meters for both. Similarly, out of the options between 300 and 800 meters for shigellosis incidence in Nha Trang, we suggest a 400 meter neighborhood. The maps produced using optimal neighborhood sizes show clear local geographic variation of the respective phenomenon suggesting the suitability of the approach. Since the ecological process may differ from one variable to another [25], different optimal neighborhood sizes are expected. The results of our analyses confirm this notion.
Measuring ecological data at a neighborhood scale to understand the spatial variability requires considerable knowledge of the phenomenon being measured [26]. For example, dissemination of an innovation may diffuse to close neighbors through literate persons. However, the media through which it diffuses is assumed spatially heterogeneous. For instance, a friendly neighborhood may accelerate the innovation, but disputes among neighbors may impede diffusion of the innovation. It would be ideal to assign weight for these social factors while measuring ecological variables, but that requires considerable knowledge about the spatial process of the phenomenon. For sanitation status, a poorly constructed latrine can be an important source of pollution by spreading fecal matter to nearby areas. Therefore, a distance decay weight can be applied here considering there is an inverse relationship from the source of pollution [27].
Since spatial filtering smoothes data, average errors may be inherent in the data [28]. Such ecological bias [29] can be more apparent in a predefined geographic area than within the natural boundary created through spatially smoothed data using optimal neighborhood modeling. Ecological bias may also be present when modeling variables with large neighborhood sizes.
One of the biggest problems in spatial epidemiology and ecological exposure assessment is in identifying geographic patterns [29] through spatial interpolation. Selection of an interpolation method has strong implications on the representation of spatial patterns as well as on the accuracy of interpolated data [30]. Interpolating the data based on spatially smoothed data obtained by an optimal neighborhood size could provide more accuracy in the local variation of the phenomena being measured. The optimal neighborhood may help ecological analysis in two ways: aggregating the data (both dependent and independent variables) using optimal neighbourhood scales and performing the analysis at the ecological level, or by limiting the dependent variable at the individual level, but attaching ecological covariates (obtained through optimal neighbourhood size) to each individual [31].
A scientifically validated method is required to assist geographic research [32], and to properly use GIS technology in health and ecological studies [33]. In our paper, we have outlined a method to choose optimal neighbourhood sizes for addressing local spatial variation of disease and social determinants. The method can be useful in health and ecological studies.
Acknowledgements
This work was supported by the Diseases of the Most Impoverished Program, funded by the Bill and Melinda Gates Foundation and coordinated by the International Vaccine Institute.
==== Refs
Talbot TO Kulldroff M Teven PF Haley VB Evaluation of spatial filters to create smoothed maps of health data Statistics in Medicine 2000 18 2399 2408 10960861 10.1002/1097-0258(20000915/30)19:17/18<2399::AID-SIM577>3.3.CO;2-I
Rushton G Lolonis P Exploratory spatial analysis of birth defect rates in an urban population Statistics in Medicine 1996 15 717 726 9132899 10.1002/(SICI)1097-0258(19960415)15:7/9<717::AID-SIM243>3.0.CO;2-0
Kafadar K Smoothing geographical data, particularly rates of disease Statistics in Medicine 1996 15 2539 2560 8961462 10.1002/(SICI)1097-0258(19961215)15:23<2539::AID-SIM379>3.3.CO;2-2
Curtis A Using a spatial filter and a geographic information system to improve rabies surveillance data Emerg Infect Dis 1999 5 603 606 10511516
Alexander FE Ricketts TJ Williams J Cartwright RA Methods of mapping and identifying small clusters of rare diseases with applications to geographical epidemiology Geographical Analysis 1991 23 158 173
Lawson AB Tutorial in biostatistics: disease map reconstruction Statistics in Medicine 2001 20 2183 2204 11439429 10.1002/sim.933
Ali M Emch M Donnay JP Spatial filtering using a raster geographic information system: methods for scaling health and environmental data Health & Place 2002 8 85 92 11943581 10.1016/S1353-8292(01)00029-6
Lawson AB Statistical Methods in Spatial Epidemiology 2001 Chichester, Wiley
Pearce N The ecological fallacy strikes back J Epidemiol Community Health 2000 54 326 7 10814650 10.1136/jech.54.5.326
Twigg L Moon G Jones K Predicting small-area health-related behaviour: a comparison of smoking and drinking indicators Soc Sci Med 2000 50 1109 1120 10714931 10.1016/S0277-9536(99)00359-7
Mercer WB Hall AD The experimental error of field trials J Agricult Sci 1991 4 107 132
Horne JK Schneider DC Spatial variance in ecology Oikos 1995 74 18 26
Diez Roux AV Investigating neighborhood and area effects on health Am J Public Health 2001 91 1783 1789 11684601
Scott DW Multivariate density estimation: theory, practice, and visualization 1992 New York, J Wiley
Diez Roux AV The study of group-level factors in epidemiology: rethinking variables, study designs, and analytical approaches Epidemiol Rev 2004 26 104 111 15234951 10.1093/epirev/mxh006
Bailey TC Gatrell AC Interactive Spatial Data Analysis 1995 Longman, Harlow
Naficy AB Trach DD Ke NT Chuc NT Sorkin A Rao MR Sy TH Thiem VD Canh DG Mahoney RT Holmgren J Ivanoff B Clemens JD Cost of immunization with a locally produced, oral cholera vaccine in Viet Nam Vaccine 2001 19 3720 3725 11395206 10.1016/S0264-410X(01)00089-5
Tuyet DT Thiem VD von Seidlein L Chowdhury A Park E Canh DG Chien BT Tung TV Naficy A Rao MR Ali M Lee H Sy TH Nichibuchi M Clemens J Trach DD Clinical, epidemiological, and socioeconomic analysis of an outbreak of Vibrio parahaemolyticus in Khanh Hoa Province, Vietnam The Journal of Infectious Diseases 2002 186 1615 20 12447738 10.1086/345731
Kim YB Okuda J Matsumoto C Takahashi N Hashimoto S Nishibuchi M Identification of Vibrio parahaemolyticus strains at the species level by PCR targeted to the toxR gene Journal of Clinical Microbiology 1999 37 1173 1177 10074546
Clemens J Kotloff KL Kay B Generic protocol to estimate the burden of Shigella diarrhoea and dysenteric mortality WHO/V&B/9926 Geneva: WHO 1999
Winer BJ Statistical principles in experimental design 1992 New York: McGraw-Hill
Berke O Exploratory disease mapping: kriging the spatial risk function from regional count data Int J Health Geogr 2004 3 18 15333131 10.1186/1476-072X-3-18
Cressie NAC Statistics for spatial data 1991 John Wiley & Sons Inc
Oliver MA Webster R Kriging A Method of Interpolation for Geographical Information Systems International Journal of Geographical Information Systems 1990 4 313 32
Fortin M-J Jacquez GM Randomization tests and spatially autocorrelated data Bulletin of the Ecological Society of America 2000 81 201 205
Eastman JR Guide to GIS and image processing 1999 2 Worcester, MA: Clark Labs, Clark University
Briggs DJ Elliott P The use of geographical information systems in studies on environment and health World Health Statistics Quarterly 1995 48 85 94
Kafadar K Simultaneous smoothing and adjusting mortality rates in U.S. counties: melanoma in white females and white males Statistics in Medicine 1999 18 3167 3188 10602143 10.1002/(SICI)1097-0258(19991215)18:23<3167::AID-SIM308>3.0.CO;2-N
Susser M The logic in ecological: I.The logic of analysis Am J Public Health 1994 84 825 829 8179056
Goovaerts P Jacquez GM Accounting for regional background and population size in the detection of spatial clusters and outliers using geostatistical filtering and spatial neutral models: the case of lung cancer in Long Island, New York Int J Health Geogr 2004 3 14 15272930 10.1186/1476-072X-3-14
Bucher F Craglia M, Onsrud H Using extended exploratory data analysis for the section of an appropriate interpolation model Geographic Information Research: Trans-Atlantic Perspectives 1999 London: Taylor and Francis 391 403
Wen SW Demissie K August D Rhoads GG Level of aggregation for optimal epidemiological analysis: the case of time to surgery and unnecessary removal of the normal appendix J Epidemiol Community Health 2001 55 198 203 11160175 10.1136/jech.55.3.198
Scholten HJ de Lepper MJC The benefits of the application of geographical information systems in public environment health World Health Statistics Quarterly 1991 44 160 170
Onsrud HJ Pinto JK Diffusion of geographic information innovations Int J Geographic Information Systems 1991 5 447 467
| 15927082 | PMC1156930 | CC BY | 2021-01-04 16:39:06 | no | Int J Health Geogr. 2005 Jun 1; 4:12 | utf-8 | Int J Health Geogr | 2,005 | 10.1186/1476-072X-4-12 | oa_comm |
==== Front
Immun AgeingImmunity & ageing : I & A1742-4933BioMed Central London 1742-4933-2-91592708010.1186/1742-4933-2-9ResearchTotal and specific serum IgE decreases with age in patients with allergic rhinitis, asthma and insect allergy but not in patients with atopic dermatitis Mediaty Anja [email protected] Karsten [email protected] Department of Dermatology, University Hospital Eppendorf, Martinistr. 52, 20246 Hamburg, Germany2005 31 5 2005 2 9 9 31 1 2005 31 5 2005 Copyright © 2005 Mediaty and Neuber; licensee BioMed Central Ltd.2005Mediaty and Neuber; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Concerning allergic diseases, the incidence of allergic symptoms, as well as their severity, seems to decrease with age. The decline of onset of allergic symptoms observed in ageing might result from a decrease of serum total and specific IgE. Atopic disorders are complex diseases that involve interactions among several physiological systems, e.g. skin, lung, mucosae, and the immune system. It was the aim of this study to compare the effects of age on total and specific IgE in patients with atopic dermatitis (AD), allergic rhinitis or asthma, and insect allergy, respectively.
The study population consisted of 559 individuals (male: 229 and female: 330). Total and allergen specific IgE was measured in every individual. From the whole study population, 113 patients suffered from atopic dermatitis (AD), 132 had allergic rhinitis or asthma, and 314 were tested because of insect allergy. Total and specific serum IgE was significantly decreased as a function of age in patients with allergic rhinitis and asthma and with insect allergy. In contrast, no significant decrease of total and specific serum IgE in old individuals with AD was observed. Additionally, in the group of patients with a total IgE < 300 kU/l a reduction of total serum IgE was significantly correlated with age. In contrast, patients with IgE levels > 300 kU/l showed no correlation with age.
Immunosenescence does not affect increased IgE levels in atopic patients with AD and/or high serum IgE levels indicating that in these subgroups of patients the atopic propensity remains into advanced age. One may hypothesize that either onset of allergic sensitization during life or the kind of atopic disease influences the correlation between age and IgE synthesis.
==== Body
Introduction
Concerning allergic diseases, the incidence of onset of allergic symptoms, as well as their severity seems to decrease with age [1]. Atopy is a relatively common, adverse humoral immune system response to common environmental agents (allergens) involving the production of allergen-specific IgE. Epidemiological investigations of allergen sensitivity in a community-based population [2-5] and an industrial setting [6,7] show that the propensity for and the relative incidence of atopic disorders tend to change with age. Serum total IgE values decline with age in the general population, and there are significantly fewer cases of atopy among elderly subjects (60 years and older) compared with younger subjects. However, atopic disorders are complex diseases that involve interactions among several physiological systems, e.g. skin, lung, mucosae, and the immune system.
The immune system undergoes characteristic changes with ageing. Most tests of T cell function are depressed in elderly individuals and the deterioration of the immune systems is believed to contribute to morbidity and mortality in man [8].
Age-associated alterations in the proportions of T cell subsets have been well documented in humans. There are clearly more CD4+ CD45RO+ "memory-phenotype" cells and less CD45RA+ "naïve-phenotype" cells in peripheral blood mononuclear cells (PBMC) from elderly individuals. The accumulation of CD45RO+ memory cells results in a reduced ability to respond to new antigens, and a retained ability to respond to recall antigens, as long as the memory cells remained present and functional [9-11].
Data on age-associated alterations in cytokine secretion in human are inconsistent. Th1 cells are characterised by their ability to secrete IFNγ, IL-2, TNF-α, IL-12 and IL-15, whereas Th2 cells are characterised by IL-4, IL-5, IL-6, IL-10 and IL-13 secretion. Infant humans exhibit impaired cellular but strong humoral immunity, and are in a type 2-dominant state. Soon thereafter, a type 1-state becomes dominant and persists in healthy humans until mid-to-later life, at which time a dominant type 2 cytokine profile may again emerge [12]. Paradoxically, despite declining T cell function, ageing does nor appear always to be associated with decreased immune reaction. For instance, there are increased levels of autoantibodies in the aged [13], resulting in autoimmune diseases (e.g. bullous pemphigoid).
Current thought posits that age-associated changes in B-cell function are primarily due to changes in the T-cell compartment and a resultant dysregulation on B-cell function [14]. Altered function is primarily associated with exposure to "new" antigenic challenges, as for example the less efficient therapeutic value of inoculation regimes among the elderly population. Although both T cells and B cells are involved with the onset of atopy, the effector mechanisms are principally humorally driven and focused on the production and regulation of IgE. To what extent this regulation is T cell dependent or T cell independent is not well stablished. Recently, it has been demonstrated that in healthy non-atopic individuals no differences in either serum total IgE and soluble CD23 or Th2 related cytokines (IL-4, IL-10 and IL-13) between young and old subjects were observed [15].
However, there are not many studies investigating the relationship between age and IgE in different atopic diseases. In this study, we evaluated the effects of age on total and specific IgE in patients with atopic dermatitis (AD), allergic rhinitis or asthma, and insect allergy.
Methods
Study population
The population of this retrospective study consisted of 559 individuals (male: 229 and female: 330) randomly selected from a data bank of patients diagnosed for one of the relevant diseases. The diagnoses were made on the basis of the case history and the clinical examination. Total and/or allergen specific IgE was measured in every individual. From the whole study population, 113 patients suffered from AD, 132 had allergic rhinitis or asthma, and 314 were tested because of insect allergy.
Determination of serum IgE
Total and allergen specific serum IgE was measured by using the UniCAP System (Pharmacia, Freiburg, Germany) according to the instructions of the manufacturer. Directly after collection of venous blood, the blood was centrifuged and serum was stored at -20°C until testing. Total serum IgE levels < 300 kU/l are in the normal or borderline range, amounts ≥ 300 kU/l were classified as pathological [17]. Serum IgE specific was determined at least for one of the following antigens: Dermatophagoides pteronyssinus and farinae (Der p 1 and Der f 1), Betula verrucosa (Bet v 1), Phleum pratense (Phl p 1), Felis domesticus (Fel d 1), and hymenoptera venoms. In patients with allergic asthma or rhinitis and in patients with AD Der p 1 and Der f 1, Bet v 1, Phl p 1, and Fel d 1 were measured. In patients with insect allergy specific IgE to hymenoptera venoms was tested. For patients polysensitized against several allergens, the highest value of specific IgE was selected and only this value was used for statistical analysis.
Statistical analysis
Statistical analysis for paired comparisons of total and specific IgE between different patient groups was conducted by Student's t test. In order to assess the relationship between IgE and age, we calculated Kendall's tau-b and the corresponding p-values to test the hypothesis that both variables are not associated [18]. Additionally, we used simple linear regression to evaluate age-associated changes in serum IgE. Statistical analysis was performed by using the software program SPSS 9.0.
Results
Patients characteristics
The study population consisted of 559 patients suffering from either AD (n = 113), allergic asthma or rhinitis (n = 132) or insect allergy (n = 314) tested for total and specific serum IgE (Table 1). Mean age of patients with AD was 23.5 years, whereas mean age of the patients with allergic rhinitis, asthma or insect allergy was 36.7 years and 44.2 years, respectively (Table 1). In the group of patients > 65 year old 6 (5.3%) had AD, 24 (18.2%) sufferd from rhinitis or asthma, and 105 (33.4%) from insect allergy. Twentytwo patients with insect allergy were younger than 10 years. In the group of patients with insect allergy older than 65 years the onset of symptoms was always over 60 years but not in the group of patients with AD and allergic rhinitis or asthma where symptoms onset always started earlier. Independent of age, the highest amounts of total and specific IgE (Figure 1) were found in patients with AD compared with the other patient groups (p < 0.01; p < 0.02).
Table 1 Correlation between age and IgE. Characteristics of patients and Kendall's tau-b for relationship between IgE and age.
Mean Age ± SD (range) Age (yr) Number of patients (%) Total IgE [kU/l] vs Age Specific IgE [kU/l] vs Age
>65 >65 male female τb (n) p τb (n) p
All patients 36.9 ± 22.6 (0.5 – 93) 424 135 229 (40.9) 330 (59.0) -0.262 (553) < 0.0001 -0.176 (485) < 0.0001
Atopic dermatitis 23.6 ± 20.1 (0.5 – 93) 107 6 53 (46.9) 60 (53.1) -0.038 (112) = 0.561 0.034 (68) = 0.696
Allergic Rhinitis/Asthma 36.7 ± 19.6 (3.0 – 79) 108 24 50 (37.9) 82 (62.1) -0.219 (132) < -0.0001 -0.320 (107) < 0.0001
Insect Allergy 44.2 ± 21.7 (2.0 – 82) 209 105 126 (40.1) 188 (59.9) -0.259 (309) < 0.0001 -0.084 (310) = 0.029
Figure 1 Total and specific IgE levels. Means ± SEM of total and specific IgE [kU/l] for Der p 1/f 1, Bet v 1, Phl p 1, and Fel d 1 in patients with (a) AD and with (b) allergic asthma or rhinitis as well as for (c) wasp and bee allergen in patients with insect allergy are shown. Total and specific IgE is sicnificantly increased in patients with AD compared with patients suffering from allergic asthma or rhinitis and insect allergy.
Ageing and total serum IgE
The total IgE had a negative correlation (p < 0.0001) with age in all patients as well as in the subgroups of patients with allergic rhinitis and asthma, and insect allergy. In contrast, no significant decrease of total serum IgE as a function of age was observed in patients with AD (Table 1).
Figure 2 shows the values of total serum IgE (kU/l) against donors age for patients with AD, allergic asthma or rhinitis, and for insect allergy. For patients with allergic asthma and rhinitis (F(1,130) = 10.850, p = 0.001) as well as for patients with insect allergy (F(1,307) = 16.236, p <0.0001), the slopes of the lines derived from simple linear regression do vary significantly from zero, but do not for patients with AD (F(1,110) = 1,734, p = 0.191).
Figure 2 Correlation between age and IgE. Multiple scatter plots of donor age versus total (●) and specific (○) serum IgE. (a) patients with AD, (b) patients with allergic asthma or rhinitis; (c) patients with insect allergy. R1, line for total IgE and R2 for specific IgE derived from simple linear regression.
Additionally, in the group of patients with a total IgE < 300 kU/l a reduction of total serum IgE was significantly correlated with age (p <0.0001). In contrast, patients with IgE levels > 300 kU/l showed no correlation with age (Table 2). From the patients with IgE levels > 300 kU/l, 16 (11.9%) were over 65 years old. Linear regression analysis revealed F(1,400) = 20.955 (p <0.0001) for total IgE levels < 300 kU/l but were not significant for IgE levels > 300 kU/l (F(1,149) = 1,138, p = 0.288)
Table 2 Correlation between age and IgE. Kendall's tau-b for relationship between IgE (< 300 kU/l or > 300 kU/l) and age.
Total IgE Mean Age ± SD (range) Age (yr) Number of patients (%) Total IgE [kU/l] vs Age
< 65 > 65 male female τb (n) p
< 300 kU/l 42.9 ± 21.4 (1.0 – 82) 274 118 148 (36.8) 254 (63.2) -0.151 (402) < 0.0001
> 300 kU/l 26.0 ± 20.1 (0.5 – 93) 151 16 80 (53.0) 71 (47.0) -0.055 (151) = 0.320
Ageing and allergen specific IgE
Figure 1 shows specific IgE values for each tested allergen in the different disease groups. No significant differences of mean IgE values between the different allergens could be detected except in patients with insect allergy. In patients with insect allergy only wasp and bee allergen were tested and the mean specific IgE for bee was significantly (p = 0.02) lower compared with wasp specific IgE. The majority of these patients (91.7%) had immunotherapy because of wasp allergy, whereas 8.3% had immunotherapy because of clinical relevant sensitization (Table 3). In patients with atopic dermatitis or with rhinitis/allergic asthma relevant sensitization (0.35 kU/l) against the different allergens was distributed homogenously. Only sensitization to Fel d 1 was less frequent in patients with allergic rhinitis or asthma (Table 3).
Table 3 Frequencies of allergens in the disease groups. Number of subjects in each disease group who were positive at ≤ 0.35 kU/l to the tested allergens.
Disease Atopic Dermatitis Allergic Rhinitis and Asthma Insect Allergy
Allergen (n = 113) (n = 132) (n = 314)
Der p 1 and f 1 26 (23.0%) 39 (29.5%) --
Bet v 1 27 (23.9%) 43 (32.6%) --
Phl p 1 24 (21.2%) 40 (30.3%) --
Fel d 1 22 (19.5%) 23 (17.4%) --
Wasp -- -- 288 (91.7%)
Bee -- -- 26 (8.3%)
Allergen specific IgE was significantly decreased in the elderly suffering from allergic rhinitis, allergic asthma (p < 0.0001) and insect allergy (p = 0.029), respectively. On the other hand, no correlation between specific IgE and age was observed in patients with AD (Table 1).
For patients with allergic asthma and rhinitis (F(1,130) = 26.437, p = 0.0001) the slopes of the lines derived from simple linear regression do vary significantly from zero (Figure 2). For patients with insect allergy (F(1,308) = 0.110, p = 0.740) and for patients with AD (F(1,68) = 0,283, p = 0.596), linear regression did not reveal significant variation from zero for allergen specific IgE (Figure 2).
Discussion
From a clinical perspective, supported by epidemiological investigations, there would appear to be a decline with age in both the incidence and severity of atopic diseases, particularly among the elderly population who are 60 years and older [2-7]. Atopic incidence declines, symptoms severity declines, and there would appear to be a general humoral alteration of the propensity for atopy reflected by age-associated declines in serum total IgE values. However, the results of this study demonstrate that the association between age and serum IgE is dependent on the type of atopic or allergic disease and is also dependent on the amount of total serum IgE.
The results of our study are supported by data that have been published recently by Jackola et al. [16]. In patients with allergic asthma they observed a decline with age in serum total IgE values. Moreover, they showed that among those asthmatic individuals sensitized to ragweed antigen, there was no age-associated change in IgE levels specific to the major ragweed allergen. This robustness of the underlying atopic mechanism, represented by specific IgE was also observed in our patients with AD but not in the patients with allergic asthma or rhinitis and in patients with insect allergy. Concerning insect allergy, this difference might be partly explained by the later onset of sensitization against hymenoptera venoms during life and thus, other immunological mechanisms regulating the production of specific IgE in patients with insect allergy, which has been defined as hypersensitivity allergic nonatopic IgE-mediated disease [17]. This assumption is supported by the fact that in the group of patients with insect allergy older than 65 years the onset of symptoms was always over 60 years but not in the group of patients with AD and allergic rhinitis or asthma.
The differences according age-association of allergen-specific IgE in patients with asthma between the two studies might be the consequence of patient's selection. The subgroup of patients with asthma in our study probably was more heterogenous compared with the population investigated by Jackola and co-workers. Moreover, it is not clear whether asthma patients with additional manifestations of other atopic diseases (e.g. AD or rhinitis) were excluded in their study or not.
As expected, the numbers of patients with AD older than 60 years were rather small. However, a higher proportion of AD patients were over 40 years old indicating that in this group of adults IgE production is unaffected by age.
The results of our study support the assumption that age-association of total and specific serum IgE is dependent on the type of atopic disease the patients are suffering from and the age at disease onset.
Conclusion
In summary, this study shows that total and allergen specific IgE production is reduced in the elderly with the exception of old patients with either high serum IgE or AD, indicating that atopic mechanisms underlying AD or other atopic diseases with high serum IgE are particularly robust, and the atopic propensity among these patients remains into advanced age. However, further studies are required to clarify the immunological mechanisms which are responsible for IgE synthesis during immunosenscence.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AM participated in the design of the study, performed the statistical analysis and drafted the manuscript. KN conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.
Acknowledgements
We thank Dr. Bubenheim (Institute for Medical Biometry and Epidemiology, University Hospital Hamburg Eppendorf) for help with statistical analysis. Gerlinde Finger and Birgit Maehnss are thanked for excellent technical assistance. This work was supported by grants from the Leidenberger-Müller-Foundation and from the Buch-Foundation, Germany.
==== Refs
Hanneuse Y Delespesse G Hudson D de Halleux F Jacques JM Influence of ageing on IgE-mediated reactions in allergic patients Clin Allergy 1978 8 165 174 417878
Barbee RA Brown WG Kaltenborn W Halonen M Allergen skin-test reactivity in a community population sample: correlation with age, histamine skin reactions and total serum immunoglobulin E J Allergy Clin Immunol 1981 68 15 19 7240595
Barbee RA Halonen M Lebowitz M Burrows B Distribution of IgE in a community population sample: correlations with age, sex, and allergen skin test reactivity J Allergy Clin Immunol 1981 68 106 111 7251998
Barbee RA Halonen M Kaltenborn W Lebowitz M Burrows B A longitudinal study of serum IgE in a community cohort: correlations with age, sex, smoking, and atopic status J Allergy Clin Immunol 1987 79 919 927 3584747 10.1016/0091-6749(87)90241-7
Barbee RA Kaltenborn W Lebowitz MD Burrows B Longitudinal changes in allergen skin test reactivity in a community population sample J Allergy Clin Immunol 1987 79 16 24 3492524 10.1016/0091-6749(87)90241-7
Freidhoff LR Meyers DA Bias WB Chase GA Hussain R Marsh DG A genetic-epidemiologic study of human immune responsiveness to allergens in an industrial population: I. Epidemiology of reported allergy and skin-test positivity Am J Med Genet 1981 9 323 340 7294070
Freidhoff LR Meyers DA Marsh DG A genetic-epidemiologic study of human immune responsiveness to allergens in an industrial population. II. The associations among skin sensitivity, total serum IgE, age, sex, and the reporting of allergies in a stratified random sample J Allergy Clin Immunol 1984 73 490 499 6707392 10.1016/0091-6749(84)90360-9
Pawelec G Barnett Y Forsey R Frasca D Globerson A McLeod J Caruso C Franceschi C Fulop T Gupta S Mariani E Mocchegiani E Solana R T cells and aging, January 2002 update Front Biosci 2002 7 d1056 183 11991846
Hannet I Erkeller-Yuksel F Lydyard P Deneys V Bruyere MD Developmental and maturational changes in human blood lymphocyte subpopulations Immunol Today 1992 13 215 218 1627249 10.1016/0167-5699(92)90157-3
Gabriel H Schmitt B Kindermann W Age-related increase of CD45RO+ lymphocytes in physically active adults Eur J Immunol 1993 23 2704 2706 8104800
Flurkey K Stadecker M Miller RA Memory lymphocyte-T hyporesponsiveness to non-cognate stimuli - a key factor in age-related immunodeficiency Eur J Immunol 1992 22 931 935 1532363
Shearer GM Th1/Th2 changes in aging Mech Ageing Dev 1997 94 1 5 9147355 10.1016/S0047-6374(96)01849-0
Klinman NR Kline GH The B-cell biology of aging Immunol Rev 1997 160 103 114 9476669
LeMaoult J Szabo P Weksler ME Effect of age on humoral immunity, selection of the B-cell repertoire and B-cell development Immunol Rev 1997 160 115 126 9476670
Di Lorenzo G Pacor ML Esposito Pellitteri M Listi F Colombo A Candore G Mansueto P Lo Bianco C Ditta V Battista Rini G Caruso C A study of age-related IgE pathophysiological changes Mech Ageing Dev 2003 124 445 448 12714251 10.1016/S0047-6374(03)00020-4
Kendall M Gibbons JD Rank Correlation Methods 1990 5 London, Edward Arnold 220
Jackola DR Pierson-Mullany LK Daniels LR Corazalla E Rosenberg A Blumenthal MN Robustness into advanced age of atopy-specific mechanisms in atopy-prone families J Gerontol A Biol Sci Med Sci 2003 58 99 107 12586846
Johansson SG Hourihane JO Bousquet J Bruijnzeel-Koomen C Dreborg S Haahtela T Kowalski ML Mygind N Ring J van Cauwenberge P van Hage-Hamsten M Wuthrich B A revised nomenclature for allergy. An EAACI position statement from the EAACI nomenclature task force Allergy 2001 56 813 824 11551246 10.1034/j.1398-9995.2001.t01-1-00001.x
| 15927080 | PMC1156931 | CC BY | 2021-01-04 16:36:32 | no | Immun Ageing. 2005 May 31; 2:9 | utf-8 | Immun Ageing | 2,005 | 10.1186/1742-4933-2-9 | oa_comm |
==== Front
J Inflamm (Lond)Journal of Inflammation (London, England)1476-9255BioMed Central London 1476-9255-2-41592152610.1186/1476-9255-2-4ResearchCXCR2 is critical for dsRNA-induced lung injury: relevance to viral lung infection Londhe Vedang A [email protected] John A [email protected] Michael P [email protected] Marie D [email protected] Ying Ying [email protected] Robert M [email protected] Department of Pediatrics, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA2 Division of Pulmonary and Critical Care Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA3 Department of Pathology and Laboratory Medicine, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA2005 28 5 2005 2 4 4 2 12 2004 28 5 2005 Copyright © 2005 Londhe et al; licensee BioMed Central Ltd.2005Londhe et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Respiratory viral infections are characterized by the infiltration of leukocytes, including activated neutrophils into the lung that can lead to sustained lung injury and potentially contribute to chronic lung disease. Specific mechanisms recruiting neutrophils to the lung during virus-induced lung inflammation and injury have not been fully elucidated. Since CXCL1 and CXCL2/3, acting through CXCR2, are potent neutrophil chemoattractants, we investigated their role in dsRNA-induced lung injury, where dsRNA (Poly IC) is a well-described synthetic agent mimicking acute viral infection.
Methods
We used 6–8 week old female BALB/c mice to intratracheally inject either single-stranded (ssRNA) or double-stranded RNA (dsRNA) into the airways. The lungs were then harvested at designated timepoints to characterize the elicited chemokine response and resultant lung injury following dsRNA exposure as demonstrated qualititatively by histopathologic analysis, and quantitatively by FACS, protein, and mRNA analysis of BAL fluid and tissue samples. We then repeated the experiments by first pretreating mice with an anti-PMN or corresponding control antibody, and then subsequently pretreating a separate cohort of mice with an anti-CXCR2 or corresponding control antibody prior to dsRNA exposure.
Results
Intratracheal dsRNA led to significant increases in neutrophil infiltration and lung injury in BALB/c mice at 72 h following dsRNA, but not in response to ssRNA (Poly C; control) treatment. Expression of CXCR2 ligands and CXCR2 paralleled neutrophil recruitment to the lung. Neutrophil depletion studies significantly reduced neutrophil infiltration and lung injury in response to dsRNA when mice were pretreated with an anti-PMN monoclonal Ab. Furthermore, inhibition of CXCR2 ligands/CXCR2 interaction by pretreating dsRNA-exposed mice with an anti-CXCR2 neutralizing Ab also significantly attenuated neutrophil sequestration and lung injury.
Conclusion
These findings demonstrate that CXC chemokine ligand/CXCR2 biological axis is critical during the pathogenesis of dsRNA-induced lung injury relevant to acute viral infections.
chemokinesneutrophilsviral infectionlung injury.
==== Body
Background
Viral infections of the respiratory tract are a cause of the common cold and flu in children and adults. These infections may predispose certain patients to develop chronic respiratory disorders such as asthma, chronic obstructive pulmonary disease (COPD), pulmonary fibrosis, and bronchopulmonary dysplasia (BPD) [1]. Clinical symptoms include mucus secretion and altered airway reactivity and are hallmarked by the recruitment of inflammatory cells with resultant changes to the airway epithelial cell lining. Inflammation can also extend further into the lung to cause parenchymal disease that is characteristic of viral pneumonia as has been observed recently in severe acute respiratory syndrome (SARS) [2]. Inflammatory cell recruitment in large part is elicited by the generation of chemokines (chemotactic cytokines) that are also important in establishing a pro-inflammatory environment underlying chronic respiratory disorders such as asthma, COPD, cystic fibrosis, pulmonary fibrosis, and BPD [1,3-10].
Inflammatory changes due to viral infection result from the host immune response rather than secondary to viral replication or the viral particles themselves [11-15]. Viral infections of epithelial cells are characterized by the generation of the pro-inflammatory molecule double-stranded RNA (dsRNA) during intracellular replication of viruses. When studied in human epithelial cell lines in vitro, dsRNA triggers an innate immune response in host cells via generation of cytokines and chemokines involved in inflammatory cell recruitment. Specifically, dsRNA has been shown to induce activation of the neutrophil chemoattractant, interleukin-8 (IL-8/CXCL8), and regulated on activation, normal T cells expressed and secreted (RANTES) [16]. In human subjects in vivo, elevated tracheal IL-8/CXCL8 levels and neutrophil accumulation are found in airways of patients with asthma, COPD, and viral infection. While animal models in vivo have largely studied the systemic effects of intraperitoneal dsRNA treatment [17], there is a paucity of information on characterization and role of chemokines in lung inflammation and injury following intratracheal dsRNA instillation.
Murine KC/CXCL1 and MIP-2/CXCL2/3 are Glutamic acid-Leucine-Arginine-positive (ELR-positive) CXC chemokines; are structural homologs of human GRO-α/CXCL1 and GRO-β/γ/CXCL2/3, respectively; and are functional homologs of human CXC chemokines, such as IL-8/CXCL8, ENA-78/CXCL5, and GRO-α/β/γ/CXCL1/2/3 [18-21]. Both murine chemokines share the ability to signal through a G protein-coupled receptor, CXCR2 [18-20]. Their human structural and functional homologs have been associated with asthma, COPD, and viral infections of the lung [22,23].
In the current study, we hypothesized that the early inflammation and resultant lung injury from intratracheal dsRNA treatment is due, in part, to the expression of ELR-positive CXC chemokines through their interaction with their major receptor, CXCR2. To test this hypothesis, we injected dsRNA intratracheally into 6–8 week old female BALB/c mice to measure neutrophil and chemokine responses and resultant injury in the airway and lung tissue compartments. We then blocked this response by pretreating animals with antibodies to specifically neutralize neutrophil recruitment in a chemokine-dependent manner and thereby decreased lung inflammation and injury. Our animal model demonstrates the critical role of CXCR2 ligands/CXCR2 in acute lung inflammation and injury due to intratracheal dsRNA.
Methods
Reagents
RNA instillation: Double-stranded RNA (dsRNA, Poly IC) and single-stranded RNA (ssRNA, Poly C) were purchased from Sigma-Aldrich Corp. (St. Louis, Mo.) and reconstituted in sterile normal saline (20 μg/μl) and stored at 4°C prior to use.
Enzyme-linked immunoadsorption assay (ELISA) experiments
Capture and Detection antibodies to murine KC/CXCL1 and murine MIP-2/CXCL2/3 were purchased as DuoSet® from R&D Systems (Minneapolis, MN).
Neutralization studies: Purified rat anti-mouse Ly-6G (Gr-1) mAb (clone RB6-8C5) was purchased from BD Pharmingen (San Diego, CA) and was used for neutrophil depletion studies as previously described [24]. Polyclonal goat anti-murine CXCR2 was produced by the immunization of a goat with a peptide containing the ligand-binding sequence Met-Gly-Glu-Phe-Lys-Val-Asp-Lys-Phe-Asn-Ile-Glu-Asp-Phe-Phe-Ser-Gly of CXCR2 [24-31]. The goat was immunized with CXCR2 in multiple intradermal sites with complete Freund's adjuvant (CFA) followed by at least 3 boosts of CXCR2 in incomplete Freund's adjuvant (IFA) as previously described. [24-31]. Direct ELISA was used to evaluate antisera titers, and sera was used for Western blot, ELISA and neutralization assays when titers had reached greater than 1/1,000,000. The CXCR2 protein sequence has been shown to contain the ligand-binding portion of the CXCR2 receptor [24-26,32]. The anti-CXCR2 antibodies have been used previously to block mouse CXCR2 in vivo, and has been shown to detect CXCR2 by Western blot and fluorescence-activated cell sorting analysis of neutrophils in vivo [24-26,32]. The anti-CXCR2 antibody has been shown to be neutralizing using both in vitro neutrophil chemotaxis assay and in vivo by abrogating the influx of neutrophils into the peritoneum of normal mice in response to exogenous ELR-positive murine CXC chemokines [24-26,32]. In vivo administration of anti-CXCR2 antibodies inhibited pulmonary neutrophil sequestration in murine models of Aspergillosis, Nocardia, and Pseudomonas pneumonia and prevented the influx of neutrophils in urine and the kidney in a murine model of Escherichia coli urinary tract infection [24-26,32]. Moreover, intraperitoneal administration of this antibody did not alter peripheral blood neutrophil counts [24-26,32]. 1 ml of antiserum against mCXCR2 and control antibody is approximately 10 mg of IgG.
Murine model of dsRNA-induced lung injury
We used 6–8 week old female BALB/c mice to intratracheally inject either single-stranded (ssRNA) or double-stranded RNA (dsRNA) using a modification of a previously described method [30,33-36]. Mice were anesthetized with ketamine (60 mg/kg) intraperitoneally; then, under sterile conditions, the anterior neck soft tissue was dissected to expose the trachea and 50 μl RNA (20 μg/μl; 40 μg/g mouse wt) was injected via 26 gauge tuberculin needle and syringe attached to a Stepper® microinjector (Indicon, Inc., Brookfield, CT) into the trachea under direct visualization. Immediately following the instillation, the skin was apposed and closed using tissue adhesive and the mice were allowed to recover from anesthesia prior to replacement into their cages.
In separate experiments, animals received either 1 ml of goat polyclonal anti-murine CXCR2, 1 ml of normal goat serum (NGS) control antibody, or 0.5 ml rat anti-mouse Ly-6G mAb or corresponding control intraperitoneally 24 hours before intratracheal injection and daily until time of sacrifice as previously described [37].
Lung bronchoalveolar lavage and tissue preparation
At time of sacrifice, 72 h following intratracheal dsRNA or ssRNA treatment, mice were euthanized using intraperitoneal Pentobarbital (100 mg/kg) and a heparinized sample of blood was harvested. The thoracic cavity was then exposed and lungs were perfused free of blood with 1 ml 0.9% normal saline via the spontaneously beating right ventricle under constant pressure of 25 cm H20. A 26 gauge butterfly needle was used to cannulate the trachea and bronchoalveolar lavage (BAL) was performed by instilling 1 ml PBS + 5 mM EDTA solution as previously described [38]. Lungs were lavaged under constant pressure of 25 cm H20 and retrieved solutions were centrifuged at 900 × g for 15 min. The cell-free supernatants were assayed by specific ELISAs and collected cells were analyzed for total cell counts and cytospin differentials. Lung tissue was then processed for the following: calculation of lung edema; microvascular permeability; mRNA; ELISA analysis; and histopathological and immunohistochemical analysis by fixing in 4% paraformaldehyde at 25 to 30 cm H2O pressure and embedding in paraffin.
Immunolocalization of TLR3
Paraffin-embedded tissues from dsRNA-treated and ssRNA-treated lungs were processed for immunohistochemical localization of Toll-like receptor 3 (TLR3) expression using a method previously described [33,39]. Briefly, tissue sections were dewaxed with xylene and rehydrated through graded concentrations of ethanol. Tissue nonspecific binding sites were blocked using Power Block® (BioGenex, San Ramon, CA). Tissue sections were overlaid with 1:50 dilution of either control (goat) or polyclonal goat anti-TLR3 antibody (Santa Cruz Biotechnology Inc., Santa Cruz, CA). The tissue sections were washed with TRIS-buffered saline and then incubated for 60 min with secondary biotinylated antibody. The tissue sections were then washed in TRIS-buffered saline and incubated with alkaline phosphatase conjugated to streptavidin (BioGenex). Tissue sections were then incubated with Vectastain ABC reagent (Vector Laboratories, Burlingame, CA) followed by the peroxidase substrate, DAB reagent (Vector Laboratories). After optimal color development, tissue sections were immersed in sterile water, counterstained with Lerners hematoxyin, and cover slipped using an aqueous mounting solution.
Total RNA isolation and real-time quantitative PCR
Total cellular RNA from lung tissue was isolated as previous described [30,31]. Total RNA was determined and 1 ug of total RNA was reversed transcribed into cDNA and amplified using TaqMan Gene Expression Quantification assays (Applied Biosystems (Foster City, CA) Kit 4304134). cDNA was amplified and quantified using the TaqMan 7700 Sequence Detection System and specific primers for murine CXCL1, murine CXCL2/3, murine CXCR2 and a housekeeping gene18S. The primers used were 5'-TGA-GCT-GCG-CTG-TCA-GTG-CCT-3' (sense) and 5'-AGA-AGC-CAG-CGT-TCA-CCA-GA-3' (antisense) for CXCL1 (259 bp) and 5'-GCT-GGC-CAC-CAA-CCA-CCA-GG-3' (sense) and 5'-AGC-GAG-GCA-CAT-CAG-GTA-CG-3' (antisense) for murine CXCL2/3 (359 bp). Predeveloped assay reagents (Applied Biosystems Kit 4304134) were used for murine CXCR2 and the housekeeping gene, 18S. Quantitative analysis of gene expression was done using the comparative CT (ΔCT) methods, in which CT is the threshold cycle number (the minimum number of cycles needed before the product can be detected)[40,41]. The arithmetic formula for the ΔCT method is the difference in threshold cycles for a target, (i.e., CXCR2) and an endogenous reference (i.e., housekeeping gene 18S). The amount of target normalized to an endogenous reference (i.e., CXCR2 in dsRNA-treated animals) and relative to a calibration normalized to an endogenous reference (i.e., CXCR2 in ssRNA-treated controls) is given by 2-ΔΔCT [40,41]. The following is an example for comparing CXCR2 expression from dsRNA-treated animals and ssRNA-treated controls. Both CXCR2 from dsRNA-treated and ssRNA-treated controls are normalized to 18S: ΔΔCT = ΔCT (CXCR2 expression from dsRNA-treated animals normalized to endogenous 18S)-ΔCT (CXCR2 expression from ssRNA-treated controls normalized to endogenous 18S). The calculation of 2-ΔΔCT then gives a relative value when comparing the target with the calibrator, which we designate in this context as fold increase of dsRNA-treated animals to ssRNA-treated controls of the target mRNA relative quantification.
Evans blue microvascular permeability and wet:dry analysis of lung edema
Microvascular permeability related to lung injury was measured using a modification of the Evans blue dye extravasation technique, as previously described [30,42]. Extravasation of Evans blue (Sigma-Aldrich) into the extravascular compartment was used as a quantitative measure of lung injury and changes in microvasculature permeability. Briefly, each animal received 20 mg/kg Evans blue (pH 7.34) by tail vein injection 3 h before sacrifice. At the time of sacrifice, a heparinized sample of blood was harvested, and plasma was removed by centrifugation. Six lungs from each group were perfused free of blood with 1 ml 0.9% normal saline via the spontaneously beating right ventricle and removed from the thoracic cavity. The trachea, mainstem bronchi, and surrounding mediastinal structures were removed. Evans blue was extracted from pulmonary tissues after homogenization in 1 ml of 0.9% normal saline. This volume was added to 2 vol of deionized formamide and incubated at 60C for 12 h. The supernatant was separated by centrifugation at 2000 × G for 30 min. Evans blue in the plasma and lung tissue was quantitated by dual-wavelength spectrophotometric analysis at 620 and 740 nm [43]. This method corrects the specimen's absorbance at 620 nm for the absorbance of contaminating heme pigments, using the following formula: corrected absorbance at 620 nm = actual absorbance at 620 nm – (1.426(absorbance at 740) + 0.03). We calculated a permeability index by dividing the correct pulmonary plasma Evans blue absorbance at 620 nm; this index reflects the degree of extravasation of Evans blue into the extravascular pulmonary tissue compartment.
To quantitate lung edema following dsRNA treatment, wet to dry weight ratios were obtained by ligating the lungs away from the hilum as previously described [40]. The lungs were blotted dry and weighed. They were then desiccated by incubation at 130°C overnight in a vacuum oven and re-weighed to determine their dry weight. The wet to dry ratio was then calculated.
KC/CXCL1 and MIP-2/CXCL2/3 ELISAs
KC/CXCL1 or MIP-2/CXCL2/3 protein was quantitated using a modification of a double ligand method as previously described [30,31,40,41]. Briefly, flat-bottomed 96 well microtiter plates (Nunc Immuno-Plate I 96-F) were coated with 50 μl/well of capture antibody to murine KC/CXCL1 or MIP-2/CXCL2/3 (2 ug/ml in sterile phosphate buffered saline (PBS), for 12 hrs at room temperature and then washed with phosphate buffered saline (PBS), pH 7.5, 0.05% Tween-20 (wash buffer). Microtiter plate nonspecific binding sites were blocked with 2% BSA in PBS and incubated for 60 minutes at 37°C. Plates were washed three times with wash buffer and samples or standard were added, followed by incubation for 1 hour at 37°C. Plates were washed three times and 50 μl/well of detection antibody for murine KC/CXCL1 and MIP-2/CXCL2/3 antibodies added, and plates were incubated for 45 minutes at 37°C. Plates were washed three times, streptavidin-peroxidase conjugate (Jackson Laboratories, West Grove, PA) added, and the plates incubated for 30 minutes at 37°C. Plates were washed three times and TMB (3,3,'5,5'-tetramethylbenzidine) chromogen substrate (Kirkegaard & Perry Laboratories, Gaithersburg, MD) added. The plates were incubated at room temperature to the desired extinction, and the reaction terminated with 3 M H2SO4 solution. Plates were read at 450 nm in an automated microplate reader (Bio-Tek Instruments, Inc., Winooski, VT). Standards were 1/2-log dilutions of either KC/CXCL1 or MIP-2/CXCL2/3 from 100 ng to 1 pg/ml (50 ul/well). This ELISA method consistently detected specific chemokine concentrations in a linear fashion greater than 50 pg/ml. KC/CXCL1 and MIP-2/CXCL2/3 were specific in our sandwich ELISA without cross-reactivity to a panel of cytokines including murine C10, JE, MIP-1α, MIP-1β, human GROα, GROβ, GROγ, RANTES, and members of the CXC and CC chemokine families.
FACS analysis of lung neutrophils
Whole lung single cell suspensions were made from harvested lungs from four mice per group using a method, as previously described [40]. Single cell suspensions (5 × 106 cells /ml) were stained with Abs: Tricolor-conjugated (BD Biosciences, Franklin Lakes, NJ) anti-murine CD45 (Caltag Laboratories, South San Francisco, CA), FITC-conjugated anti-murine MOMA-2 (macrophage surface marker; Seratec, Raleigh, NC), R-Phycoerythrin (R-PE) conjugated Rat anti-murine Ly-6G (neutrophil surface marker) and R-PE-conjugated mouse anti-murine CD3e (lymphocyte surface marker) (BD Biosciences). Dual-color-stained cell suspensions were analyzed on a FACScan flow cytometer (Becton Dickinson Immunocytometry Systems, San Jose, CA) using CellQuest 3.2.1f1 software (BD Immunocytometry Systems).
Statistical analysis
Data were analyzed using the Microsoft® Excel 2000 statistical package (Microsoft Corporation, USA). Two group comparisons were evaluated using the unpaired Students t test. Three group comparisons were evaluated by the ANOVA test with the post hoc analysis (i.e. Bonferroni/Dunn). Data were expressed as mean ± SEM.
Results
DsRNA induces expression of TLR3 on airway epithelial cells
Since in vitro studies in human epithelial cells have demonstrated that dsRNA induces the generation of chemokines involved in leukocyte recruitment, we performed in vivo studies using a murine model system of intratracheal dsRNA-induced inflammation and injury to mimic an acute viral infection and thus dissect the mechanisms related to this process. The putative receptor for dsRNA has been identified as Toll-like receptor 3 (TLR3), thus we first determined whether dsRNA treatment was associated with expression of TLR3. Immunolocalization using goat anti-murine TLR3 Ab showed markedly increased expression of TLR3 localized to the surface airway epithelium of dsRNA-treated lungs as compared to ssRNA-treated controls at 72 h. Specificity to TLR3 was demonstrated by lack of staining on control goat IgG-stained sections from both dsRNA and ssRNA-treated groups (Fig. 1).
Figure 1 DsRNA induces expression of TLR3 on airway epithelial cells. Immunohistochemistry of lungs at 72 h following ssRNA or dsRNA treatment. Slides were stained with either control goat IgG or anti-TLR3 Ab (100X) (n = 4 mice per group).
DsRNA induces lung neutrophil infiltration
Mice treated with intratracheal dsRNA were noted to have significant intraparenchymal and airway leukocyte infiltration at 72 h following treatment as compared to naïve and ssRNA-treated controls as demonstrated by histopathology (Fig. 2A). No significant differences in cellular infiltrates were noted among the histopathologic groups at earlier timepoints 4, 12, 24, and 48 h following dsRNA treatment (data not shown). FACS and BAL analysis at 72 h following dsRNA treatment specifically showed that dsRNA induced significant neutrophil recruitment compared to controls (Fig. 2B and 2C).
Figure 2 DsRNA induces lung neutrophil infiltration. (A) Histopathology of naïve lung and at 72 h following treatment with ssRNA or dsRNA. Representative photomicrographs with H & E staining (100X; n = 4 mice per group). (B) Total lung cells via FACS analysis of whole-lung single-cell suspensions at 72 h (n = 4 mice per group; *p < 0.05). (C) Total cells via BAL fluid cell count at 72 h (n = 4 mice per group; *p < 0.05).
DsRNA induces increased lung injury
To determine if the influx of neutrophils into lung airways and parenchyma results in lung injury, we measured two markers of lung injury to quantify changes in lung edema and lung vascular permeability. Results showed a significant increase in the wet:dry ratio in dsRNA-treated lungs as compared to controls at 72 h following dsRNA treatment (Fig. 3A). Similarly, measurement of the lung:plasma extravasation ratio of Evans blue dye also showed a significant increase in microvascular permeability in dsRNA-treated lungs as compared to controls (Fig. 3B).
Figure 3 DsRNA induces lung injury. (A) Wet:Dry ratio at 72 h (n = 6 mice per group; *p < 0.05). (B) Evans blue permeability index at 72 h (n = 6 mice per group; *p < 0.05).
Neutrophil depletion decreases dsRNA-induced lung inflammation and injury
With the finding that an increase in lung neutrophils coincided with an increase in lung injury in dsRNA-treated lungs, we next attempted to determine if this was a causal relationship by inducing neutropenia and measuring changes in lung neutrophil influx and lung injury. Mice were passively immunized with specific anti-mouse Ly-6G mAb or corresponding control at -24 h as well as 0, 24, and 48 h following dsRNA treatment. Lungs were harvested at 72 h and results showed that animals pretreated with neutrophil depleting mAb had a significant decrease in total neutrophil counts in both lung tissue and airways as compared to controls as reflected by FACS analysis and BAL (Fig. 4A and 4B). Importantly, BAL samples of animals pretreated with anti-mouse Ly-6G mAb antibody showed a specific reduction in neutrophil number but no reduction in monocyte numbers (data not shown). Furthermore, neutrophil depletion resulted in decreased lung microvascular permeability back to baseline values (Fig. 4C).
Figure 4 Neutrophil depletion decreases dsRNA-induced lung inflammation and injury. (A) Total lung PMNs via FACS analysis of whole-lung single-cell suspensions at 72 h. (n = 4 mice per group; *p < 0.05). (B) Total PMNs via BAL fluid cell count at 72 h (n = 4 mice per group; *p < 0.05). (C) Evans blue permeability index at 72 h (n = 6 mice per group; p < 0.05). Mice were pretreated control Ab or anti-PMN Ab at times -24, 0, 24, and 48 h following IT dsRNA treatment.
DsRNA induces elevated KC/CXCL1 and MIP-2/CXCL2/3 mRNA and protein levels
Since neutrophil influx was shown to result in lung injury in dsRNA-treated lungs, we next identified which specific factors, such as chemokines, were responsible for neutrophil recruitment. We focused on the ELR+ chemokines KC/CXCL1 and MIP-2/CXCL2/3, which are known to have neutrophil chemoattractant properties. DsRNA treatment resulted in significant increases in lung KC/CXCL1 mRNA levels as well as in protein levels from whole lung homogenates and BAL when compared to controls (Fig. 5A, C, and 5E) at 72 h following dsRNA treatment. Similarly, mRNA levels and lung protein expression of MIP-2/CXCL2/3 were also significantly elevated with an increasing trend noted in BAL protein from dsRNA-treated lungs as compared to controls (Fig. 5B, D, and 5F). Levels of CXCR2 chemokine ligands at earlier timepoints showed only a small increase (two-fold) in the induction of KC/CXCL1 and MIP-2/CXCL2/3 as early as 4 h following dsRNA-exposure (data not shown), but the maximal increase (>ten-fold for KC/CXCL1 and >four-fold for MIP-2/CXCL2/3) occurred at 72 h following dsRNA-exposure.
Figure 5 DsRNA induces elevated KC/CXCL1 and MIP-2/CXCL2/3 mRNA and protein levels. (A and B) Quantitative levels of CXCL1 and CXCL2/3 mRNA, respectively, in naïve lung and at 72 h following treatment with ssRNA or dsRNA (n = 6 mice per group; *p < 0.05). (C and D) Quantitative levels of CXCL1 and CXCL2/3 protein, respectively, in the lungs at 72 h (n = 6 mice per group; *p < 0.05). (E and F) Quantitative levels of CXCL1 and CXCL2/3 protein, respectively, in BAL fluid at 72 h (n = 6 mice per group; *p < 0.05).
DsRNA increases CXCR2 expression in lungs
CXCR2 is the shared cellular receptor for the murine CXC chemokine ligands KC/CXCL1 and MIP-2/CXCL2/3 [18-20]. The finding of increased levels of KC/CXCL1 and MIP-2/CXCL2/3 associated with dsRNA-induced neutrophil sequestration and lung injury led us to evaluate the expression of CXCR2 mRNA in the lungs of these animals. Lung homogenates from dsRNA-treated animals had significantly increased CXCR2 mRNA expression as compared to controls (Fig. 6). The expression of CXCR2 mRNA paralleled its ligand expression, neutrophil sequestration, and lung injury at 72 h following dsRNA treatment (Figs. 2, 3, and 5).
Figure 6 DsRNA increases CXCR2 expression in lungs. Quantitative real-time PCR was determined by TaqMan analysis for CXCR2 mRNA from naïve lung and at 72 h following ssRNA or dsRNA treatment (n = 6 mice per group; *p < 0.05).
Inhibition of CXCR2 inhibits infiltration of neutrophils and attenuates dsRNA-induced lung injury
To better understand the mechanism partly responsible for dsRNA-induced lung neutrophilia and injury, we determined whether inhibiting CXCR2 ligand interaction with CXCR2 significantly decreased neutrophil recruitment during the pathogenesis of dsRNA-induced lung injury. Mice were passively immunized with specific neutralizing anti-murine CXCR2 or with control antibody at -24 h, as well as 0, 24, and 48 h following dsRNA treatment. Lungs were harvested at 72 h and results showed that BAL neutrophil counts from animals that received anti-CXCR2 Ab were significantly reduced as compared to control animals that received normal goat serum (Fig. 7A). Furthermore, measurement of lung edema and lung microvascular permeability also showed significant decreases in wet:dry and Evans blue extravasation in mice treated with anti-CXCR2 compared to NGS-treated controls (Fig. 7B and 7C). Finally, histopathologic comparison of lung fields from anti-CXCR2 pretreated mice showed marked reduction in leukocytic infiltrate as compared to NGS-pretreated controls (Fig. 7D).
Figure 7 Inhibition of CXCR2 inhibits dsRNA-induced neutrophil recruitment and attenuates dsRNA-induced lung injury. (A) Total cells via BAL fluid cell count at 72 h (n = 4; *p < 0.05). (B) Wet:Dry ratio at 72 h (n = 6 mice per group; *p < 0.05). (C) Evans blue permeability index at 72 h (n = 6 mice per group; *p < 0.05). (D) Histopathology of lungs at 72 h following dsRNA treatment. Representative photomicrographs with H & E staining (100X; n = 4 mice per group). Mice were pretreated with normal goat serum (control Ab) or anti-CXCR2 Ab at times -24, 0, 24, and 48 h following IT dsRNA treatment.
Discussion
Respiratory viral infections are characterized by a two-component immune response comprised of an innate component that is fully functional before viral entry into the epithelium and an adaptive component that develops in response to the continued presence of the virus [1]. The innate or acute inflammatory component is associated with a predominance of infiltrating neutrophils. While many viral infections of the lung are self-limiting, the associated lung injury due to this initial event may be critical in establishing a pro-inflammatory environment underlying certain chronic respiratory disorders such as asthma, COPD, cystic fibrosis, pulmonary fibrosis, viral pneumonia, and bronchopulmonary dysplasia [1,3-10]. The host's inflammatory response to viral replication leads to pulmonary pathology hallmarked by leukocyte infiltrate and resultant microvascular leak that progresses to lung edema and the clinical signs of pneumonia. As the lung injury persists or recurs with multiple subsequent viral infections, pulmonary vascular and airway remodeling may occur and eventually lead to development of airway hyper-reactivity and/or interstitial fibrosis [44]. Viral infections are mediated by dsRNA, a proinflammatory molecule generated during viral replication. DsRNA binds to its cell surface receptor, TLR3, and activates the production of downstream gene products, such as CXC chemokines. In this study, we hypothesized that the interaction between CXCR2 and ELR-positive CXC chemokines expressed during dsRNA-induced lung inflammation is critical in mediating neutrophil recruitment, a pivotal process required for dsRNA induced lung injury in viral infections.
Previous studies have demonstrated that mice exposed to a live virus via intranasal inoculation generate a systemic acute-phase response with maximal pulmonary chemokine response at one week that includes both CC and CXC chemokines and is mouse strain-dependent (response in BALB/c greater than in C57BL/6 mice) [45,46]. A similar study using intratracheal delivery of live virus also demonstrated marked pulmonary pathology including mucous cell metaplasia and airway epithelial remodeling [47]. Another study focusing specifically on the effects of intratracheal dsRNA at a low dose found similar systemic inflammatory effects but specific pulmonary effects only when dsRNA was delivered in conjunction with IFNγ [48]. Finally, a recent study examined the effects of inhibiting the CC chemokine receptor, CCR1, during live-virus exposure in mice and showed that mortality during pneumovirus infection was decreased [49]. The present study extends these findings by first determining the effects of high-dose intratracheal dsRNA alone on neutrophil recruitment and lung injury in BALB/c mice and then subsequently blocking these effects via inhibition of the CXC chemokine receptor, CXCR2.
To determine the effects of dsRNA in vivo, we first characterized our murine model by performing a time-course of dsRNA to observe histopathologic changes at 0, 4, 12, 24, 48 and 72 h following intratracheal dsRNA delivery. We used a maximal dose of dsRNA at 40 μg/g mouse wt after initial studies at lower doses (4 μg/g and 20 μg/g) showed minimal leukocytic infiltrate. Furthermore, we are aware of only one other publication that describes intratracheal dsRNA delivery that showed no effects when used alone at low concentration (10 ug/g) [48]. Histopathological analysis demonstrated a significantly increased leukocytic infiltrate at 72 h following intratracheal dsRNA as compared to earlier time points. We thus chose this observation as the basis to focus upon the 72 h timepoint. Other studies using administration of dsRNA in BALB/c mice in vivo have also shown a maximal innate immune response starting at 72 h following dsRNA exposure [50] Further characterization at 72 h following intratracheal dsRNA showed that neutrophils were a predominant cell type and that there was an associated injury to alveolar-capillary membrane integrity as shown by increased lung edema and microvascular leak.
Having characterized the histopathological damage caused by intratracheal dsRNA, we then focused on the underlying mechanisms responsible for promoting the inflammation and subsequent lung injury. Our findings of a significant increase in neutrophil infiltration following dsRNA treatment are consistent with findings from previous studies using a live virus that also resulted in early neutrophil infiltration [47]. In order to determine whether neutrophil influx was causally linked to the lung injury observed in our system, we performed studies using a monoclonal antibody to the Ly6G antigen on the surface of mouse granulocytes to specifically deplete neutrophils. These results showed decreased numbers of lung neutrophils by FACS analysis and BAL as expected with no reduction in monocyte number, and also showed a decrease in lung microvascular leak and therefore decreased lung injury associated with decreased neutrophil recruitment. However, the molecular and cellular mechanisms involved in recruiting these neutrophils remained to be fully elucidated.
Elegant in vitro studies have demonstrated that dsRNA can induce IL-8 expression from human bronchial epithelial cells [16]. Furthermore, in vivo studies using live virus have also demonstrated that CXC chemokines are generated during lung inflammation [45-47]. Having demonstrated that lung injury due to intratracheal dsRNA is dependent on neutrophil infiltration, we determined that CXCL1 and CXCL2/3 expression was significantly greater in the lungs of dsRNA-treated mice than in lungs of ssRNA-treated controls. Furthermore, expression of CXCR2 mRNA was similarly increased in parallel to the production of both CXCL1 and CXCL2/3 ligands and neutrophil sequestration during dsRNA-induced lung inflammation. Other studies of inflammatory diseases, such as ventilator-induced lung injury, pneumonia, and hyperoxia-induced lung injury have demonstrated the importance of CXCR2 expression and its role in neutrophil recruitment during the pathogenesis of these diseases [24,40,41,51]. Collectively, these studies demonstrate that augmented levels of CXCR2 ligands are important in the recruitment of neutrophils during the pathogenesis of inflammatory diseases and suggest that the interaction between CXCR2 ligands and CXCR2 may be pivotal in the recruitment of neutrophils to the lung during dsRNA-induced lung injury.
Based on these findings, we performed proof of principle studies in vivo using an antibody-mediated neutralization strategy to evaluate the direct role for CXCL1 and CXCL2/3 ligands and their interaction with CXCR2 during the pathogenesis of dsRNA-induced lung injury. The anti-CXCR2 Ab-treated mice demonstrated significant reductions specific toneutrophil infiltration that were paralleled by a decrease in lung injury. This suggests that if other leukocytes are recruited there are redundant pathways involved in their recruitment that are CXCR2-independent. Our findings are similar to previous neutralization studies in virus-infected mice that showed functional antagonism to the chemokine receptor CCR1 reduced mortality [49]. Our study is the first to demonstrate that neutralization of CXCR2 results in decreased neutrophil recruitment and lung injury induced by intratracheal dsRNA.
TLR3 is a pattern-recognition cell surface receptor that is responsible for specifically recognizing extracellular dsRNA [52]. It is a member of a family of toll-like receptors (TLRs) that aid the host in combating infections when microbial pathogens bind to their specific toll-like receptor to trigger NFκB and downstream generation of cytokines and costimulatory molecules during the innate immune attack. TLR3 thus plays an important role in host defense against viral infections. Our study showed that dsRNA treatment induced the expression of TLR3 on the cell surface of airways as compared to ssRNA-treated controls animals. The upregulation of TLR3 in dsRNA-exposed mouse airways suggests a positive feedback system in which epithelial cells exposed to dsRNA increase TLR3 expression on their cell surface. A similar upregulation of TLR3 has previously been described in epithelial cells in vitro treated with Respiratory syncytial virus (RSV)[53]. These findings are also consistent with in vivo studies using monoclonal antibody to TLR3 and genetic knockout mice of TLR3 [52,54] that establish the role of TLR3 in dsRNA recognition and raise the interesting possibility that TLR3 may be involved in the signaling pathway mediating dsRNA induction of CXC chemokines.
Conclusion
In conclusion, we have demonstrated that the biological axis of CXCR2 ligand/CXCR2 signaling plays a pivotal role in mediating neutrophil recruitment and dsRNA-induced lung injury. This mechanism may be critical for the promotion of further lung injury and remodeling in patients who eventually develop chronic lung diseases such as asthma, COPD, cystic fibrosis, pulmonary fibrosis or bronchopulmonary dysplasia. The findings of the current study support the contention that CXCR2 ligands and CXCR2 mediate dsRNA-induced lung injury, and may be a therapeutic target to attenuate this pathology.
Competing interests
The author(s) declare that they have no competing interests.
Authors' Contributions
VL designed and carried out all experiments and drafted the manuscript; JB and MK contributed to analysis and interpretation of data; MB and YX contributed to performing animal experiments; and RS provided final analysis and interpretation of data.
Acknowledgements
This work was supported in part by the National Institutes of Health (NIH), grants HL66027 and P50HL67665 (to R.S.), National Research Service Award (NRSA) Institutional Training Grant, T32 HD07549 (to V.L.) and the NIH-sponsored UCLA Child Health Research Career Development K12 Award (CHRCDA) (to V.L.).
==== Refs
Hogg JC Role of latent viral infections in chronic obstructive pulmonary disease and asthma Am J Respir Crit Care Med 2001 164 S71 5 11734471
Cheung OY Chan JW Ng CK Koo CK The spectrum of pathological changes in severe acute respiratory syndrome (SARS) Histopathology 2004 45 119 124 15279629 10.1111/j.1365-2559.2004.01926.x
Gern JE Busse WW The role of viral infections in the natural history of asthma J Allergy Clin Immunol 2000 106 201 212 10932062 10.1067/mai.2000.108604
Gern JE Viral respiratory infection and the link to asthma Pediatr Infect Dis J 2004 23 S78 86 14730274
Hogg JC Chu F Utokaparch S Woods R Elliott WM Buzatu L Cherniack RM Rogers RM Sciurba FC Coxson HO Pare PD The nature of small-airway obstruction in chronic obstructive pulmonary disease N Engl J Med 2004 350 2645 2653 15215480 10.1056/NEJMoa032158
Hogg JC The adenovirus and bronchopulmonary dysplasia: an association that could be causal or coincidental Pediatr Res 2000 47 175 176 10674341
Couroucli XI Welty SE Ramsay PL Wearden ME Fuentes-Garcia FJ Ni J Jacobs TN Towbin JA Bowles NE Detection of microorganisms in the tracheal aspirates of preterm infants by polymerase chain reaction: association of adenovirus infection with bronchopulmonary dysplasia Pediatr Res 2000 47 225 232 10674351
Wat D Doull I Respiratory virus infections in cystic fibrosis Paediatr Respir Rev 2003 4 172 177 12880751 10.1016/S1526-0542(03)00059-9
Lok SS Egan JJ Viruses and idiopathic pulmonary fibrosis Monaldi Arch Chest Dis 2000 55 146 150 10949877
Tang YW Johnson JE Browning PJ Cruz-Gervis RA Davis A Graham BS Brigham KL Oates JAJ Loyd JE Stecenko AA Herpesvirus DNA is consistently detected in lungs of patients with idiopathic pulmonary fibrosis J Clin Microbiol 2003 41 2633 2640 12791891 10.1128/JCM.41.6.2633-2640.2003
Subauste MC Jacoby DB Richards SM Proud D Infection of a human respiratory epithelial cell line with rhinovirus. Induction of cytokine release and modulation of susceptibility to infection by cytokine exposure J Clin Invest 1995 96 549 557 7615827
Schroth MK Grimm E Frindt P Galagan DM Konno SI Love R Gern JE Rhinovirus replication causes RANTES production in primary bronchial epithelial cells Am J Respir Cell Mol Biol 1999 20 1220 1228 10340941
Majde JA Viral double-stranded RNA, cytokines, and the flu J Interferon Cytokine Res 2000 20 259 272 10762073 10.1089/107999000312397
Jacobs BL Langland JO When two strands are better than one: the mediators and modulators of the cellular responses to double-stranded RNA Virology 1996 219 339 349 8638399 10.1006/viro.1996.0259
Gern JE Vrtis R Grindle KA Swenson C Busse WW Relationship of upper and lower airway cytokines to outcome of experimental rhinovirus infection Am J Respir Crit Care Med 2000 162 2226 2231 11112143
Gern JE French DA Grindle KA Brockman-Schneider RA Konno SI Busse WW Double-stranded RNA induces the synthesis of specific chemokines by bronchial epithelial cells Am J Respir Cell Mol Biol 2002
Guha-Thakurta N Majde JA Early induction of proinflammatory cytokine and type I interferon mRNAs following Newcastle disease virus, poly [rI:rC], or low-dose LPS challenge of the mouse J Interferon Cytokine Res 1997 17 197 204 9142648
Bozic CR Gerard NP von Uexkull-Guldenband C Kolakowski LFJ Conklyn MJ Breslow R Showell HJ Gerard C The murine interleukin 8 type B receptor homologue and its ligands. Expression and biological characterization J Biol Chem 1994 269 29355 29358 7961909
Lee J Cacalano G Camerato T Toy K Moore MW Wood WI Chemokine binding and activities mediated by the mouse IL-8 receptor J Immunol 1995 155 2158 2164 7636264
Cerretti DP Nelson N Kozlosky CJ Morrissey PJ Copeland NG Gilbert DJ Jenkins NA Dosik JK Mock BA The murine homologue of the human interleukin-8 receptor type B maps near the Ity-Lsh-Bcg disease resistance locus Genomics 1993 18 410 413 8288247 10.1006/geno.1993.1486
Harada A Kuno K Nomura H Mukaida N Murakami S Matsushima K Cloning of a cDNA encoding a mouse homolog of the interleukin-8 receptor Gene 1994 142 297 300 8194768 10.1016/0378-1119(94)90278-X
Barnes PJ Potential novel therapies for chronic obstructive pulmonary disease Novartis Found Symp 2001 234 255 67; discussion 267-72 11199100
Gordon JR Swystun VA Li F Zhang X Davis BE Hull P Cockcroft DW Regular salbutamol use increases CXCL8 responses in asthma: relationship to the eosinophil response Eur Respir J 2003 22 118 126 12882461 10.1183/09031936.03.00031102
Mehrad B Strieter RM Moore TA Tsai WC Lira SA Standiford TJ CXC chemokine receptor-2 ligands are necessary components of neutrophil-mediated host defense in invasive pulmonary aspergillosis J Immunol 1999 163 6086 6094 10570298
Moore TA Newstead MW Strieter RM Mehrad B Beaman BL Standiford TJ Bacterial clearance and survival are dependent on CXC chemokine receptor-2 ligands in a murine model of pulmonary Nocardia asteroides infection J Immunol 2000 164 908 915 10623839
Tsai WC Strieter RM Mehrad B Newstead MW Zeng X Standiford TJ CXC chemokine receptor CXCR2 is essential for protective innate host response in murine Pseudomonas aeruginosa pneumonia Infect Immun 2000 68 4289 4296 10858247 10.1128/IAI.68.7.4289-4296.2000
Greenberger MJ Strieter RM Kunkel SL Danforth JM Laichalk LL McGillicuddy DC Standiford TJ Neutralization of macrophage inflammatory protein-2 attenuates neutrophil recruitment and bacterial clearance in murine Klebsiella pneumonia J Infect Dis 1996 173 159 165 8537653
Huffnagle GB Strieter RM Standiford TJ McDonald RA Burdick MD Kunkel SL Toews GB The role of monocyte chemotactic protein-1 (MCP-1) in the recruitment of monocytes and CD4+ T cells during a pulmonary Cryptococcus neoformans infection J Immunol 1995 155 4790 4797 7594481
Huffnagle GB Strieter RM McNeil LK McDonald RA Burdick MD Kunkel SL Toews GB Macrophage inflammatory protein-1alpha (MIP-1alpha) is required for the efferent phase of pulmonary cell-mediated immunity to a Cryptococcus neoformans infection J Immunol 1997 159 318 327 9200469
Keane MP Belperio JA Moore TA Moore BB Arenberg DA Smith RE Burdick MD Kunkel SL Strieter RM Neutralization of the CXC chemokine, macrophage inflammatory protein-2, attenuates bleomycin-induced pulmonary fibrosis J Immunol 1999 162 5511 5518 10228032
Belperio JA Burdick MD Keane MP Xue YY Lynch JP Daugherty BL Kunkel SL Strieter RM The role of the CC chemokine, RANTES, in acute lung allograft rejection J Immunol 2000 165 461 472 10861085
Olszyna DP Florquin S Sewnath M Branger J Speelman P van Deventer SJ Strieter RM van der Poll T CXC chemokine receptor 2 contributes to host defense in murine urinary tract infection J Infect Dis 2001 184 301 307 11443555 10.1086/322030
Belperio JA Dy M Burdick MD Xue YY Li K Elias JA Keane MP Interaction of IL-13 and C10 in the pathogenesis of bleomycin-induced pulmonary fibrosis Am J Respir Cell Mol Biol 2002 27 419 427 12356575
Huang M Sharma S Zhu LX Keane MP Luo J Zhang L Burdick MD Lin YQ Dohadwala M Gardner B Batra RK Strieter RM Dubinett SM IL-7 inhibits fibroblast TGF-beta production and signaling in pulmonary fibrosis J Clin Invest 2002 109 931 937 11927620 10.1172/JCI200214685
Keane MP Belperio JA Burdick MD Strieter RM IL-12 attenuates bleomycin-induced pulmonary fibrosis Am J Physiol Lung Cell Mol Physiol 2001 281 L92 7 11404251
Keane MP Belperio JA Arenberg DA Burdick MD Xu ZJ Xue YY Strieter RM IFN-gamma-inducible protein-10 attenuates bleomycin-induced pulmonary fibrosis via inhibition of angiogenesis J Immunol 1999 163 5686 5692 10553099
Garcia-Ramallo E Marques T Prats N Beleta J Kunkel SL Godessart N Resident cell chemokine expression serves as the major mechanism for leukocyte recruitment during local inflammation J Immunol 2002 169 6467 6473 12444156
Laichalk LL Kunkel SL Strieter RM Danforth JM Bailie MB Standiford TJ Tumor necrosis factor mediates lung antibacterial host defense in murine Klebsiella pneumonia Infect Immun 1996 64 5211 5218 8945568
Arenberg DA Kunkel SL Polverini PJ Glass M Burdick MD Strieter RM Inhibition of interleukin-8 reduces tumorigenesis of human non-small cell lung cancer in SCID mice J Clin Invest 1996 97 2792 2802 8675690
Belperio JA Keane MP Burdick MD Londhe V Xue YY Li K Phillips RJ Strieter RM Critical role for CXCR2 and CXCR2 ligands during the pathogenesis of ventilator-induced lung injury J Clin Invest 2002 110 1703 1716 12464676 10.1172/JCI200215849
Sue RD Belperio JA Burdick MD Murray LA Xue YY Dy MC Kwon JJ Keane MP Strieter RM CXCR2 is critical to hyperoxia-induced lung injury J Immunol 2004 172 3860 3868 15004193
Green TP Johnson DE Marchessault RP Gatto CW Transvascular flux and tissue accrual of Evans blue: effects of endotoxin and histamine J Lab Clin Med 1988 111 173 183 3276800
Saria A Lundberg JM Evans blue fluorescence: quantitative and morphological evaluation of vascular permeability in animal tissues J Neurosci Methods 1983 8 41 49 6876872 10.1016/0165-0270(83)90050-X
Vassallo R Viral-induced inflammation in interstitial lung diseases Semin Respir Infect 2003 18 55 60 12652455 10.1053/srin.2003.50008
Weinberg JB Lutzke ML Alfinito R Rochford R Mouse strain differences in the chemokine response to acute lung infection with a murine gammaherpesvirus Viral Immunol 2004 17 69 77 15018663 10.1089/088282404322875467
Weinberg JB Lutzke ML Efstathiou S Kunkel SL Rochford R Elevated chemokine responses are maintained in lungs after clearance of viral infection J Virol 2002 76 10518 10523 12239330 10.1128/JVI.76.20.10518-10523.2002
Kajon AE Gigliotti AP Harrod KS Acute inflammatory response and remodeling of airway epithelium after subspecies B1 human adenovirus infection of the mouse lower respiratory tract J Med Virol 2003 71 233 244 12938198 10.1002/jmv.10475
Traynor TR Majde JA Bohnet SG Krueger JM Intratracheal double-stranded RNA plus interferon-gamma: a model for analysis of the acute phase response to respiratory viral infections Life Sci 2004 74 2563 2576 15010266 10.1016/j.lfs.2003.10.010
Bonville CA Lau VK DeLeon JM Gao JL Easton AJ Rosenberg HF Domachowske JB Functional antagonism of chemokine receptor CCR1 reduces mortality in acute pneumovirus infection in vivo J Virol 2004 78 7984 7989 15254170 10.1128/JVI.78.15.7984-7989.2004
Zare F Bokarewa M Nenonen N Bergstrom T Alexopoulou L Flavell RA Tarkowski A Arthritogenic properties of double-stranded (viral) RNA J Immunol 2004 172 5656 5663 15100310
Auten RL Richardson RM White JR Mason SN Vozzelli MA Whorton MH Nonpeptide CXCR2 antagonist prevents neutrophil accumulation in hyperoxia-exposed newborn rats J Pharmacol Exp Ther 2001 299 90 95 11561067
Alexopoulou L Holt AC Medzhitov R Flavell RA Recognition of double-stranded RNA and activation of NF-kappaB by Toll-like receptor 3 Nature 2001 413 732 738 11607032 10.1038/35099560
Rudd BD Burstein E Duckett CS Li X Lukacs NW Differential Role for TLR3 in Respiratory Syncytial Virus-Induced Chemokine Expression J Virol 2005 79 3350 3357 15731229 10.1128/JVI.79.6.3350-3357.2005
Matsumoto M Kikkawa S Kohase M Miyake K Seya T Establishment of a monoclonal antibody against human Toll-like receptor 3 that blocks double-stranded RNA-mediated signaling Biochem Biophys Res Commun 2002 293 1364 1369 12054664 10.1016/S0006-291X(02)00380-7
| 15921526 | PMC1156932 | CC BY | 2021-01-04 16:36:23 | no | J Inflamm (Lond). 2005 May 28; 2:4 | utf-8 | J Inflamm (Lond) | 2,005 | 10.1186/1476-9255-2-4 | oa_comm |
==== Front
J Neuroengineering RehabilJournal of NeuroEngineering and Rehabilitation1743-0003BioMed Central London 1743-0003-2-111592705610.1186/1743-0003-2-11ResearchAnalysis of right anterolateral impacts: the effect of head rotation on the cervical muscle whiplash response Kumar Shrawan [email protected] Robert [email protected] Yogesh [email protected] Physical Therapy, University of Alberta, 3–75 Corbett Hall, Edmonton, Alberta T6G 2G4, Canada2 Department of Medicine, University of Alberta, Edmonton, Alberta T6G 2B7, Canada3 Physical Therapy, University of Alberta, 3–78 Corbett Hall, Edmonton, Alberta T6G 2G4, Canada2005 31 5 2005 2 11 11 26 11 2004 31 5 2005 Copyright © 2005 Kumar et al; licensee BioMed Central Ltd.2005Kumar et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The cervical muscles are considered a potential site of whiplash injury, and there are many impact scenarios for whiplash injury. There is a need to understand the cervical muscle response under non-conventional whiplash impact scenarios, including variable head position and impact direction.
Methods
Twenty healthy volunteers underwent right anterolateral impacts of 4.0, 7.6, 10.7, and 13.0 m/s2 peak acceleration, each with the head rotated to the left, then the head rotated to the right in a random order of impact severities. Bilateral electromyograms of the sternocleidomastoids, trapezii, and splenii capitis following impact were measured.
Results
At a peak acceleration of 13.0 m/s2, with the head rotated to the right, the right trapezius generated 61% of its maximal voluntary contraction electromyogram (MVC EMG), while all other muscles generated 31% or less of this variable (31% for the left trapezius, 13% for the right spleinus. capitis, and 16% for the left splenius capitis). The sternocleidomastoids muscles also tended to show an asymmetric EMG response, with the left sternocleidomastoid (the one responsible for head rotation to the right) generating a higher percentage (26%) of its MVC EMG than the left sternocleidomastoid (4%) (p < 0.05). When the head is rotated to the left, under these same conditions, the results are reversed even though the impact direction remains right anterolateral.
Conclusion
The EMG response to a right anterolateral impact is highly dependent on the head position. The sternocleidomastoid responsible for the direction of head rotation and the trapezius ipsilateral to the direction of head rotation generate the most EMG activity.
Cervical musclesElectromyographyAccelerationAnterolateral impactsWhiplash
==== Body
Background
Although many diagnostic efforts over the decades have aimed at objectively identifying the acute whiplash injury that is often labelled as "soft tissue injury" or "neck sprain", with the exception of a few case reports and excluding spinal cord or bony injury, the pathology of the acute whiplash injury remains elusive [1]. In the absence of an identifiable injury, efforts have simultaneously focused on development of better preventative measures and treatment approaches. Even without knowing what the acute whiplash injury is, for example, knowing more of the human response to whiplash type impacts led to the introduction of head restraints in 1969[2] and further innovations of head restraints have followed as the knowledge has increased [3]. Most efforts to understand the whiplash injury mechanism have focused on rear impacts [4-11]. Although it has been traditionally reported that rear-impacts account for most cases of whiplash injury, epidemiological evidence suggests that rear, lateral, and frontal collisions account for whiplash injury in roughly equal proportions [12].
Frontal collisions thus require more investigative attention, and yet there are a number of variables to consider in terms of understanding how the cervical muscles respond to a whiplash-type frontal impact. First, not all collision victims have their head in the neutral (facing forward) position. We recently reported on the effect of head rotation in straight-on frontal impacts [13], and compared this to the head in neutral position in a frontal impact [14]. With the head in neutral position, a frontal impact causes the greatest EMG activity to be generated symmetrically in the trapezii, which have an EMG activity that is 30–50% of their maximal voluntary contraction (MVC EMG). In a frontal impact with head rotated to the left, however, the left trapezius generated 77% of its maximal voluntary contraction (MVC) EMG (more than double the response of other muscles). In comparison, the right trapezius generated only 33% of its MVC. The right sternocleidomastoid (25%) and left splenius muscles (32%), the ones responsible for head rotation to the left, were more active than their counterparts. On the other hand, with the head rotated to the right, the right trapezius generated 71% of its MVC EMG, while the left trapezius generated only 30% of this value. Again, the left sternocleidomastoid (27% of its MVC EMG) and right splenius (28% of its MVC EMG), being responsible for head rotation to the right, were more active than their counterparts. Thus, head rotation produces an asymmetric EMG response.
Then there is the direction of impact. Frontal impacts are not always straight-on impacts. We have considered the example of a right anterolateral impact [15], and the results confirm the importance of direction of impact on the cervical muscle response. When the impact is a right anterolateral impact, the left trapezius still generated the greatest EMG, up to 83% of the maximal voluntary contraction EMG, and the left splenius capitis instead became more active and reached a level of 46% of this variable [15]. This is greater than the response of the splenius capitis in straight-on frontal impacts. Thus, direction of impact also determines which muscles respond and the proportionality of the response among the different muscle groups.
The question is whether head rotation in anterolateral impacts will increase or decrease the EMG activity, and how. We thus undertook a study to assess the cervical muscle response in right anterolateral impacts, but with the head rotated to either the left or right at the time of impact. This is part of a series of experiments to approach the more complex impact scenarios of varying directions and head positions.
Materials and methods
Sample
The methods for this study of offset frontal impacts are the same as that used previously for our previous right anterolateral and frontal impact studies [13-15]. Twenty healthy normal subjects (10 males, 10 females, all right-hand dominant) with no history of whiplash injury and no cervical spine pain during the preceding 12 months volunteered for the study. The study was approved by the University Research Ethics Board. The twenty subjects had a mean age of 23.6 ± 3.0 years, a mean height of 172 ± 7.7 cm, and a mean weight of 69 ± 13.9 kg.
Tasks and Data Collection
Active surface electrodes with 10 times on-site amplification were placed on the belly of the sternocleidomastoids, upper trapezius at C4 level, and splenius capitis in the triangle between sternocleidomastoids and trapezii bilaterally. The fully-isolated amplifier had additional gain settings up to 10, 000 times with frequency response DC-5 kHz and common mode rejection ratio of 92 dB. Before calibrating sled acceleration, the cervical strength of the volunteers was measured to develop force-EMG calibration factor [16,17]. The seated and stabilized subjects exerted their maximum isometric effort in attempted flexion, extension, and lateral flexion to the left and the right for force-EMG calibration, as described by Kumar et al.[16,17]. The acceleration device consisted of an acceleration platform and a sled. The full details of the device and the electromyography data collection are given by Kumar et al.[7] and the device is as shown in Fig. 1. After the experiment was discussed and informed consent obtained, the age, weight, and height of each volunteer was recorded. The volunteers then were seated on the chair and stabilized in neutral spinal posture. The chair was rigid so as to minimize any effect of elastic properties of the chair following acceleration. Subjects were then outfitted with triaxial accelerometers (Model # CXL04M3, Crossbow technology, Inc., San Jose, California, U. S. A.) on their glabella and the first thoracic spinous process. Another triaxial accelerometer was mounted on the sled, not the chair. The accelerometers had a full scale nonlinearity of 0.2%, dynamic range of ± 5 g, with a sensitivity of 500 mV/g, resolution of 5 mg within a bandwidth of DC-100 Hz, and a silicon micromachined capacitive beam that was quite rugged and extremely small in die area. The subjects were then exposed to right anterolateral impacts (offset from a frontal impact by 45 degrees) with their head rotated 45 degrees to their left and right at accelerations of 4.0, 7.6, 10.7, and 13.0 m/s2 generated in a random order by a pneumatic piston. To release the piston the solenoid of the pneumatic system was activated by an electronic impulse which was recorded for timing reference. Upon delivery of impact by the pneumatic piston, the sled moved on two parallel tracks mounted 60 cm. apart. The coefficient of friction of the tracks was 0.03 which allowed for smooth gliding of the sled on the rails. The opposite end of the track was equipped with non-linear springs and high density rubber stopper to prevent the subject from sliding off the platform. Each subject effectively underwent 4 levels of accelerative impacts under two conditions of head rotation, for one direction of impact (a total of 8 impacts). The head rotation itself did not place the head in a more forward position. Although the subjects are asked to rotate their head prior to impact, nothing was done to fix the position, and the head is free to move after impact. The accelerations involved in this experiment were low enough that injury was not expected. The acceleration was delivered in a way that mimicked the time course seen in motor vehicle collisions and occurred fast enough to produce eccentric muscle contractions. The acceleration impulse reached its peak value in 33 ms. Subjects were asked to report any headache or other aches they experienced in the days following the impacts.
Figure 1 Illustration of the sled device for whiplash-type impacts.
Data analysis
The data on the peak and average accelerations in all three axes of the sled, shoulder, and head for all four levels of accelerative impacts were measured. The gravity bias was eliminated by subtracting this value from the accelerometer readings. The onset of acceleration was measured by dropping the ascending slope line on the base line. The point of intersection of these lines was considered as onset of acceleration. In the analysis, the sample of volunteers was collapsed across gender because preliminary analysis showed no statistically significant differences in the EMG amplitudes between the men and women. The sled velocity and its acceleration subsequent to the pneumatic piston impact and the rubber stopper impact were measured. All timing data (time to onset of EMG and peak EMG) were referred to the solenoid of the piston firing. The time of the peak accelerations of sled and head were measured. Also, the time relations of the onset and peak of the EMG were measured and analyzed. The time to onset was determined when the EMG perturbation reached 2% of the peak EMG value to avoid false positives due to tonic activity. This method was chosen to avoid any false positives due to tonic EMG. This method was in agreement with projection of the line of slope on the baseline. EMG amplitudes were normalised against the subjects' maximal voluntary contraction electromyogram. The ratio percentage of the EMG amplitude versus the maximal contraction normalised EMG activity for that subject allowed us to determine the force equivalent generated due to the impact for each muscle.
Statistical analysis was performed using the SPSS statistical package (SPSS Inc., Chicago, IL) to calculate descriptive statistics, correlation analysis between EMG and head acceleration, analysis of variance (ANOVA) of the time to EMG onset, time to peak EMG, average EMG, and the force equivalents. Additionally, a linear regression analysis was carried out for the kinematic variables of head displacement, head velocity and head acceleration and EMG variables on the peak of the sled acceleration. Initially, all regressions were carried out to the level of exposure and subsequently they were extrapolated to twice the level of acceleration used in the study. The purpose of the regression analysis was to see if using the acceleration of the sled – one could predict the head acceleration and EMG response. The regression analysis was carried out using linear and non-linear functions. The linear regression was found to be the best fit, perhaps because the input acceleration impulse was non-linear.
Results
Head acceleration
The kinematic response of the head to the four levels of applied acceleration are shown in Fig. 2. As anticipated, an increase in applied acceleration resulted in an increase in excursion of the head and accompanying accelerations (p < 0.05). The accelerations in these impacts were not associated with any reported symptoms in the volunteers.
Figure 2 Head acceleration in the x, y, and z axes of one subject in response to the level of applied acceleration. The z-axis is parallel, the x-axis orthogonal, and the y-axis vertical to the direction of travel. Head X, head acceleration in the x-axis; Head Y, head acceleration in the y-axis; Head Z, head acceleration in the z-axis.
Electromyogram amplitude
In a right anterolateral impact, with the head rotated 45 degrees to the right or left, the trapezius muscle ipsilateral to the direction of head rotation showed the greatest EMG response (p < 0.05). The sternocleidomastoid muscles responsible for the head rotation each showed more EMG response to the pertubation than their counterparts (p < 0.05).
At a peak acceleration of 13.0 m/s2, for example with the head rotated to the right, the right trapezius generated 61% of its maximal voluntary contraction electromyogram, while all other muscles generated 31% or less of this variable. Though they generated less EMG activity, the sternocleidomastoids muscles also tended to show an asymmetric EMG response, with the left sternocleidomastoid (the one responsible for head rotation to the right) generating a higher percentage (26%) of its maximal voluntary contraction electromyogram than the right sternocleidomastoid (4%) (p < 0.05). When the head is rotated to the left, under these same conditions, the EMG results are reversed even though the impact direction remains right anterolateral. When looking left, the left trapezius generated 51% of its maximal voluntary contraction electromyogram, with only 14% of the maximal voluntary contraction for the right trapezius, and less than 25% for the remaining muscles. The sternocleidomastoid muscles in this case still showed an asymmetric EMG response, with the right sternocleidomastoid (the one responsible for head rotation to the left) generating a higher percentage (22%) of its maximal voluntary contraction electromyogram than the left sternocleidomastoid (4%) (p < 0.05).
The normalized EMG for the sternocleidomastoid (SCM), splenius capitis (SPL) and trapezius (TRP) muscles are shown in Fig. 3. As the level of applied acceleration in the impact increased, the magnitude of the EMG recorded from the trapezius ipsilateral to the head rotation increased progressively and disproportionately compared to other muscles (p < 0.05). The reverse occurred when the head was rotated to the left, where the left TRP instead generated 77% of its MVC and again the remaining muscles generated 33% or less of their MVC. Figure 4 also compares these responses at the highest level of acceleration to the cervical muscle responses with the head in neutral position. The results indicate that head rotation affected the muscle response independent of direction of impact. Although the data concerning EMG responses with the head in neutral posture are from a different group of subjects, the methodology of always normalizing the EMG response to an individual's maximal voluntary contraction helps to adjust for these variables (i.e, gender, stature and age affects maximal voluntary contraction, and EMG responses should thus be normalized before making comparisons among individuals or groups). Thus, we were able to compare normalized populations from different studies, each group undergoing the same experimental protocols are used.
Figure 3 Normalized average and peak electromyogram (EMG) (percentage of isometric maximal voluntary contraction), force equivalent of EMG (N), and head rotated right or left, and applied acceleration. LSCM, left sternocleidomastoid; RSCM, right sternocleidomastoid; LSPL, left splenius capitis; RSPL, right splenius capitis; LTRP, left trapezius; RTRP, right trapezius.
Figure 4 Normalized peak electromyogram (EMG) (percentage of isometric maximal voluntary contraction), for head in neutral position, rotated right, or rotated left, at an applied acceleration of 13.0 m/s2. LSCM, left sternocleidomastoid; RSCM, right sternocleidomastoid; LSPL, left splenius capitis; RSPL, right splenius capitis; LTRP, left trapezius; RTRP, right trapezius.
Timing
The time to onset of the sled, shoulder, and head acceleration onset in the z-axis (axis along impact direction) and the EMG signals of the six muscles examined for head rotated to the left or right are presented in Table 1. The timing data is in relation to firing of the solenoid of the piston. The time to onset of the sled, torso, and head acceleration decreased with increased applied acceleration (p < 0.05). Similarly, the time to onset of the EMG show a trend (p > 0.05) for all muscles to decrease with increased applied acceleration. The mean times at which peak EMG occurred for all the experimental conditions are presented in Table 2, and also show a trend to earlier times of peak activity with increasing acceleration, though this again did not reach statistical significance.
Table 1 Mean Time to Onset (msec) of Acceleration and of Muscle EMG From the Firing of the Solenoid of the Pneumatic Piston
Muscle
Sternocleidomastoid Splenius Capitis Trapezius
Acceleration (m/s2) Sled Shoulder Head Left Right Left Right Left Right
Right Head Rotation
4.0 44 (19) 65 (32) 85 (17) 199 (116) 224 (136) 125 (45) 104 (52) 105 (44) 108 (48)
7.6 34 (10) 52 (18) 61 (21) 177 (81) 197 (143) 109 (33) 97 (40) 104 (42) 96 (46)
10.7 30 (11) 42 (14) 55 (21) 170 (49) 141 (109) 104 (42) 96 (47) 97 (33) 92 (41)
13.0 26 (11) 35 (15) 52 (21) 132 (60) 125 (63) 91 (22) 93 (36) 89 (30) 90 (28)
Left Head Rotation
4.0 48 (21) 64 (26) 97 (22) 185 (61) 222 (50) 114 (43) 196 (105) 137 (35) 180 (55)
7.6 31 (15) 49 (22) 71 (25) 99 (45) 194 (45) 98 (37) 172 (78) 106 (45) 114 (44)
10.7 29 (14) 43 (12) 65 (22) 86 (47) 181 (77) 94 (35) 163 (107) 98 (41) 110 (48)
13.0 27 (11) 42 (19) 64 (19) 79 (48) 180 (70) 85 (27) 138 (48) 78 (29) 101 (36)
Times for the sled, shoulder, and head represent the time at which acceleration in z-axis (direction of travel) began. Times for the cervical muscles represent the time to onset for EMG activity. Values in parentheses represent one standard deviation.
Table 2 Mean Time (msec) at Which Peak Electromyogram Occurred After the Firing of the Solenoid of the Pneumatic Piston
Muscle EMG
Sternocleidomastoid Splenius Capitis Trapezius
Acceleration (m/s2) Left Right Left Right Left Right
Right Head Rotation
4.0 479 (298) 599 (374) 247 (46) 264 (374) 223 (20) 228 (28)
7.6 379 (281) 569 (263) 225 (36) 224 (32) 211 (28) 227 (24)
10.7 363 (212) 547 (414) 219 (36) 219 (30) 206 (31) 224 (35)
13.0 321 (225) 521 (349) 210 (35) 211 (23) 196 (30) 210 (26)
Left Head Rotation
4.0 526 (342) 687 (433) 243 (34) 822 (511) 281 (90) 664 (255)
7.6 255 (72) 576 (141) 227 (19) 704 (365) 267 (42) 262 (60)
10.7 245 (34) 521 (240) 223 (27) 631 (225) 256 (57) 246 (58)
13.0 244 (25) 510 (284) 215 (32) 608 (208) 249 (52) 218 (55)
Values in parentheses represent one standard deviation.
The relationship between the force equivalent EMG response of each muscle and the head acceleration are shown in Table 3. To obtain the force equivalency of a muscle response due to impact, we first performed a linear regression analysis on the graded EMG data obtained in the maximal voluntary contraction trials. This resulted inan equation for force/emg ratio. EMG values from each muscle as measured in this impact study were then entered into the equation, giving us a force equivalent value (Newtons) for each muscle as shown in Table 3. The kinematic responses show that very-low velocity impacts produce less force equivalent than the maximal voluntary contraction for the same subject. The head accelerations were correspondingly lower than the sled accelerations in this experiment. For very-low velocity impacts, this is to be expected, as it is usually only when the sled acceleration exceeds 5 g's that head acceleration begins to exceed sled acceleration. This experiment involved less than 2 g accelerations.
Table 3 Mean Force Equivalents (Newtons, N) and Mean Head Accelerations at Time of Maximal EMG in Direction of Travel for Right Anterolateral Impact.
Force Equivalents for Muscle (N)
Sternocleidomastoid Splenius Capitis Trapezius
Sled Acceleration (m/s2) Head Acceleration (m/s2) Left Right Left Right Left Right
Right Head Rotation
4.0 3.6 (0.8) 9 (4) 3 (2) 19 (7) 19 (8) 11 (4) 18 (6)
7.6 6.1 (1.0) 10 (5) 5 (2) 21 (14) 22 (10) 18 (7) 21 (10)
10.7 8.0 (1.1) 11 (6) 6 (2) 23 (10) 26 (9) 21 (5) 27 (11)
13.0 9.7 (1.4) 12 (7) 7 (5) 26 (10) 18 (16) 23 (9) 28 (11)
Left Head Rotation
4.0 4.3 (0.7) 4 (2) 7 (5) 19 (13) 11 (6) 17 (6) 10 (4)
7.6 7.7 (1.3) 4 (3) 10 (6) 29 (13) 12 (8) 22 (7) 11 (6)
10.7 10.0 (1.3) 5 (4) 11 (8) 33 (19) 17 (7) 29 (10) 12 (6)
13.0 11.7 (1.8) 6 (5) 13 (7) 34 (17) 19 (8) 35 (14) 13 (6)
Values in parentheses represent one standard deviation.
Regression analyses
The applied acceleration, and the muscles examined had significant main effects on the peak EMG activity (p < 0.05) as shown in Table 4. We used a linear regression model to plot the available data and extrapolate from the experimental accelerations to accelerations on the order of 30 m/s2. Initially, regression analyses were performed only up to 13.0 m/s2 using a linear function. The kinematic variables of head displacement, velocity, and acceleration in response to applied acceleration were calculated (see Fig. 5.). Additionally, we also regressed the EMG magnitudes on acceleration. The responses of the left and right muscle groups were extrapolated to more than twice the applied acceleration value.
Table 4 ANOVA table for Peak EMG (μV) by Muscles and Applied Acceleration.
Source df F Sig.
Right Head
Rotation applied acceleration 3 13.38732 0.00
muscle 5 64.17247 0.00
Left Head
Rotation applied acceleration 3 18.76792 0.00
muscle 5 87.74690 0.00
Figure 5 Extrapolated regression plots of the effect that applied acceleration has on the head motion variables of displacement (A) (mm), velocity (B) (m/s), and acceleration (C) obtained (m/s2).
Discussion
The chief purpose of this study was to see what effect head rotation had on muscle responses in a right anterolateral impact. When the head was in neutral position in a previous study of right anterolateral impact [15], the left trapezius generated the greatest EMG, up to 83% of the maximal voluntary contraction EMG, and the left splenius capitis instead became more active and reached a level of 46% of this variable. In the current study, having kept the impact direction constant, but varying head rotation to right or left we see that the muscles responsible for head rotation (the contralateral sternocleidomastoid), and those which are likely stretched by this rotation (the ipsilateral trapezius), are most active and differ from their counterparts.
Although one might predict this, the human response to impacts and the neck structure is seemingly complex enough that it cannot always be assumed to be as one predicts. Our study methodology allowed for direct testing of the response rather than assumptions. There is no direct way to measure forces exerted by muscles due to neck perturbation and subsequent muscle activity, examining the EMG activity generated allows one to compare this to EMG activity in voluntary contractions. This in turn allows one to relate the muscle responses to normal muscle forces in various physiological ranges of activity. Because one cannot test the higher accelerations for ethical reasons, the best one can do currently is to compare to the small volunteer studies that were done previously. Further studies with larger samples and perhaps somewhat higher accelerations (within ethical limits) will allow to determine further how reasonable these extrapolations are. The projected values are hypothetical and likely to be affected by the ligaments and joint geometry in a manner different from that recorded in the experiment.
In frontal impacts, the direction of impact, anterolateral or straight-on, determines the muscle response, but so too does the occupant's head position, rotated right or left, at the time of impact. Anecdotally at least, whiplash patients report both offset impacts and also may report head rotation to the left or right at the time of impact. These patients also tend to emphasize the unilateral nature of their neck pain, but it remains to be seen in epidemiological studies if this is true. The evidence from low-velocity impacts studies does point in the direction of differential injury risks to different muscles depending on the impact conditions. This is in keeping with other studies of the pattern of muscle activation. Gabriel et al.[19] assessed maximal static strength and bilateral EMG activity associated with force exerted in the direction of the anatomic reference planes, as well as for planes at 30° intervals between the anatomic reference planes. In extending previous work in this area [19,20], Gabriel et al. observed that right-hand dominant subjects have the greatest strength directed to the right side of the body. For this reason, it is important to normalize EMG responses to impact to the subject's maximal voluntary contraction EMG, to account for directional and other confounders. Also, they showed that the SCM muscles are an agonist for static contractions with force exerted in a direction that corresponded to flexion, and a synergist for a force direction associated with lateral bending. It is thus expected that an anterolateral impact will generate the greatest response from the SCMs, and this is consistent with our findings.
Whether or not the pathology of the acute whiplash injury is known, measures to prevent this injury or understand its nature may well be advanced by understanding both the cervical muscle responses and the head kinematics in response to whiplash-type impacts. The difficulty is that besides individual subject characteristics, there are many collision parameters which may affect the pattern of response, including severity of impact, direction of impact, awareness of impending impact, head position, seat design and restraint systems. We have, however, begun the process of a larger series of investigations by showing what effect increasing acceleration, impact direction, head rotation and expectation has on muscle responses when other factors are held constant (i.e. seat and restraint type) [7,13-15]. Future studies can build on this and determine how different seat design or other factors that exist in vehicles affect muscle responses when things such as acceleration, expectation, and direction, for example, are held constant. EMG studies also allow one to examine muscle group responses and patterns, rather than simply describe head or other body region accelerations. The experimental design we have used to study neck perturbations to very low-velocity change is not intended to mimic vehicle occupancy, but rather to allow for the initial exploration of the role of EMG in assessing neck perturbations.
Abbreviations
MVC (Maximal Voluntary Contraction); EMG (Electromyogram); cm (Centrimetres); dB (decibels); C4 (fourth cervical vertebra); mV/g (Millivolts per gram); Hz (Hertz); kHz (kilohertz); g (acceleration due to gravity); m/s2 (metres per second per second); kg (kilograms); SCM (Sternocleidomstoid); TRP (Trapezius); SPL (Splenius capitis)
Competing Interests
The author(s) declare that they have no competing interests.
Authors' Contributions
SK made substantial contributions to conception and design, to acquisition of data, and analysis and interpretation of data, was involved in drafting the article and revising it critically for important intellectual content. RF made substantial contributions to analysis and interpretation of data, and was involved in drafting the article and revising it critically for important intellectual content. YN made substantial contributions to acquisition of data, and analysis and interpretation of data. All authors read and approved the final manuscript.
Acknowledgements
There was no external funding source for this research
==== Refs
Ferrari R The Whiplash Encyclopedia The Facts and Myths of Whiplash 1999 Gaithersburg, Maryland: Aspen Publishers Inc 449 470
Ruedmann AD Jr Automobile safety device – headrest to prevent whiplash injury JAMA 1969 164 1889
Jakobsson L Lundell B Norin H Isaksson-Hellman I WHIPS – Volvo's whiplash protection study Accid Anal Prev 2000 32 307 319 10688487 10.1016/S0001-4575(99)00107-4
Brault JR Wheeler JB Siegmund GP Brault EJ Weber M Peuker C Wortler K. Clinical response of human subjects to rear-end automobile collisions Arch Phys Med Rehabil 1998 79 72 80 9440422 10.1016/S0003-9993(98)90212-X
Castro WH Schilgen M Meyer S Weber M Peuker C Wortler K Do "whiplash injuries" occur in low-speed rear impacts ? Euro Spine J 1997 6 366 375
Kumar S Narayan Y Amell T Role of awareness in head-neck acceleration in low velocity rear-end impacts Accid Anal Prev 2000 32 233 241 10688479 10.1016/S0001-4575(99)00114-1
Kumar S Narayan Y Amell T An electromyographic study of low-velocity rear-end impacts Spine 2002 27 1044 1055 12004171 10.1097/00007632-200205150-00009
Magnusson ML Pope MH Hasselquist L Bolte KM Ross M Goel VK Lee JS Spratt K Clark CR Wilder DG Cervical electromyographic activity during low-speed rear-end impact Euro Spine J 1998 8 118 125 10.1007/s005860050140
Siegmund GP Sanderson DJ Myers BS Inglis JT Awareness affects the response of human subjects exposed to a single whiplash-like perturbation Spine 2003 28 671 679 12671354 10.1097/00007632-200304010-00010
Szabo TJ Welcher J Anderson RD Human occupant kinematic response to low speed rear-end impacts. Proceedings of the Thirty Eighth Stapp Car Crash Conference Warrendale, Pennsylvania: Society of Automotive Engineers 1994 23 35 SAE 940532
Szabo TJ Welcher J Dynamics of low speed crash tests with energy absorbing bumpers Warrendale, Pennsylvania: Society of Automotive Engineers 1992 101 1367 1375 SAE 921573
Cassidy JD Carroll LJ Cote P Lemstra M Berglund A Nygren A Effect of eliminating compensation for pain and suffering on the outcome of insurance claims for whiplash injury N Engl J Med 2000 342 1179 1186 10770984 10.1056/NEJM200004203421606
Kumar S Ferrari R Narayan Y Turning away from whiplash. An EMG study of head rotation in whiplash impact J Orthop Res
Kumar S Narayan Y Amell T Analysis of low-velocity frontal impacts Clin Biomech 2003 18 694 703 10.1016/S0268-0033(03)00137-2
Kumar S Ferrari R Narayan Y Cervical muscle response to whiplash-type right anterolateral impacts Euro Spine J 2004 13 398 407
Kumar S Narayan Y Amell T Cervical strength of young adults in sagittal, coronal, and intermediate planes Clin Biomech 2001 6 380 388 10.1016/S0268-0033(01)00023-7
Kumar S Narayan Y Amell T Ferrari R Electromyography of superficial cervical muscles with exertions in sagittal, coronal, and oblique planes Euro Spine J 2002 11 27 37 10.1007/s005860100318
Kumar S Narayan Y Amell T An electromyographic study of low-velocity rear-end impacts Spine 2002 27 1044 1055 12004171 10.1097/00007632-200205150-00009
Spitzer WO Skovron ML Salmi LR Cassidy JD Duranceau J Suissa S Zeiss E Scientific monograph of the Quebec Task Force on Whiplash-Associated Disorders Spine 1995 120 1S 73S 7604354
Gabriel DA Matsumoto JY Davis DH Currier BL An KN Multidirectional neck strength and electromyographic activity for normal controls Clin Biomech 2004 19 653 658 10.1016/j.clinbiomech.2004.04.016
Vasavada AN Peterson BW Delp SL Three-dimensional spatial tuning of neck muscle activation in humans Exp Brain Res 2002 147 437 448 12444475 10.1007/s00221-002-1275-6
| 15927056 | PMC1156933 | CC BY | 2021-01-04 16:37:41 | no | J Neuroengineering Rehabil. 2005 May 31; 2:11 | utf-8 | J Neuroeng Rehabil | 2,005 | 10.1186/1743-0003-2-11 | oa_comm |
==== Front
Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-241590452010.1186/1475-2875-4-24ReviewIntrarectal quinine for treating Plasmodium falciparum malaria: a systematic review Eisenhut Michael [email protected] Aika [email protected] Harriet G [email protected] Arrowe Park Hospital, Wirral Hospital NHS Trust, Upton, Wirral, UK2 Paediatric Department, Glan Clwyd Hospital, Clwyd, UK3 International Health Research Group, Liverpool School of Tropical Medicine, Liverpool, UK2005 18 5 2005 4 24 24 1 2 2005 18 5 2005 Copyright © 2005 Eisenhut et al; licensee BioMed Central Ltd.2005Eisenhut et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In children with malaria caused by Plasmodium falciparum, quinine administered rectally may be easier to use and less painful than intramuscular or intravenous administration. The objective of this review was to compare the effectiveness of intrarectal with intravenous or intramuscular quinine for treating falciparum malaria.
Methods
All randomized and quasi-randomized controlled trials comparing intrarectal with intramuscular or intravenous quinine for treating people with falciparum malaria located through the following sources were included: Cochrane Infectious Diseases Group Specialized Register, CENTRAL, MEDLINE, EMBASE, LILACS and CINAHL. Trial quality was assessed and data, including adverse event data, were extracted. Dichotomous data were analysed using odds ratios and continuous data using weighted mean difference.
Results
Eight randomized controlled trials (1,247 children) fulfilled the inclusion criteria. The same principal investigator led seven of the trials. Five compared intrarectal with intravenous quinine, and six compared intrarectal with intramuscular treatment. No statistically significant difference was detected for death, parasite clearance by 48 hours and seven days, parasite and fever clearance time, coma recovery time, duration of hospitalization and time before drinking began. One trial (898 children) reported that intrarectal was less painful than intramuscular administration.
Conclusion
No difference in the effect on parasites and clinical illness was detected for the use of intrarectal quinine compared with other routes, but most trials were small. Pain during application may be less with intrarectal quinine. Further larger trials, in patients with severe malaria and in adults, are required before the intrarectal route could be recommended.
==== Body
Background
Plasmodium falciparum malaria often causes serious illness, particularly in Africa, South-East Asia and South America. An estimated 200 million episodes of clinical malaria and two million deaths occur in children under five years old in Africa every year [1].
Uncomplicated malarial illness is usually treated with oral drugs [2]. But as vomiting is a prominent feature in 30 to 50 per cent of people with P. falciparum malaria [3-6], those who present to hospital with persistent vomiting (regardless of severity of disease) or severe malaria require other routes of administration, such as intravenous infusion, intramuscular injection [7] or via the nasogastric route. These different routes of administration require trained staff and equipment, which may be in short supply in low- and middle-income countries.
Despite emerging resistance to commonly used drugs, such as chloroquine and mefloquine, malaria parasites remain sensitive to quinine in Africa [8,9]. In some parts of South-East Asia, however, decreasing sensitivity to quinine has been detected [10].
Although intramuscular injection is the most common route of quinine administration used in low-and middle-income countries, side effects have been reported [11]. In some of these countries, it is the most common cause of lower limb paralysis when administered mistakenly into the sciatic nerve [12-14]. Other reported harmful effects of intramuscular quinine injections are bacterial and viral infections including tetanus [15], poliomyelitis [16,17] and HIV [18,19]. An alternative to intravenous and intramuscular administration is, therefore, worth evaluating.
The intrarectal route has been used to give quinine [20]. Health workers with minimal training can give intrarectal quinine to people who are either vomiting or comatose. This provides early treatment of the illness and is one of the strategies of the World Health Organization initiative 'Roll Back Malaria' . However, disadvantages of using the intrarectal route are local irritation, diarrhoea and expulsion of the medication [21]. The likelihood of intrarectal irritation has been reduced by the development of less acidic quinine gluconate (in Quinimax®). People may also reject suppositories and other intrarectal formulations in preference for the intramuscular route because injections are perceived as a more effective treatment, particularly in people who are seriously ill [19].
This review summarizes existing trials that compare the effectiveness and safety of intrarectal quinine with other routes of administration in people with malaria caused by P. falciparum.
The paper is based on a Cochrane review published in The Cochrane Library 2005, Issue 1 [22].
Cochrane reviews are regularly updated as new evidence emerges and in response to comments and criticisms, and The Cochrane Library should be consulted for the most recent version of the review. The conclusions presented are the opinions of review authors and are not necessarily shared by The Cochrane Collaboration.
Methods
The review's inclusion criteria were randomized and quasi-randomized controlled trials comparing intrarectal with intravenous or intramuscular quinine administration in patients with P. falciparum malaria (uncomplicated or severe, and confirmed by a blood-slide examination). Quinine could be used as a single therapy or in combination.
The review's primary outcome measure was death, but parasite clearance by 48 hours (number of participants free of parasites by 48 hours), parasite clearance by day seven (number of participants free of parasites by day seven), parasite clearance time, fever clearance time, duration of hospitalization, coma recovery time, time lapse before drinking or eating, and adverse events (serious events that resulted in death, were life-threatening, required hospitalization or resulted in discontinuation of treatment, such as local pain, abscess formation, and paralysis, mild or moderate adverse events, as classified or defined by trial investigators, including vertigo and tinnitus) were also measured.
The search strategy aimed to identify all relevant trials regardless of language or publication status (published, unpublished, in press and in progress). The following search terms were used: 'quinine', 'Quinimax®', 'cinchona alkaloids', 'cinchona alkaloid', 'suppositories', 'suppository', 'rectal drug administration', 'administration, rectal', 'intrarectal', 'rectal', 'rectum' and 'malaria'. The following databases were searched: Cochrane Infectious Diseases Group Specialized Register (January 2005), Cochrane Central Register of Controlled Trials (CENTRAL) published in The Cochrane Library (Issue 4, 2004), MEDLINE (1966 to January 2005), EMBASE (1974 to January 2005), LILACS (1982 to January 2005) and CINAHL (1982 to January 2005). The following conference proceedings were also searched for relevant abstracts: The Third Multilateral Initiative on Malaria Pan-African Conference, 18 to 22 November, 2002, Arusha, Tanzania, and the Third European Congress on Tropical Medicine and International Health held in Lisbon, Portugal (September, 2002). For unpublished or ongoing trials, individual researchers working in the field and the pharmaceutical company Sanofi-Synthélabo, which manufactured Quinimax® suppositories and intrarectal cream, were contacted. The reference lists of all studies identified by the above methods were also checked.
The full reports for all potentially relevant trials were retrieved and independently assessed for their eligibility. For methodological quality, allocation sequence and allocation concealment were independently assessed to be adequate, inadequate or unclear according to established guidelines [23]. It was noted whether the participant, carer or outcome assessor was blind to the intervention, and the inclusion of all randomized participants in the final analysis was considered to be adequate if greater than 90%. Data were independently extracted using standard forms, and trial authors were contacted where additional unpublished data were required. Data were analysed using Review Manager 4.2 software and outcome measures were compared using odds ratios (OR) for dichotomous data and weighted mean difference (WMD) for continuous data, both with 95% confidence intervals. The fixed-effect model was used for those without statistically significant heterogeneity (see below). Data were pooled on the same interventions (same route of administration and drug regimen), where appropriate and separate analyses for the intravenous and intramuscular control regimens were conducted. Adverse event data were presented in a table, a meta-analysis and in a narrative summary of the findings.
Heterogeneity was assessed by visually examining the forest plots (for overlapping confidence intervals and outliers) and using the chi-squared test for heterogeneity with a 10% level of statistical significance. Because statistically significant heterogeneity was detected for diarrhoea (an adverse event), the DerSimonian and Laird random-effects model was used to pool data for this outcome.
It was intended to use subgroup analyses or meta-regression to explore participant age (less than five years versus five years or older), disease severity (uncomplicated versus severe) and different galenic quinine formulations (solution, intrarectal cream or suppositories) as potential sources of heterogeneity, but this was not possible because of the uniformity of the age of participants (children less than 15 years of age only) and the small number of trials of people with severe disease and different galenic formulations. It was intended to investigate publication bias using funnel plots, but it was considered to be inappropriate in view of the small number of trials included.
Results
Description of studies
The search strategy identified 14 potentially relevant studies. Eight randomized controlled trials (involving 1,247 children) [20,24-30], one of which was quasi-randomized [24], fulfilled the inclusion criteria. The same principal investigator led seven of the trials. The trials recruited children up to 15 years of age who were hospital in-patients in Burkina Faso [26,30], Niger [20,24,25,27,29] or Togo [28]. The criteria for inclusion in the trial were: the degree of parasitaemia (>1,000 asexual P. falciparum stages/μl) [20,24-27], vomiting [20,24,26,28] and severe malaria [25,29]. All the trials had diarrhoea as an exclusion criterion; some also used treatment with antimalarial drugs before admission (seven trials), other documented causes of fever (six trials) and forms of severe malaria (five trials). Details on the number of participants, interventions and outcomes investigated are listed in Table 1. Five trials (227 children) compared intrarectal with intravenously administered quinine. Four trials (179 children) compared intrarectal quinine administered for two to three days with intravenous quinine administered for the same duration [20,24-26], and one trial compared single doses of intrarectal quinine and intravenous quinine that were followed by a three-day course of oral quinine [27]. Participants that had a two-day quinine course completed a total of five days of treatment with oral chloroquine [26] or quinine [27]. Six trials (1,122 children) compared intrarectal with intramuscular quinine. Four trials compared intrarectal and intramuscular quinine administered for three days [20,24,28,30], one trial did not mention the duration of treatment [29] and one trial gave a single dose of intrarectal or intramuscular quinine followed by three days of oral quinine [27].
Table 1 Trial participants, interventions and outcomes
Reference Participants Quinine: routes of administration and doses* Outcomes
[20] 21 children aged 2 to 14 years in Niger (1) Intrarectal (8 mg/kg; gluconate cream)
(2) Intramuscular (4.7 mg/kg)
(3) Intravenous (4.7 mg/kg)
All 8 hourly for 3 days Death, parasite clearance by day 7, adverse events
[24] 66 children aged 2 to 15 years in Niger (1) Intrarectal (11.8 mg/kg; given as Quinimax® diluted with water in a syringe)
(2) Intravenous (7.4 mg/kg)
(3) Intramuscular (7.4 mg/kg)
All 12 hourly for 3 days Death, parasite clearance time, parasite clearance by day 7, fever clearance time, adverse events
[25] 77 children aged 2 to 15 years in Niger (1) Intrarectal (11.8 mg/kg once then 8.8 mg/kg 8 hourly for 2 days; Quinimax® solution diluted in water via a syringe)
(2) Intravenous (4.7 mg/kg 8 hourly for 2 days) Death, parasite clearance time, fever clearance time, days in hospital, coma recovery time, time until drinking began, adverse events
[26] 48 children aged 2 to 15 years in Burkina Faso (1) Intrarectal (bichlorhydrate diluted in a syringe)
(2) Intrarectal (cinchona alkaloid diluted in a syringe)
(3) Intravenous (bichlorhydrate)
(4) Intravenous (cinchona alkaloid)
All 8 mg/kg quinine 8 hourly for 2 days Parasite clearance by 48 hours and 7 days
[27] 15 children aged 2 to 14 years in Niger (1) Intravenous (4.74 mg/kg)
(2) Intramuscular (4.74 mg/kg)
(3) Intrarectal (11.85 mg/kg)
All single dose Death, parasite clearance by day 7, adverse events
[28] 64 children aged 0 to 15 years in Togo (1) Intrarectal (15 mg/kg: Quinimax® solution diluted in a syringe)
(2) Intramuscular (12.5 mg/kg) Both 12 hourly for 3 days Death, parasite clearance by day 2, adverse events
[29] 58 children aged 2 to 15 years in Niger (1) Intrarectal (17.9 mg/kg once, then 11.75 mg/kg 12 hourly; Quinimax® diluted in a syringe
(2) Intramuscular (7.5 mg/kg 12 hourly) Death, fever clearance time, duration of hospitalization, coma recovery time
[30] 898 children aged 1 to 15 in Burkina Faso (1) Intrarectal (20 mg/kg; Quinimax® diluted in a syringe)
(2) Intramuscular (12.5 mg/kg) Both 12 hourly for 3 days Death, fever clearance time, adverse events
*Quinine dose given as quinine base
Methodological quality of included studies
Four trials did not describe the method used to generate the allocation sequence [20,26,28,30], three trials used random-numbers tables [25,27,29] and one trial used alternate allocation (quasi-randomization) [24]. None used procedures to conceal allocation. The nature of the interventions used did not allow blinding of participants and carers, and none of the outcome assessors were blinded. One trial excluded one participant (1.3%) from the analysis [25]. Another trial [24] could only analyse the parasite clearance time for 20 out of the 66 trial participants (30%) without providing a reason for the missing participants. The other six trials did not report on any exclusion or drop out of randomized participants. None of the trials analysed data on an intention-to-treat basis.
None of the trials reviewed contained a power calculation to determine the number of participants required to achieve sufficient statistical power for a clinically meaningful difference in effect of the different modes of administration to be detected.
Outcomes
Each trial reported on at least one of the review's pre-specified outcomes, but none reported on the time lapsed before eating resumed. They also reported on other outcomes not analysed in this systematic review: time for parasitaemia to fall by 50% (three trials), percentage of initial parasitaemia after 24 hours (one trial) and 48 hours (three trials), time before one could sit (one trial); time before one could walk (one trial) and time for the body temperature to fall below 37.5°C (one trial).
Effects of intrarectally versus intravenously administered quinine
There was no statistically significant difference between the interventions for the number of deaths (100 participants, three trials) (Figure 1), parasite clearance at 48 hours (24 participants, one trial), parasite clearance time (76 participants, one trial), fever clearance time (76 participants, one trial) (Figure 2), duration of hospitalization (76 participants, one trial) coma recovery time (76 participants, one trial) and time until drinking began (76 participants, one trial). Participants in two trials were reported to have cleared their parasites by 48 hours [26,28] while four trials reported that all participants had cleared their parasites by day seven [20,21,24,26].
Figure 1 Meta-analysis of effects of intrarectal, intravenous and intramuscular quinine on death
Figure 2 Meta-analysis of effects of intrarectal, intravenous and intramuscular quinine on fever clearance time (hours)
Two trials reported on adverse events and specifically mentioned the absence of rectal irritation and diarrhoea [25,26]. Barennes [26] also reported mucoid stools in four children in the intrarectal group.
Effects of intrarectally versus intramuscularly administered quinine
There was no statistically significant difference between the interventions in the number of deaths (1,110 participants, six trials) (Figure 1), parasite clearance at 48 hours (64 participants, one trial), fever clearance time (1,022 participants, three trials) (Figure 2), duration of hospitalization (58 participants, one trial) and coma recovery time (58 participants, one trial). Four trials (154 participants) also reported that no deaths occurred. Parasite clearance time was statistically significantly longer in the participants treated with intrarectal quinine (mean 46.5 (standard deviation 22.0) hours) compared with those treated with intramuscular quinine (27.4 (9.5) hours) (WMD 19.10 hours, 5.20 to 33.00; 20 participants, one trial) [24]. However, all ninety participants in three trials [20,24,27] were reported to have cleared their parasites by day seven.
Data on adverse events were accessible for statistical analysis in three trials [23,27,29]. Assimadi [28] reported no statistically significant difference between painful swelling at the site of application (OR 0.13, 0.01 to 2.62, random-effects model) and pain at the site of application after administration (OR 0.1, 0.01 to 1.89, random-effects model). Barennes [30] reported that pain during administration occurred in four out of 450 participants given intrarectal quinine and 412 out of 448 participants given intramuscular quinine (OR 0.00, 0.00 to 0.00), with a test result for overall effect of Z = 13.46 (P < 0.00001). There was no statistically significant difference in the number of participants with mild diarrhoea between the groups (1,022 participants, three trials). The largest trial [30] also documented adverse events affecting stool consistency and content, pain in the rectum, effects on the rectal mucosa, as well as effects specific to intramuscular administration (Table 2).
Table 2 Descriptive adverse event data from Barennes [30]
Adverse event Intrarectal (number affected/total number of patients, %) Intramuscular (number affected/total number of patients, %)
Liquid stool (<3/day) 109/450 (24.2) 7/448 (1.5)
Mucoid stools 296/450 (65.7) 23/448 (5.10)
Blood in stool 20/450 (4.4) 3/448 (0.7)
Painful contractions 64/450 (14.2) No data
Inflammation at the injection site No data 358/448 (80)
Tenesmus 56/450 (12.4) No data
Number investigated by anoscopy with a single microulceration healing within 24 hours 4/259 (1.5) No data
Multiple microulcerations recovering by day 7 1/259 (0.4) No data
Difficulty in walking No data 67/448 (15)
Sciatic paresthesia No data 1/448 (0.2)
Fever recurrence due to inflammation or infection of the injection site No data 30/448 (6.6)
Three trials that reported on adverse events did not separate the results for the intrarectal, intramuscular and intravenous groups [20,24,27]. They specifically commented on the absence of rectal irritation (all three trials) and diarrhoea [20,27]. Barennes [20] also observed slight pain at the injection site in the intramuscular group.
Discussion
Seven out of the eight trials that met the inclusion criteria consisted of less than 80 participants. This small number of participants increased the probability of missing a clinically important difference between groups. Only for the outcomes of death, fever clearance time and mild diarrhoea (an adverse event) were there two or more trials available for a meta-analysis. Only three of the trials documented the use of adequate randomization; adequate randomization was particularly important because blinding of participant and carer was not possible. This has increased the risk of a selection bias. All but one trial were conducted under the lead of one author, H. Barennes. The only pharmaceutical company producing an intrarectal quinine preparation, Quinimax®, sponsored four of the trials [24,25,29,30].
There was no statistically significant difference between intrarectal and parenteral quinine administration in terms of death, course of P. falciparum malaria or diarrhoea. Intrarectal administration also had the benefit of being less painful.
Parasite clearance time was longer in participants given quinine intrarectally as compared with intramuscular treatment in one trial [24], but it was not different when intrarectal administration was compared with intravenous administration in another trial [25]. This may have been because parasitaemia in the trial Barennes published in 1995 [24] was three times higher at baseline in the intrarectal group.
Statistically significant heterogeneity was observed when analysing the mild diarrhoea adverse event outcome. This may have been due to different definitions of diarrhoea, which was only clearly defined in one trial [30], and the large weight attributed in the meta-analysis to one small trial in which two out of five participants in the control group were affected [24]. Persistent pain at the injection site due to inflammation with the recurrence of fever seemed to be common with intramuscular injection and is an adverse effect not observed with the intrarectal route. This should be taken into consideration in the design of future trials comparing the two modes of administration. The occurrence of rectal mucosal ulcerations with intrarectal administration and its significance should also be assessed in all future trials. Adverse effects unique to the methods of intramuscular administration (sciatic nerve injury, infections with other viral and bacterial pathogens through contaminated needles) or intravenous injection (infections) are absent in intrarectal administration and cannot be addressed in a trial setting where administration is performed by trained personnel with adequate supply of consumables.
All trials included in the review had only children as participants, therefore, the results may not be applicable to adults.
Only two small randomized controlled trials (135 children) comparing intrarectal with intravenous [25] or intramuscular treatment [29] had participants with severe malaria. Limited data are, therefore, available on the effectiveness of intrarectal quinine in life-threatening forms of malaria.
Conclusion
In a series of unconcealed trials from one research group, intrarectal administration appeared to be superior to intramuscular or intravenous administration in terms of reduced pain during administration. No difference in the effect on parasites and clinical illness was detected for the use of intrarectal quinine compared with other routes, but most trials were small. Thus, intrarectal application may be superior for uncomplicated falciparum malaria in children in cases in which administration of antimalarial drugs by mouth is not possible. There is insufficient evidence of the effectiveness of intrarectal quinine in severe falciparum malaria in children.
Further large-scale concealed randomized controlled trials are required to investigate intrarectally administered quinine in severe falciparum malaria in children and in all forms of falciparum malaria in adults. Further trials should focus on adverse effects including short-term and long-term effects on the rectal mucosa with intrarectal administration. Trials investigating the use of intrarectal quinine in the primary care setting, its role in preventing hospital admission and early treatment in the community, preventing complications associated with late presentation at healthcare facilities, are also desirable.
Authors' contributions
Michael Eisenhut and Aika Omari developed the protocol, searched the literature and extracted data. Michael Eisenhut wrote the review, which Harriet G MacLehose critically revised. All authors read and approved the final manuscript.
Acknowledgements
We thank all members of the Cochrane Infectious Diseases Group at the Liverpool School of Tropical Medicine for their generous support in the review and criticism of our work, and support with literature retrieval. We would like to thank Dr Tobias Eisenhut and Dr Ai Koyanagi for translating and extracting data from the French-language articles.
==== Refs
Greenwood B Marsh K Snow R Why do some African children develop severe malaria? Parasitol Today 1991 7 277 281 15463389 10.1016/0169-4758(91)90096-7
Severe falciparum malaria. World Health Organization, Communicable Diseases Cluster Trans R Soc Trop Med Hyg 2000 94 S1 S90 11103309
Anothay O Pongvongsa T Maharat N Sirivichayakul C Chantavanich P Silachamroon U Looareesuwan S Clinical presentation of childhood malaria in Savannakhet province, Lao PDR Southeast Asian J Trop Med Public Health 2000 31 85 90 11414467
Kortepeter M Brown JD A review of 79 patients with malaria seen at a military hospital in Hawaii from 1979–95 Mil Med 1998 163 84 89 9503898
Piarroux R Dumon H Garnier JM Lehmann M Unal D Quilici M Choice of therapy for imported cases of falciparum malaria in children: a retrospective study of 100 cases seen in Marseilles, France Trans R Soc Trop Med Hyg 1993 87 72 74 8465403 10.1016/0035-9203(93)90429-T
Sowunmi A Ogundahunsi OA Falade CO Gbotosho GO Oduola AM Gastrointestinal manifestations of acute falciparum malaria in children Acta Trop 2000 74 73 76 10643910 10.1016/S0001-706X(99)00043-1
White NJ Looreesuwan S Warrell DA Warrell MJ Bunnag D Harinasuta T Quinine pharmacokinetics and toxicity in cerebral and uncomplicated falciparum malaria Am J Med 1982 73 564 572 6751085 10.1016/0002-9343(82)90337-0
Bjorkman A Quinine-resistant malaria in Africa Trans R Soc Trop Med Hyg 1991 85 138 139 2068747 10.1016/0035-9203(91)90193-3
Jelinek T Schelbert P Loescher T Eichenlaub D Quinine resistant falciparum malaria acquired in East Africa Trop Med Parasitol 1995 46 38 40 7631126
Global Partnership to Roll Back Malaria. The use of antimalarial drugs: report of a WHO informal consultation, 13–17 November 2000 2001 Geneva: World Health Organization
Barennes H Les injections intramusculaires chez l'enfant en Afrique subsaharienne, à propos d'une pathologie souvent méconnue: les complications liées aux injections intramusculaires de quinine Bull Soc Pathol Exot 1999 92 33 37 10214519
Barennes H Renouil M Macaigne F Raoseta V Raoeline Y Efficacité de la quinine injectable diluée en intra-rectal dans le paludisme de l'enfant et excrétion urinaire de la quinine Médecine d'Afrique Noire 1993 40 401 404
Bourrel P Souvestre R Les lésions du nerf sciatique par injections intra-fessières de quinine Med Trop 1982 42 209 221
Carayon A Blanc JF Caron JJ Accidents nerveux par infections intra-fessières d'antipaludiques. Indication de la neurolyse des troncs plexuels sciatiques Bull Soc Méd Afr Noire Lang Franç 1960 52 112 123 15023494
Yen LM Dao LM Day NP Waller DJ Bethell DB Son LH Hien TT White NJ Role of quinine in the high mortality of intramuscular injection tetanus Lancet 1994 344 786 787 7916074 10.1016/S0140-6736(94)92342-6
Wyatt HV Poliomyelitis in developing countries: lower limb paralysis and infections Trans R Soc Trop Med Hyg 1989 83 545 549 2617611 10.1016/0035-9203(89)90286-1
Wyatt HV Mahadevan S Srinivasan S Unnecessary injections and paralytic poliomyelitis in India Trans R Soc Trop Med Hyg 1992 86 546 549 1475830 10.1016/0035-9203(92)90104-K
Hoelscher M Riedner G Hemed Y Wagner H Korte R von Sonnenburg F Estimating the number of HIV transmissions through reused syringes and needles in the Mbeya Region, Tanzania AIDS 1994 8 1609 1615 7848599
Wyatt HV The popularity of injections in the Third World: origins and consequences for poliomyelitis Soc Sci Med 1984 19 911 915 6515426 10.1016/0277-9536(84)90320-4
Barennes H Pussard E Mahaman Sani A Clavier F Kahiatani F Ganic G Henzel D Ravinet L Verdier F Efficacy and pharmacokinetics of a new intrarectal quinine formulation in children with Plasmodium falciparum malaria Br J Clin Pharmacol 1996 41 389 395 8735679 10.1046/j.1365-2125.1996.03246.x
Barennes H Mahaman Sani A Kahia Tani F Meda H Khenine A Tolérance de la quinine administrée en solution intrarectale chez l'enfant en Afrique Francophone Med Trop 1999 59 383 388
Eisenhut M Omari AAA Intrarectal quinine for treating Plasmodium falciparum malaria The Cochrane Database of Systematic Reviews 2005 Art. No.: CD 004009. DOI: 10.1002/14651858/CD004009.pub2
Juni P Altman DG Egger M Systematic reviews in health care: Assessing the quality of controlled clinical trials BMJ 2001 323 42 46 11440947 10.1136/bmj.323.7303.42
Barennes H Kahiatani D Pussard E Clavier F Meynard D Njifountawouo S Verdier F Intrarectal Quinimax (an association of Cinchona alkaloids) for the treatment of Plasmodium falciparum malaria in children in Niger: efficacy and pharmacokinetics Trans R Soc Trop Med Hyg 1995 89 418 421 7570885 10.1016/0035-9203(95)90036-5
Barennes H Munjakazi J Verdier F Clavier F Pussard E An open randomized clinical study of intrarectal versus infused Quinimax for the treatment of childhood cerebral malaria in Niger Trans R Soc Trop Med Hyg 1998 92 437 440 9850403 10.1016/S0035-9203(98)91083-5
Barennes H Sterlingot H Nagot N Meda H Kabore M Sanou M Nacro B Bouree P Pussard E Intrarectal pharmacokinetics of two formulations of quinine in children with falciparum malaria Eur J Clin Pharmacol 2003 58 649 652 12610739
Barennes H Verdier F Clavier F Pussard E Pharmacokinetics of Quinimax suppositories in children with malaria: a preliminary study Clinical Drug Investigation 1999 17 287 291
Assimadi JK Gbadoe AD Agbodjan-Djossou O Larsen SE Kusiaku K Lawson-Evi K Redah D Adjogble A Gayibor A Quinine injectable diluée en intramusculaire et en intrarectal: efficacité et tolérance comparées dans le traitement du paludisme de l'enfant Med Trop 2002 62 158 162
Barennes H Kailou D Pussard E Munjakazi JM Fernan M Sherouat H Sanda A Clavier F Verdier F Administration intrarectale de la quinine: un traitement précoce du paludisme grave de l'enfant? Santé 2001 11 145 153
Barennes H Intrarectal administration of Quinimax® in children: bioavailability, efficacy and tolerance Symposium workbook, The Third MIM Pan-African Conference on Malaria, 17–22 November, Arusha, Tanzania 2002 16 23
| 15904520 | PMC1156934 | CC BY | 2021-01-04 16:24:13 | no | Malar J. 2005 May 18; 4:24 | utf-8 | Malar J | 2,005 | 10.1186/1475-2875-4-24 | oa_comm |
==== Front
Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-251591890910.1186/1475-2875-4-25ResearchExperimental hut evaluation of bednets treated with an organophosphate (chlorpyrifos-methyl) or a pyrethroid (lambdacyhalothrin) alone and in combination against insecticide-resistant Anopheles gambiae and Culex quinquefasciatus mosquitoes Asidi Alex N [email protected]' Guessan Raphael 12Raphael.N¿[email protected] Alphonsine A [email protected] Christopher F [email protected] Jean-Marc [email protected] Fabrice [email protected] Vincent [email protected] Frédéric [email protected] Morteza [email protected] Mark W [email protected] London School of Hygiene & Tropical Medicine, London, WC1E 7HT, UK2 Centre Pierre Richet, 01 PO Box 1500, Bouaké 01, Côte d'Ivoire, France3 Laboratoire de Lutte contre les Insectes Nuisibles (LIN), 911 Avenue Agropolis, PO Box 64501, 34394 Montpellier Cedex 5, France4 World Health Organization, 27 Avenue Appia, CH-1211 Geneva 27, Switzerland2005 26 5 2005 4 25 25 22 12 2004 26 5 2005 Copyright © 2005 Asidi et al; licensee BioMed Central Ltd.2005Asidi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Pyrethroid resistant mosquitoes are becoming increasingly common in parts of Africa. It is important to identify alternative insecticides which, if necessary, could be used to replace or supplement the pyrethroids for use on treated nets. Certain compounds of an earlier generation of insecticides, the organophosphates may have potential as net treatments.
Methods
Comparative studies of chlorpyrifos-methyl (CM), an organophosphate with low mammalian toxicity, and lambdacyhalothrin (L), a pyrethroid, were conducted in experimental huts in Côte d'Ivoire, West Africa. Anopheles gambiae and Culex quinquefasciatus mosquitoes from the area are resistant to pyrethroids and organophosphates (kdr and insensitive acetylcholinesterase Ace.1R). Several treatments and application rates on intact or holed nets were evaluated, including single treatments, mixtures, and differential wall/ceiling treatments.
Results and Conclusion
All of the treatments were effective in reducing blood feeding from sleepers under the nets and in killing both species of mosquito, despite the presence of the kdr and Ace.1R genes at high frequency. In most cases, the effects of the various treatments did not differ significantly. Five washes of the nets in soap solution did not reduce the impact of the insecticides on A. gambiae mortality, but did lead to an increase in blood feeding. The three combinations performed no differently from the single insecticide treatments, but the low dose mixture performed encouragingly well indicating that such combinations might be used for controlling insecticide resistant mosquitoes. Mortality of mosquitoes that carried both Ace.1R and Ace.1S genes did not differ significantly from mosquitoes that carried only Ace.1S genes on any of the treated nets, indicating that the Ace.1R allele does not confer effective resistance to chlorpyrifos-methyl under the realistic conditions of an experimental hut.
==== Body
Background
Insecticide-treated nets (ITN) are an important component of the Roll Back Malaria campaign to reduce malaria morbidity and mortality in Africa. Pyrethroids are the only group of insecticides currently recommended for use on nets [1]. In recent years, pyrethroid resistance has become widespread among anopheline mosquitoes in western Africa and has also arisen in eastern and southern Africa [2-6]. Pyrethroid resistance has evolved concurrently in the filariasis vector and nuisance mosquito Culex quinquefasciatus [7,8].
Initial alarm over the rapid spread of the kdr allele responsible for pyrethoid resistance [9] has been tempered by recent evidence which indicates that nets incorporating permethrin continue to reduce malaria transmission and morbidity in an area known to have a high frequency of kdr [10]. Experimental hut studies in a similar area confirm that pyrethroid treated nets continue to kill pyrethroid resistant mosquitoes [11-14]. However, in another PCR genotyping hut study by Kolaczinski et al [15] showed a significantly higher frequency of the kdr gene among survivors of alphacypermethrin- or etofenprox-treated nets than in mosquitoes which were killed by these treatments. In Kenya, a different form of the kdr gene was found in Anopheles gambiae by Ranson et al [5], in an area where one of the most successful large-scale ITN trials was subsequently carried out [16,17]. These findings may allay initial fears but it would be complacent to assume that pyrethroid-treated nets will remain effective indefinitely. Selection of supplementary resistance mechanisms could tip the balance towards control failure [3,18,19], as happened with indoor residual spraying in South Africa owing to selection of a metabolic form of resistance in Anopheles funestus which required switching back to DDT to restore malaria control [6]. The danger of pyrethroid resistance is apparent in Tanzanian C. quinquefasciatus. It effectively prevents mortality with pyrethroid treated nets in experimental huts [20] and prevents Culex population suppression when ITNs are used by whole communities, in contrast to high mortality and population suppression of susceptible A. gambiae populations under these conditions [21].
Finding an alternative to the pyrethroids has, therefore, become a priority [22]. Some members of the earlier generation of insecticides, the organophosphates (OPs) and carbamates, although developed primarily for agricultural use and for indoor residual spraying, may have potential as net treatments. If there is a complete lack of cross-resistance to pyrethroid resistant mosquitoes, use of such compounds in combination with pyrethroids may provide an opportunity for resistance management [23,24].
Experimental hut studies have shown that the performance of the carbamate carbosulfan on nets against pyrethroid- resistant Anopheles and carbamate- resistant Culex mosquitoes is equivalent or better than that of pyrethroids against susceptible mosquitoes [14,15]. Because of a perceived risk of human toxicity with a carbosulfan breakdown product, carbofuran, some doubt has been cast over the suitability of carbosulfan as a net treatment [13,25]. Chlorpyrifos-methyl (Reldan®, Dow AgroSciences), an OP used in agriculture to control stored product pests, may be more suitable, being classified by WHO as unlikely to present acute hazard in normal use, whereas carbosulfan and the majority of pyrethroids recommended for treatment of mosquito nets are classified by WHO as Class II, i.e. moderately hazardous [26].
Broad spectrum resistance to organophosphates and carbamates caused by insensitive acetylcholinesterase mechanisms have been identified in A. gambiae and C. quinquefasciatus from Côte d'Ivoire [4,27,28]. To assess the potential of chlorpyrifos-methyl to control resistant anophelines and culicines, the performance of nets treated with chlorpyrifos-methyl, lambdacyhalothrin and various combinations of the two was compared in experimental huts near Bouaké, Côte d'Ivoire.
Materials and methods
Study area and experimental huts
The treated nets were evaluated in 11 experimental huts at a field site of the Institut Pierre Richet in Yaokoffikro, Bouaké, Côte d'Ivoire. The huts were situated near rice and vegetable fields and arrayed in two rows with a 5 metre gap between huts. The style of the hut was typical of the region. Each was made from concrete bricks, with a corrugated iron roof, a ceiling of thick polyethylene sheeting, and a concrete base surrounded by a water-filled channel to prevent entry of ants [11]. Mosquito access was via 4 window slits constructed from pieces of plywood, fixed at an angle to create a funnel with a 1 cm wide gap. Mosquitoes had to fly upward to enter through the gap and downwards to exit; this precluded or, at worst, limited exodus though the aperture enabling the majority of entering mosquitoes to be accounted for. A verandah trap made of polyethylene sheeting and screening mesh measuring 2 m long, 1.5 m wide, and 1.5 m high, projected from the back wall of each hut. Movement of mosquitoes between hut and verandah was unimpeded. All huts were thoroughly cleaned before the trial. Sheets were laid over the floor each night to ease the collecting of knocked down mosquitoes in the morning.
Bednets
The nets were made from white 100-denier polyester (SiamDutch Mosquito Netting Co., Thailand). They measured 1.9 m long, 1.8 m wide and 1.5 m high, and had a surface area of 14.5 m2. To simulate badly torn nets, 80 holes, each measuring 2 × 2 cm, were cut in the sides and ends of all but two of the nets.
Insecticide treatment
The insecticides used were:
chlorpyrifos-methyl 38.8% CS ('Reldan GF934', Dow AgroSciences), an experimental microencapsulated suspension, designed for controlled residual activity and short outdoor persistence).
lambdacyhalothrin 2.5% CS, ('Icon', Syngenta UK), a commercial microencapsulated formulation.
Impregnation of nets was carried out using the formula of Pleass et al [29] to calculate the amount of insecticide needed. The nine treatments and target application rates were:
lambdacyalothrin 18 mg/m2, holed, unwashed
lambdacyalothrin 18 mg/m2, holed, washed with soap 5 times
chlorpyrifos-methyl 100 mg/m2, holed, unwashed
chlorpyrifos-methyl 100 mg/m2, intact, washed with soap 5 times
chlorpyrifos-methyl 250 mg/m2, holed, unwashed
chlorpyrifos-methyl 250 mg/m2, holed, washed with soap 5 times
mixture of chlorpyrifos-methyl 100 mg/m2 and lambdacyalothrin 18 mg/m2, holed, unwashed
mixture of chlorpyrifos-methyl 25 mg/m2 and lambdacyalothrin 4.5 mg/m2, holed, unwashed
chlorpyrifos-methyl 100 mg/m2 on ceiling of net and lambdacyalothrin 18 mg/m2 on walls, holed, unwashed. Differential treatment of the ceiling and walls of a net has been called "two-in-one" by Guillet et al [25] and a "mosaic" by others [13]. However, the term mosaic is conventionally used to describe a resistance management strategy in which two or more chemicals are used for spraying different sectors of countryside, the sectors being large enough to contain their "own" mosquito populations in which different resistance gene frequencies would evolve as a result of different selection pressures, with only limited exchange of genes by immigration. Thus, in this paper we have preferred to adopt the term "two-in-one", suggested by Guillet et al [25].
untreated control, intact
untreated control, holed.
The application rate of 18 mg/m2 lambdacyalothrin was the same as that used by Asidi et al [14] and was in the application dose range proposed by WHO. The 100 mg/m2 chlorpyrifos-methyl (CM) treatment was twice the dosage needed to kill 100% of a laboratory OP resistant strain (DUBAI 234) in a 3 min exposure bioassay test. To assess the effect of repeated washing, a higher application rate of CM was used initially (250 mg/m2). Washing was done by hand using a palm-oil soap 'Maxi Mousse', and was repeated five times with a one-day interval between washes. The effect of holed versus intact nets was assessed using untreated and 100 mg/m2 CM treatments. The mixture of lambdacyhalothin and chlorpyrifos-methyl was prepared with the two CS formulations mixed in water. The intention with the low dose mixture (25 mg/m2 CM and 4.5 mg/m2 lambdacyhalothin) was to examine whether efficacy would be maintained after a period of simulated insecticide decay or whether differential decay might lead to selection of Ace.1R or kdr resistance genotypes. The "two-in-one" net was prepared by cutting the ceiling from the walls of the net, dipping the two sections in chlorpyrifos-methyl and lambdacyhalothin respectively, and then sewing the two sections together again. Untreated intact and holed nets were used as controls.
Sleepers and mosquito collections
The treatments were randomly allocated to the eleven experimental huts. Eleven adult men were paid to sleep in the huts each night from 20.00 to 05.00 hours and to collect mosquitoes in the mornings. The sleepers/collectors were experienced in collecting mosquitoes, gave informed consent and were given malaria chemoprophylaxis.
The trial ran for only 33 nights over 6 weeks (from 15 August 2002). The trial was planned to run for 44 nights but had to be curtailed owing to political unrest. The sleepers were rotated between huts to correct for possible variation in individual attractiveness. The nets were not rotated for fear of cross-contaminating huts with different treatments. There was a risk in this experimental design of not being able to separate possible confounding factors due to variation in hut attractiveness independent from that of treatment. However, baseline measurements indicated that the huts were comparable in attractiveness (Table 1). Sleepers were questioned each day during the first two weeks to find out whether they experienced any side effects from using nets.
Table 1 Mean numbers of mosquitoes collected per night over 10 nights before installation of treated nets.
Hut number An. gambiae Other anophelines Culex spp. Aedes spp. Mansonia spp. Overall means
1 4.2 0.2 3.9 0.6 5.2 14.1
2 5.3 0.1 2.7 0.8 3.9 12.9
3 6.5 0.6 1.6 0.5 5.4 14.6
4 5.1 0.3 2.9 0.6 3.8 12.7
5 6.1 0.4 2.8 1.0 3.6 13.9
6 5.4 0.5 1.9 0.3 3.9 12.0
7 5.1 0.5 2.6 0.2 4.4 12.8
8 4.9 0.6 0.9 0.6 4.8 11.8
9 6.2 0.4 2.6 0.6 3.7 13.5
10 6.5 0.2 3.8 0.5 3.3 14.4
Each dawn, the huts were searched and all mosquitoes were collected from the floors, walls, and ceilings of rooms, verandahs and nets. Mosquitoes were identified and scored as blood-fed or unfed and dead or alive. Male mosquitoes were not recorded. Live females were held in netted plastic cups and supplied with 10% honey solution for 24 h before recording any delayed mortality. All A. gambiae and C. quinquefasciatus were kept for determination of kdr and Ace.1R genotypes.
DNA diagnostic test for the pyrethroid resistance kdr and insensitive acetylcholinesterase G119S mutations in single A. gambiae
Genomic DNA was extracted from single mosquitoes according to Collins et al [30]. For determination of acetylcholinesterase mutation, the DNA was PCR amplified with the degenerate primers Moustdir1 5'CCGGGNGCSACYAT-GTGGAA3' and Moustrev1 5'ACGATMACGTTCTCYTCCGA3' for thirty cycles (94°C for 30 seconds, 52°C for 30 seconds and 72°C for 1 minute), the PCR fragments were then digested with AluI restriction enzyme according to the manufacturer's instructions and fractionated on a 2% agarose gel according to Weil et al [31]. Genotypes for pyrethroid resistance kdr "leu-phe" mutation were determined according to Martinez-Torres et al [9].
Data analysis
The effect of each treatment was assessed relative to the control in terms of deterrency (the number of mosquitoes caught in each hut), excito-repellency (the proportion of mosquitoes in the verandah traps), blood feeding inhibition and mortality rates. Proportional data were analysed using logistic regression (STATA 6 software). Comparisons between treatments were made by successively dropping treatments from the overall comparison and this process allowed each treatment to be compared with every other. Owing to non-normality of the data the numbers of blood-fed and dead mosquitoes and overall totals collected from each hut were compared using Wilcoxon rank sum non-parametric tests. Genotype frequencies were tested using χ2 or Fisher's exact test.
Results
Mosquito abundance
To assess any difference in the attractiveness of the huts to mosquitoes, preliminary collections were carried out over 100 hut-nights (10 huts × 10 nights) from late July 2002, with the sleepers/collectors being rotated between huts on successive nights (the 11th hut was being used for another purpose at this time). A total of 1328 mosquitoes were recorded of which 41.6% were A. gambiae, 19.4% Culex spp., 31.6% Mansonia spp., 4.3% Aedes spp., 2% A. funestus, 0.9% Anopheles pharaonsis, 0.5% Anopheles coustani, 0.2% Coquillettidia crystata and 0.1% Eretmapodites. There were no significant differences between the huts in the numbers of each species collected (F = 0.411 df = 9,9 P = 0.926) (Table 1).
Efficacy of treatment
From 363 hut-nights collections (33 days × 11 huts) a total of 5,639 mosquitoes were recorded of which 10% were A. gambiae and 45% Culex spp. (C. quinquefasciatus predominated). The markedly lower A. gambiae frequency than that mentioned in the previous paragraph was presumably due to a seasonal change. Of the remaining species 23% were A. coustani, 35% Mansonia spp., 6% Aedes spp., 0.4% A. funestus, and 1% A. pharaoensis with Coquillettidia crystata and Eretmapodites being present in small numbers. Only the malaria vector A. gambiae and the nuisance mosquito Culex spp. were subjected to analysis. Additional file 1 and Additional file 2 summarize the data for A. gambiae and Culex spp., respectively. All types of insecticide treatment appeared to deter entry of mosquitoes into huts, compared to that of untreated nets. Estimated deterrency among A. gambiae ranged from 30% with the low dosage mixture to 71% with lambdacyhothrin. However, the differences between treatments never reached statistical significance among the relatively small numbers of A. gambiae caught. Percentage deterrency among Culex spp. was less than the estimates for A. gambiae, but was more often statistically significant among the larger numbers caught. Estimated deterrency of Culex spp. ranged from 36% with the washed chlorpyrifos-methyl treated net to 58% with the unwashed lambdacyhalothrin net but the differences between unwashed and washed nets were not significant.
Additional file 1 and Additional file 2 show data on mosquito feeding and mortality expressed either as mean numbers per daily collection (cols. 9 and 12) or as percentages among those that entered the huts (cols. 6–8 and 10–11).
Untreated nets failed to give protection when holed: 39% of A. gambiae and 25% of Culex spp. were blood-fed, whereas with the intact net only 7% of A. gambiae and 2% of Culex were blood-fed. Treatment with chlorpyrifos-methyl or lambdacyhalothrin restored the capacity of holed nets to inhibit blood feeding. After five washes of the lambdacyhalothrin (18 mg/m2) or chlorpyrifos-methyl (250 mg/m2) treated nets, the inhibition of blood feeding shown by in A. gambiae was no longer statistically significant. Among the Culex, on the other hand, blood feeding remained significantly inhibited after washing, both for lambdacyhalothrin and chlorpyrifos-methyl treated nets. The proportion of A. gambiae and Culex killed by lambdacyhalothrin or chlorpyrifos-methyl treatments were similar before and after washing, indicating that the treatments remained highly insecticidal even after 5 washes.
The treatments deterred A. gambiae from entering the holed nets. The proportions collected from inside the nets were 20.5% for the untreated nets, 10.0% for the chlorpyrifos-methyl treated nets and 9.0% for the lambdacyhalothrin treated nets (P < 0.01). No such effect was observed against Culex, the proportion found inside the nets showing little variation (8–9%) between the untreated nets and nets treated with chlorpyrifos-methyl or lambdacyhalothrin.
The mortality was consistently high among A. gambiae and Culex regardless of the treatment used. An application rate of 100 mg/m2 chlorpyrifos-methyl was sufficient to kill 58% of A. gambiae and 68% of Culex (despite the fact that many of the mosquitoes carried genes for Ace.1R), and reduced the rate of blood feeding through the holed nets by 62% among A. gambiae and 88% among Culex. An increase in the application rate of chlorpyrifos-methyl from 100 to 250 mg/m2 did not significantly increase the mortality rates of A. gambiae and Culex, but with the higher dose there was complete inhibition of blood feeding of A. gambiae. Most of the mortality was evident by dawn and for most types of treatment less than 10% of deaths occurred over the next 24 h. The two-in-one and the high and low dose mixtures showed equivalent efficacy. When compared to each insecticide applied as single treatments the two-in-one and high dose mixture did not appear to give either better protection against blood feeding or higher mosquito mortality. The low dose mixture was associated with mortality and blood-feeding rates similar to that observed with the high dose mixture and two-in-one treatments.
All treatments increased the proportion of mosquitoes in the verandah traps. The magnitude of the excito-repellent effect ranged from 16% to 40%. By this criterion, lamdacyhalothrin and chlorpyrifos-methyl were equally repellent, and both Culex and A. gambiae showed a similar tendency to be repelled. With the greater numbers of mosquitoes collected, the "insecticide-induced exophily" tended to be significant for Culex, but it was not significant with the small A. gambiae samples.
Side-effects of treatments
In 154 interviews with sleepers recorded during the first two weeks of the trial, 14 (9.1%) recorded sneezing and 5 five recorded severe headaches. Such symptoms occurred during the first 10 days after treating the nets. The complaints of headache were made, in every instance, after a night spent under the unwashed net treated with 250 mg/m2 chlorpyrifos-methyl.
Genotype frequencies
The frequency of the kdr gene in A. gambiae from the control huts was above 90%. With this high frequency the sample sizes were considered too low to test satisfactorily for any selective effect by treatments, and, therefore, the collections from the treatment huts were not tested any further for kdr. The high mortality obtained with the lambdacyhalothrin treated nets indicates that this pyrethroid, when present on nets, is still effective in controlling many kdr resistant mosquitoes.
The results of Ace.1 genotyping of 298 A. gambiae are grouped into those found live or dead in the huts with untreated control nets, or with nets treated with chlorpyrifos-methyl, lambdacyhalothrin or combinations of these two insecticides (Table 2). The numbers identified were two mosquitoes with Ace.1R only, 243 mosquitoes with Ace.1R plus Ace.1S and 53 mosquitoes with Ace.1S only. Assuming that the first type were homozygous for Ace.1R, the third type were homozygous for Ace.1S and the second type were heterozygous, the frequencies of the Ace.1R and Ace.1S alleles would be 0.41 and 0.59 respectively. According to conventional Hardy-Weinberg ratios (P2, 2PQ, Q2) the frequency of Ace.1R/Ace.1R genotype was expected to be 17.2% rather than the 0.7% actually observed. This indicates a strong fitness disadvantage associated with Ace.1R homozygotes or some alternative genetic explanation.
Table 2 Acetylcholinesterase genotypes as determined by PCR on Anopheles gambiae collected from the experimental huts.
Type of Treatment Live/dead at collection Ace1R/Ace1R No. Ace1R/Ace1S Ace1S/Ace1S
No. % survival No. % survival
Untreated Control live 0 27 64%a 9 100%b
dead 0 15 0
total 0 42 9
Chlorpyrifos- methyl live 1 23 26%c 11 46%c
dead 1 65 13
total 2 88 24
Lambda- cyhalothrin live 0 16 34%c 2 14%c
dead 0 31 12
total 0 47 14
Mixture & two-in-one live 0 21 47%c 1 17%c
dead 0 45 5
total 0 66 6
Note percentages accompanied by different letter superscripts differ significantly by χ2 or Fisher's exact test
Survival rates of Ace.1R/Ace.1S in huts with untreated nets were lower than that of Ace.1S/Ace.1S suggesting a fitness disadvantage. There was no apparent difference between the survival rates of Ace.1R/Ace.1S and Ace.1S/Ace.1S in huts with chlorpyrifos-methyl treated nets or other treatments.
Discussion
Where ITN-induced deterrency occurs, it will contribute to reducing human/vector contact, and in households where not everyone has access to a net, even non-users would be expected to gain some indirect protection from mosquito biting. However, deterrency will reduce the number of mosquitoes killed and, hence, the potential for a community-wide impact on the infective biting population. Thus, the mean numbers found fed and dead in the huts are more realistic indicators of the relative impact of the treated nets than are the columns showing the proportions fed and dead. However, the latter are also of interest in giving an idea of the extent of the feeding inhibition and insecticidal effects of the various treatments on mosquitoes exposed to them.
Nets treated with chlorpyrifos-methyl appeared to induce a level of deterrency against A. gambiae similar to that of the pyrethroid lambdacyhalothrin but the effects were not statistically significant. Compared to other hut studies in Côte d'Ivoire, the deterrency shown by chlorpyrifos-methyl against C. quinquefasciatus was greater than that shown by Olyset nets [12] or by nets treated with permethrin 500 mg/m2 [32] or carbosulfan 200 and 300 mg/m2 [13,14,25]. The deterrency of chlorpyrifos-methyl was still evidence presently after five washes. This contrasts with another type of OP, pirimiphos -methyl which was undetectable on nets after only 3 three washes [33]. The difference partly lies with the formulations used. The microencapsulated formulation was specially developed to bind chlorpyrifos-methyl more strongly and to release it more gradually compared to the emulsifiable concentrate used with pirimiphos -methyl. The manufacturer of chlorpyrifos-methyl (Dow AgroSciences) has recently improved the release characteristics and wash-fastness of the microencapsulation so that insecticide persistence and performance compares favourably with pyrethroids and DDT (Rowland & Yates, unpublished). The microencapsulated formulation used for lambdacyhalothrin has been shown earlier to withstand at least 5 washes [21,34] and this was re-confirmed in the present trials.
Only a small proportion of mosquitoes in the hut with the intact, untreated net had blood-fed. The protective effect of untreated nets was lost when they were holed. This re-confirms the results in experimental huts of Lines et al [35] and Curtis et al [20] and data on malaria in children sleeping under untreated nets which were either intact or torn [36]. It emphasizes the point that ITN programmes which fail to ensure that nets are and remain effectively insecticidal cannot expect to achieve an impact on malaria when the nets become torn, as they inevitably will.
Treatment with chlorpyrifos-methyl or lambdacyalothrin restored the protectiveness of torn nets against Culex and A. gambiae, as has been demonstrated with various pyrethroid treatments [20,32,35,37].
Whether or not A. gambiae feeds through chlorpyrifos-methyl treated nets seems to depend upon the concentration of insecticide used, the observed blood feeding rate being zero for the net treated with 250 mg/m2 and 14.6% with 100 mg/m2. Further tests with intermediate concentrations would be needed to confirm a dosage trend. Mortality also appeared to be dosage-dependent, but the difference between the doses was non-significant among the small numbers collected.
Adverse effects involving headache and sneezing were associated with exposure to nets recently treated with 250 mg/m2 chlorpyrifos-methyl. Such effects were not apparent with nets treated with 100 mg/m2 chlorpyrifos-methyl or with washed nets or with nets used two weeks after treatment. Symptoms appeared to be dosage-dependent. With lambdacyhalothrin treated nets, sneezing was reported with a 30 mg/m2 dosage [38] and 20 mg/m2 [21] but not at 10 mg/m2 [38] or with the dosage of lambdacyhalothrin used in the present trial (18 mg/m2). While the toxicological profile of chlorpyrifos-methyl is favourable, a more comprehensive hazard assessment is warranted before considering community trials [39,40]. To avoid the side effects which were reported at 250 mg/m2, it is proposed that a dose of 100 mg/m2 should not be exceeded.
Resistance due to insensitive acetylcholinesterase has arisen independently through a point mutation to G119S (glycine to serine substitution) of the Ace.1 gene on several occasions in A. gambiae, Anopheles albimanus, C. quinquefaciatus and Culex pipiens [31]. This study confirms the finding of Weill et al [31] on the same population of A. gambiae of there being far fewer homozygotes for the Ace.1R than would be expected from the Hardy Weinberg ratio, indicating extremely low viability of these homozygotes. In the huts with untreated nets, the observed survival of heterozygotes for this gene was significantly less than that of homozygotes for Ace.1S. With such strong selection pressure against Ace.1R at the adult stage in homozygotes and heterozygotes, this gene could apparently only persist in the population if there is very strong selection against Ace.1S in the immature stages. A more likely explanation is one of gene duplication having occurred at the acetylcholinesterase locus. Such duplication of the Ace.1 locus is common in C. pipiens [31], and the existence of such duplication in A. gambiae could result in Ace.1S and Ace.1R being present at diffent loci on the same chromosome. Insects with one locus for Ace.1S and one for Ace.1R could be included among those labelled Ace.1R/Ace.1S in table 4. The rare individuals labelled Ace.1R/Ace.1R in Table 2 would be those where both Ace.1 loci and all alleles were of the R type.
The survival of all genotypes was significantly lower in huts with treated nets than with untreated nets, and the survival of mosquitoes with at least one Ace.1R allele and those with only Ace.1S did not differ significantly on any of the treatments. Thus under the realistic conditions of the experimental huts the Ace.1R allele does not give effective resistance to chlorpyrifos-methyl.
The present study is the first experimental hut trial of chlorpyrifos-methyl on nets. Other recent trials with non-pyrethroids have focused on carbosulfan which, like chlorpyrifos-methyl, is an acetylcholinesterase inhibitor known to be effective against kdr resistant A. gambiae and C. quinquefasciatus [13-15,25]. Quite apart from its potentially toxic breakdown product, carbofuran [25], carbosulfan has the disadvantage of poorer wash fastness compared to alphacyano-pyrethroids [14]. The rate of conversion of carbosulfan to carbofuran and the hazard this presents to net users has yet to be established, and the inferior wash fastness might conceivably be improved through formulation technology. But in view of these uncertainties, chlorpyrifos-methyl seems the better prospect at the present time. Chlorpyrifos-methyl should be subjected to a full hazard assessment before being used on a community scale [40].
There are several potential advantages to using combinations of insecticide on nets: a) a reduced or delayed selection of resistance alleles, b) an improved control of resistant populations, c) an improved efficacy if the two components are synergistic, with possible cost savings and improved safety if the dosage can be substantially reduced as a result. To delay the selection of resistance with a mixture would require no cross-resistance between the two components and 100% mortality of insects carrying resistance to one of the components on being exposed simultaneously to the other component [23,41]. The situation in the Bouaké region of Côte d'Ivoire is not favourable to classical resistance management of genes that have an impact on chemical control, because kdr and Ace.1R are already present at high frequency. Elsewhere in Africa, where neither mechanism is present a strategy of using mixtures of lambdacyhalothrin and chlorpyrifos-methyl might have potential for resistance management. There is some evidence in laboratory studies for synergy between pyrethroids and OPs/carbamates against susceptible A. gambiae adults [42] and resistant C. quinquefasciatus larvae bearing kdr and elevated oxidase mechanisms [43], but not against adults with site-insensitivity resistance mechanisms [44]. The encouragingly high efficacy of the low dosage mixture could be due to synergism, but to test this hypothesis a comparison should be made in experimental huts with the low dosages applied individually to nets.
Both chlorpyrifos-methyl and lambdacyhalothrin treated nets and combinations of these insecticides were shown in the realistic conditions of experimental huts to be equally effective in preventing blood feeding by, and killing of, A. gambiae and C. quinquefasciatus, which laboratory and molecular studies suggested would have shown pyrethroid and organophosphate resistance. Thus, these resistance genes may have little or no practical significance. To demonstrate this unequivocally, it is desirable to examine the impact of mixtures and single insecticide treatments on both malaria transmission and the selection of resistance alleles in an area where resistance to OPs and pyrethroids in A. gambiae and C. quinquefasciatus exist at low-to-intermediate frequencies. Chlorpyrifos-methyl should also be tested on nets and as an indoor residual spray treatment against populations of pyrethroid resistant C. quinquefasciatus (in Tanzania) and A. funestus (in South Africa), respectively, that have demonstrated practical impact in the field [6,21]. Chlorpyrifos-methyl could be useful for managing such problematic examples of resistance if it could be shown that these pyrethroid resistance mechanisms confer no cross- resistance to organophosphates.
Authors' contributions
ANA & RNG conducted the field work, summarizsed the data, and wrote the first report.
AAK, FC, VC conducted or supervised the genotyping and interpreted the results.
MWR designed the study, supervised RNG and ANA, analysed the data and wrote the manuscript.
CFC co-designed the study, co-supervised ANA and contributed to the manuscript.
JMH co-designed the study and supervised FC and VC and contributed to the manuscript.
FD designed the experimental huts.
MZ established the collaboration with Dow AgroSciences, coordinated the evaluation of chlorpyrifos-methyl, and contributed to study design and writing of the manuscript.
Supplementary Material
Additional File 1
Summary data of Anopheles gambiae collected from experimental huts over 33 nights at Yaokoffikro.
Click here for file
Additional File 2
Summary data of Culex spp. collected from experimental huts over 33 nights at Yaokoffikro.
Click here for file
Acknowledgements
We wish to thank all the staff at the Institut Pierre Richet, Bouaké, Côte d'Ivoire for their hard work during the trial and for their continuing commitment despite extremely difficult conditions since the outbreak of civil war. We also thank Dr Driss Kelili of Dow AgroSciences for providing the formulation of chlorpyrifos-methyl and for his continuing encouragement, and Dr Alan Buckle of Syngenta for the formulation of lambdacyhalothrin. This study was funded by the Gates Malaria Partnership. Mention of specific companies or products does not imply that they are recommended or endorsed by WHO.
==== Refs
Zaim M Aitio A Nakashima N Safety of pyrethroid-treated nets Med Vet Entomol 2000 14 1 5 10759305 10.1046/j.1365-2915.2000.00211.x
Elissa N Mouchet J Rivière F Meunier JY Yao K Resistance of Anopheles gambiae s.s to pyrethroids in Côte d'Ivoire Ann Soc Belge Méd Trop 1993 73 291 294
Chandre F Darriet F Manguin S Brengues C Carnevale P Guillet P Pyrethroid cross resistance spectrum among populations of Anopheles gambiae s.s. from Côte d'Ivoire J Am Mosq Control Assoc 1999 15 53 59 10342269
Chandre F Darriet F Manga L Akogbeto M Faye O Mouchet J Guillet P Status of pyrethroid resistance in Anopheles gambiae sensu lato Bull World Health Organ 1999 77 230 234 10212513
Ranson H Jensen B Vulule JM Wang X Hemingway J Collins FH Identification of a point mutation in the voltage-gated sodium channel gene of Kenyan Anopheles gambiae associated with resistance to DDT and pyrethroids Insect Mol Biol 2000 9 491 497 11029667 10.1046/j.1365-2583.2000.00209.x
Hargreaves K Koekemoer LL Brooke B Hunt RH Mthembu J Coetzee M Anopheles funestus resistant to pyrethroid insecticides in South Africa Med Vet Entomol 2000 14 181 189 10872862 10.1046/j.1365-2915.2000.00234.x
Chandre F Darriet F Darder M Cuany A Doannio JMC Pasteur N Guillet P Pyrethroid resistance in Culex quinquefasciatus from West Africa Med Vet Entomol 1998 12 359 366 9824819 10.1046/j.1365-2915.1998.00120.x
Khayrandish A Wood RJ A multiple basis for insecticide resistance in a strain of Culex quinquefasciatus (Diptera, Culicidae) from Muheza, Tanzania, studied as resistance declined Bull Ent Res 1993 83 75 86
Martinez-Torres DF Chandre F Williamson MS Darriet F Bergé JB Devonshire AL Guillet P Pasteur N Pauron D Molecular characterization of pyrethroid knockdown resistance (kdr) in the major malaria vector Anopheles gambiae s.s Insect Mol Biol 1998 7 179 184 9535162 10.1046/j.1365-2583.1998.72062.x
Henry MC Doannio JMC Darriet F Nzeyimana I Carnevale P Efficacité des moustiquaires pré-imprégnées de perméthrine Olyset Net® en zone de résistance des vecteurs aux pyréthrinoides. II-Evaluation parasitoclinique Méd Trop 1999 59 355 357
Darriet F N'Guessan R Koffi AA Konan L Doannio JMC Chandre F Carnevale P Impact de la résistance aux pyréthrinoïdes sur l'efficacité des moustiquaires imprégnées dans la prévention du paludisme: résultats des essais en cases expérimentales avec la deltaméthrine SC Bull Soc Path Exot 2000 93 131 134 10863621
N'Guessan R Darriet F Doannio JMC Chandre F Carnevale P Olyset Net® efficacy against pyrethroid-resistant Anopheles gambiae and Culex quinquefasciatus after 3 years' field use in Côte d'Ivoire Med Vet Entomol 2001 15 97 104 11297108 10.1046/j.1365-2915.2001.00284.x
Hougard JM Corbel V N'Guessan R Darriet F Chandre F Akogbéto M Baldet T Guillet P Carnevale P Traoré-Lamizana M Efficacy of mosquito nets treated with insecticide mixtures against insecticide resistant Anopheles gambiae and Culex quinquefasciatus (Diptera: Culicidae) in Côte d'Ivoire Bull Ent Res 2003 93 491 498 10.1079/BER2003261
Asidi AN N'Guessan R Hutchinson RA Traoré-Lamizana M Carnevale P Curtis CF Experimental hut comparisons of nets treated with carbamate or pyrethroid insecticides, washed or unwashed, against pyrethroid-resistant mosquitoes Med Vet Entomol 2004 18 134 140 15189238 10.1111/j.0269-283X.2004.00485.x
Kolaczinski JH Fanello C Herve JP Conway DJ Carnevale P Curtis CF Experimental and molecular genetic analysis of the impact of pyrethroid and non-pyrethroid insecticide impregnated bednets for mosquito control in an area of pyrethroid resistance Bull Ent Res 2000 90 125 132
Phillips-Howard PA Nahlen BL Kolzac MS Hightower AW ter Kuile FO Alaii JA Gimnig JE Arudo J Vulule JM Odhacha A Kachur SP Schoute E Rosen DH Sexton JD Oloo AJ Hawley WA Efficacy of permethrin-treated bed nets on prevention of morbidity in young children in an area of high perennial malaria transmission in western Kenya Am J Trop Med Hyg 2003 68 23 29 12749482
Hawley WA Phillips-Howard PA ter Kuile FO Terlouw DJ Vulule JM Ombok M Nahlen BL Gimnig JE Kariuki SK Kolzac MS Hightower AW Community-wide effects of permethrin-treated bed nets on child mortality and malaria morbidity in western Kenya Am J Trop Med Hyg 2003 68 121 127 12749495
Malcom CA Current status of pyrethroid resistance on anophelines Parasitol Today 1988 4 S12 S14
Denholm I Rowland MW Tactics for managing resistance in arthopods: theory and practice Ann Rev Entomol 1992 37 91 112 1539942 10.1146/annurev.en.37.010192.000515
Curtis CF Myamba J Wilkes TJ Comparison of different insecticides and fabrics for anti-mosquito bednets and curtains Med Vet Entomol 1996 10 1 11 8834736
Maxwell CA Myamba J Njunwa KJ Greenwood BM Curtis CF Comparison of bednets impregnated with different pyrethroids for their impact on mosquitoes and on re-infection with malaria after clearance of pre-existing infections with chlorproguanil-dapsone Trans R Soc Trop Med Hyg 1999 93 4 11 10492776 10.1016/S0035-9203(99)90158-X
Zaim M Guillet P Alternative insecticides an urgent need Trends Parasitol 2002 18 161 163 11998703 10.1016/S1471-4922(01)02220-6
Curtis CF Theoretical models of the use of insecticide mixtures for management of resistance Bull Ent Res 1985 75 259 265
Tabashnik BE Managing resistance with multiple pesticide tactics: theory, evidence and recommendations J Econ Entomol 1989 82 1263 1269 2689487
Guillet P N'Guessan R Darriet F Traoré-Lamizana M Chandre F Carnevale P Combined pyrethroid and carbamate «two in one» treatedmosquito nets: field efficacy against pyrethroid-resistant Anopheles gambiae and Culex quinquefasciatus Med Vet Entomol 2001 15 105 112 11297094 10.1046/j.1365-2915.2001.00288.x
WHO The WHO recommended classification of pesticides by hazard and guidelines to classification 2002 Geneva, World Health Organization 57
Chandre F Darriet F Doannio JMC Riviere F Pasteur N Guillet P Distribution of organophosphate and carbamate resistance in C ulex pipiens quinquefasciatus (Diptera: Culicidae) from West Africa J Med Entomol 1997 34 664 671 9439121
N'Guessan R Darriet F Guillet P Carnevale P Lamizana MT Corbel V Koffi AA Chandre F Resistance to carbosulfan in field populations of Anopheles gambiae from Côte d'Ivoire based on reduced sensitivity of acetylcholinesterase Med Vet Entomol 2003 17 19 25 12680920 10.1046/j.1365-2915.2003.00406.x
Pleass RJ Armstrong JRM Curtis CF Jawara M Lindsay SW Comparison of permethrin treatments for bednets in The Gambia Bull Ent Res 1993 83 133 140
Collins FH Finnerty V Petrarca V Ribosomal DNA probes differentiate five cryptic species in the Anopheles gambiae complex Parasitology 1988 30 231 240
Weill M Malcolm C Chandre F Mogensen K Berthomieu A Marquine M Raymond M The unique mutation in ace-1 giving high insecticide resistance is easily detectable in mosquito vectors Insect Mol Biol 2004 13 1 7 14728661 10.1111/j.1365-2583.2004.00452.x
Darriet F N'Guessan R Carnevale P Evaluations en cases-pièges de l'effet protecteur de moustiquaires non imprégnées d'insecticides dans la prévention des piqûres d' Anopheles gambiae s.s Cahiers Santé 2000 10 413 417
Miller JE Lindsay SW Armstrong JR Experimental hut trials of bednets impregnated with synthetic pyrethroid or organophosphate insecticide for mosquito control in the Gambia Med Vet Entomol 1991 5 465 467 1685337
Curtis CF Maxwell CA Finch RJ Njunwa KJ A comparison of use of a pyrethroid either for house spraying or for bednet treatment against malaria vectors Trop Med Int Health 1998 3 619 631 9735932
Lines JD Myamba J Curtis CF Experimental hut trials of permethrin-impregnated mosquito nets and eave curtains against malaria vectors in Tanzania Med Vet Entomol 1987 1 37 51 2979519
Mwangi TW Ross A Marsh K Snow RW The effects of untreated bednets on malaria infection and morbidity on the Kenyan coast Trans R Soc Trop Med Hyg 2003 97 369 372 15259458 10.1016/S0035-9203(03)90056-3
Carnevale P Bitsindou P Diomandé L Robert V Insecticide impregnation can restore the efficiency of torn bednets and reduce man-vector contact in malaria endemic areas Trans R Soc Trop Med Hyg 1992 86 362 364 1440802 10.1016/0035-9203(92)90219-3
Njunwa KJ Lines JD Magesa SM Mnzava AEP Wilkes TJ Alilio M Kivumbi K Curtis CF Trial of pyrethroid impregnated bednets in an area of Tanzania holoendemic for malaria. Part 1. Operational methods and acceptability Acta Trop 1991 49 87 96 1680283 10.1016/0001-706X(91)90056-P
Tomlin CDS The Pesticide Manual, a World Compendium 2003 12 (British Crop Protection Council), London, United Kingdom
WHO A generic risk assessment model for insecticide treatment and subsequent use of mosquito nets 2004 WHO/CDS/WHOPES/GCDPP/2004.6.
Mani GS Evolution of resistance in the presence of two insecticides Genetics 1985 109 761 783 3988040
Corbel V Darriet F Chandre F Hougard JM Insecticide mixtures for mosquito net impregnation against malaria vectors Parasite 2002 9 255 259 12375369
Corbel V Raymond M Chandre F Darriet F Hougard JM Efficacy of insecticide mixtures against larvae of Culex quinquefasciatus Say (Diptera: Culicidae) resistant to pyrethroids and carbamates Pest Management Science 2004 60 375 380 15119600 10.1002/ps.809
Darriet F Corbel V Hougard JM Efficacy of mosquito nets treated with a pyrethroid-organophosphorous mixture against kdr- and kdr+ malaria vectors (Anopheles gambiae) Parasite 2003 10 359 362 14710633
| 15918909 | PMC1156935 | CC BY | 2021-01-04 16:24:13 | no | Malar J. 2005 May 26; 4:25 | utf-8 | Malar J | 2,005 | 10.1186/1475-2875-4-25 | oa_comm |
==== Front
Microb Cell FactMicrobial Cell Factories1475-2859BioMed Central London 1475-2859-4-141590451810.1186/1475-2859-4-14ReviewImprovement of Escherichia coli production strains by modification of the phosphoenolpyruvate:sugar phosphotransferase system Gosset Guillermo [email protected] Departamento de Ingeniería Celular y Biocatálisis, Instituto de Biotecnología, Universidad Nacional Autónoma de México, Apdo. Postal 510-3, Cuernavaca, Mor. 62250, México2005 16 5 2005 4 14 14 12 4 2005 16 5 2005 Copyright © 2005 Gosset; licensee BioMed Central Ltd.2005Gosset; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The application of metabolic engineering in Escherichia coli has resulted in the generation of strains with the capacity to produce metabolites of commercial interest. Biotechnological processes with these engineered strains frequently employ culture media containing glucose as the carbon and energy source. In E. coli, the phosphoenolpyruvate:sugar phosphotransferase system (PTS) transports glucose when this sugar is present at concentrations like those used in production fermentations. This protein system is involved in phosphoenolpyruvate-dependent sugar transport, therefore, its activity has an important impact on carbon flux distribution in the phosphoenolpyruvate and pyruvate nodes. Furthermore, PTS has a very important role in carbon catabolite repression. The properties of PTS impose metabolic and regulatory constraints that can hinder strain productivity. For this reason, PTS has been a target for modification with the purpose of strain improvement. In this review, PTS characteristics most relevant to strain performance and the different strategies of PTS modification for strain improvement are discussed. Functional replacement of PTS by alternative phosphoenolpyruvate-independent uptake and phosphorylation activities has resulted in significant improvements in product yield from glucose and productivity for several classes of metabolites. In addition, inactivation of PTS components has been applied successfully as a strategy to abolish carbon catabolite repression, resulting in E. coli strains that use more efficiently sugar mixtures, such as those obtained from lignocellulosic hydrolysates.
==== Body
Review
Metabolic engineering can be defined as the purposeful modification of cellular activities with the aim of strain improvement [1]. Development of microbial strains for the production of metabolites is based primarily on the application of recombinant DNA technology to alter the properties of the metabolic network by modifying the level of activity or the properties of specific enzymes. These principles have been applied to the generation of a large number of Escherichia coli strains, designed for the production of commercially important compounds [2]. Cultures with these engineered strains usually employ media containing glucose. This sugar is nowadays the most utilized raw material in industrial fermentations with E. coli, mostly because it is relatively inexpensive and it is the preferred carbon and energy source for this bacterium. As a component of culture media, glucose provides carbon atoms for biomass and product generation. The cell's capacity to uptake and metabolize this carbohydrate has a profound impact on its growth rate and productivity. Thus, it can be expected that modifications to glucose transport systems should have and important impact on the cell's physiology and this, in turn, can either improve or become detrimental in an industrial production context. The purpose of this review is to summarize the characteristics of glucose uptake systems in E. coli and discuss examples where their modification in wild type or engineered production strains has resulted in improved performance.
Glucose transport systems in Escherichia coli
A distinctive feature of E. coli and other gram-negative bacteria is the presence of two concentric membranes surrounding its cytoplasm. The space between these two membranes is the periplasm (Fig. 1). The outer and cytoplasmic membranes constitute a hydrophobic barrier to polar compounds. To control the inward and outward flow of molecules across these barriers, the bacterial cell synthesizes proteins that form channels. The outer membrane constitutes the first barrier to the entry of carbohydrates, E. coli contains about 105 channels formed by specialized proteins called porins [3]. The proteins OmpC and OmpF are the most abundant porins present under typical laboratory growth conditions, representing up to 2% of the total cellular protein [4]. Their relative abundance is influenced by various factors such as medium osmolarity, temperature and growth phase [5-7]. It has been shown that these two porins constitute the main entry channels for glucose into the periplasm when this sugar is present at a concentration higher than 0.2 mM [8,9]. In experiments with reconstituted liposomes, the diffusion rate for glucose was found to be about twofold higher through OmpF than through the OmpC channel [10]. Under conditions of glucose limitation, synthesis of the outer membrane glycoporin LamB is induced [9]. This protein permeates several carbohydrates including maltose, maltodextrins and glucose [11]. It has been demonstrated that under external submicromolar concentrations of glucose, LamB contributes about 70% of the total glucose import capacity of the cell [9].
Figure 1 The phosphoenolpyruvate:sugar phosphotransferase system and other glucose transport systems in Escherichia coli.
Glucose diffusion by porins through the outer membrane is a passive process. The driving force for glucose internalization into the periplasm is its active transport into the cytoplasm. Due to the presence of active transport systems in the cytoplasmic membrane, it can be assumed that glucose concentration in the periplasm is very low. Once inside the periplasm, glucose can be internalized into the cytoplasm by the phosphoenolpyruvate:sugar phosphotransferase system (PTS). This protein system belongs to the group translocator family of transporters, which are widespread in bacteria and absent in Archaea and eukaryotic organisms [12,13]. PTS participates in the transport and phosphorylation of several sugars. The system is composed of the soluble and non sugar-specific protein components Enzyme I (EI) and the phosphohistidine carrier protein (HPr), encoded by genes ptsHI (Fig. 1). These proteins relay a phosphoryl group from PEP to the sugar-specific enzymes IIA and IIB. The last component of this system, IIC (in some cases also IID), is an integral membrane protein permease that recognizes and transports the sugar molecules, which are phosphorylated by component IIB. There are 21 different identified enzyme II complexes encoded in the E. coli chromosome, that are involved in the transport of about 20 different carbohydrates [14]. In E. coli, the enzyme II complexes IIGlc and IIMan are involved in glucose import. The glucose-specific IIGlc complex is composed of the soluble IIAGlc enzyme and the integral membrane permease IICBGlc, encoded by genes crr and ptsG, respectively. Reported Km and Vmax values for IIGlc with glucose as substrate are in the range of 3–10 μM and 126 μmol min-1 g-1, respectively [15,16]. The mannose IIMan complex is composed of the IIABMan homodimer enzyme and the integral membrane permease IICDMan. These proteins are encoded in the manXYZ operon. In addition to mannose, these proteins can also transport with similar efficiently glucose, fructose, N-acetylglucosamine and glucosamine [17]. Reported Km and Vmax values for IIMan with glucose as substrate are 15 μM and 72 μmol min-1 g-1, respectively [16]. In a wild type strain growing with glucose as the carbon source, ptsG is induced and the manXYZ operon is repressed. Inactivation of the genes encoding the IIGlc complex abolishes repression of manXYZ and the resulting mutant now transports glucose with the IIMan complex, displaying a growth rate corresponding to approximately 84% of that observed in a wild type strain [18].
Glucose can also be actively transported into the cytoplasm by systems that are normally involved in galactose internalization. The genes coding for these transporters and for enzymes involved in galactose metabolism are normally induced by the presence of galactose in the growth medium. However, it has been demonstrated that under growth conditions where external glucose concentration is lower than 1 μM, the genes encoding these systems are maximally induced in the absence of galactose [19,20]. A similar response has been observed in an E. coli strain lacking the PTS and growing in a medium with a relatively high glucose concentration (2 g/l) [21]. Analysis of strains growing in glucose-limited conditions revealed that induction of these genes is caused by the intracellular synthesis of galactose that functions as an autoinducer of the system [19]. One of the genes induced under conditions of glucose limitation is galP, that encodes the low affinity galactose:H+ symporter GalP. It constitutes an integral membrane protein belonging to the major facilitator superfamily (MFS) of the electrochemical potential-driven class of transporters. The GalP protein can import glucose with a Km of 10.2 μM and a Vmax of 15.6 μmol min-1 g-1 [22]. The genes in the mglABC operon encode an ATP-binding protein, a galactose/glucose periplasmic binding protein and an integral membrane transporter protein, respectively. These proteins constitute the Mgl system, that is also involved in galactose/glucose (methyl galactoside) import. This high affinity porter belongs to the ATP-binding cassette (ABC) superfamily of the primary active class of transporters. The Mgl proteins together with LamB constitute a high-affinity glucose transport system that is induced when this sugar is present at a very low concentration [19]. Glucose internalized by GalP or the Mgl systems must be phosphorylated to enter the Embden-Meyerhof-Parnas (EMP) glycolytic pathway. The enzyme glucokinase, encoded by glk, catalyzes the ATP-dependent phosphorylation of glucose in the cytoplasm [23]. Glucokinase activity in E. coli is not essential when glucose is abundant and it is transported by PTS. However, in conditions of glucose limitation or in a mutant lacking PTS, inactivation of glk severely impairs growth capacity [17].
As it can be observed in figure 1 and Table 1, the energetic costs of importing glucose differ significantly among these systems. Since glucose must be phosphorylated in order to enter the EMP glycolytic pathway, a fair comparison must include transport and phosphorylation reactions. It can be observed that PTS is the most efficient system as it consumes one mol of PEP for each mol of internalized and phosphorylated glucose. The energetic equivalent of one mol of PEP is one mol of ATP, since the conversion of PEP into pyruvate (PYR) by pyruvate kinase would yield a mol of ATP by substrate-level phosphorylation. The high-affinity Mgl-glucokinase system is the most expensive energetically, as it consumes two mol of ATP for every mol of glucose that is internalized and phosphorylated. Finally, GalP and glucokinase transports and phosphorylates glucose at the expense of one mol of H+ that is internalized into the cytoplasm and one mol of ATP.
Table 1 Kinetic parameters for glucose transporters and energetic costs for glucose internalization and phosphorylation*.
Transporter Km Vmax Energetic costsa
IIGlc complex 3–10 μM 126 μmol min-1 g-1 1 PEP
IIMan complex 15 μM 72 μmol min-1 g-1 1 PEP
GalP 10.2 μM 15.6 μmol min-1 g-1 1 H+ + 1 ATP
MglABC N.D N.D. 2 ATP
Glf 4.1 mM 75 μmol min-1 g-1 1 ATP
* N.D., not determined.
a The energetic cost equivalent of 1 mol of PEP is 1 mol of ATP.
Laboratory and industrial scale cultures with E. coli employ media containing glucose in a wide range of concentrations. Furthermore, feeding strategies are usually implemented to precisely control glucose concentration through different stages in a production process. However, even under these different scenarios, glucose will almost always be present at a concentration that can be transported efficiently by PTS. Thus, it can be assumed that in industrial production fermentations and most laboratory-scale cultures, glucose uptake and phosphorylation will be dependent almost entirely on the IIGlc PTS complex. Notwithstanding its efficiency as a transport system, PTS imposes several physiological constraints that can hinder some production processes. In the next sections, the characteristics of PTS that can limit productivity will be discussed and strategies to overcome them will be presented.
Increasing PEP metabolic availability for the production of shikimate pathway intermediates and amino acids
By coupling glucose internalization to PEP-dependent phosphorylation, the PTS provides a tight linkage between sugar transport and its subsequent metabolism. Fig. 2 shows central metabolism reactions related to PTS activity. It can be observed that PEP is a link between the EMP glycolytic pathway and PTS, together forming a phosphorylation circuit. In addition to its role as a phosphate donor for PTS, PEP participates in several metabolic reactions. It is a precursor in several biosynthetic pathways and also participates directly in energy-generating reactions such as substrate-level phosphorylation of ADP or indirectly as a precursor of acetyl-CoA. The relative carbon flux originating from the PEP node into the different metabolic pathways has been determined experimentally. When E. coli grows in minimal medium containing glucose as the carbon source, PTS consumes 50% of available PEP, whereas the reactions catalyzed by PEP carboxylase, pyruvate kinases, UDP-N-acetylglucosamine enolpyruvyl transferase and 3-deoxy-D-arabino-heptulosonate 7-phosphate (DAHP) synthase consume approximately 16%, 15%, 16% and 3%, respectively [24-26].
Figure 2 Central pathways related to PTS glucose transport and metabolism.
The shikimate (SHIK) or common aromatic pathway is a source of compounds with commercial applications that include the aromatic amino acids tryptophan, tyrosine and phenylalanine. Carbon flux into the common aromatic pathway starts with the condensation of D-erythrose 4-phosphate (E4P) and PEP to yield DAHP, in a reaction catalyzed by the enzyme DAHP synthase (Fig. 2). It has been determined that carbon flux partitioning at the PEP node is the major determinant of yield for aromatic compounds synthesized from glucose in E. coli [27-30]. Stoichiometric analyses of the metabolic network involved in aromatics synthesis indicates that the maximum theoretical molar yield from glucose could double in a strain that transports glucose without coupling this process to PEP utilization [31,32]. For this reason, several research groups have explored strategies for developing E. coli strains that can uptake glucose without coupling this process to PEP-dependent phosphorylation. Inactivation of the ptsHI-crr operon is the most common approach for generating PTS- strains. Strains with this genetic modification lack the general PTS proteins involved in the phosphotransfer relay to any of the PTS complexes. Therefore, the resulting PTS- strains will exhibit a very limited capacity to transport and phosphorylate glucose (PTS-Glc- phenotype), supported mainly by the GalP, Mgl and glucokinase systems [21]. The severely reduced capacity to transport glucose would be a serious drawback, rendering a PTS- Glc- strain unsuitable for production purposes as it would display very low specific growth rate and productivity. For this reason, different strategies have been investigated for the generation, from PTS- strains, of derivatives that can transport glucose efficiently by a PTS-independent mechanism.
E. coli strains lacking a functional PTS but capable of internalizing glucose at a rate similar to that of a PTS+ strain (PTS- Glc+ phenotype), have been selected for their capacity to grow rapidly using glucose as the sole carbon source in a continuous culture [33]. Characterization of these mutants has revealed that glucose is now transported and phosphorylated by an alternative transport system: galactose permease (GalP) and glucokinase (Glk), that uses ATP as phosphate donor instead of PEP [33,26]. These PTS- Glc+ strains have been modified by metabolic engineering to direct carbon flow to the SHIK pathway and they were compared to equally modified PTS+ strains with regard to yield from glucose in the synthesis of DAHP and phenylalanine. These studies showed that the yield from glucose in the synthesis of DAHP and phenylalanine increased by 65% and 57%, respectively, when compared to isogenic PTS+ strains [32-34].
Generation of PTS- Glc+ strains has also been achieved by expressing from a plasmid genes coding for non PTS-dependent glucose transporters. The first example of this approach was the expression of the genes glfZm and glkZm encoding, respectively, a glucose facilitator (Glf) and glucokinase from Zymomonas mobilis in an E. coli strain lacking functional glucose and mannose PTS complexes and glucokinase [35]. The resulting recombinant strain recovered glucose uptake capacity, increasing its specific growth rate from 0.01 to 0.53 h-1. The Z. mobilis Glf or glucose uniporter belongs to the sugar porter family of the mayor facilitator superfamily. This type of permease does not use energy during transport. Biochemical characterization of Glf revealed that it can transport glucose (Km of 4.1 mM and Vmax of 75 μmol min-1g-1) and fructose (Km of 39 mM and Vmax of 93 μmol min-1 g-1) (see Table 1) [36]. More recently, the effects of different expression levels of native galP and glk on glucose consumption and growth capacity in a PTS-Glc- E. coli strain were studied [37]. Strain VH32 is an E. coli derivative of W3110 having a deletion of the ptsHI-crr operon that displays a specific growth rate of 0.03 h-1 when using glucose as the only carbon source. When this strain was transformed with a plasmid expressing glk, no increase in growth rate was observed. In contrast, expression of galP in VH32 increased specific growth rate to 0.55 h-1 and simultaneous expression of both glk and galP caused VH32 to grow at a specific rate identical to that of the PTS+ parent strain. These results showed that glucose internalization and not phosphorylation is the main limiting factor for rapid growth on glucose for this PTS- Glc- strain. These two approaches for generating PTS- Glc+ strains have been compared with regard to their impact on the yield from glucose in the synthesis of 3-dehydroshikimic acid (DHS) and intermediates from the SHIK pathway in engineered E. coli strains [38]. In fed-batch cultures, a strain transporting glucose by PTS synthesized DHS and SHIK pathway intermediates with 33% (mol/mol) yield from glucose. Under the same culture conditions, PTS- strains expressing glfZm and glkZm or having glucose transport and phosphorylation dependent on high-level expression of native galP and glk, synthesized DHS and SHIK pathway intermediates with 41% and 43% (mol/mol) yield from glucose, respectively.
It can be expected that replacement of PTS by an ATP-dependent glucose transport and phosphorylation system should increase the yield from glucose not only for metabolites derived from the SHIK pathway, but also for some compounds having PEP as a precursor. One example is the synthesis of the amino acid asparagine (ASN). This metabolite is synthesized from aspartate (ASP), whose precursor is PEP-derived oxaloacetate (OAA) (Fig. 2). A recent study based on flux balance analysis of gene knockouts or additions in an E. coli metabolic model, predicted that replacement of PTS activity by an ATP-dependent glucose transport system should increase asparagine yield by 16.5% [39].
Reduction of overflow metabolism: acetate production reduction and increasing recombinant protein production
Acetate is one of the fermentation products generated by E. coli when it grows using glucose in oxygen-limited conditions. Pyruvate formate-lyase catalyzes the conversion of PYR and coenzyme A (CoA) into acetyl-CoA (AcCoA) and formate. The enzyme phosphotransacetylase (Pta) catalyses an acyl transfer reaction to convert AcCoA and Pi into acetyl phosphate (Ac~P) and CoA. Finally, acetate kinase (Ack) catalyzes the substrate-level phosphorylation of ADP to yield ATP and acetate. These series of reactions provide the cell with a source of ATP under conditions where aerobic respiration is not possible [40]. However, it is known that E. coli can also synthesize a significant amount of acetate under aerobic conditions [41,42]. It has been determined that this condition is the result of the combined high rates of glucose uptake by PTS and glucose catabolism by the EMP pathway that result in a rate of AcCOA synthesis surpassing the capacity of the tricarboxylic acid (TCA) cycle to completely consume this metabolite. Part of the excess AcCOA is diverted into the Ack-Pta pathway to generate acetate [43]. The accumulation of acetate in culture media is an important problem in industrial fermentations since this organic acid inhibits cell growth and recombinant protein production [44]. A common approach to reduce acetate accumulation is the application of glucose feeding strategies to limit glucose availability and the genetic modification of central metabolic pathways directly related to acetate biosynthesis [45,46]. In addition, modification of glucose transport capacity is another successful approach at reducing acetate production rate under aerobic conditions. The metabolically inert glucose analog methyl α-glucoside (α-MG) has been employed as a competitive inhibitor of PTS glucose transport [18]. In this report, the addition of 6.67 g/l of α-MG to E. coli cultures growing in complex medium supplemented with 20 g/l of glucose resulted in a 54% reduction in acetate concentration, 39% increase in the specific activity of a recombinant protein and 15% increase in the final biomass concentration, when compared to cultures lacking α-MG. In a different approach, inactivation of ptsG was evaluated as a strategy to reduce glucose uptake capacity. In this study, it was determined that cultures of a ptsG- strain growing in complex medium supplemented with 15 or 20 g/l of glucose resulted in significant reduction of acetate secretion and more than 50% increase in recombinant protein synthesis when compared to wild type strain cultures [47]. Similar results have been reported with a PTS- Glc+ strain having a deletion of the ptsHI-crr operon and obtained by selection from a continuous culture [33]. Its characterization when growing in minimal medium cultures with glucose as carbon source revealed that it accumulates 80% less acetate than an isogenic PTS+ strain [48].
Transcriptional repression of genes encoding PTS components has been reported as a successful method to decrease glucose uptake rate and acetate buildup in wild type E. coli strain cultures. The mlc gene encodes the regulatory protein Mlc that represses, among others, the ptsHI and ptsG genes [49] (Fig. 3). Cultures of a recombinant strain having the mlc gene expressed from a multicopy plasmid and growing in complex medium with 0.4% glucose, showed a 50% reduction in acetate accumulation and increased capacity to consume this acid when compared to a strain having only a chromosomal copy of mlc [50].
Figure 3 Carbon catabolic repression mechanisms in Escherichia coli.
Increasing production of glycolytic and TCA cycle intermediates
The PTS is one of several cellular activities that influences the PEP/PYR ratio and carbon flux distribution originating from these two central metabolic nodes. As the main PEP-consuming activity when E. coli grows on glucose or other PTS sugars, it can be expected that modification or elimination of PTS components should have a significant impact on carbon flux distribution in central metabolism. To ascertain some of the effects of PTS inactivation on carbon metabolism, metabolic flux analysis using 13C-labeled glucose and NMR spectroscopy has been performed [26]. This study revealed significant carbon flux distribution differences among PTS+ and PTS- strains at the EMP, pentose phosphate pathway (PPP) and TCA cycle. As an example of the results obtained, the relative carbon fluxes into the EMP pathway and the PPP corresponded to 76.6% and 22.3% for the PTS+ strain and 93.1% and 5.3% for the PTS- Glc+strain, respectively. Most of the changes observed in metabolic flux distribution in the PTS- Glc+ strain could not have been predicted based on our current knowledge about the properties of the central metabolic network in E. coli. Further experimental and theoretical analysis, now in progress by several research groups, will be required to fully understand the connections between PTS and central metabolism.
The effect of PTS replacement by the combined activities of GalP and glucokinase on glycolytic flux to fermentation products has been reported [37]. The high-level expression of galP and glk in a PTS- Glc- strain restored glucose transport and resulted in a two-fold increase in the specific rate of acetate production when compared to a PTS+ strain. Both of these strains were transformed with a plasmid carrying the Zymomonas mobilis genes pdcZm and adhBZm encoding pyruvate decarboxylase and alcohol dehydrogenase II, respectively. These enzymes generate a pathway for the synthesis of ethanol from PYR [51]. When these two strains were grown in complex medium supplemented with glucose, a two-fold increase in the specific rate of ethanol production was observed when comparing the PTS- galP+ glk+ and the PTS+ strains. These results, and those obtained when trying to reduce acetate overflow, show that modulation of galP and glk expression level in a PTS- Glc- strain allows the possibility of controlling glycolytic flux so that the rate of production of acetate and other fermentation products could be lower of higher than that of a PTS+ strain.
One example of the improvement of a strain for the production of a TCA cycle intermediate is the case of succinate. This is a valuable specialty compound that is employed as a precursor of industrial chemicals [52]. Succinate is currently produced from petroleum derivates as a raw material, but recently, considerable effort has been applied to the development of microbial strains for the biotechnological production of this metabolite. These E. coli strains have extensive modifications to central metabolic pathways, resulting in the redirection of carbon flux from the EMP and TCA pathways to enzymes isocitrate lyase (aceA) and succinyl-CoA synthetase (sucC, sucD), both leading to the production of succinate [53]. An additional modification to these strains was the inactivation of ptsG. In this study, it was determined that a strain with an inactive ptsG displayed a 22.5% increase in final succinate concentration, 16.4% increase in succinate molar yield from glucose and 22% higher specific productivity when compared with an isogenic ptsG+ strain. These results have been explained considering that inactivation of ptsG caused a lower rate of glucose consumption and acetate production, thus resulting in a more balanced and efficient metabolism [53].
Elimination of carbon catabolic repression for the simultaneous consumption of sugar mixtures
E. coli has the capacity to select, from a mixture of carbon sources, the one that affords the highest growth rate. This response is called carbon catabolite repression (CCR) and it is the result of inhibition of sugar transport capacity, enzyme activities and gene expression by the presence of a rapidly metabolizable carbon source [54]. The IIAGlc protein has a central role in CCR. When glucose is present in the medium, this protein is non-phosphorylated and in this state it binds to various non-PTS permeases, inhibiting uptake of non-PTS sugars (Fig. 3). This form of IIAGlc also binds to the enzyme glycerol kinase (GK), inhibiting its activity [55]. In addition, non-phosphorylated IIBGlc binds the Mlc repressor protein, thus relieving its repression from genes ptsHI, ptsG, mlc, manXYZ and malT [49]. When glucose is absent from the culture medium, IIAGlc and IIBGlc will be mainly in their phosphorylated state. In this condition, IIAGlc~P binds to the enzyme adenylate cyclase (AC) activating its cAMP biosynthetic capacity. Therefore, cAMP concentration increases in the cell, binding to the cAMP receptor protein (CRP) and causing the induction of catabolite-repressed genes [56]. Protein IIBGlc~P looses its capacity to bind Mlc, so this protein binds to its target operator sequences, causing repression of genes involved in glucose uptake [49]. Proteins EI and Hpr also have regulatory functions. In its non-phosphorylated state, Hpr activates glycogen phosphorylase (GP), whereas Hpr~P has a similar effect on BglG, a transcriptional activator of the bgl operon that encodes proteins involved in β-glucosidic sugars uptake and utilization [57,58]. Non-phosphorylated EI has been shown to bind to the chemotaxis protein CheA, inhibiting its autophosphorylation and thus causing smooth swimming [59].
These data shows that PTS forms part of a regulatory network involved in coordinating cellular processes related to the cell's capacity to find, select, transport and metabolize a large number of carbon sources [12-60]. In addition, it has been demonstrated that protein IIAGlc exerts negative control of expression for the gene encoding the σS subunit of RNA polymerase [59]. Therefore, it can be expected that alterations to PTS components should have wide-ranging effects on the cell's physiology. Regarding the regulatory functions of PTS, its modification for strain improvement purposes has focused on reducing or eliminating CCR.
Lignocellulose derived from agricultural residues is a potential low-cost feedstock for the production of different types of chemicals by fermentation, among them, fuel ethanol and L-lactic acid [62,63]. Hydrolysis of lignocellulose yields a mixture of sugars containing mainly glucose, arabinose and xylose [64]. Due to CCR, E. coli displays sequential sugar consumption when it is grown in media derived from lignocellulose hydrolyzates. Simultaneous consumption of sugars in a mixture would be advantageous in a fermentative production process, as this would eliminate diauxic growth, therefore, reducing operating time and increasing productivity.
Different groups have explored strategies to disrupt CCR by inactivating PTS components. Starting from an E. coli strain engineered for increased ethanol production by the introduction of genes pdcZm and adhBZm, mutant derivatives were selected on the basis of resistance to the PEP analog fosfomycin [65]. Fosfomycin is used to select for mutants that cannot transport PTS sugars and mutations usually occur in the ptsI gene [66]. Some of these mutant strains lost the capacity to ferment PTS sugars, while others retained it. The specific genetic lesions in these strains were not determined. When compared to the parental wild type strain, a mutant strain from the latter class displayed a higher rate of sugars consumption when growing in a medium supplemented with a mixture containing 30 g/l each of glucose, arabinose and xylose. In addition, this mutant produced 20% more ethanol than the wild type when growing in medium containing 120 g/l xylose as sole carbon source [65].
The effect of ptsG inactivation on the pattern of sugar mixture utilization and its impact on ethanol production has been determined [67]. In cultures performed with minimal medium containing 2 g/l each of either glucose and arabinose or glucose and xylose, a wild type strain displayed sequential glucose-pentose utilization, whereas the ptsG mutant consumed these sugars simultaneously. In both conditions, the ptsG mutant consumed the total amount of sugars in about half the time, when compared to the wild type strain. Both strains were transformed with a plasmid bearing the pdcZm and adhBZm genes and the transformants were cultured in LB medium supplemented with 4 g/l each of either glucose and xylose. Under these conditions, the ethanol yields of the ptsG and wild type strains were 3.5% and 2.9% (w/v), respectively [67]. A similar study was performed with a PTS- Glc+ strain obtained from a PTS- Glc- mutant by a continuous culture selection method [68,33]. When grown in a medium containing 1 g/l each of glucose, arabinose and xylose, the PTS- Glc+ strain consumed the total amount of sugars in the medium 16% faster than an isogenic PTS+strain. In addition, acetate produced during growth was completely consumed by the PTS- Glc+ strain, whereas 0.4 g/l remained in the PTS+ strain culture [68].
L-lactic acid is a chemical precursor with many different applications in industry [69]. E. coli strains engineered for the production of lactic acid have been developed by inactivation of the genes encoding pyruvate formate lyase and lactate dehydrogenase. The double mutant strain was then transformed with a plasmid carrying a gene encoding the L-specific lactic acid dehydrogenase from Streptococcus bovis, thus generating an engineered strain that produces only L-lactate. From this strain, a ptsG mutant was generated and compared with other isogenic strains in cultures performed with medium containing 50 g/l each of glucose and xylose. The ptsG mutant fermented 75% of the xylose, while for ptsG+ strains this value was 18–20%. Furthermore, lactate yield for the ptsG and wild type strains were 0.77 (g lactic acid/ g sugar) and 0.48 (g/g), respectively [63].
Perspectives
As shown by the examples presented here, inactivation of specific PTS components or the total functional replacement of PTS-dependent glucose transport capacity in wild type and engineered strains can result in significant improvements for the production of different classes of compounds. Although some of the strain improvements generated by PTS modification are product-specific, disruption of CCR to improve utilization of sugar mixtures from inexpensive lignocellulosic hydrolysates would be expected to be non product-specific and thus have a positive impact in several of the current biotechnological production processes for bulk and commodity compounds. It would be expected that some of the gains observed by modifying or replacing the glucose PTS components, could also be obtained when the same strategies are applied to the components for other PTS sugars. This is an issue that has not been explored yet, however, it is particularly revelant for production processes based on the use as raw materials of sugar mixtures containing different PTS sugars.
The PTS is widely distributed among Bacteria, therefore, some of the improvements observed in E. coli could be translated to other species used in industrial processes like the Gram-positive bacteria Bacillus subtilis and Corynebacterium glutamicum. Considering the significant differences between Gram-negative and Gram-positive bacteria with regard to central metabolic network architecture, PTS complex composition and CCR regulation, it remains to be determined what would be the consequences of PTS modification in this group of bacteria in an industrial production context.
The PTS is a complex system deeply integrated to the cell's physiology. Much work remains to be done to fully understand the outcome of its modifications. A genomic approach to the study of PTS- strains, including transcriptome, proteome and metabolome analysis, will be fundamental to comprehend the extent of the participation of this system in several of the cell's processes. This knowledge will be valuable to help in the definition of strain design and improvement strategies by metabolic engineering.
Acknowledgements
I wish to thank F. Bolívar and A. Martínez for the critical reading of this manuscript. This work was supported by CONACYT grants NC-230, 43243 and 37342-N and DGAPA-PAPIIT, UNAM grants IN220403-2 and IN218902.
==== Refs
Bailey JE Toward a science of metabolic engineering Science 1991 252 1668 1675 2047876
Cameron DC Tong IT Cellular and metabolic engineering. An overview Appl Biochem Biotechnol 1993 38 105 140 8346901
Nikaido H Nakae T The outer membrane of Gram-negative bacteria Adv Microb Physiol 1979 20 163 250 394591
Nikaido H Neidhardt FC Outer membrane 1996 Washington, DC: American Society for Microbiology 1325 1343
Hall MN Silhavy TJ The ompB locus and the regulation of the major outer membrane porin proteins of Escherichia coli K12 J Mol Biol 1981 146 23 43 7021856 10.1016/0022-2836(81)90364-8
Lugtenberg B Peters R Bernheimer H Berendsen W Influence of cultural conditions and mutations on the composition of the outer membrane proteins of Escherichia coli Mol Gen Genet 1976 147 251 262 787762 10.1007/BF00582876
Pratt LA Silhavy TJ The response regulator SprE controls the stability of RpoS Proc Natl Acad Sci U S A 1996 93 2488 2492 8637901 10.1073/pnas.93.6.2488
Nikaido H Vaara M Molecular basis of bacterial outer membrane permeability Microbiol Rev 1985 49 1 32 2580220
Death A Notley L Ferenci T Derepression of LamB protein facilitates outer membrane permeation of carbohydrates into Escherichia coli under conditions of nutrient stress J Bacteriol 1993 175 1475 483 8444809
Nikaido H Rosenberg EY Porin channels in Escherichia coli : studies with liposomes reconstituted from purified proteins J Bacteriol 1983 153 241 252 6294049
von Meyenburg K Nikaido H Outer membrane of gram-negative bacteria. XVII. Secificity of transport process catalyzed by the lambda-receptor protein in Escherichia coli Biochem Biophys Res Commun 1977 78 1100 1107 334177
Postma PW Lengeler JW Jacobson GR Neidhardt FC Phosphoenolpyruvate: Carbohydrate phosphotransferase systems Escherichia coli and Salmonella Cellular and Molecular Biology 1996 Washington, DC: American Society for Microbiology 1149 1174
Saier MH Vectorial metabolism and the evolution of transport systems J Bacteriol 2000 182 5029 5035 10960084 10.1128/JB.182.18.5029-5035.2000
Tchieu JH Norris V Edwards JS Saier MH Jr The complete phosphotranferase system in Escherichia coli J Mol Microbiol Biotechnol 2001 3 329 346 11361063
Misset O Blaauw M Postma PW Robillard GT Bacterial phosphoenolpyruvate-dependent phosphotransferase system. Mechanism of the transmembrane sugar translocation and phosphorylation Biochemistry 1983 22 6163 6170 6362721 10.1021/bi00295a019
Stock JB Waygood EB Meadow ND Postma PW Roseman S Sugar transport by the bacterial phosphotransferase system. The glucose receptors of the Salmonella typhimurium phosphotransferase system J Biol Chem 1982 257 14543 14552 6292227
Curtis SJ Epstein W Phosphorylation of D-glucose in Escherichia coli mutants defective in glucosephosphotransferase, mannosephosphotransferase, and glucokinase J Bacteriol 1975 122 1189 99 1097393
Chou CH Bennett GN San KY Effect of modulated glucose uptake on high-level recombinant protein production in a dense Escherichia coli culture Biotechnol Prog 1994 10 644 647 7765699 10.1021/bp00030a009
Death A Ferenci T Between feast and famine: endogenous inducer synthesis in the adaptation of Escherichia coli to growth with limiting carbohydrates J Bacteriol 1994 176 5101 5107 8051023
Liu X Ferenci T Regulation of porin-mediated outer membrane permeability by nutrient limitation in Escherichia coli J Bacteriol 1998 180 3917 3922 9683489
Flores N Flores S Escalante A de Anda R Leal L Malpica R Georgellis D Gosset G Bolivar F Adaptation for fast growth on glucose by differential expression of central carbon metabolism and gal regulon genes in an Escherichia coli strain lacking the phosphoenolpyruvate:carbohydrate phosphotransferase system Metab Eng 2005 7 70 87 15781417 10.1016/j.ymben.2004.10.002
McDonald TP Walmsley AR Henderson PJF Asparagine 394 in putative helix 11 of the galactose-H+ symport protein (GalP) from Escherichia coli is associated with the internal binding site for cytochalasin B and sugar J Biol Chem 1997 272 15189 15199 9182541 10.1074/jbc.272.24.15189
Lunin VV Li Y Schrag JD Iannuzzi P Cygler M Matte A Crystal structures of Escherichia coli ATP-dependent glucokinase and its complex with glucose J Bacteriol 2004 186 6915 6927 15466045 10.1128/JB.186.20.6915-6927.2004
Holms WH Horecker BL, Stadtman ER The central metabolic pathway of Escherichia coli: relationship between flux and control at a branch point, efficiency of conversion to biomass, and excretion of acetate Current topics in cell regulation 1986 Academic Press, Inc., New York 69 105
Valle F Muñoz E Ponce E Flores N Bolivar F Basic and applied aspects of metabolic diversity: the phosphoenolpyruvate node J Ind Microbiol 1996 17 458 462 10.1007/BF01574776
Flores S Gosset G Flores N de Graff AA Bolivar F Analysis of carbon metabolism in Escherichia coli strains with an inactive phosphotransferase system by 13C labeling and NMR spectroscopy Metab Eng 2002 4 124 137 12009792 10.1006/mben.2001.0209
Förberg C Eliaeson T Häggström L Correlation of theoretical and experimental yields of phenylalanine from non-growing cells of a rec Escherichia coli strain J Biotechnol 1988 7 319 332 10.1016/0168-1656(88)90043-0
Varma A Boesch BW Palsson BO Biochemical production capabilities of Escherichia coli Biotechnol Bioeng 1993 42 59 73 10.1002/bit.260420109
Patnaik R Liao JC Engineering of Escherichia coli central metabolism for aromatic production with near theoretical yield Appl Environ Microbiol 1994 60 3903 3908 7993080
Frost JW Draths KM Biocatalytic syntheses of aromatics from D-glucose: renewable microbial sources of aromatic compounds Annu Rev Microbiol 1995 49 557 579 8561472 10.1146/annurev.mi.49.100195.003013
Liao JC Hou SY Chao YP Pathway analysis, engineering, and physiological considerations for redirecting central metabolism Biotechnol Bioeng 1996 52 129 140 10.1002/(SICI)1097-0290(19961005)52:1<129::AID-BIT13>3.0.CO;2-J
Báez JL Bolivar F Gosset G Determination of 3-deoxy-D-arabino-heptulosonate 7-phosphate productivity and yield from glucose in Escherichia coli devoid of the glucose phosphotransferase system Biotechnol Bioeng 2001 6 530 535 10.1002/bit.1088
Flores N Yong-Xiao J Berry A Bolivar F Valle F Pathway engineering for the production of aromatic compounds in Escherichia coli Nat Biotechnol 1996 14 620 623 9630954 10.1038/nbt0596-620
Báez-Viveros JL Osuna J Hernandez-Chavez G Soberon X Bolivar F Gosset G Metabolic engineering and protein directed evolution increase the yield of L-phenylalanine synthesized from glucose in Escherichia coli Biotechnol Bioeng 2004 87 516 524 15286989 10.1002/bit.20159
Snoep JL Arfman N Yomano LP Fliege RK Conway T Ingram LO Reconstitution of glucose uptake and phosphorylation in a glucose-negative mutant of Escherichia coli by using Zymomonas mobilis genes encoding the glucose facilitator protein and glucokinase J Bacteriol 1994 176 2133 2135 8144485
Weisser P Kramer R Sahm H Sprenger GA Functional expression of the glucose transporter of Zymomonas mobilis leads to restoration of glucose and fructose uptake in Escherichia coli mutants and provides evidence for its facilitator action J Bacteriol 1995 177 3351 3354 7768841
Hernández-Montalvo V Martinez A Hernández-Chávez G Bolivar F Valle F Gosset G Expression of galP and glk in a Escherichia coli PTS mutant restores glucose transport and increases glycolytic flux to fermentation products Biotechnol Bioeng 2003 83 687 694 12889033 10.1002/bit.10702
Yi J Draths KM Li K Frost JW Altered glucose transport and shikimate pathway product yields in E. coli Biotechnol Prog 2003 19 1450 1459 14524706 10.1021/bp0340584
Burgard AP Maranas CD Probing the performance limits of the Escherichia coli metabolic network subject to gene additions or deletions Biotechnol Bioeng 2001 74 364 75 11427938 10.1002/bit.1127
Kessler D Knappe J Neidhardt FC Anaerobic dissimilation of pyruvate Escherichia coli and Salmonella Cellular and Molecular Biology 1996 Washington, DC: American Society for Microbiology 1149 1174
Han KH Lim C Hong J Acetic acid formation in Escherichia coli fermentation Biotechnol Bioeng 1992 39 663 671 10.1002/bit.260390611
Majewski RA Domach MM Simple constrained-optimization view of acetate overflow in E. coli Biotechnol Bioeng 1990 35 732 738 10.1002/bit.260350711
Delgado J Liao JC Inverse flux analysis for reduction of acetate excretion in Escherichia coli Biotechnol Prog 1997 13 361 367 9265774 10.1021/bp970047x
Shiloach J Kaufman J Guillard AS Fass R Effect of glucose supply strategy on acetate accumulation, growth, and recombinant protein production by Escherichia coli BL21 (lλDE3) and Escherichia coli JM109 Biotechnol Bioeng 1996 49 421 428 10.1002/(SICI)1097-0290(19960220)49:4<421::AID-BIT9>3.0.CO;2-R
Konstantinov K Kishimoto M Seki T Yoshida T A balanced DO-stat and its application to the control of acetic acid excretion by recombinant Escherichia coli Biotechnol Bioeng 1990 36 750 758 10.1002/bit.260360714
Yang YT San KY Bennett GN Metabolic flux analysis of E. coli deficient in the acetate production pathway and expressing the B. subtilis acetolactate synthase Metab Eng 1999 1 26 34 10935752 10.1006/mben.1998.0103
Chou CH Bennett GN San KY Effect of modified glucose uptake using genetic engineering techniques on high-level recombinant protein production in Escherichia coli dense cultures Biotechnol Bioeng 1994 44 952 960 10.1002/bit.260440811
Sigüenza R Flores N Hernández G Martínez A Bolivar F Valle F Kinetic characterization in batch and continuous culture of Escherichia coli mutants affected in phosphoenolpyruvate metabolism: differences in acetic acid production W J Microbiol Biotech 1999 15 587 592 10.1023/A:1008934810150
Plumbridge J Regulation of gene expression in the PTS in Escherichia coli : the role and interactions of Mlc Curr Opin Microbiol 2002 5 187 193 11934616 10.1016/S1369-5274(02)00296-5
Hosono K Kakuda H Ichihara S Decreasing accumulation of acetate in a rich medium by Escherichia coli on introduction of genes on a multicopy plasmid Biosci Biotechnol Biochem 1995 59 256 61 7766024
Martinez A York SW Yomano LP Pineda VL Davis FC Shelton JC Ingram LO Biosynthetic burden and plasmid burden limit expression of chromosomally integrated heterologous genes (pdc, adhB) in Escherichia coli Biotechnol Prog 1999 15 891 897 10514259 10.1021/bp990103p
Zeikus JG Jain MK Elankovan P Biotechnology of succinic acid production and markets for derived industrial products Appl Microbiol Biotechnol 1999 51 545 552 10.1007/s002530051431
Lin H Bennett GN San KY Metabolic engineering of aerobic succinate production systems in Escherichia coli to improve process productivity and achieve the maximum theoretical succinate yield Metab Eng 2005
Saier MH Ramseier TM Reizer J Neidhardt FC Regulation of carbon utilization Escherichia coli and Salmonella Cellular and Molecular Biology 1996 Washington, DC: American Society for Microbiology 1325 1343
Novotny MJ Frederickson WL Waygood EB Saier MH Jr Allosteric regulation of glycerol kinase by enzyme IIIglc of the phosphotransferase system in Escherichia coli and Salmonella typhimurium J Bacteriol 1985 162 810 816 2985549
Korner H Sofia HJ Zumft WG Phylogeny of the bacterial superfamily of Crp-Fnr transcription regulators: exploiting the metabolic spectrum by controlling alternative gene programs FEMS Microbiol Rev 2003 27 559 592 14638413 10.1016/S0168-6445(03)00066-4
Seok YJ Koo BM Sondej M Peterkofsky A Regulation of E. coli glycogen phosphorylase activity by HPr J Mol Microbiol Biotechnol 2001 3 385 393 11361069
Görke B Rak B Catabolite control of Escherichia coli regulatory protein BglG activity by antagonistically acting phosphorylations EMBO J 1999 18 3370 3379 10369677 10.1093/emboj/18.12.3370
Lux R Jahreis K Bettenbrock K Parkinson JS Lengeler JW Coupling the phosphotransferase system and the methyl-accepting chemotaxis protein-dependent chemotaxis signaling pathways of Escherichia coli Proc Natl Acad Sci 1995 92 11583 11587 8524808
Saier MH JrReizer J The bacterial phosphotransferase system: new frontiers 30 years later Mol Microbiol 1994 13 755 764 7815935
Ueguchi C Misonou N Mizuno T Negative control of rpoS expression by phosphoenolpyruvate: carbohydrate phosphotransferase system in Escherichia coli J Bacteriol 2001 183 520 527 11133945 10.1128/JB.183.2.520-527.2001
Bothast RJ Nichols NN Dien BS Fermentations with new recombinant organisms Biotechnol Prog 1999 15 867 875 10514256 10.1021/bp990087w
Dien BS Nichols NN Bothast RJ Fermentation of sugar mixtures using Escherichia coli catabolite repression mutants engineered for production of L-lactic acid J Ind Microbiol Biotechnol 2002 29 221 227 12407454 10.1038/sj.jim.7000299
Sheehan J Himmel M Enzymes, Energy, and the Environment: A Strategic Perspective on the U.S. Department of Energy's Research and Development Activities for Bioethanol Biotechnol Prog 1999 15 817 827 10514251 10.1021/bp990110d
Lindsay SE Bothast RJ Ingram LO Improved strains of recombinant Escherichia coli for ethanol production from sugar mixtures Appl Microbiol Biotechnol 1995 43 70 75 7766137 10.1007/s002530050372
Cordaro JC Melton T Stratis JP Atagun M Gladding C Hartman PE Roseman S Fosfomycin resistance: selection method for internal and extended deletions of the phosphoenolpyruvate:sugar phosphotransferase genes of Salmonella typhimurium J Bacteriol 1976 128 785 793 186449
Nichols NN Dien BS Bothast RJ Use of catabolite repression mutants for fermentation of sugar mixtures to ethanol Appl Microbiol Biotechnol 2001 56 120 125 11499918 10.1007/s002530100628
Hernández-Montalvo V Valle F Bolivar F Gosset G Characterization of sugar mixtures by an Escherichia coli mutant devoid of the phosphotransferase system Appl Microbiol Biotechnol 2001 57 186 191 11693918 10.1007/s002530100752
Jarvis L Lactic acid outlook up as polylactide nears market Chem Market Rep 2001 259 5 14
| 15904518 | PMC1156936 | CC BY | 2021-01-04 16:24:35 | no | Microb Cell Fact. 2005 May 16; 4:14 | utf-8 | Microb Cell Fact | 2,005 | 10.1186/1475-2859-4-14 | oa_comm |
==== Front
Microb Cell FactMicrobial Cell Factories1475-2859BioMed Central London 1475-2859-4-171592462310.1186/1475-2859-4-17ResearchExpression of yeast deubiquitination enzyme UBP1 analogues in E. coli Wojtowicz Anna [email protected] Anna [email protected] Grazyna [email protected] Diana [email protected] Luiza [email protected] Natalia [email protected] Andrzej [email protected] Institute of Biotechnology and Antibiotics, Staroscinska 5,02-516 Warsaw, Poland2005 30 5 2005 4 17 17 18 5 2005 30 5 2005 Copyright © 2005 Wojtowicz et al; licensee BioMed Central Ltd.2005Wojtowicz et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
It has been shown that proteins fused to ubiquitin undergo greater expression in E. coli and are easier to purify and renaturate than nonhybrid foreign proteins. However, there is no commercial source of large quantities of specific deubiquitinating proteases. This is the reason why hybrid proteins containing ubiquitin at their N-end cannot be used in large scale biotechnological processes.
Results and Conclusion
We have described the synthesis of the yeast deubiquination enzyme UBP1 muteins in E. coli. We have shown that an efficient overproduction of the enzyme in E. coli may be achieved after the introduction of several changes in the nucleotide sequence encoding UBP1. One of the conditions of an effective synthesis of the UBP1 muteins is the removal of the 5'-end sequence encoding the transmembrane region of the enzyme. The obtained variants of the enzyme may be successfully used for processing large amounts of hybrid proteins comprising ubiquitin or tagged ubiquitin at their N-ends.
==== Body
Background
Ubiquitin is composed of 76 amino-acid residues with a total molecular mass of 8.6 kDa. This protein is an element of the universal protein modification in eukaryots called ubiquitination, a phenomenon which does not occur in bacteria. In spite of that, it has been shown that proteins fused to ubiquitin undergo greater expression in E. coli and are easier to purify and renaturate than nonhybrid foreign proteins [1]. However, to take advantage of these properties of hybrid proteins in technological processes, large amounts of proteases for cleaving specifically ubiquitin from those proteins are necessary. Protease UBP1, an enzyme found in the yeast Saccharomyces cerevisiae, is a candidate for becoming such a tool. The enzyme was described in 1991 [2] and is the subject of a patent application [3]. UBP1 is a cysteine protease which cleaves ubiquitin from protein fused to its C-end. Its activity and culture conditions in E. coli have been described [2], but the problem of a larger and more efficient production of the enzyme remains unsolved and for that reason technical applications of the inventions mentioned in [3] have not been possible.
The aim of this work was to obtain an expression system for an efficient synthesis of UBP1 protease variants that would be useful in industrial processes.
Results and Discussion
We were not able to obtain an efficient expression of the full coding sequence of UBP1. This is in agreement with the results of Tobias and Varshavsky [2], who were able to detect UBP1 in E. coli only after an immunoblot analysis. A possible reason for this situation might be the presence of a transmembrane region in the N-end part of the protein, which was discovered after a computer analysis using the TMHMM v. 2.0 software package (CBS, Denmark) [4]. The region encompasses amino-acids 34–51 (Figure 1). Because of that, we decided to prepare truncated variants of the sequences encoding UBP1 without the transmebrane region. Using the PCR technique, two shortened genes were prepared without the 5' parts of the UBP1 coding region. Proteins encoded by these variants of the gene are shorter by 54 and 98 amino-acids (Figure 1). Expression of these analogues of the UBP1 gene inserted into pT7RS derivatives was analyzed. In addition, the influence of one mutation, named Q754L, within the UBP1 gene, which happened during PCR amplification, on the expression level was investigated (Figure 1A).
Figure 1 Nucleotide and amino-acid sequences of the yeast UBP1 gene and UBP1 protease [2]. A) The nucleotide residues are numbered on the left, and amino acid residues are numbered on the right of the figure. The underlined AGA or AGG arginine and TTA leucine codons were changed for CGC or CGT and CTG triplets, respectively. The CAG triplet (shown in bold) at positions 2260–2262 was changed to CTG one (Q754L mutation). The amino-acids forming active centre is shown in italics [4]. For the construction of UBI::UBPDQL or UBI::UBPD2QL coding sequences the first 162 bp or 294 bp which were removed are overlined or underlined, respectively. The 3'part of the UBP1 gene (positions: 2074–2430) was used to construct the S protein, the truncated hybrid protein which was used to determine the UBP1 variants' activity. B) Schematic structure of UBI::UBP1 construction variants. – ubiquitin, □ – UBP1, – 6xHis.
Expression of the UBP1 variants in E. coli
The obtained plasmids with the hybrid genes encoding the analogues: UBI::UBPD, UBI::UBPDQL, and UBI::UBPD2QL were used to transform the E. coli BLD21 strain. In general, several expression host strains and different culture conditions were examined (results not shown) in our mini-induction screening experiments to optimize recombinant protein expression. The cultures were grown at 37°C and 25°C until OD600 nm reached 0.5–1.8 and 1 ml samples were removed from each culture and saved as controls. Then, isopropyl β-D-thiogalactoside was added to the cultures at the desired concentration of 0.5–1.0 mM. Aliquots of cultures were taken 30, 60, 90, 120 and 150 min. post-induction (Figure 2). Cell density in these samples was measured to monitor the protein induction process and to normalize total proteins in E. coli after cell lysis in the 2xSDS sample buffer. The expression level (Figure 3) of the recombinant proteins was determined by a densitometric analysis of the electrophoretic pattern. The protein level was equal to 7.2 % and 9.8 % of the proteins visible in lines 2 and 5 (Figure 3) for UBPDQL and UBPD2QL, respectively. The results shown in Fig. 4 and 5 indicate that the fusions of ubiquitin with UBP1 analogues are active proteases during cell culture because an intensive band of ubiquitin is visible on SDS gels after electrophoresis of E. coli lysates.
Figure 2 Dependence of the expression level of the shorter analogue of UBP1, UBP1D2QL, on the time after IPTG induction as revealed by SDS electrophoresis in 12% polyacrylamide gel. E. coli BL21 (DE3) cells transformed by the plasmid pT7UPD2QL were cultured at 25°C in LB, and induced by the addition of IPTG to the final concentration of 1 mM. Aliquots of cultures were taken at: 1 – 0 min, 2 – 30 min, 3 – 60 min, 4 – 90 min, 5 – 120 min, 6 – 150 min after induction.
Figure 3 Expression level of the UBP1 analogues as revealed by SDS electrophoresis in 12% polyacrylamide gel. 1 and 2 – lysates of E. coli BLD3(DE3) bacteria transformed with pT7UPD2QL, not induced and induced with IPTG from cultures maintained in 100 ml LB media, respectively; 3 – supernatant after sonication of the bacteria culture performed in the fermenter; 4 and 5 – lysates of E. coli BLD3 bacteria transformed with pT7UPDQL, cultured, not induced and induced with IPTG, respectively.
Figure 4 Activity of UBPD2QL protease in bacterial cells, SDS electrophoresis in 12% polyacrylamide of bacterial lysates. 1 – molecular mass marker (kDa); 2 and 3 – lysates of E. coli BLD3 (DE3) bacteria transformed with pT7UPDQL2, cultured, induced and not induced with IPTG, respectively. The ubiquitin band is marked with an arrow.
Figure 5 Expression level of UBP1 analogue with Q754L mutation (UBPQL) as revealed by SDS electrophoresis in 12% polyacrylamide gel. 1 – molecular mass marker (kDa); 2 and 3 – lysates E. coli BLD21(DE3) bacteria transformed with pT7UPQL, cultured, not induced and induced with IPTG, respectively. The UBPQL protease and ubiquitin bands are marked with arrows.
The main factors influencing the expression level of UBP1 analogues and growth rate of bacteria are: the presence or absence of the Q754L mutation (Figure 5) and lack of the transmembrane region at the N-end of UBP1 (Figure 3). The lack of Q754L mutation in the UBP1 analogue gene makes an efficient production of the enzyme impossible because of a very slow culture growth rate and a low level of protease expression (results not shown). Apart from that we have shown that the removal of the transmembrane region leads to a significant reduction in growth time and an increase in the expression level (Figure 3). The period of time to reach OD600 = 1 for UBPD, UBPDQL and UBPD2QL was 48, 12 and 9 h, respectively. It appears also that the bacteria containing the vector with the shorter gene of the UBP1 analogue encoding UBPD2QL produce the highest amount of the recombinant protease (Figure 3 and 4). It was also established that the presence of the proper codons in the gene used for expression additionally shortened the period of time necessary to reach OD600 = 1 (results not shown). All things considered, the best expression system consists of the gene variant with the Q754L mutation, larger deletion of the N-end encoding sequence and proper codon usage. The best results were obtained when the bacteria culture was carried out at 25°C, the induction with IPTG begun when OD had reached 1 and when the culture time after induction was 2h (Figure 2).
Purification of the recombinant UBPD2QLHisx6 protease and fusion proteins of the type 6xHisUBI::protein
For purification, the cells from 0.5 litre culture after a 2-hour induction at 25°C in the presence of 1 mM IPTG were harvested. The pellet was resuspended in 50 ml of buffer A. All subsequent steps were performed at 4°C. The cells were disrupted by sonication on ice and the insoluble debris was removed by centrifugation for 25 min. at 11500 rpm. The cleared extract was chromatographed on a Ni-NTA Superflow column, pre-equilibrated with 10 vol. of buffer A. After loading, the column was washed with 5 vol. of buffer B and then the protease was eluted twice with elution buffer C. Protein concentrations were determined using the Bradford dye binding assay and bovine serum albumin as the standard. Both analogues of UBP1, UBPDQL and UBPD2QL, were purified in the same manner. Digestion of the two substrates, 6xHisUBI::S and UBI::K, shows that the preparations of the two different UBP1 analogues are active proteases, but the activity of the shorter one (UBP1DC2) is higher (Figure 6). Stability of both the UBPDQL and UBPD2QL analogues during long exposure to 37°C was also investigated. It appeared that both enzymes were active after 24 h incubation (results not shown).
Figure 6 Activity of the UBP1 analogues, UBPDQL and UBPD2QL, as revealed by SDS electrophoresis in 12% polyacrylamide gel. 1 and 2 – the substrate 6xHisUBI::S digested with UBPDQL and UBPD2QL, respectively; 3 – the undigested substrate, 6xHisUBI::S; 4 – molecular mass marker (kDa); 5 and 6 – the substrate UBI::K digested with UBPD2QL and UBPDQL; 7 – the undigested substrate, UBI::K.
We obtained 13.8 mg of purified UBPDQL and 20 mg of purified UBPD2QL from 1 l of E. coli culture. The amounts represent 6.3% and 7.4% of the total protein obtained after the destruction of bacterial cells, and correspond to 710 U of the purified UBPDQL and 1650 U of UBPD2QL, respectively.
We would like to stress that the presence of 6xHis at the C-end of UBP1 analogues may greatly facilitate purification of the recombinant protein containing ubiquitin tagged at its N-end with 6 histidine residues. We propose a three steps procedure: 1) purification of the fusion protein on a Ni-NTA column followed by dialysis; 2) digestion of the purified fusion with a UBP1 analogue, and 3) chromatography of the digestion mixture on the same column. The protein 6xHisUBI::S, used for activity determination, was treated in this way (Figure 7). SDS-PAGE image analysis showed that this simple procedure leads to an almost pure peptide because of the undigested fusion protein. His-tagged ubiquitin and protease bind to the Ni-NTA column while after the digestion of the fusion, the liberated protein flows out of the Ni-NTA column. Such an approach greatly simplifies the purification of the fusion proteins and the limited number of purification stages greatly enhances the efficiency of the whole process.
Figure 7 Digestion of the 6xHisUBI::S fusion with UBPD2QL and purification of peptide S on an Ni-NTA column. SDS electrophoresis in 12% polyacrylamide gel. 1 – molecular mass marker (kDa), 2 – 6HisUbi::S purified in a chromatography column containing Ni-NTA medium, 3 – 6xHisUBI::S digested with protease UBPD2QL, 4 – S protein purified in a chromatography column filled with Ni-NTA medium.
It should be noted that a solution similar to ours, produced artificially as a result of UBP1 gene truncation [6], was observed by Schmitz et al. [7], who discovered two forms of UBP1 in the yeast cells, the anchored and the soluble one.
Conclusion
Our results show that an efficient expression of the unmodified yeast UBP1 protease gene in E. coli in the presented expression system is impossible. The most important change to be introduced into the UBP1 gene is the Q754L mutation leading to the replacement of glutamine by leucine at position 754 in the aa sequence in the yeast UBP1. The removal of the transmembrane N-end region of UBP1 improves the level of expression of a properly truncated gene. Both of these changes in the gene decrease the time needed for bacteria cultivation.
We have shown that using protease analogues, tagged at the C-end by 6xHis, for cleaving fusion proteins containing N-end His-tagged ubiqutin of the type 6xHisUBI::protein facilitates purification of the protein present in the hybrid to a great extent.
The high expression level in E. coli of our UBP1 analogues allows the use of ubiquitin::protein fusion in large scale production of recombinant proteins.
Methods
Bacterial Strains, Plasmids, Enzymes, and Reagents
Saccharomyces cerevisiae, strain W303, was used as a source of the DNA for PCR amplification. The E. coli DH5α, NM522 strains were used for transformations to obtain recombinant plasmids. The E. coli BLD21 (DE3) strain was applied to achieve expression of the His6-tagged analogues of the UBP1 protease. The plasmid pT7RS (GenBank Accession No. AY923866), containing ArgU tRNA (UCU) gene and the transcription terminator of phage T7 RNA polymerase, was used for the construction of the expression system. The E. coli BLD21 (DE3) cells with the pT7RS plasmid derivatives were cultured aerobically at 25°C or 37°C in LB medium supplemented with 50 μg/ml ampicilin. Restriction, modification enzymes were purchased from Amersham Pharmacia Biotech. The reagents for PCR and Ni-NTA Superflow columns were obtained from Qiagen. IPTG, agarose, polyacrylamide and the reagents for protein purification were purchased from Sigma-Aldrich ChemieGmbH, Steinheim, Germany. For site-directed mutagenesis the Stratagene kit was used (Cat. No. 200518-5). We used the GelScan v. 1.45 software package for the densitometric analysis of the electrophoretic image.
The buffers used for the purification of UBP1 analogues were: buffer A (50 mM buffer phos. pH 8.0; 0.3 M NaCl; 10 mM imidazol; 10 mM 2-β-mercapthoethanol; 10% glicerol; 0,1% Triton), buffer B (50 mM phosphate buffer, pH 8.0; 0.3 M NaCl; 40 mM imidazol; 10 mM 2-β-mercapthoethanol; 10% glicerol; 0,1% Triton) and buffer C (50 mM phosphate buffer pH 8.0; 0,3 M NaCl; 300 mM imidazol; 10 mM 2-β-mercapthoethanol; 10% glicerol; 0,1% Triton).
Plasmid construction
To facilitate subsequent purification of the UBP1 variants or fusion proteins, the pT7CH, pT7NH, pT7U and pT7NHU plasmids were constructed. The first one contains in the 3' polilinker part the 6 histidines coding sequences followed by a TAA stop triplet. The second one was constructed for the addition of 6 histidines tags to the N-ends of hybrid proteins. It contains 6 histidines coding sequences following an ATG triplet.
The pT7CH was obtained by inserting a short double stranded DNA fragment formed by two synthetic oligonucleotides, HIST7G and HIST7D (Table 1), into the pT7RS plasmid digested with EcoRI and HindIII restriction nucleases. Similarly, the pT7NH was constructed by inserting a double stranded DNA fragment formed by 6HisG and 6HisD oligonucleotides (Table 1) into pT7RS digested with NdeI and EcoRI.
Table 1 Primers for the construction of pT7RS and UBP1 derivatives
Primersa Sequencesb Restriction site
HIST7G 5' AATTCGATATCGTCGACGGATCCCATCATCACCATCACCATTAAAAT 3' EcoRI
HIST7D 5' AGCTATTTTAATGGTGATGGTGATGATGGGATCCGTCGACGATATCG 3' BamHI, SalI, EcoRV,
6HisG 5' TATGGCACATCATCACCATCACCACTCTGGTTCTG 3' NdeI
6HisD 5' AATTCAGAACCAGAGTGGTGATGGTGATGATGTGCCA 3' EcoRI
UB1G 5' GGGGAATTCATATGCAGATTTTCGTCAAAACTTTG 3' EcoRI and NdeI
UBID2 5' GGGGATCCTTAATGCTCTTCACCACCGCGGAGTCTTAAG 3' BamHI and SacII
UBP1G 5' AGACTCCGCGGTGGTGATTTGTTTATTGAAAGCAAGATA 3' SacII
UBP1D 5' GGGGATCCTTAGTTTACATCTTTACCAGAAATA 3' BamHI
UBP1MG 5'GGCATAGTAGTATTTTTTTACCGCGGTGGTGACCATCTAAACTACATTGT 3' SacII
UBP1MD 5'ACAATGTAGTTTAGATGGTCACCACCGCGGTAAAAAAATACTACTATGCC 3' SacII
SkrutG 5'AAAACCGCGGTGGTTTCATTGCTGGTTTA 3' SacII
SkrutD 5'GGAAGAATTCTTGCGCGTCCTC 3' EcoRI
UBP1GC 5'-GATTTGGAAGCTATTCAGAGTAATAATGAAG-3'
UBP1DC 5'-CTTCATTATTACTCTGAATAGCTTCCAAATC-3'
KalaG 5'-GATCCAGGCGATAAAGACGGTGATGGCTATATTTCTGCTGCTGAAGCTATGGCTTA-3' BamHI
KalaD 5'-AGCTTAAGCCATAGCTTCAGCAGCAGAAATATAGCCATCACCGTCTTTATCGCCTG-3' HindIII
a The names of primers contain the letters: G – sense or D – antisense.
b The restriction sites are underlined or overlined. The histidine and leucine codons are in bold face.
The pT7U plasmid contains a nucleotide sequence encoding modified yeast ubiqutin with the SacII restriction nuclease recognition sequence near the 3' end of the sequence, facilitating construction of different hybrid genes. Primers UB1G and UBID2 (Table 1) were used to amplify and modify the ubiquitin gene. The plasmid was used to express the fusion proteins: ubiquitin::modified UBP1.
The pT7NHU plasmid contains a synthetic nucleotide sequence encoding ubiquitin with the codons used most frequently in the E. coli genome, inserted into the pT7NH plasmid. This plasmid was used to express a hybrid gene for the synthesis of the substrate for the determination of UBP1 activity.
The UBP1 protease gene was obtained using PCR. For amplification, the UBP1G and UBP1D primers were used (Table 1). The oligonucleotides contained recognition sites for the restriction endonucleases SacII and BamHI, respectively. The PCR was performed in a 50 μl reaction volume with a buffer containing 50 mM KCl, 1.75 mM MgCl2, 0.02 mM of each dNTP, 10 mM Tris-HCl (pH 8.9), 100 pM of each primer, the enzyme mixture of Taq and Pwo DNA polymerases (Expand Long Template PCR System, Roche) and 100 ng of total DNA of S. cerevisiae as a template for 30 cycles using Eppendorf 5330 thermocycler. Each cycle consisted of 15 sec at 94°C, 15 sec at 56°C, and 2 min at 72°C. The amplified 2430-bp-long DNA fragment was isolated by 1% agarose gel electrophoresis using Gel-Out kit (Kucharczyk T. E. Co, Warsaw), digested with BamHI and SacII restriction nucleases, purified and ligated to BamHI and SacII digested pT7U vector. Ligation products were transformed into the NM522 E. coli strain. Plasmid DNA was isolated using the alkaline method [8]. Next, the 2430 bp UBP1 gene was excised from the recombinant plasmid using the restriction enzymes NdeI and BamHI, and the DNA fragment encoding hybrid gene ubiquitin::UBP1 was recloned into the pT7CH plasmid at the NdeI and BamHI sites. In this way the pT7UPQL vector was obtained (Table 2).
Table 2 Plasmids and genes used to construct vectors of the pT7RS derivatives
Vectors Gene cloned Result ing vectors Expression products Name of the protein used in the text
pT7SR Ubiquitin pT7U Ubiquitin UBI
pT7U K pT7UK UBI::K UBI::K
UBP1 pT7UP UBI::UBP1 UBP1
pT7CH UBI::UBP1 with Q754L mutation pT7UPQL UBI::UBPQL::6xHis UBPQL
UBI::UBP1(55-809) without Q754L mutation pT7UPD UBI::UBPD::6xHis UBPD
UBI::UBP1(55-809) with Q754L mutation pT7UPDQL UBI::UBPDQL::6xHis UBPDQL
pT7CH UBI::UBP1(99-809) with Q754L mutation pT7UPD2QL UBI::UBPD2QL::6xHis UBPD2QL
pT7NH Ubiquitin pT7NHU 6xHis::UBI 6xHisUBI
pT7NHU S pT7NHUS 6xHis::UBI::S 6xHisUBI::S
UBP1 variants
PCR was used to remove the transmembrane domain from the UBP1 gene. Two variants were obtained. In the first variant, a 162 bp fragment was removed from the 5' part of the coding sequence. For this purpose, site-directed mutagenesis was used with the primers UBP1MG and UBP1MD (Table 1). The shortened gene was modified by the addition of '5-GGTGGT-3', the sequence encoding Gly-Gly, the C-end amino-acids of ubiquitin, and the SacII nuclease recognition sequence. We called this mutein UBPDQL (Figure 1).
The second, shorter variant of UBP1 was prepared by PCR using SkrutG and SkrutD primers (Table 1). In this way an additional 132 bp long DNA fragment was removed. The new variant of the protease consisted of 711 aa (Figure 1). We named this protease variant UBPD2QL. The two shorter variants of the UBP1 gene were inserted into the pT7UCH expression plasmid. The obtained plasmids were designated pT7UPDQL and pT7UPD2QL, and used for the synthesis of UBPDQL and UBPD2QL proteases, respectively (Table 2).
Both variants of the modified gene contain the same mutation leading to CAG to CTG codon change (gln → leu, Figure 1A), which appeared after the first amplification of the UBP1 gene. This mutation was removed using the site-directed kit with the primers UBP1GC and UBP1DC (Table 1).
To circumvent the codon usage problem [9], the UBP1 protease gene was modified through the exchange of certain argining codons (AGA or AGG for CGT or CGC) and leucine codons (TTA for CTG). Stratagene mutagenesis kit with turbo-polymerase was used for site-directed mutagenesis. In this way the following replacements were made: arginine's codons in positions 96, 334, 425, 476, 482, 487, 613, 702, 705, 710, 796 and 801, and leucine codons in positions 354, 356, 358, 367, 369, 372, 373, 469 and 470 of the UBP1 amino-acid sequence (Figure 1A).
Determination of the activity of the UBI::UBP1 protease variants
In order to determine the protease activity of the UBP1 analogues, two hybrid proteins were obtained. To obtain the first one, the 354 bp long DNA fragment, encoding the C-terminal end of the UBP1 protease named S, was cloned into the pT7NHU plasmid (Figure 1A, B). The second one consists of yeast ubiquitin followed by: AspProGlyAspLysAspGlyAspGlyTyrIleSerAlaAlaGluAlaMetAla-, a peptide analogous to the IIId calcium-binding loop of calmodulin [10]. The sequence encoding this fusion peptide was obtained by ligation of the yeast ubiquitin gene with synthetic oligonucleotides: KalaG and KalaD (Table 1), and cloned into pT7U. Both plasmids were used to transform E. coli BLD21 (DE3) cells with the aim of obtaining the hybrid proteins 6xHisUBI::S and UBI::K (Table 2). Both fusion proteins were soluble during the synthesis in E. coli. The 6xHisUBI::S was purified by Ni-NTA affinity chromatographs. The UBI::K protein was purified by ion exchange chromatography on a column of DEAE-Sepharose Fast Flow (Pharmacia LKB), followed by NiCl affinity chromatography using Chelating Sepharose Fast Flow (Pharmacia LKB). In both cases, the expression level and purity of the hybrid proteins were high enough (data not shown) to be used for activity determination. The reactions were performed in a volume of 50 μl at 37°C for 30 min. in a buffer of the following composition: 20 mM phosphate pH 7.5, 2 mM DDT, 1 mM EDTA. 4 μg (380 pM) of the substrate UBI::K or 2 μg (87 pM) of 6xHisUBI::S were digested with 1.5 μg (18.2 pM) of the protease variants. The digestion reactions were stopped by heating at 100°C for 3 min. in the presence of SDS, and the digestion products were analyzed using SDS-PAGE (12%) (Figure 6). The unit (U) of enzyme activity is defined as the amount (of the enzyme) which will catalyze the transformation of 1 micromole of the substrate per minute under standard conditions.
Abbreviations used
aa – amino-acid residue
bp – base pair(s)
IPTG – isopropyl-β-D-galactoside
kDa – kilodalton
LB – Luria-Bertani
PCR – polymerase chain reaction
U – unit
UBI – ubiquitin
UBP – ubiquitin-specific protease (Ubp)
SDS – sodium dodecyl sulfate
SDS-PAGE – sodium dodecyl sulfate-polyacrylamide gel electrophoresis
Authors' contributions
AW carried out the molecular genetics studies, participated in the purification of the proteins and drafted the manuscript. AMP carried out the purification of the UBP1 analogues and helped to draft the manuscript. DMS participated in the construction of the UBPDQL gene. GP was involved in the purification of the UBPDQL protease. LCh constructed the synthetic gene of ubiquitin. NŁ was involved in the purification of the UBI::K protein. The study was conceived and coordinated by AP, who also helped to draft the manuscript.
All the authors read and approved the final manuscript.
Acknowledgements
This work was supported by State Committee for Scientific Research, Projekt Celowy nr 6T090462001 C/5592.
==== Refs
Baker R Protein expression using ubiquitin fusion and cleavage Curr Opin Biotechnol 1996 7 541 6 8939629 10.1016/S0958-1669(96)80059-0
Tobias J Varshavsky A Cloning and functional analysis of the ubiquitin-specific protease gene UBP1 of Saccharomyces cerevisiae J Biol Chem 1991 266 12021 8 2050695
Varshavsky A Tobias J Ubiquitin – specific protease 1991 Patent WO9117245.
Sonnhammer EL Eddy SR Birney E Bateman A Durbin R Pfam: multiple sequence alignments and HMM-profiles of protein domains Nucleic Acids Res 1998 26 320 2 9399864 10.1093/nar/26.1.320
Baker R Tobias T Varshavsky A Ubiquitin-specific proteases of Saccharomyces cerevisiae. Cloning of UBP2 and UBP3, and functional analysis of the UBP gene family J Biol Chem 1992 267 23364 75 1429680
Plucienniczak A Wojtowicz A Mikiewicz-Sygula D Plucienniczak G A UBP1 protease mutant, and its coding sequence, their application and methods of production 2004 Application of Patent WO 2004/097011.
Schmitz C Kinner A Kolling R The Deubiquitinating Enzyme Ubp1 Affects Sorting of the ATP-binding Cassette-Transporter Ste6 in the Endocytic Pathway Mol Biol Cell 2005
Birnboim HC Doly J A rapid alkaline extraction procedure for screening recombinant plasmid DNA Nucleic Acids Res 1979 24 1513 23 388356
Kane JF Effects of rare codon clusters on high-level expression of heterologous proteins in Escherichia coli Curr Opin Biotechnol 1995 6 494 500 7579660 10.1016/0958-1669(95)80082-4
Goch G Kozlowska H Wojtowicz A Bierzynski A A comparative CD and fluorescence study of a series of model calcium-binding peptides Acta Biochim Pol 1999 46 673 7 10698275
| 15924623 | PMC1156937 | CC BY | 2021-01-04 16:24:35 | no | Microb Cell Fact. 2005 May 30; 4:17 | utf-8 | Microb Cell Fact | 2,005 | 10.1186/1475-2859-4-17 | oa_comm |
==== Front
Mol CancerMolecular Cancer1476-4598BioMed Central London 1476-4598-4-181588245510.1186/1476-4598-4-18ResearchABCC5, ERCC2, XPA and XRCC1 transcript abundance levels correlate with cisplatin chemoresistance in non-small cell lung cancer cell lines Weaver David A [email protected] Erin L [email protected] Kristy A [email protected] Fadel [email protected] Sadik A [email protected] James C [email protected] Department of Medicine, Medical College of Ohio, 3055 Arlington Ave., Toledo, OH 43699, USA2005 9 5 2005 4 18 18 13 12 2004 9 5 2005 Copyright © 2005 Weaver et al; licensee BioMed Central Ltd.2005Weaver et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Although 40–50% of non-small cell lung cancer (NSCLC) tumors respond to cisplatin chemotherapy, there currently is no way to prospectively identify potential responders. The purpose of this study was to determine whether transcript abundance (TA) levels of twelve selected DNA repair or multi-drug resistance genes (LIG1, ERCC2, ERCC3, DDIT3, ABCC1, ABCC4, ABCC5, ABCC10, GTF2H2, XPA, XPC and XRCC1) were associated with cisplatin chemoresistance and could therefore contribute to the development of a predictive marker. Standardized RT (StaRT)-PCR, was employed to assess these genes in a set of NSCLC cell lines with a previously published range of sensitivity to cisplatin. Data were obtained in the form of target gene molecules relative to 106 β-actin (ACTB) molecules. To cancel the effect of ACTB variation among the different cell lines individual gene expression values were incorporated into ratios of one gene to another. Each two-gene ratio was compared as a single variable to chemoresistance for each of eight NSCLC cell lines using multiple regression. In an effort to validate these results, six additional lines then were evaluated.
Results
Following validation, single variable models best correlated with chemoresistance (p < 0.001), were ERCC2/XPC, ABCC5/GTF2H2, ERCC2/GTF2H2, XPA/XPC and XRCC1/XPC. All single variable models were examined hierarchically to achieve two variable models. The two variable model with the highest correlation was (ABCC5/GTF2H2, ERCC2/GTF2H2) with an R2 value of 0.96 (p < 0.001).
Conclusion
These results provide markers suitable for assessment of small fine needle aspirate biopsies in an effort to prospectively identify cisplatin resistant tumors.
==== Body
Background
Non-small cell lung cancer (NSCLC) is the most common type of bronchogenic carcinoma. Although chemotherapeutic regimens with greater efficacy continue to be developed, the best regimens presently give an overall response rate of only 30–50%. Lack of response is attributable to resistance that is present de novo or develops in response to treatment. If the resistance to drugs could be surmounted or if the most effective drug candidates for treatment could be better determined, the impact in terms of survival would be substantial. Because mechanisms of chemoresistance likely involve multiple gene products, we hypothesize that patterns of individual gene expression and/or indices comprising the expression values of multiple genes will provide more effective markers of chemoresistant NSCLC tumors than values of individual genes.
Currently, cisplatin and carboplatin are among the most widely used cytotoxic anticancer drugs. However, resistance to these drugs through de novo or induced mechanisms undermines their curative potential [1]. Recently, understanding regarding potential modes of chemoresistance to platinum compounds has been obtained through studies correlating cytotoxicity with nucleotide excision-repair (NER) [2-7] or drug uptake/efflux [7-13]. In this study, we investigated whether de novo gene expression differences are correlated with a predisposition of NSCLC tumors to chemoresistance.
Current advances in technology, including microarrays and quantitative RT-PCR methods, enable classification of cancer types on the basis of TA levels rather than histomorphology [14,15]. For example, these techniques enable the discovery of predictive markers based on TA profiles. Microarray screening analysis currently is being investigated to predict chemotherapeutic sensitivity based on TA profiles [16-18]. An advantage of microarray analysis is that thousands of genes may be simultaneously evaluated. However, it is generally recognized that, due to lack of standardization, relatively low sensitivity and relatively poor lower thresholds of detection, microarray assessments need to be confirmed with follow-up quantitative methods. StaRT-PCR is a method that enables rapid, sensitive, reproducible, standardized, quantitative measurements for many genes simultaneously [19,49,50].
Briefly, in StaRT-PCR, the TA level of each gene is made relative to an internal standard (IS) within a standardized mixture of internal standards (SMIS). Known concentrations of these mixtures are combined with cDNA samples in a master mixture for PCR amplification. This enables quantitative measurement of gene expression while controlling for inter-sample, inter-experimental and loading differences. With StaRT-PCR, due to the presence of the SMIS, the measurements are quantitative and quality-controlled when measured either kinetically or at endpoint [51,52]. In other words, measurement of each TA value relative to a known quantity of internal standard controls for variation in amplification efficiency in early, log-linear, and plateau phases of PCR [53].
In an initial survey, StaRT-PCR was used to measure expression of 35 genes involved in DNA repair, multi-drug resistance, cell cycling and apoptosis in two cell lines previously reported to be the least (H460) and most (H1435) chemoresistant among 20 NSCLC cell lines [20]. It was determined that genes involved in DNA repair (ERCC2, XRCC1) and drug influx/efflux (ABCC5) were associated with chemoresistance. The number of genes from each of these two categories was expanded to include additional representative genes associated with generalized DNA damage recognition and repair (DDIT3), associated specifically with NER (LIG1, ERCC3, GTF2H2, XPA, XPC), or associated with drug transport (ABCC1, ABCC4, ABCC10). Expression of these twelve genes was measured in eight NSCLC cell lines with variable cisplatin resistance [20]. StaRT-PCR data were obtained using ACTB as a reference gene. Thus, data were reported in the form of mRNA molecules/106 ACTB molecules. These data then were combined into interactive transcript abundance indices (ITAI) by placing one or more genes directly associated with the phenotype on the numerator and one or more genes negatively associated with the phenotype on the denominator [19,21]. It is reasonable to expect that optimal predictors of phenotypes are more likely to be discovered among ITAI than among expression levels of individual genes. This has been demonstrated for certain cancer-related phenotypes [19,21-23]. A further advantage of ITAI is that they control for previously observed variation in the reference gene value (in this case, ACTB) from one cell line to another [19,21]. When a single gene in the numerator is divided by another single gene in the denominator, the reference value mathematically cancels out. The ITAI values were compared to cisplatin chemoresistance among the eight NSCLC cell lines with variable resistance. Results then were validated in an additional six NSCLC cell lines.
Results
Reproducibility
Among the gene expression measurements for which three or more replicate values were obtained, the mean coefficient of variation was 38.5% (see Additional file 1). This is similar to the reproducibility observed in other gene expression studies using the StaRT-PCR method [19,22]. Recently, through implementation of robotic liquid handlers, automation software, and standard operating procedures in the NCI funded (CA95806) Standardized Expression Measurement (SEM) Center, variation among replicates has been reduced to a CV of less than 10% [50].
Individual gene expression measurements and chemoresistance
The results of the direct comparison of individual gene expression mean values versus cisplatin chemoresistance are presented in Table 1. All StaRT-PCR data values were in the form of molecules/106 ACTB molecules (see Additional file 1). For 8/12 genes assessed, the correlation was significant (p < 0.05).
Table 1 Correlation between each of twelve putative chemoresistance transcript abundance values and chemoresistance among NSCLC cell lines. Eight of the 12 selected multi-drug resistance and DNA repair genes were significantly correlated with chemoresistance among the first group of 8 NSCLC cell lines evaluated (Group 1). In order to validate these results, an additional six lines were evaluated (Group 2).
Group 1a Groups 1 and 2b
Gene R2 p value R2 p value
ABCC5 0.93 0.0001 0.85 0.0001
ERCC2 0.85 0.0012 0.82 0.0001
XPA 0.81 0.0023 0.75 0.0001
LIG1 0.78 0.0034 NAc NA
DDIT3 0.68 0.0120 NA NA
ABCC1 0.64 0.0164 NA NA
XRCC1 0.56 0.0318 0.57 0.0019
ERCC3 0.56 0.0319 NA NA
ABCC4 0.37 0.1081 NA NA
ABCC10 0.06 0.5641 NA NA
GTF2H2 0.02 0.7424 0.04 0.5183
XPC 0.00 0.9851 0.02 0.6654
a Initial 8 NSCLC cell lines analyzed. b All 14 NSCLC cell lines analyzed. cNot Assessed.
Establishment of inter-active transcript abundance indices
ITAI were established as balanced ratios comprising every possible combination with one gene divided by the TA value of another gene for data obtained from each of the initial eight NSCLC cell lines (Group 1). Each TA value was calculated as molecules/106 ACTB molecules. Thus, in these ITAI, the effect of the reference gene, ACTB, is cancelled. For example: ERCC2 molecules/106 ACTB molecules ÷ XPC molecules/106 ACTB molecules = ERCC2 molecules/XPC molecules. Bivariate analysis of each two-gene ratio versus corresponding cisplatin IC50 chemoresistance value was conducted among the eight cell lines (see Additional file 2). There were 12 genes assessed and 11 sets of ratios for each gene as the numerator resulting in 132 ratios. The data from bivariate analyses then were ranked in descending order such that the ratio set listed first was that for which the mean value for correlation with chemoresistance was highest, and the ratio set listed last was that for which the mean r value for correlation with chemoresistance was lowest. Thus, the ratio set with ERCC2 in the numerator is listed first because the mean r value for the ratios between ERCC2 and each of the other eleven genes was the most positive among the twelve genes evaluated. In contrast, the ratio set with XPC in the numerator is listed last because the ratios between XPC and each of the other 11 genes had the most negative correlation with chemoresistance.
Modelling of gene expression with chemoresistance
The ratios ERCC2/XPC, ABCC5/GTF2H2, ERCC2/XRCC1, ERCC2/GTF2H2, XPA/XPC, XRCC1/XPC, and ABCC5/XPC were the best (i.e. those single variable models with highest R2 identified in the initial eight NSCLC cell lines by simple linear regression (see Additional file 2). The effect of adding a second variable into the model was then assessed. The best two variable model was (ABCC5/GTF2H2, ERCC2/GTF2H2) with an R2 value of 0.96.
Validation of Models
We tested our single and two variable models in an additional six NSCLC cell lines (Table 2). In statistical analysis of the combined data for all 14 NSCLC cell lines, the p value improved or stayed the same for three of the single variable models (ERCC2/XPC, ABCC5/GTF2H2, XRCC1/XPC), as well as the two variable model. The decline in p value for ERCC2/GTF2H2 and XPA/XPC was not significant. In contrast, ERCC2/XRCC1 was no longer significantly associated with chemoresistance, and the p value declined substantially for ABCC5/XPC.
Table 2 Bivariate correlation between two-transcript abundance ratios and chemoresistance in a validation set of NSCLC cell lines. The ratios best correlated with chemoresistance from Additional file 2 were evaluated in an additional six lines. The effect of adding a second variable into the model was assessed. ERCC2/XPC was no longer correlated with chemoresistance and ABCC5/XPC had substantially lower p value. The other single variable models and the two variable model were validated.
Model Group 1a Groups 1 and 2b
Single variable Model r2 p value Model r2 p value
ERCC2/XPC 0.91 0.0002 0.69 0.0002
ABCC5/GTF2H2 0.91 0.0002 0.88 0.0001
ERCC2/XRCC1 0.90 0.0003 NSc
ERCC2/GTF2H2 0.90 0.0004 0.63 0.0007
XPA/XPC 0.89 0.0004 0.62 0.0008
XRCC1/XPC 0.88 0.0005 0.75 0.0001
ABCC5/XPC 0.88 0.0006 0.29 0.0461
Two variable Model r2 p value Model r2 p value
ABCC5/GTF2H2 and ERCC2/GTF2H2 0.96 0.0003 0.91 0.0001
a Initial 8 NSCLC cell lines analyzed. b All 14 NSCLC cell lines analyzed. cnot significant (p > 0.05).
Discussion
The results obtained by measuring gene expression with StaRT-PCR, incorporating values for individual genes into ITAI, and correlating ITAI with chemoresistance led us to propose several models as potential predictors of cisplatin chemoresistance in cultured NSCLC cells. These models comprise genes that have been associated with cisplatin chemoresistance in previous studies including ABCC5 [13], and XPA [4,24].
Experimental results suggest that increased expression of ABCC5, also known as MRP5, is associated with exposure to platinum drugs in lung cancer in vivo and/or the chronic stress response to xenobiotics [13]. Thus, increased resistance to platinum drugs with increased ABCC5 levels may be due to glutathione S-platinum complex efflux. Increased efflux of platinum drugs could result in lower levels of drug available to form damaging DNA-platinum drug adducts.
XPA and ERCC2 are components of the nucleotide excision repair (NER) mechanism, which generally is recognized as the major repair response to DNA damage induced by chemotherapeutic agents such as cisplatin [1,3,7]. In NER, XPA is the main DNA lesion recognition protein [25], is the key element in assembly of the NER complex by recruiting several other proteins to the lesion site [26] and XPA levels are rate-limiting for NER [4,27]. Enhanced NER gene expression is a major cause of resistance to cisplatin and other DNA-damaging chemotherapeutic agents [3,28] and over expression of the XPA gene component of NER has been associated with resistance to cisplatin in human ovarian cancer [4,24]. ERCC2 specifically is a component of the transcription factor IIH (TFIIH) that consists of seven polypeptides [29,30] and in its entirety is a repair factor [31-33]. In NER, ERCC2 (or XPD) is essential for TFIIH helicase activity [34] and it has been demonstrated more recently that ERCC2 interacts specifically with GTF2H2 (or p44) and that this interaction results in the stimulation of the 5' to 3' helicase activity [35]. In at least some other tissues, ERCC1 is associated with cisplatin resistance, while ERCC2 is not [36,37]. Thus, our data support the importance of excision repair in cisplatin resistance, but suggest that there is inter-tissue variation in the excision repair genes that are responsible for de novo cisplatin resistance.
XRCC1 has long been recognized as a key component of the base excision repair (BER) pathway, acting as a "scaffold" for the coordination of other BER proteins at the sites of base damage during repair [38-40]. It has been shown that polymorphisms in XRCC1, while in themselves are not associated with increased risk of lung cancer, have shown an increased risk of lung cancer in a supermultiplicative manner when associated with polymorphisms in another component of BER, poly (ADP-ribose) polymerase family, member 1 transfersase (PARP1) [41]. XRCC1 has also recently been proposed as a component of an alternative nonhomologous end-joining route of DNA double-stranded breaks (DSBs), that complements the predominant repair pathway of DNA-dependent protein kinase (DNA-PK) and X-ray repair complementing defective repair in Chinese hamster cells 4 (XRCC4)-DNA ligase IV complex [42]. Although the NER pathway is the major repair mechanism for cisplatin-DNA adducts, our data supports the proposal of overlapping repair pathways involved in alternative repair of cisplatin adducts, such as the BER pathway. XRCC1 may also be involved in the repair of other types of DNA damage caused by cisplatin including DSBs.
Selection of a stable reference for the amount of sample loaded for each gene expression measurement is important to ensure measurement accuracy and reproducibility. With microarray analysis, because thousands of genes are assessed simultaneously, an index of all genes measured provides a stable reference for the amount of sample loaded from one microarray to another. In quantitative RT-PCR studies, typically, a single non-regulated gene is used as a loading reference, such as ACTB, GAPD, cyclophilin or ribosomal RNA. However, all of these genes have been reported to vary among multiple samples. One way to assess inter-sample variation in reference gene expression among multiple samples is to compare variation between two reference genes. In our experience, ACTB and GAPD vary 50-fold relative to each other among bronchial epithelial cells (BEC) and even more between BEC and other cell types [19,44]. In situations where limited numbers of genes are measured (< 200), an index of all genes for the normalization of data is not sufficiently stable. In order to eliminate the effect of unknown variation in the reference gene expression among samples, we analyzed balanced ratios of one gene expression value obtained by StaRT-PCR to another. These balanced ratios did not represent actual cellular concentration changes of the individual genes comprising the ratio, but related the expression of one gene to another and could be used for comparison with phenotypic determinants such as chemoresistance. In this study, ITAI analysis (Table 2) confirmed most of the results obtained by analysis of individual gene expression values relative to chemoresistance (Table 1). This suggests that variation in ACTB among this group of cDNA samples was not significant. However, in our experience inter-sample variation in ACTB expression is greater among primary samples. Thus, we will continue to use ITAI to remove doubt regarding potential effect of variation in reference gene expression whenever possible.
As is presented in Table 2, by evaluating an empirically derived set of balanced ratios (ITAI) derived from expression values for all of the genes measured, it is possible to establish a hierarchy regarding the strength of association between a set of genes and a phenotype.
Conclusion
In summary, the association of ERCC2, ABCC5, XPA, and XRCC1 with chemoresistance was established through a sequential process involving a) screening genes representing many different functional classes, b) evaluating an expanded group of genes represented by those that were positively associated in the first round, c) identification of outliers (see Additional file 2), d) model building and e) validation (Table 2). Although only two of the 35 genes assessed in the first round were correlated with chemoresistance, 8/12 of the selected DNA repair and MDR genes were correlated. The models established in this study demonstrate the importance of evaluating the interaction among multiple genes representing multiple pathways involved in cisplatin chemoresistance. These models will be tested through a blinded study of gene expression levels of the identified potential markers in samples consisting of fine needle aspirate (FNA) biopsies from patients with various treatment outcomes.
Methods
Cell culture
Non-small cell lung cancer (NSCLC) cell lines with varying levels of cisplatin chemoresistance, H460, H1155, H23, H838, H1334, H1437, H1355, H1435, H358, H322, H441, H522, H226 and H647, were obtained from the American Type Culture Collection (Rockville, MD). The previously reported [20] cisplatin IC50 concentration for each line is provided in Table 3. All cells were incubated in RPMI-1640 medium (Biofluids, Inc., Rockville, MD) containing 10% fetal bovine serum (FBS) and 1 mM glutamine at 37°C in the presence of 5% CO2. Proliferative, subconfluent cultures were obtained for RNA extractions and subsequent analyses.
Table 3 Cisplatin chemoresistance in 14 non-small cell cancer cell lines
NSCLC cell line Cisplatin IC50 (μM)
Group 1b NCI-H460 0.52 ± 0.04
NCI-H1155 0.90 ± 0.20
NCI-H23 2.09 ± 0.32
NCI-H838 3.86 ± 0.31
NCI-H1334 5.15 ± 0.47
NCI-H1437 5.90 ± 1.61
NCI-H1355 6.74 ± 1.29
NCI-H1435 22.86 ± 2.36
Group 2c NCI-H358 1.16 ± 0.24
NCI-H322 2.85 ± 0.22
NCI-H441 3.38 ± 0.99
NCI-H522 3.53 ± 0.56
NCI-H226 5.05 ± 0.95
NCI-H647 7.27 ± 1.14
a Previously published results by Chun-Ming Tsai, et.al. [20]
b Initial set of 8 NSCLC cell lines evaluated.
c Additional set of 6 NSCLC cell lines evaluated.
Reagents
10X PCR buffer for the Rapidcycler (500 mM Tris, pH 8.3; 2.5 mg/μl BSA; 30 mM MgCl2) was obtained from Idaho Technology, Inc. (Idaho Falls, ID). Taq polymerase (5 U/μl), oligo dT primers, RNasin (25 U/μl) and dNTPs were obtained from Promega (Madison, WI). M-MLV reverse transcriptase (200 U/μl) and 5X first strand buffer (250 mM Tris-HCl, pH 8.3; 375 mM KCl; 15 mM MgCl2; 50 mM DTT) were obtained from Gibco BRL (Gaithersburg, MD). DNA 7500 Assay kits containing dye, matrix and standards were obtained from Agilent Technologies, Inc. (Palo Alto, CA). All other chemicals and reagents were molecular biology grade.
RNA extraction and reverse transcription
Total RNA was isolated from cell cultures by a TriReagent protocol (Molecular Research Center, Inc., Cincinnati, OH) [43]. Following extraction, approximately 1 μg of total RNA for each cell line was reverse-transcribed using M-MLV reverse-transcriptase and an oligo dT primer as previously described [44].
Quantitative standardized RT (StaRT)-PCR
Gene expression was determined using previously published quantitative StaRT-PCR protocols [19,44-50]. Briefly, a master mixture containing buffer, MgCl2, dNTPs, sample cDNA, Taq polymerase and SMIS was prepared and 9 μl aliquots dispensed into 0.6 ml microfuge tubes containing 1 μl of gene-specific primers. A SMIS comprises gene-specific IS's for each gene at defined concentrations relative to one another. The mixture includes IS's for reference (or housekeeping genes) to control for cDNA loading and to simplify normalization of all gene data. All primers used for PCR and those used in the construction of the CTs, are listed in Additional file 3. PCR reactions mixtures were subjected to 35 cycles of PCR with 5 seconds of denaturation at 94°C, 10 seconds of annealing at 58°C and 15 seconds of elongation at 72°C in a Rapidcycler (Idaho Technology, Inc.). PCR products were electrophoretically separated and quantified in an Agilent 2100 Bioanalyzer (Agilent Technologies, Inc.) with the DNA 7500 Assay kit. The area under the curve (as calculated by Agilent software) for each native template (NT) and IS peak was used in all calculations. Representative electropherograms of each gene assessed are presented in Additional file 4. The NT/IS ratio for a reference gene, ACTB, and the NT/IS ratios for each target gene were calculated. The initial number of NT molecules for each gene then could be determined from these ratios because the initial number of IS molecules added into the PCR reaction was known. To normalize measurements and control for sample-to-sample variation and inter-experimental loading, the calculated number of target gene molecules was divided by the calculated number of ACTB molecules. A size correction was employed to correct for fluorescence intensity differences affecting the measured area under the curve [19,48].
Statistical analyses
Ratios of one gene to another, from each of the initial eight NSCLC cell lines, were subjected to multiple regression analysis using SAS 6.12 (SAS Institute Inc., Cary, NC) to determine the combination of genes that best predict cisplatin resistance. Each ratio was compared separately to chemoresistance and ratios with significant correlation to resistance (R2 ≥ 0.88, p < 0.001) then were examined hierarchically to achieve two variable models based on the highest R2 values. Following assessment of an additional 6 cell lines, results for all 14 NSCLC cell lines were combined and also subjected to analysis as described.
Abbreviations
LIG1 – DNA ligase I, ATP-independent; ERCC2 – excision repair cross-complementing rodent repair deficiency (ERCC), complementation group 2; ERCC3 – ERCC, complementation group 3; DDIT3 – DNA-damage-inducible transcript 3; ABCC1 – ATP-binding cassette, subfamily C (ABCC), member 1; ABCC4 – ABCC, member 4; ABCC5 – ABCC, member 5; ABCC10 – ABCC, member 10; GTF2H2 – general transcription factor IIH, polypeptide 2; XPA – xeroderma pigmentosum (XP), complementation group A; XPC – XP, complementation group C and XRCC1 – X-ray repair complementing defective repair in Chinese hamster cells 1.
Authors' contributions
DAW participated in study design, cell culture, transcript abundance analysis and data interpretation and drafted the manuscript. ELC participated in data interpretation and manuscript preparation. KAW participated in cell culture and study design. FE conducted gene expression experiments. SAK conducted statistical analyses and participated in study design and data interpretation. JCW conceived of the study, participated in study design and data interpretation, and critically reviewed the manuscript.
Competing interests
DAW, ELC, KAW and JCW each have significant equity interest in Gene Express, Inc. which produces and markets StaRT-PCR reagents used in these studies.
Supplementary Material
Additional File 1
StaRT-PCR Average Transcript Abundance Data
Click here for file
Additional File 2
Bivariate correlation between two-transcript abundance ratios and chemoresistance in the eight NSCLC lines of Group 1
Click here for file
Additional File 3
Primers used for StaRT-PCR analysis
Click here for file
Additional File 4
Representative electropherograms for detminations of transcript abundance
Click here for file
==== Refs
Perez RP Cellular and molecular determinants of cisplatin resistance Eur J Cancer 1998 34 1535 1542 9893624 10.1016/S0959-8049(98)00227-5
Dijt F Fitchinger-Schepman AM Berends F Reedijk J Formation and repair of cisplatin-induced adducts to DNA in cultured normal and repair-deficient human fibroblasts Cancer Res 1988 48 6058 6062 3167856
Zamble DB Lippard SJ Cisplatin and DNA repair in cancer chemotherapy Trends Biochem Sci 1995 20 435 439 8533159 10.1016/S0968-0004(00)89095-7
States JC Reed E Enhanced XPA mRNA levels in cisplatin-resistant human ovarian cancer are not associated with XPA mutations or gene amplifications Cancer Lett 1996 108 233 237 8973600 10.1016/S0304-3835(96)04428-X
Ferry KV Fink D Johnson SW Hamilton TC Howell SB Quantitation of platinum-DNA adduct repair in mismatch repair deficient and proficient human colorectal cancer cell lines using an in vitro DNA repair assay [abstract] Proc Am Assoc Cancer Res 1997 38 359
Jordan P Carmo-Fonseca M Molecular mechanisms involved in cisplatin cytotoxicity Cell Mol Life Sci 2000 57 1229 1235 11028915
Kartalou M Essigmann JM Mechanisms of resistance to cisplatin Mutat Res 2001 478 23 43 11406167
Berger W Elbling L Hauptmann E Micksche M Expression of the multidrug resistance-associated protein (MRP) and chemoresistance of human non-small-cell lung cancer cells Int J Cancer 1997 73 84 93 9334814 10.1002/(SICI)1097-0215(19970926)73:1<84::AID-IJC14>3.0.CO;2-5
Borst P Kool M Evers R Do cMOAT (MRP2), other MRP homologues, and LRP play a role in MDR? Cancer Biol 1997 8 205 213 10.1006/scbi.1997.0071
Young LC Campling BG Voskoglou-Nomikos T Cole SPC Deeley RG Gerlach JH Expression of multidrug resistance protein-related genes in lung cancer: correlation with drug response Clin Cancer Res 1999 5 673 680 10100721
Berger W Elbling L Micksche M Expression of the major vault protein LRP in human non-small-cell lung cancer cells: activation by short-term exposure to antineoplastic drugs Int J Cancer 2000 88 293 300 11004683 10.1002/1097-0215(20001015)88:2<293::AID-IJC23>3.3.CO;2-J
Borst P Evers R Kool M Wijnholds J A family of drug transporters: the multidrug resistance-associated proteins J Nat Cancer Inst 2000 92 1295 1302 10944550 10.1093/jnci/92.16.1295
Oguri T Isobe T Suzuki T Nishio K Fujiwara Y Katoh O Yamakido M Increased expression of the MRP5 gene is associated with exposure to platinum drugs in lung cancer Int J Cancer 2000 86 95 100 10728601 10.1002/(SICI)1097-0215(20000401)86:1<95::AID-IJC15>3.0.CO;2-G
Golub TR Slonim DK Tamayo P Huard C Gaasenbeek M Mesirov JP Coller H Loh ML Downing JR Caligiuri MA Bloomfield CD Lander ES Molecular classification of cancer: class discovery and class prediction by gene expression monitoring Science 1999 286 531 537 10521349 10.1126/science.286.5439.531
Alizadeh AA Eisen MB Davis RE Ma C Lossos IS Rosenwald A Boldrick JC Sabet H Tran T Yu X Powell JI Yang L Marti GE Moore T Hudson J JrLu L Lewis DB Tibshirani R Sherlock G Chan WC Greiner TC Weisenburger DD Armitage JO Warnke R Levy R Wilson W Grever MR Byrd JC Botstein D Brown PO Staudt LM Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling Nature 2000 403 503 511 10676951 10.1038/35000501
Scherf U Ross DT Waltham M Smith LH Lee JK Tanabe L Kohn KW Reinhold WC Myers TG Andrews DT Scudiero DA Eisen MB Sausville EA Pommier Y Botstein D Brown PO Weinstein JN A gene expression database for the molecular pharmacology of cancer Nat Genet 2000 24 236 244 10700175 10.1038/73439
Kihara C Tsunoda T Tanaka T Yamana H Furukawa Y Ono K Kitahara O Zembutsu H Yanagawa R Hirata K Takagi T Nakamura Y Prediction of sensitivity of esophageal tumors to adjuvant chemotherapy by cDNA microarray analysis of gene-expression profiles Cancer Res 2001 61 6474 6479 11522643
Zembutsu H Ohnishi Y Tsunoda T Furukawa Y Katagiri T Ueyama Y Tamaoki N Nomura T Kitahara O Yanagawa R Hirata K Nakamura Y Genome-wide cDNA microarray screening to correlate gene expression profiles with sensitivity of 85 human cancer xenografts to anticancer drugs Cancer Res 2002 62 518 527 11809704
Willey JC Crawford EL Jackson CM Weaver DA Hoban JC Khuder SA DeMuth JP Expression measurement of many genes simultaneously by quantitative RT-PCR using standardized mixtures of competitive templates Am J Respir Cell Mol Biol 1998 19 6 17 9651175
Tsai CM Chang KT Wu LH Chen JY Gazdar AF Mitsudomi T Chen MH Perng RP Correlations between intrinsic chemoresistance and HER-2/neu gene expression, p53 mutations, and cell proliferation characteristics in non-small cell lung cancer cell lines Cancer Res 1996 56 206 209 8548764
DeMuth JP Jackson CM Weaver DA Crawford EL Durzinsky DS Durham SJ Zaher A Phillips ER Khuder SA Willey JC The gene expression index c-myc x E2F1/p21 is highly predictive of malignant phenotype in human bronchial epithelial cells Am J Respir Cell Mol Biol 1998 19 18 24 9651176
Crawford EL Khuder SA Durham SJ Frampton M Utell M Thilly WG Weaver DA Ferencak WJ Jennings CA Hammersley JR Olson DA Willey JC Normal bronchial epithelial cell expression of glutathione transferase P1, glutathione transferase M3, and glutathione peroxidase is low in subjects with bronchogenic carcinoma Cancer Res 2000 60 1609 1618 10749130
Rots JG Willey JC Jansen G Van Zantwijk CH Noordhuis P DeMuth JP Kuiper E Veerman AJ Pieters R Peters GJ mRNA expression levels of methotrexate resistance-related proteins in childhood leukemia as determined by a standardized competitive template-based RT-PCR method Leukemia 2000 14 2166 2175 11187907 10.1038/sj.leu.2401943
Dabholkar M Vionnet J Bostick-Bruton F Yu JJ Reed E Messenger RNA levels of XPAC and ERCC1 in ovarian cancer tissue correlate with response to platinum-based chemotherapy J Clin Invest 1994 94 703 708 8040325
Asahina H Kuraoka I Shirakawa M Morita EH Miura N Miyamoto I Ohtsuka E Okada Y Tanaka K The XPA protein is a zinc metalloprotein with an ability to recognize various kinds of DNA damage Mutat Res DNA Repair 1994 315 229 237 7526200 10.1016/0921-8777(94)90034-5
Li L Peterson CA Lu X Legerski RJ Mutations in XPA that prevent association with ERCC1 are defective in nucleotide excision repair Mol Cell Biol 1995 15 1993 1998 7891694
Mu D Park CH Matsunaga T Hsu DS Reardon JT Sancar A Reconstitution of human DNA repair excision nuclease in a highly defined system J Biol Chem 1995 270 2415 2418 7852297 10.1074/jbc.270.6.2415
Reed E Devita VT Jr, Hellman S, Rosenberg SA Anticancer drugs: platinum analogs Principles and Practice of Oncology 1993 Philadelphia, PA: Lippincott 390 399
Mu D Hsu DS Sancar A Reaction mechanism of human DNA repair excision nuclease J Biol Chem 1996 271 8285 8294 8626523 10.1074/jbc.271.32.19451
Schaeffer L Moncollin V Roy R Staub A Mezzina M Sarasin A Weeda G Hoeijmakers JH Egly JM The ERCC2/DNA repair protein is associated with the class II BTF2/TFIIH transcription factor EMBO J 1994 13 2388 2392 8194528
Drapkin R Reardon JT Ansari A Huang JC Zawel L Ahn K Sancar A Reinberg D Dual role of TFIIH in DNA excision repair and in transcription by RNA polymerase II Nature 1994 368 769 772 8152490 10.1038/368769a0
Wang Z Svejstrup JQ Feaver WJ Wu X Kornberg RD Friedberg EC Transcription factor b (TFIIH) is required during nucleotide-excision repair in yeast Nature 1994 368 74 76 8107888 10.1038/368074a0
Prakash S Sung P Prakash L DNA repair genes and proteins of Saccharomyces cerevisiae Annu Rev Genet 1993 27 33 70 8122907 10.1146/annurev.ge.27.120193.000341
Coin F Marinoni JC Rodolfo C Fribourg S Pedrinin AM Egly JM Mutations in the XPD helicase gene result in XP and TTD phenotypes, preventing interaction between XPD and the p44 subunit of TFIIH Nature Genet 1998 20 184 188 9771713 10.1038/2491
Yen L Woo A Christopoulopoulos G Batist G Panasci L Roy R Mitra S Alaoui-Jamili MA Enhanced host cell reactivation capacity and expression of DNA repair genes in human breast cancer cells resistant to bi-functional alkylating agents Mutation Research 1995 337 179 189 7491121 10.1016/0921-8777(95)00022-C
Damia G Guidi G D'Incalci M Expression of genes involved in nucleotide excision repair and sensitivity to cisplatin and melphalan in human cancer cell lines European Journal of Cancer 1998 34 1783 1788 9893669 10.1016/S0959-8049(98)00190-7
Dabholkar M Bostick-Bruton F Weber C Bohr VA Egwuagu C Reed E ERCC1 and ERCC2 expression in malignant tissues from ovarian cancer patients Journal of the National Cancer Institute 1992 84 1512 1517 1433335
Hoeijmakers JHJ Genome maintenance mechanisms for preventing cancer Nature 2001 411 366 374 11357144 10.1038/35077232
Wood RD Mitchell M Sgouros J Lindahl T Human DNA Repair Genes Science 2001 291 1284 1289 11181991 10.1126/science.1056154
Sancar A Lindsey-Boltz LA Ünsal-KaÇmaz K Linn S Molecular mechanisms of mammalian DNA repair and the DNA damage checkpoints Annu Rev Biochem 2004 73 39 85 15189136 10.1146/annurev.biochem.73.011303.073723
Zhang X Xiaoping M Liang G Hao B Wang Y Tan W Li Y Guo Y He F Wei Q Lin D Polymorphisms in DNA base excision repair genes ADPRT and XRCC1 and risk of lung cancer Cancer Res 2005 65 722 726 15705867
Audebert M Salles B Calsou P Involvement of Poly (ADP-ribose) Polymerase-1 and XRCC1/DNA Ligase III in an alternative route for DNA double-strand breaks rejoining J Biol Chem 2004 279 55117 55126 15498778 10.1074/jbc.M404524200
Chomczynski P A reagent for the single-step simultaneous isolation of RNA, DNA and proteins from cell and tissue samples Biotechniques 1993 15 536 537 7692896
Willey JC Coy E Brolly C Utell MJ Frampton MW Hammersley J Thilly WG Olson D Cairns K Xenobiotic metabolism enzyme gene expression in human bronchial epithelial and alveolar macrophage cells Am J Respir Cell Mol Biol 1996 14 262 271 8845177
Apostolakos MJ Schuermann WH Frampton MW Utell MJ Willey JC Measurement of gene expression by multiplex competitive polymerase chain reaction Anal Biochem 1993 213 277 284 8238902 10.1006/abio.1993.1421
Willey JC Coy EL Frampton MW Torres A Apostolakos MJ Hoehn G Schuermann WH Thilly WG Olson DE Hammersley JR Crespi CL Utell MJ Quantitative RT-PCR measurement of cytochromes p450 1A1, 1B1, and 2B7, microsomal epoxide hydrolase, and NADPH oxidoreductase expression in lung cells of smokers and non-smokers Am J Respir Cell Mol Biol 1997 17 114 124 9224217
Celi FS Zenilman ME Shuldiner AR A rapid and versatile method to synthesize internal standards for competitive PCR Nucleic Acids Research 1993 21 1047 8451177
Crawford EL Peters GJ Noordhuis P Rots MG Vondracek M Grafström RC Lieuallen K Lennon G Zahorchak RJ Georgeson MJ Wali A Lechner JF Fan PS Kahaleh MB Khuder SA Warner KA Weaver DA Willey JC Reproducible gene expression measurement among multiple laboratories obtained in a blinded study using Standardized RT (StaRT)-PCR Molecular Diagnosis 2001 6 217 225 11774186 10.1054/modi.2001.29789
Willey JC Crawford EL Knight CA Warner KA Motten CR Herness Peters E Zahorchak RJ Graves TG Bergman JR Weaver DA Vondracek M Grafström RC Bustin SA Use of standardized mixtures of internal standards in quantitative RT-PCR to ensure quality control and develop a standardized gene expression database A-Z of Quantitative PCR 2004 International University Line, La Jolla, CA 545 576
Willey JC Crawford EL Knight CR Warner KA Motten CA Herness EA Zahorchak RJ Graves TG Standardized RT-PCR and the standardized expression measurement center Methods Mol Biol 2004 258 13 41 14970455
Pagliarulo V George B Beil SJ Groshen S Laird PW Cai J Willey J Cote RJ Datar RH Sensitivity and reproducibility of standardized-competitive RT-PCR for transcript quantification and its comparison with real time RT-PCR Mol Cancer 2004 3 5 14741054 10.1186/1476-4598-3-5
Millson A Suli A Hartung L Kunitake S Bennett A Nordberg MCL Hanna W Wittwer CT Arun S Lyon E Comparison of two quantitative polymerase chain reaction methods for detecting HER2/neu amplification J Mol Diagn 2003 184-90
Gilliland G Perrin S Blanchard K Bunn HF Analysis of cytokine mRNA and DNA: Detection and quantification by competitive polymerase chain reaction Proc Natl Acad Sci USA 1990 87 2725 2729 2181447
| 15882455 | PMC1156938 | CC BY | 2021-01-04 16:36:35 | no | Mol Cancer. 2005 May 9; 4:18 | utf-8 | Mol Cancer | 2,005 | 10.1186/1476-4598-4-18 | oa_comm |
==== Front
Mol PainMolecular Pain1744-8069BioMed Central London 1744-8069-1-181591890010.1186/1744-8069-1-18MethodologySpared nerve injury rats exhibit thermal hyperalgesia on an automated operant dynamic thermal escape Task Baliki Marwan [email protected] Oscar [email protected] Dante R [email protected] A Vania [email protected] Department of Physiology, Northwestern University Medical School, Chicago, IL, 60611, USA2005 26 5 2005 1 18 18 8 4 2005 26 5 2005 Copyright © 2005 Baliki et al; licensee BioMed Central Ltd.2005Baliki et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Well-established methods are available to measure thermal and mechanical sensitivity in awake behaving rats. However, they require experimenter manipulations and tend to emphasize reflexive behaviors. Here we introduce a new behavioral test, with which we examine thermal sensitivity of rats with neuropathic injury. We contrast thermal hyperalgesia between spared nerve injury and chronic constriction injury rats. This device is a fully automated thermal sensitivity assessment tool designed to emphasize integrated learned responses to thermal painful and non-painful stimuli that are applied dynamically to a surface on which the animal is standing. It documents escape behavior in awake, unrestrained animals to innocuous and noxious heating of the floor where the animal is located. Animals learn to minimize pain by escaping to the opposite non-heated side; escape latency is recorded. On this device, thermal stimulus-response curves showed > 6°C leftward shift in both groups of neuropathic rats. In contrast, when these animals were tested on hotplate the stimulus-response shift was < 2°C. Spared nerve injury rats showed even less evidence for thermal hyperalgesia when thermal sensitivity was tested by measuring paw withdrawal to infrared heating, plantar test. The implications of test dependent magnitude of thermal hyperalgesia are discussed from the viewpoint of the tests used, as well as the animal models studied. It is argued that the dynamic thermal operant task reveals the relevance of the neuropathic injury associated pain-like behavior in relation to the whole organism.
Neuropathic PainSpared Nerve InjuryChronic Constriction InjuryAllodyniaPain Behaviorratmouse
==== Body
1. Introduction
The discovery of animal models that exhibit different elements of clinical pain syndromes, coupled with advances in tools for quantification of pain behavior, have greatly advanced understanding of mechanisms involved in acute and chronic pain. Pain in animals can only be determined by evaluating behavioral cues. Animal studies measure two types of pain behavior: simple withdrawal reflexes and complex voluntary and intentional behaviors [1]. Various methods have been used in assessing reflexive pain behaviors such as tail flick, limb withdrawal, or orofacial reflexes in response to acute painful stimuli. However, reflex behavior as a measure of pain perception has long been debated among pain researchers. Chapman [1] argued that they could be merely a measure of reflex activity instead of true pain sensation since reflexes can be exerted in spinalized or anesthetized animals. For example, the tail flick and paw withdrawal responses can be elicited in spinal animals [2] and therefore represent spinal reflexes; while vocalization [3] and paw licking on the hotplate test can be elicited in decerebrate animals [3-6] and therefore represent spinal-bulbospinal reflexes. Other investigators also suggest that changes in reflex activity might be due to alterations in motor as well as sensory processing [7,8]. Considerably less effort has been dedicated to measure supraspinal nocifensive behaviors that require integrated behavior, which may be more dependent on cortical activity.
The most commonly used stimulus modalities are electrical, mechanical and thermal. The adequacy and shortcomings of each of these modalities has also been widely debated. Electrical stimulation has been criticized since skin receptors are bypassed and synchronous afferent patterns are generated [9]. Although stimulus location and current density can be well controlled with electrical stimulation, it usually requires restraining the animal thus leading to high levels of anxiety and stress that are known to exhibit modulatory effects on pain sensitivity [10-12]. Commonly used mechanical tests utilize the assessment of paw withdrawal latencies and/or observations of guarding behavior to certain mechanical stimuli [13], such as thresholds to withdrawal to von Frey filaments and to pin prick. A present difficulty with mechanical tests is that the characteristics of mechanical nociception (e.g. combinations of diameter and force in the von Frey test) remain unknown and variable across body sites [14]. Thermal stimulation has been extensively used in assessing pain behavior in animals [9] and it remains the basis of most pain assessment tools, like the tail flick test [15], the hind-limb withdrawal plantar test [16] and the hotplate test [17,18]. The advantages of using heat are the relative constancy of its threshold across body sites, and the extensive research done in psychophysical and physiological studies that has defined and established the range of temperatures that produce heat nociception and its underlying mechanisms. Thus, responses to painful thermal stimuli remain one of the best behavioral tools for studying pain in animals.
The most widely used animal pain assessing tools entail manually recording the duration of animals' limb, or other body parts, withdrawal after applying acute noxious or non-noxious stimuli, or to manually measure and record the latency of licking behaviors, most commonly on the hot plate test (see [19] for latest review). Recent studies examined the interaction between the lab environment and measures of pain behaviors in rodents [20,21]. Chesler et al. [20] compared more than 5,000 data points collected from adult mice, where tail withdrawal latencies were measured for a 49 °C stimulus. A major source of the variance was attributed to experimenter-related variability. Given that all commonly used pain behavior measures depend on the interaction between the experimenter and the animal, the experimenter bias and inability to perform the measures in a blinded fashion remain important shortcomings. Therefore, we developed: 1) A fully automated thermal pain behavior assessment tool, 2) to evaluate pain behavior in response to thermal innocuous and noxious stimuli, 3) in awake unrestrained rodents in a task that requires learning, and 4) to produce a more objective measure of pain behavior.
Patrick Wall and colleagues were the first to recognize that injury to a peripheral nerve may generate pathophysiological mechanisms that are different from those elicited by discrete acute noxious stimuli [22]. Since then, many important animal models for acute and chronic pain have been established and proved useful in unraveling their mechanisms. Here we examine pain behavior in two animal models for neuropathic pain, the chronic constrictive injury (CCI) and the spared nerve injury (SNI) models, and compare the results to standard pain assessing tools. The CCI model was developed by Bennett and Xie [23] and has been extensively studied since; it mimics important clinical chronic pain symptoms such as heat hyperalgesia and mechanical allodynia. SNI model was more recently established by Décosterd and Woolf [24], and was shown to robustly express mechanical and cold allodynia. However, in contrast to the CCI model, SNI rats did not show heat hyperalgesia on the plantar test. Here we test both neuropathic models for mechanical, cold and thermal sensitivity. We examine thermal sensitivity using three tests: plantar, hotplate, and our new automated operant dynamic thermal paradigm (AlgoTrack).
Results
Rats subjected to SNI and CCI showed signs of neuropathy on the operated paw 7 days after induction of injury. The neuropathic behavior included abnormal positioning of the paw (inversion) and signs of spontaneous pain such as shaking, licking and biting of the injured paw. Manifestations of the neuropathic pain were also evident from the behavioral pain tests. Injured rats also displayed trophic changes, which were most evident as abnormal growth of the nails and decreased grooming behavior. The operated animals (SNI, CCI, and sham) did not exhibit any evident motor dysfunctions post surgery.
Mechanical sensitivity
Mechanical paw-withdrawal thresholds, in grams, of the injured (left) paw decreased significantly in both SNI and CCI animals as compared to the control (right) paw and to sham, 7 days post ligation. This attenuation in mechanical thresholds peaked during the second week following injury and was maintained throughout the period of testing (Fig. 1). 3-way ANOVA for mechanical thresholds as a function of animal grouping, day of test, and paw tested was highly significant for all 3 factors (for animal grouping F2,199 = 370.; for test day F3,199 = 157.; and for paw tested F3,199 = 1407.; p < 10-5 for all 3 factors), for their pair-wise interactions (p < 10-5) and for their 3 way interaction (p < 10-5).
Figure 1 Mechanical sensitivity, force (g) required for 50% threshold for paw withdrawal, as a function of time in CCI (A), SNI (B) and sham (C) rats. Mechanical paw-withdrawal thresholds of the ligated paw were significantly attenuated after ligation (indicated by an arrow) in the CCI and SNI models as compared to the right paw (control), and to sham. In both groups, increased mechanical sensitivity persisted for 24 days after ligation.
Cold sensitivity
Both SNI and CCI rats exhibited significant changes in cold sensitivity after ligation, although the magnitude of change in SNI rats was ten times larger than in CCI rats. Paw-withdrawal duration to cold, as measured by the acetone drop test increased at day 7, peaked at day 12, and was maintained for 24 days following ligation (Fig. 2). 3-way ANOVA for paw-withdrawal duration as a function of animal grouping, day of test, and paw tested was highly significant for all 3 factors (for animal grouping F2,199 = 997.; for test day F3,199 = 137.; and for paw tested F1,199 = 1142.; p < 10-5 for all 3 factors), for their pair-wise interactions (p < 10-5) and for their 3 way interaction (p < 10-5).
Figure 2 Cold sensitivity, paw-withdrawal duration to acetone, as a function of time in CCI (A), SNI (B) and sham (C) rats. Paw withdrawal duration to acetone applied to the ligated paw was significantly increased after ligation (indicated by an arrow) in SNI and CCI animals as compared to the right paw (control) and to sham. Increased cold sensitivity was maintained for 24 days after ligation in both groups. Cold sensitivity change was much smaller in CCI animals than in SNI rats.
Plantar test responses
Both SNI and CCI rats exhibited significant increases in thermal sensitivity measured by plantar test after ligation, although the magnitude of change in SNI rats was half as large as in CCI rats. Paw-withdrawal latency to IR heat decreased by day 7 and was maintained for 24 days after ligation (Fig. 3). 4-way ANOVA for paw-withdrawal latency as a function of animal grouping, day of test, paw tested, and IR heat intensity (IR = 30, 70) was highly significant for all 4 factors (for animal grouping F2,396 = 13.8; for test day F3,396 = 6.2.; for paw tested F1,396 = 34; and for heat intensity F1,396 = 6393.7; p < 10-5 for all 4 factors). All pair-wise interactions were significant except animal grouping with heat intensity. All 3-way interactions were not significant except for the animal group, paw, and heat intensity interaction (F2,396 = 3.8; p < 0.02). The 4-way interaction was not significant. Also, the contrast between paws (injured and non-injured) in SNI rats was not significant for higher intensity heat (IR = 70 F1,396 = 1.2, p < 0.27) in contrast to lower intensity heat in SNI rats (IR = 30 F1,396 = 9.6, p < 0.002) and to CCI (IR = 70 F1,396 = 56.6, p < 10-5; IR = 30 F1,396 = 19.8, p < 0.01).
Figure 3 Plantar test for thermal sensitivity, paw-withdrawal latencies, as a function of time in CCI, SNI and sham rats, performed at 2 Infra Red (IR) intensities, 30 (A, B, and C) and 70 (D, E, and F). Paw withdrawal latencies of the ligated paw decreased in CCI and SNI rats 7 days post ligation as compared to the control (right) paw and to sham animals. Thermal hyperalgesia, as assessed on plantar test, was smaller in SNI than CCI rats.
Hot plate test responses
Hot plate test was performed for 4 different temperatures (40, 44, 48, and 52°C). At 40 and 44°C, most animals did not exhibit a positive response within the cut-off limits (30 seconds) pre and post ligation. However, for 48 and 52°C, animals responded within the 30 seconds limit and the response was temperature dependent (Fig. 4). 3-way ANOVA for hind-paw withdrawal latency as a function of animal grouping, day of test, and floor temperature (40, 44, 48, and 52°C) was highly significant for all 3 factors (for animal grouping F2,392 = 224.; for test day F3,392 = 80.; and for floor temperature F3,392 = 2,442.; p < 10-5 for all 3 factors). All pair-wise interactions and the 3-way interaction were significant. Contrasts between pre- and post surgery hind-paw withdrawal latencies were significant for CCI (F1,392 = 56.7, p < 10-5) and SNI (F1,392 = 13.8, p < 0.0002) but not for sham (F1,392 = 0.01, p > .9) animals. The largest difference between pre- and post surgery responses were observed at 48°C stimuli in CCI rats (F1,392 = 56.7, p < 10-5). In SNI rats, pre- vs. post-surgery responses were significantly different only at 44°C (F1,392 = 5.2, p < 0.03) and 48°C stimuli (F1,392 = 8.6, p < 0.004). The contrasts between CCI and sham, and between SNI and sham were significant across days and temperatures tested (CCI vs. sham F1,392 = 79., p < 10-5; SNI vs. sham F1,392 = 6.6, p < 0.01). Over the post-surgery test days (7, 17, 24), SNI vs. sham contrast was not significant at 40°C and 52°C, but was significant at 44°C (F1,392 = 4.8, p < 0.03) and 48°C (F1,392 = 47.5, p < 10-5). In comparison, CCI vs. sham contrasts, over post-surgery days, were significant for all temperatures tested (for 40°C F1,392 = 8.3, p < 0.004; for 44°C F1,392 = 6.3, p < 0.01; for 48°C F1,392 = 142., p < 10-5; for 52°C F1,392 = 5.6, p < 0.02).
Figure 4 Hotplate test for thermal sensitivity, paw withdrawal latencies, as a function of time in CCI (A), SNI (B) and sham (C) rats, performed at 4 temperatures (40, 44, 48 and 52°C). Response latency decreases were most evident in CCI rats when tested at 48°C.
Operant dynamic heat sensitivity test – AlgoTrack – responses
After appropriate training on this task (see below), all three animal groups (CCI, SNI, and sham) show a systematic decrease of escape latency with increasing temperatures applied to the floor (Fig. 5). Moreover, escape latencies for both SNI and CCI exhibited significant attenuation at all 4 temperatures after surgery as compared to pre-surgery values within each group and compared to sham (Fig. 5). 3-way ANOVA for escape latency as a function of animal grouping, day of test, and temperature tested was highly significant for all 3 factors (for animal grouping F2,392 = 107; for test day F3,392 = 76.62; and for test temperature F3,392 = 666.2; p < 10-5 for all 3 factors), for all their pair-wise interactions (p < 10-5) and their 3 way interaction (p < 0.01). Contrasts between pre- and post surgery escape latencies were significant for CCI (F1,392 = 195, p < 10-5) and SNI (F1,392 = 148, p < 10-5) but not for sham (F1,392 = 0.96, p > .32) animals. Differences between pre- and post-surgery escape latencies in both SNI and CCI was highly significantly different at all 4 temperatures tested (F-tests, p < 10-5). The contrasts between CCI and sham, and between SNI and sham were significant across days and temperatures tested (CCI vs. sham F1,392 = 184, p < 10-5; SNI vs. sham F1,392 = 144, p < 10-5). Over the post-surgery test days (7, 17, 24), CCI vs. sham contrasts, over post-surgery days, were significant for all temperatures tested (for 40°C F1,392 = 53.7, p < 10-5; for 44°C F1,392 = 148.5, p < 10-5; for 48°C F1,392 = 105.7, p < 10-5; for 52°C F1,392 = 5.9, p < 0.01). Similarly, SNI vs. sham contrast was significant for all temperatures tested (for 40°C F1,392 = 35, p < 10-5; for 44°C F1,392 = 105, p < 10-5; for 48°C F1,392 = 99.7, p < 10-5; for 52°C F1,392 = 6.25, p < 0.01).
Figure 5 AlgoTrack test for thermal sensitivity, escape latencies, as a function of time in CCI (A), SNI (B) and sham (C) rats. The test was performed for all animals at 4 temperatures (40, 44, 48 and 52°C). Escape latencies were significantly reduced in CCI and SNI animals at all temperatures following ligation as compared to sham, and as compared to pre-ligation. The decrease in escape latencies was maximal at day 12 post-ligation and was maintained throughout the period of testing.
Comparison between hotplate and AlgoTrack stimulus-response curves
Given that hotplate and AlgoTrack tests are the only supraspinal responses examined in this study, direct comparison between them is informative. CCI and SNI, but not sham, animals show similar leftward shifts in stimulus-response curves on both tests (Fig. 6). However, on the hotplate this leftward shift is only 2°C, while on AlgoTrack it is larger than 6°C.
Figure 6 Comparison of the temperature-escape/paw withdrawal responses between AlgoTrack and Hotplate tests in CCI (A, D), SNI (B, E), and sham (C, F) animals tested at 3 days prior and at 7 and 17 days post-ligation. On the hotplate test, paw withdrawal latency decreases post-ligation are small and observed mainly in CCI rats. On the AlgoTrack test, both CCI (A) and SNI (B) rats exhibit post-ligation attenuation in their escape times at all temperatures tested.
Comparing between Plantar, Hotplate and AlgoTrack for thermal hyperalgesia
Plantar, hotplate and AlgoTrack tests assess thermal pain behavior. To compare the outcomes on these tests, we use F-values obtained for contrasts examined in planned comparisons. Given that the largest differences between the three thermal tasks were seen in SNI animals, the F-values for pre- vs. post-ligation thermal responses, and for SNI vs. sham post-ligation thermal responses are summarized in figure 7. Behavioral outcomes on AlgoTrack show the largest thermal hyperalgesia, followed by outcomes on Hotplate, while outcomes on Plantar test show the least amount of thermal hyperalgesia.
Figure 7 Comparing detection of thermal sensitivity changes between Plantar, Hotplate, and AlgoTrack tests in SNI rats (F-values in ANOVA planned comparisons). A) F-values for contrasting thermal pain behavior pre-ligation (3 days prior) to post-ligation (days 7, 17 and 24) on the Plantar test (P; for both 30 and 70 IR intensities); on the Hotplate test (H; for all test temperatures: 40, 44, 48, 52°C); and on the AlgoTrack test (for the same test temperatures). B) F-values for contrasting between SNI rats and sham rats on AlgoTrack test (gray bars) and Hotplate test (black bars) in post-ligation days (7, 17 and 24), for each applied temperature indicated. C) F-values for contrasting between SNI rats and sham rats on Plantar test in post-ligation days, for the IR intensities indicated. The y-axis is in log scale and covers 4 decades. The dashed line is for F = 2.0, which delineates threshold for significance.
Performance on AlgoTrack is learned behavior
Unlike the hotplate and plantar tests, and because AlgoTrack involves more complex operant behavior, hence proper assessment of thermal responses on this device requires an initial period of training. Rats seem to need at least 4 testing sessions over a 2-week period to learn that escaping to the other plate eliminates the stimulus. These are seen in the initial escape values in figure 5. Figure 8 illustrates the effects of learning by showing that the variance of escape latencies in the sham rats decreased by two magnitudes over six test-sessions. Moreover, examining individual animal behavior shows that each animal undergoes a unique behavioral pattern of change and yet eventually settles to stereotyped escape behavior where escape latencies become predictable by the applied temperature and similar between animals (Fig. 9).
Figure 8 AlgoTrack escape latency variance in sham rats (n = 8) for 4 temperatures (40, 44, 48 and 52°C) as a function of test session. Variances for all temperatures tested decreased by the fourth testing session, and were maintained through further testing.
Figure 9 Algotrack individual responses of 4 sham rats for 4 temperatures 40 (A), 44 (B), 48 (C) and 52°C (D) plotted against test session. Illustrated are the first 6 sessions these animals were ever exposed to this paradigm. Each animal exhibits its own pattern of learning throughout the first few testing sessions, and then stabilizes to well-defined escape latencies for the different temperatures.
Discussion
The main observation of the present result is that the presence and magnitude of thermal hyperalgesia, as assessed in two types of neuropathic rats, critically depends on the type of test used. Our automated, operant, dynamic thermal sensitivity assessment, AlgoTrack, seems most sensitive, hotplate test shows lower sensitivity, and the plantar test seems to be the least sensitive. We also show the ability to follow thermal stimulus-response curves in awake, freely moving rats, repeatedly for weeks, using our automated device.
Comparison between pain assessment tools
A number of groups have attempted to develop automated behavior assessment tools to measure pain responses in animals in order to exclude bias from experimenters' judgments. Jett and Michelson [25] and Duysens and Gybles [26] developed a computer-driven force detector to measure animals' motor response accurately after painful stimuli. Jourdan et al. [27] introduced an automated pain scoring technique based on continuous video recording of movements of rats after formalin injection. Möller et al. [28] constructed an electronic pressure stimulator to standardize mechanical force applied to animals. The noxious heat stimulator first described by Hargreaves et al. [16], made a great impact on thermal pain behavior measurement due to several advantages: 1) animals are stimulated without restraining, thus minimizing stress and unnecessary sympathetic and hormonal response, 2) stimulus application and response measures are fully automated thereby minimizing human involvement in the assessment, and 3) the stimulus is applied to each limb separately. Although these advances have improved our understanding of animal pain behavior, these methods emphasize reflexive aspects of pain behavior. Recently LaBuda and Fuchs [29] described cognitive avoidance behavior test paradigm based on the combination of place preference and mechanical stimulation. Even though the paradigm does not eliminate human involvement, it does assess integrated behavior contrasting pain with aversion, and thus seems a very useful tool. Mauderli et al. [14] describe an alternative thermal pain assessment tool in awake, behaving rats. The paradigm is based on two boxes, one is used for thermal escape behavior, modulated by light/dark preference; while a second box provides a control task which measures escape latencies from bright light, controlling for generalized effects on aversion. Similarly to our device, the paradigm of Mauderli et al. [14] assesses pain behavior that depends on learned responses in the awake, behaving rat. Their paradigm also provides the means for dissociating the effects of analgesics on motor behavior from that of antinociception. The Mauderli et al. [14] and LaBuda and Fuchs [29] devices are complimentary to each other, contrasting mechanical and thermal pain with aversion. Yet outcomes from these devices have not been compared directly. Given the present results, we predict that outcomes on these devices may be different depending on type of injury giving rise to the pain behavior. We dub our device as a dynamic operant behavior task, to distinguish it from the paradigm of Mauderli et al. [14], since in our paradigm the animal responds to changes in skin temperature, while in the Mauderli paradigm the animal is actively exploring an environment where the floor temperature is kept constant.
Given that the our paradigm is dependent on learned behavior and on motor responses where the animal has to assess its body position relative to the track and find the appropriate escape path, implies that the responses are due to highly integrated behavior, which most likely depends on cortical circuitry. Obviously anesthetized or decorticated animals cannot do this task, which is not true for reflex-dependent pain assessment tools, e.g. the plantar or tail flick tests. Our device (AlgoTrack) is similar in design to a shuttle box introduced by Berkeley and Parmer [30] who used electrical shock applied through the floor for induction of pain and latency to escape over a raised barrier as a measure of pain behavior in awake, unrestrained, trained cats. Their study is important since it showed that lesions limited to the somatosensory cortex increased escape response thresholds. Hence demonstrating at least partial dependence of such behaviors on cortical circuits. The paradigm introduced by Berkeley and Parmer [30] has not been used in pain assessment since, perhaps because it was based on responses to electrical shock (electrical-shock induced escape behavior remains popular in studies of brain substrates underlying aversive classical conditional learning, e.g. LeDoux, [31], mainly due to the well-defined temporal relationship between conditioned and unconditioned stimuli), and perhaps due to the lack of convincing evidence at the time for involvement of the cerebral cortex in pain behavior/perception. Our device also shares similarities with both hotplate and plantar devices since it assesses escape times for thermal stimuli. Of course the plantar test is based on local monosynaptic reflex circuitry, while hotplate and AlgoTrack involve more complex behavior. It should be emphasized that the monitoring of local sign in the plantar test is a main advantage since the measure can distinguish thermal sensitivity differences between injured and uninjured limbs. The AlgoTrack sacrifices this differentiation to assess the significance of thermal stimuli to the whole organism, in a completely automated fashion. It is based on the assumption that the escape behavior is triggered primarily through tissue that is most sensitive to the stimulus, in this case the injured limb. It should be mentioned that AlgoTrack is rather labor intensive as compared to simple reflexive tests, making it difficult to implement in high throughput studies. It also requires ability to learn and this may become a confounder in genetic manipulations where learning may be impaired.
Caveats regarding learning and over-learning
Our results show that AlgoTrack requires an initial period of training of about 4 sessions, over a 2-week interval. In the first training trials, escape behavior exhibits large variability because the animals need to discover that moving to the opposite side would eliminate the stimulus, and often exhibit agitated behavior including licking, jumping, as well as exploring. Moreover, in the initial trials many animals exhibit side preference. After training the overall variance decreased, and the mean escape duration was very similar across different animals. This dependence on learning indicates involvement of cortical circuitry.
Various precautions were implemented to limit or decrease over-learned behavior: 1. Non-noxious, or minimally noxious, thermal stimuli (40 and 44°C) were inter-mixed with more noxious stimuli (>45°C), to reduce stimulus predictability. 2. Thermal stimuli were presented in a pseudo-random sequence. 3. Behavioral assessment was limited to two test sessions per week. The results indicate that we have overcome these difficulties since in the sham-operated rats there is a relatively constant stimulus-response curve for the duration of the study (38 days), past the initial period of training. Not presented are observations we have made in other animals where more frequent testing on the AlgoTrack (> 4 tests per week) resulted in further decreases in escape latencies and decreased sensitivity to stimulus temperature variations.
Neuropathic pain behavior and differential thermal hyperalgesia
Overall the quantitative results assessing pain-like behavior shown for CCI and SNI rats closely agree to earlier results, demonstrating that both groups show increased mechanical sensitivity as determined by decreased thresholds to von Frey filaments. SNI animals also exhibited cold sensitivity as measured by the acetone drop test [24,32,33]. CCI animals showed thermal hyperalgesia as measured by the plantar and hotplate test [23,34-36]. The new result is the illustration that both groups exhibit more profound thermal hyperalgesia than suspected in the past when they are tested on our automated paradigm (AlgoTrack). The F-values for comparing pain behavior in SNI animals pre- to post-ligation indicates a 10-fold increase in detectability of hyperalgesia by AlgoTrack as compared to plantar or hotplate tests. Moreover, detectability of hyperalgesia by AlgoTrack of SNI rats as compared to sham post-ligation also show consistently higher F-values than on hotplate or plantar tests.
Multiple reasons may underlie the observed thermal sensitivity differences. The differences may be a reflection of the specifics of executing the different tests, and/or a reflection of the animal models studied. Regarding the details of the tests: Since our paradigm is completely automated its results must be considered the most reliable. On the other hand, the plantar test is very simple to deliver and the outcome measure is automated. It is possible that the application of the heat during the plantar test is more compromised because of the abnormal paw position that SNI animals exhibit, and because of the large tactile sensitivity differences these animals show between lateral and medial aspects of the paw. Alternatively, or in addition the thermal sensitivity differences may reflect the level of the central nervous system circuitry assessed by each test and its significance to the pain model understudy. The existence of multiple distinct central nociceptive circuits has been suggested in the past [19,37-39]. We presume that the more complex behavioral requirement translates into engaging higher central circuitry; as a result the plantar test reflects local reflexes, the hotplate bulbo-spinal pathways, while the AlgoTrack may require cortical nociceptive circuitry. Therefore, our results can be interpreted as suggesting that, in neuropathic injury rats, thermal hyperalgesia becomes more prominent when the task requires more integrated behavior. This in turn implies that 'central sensitization' in these models more prominently includes supraspinal nociceptive circuitry. Based on this interpretation, we predict that pain models impinging more directly on peripheral processes, such as inflammatory conditions, might exhibit a test-dependence pattern of thermal hyperalgesia opposite in sensitivity to that observed for neuropathic animals. These hypotheses remain to be tested.
Methods
Hardware for automated thermal sensitivity testing
The device consists of a small rectangular box placed upon a metallic floor with 2 separate heating plates. The heating elements are made with flexible etched foil resistance, and covered with self-adhesive aluminum foil that maximizes thermal conductance to the plates. Plate temperatures increase at a rate of 5°C per second and are controlled by microprocessor-based temperature controllers (Odgen, Chicago, IL.). Initially both plates are set to a warm, comfortable temperature (36°C). A motion detector based on an array of infrared sensors, identifies on which of the two plates the animal is located. This plate is heated to a desired temperature. Upon perceiving the thermal stimulus, the animal typically escapes to the opposite non-heated plate. The infrared sensors detect the crossing of the animal and both escape latency and temperature are displayed. The controller switches the heated plate to the baseline temperature and, the whole operation is repeated after a 5-minute inter-trial interval.
Animals
Twenty-six male Sprague-Dawley rats (250–350 grams) were used throughout this study. They were maintained in a climate-controlled environment on a 12-hour light/dark cycle with free access to food and water. All tests were performed in the light period and were approved by the Animal Care and Use Committee (ACUC) at Northwestern University, Chicago, and were in accordance with the NIH for the ethical use of laboratory animals and the IASP for use of conscious animals in pain research [40].
Surgery
Animals were anesthetized using ketamine hydrochloride (45 mg/kg, i.p.) and xylazine (10 mg/kg i.p.) After surgery, all wounds were sutured using a non-absorbable surgical suture (3-0 silk, REF 1124-11, Sherwood Medical, USA), treated with a topical antibiotic ointment, and were rested for seven days before further testing. The SNI model was implemented in 10 animals. The left sciatic nerve was exposed at the level of its trifurcation into the sural, tibial and common peroneal nerves. Each of the tibial and common peroneal nerve was tightly ligated by two knots 4 mm apart using 6.0 silk and then completely severed in between, leaving the sural nerve intact [24]. The CCI model was implemented in a second group of animals (n = 8). In these animals, the left sciatic nerve was exposed above the level of trifurcation and four loose knots were carefully applied to the nerve using absorbable gut [23]. An additional 8 animals were used as sham, where the sciatic nerve was exposed but not manipulated.
Pain behavioral testing
All operated animals (CCI, SNI and sham rats) were studied behaviorally from 2 weeks prior to surgery up to 4 weeks post surgery. Mechanical and cold sensitivities as well as thermal responses on AlgoTrack, Hotplate, and Plantar test were examined in all animals, at all time points.
Mechanical sensitivity
Mechanical sensitivity of the hind paw was measured by determining withdrawal thresholds to Von Frey filaments. A set of 18 filaments (Stoelting, Chicago, IL.), marked from 1.65 to 6.5, was used. The respective bending forces were in the range of 0.005 to 125.892 g. The animals were placed individually in a small (35 × 20 × 15 cm) plastic cage with an open wire mesh bottom. Before testing, the rats were left in the test cages for 15–20 min so that their grooming and exploratory behaviors cease and all four paws were placed on the ground. All tests were performed on the right (control) and left (ligated) hind paws. Von Frey filaments were applied perpendicularly to the planter surface of the paw with an upward force just sufficient to bend the microfilament. When testing SNI animals, special care was taken to stimulate the lateral plantar surface, which is the area of the skin innervated by the sural nerve [24]. Paw withdrawals due to locomotion or weight shifting were not counted and such trials were repeated. The 50% threshold for each paw withdrawal was calculated as described by Chaplan et al. [13].
Cold sensitivity
Cold allodynia was measured by the acetone drop test described by Décosterd and Woolf [24]. A blunt needle connected to a syringe was used to drop 50 μl of acetone on the paw and the duration (in seconds) of the paw withdrawal was recorded. Minimal and maximal cut-offs were assigned at 0.5 and 20 sec, respectively. Paw withdrawals due to locomotion or weight shifting were not counted and such trials were repeated.
Plantar test
Animals were placed in an acrylic box with glass pane floor and the plantar surface of their hind paw was exposed to a beam of infrared radiant heat (Ugo Basile, Stoelting, Chicago, IL; [16]). The paw withdrawal latencies were recorded at 2 different infrared intensities, 30 and 70, and were measured twice per session, separated by a minimum interval of 5 minutes. Minimum and maximum cut-offs were assigned at 1 and 30 seconds, respectively. Again, paw withdrawals due to locomotion or weight shifting were not counted and the trials repeated.
Hotplate test
Animals were gently dropped into an acrylic box with a metal floor that was preheated to a certain temperature [17]. Paw withdrawal latency was calculated using a timer that was started when the animal is released onto the preheated plate and stopped at the moment of withdrawal, shaking, or licking of either hind paw. Paw withdrawal latencies for each animal were calculated for 4 different temperatures (40, 44, 48, and 52°C). All animals were tested once for each temperature per session in a random sequence.
Automated dynamic thermal sensitivity test (AlgoTrack)
Animals were gently dropped into the chamber with both plates preheated to 36°C and allowed to rest for 5 minutes. Then the plate on which the rat is rested is heated up to the desired temperature. Escape latencies were recorded for all the animals at 4 different temperatures (40, 44, 48, and 52°C). Animals were tested twice for each temperature per session, with the temperatures presented in a randomized sequence. The maximum thermal stimulus duration was 30 seconds. If an animal did not escape a stimulus in this time duration, the stimulus was turned off and the animal's response recorded as escape at 30 seconds. Animals tested on this paradigm did not exhibit any signs of tissue injury or burns at all temperatures tested, throughout the period of the study.
Data analysis
All data are presented as group averages ± SEM. The baseline value for all tests between pre- and post neuropathic injury used the last test session prior to operation (day -3). The differences between animal groupings (CCI, SNI, sham), day of test (-3, 7, 17, and 24 days), ligated and non-ligated hind-paws (left and right), and different intensities of stimuli were compared with multi-way between-groups fixed-effects analysis of variance (ANOVA). Planned comparisons were tested using contrast analysis (F-tests), and post-hoc comparisons were done using Turkey's honest test for post-hoc multiple comparisons for unequal N. Statistical significance was considered at p < 0.05 (Statistica, StatSoft, Inc.).
Competing interests
The author(s) declare that they have no competing interests.
Acknowledgements
The authors thank Youngsoo Kim, Sonya Dave, and Ola Alade for their contributions to the study. The study was partially funded by Emma Gale Harris fund and by NIH NINDS NS 42660.
==== Refs
Chapman CR Casey KL Dubner R Foley KM Gracely RH Reading AE Pain measurement: an overview Pain 1985 22 1 31 4011282 10.1016/0304-3959(85)90145-9
Franklin KBJ Abbott FV Boultoun M, Baker GB and Greenshaw AJ Techniques for assessing the effects of drugs on nociceptive responses Neuromethods, Psychopharmacology 1989 Clifton, The Humana Press 145 215
Caroll MN LIM RK Observations on the neuropharmacology of morphine and morphinelike analgesia Arch Int Pharmacodyn Ther 1960 125 383 403 13849328
Berridge KC Progressive degradation of serial grooming chains by descending decerebration Behav Brain Res 1989 33 241 253 2757783
Matthies BK Franklin KB Effects of partial decortication on opioid analgesia in the formalin test Behav Brain Res 1995 67 59 66 7748501 10.1016/0166-4328(94)00104-N
Woolf CJ Long term alterations in the excitability of the flexion reflex produced by peripheral tissue injury in the chronic decerebrate rat Pain 1984 18 325 343 6728499 10.1016/0304-3959(84)90045-9
McGrath PA Sharav Y Dubner R Gracely RH Masseter inhibitory periods and sensations evoked by electrical tooth pulp stimulation Pain 1981 10 1 17 7232006 10.1016/0304-3959(81)90041-5
Willer JC Boureau F Albe-Fessard D Supraspinal influences on nociceptive flexion reflex and pain sensation in man Brain Res 1979 179 61 68 509233 10.1016/0006-8993(79)90489-X
Dubner R Wall PD and Melzack R Methods of assessing pain in animals. Textbook of Pain 1989 Edinburgh, Churchville Livingstone 247 256
Basbaum AI Fields HL Endogenous pain control systems: brainstem spinal pathways and endorphin circuitry Annu Rev Neurosci 1984 7 309 338 6143527 10.1146/annurev.ne.07.030184.001521
Caceres C Burns JW Cardiovascular reactivity to psychological stress may enhance subsequent pain sensitivity Pain 1997 69 237 244 9085297 10.1016/S0304-3959(96)03289-7
Maier SF Wiertelak EP Watkins LR Endogenous pain facilitatory systems: antianalgesia and hyperalgesia J Pain 1992 1 191 198
Chaplan SR Bach FW Pogrel JW Chung JM Yaksh TL Quantitative assessment of tactile allodynia in the rat paw J Neurosci Methods 1994 53 55 63 7990513 10.1016/0165-0270(94)90144-9
Mauderli AP Acosta-Rua A Vierck CJ An operant assay of thermal pain in conscious, unrestrained rats J Neurosci Methods 2000 97 19 29 10771071 10.1016/S0165-0270(00)00160-6
D' Amour FE Smith D A method for determining loss of pain sensation. J Pharmacol Exp Ther 1941 72 74 79
Hargreaves K Dubner R Brown F Flores C Joris J A new and sensitive method for measuring thermal nociception in cutaneous hyperalgesia Pain 1988 32 77 88 3340425 10.1016/0304-3959(88)90026-7
Espejo EF Mir D Structure of the rat's behaviour in the hot plate test Behav Brain Res 1993 56 171 176 8240711 10.1016/0166-4328(93)90035-O
Woolfe G Macdonald AD The evaluation of the anelgesic action of pethidine hydrochloride (Dermol) J Pharmacol Exp Ther 1944 80 300 307
Mogil JS Wilson SG Wan Y Kruger L Assessing nociception in murine subjects Methods in Pain Research 2001 Boca Raton, CRC Press 11 39
Chesler EJ Wilson SG Lariviere WR Rodriguez-Zas SL Mogil JS Influences of laboratory environment on behavior
1 Nat Neurosci 2002 5 1101 1102 12403996 10.1038/nn1102-1101
Shir Y Seltzer Z Heat hyperalgesia following partial sciatic ligation in rats: interacting nature and nurture Neuroreport 2001 12 809 813 11277588 10.1097/00001756-200103260-00038
Wall PD Gutnick M Ongoing activity in peripheral nerves: the physiology and pharmacology of impulses originating from a neuroma Exp Neurol 1974 43 580 593 4827166 10.1016/0014-4886(74)90197-6
Bennett GJ Xie YK A peripheral mononeuropathy in rat that produces disorders of pain sensation like those seen in man Pain 1988 33 87 107 2837713 10.1016/0304-3959(88)90209-6
Decosterd I Woolf CJ Spared nerve injury: an animal model of persistent peripheral neuropathic pain Pain 2000 87 149 158 10924808 10.1016/S0304-3959(00)00276-1
Jett MF Michelson S The formalin test in rat: validation of an automated system Pain 1996 64 19 25 8867244 10.1016/0304-3959(95)00098-4
Duysens J Gybels J The automated measurement of hindlimb flexor reflex of the rat as a substitute for the tail-flick assay J Neurosci Methods 1988 24 73 79 3386305 10.1016/0165-0270(88)90035-0
Jourdan D Ardid D Bardin L Bardin M Neuzeret D Lanphouthacoul L Eschalier A A new automated method of pain scoring in the formalin test in rats Pain 1997 71 265 270 9231869 10.1016/S0304-3959(97)03366-6
Moller KA Johansson B Berge OG Assessing mechanical allodynia in the rat paw with a new electronic algometer
1 J Neurosci Methods 1998 84 41 47 9821632 10.1016/S0165-0270(98)00083-1
Labuda CJ Fuchs PN Low dose aspirin attenuates escape/avoidance behavior, but does not reduce mechanical hyperalgesia in a rodent model of inflammatory pain Neurosci Lett 2001 304 137 140 11343821 10.1016/S0304-3940(01)01787-6
Berkley KJ Parmer R Somatosensory cortical involvement in responses to noxious stimulation in the cat Exp Brain Res 1974 20 363 374 4426358 10.1007/BF00237381
LeDoux JE Emotional memory systems in the brain Behav Brain Res 1993 58 69 79 8136051 10.1016/0166-4328(93)90091-4
El Khoury C Hawwa N Baliki M Atweh SF Jabbur SJ Saade NE Attenuation of neuropathic pain by segmental and supraspinal activation of the dorsal column system in awake rats
1 Neuroscience 2002 112 541 553 12074897 10.1016/S0306-4522(02)00111-2
Erichsen HK Blackburn-Munro G Pharmacological characterisation of the spared nerve injury model of neuropathic pain Pain 2002 98 151 161 12098627 10.1016/S0304-3959(02)00039-8
Attal N Jazat F Kayser V Guilbaud G Further evidence for 'pain-related' behaviours in a model of unilateral peripheral mononeuropathy Pain 1990 41 235 251 2164179 10.1016/0304-3959(90)90022-6
Coggeshall RE Dougherty PM Pover CM Carlton SM Is large myelinated fiber loss associated with hyperalgesia in a model of experimental peripheral neuropathy in the rat? Pain 1993 52 233 242 8384361 10.1016/0304-3959(93)90136-D
Jacobs JM Ro LS A morphological study of experimental mononeuropathy in the rat: early ischemic changes J Neurol Sci 1994 127 143 152 7707073 10.1016/0022-510X(94)90066-3
Jensen TS Yaksh TL Comparison of the antinociceptive effect of morphine and glutamate at coincidental sites in the periaqueductal gray and medial medulla in rats Brain Res 1989 476 1 9 2563331 10.1016/0006-8993(89)91529-1
Konig M Zimmer AM Steiner H Holmes PV Crawley JN Brownstein MJ Zimmer A Pain responses, anxiety and aggression in mice deficient in pre-proenkephalin Nature 1996 383 535 538 8849726 10.1038/383535a0
Pastoriza LN Morrow TJ Casey KL Medial frontal cortex lesions selectively attenuate the hot plate response: possible nocifensive apraxia in the rat Pain 1996 64 11 17 8867243 10.1016/0304-3959(95)00070-4
Zimmermann M Ethical guidelines for investigations of experimental pain in conscious animals Pain 1983 16 109 110 6877845 10.1016/0304-3959(83)90201-4
| 15918900 | PMC1156939 | CC BY | 2021-01-04 16:40:08 | no | Mol Pain. 2005 May 26; 1:18 | utf-8 | Mol Pain | 2,005 | 10.1186/1744-8069-1-18 | oa_comm |
==== Front
Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-151587173610.1186/1475-2891-4-15ResearchWeight reduction is not a major reason for improvement in rheumatoid arthritis from lacto-vegetarian, vegan or Mediterranean diets Sköldstam Lars [email protected] Lars [email protected] Linda [email protected] Gunnar [email protected] Department of Medicine, County Hospital, Visby, Sweden2 Department of Clinical Physiology, County Hospital, Kalmar, Sweden3 Department of Food and Nutrition, Umeå University, SE-901 87 Umeå, Sweden2005 4 5 2005 4 15 15 4 10 2004 4 5 2005 Copyright © 2005 Sköldstam et al; licensee BioMed Central Ltd.2005Sköldstam et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Objectives
Several investigators have reported that clinical improvements of patients with rheumatoid arthritis (RA), from participating in therapeutic diet intervention studies, have been accompanied by loss of body weight. This has raised the question whether weight reduction per se can improve RA. In order to test this hypothesis, three previously conducted diet intervention studies, comprising 95 patients with RA, were pooled. Together with Age, Gender, and Disease Duration, change during the test period in body weight, characterised dichotomously as reduction or no reduction (dichoΔBody Weight), as well as Diet (dichotomously as ordinary diet or test diet), were the independent variables. Dependent variables were the difference (Δ) from baseline to conclusion of the study in five different disease outcome measures. ΔESR and ΔPain Score were both characterised numerically and dichotomously (improvement or no improvement). ΔAcute Phase Response, ΔPhysical Function, and ΔTender Joint Count were characterised dichotomously only. Multiple logistic regression was used to analyse associations between the independent and the disease outcome variables.
Results
Statistically significant correlations were found between Diet and three disease outcome variables i.e. ΔAcute-Phase Response, ΔPain Score, and ΔPhysical Function. Δ Body Weight was univariately only correlated to ΔAcute-Phase Response but not significant when diet was taken into account.
Conclusion
Body weight reduction did not significantly contribute to the improvement in rheumatoid arthritis when eating lacto-vegetarian, vegan or Mediterranean diets.
Rheumatoid Arthritistherapydiet interventionlacto-vegetarian dietvegan dietMediterranean dietbody weight
==== Body
Introduction
We have recently found that patients with rheumatoid arthritis (RA) improved significantly in disease activity from eating a modified Cretan Mediterranean diet (MD) [1]. Improvement was seen in 9 of 14 efficacy variables. The RA disease activity as measured by the "Disease Activity Score" DAS28 [2] was reduced by 13 %. In comparison the control group showed no change. The favourable outcome of the diet intervention group indicated that the MD had been suppressive to RA inflammation. We believe that the anti inflammatory effect of the MD was mediated by its different lipid profile [1] in conjunction with its high content of fresh fruits and green vegetables. However, at the end of the experiment the patients of the MD group, but not the control group, had lost 3 kg in weight (p < 0.001 between groups), although our ambition had been to prescribe a diet that was isocaloric compared to the patients' previous food intake. This unexpected weight loss raised the question whether a reduced energy intake, could have been yet another anti-inflammatory factor? That diet intervention studies may induce unexpected weight reduction to patients with RA is a finding shared by others [3-9]. Some researchers [7,8] have even reported a statistically significant correlation between change in body weight with change in arthritis measurements.
The aim of the present study was to investigate whether weight reduction could have anti inflammatory effect on RA. Data from three previously performed diet intervention studies were pooled for the analysis.
Methods
Design
Data from three different diet intervention studies, which we had previously conducted on patients with rheumatoid arthritis [1,3,4] were pooled and reanalysed.
Two studies (1st and 3rd) had a prospective, randomised, and parallel design. The 2nd was a prospective, cross over study. Nor the participating patients or the clinical outcome valuators were blinded with respect to the study designs.
Patients
The pooled patient material comprises 95 Caucasians with RA according to the American College of Rheumatology (ACR) criteria from 1984. All but one had active disease. The three populations showed equal distributions with respect to sex, age, disease duration, functional capacity and stage of RA disease (Table 1).
Table 1 Baseline characteristics of patients who completed the trials.
Diet group (n = 60) Control group (n = 42)
Age (years) 54.5 (33–73) 57.0 (35–75)
Gender (M/F)* 11/49 7/35
Weight (kg) 72.1 (40–109) 69.4 (41–102)
Disease duration (years) 13.0 (0.5–59) 11.9 (2–35)
Footnotes to table:
Data are presented as mean (range) unless otherwise stated.
* Number of male and females, respectively
The original studies
In the first study [3] 24 RA patients concluded the trial. 14 tested a lacto-vegetarian diet, and 10 were controls. Eight from the diet group reported less pain at the end of the experiment. As a group they lost 2.6 kg (p < 0.001) in body weight, but showed no statistically significant change in disease outcome measures. The control group stayed unchanged.
In the second study [4], which had a cross over design, 20 RA patients concluded the trial. All were initially followed for a control period of two or five months, 13 and 7 patients, respectively. Thereafter, all patients adopted a vegan diet for 4 months. During the vegan diet period the group of 20 patients obtained statistically significant reduction in pain score, increase in functional capacity, and a loss of 4.8 kg (p < 0.001) in body weight. No change took place during the control periods. Only those 7 patients who had had a five months long control period were included as controls in the present pooled analysis.
In the third study [1] 51 RA-patients concluded the trial. 26 tested a Cretan Mediterranean diet (MD) and 25 were controls. As a group the MD patients improved in 9 out of 14 disease outcome measures and lost 3 kg (p < 0.001) in body weight.
The pooled patient material
Number of patients from the three studies is 95. Seven of these patients were studied in a cross-over design (2nd study), and therefore, included both in the control and the intervention group giving a total number of 95+7. The patient demographics are shown in Table 1.
Diets
The diets tested were lacto-vegetarian [3], strictly vegetarian, vegan [4] and a modified Cretan Mediterranean diet [1].
Measurements
Five different measures of RA disease activity were identified, that had been assessed in all three studies. Two had been identically assessed in all three studies. These were the ESR (the Westergren erythrocyte sedimentation rate), and the Pain Score, which was the patient's self perceived pain severity as evaluated on a visual analogue scale (VAS, 0–100 mm).
The three measures, which were not similarly assessed and therefore not directly comparable, were firstly the blood plasma Acute-Phase Response [10], which in the 1st study was measured with the reaction in plasma concentration of orosomucoid, and in the 2nd and 3rd with the corresponding reaction of C-reactive protein (CRP). The second was Physical Function as evaluated and recorded by the patient with a graded questionnaire of difficulties in performing specified tasks of daily living. A locally constructed questionnaire was used in the 1st study, a not validated Swedish version of the Stanford-health assessment questionnaire (HAQ) in the 2nd study, and the official Swedish version of HAQ [11] in the third study. The third dissimilarly assessed outcome measure was the Tender Joint Count [2], which in the 1st study was the Ritchie joint index [12], and in the 2nd and 3rd study was the number of tender joints from palpation of 40, and 28 peripheral joints, respectively.
Due to different techniques of assessing these last three measures, change (Δ) obtained with these variables from baseline to the end of the experiments, could not be directly compared from the numerical results. These changes had to be converted into dichotomous values to enable a relevant statistical analysis. Change from baseline to the end of each study was thus characterised as improvement, or no improvement. These dichotomous variables are here called: dichoΔAcute-Phase Response, dichoΔPhysical Function and dichoΔTender Joint Count.
Tested variables
Patient demographics
Gender, age and disease duration were used as independent variables. Age was characterised with respect to the median value, dichotomously, as <56 yrs or ≥ 56 yrs (dichoAge). Similarly, disease duration was characterised as <9.5 yrs or ≥ 9.5 yrs (dichoDisease Duration). Also the diet intervention was handled as an independent variable, and was characterised dichotomously, either as Case (lacto-vegetarian, vegan diet and Mediterranean) or as Control. The only independent outcome variable was the difference (Δ) from baseline to end of the study in body weight, and was characterised dichotomously as reduction or no reduction (dichoΔBody Weight).
Disease outcome variables
For evaluating effects that the dietary interventions might have had on RA disease activity, the difference (Δ) from baseline to end of study, in each of the five disease measures was calculated. For ESR and for Pain Score this was done numerically (numΔESR and numΔPain Score) but also, in the logistic regression, dichotomously characterised as improvement, or no improvement (dichoΔESR and dichoΔPain Score, respectively).
The other three RA disease measures were only handled dichotomously. They were: dichoΔAcute-Phase Response, dichoΔPhysical Function and dichoΔTender Joint Count, and were characterized as improvement /no improvement.
Statistics
Baseline characteristics of the two diet groups were tested with non parametric tests (Mann-Whitney's U-test, and for gender Fischer's exact test; Table 1). Student's t-test and Mann-Whitney's nonparametric U-test were used to analyse group differences of change in numerical values of ΔESR and ΔPain Score, respectively.
The dichotomous variables were only analysed with logistic regression. All independent variables considered potentially significant were initially included in the model followed by a step-wise deletion of the least significant variable until a significant level of 0.05 or less. The statistical analyses were made using a commercially available computer programme (STATISTICA, StatSoft®, Tulsa, USA) and a p-value less than 0.05 was considered statistically significant.
Results
Baseline values before diet intervention
At baseline, no significant differences in Age, Gender, Body Weight or Disease Duration were found between the two groups (Table 1) or in the disease measures ESR and Pain Score.
Change obtained from baseline to the end of the diet intervention period
Univariate analysis
Gender, Age, or Disease Duration were not associated with any of the outcome variables (p-values 0.2–1.0; univariate logistic regression; Table 2). The diet regimens rendered the diet group a weight fall of on average 3.5 kg, which is highly significant compared to the controls, who lost 0.1 kg (p < 0.001; Student's t-test). Body Weight reduction was univariately only correlated to improvement in dichoΔAcute-Phase Response (Table 2). Of the other outcome variables, ESR raised from 30.7 +/- 22 to 31.5 +/- 23 for the diet group and from 34.1 +/- 21 to 36.1 +/- 27 for the control group. These changes were not statistically significant (Student's t-test or Mann-Whitney's U-test) and did not correlate significantly with ΔBody Weight or dichoΔBody Weight. These findings were in accordance with the univariate logistic regression using the dichotomous variables (Table 2).
Table 2 Univariate logistic regression results (Odds Ratios (OR) and corresponding p-values), showing the associations between Gender, Age, Diet, Disease Duration and ΔBody Weight on the one hand, and the five different dichoΔ outcome variables on the other. Bolded values are statistically significant.
Independent variable
OR Gender Age (yrs) Diet Duration (yrs) ΔBody Weight
Effect variable M F p <56 ≥56 p Control Case p <9.5 ≥9.5 p <0 ≥0 p
dichoΔESR 1.00 1.02 0.98 1.00 1.22 0.62 1.00 1.29 0.55 1.00 1.00 1.00 1.00 1.64 0.29
dichoΔAcute-PR 1.00 0.60 0.33 1.00 0.74 0.45 1.00 3.27 0.007 1.00 1.17 0.69 1.00 2.85 0.03
dichoΔPain Score 1.00 0.47 0.18 1.00 0.65 0.29 1.00 3.43 0.004 1.00 1.27 0.55 1.00 2.10 0.10
dichoΔPhysic. F 1.00 0.86 0.78 1.00 0.74 0.45 1.00 4.22 0.002 1.00 1.38 0.42 1.00 2.16 0.10
dichoΔTender JC 1.00 0.93 0.90 1.00 1.54 0.28 1.00 1.54 0.29 1.00 1.22 0.62 1.00 1.77 0.20
Footnotes to table:
ESR = The Westergren erythrocyte sedimentation rate
Acute-PR = The Acute-Phase Response (Orosomucoid or C-reactive protein (CRP))
Pain Score = the patient's self perceived pain severity as evaluated on a visual analogue scale
Physic. F = Physical Function assessed by self completed questionnaires on the degree of difficulty in performing specified tasks of daily living
Tender JC = Tender Joint Count = number of painful joints at rest with pressure.
ΔPain Score showed a decrease of 10 units for the diet group and an increase of 2 units for the control group. This difference between groups was statistically significant (p = 0.011; Mann-Whitney's U-test) and was confirmed with the dichoΔPain Score difference (p = 0.007) (Table 2).
Using logistic regression, there were statistically significant correlations between diet and the results from three outcome variables. These were the dichoΔAcute-Phase Response, the dichoΔPain Score and the dichoΔPhysical Function, but not with the dichoΔESR or the dichoΔTender Joint Count (Table 2).
Multivariate analyses
The significant correlations using multiple logistic regression were limited to the association of dichoDiet to the three outcome variables dichoΔAcute-Phase Response, the dichoΔPain Score and the dichoΔPhysical Function, but not to the dichoΔESR or the dichoΔTender Joint Count (Table 3). Hence, body weight reduction (dichoΔBody Weight) was not statistically significantly coupled to any outcome variables when diet was taken into account.
Table 3 Multiple logistic regression results (Odds ratios (OR) and 95% confidence intervals(95% CI)), showing the association between the only statistically significant independentvariable, diet on the one hand, and the three disease outcome variables, which showedstatistically significant correlation to type of diet, on the other.
Control diet Intervention diet
Imp/no imp Imp (%) OR Imp/no imp Imp (%) OR 95% CI p
dichoΔAcute PR 12/30 29 1.00 34/26 57 3.27 1.39 – 7.67 0.007
dichoΔPain Score 17/25 40 1.00 42/18 70 3.43 1.49 – 7.93 0.005
dichoΔPhysic. F 10/31 24 1.00 34/25 58 4.22 1.73 – 10.3 0.002
Footnotes to table:
ESR = The Westergren erythrocyte sedimentation rate
Acute-PR = The Acute-Phase Response (orosomucoid or C-reactive protein (CRP))
Pain Score = the patient's self perceived pain severity as evaluated on a visual analogue scale
Physic. F = Physical Function assessed by self completed questionnaires on the degree of difficulty in performing specified tasks of daily living.
Discussion
The time intervals between the 1st and 2nd, and between the 2nd and 3rd studies were approximately ten years. One of the authors was local clinical research coordinator for all three projects. The studies were conducted in three different regions of Sweden. However, the patient populations were very much alike, as all patients were recruited from similar clinical settings at the central out patient rheumatology clinic of each region.
In all three studies the patients had been aggressively treated pharmacologically according to the internationally recommended guidelines that were prevailing at the time of each study. As a consequence of the more efficient disease modifying drugs of the 1990ths, the arthritis activity was better controlled in the 3rd study than in the two earlier studies. With all three studies the participating patients had showed active interest for the kind of diet that was under investigation. At the time of the first two studies, vegetarian diets were advocated by laymen, while in the 1990ths much attention was focused on Mediterranean diets. Although different, these three diets shared some common characteristics which we believe were of importance with respect to control of inflammation. Compared to ordinary western diets they contained less of saturated fats from meat and dairy products. They had more of fresh fruits and of green vegetables. The MD was also rich in fats from sea foods, which was not a feature of the two vegetarian variants.
Another consequence of the time difference between the three studies was that the RA measures were not completely identical. They had successively been changed in line with the international recommendations of good scientific standards. Thus, comparable absolute baseline estimations of pain-score, acute phase response, physical function or tender joint count were not available due to their different definitions during the three intervention studies. Nevertheless, as explained in "methods", for the present analysis the outcome results of these variables were well defined, as they were dichotomously characterised as improvement, or no improvement.
The cross over design of the 2nd study, with 7 patients assessed first as controls and later as diet patients, weakens the statistical power of the pooled statistical analysis only marginally.
In the three studies of ours each intervention i.e. with lacto-vegetarian, vegan or modified Cretan Mediterranean diet had rendered the patients a weight fall of on the average 2.4 kg over a trial length of 3–4 months. In the pooled analysis of change in body weight versus change in clinical outcome measurements the results were fairly clear. Apart from a univariate correlation between improvement in acute phase response and reduction in body weight, no statistically significant correlation was seen with the multiple logistic regression analysis between weight loss and the concomitantly obtained change in RA disease measurements. Thus the body weight loss seemed to have had no statistical significant anti inflammatory effect. However, although the number of patients was increased from pooling three different studies, the total amount of data was still relatively small. Therefore a possible anti inflammatory effect of weight reduction should not be completely discarded. Furthermore, the potential effect that weight reduction may have, should ideally be studied by itself, i.e. the test and control diets should differ only with respect to their contents of energy.
As for comparison of diet intervention versus control diet, the multiple logistic regression analysis, showed highly significant statistical correlations between diet intervention and improvement in RA outcome variables. These data indicate that dietary factors may have a potential role in treatment of RA. With regards to vegan diets, our observation is supported by the results of at least two independent, randomised and controlled studies [6,9].
Of course, the beneficial factors that the tested diets share need to be identified. A noteworthy candidate is their relatively low content of saturated fats [13]. Another is their high content of fresh fruits and vegetables.
Patients with active RA tend to have an abnormally increased peripheral insulin resistance [14]. Deliberate weight reduction in a group of patients with gout [15] was accompanied by reduction of insulin resistance and less numbers of inflammatory events. Whether this strategy for gout would work for patients with RA, and improve their arthritis, is an important clinical question. Some recent observational data indicate that overweight and obesity might be a risk factor for RA [16]. From experimental studies in mice [17] it is known that long-term pure energy under-nutrition has anti-inflammatory effects. In humans, deliberately undertaken short term fasting is well known to induce immune suppression and improvement in RA disease activity [18]. There are no controlled long-term studies of reduced energy intake without mal-nutrition on patients with RA. However, Iwasahige et al. [19] recently conducted a regiment for 54 days of caloric restriction combined with fasting in ten patients with RA. The patients lost in weight, and interestingly, the composite disease activity score of Lansbury was significantly reduced.
We know of two other small studies on deliberately undertaken weight reduction in patients with RA. In an uncontrolled pilot study [20] with 19 overweight patients with RA, Danish researchers had instructed the patients to lower their energy intake by 30% to achieve weight reduction. After a period of 12 weeks, the mean weight loss was 4.5 kg. No change was obtained in joint pain, morning stiffness, number of tender joints, or in sedimentation rate. Neither did Gordon et al. from their uncontrolled pilot study report any short-term favourable effects from assisting obese RA patients to reduce their body weight [21].
In conclusion, it seems as weight reduction strategies have little if any influence on RA inflammation. Perhaps this is not surprising with regards to what is already known from research on the metabolic syndrome and on atherosclerotic vascular disease, where isolated overweight is rank as a less important factor of risk [22].
Before conclusion of this discussion we need to remember that there is a controversy in letting patients with RA test an experimental regiment, which would involve prolonged reduction of their energy intake. During the course of RA most RA patients including those with overweight will develop a muscle wasting condition known as rheumatoid cachexia [22]. Although not directly fatal, this form of cachexia is believed to contribute to co-morbidity and reduced life expectancy. It is caused by the rheumatoid inflammation by itself and is refractory to nutritional therapy.
==== Refs
Sköldstam L Hagfors L Johansson G Rheumatoid arthritis and Mediterranean diet EULAR Ann Rheum Dis 2003 62 208 214 10.1136/ard.62.3.208
Riel PL van Gestel AM Clinical outcome measures in rheumatoid arthritis Ann Rheum Dis 2000 28 31 10.1136/ard.59.suppl_1.i28
Sköldstam L Larsson L Lindström FD Effects of fasting and laktovegetarian diet on rheumatoid arthritis Scand J Rheumatol 1979 8 249 255 534318
Sköldstam L Fasting and vegan diet in rheumatoid arthritis Scand J Rheum 1986 15 219 221 3749829
Darlington LG Ramsey NW Mansfield JR Placebo-controlled, blind study of dietary manipulation therapy in rheumatoid arthritis Lancet 1986 1 236 238 2868255 10.1016/S0140-6736(86)90774-9
Kjeldsen-Kragh J Haugen M Borchgrevink CF Laerum E Eek M Mowinkel P Hovi K Førre Ø Controlled trial of fasting and one-year vegetarian diet in rheumatoid arthritis Lancet 1991 338 899 902 1681264 10.1016/0140-6736(91)91770-U
Hansen GV Nielsen L Kluger E Thysen M Emmertsen H Stengaard-Pedersen K Hansen EL Unger B Andersen PW Nutritional status of Danish rheumatoid arthritis patients and effects of diet adjustment in energy intake, fish-meal, and antioxidants Scand J Rheumatol 1996 25 325 330 8921927
Sarzi-Puttin P Comi D Boccassini L Muzzupappa S Turiel M Panni B Salvaggio A Diet therapy for rheumatoid arthritis Scand J Rheumatol 2000 29 302 307 11093596 10.1080/030097400447688
Hafström I Ringertz B Spångberg A von Zweigbergk L Brannemark S Nylander I Rönnelid J Laasonen L Klareskog L A vegan diet free of gluten improves the signs and symptoms of rheumatoid arthritis: the effects on arthritis correlate with a reduction in antibodies to food antigens Rheumatology 2001 40 1175 1179 11600749 10.1093/rheumatology/40.10.1175
Emery P Luqmani R The validity of surrogate markers in rheumatic disease Br J Rheumatol 1993 32 3 8 7685227
Ekdahl C Eberhardt K Andersson SI Svensson B Assessing disability in patients with rheumatoid arthritis: use of a Swedish version of the Stanford-Health Assessment Questionnaire Scand J Rheumatol 1988 17 263 271 3187457
Ritchie DM Boyle JA McInnnes JM Jasani MK Dalakos TG Grieveson P Buchanan WW Clinical studies with an articular index for the assessments of joint tenderness in patients with rheumatoid arthritis Q J Med 1968 147 393 406 4877784
Adam O Beringer C Kless C Lemmen C Adam A Wiseman M Adam P Klimmek R Forth W Anti-inflammatory effects of a low arachidonic acid diet and oil in patients with rheumatoid arthritis Rheumatol Int 2003 23 27 36 12548439
Svensson KLG Lundqvist G Wide L Hällgren R Impaired Glucos handling in active rheumatoid arthritis: effects of corticosteroids and rheumatic treatment Metabolism 1987 36 944 948 3309546 10.1016/0026-0495(87)90129-6
Dessein PH Shipton AE Joffe BI Ramokgadi J Beneficial effects of weight loss with moderate calorie / carbohyudrate restriction, and increased proportional intake of protein and unsaturated fat on serum urate and lipoprotein levels in gout: a pilot study Ann Rheum Dis 2000 59 539 543 10873964 10.1136/ard.59.7.539
Symmons DPM Bankhead CR Harrison BJ Brennan P Barrett EM Scott DG Stilman AJ Blood transfusion, smoking, and obesity as risk factors for the developement of rheumatoid arthritis: results from a primary care-based incident case control study in Norfolk, England Athritis Rheum 1997 40 1955 1961
Hart RW Dixit R Seng J Turturro A Leakey JE Feuers R Duffy P Buffington C Cowan G Lewis S Pipkin J Li SY Adaptive role of caloric intake on the disease processes Toxicological Science 1999 3 12
Fraser DA Thoen J Reseland JJ Førre Ø Kjeldsen-Kragh J Decreased CD4+ lymphocyte activation and increased interleukin-4 production in peripheral blood of rheumatoid arthritis patients after acute starvation Clin Rheumatol 1999 18 394 401 10524554 10.1007/s100670050125
Iwashige K Kouda K Kouda M Horiuchi K Takahashi M Nagano A Tanaka T Tacheuchi H Calorie restricted diet and urinary pentosidine in patients with rheumatoid artritis J Physiol Anthropol Appl Human Sci 2004 23 19 24 14757997 10.2114/jpa.23.19
Engelhart M Kondrup J Hoie LH Andersen V Kristensen JH Heitmann BL Weight reduction in obese patients with Rheumatoid Arthritis with preservation of body cell mass and improvement of physical fitness Clin Exp Rheumatol 1996 14 289 293 8809443
Gordon MM Thomsson EA Madhok R Capell H Can intervention modify adverse lifestyle variables in a rheumatoid population? Results of a pilot study Ann Rheum Dis 2002 61 66 69 11779763 10.1136/ard.61.1.66
Haapanen-Niemi N Vouri I Pasanen M Public health burden of coronary heart disease risk factors among middle aged and elderly men Prev Med 1999 28 343 348 10090863 10.1006/pmed.1998.0426
Walsmith J Roubenoff R Cachexia in rheumatoid arthritis Int J Cardiol 2002 85 89 99 12163213 10.1016/S0167-5273(02)00237-1
| 15871736 | PMC1156940 | CC BY | 2021-01-04 16:39:31 | no | Nutr J. 2005 May 4; 4:15 | utf-8 | Nutr J | 2,005 | 10.1186/1475-2891-4-15 | oa_comm |
==== Front
Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-161588820310.1186/1475-2891-4-16ResearchPlasma retinol, carotene and vitamin E concentrations and lung function in a crocidolite-exposed cohort from Wittenoom, Western Australia: a cohort study Alfonso Helman S [email protected] Lin [email protected] Klerk Nicholas H [email protected] Gina [email protected] John [email protected] Nola [email protected] A William [email protected] School of Population Health, University of Western Australia, Perth, Western Australia, Australia2 School of Public Health, Curtin University of Technology, Perth, Western Australia3 Department of Respiratory Medicine, Sir Charles Gairdner Hospital, Perth, Western Australia, Australia4 Department of Biostatistics and Genetic Epidemiology, Telethon Institute for Child Health Research, Perth, Western Australia, Australia5 Clinical Biochemistry, PathCentre, Western Australia2005 11 5 2005 4 16 16 4 2 2005 11 5 2005 Copyright © 2005 Alfonso et al; licensee BioMed Central Ltd.2005Alfonso et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Increased rates of death from asbestos related diseases have been reported for people previously employed in the mining and milling operations at Wittenoom (Western Australia), and people who lived in the nearby town, where they were environmentally exposed to crocidolite.
Methods
Annual measurements of forced expiratory volume in 1 second (FEV1) and forced vital capacity (FVC) and plasma concentrations of retinol, carotene and vitamin E have been made since 1992. Mixed effects models were used to examine the associations between lung function and the plasma vitamin levels of retinol, carotene and vitamin E.
Results
After adjusting for potential confounders, higher plasma retinol and carotene concentrations were significantly associated with higher levels of lung function at entry into the study, while vitamin E concentrations were associated with lower entry lung function. Retinol was associated with a less steep decline of lung function over time, while carotene concentrations were associated with an increased decline of lung function over time and vitamin E levels were not associated with changes of lung function over time.
Conclusion
These results support a beneficial relationship between plasma concentrations of retinol on the levels and rates of change of lung function, while showing no such consistent beneficial effect for plasma levels of beta-carotene or vitamin E.
==== Body
Introduction
Respiratory impairment is an important cause of disability in asbestos-exposed people. Cross-sectional and longitudinal studies have shown that greater intensity and duration of asbestos exposure are associated with greater impairment of pulmonary function [1]. A protective effect of fruit and vegetable consumption on lung function has been reported, and a number of studies have found positive associations between pulmonary function and dietary intake and blood levels of both carotenoids and retinoids [2-4].
Crocidolite (blue asbestos) was mined at Wittenoom, Western Australia from 1943 until 1966. Approximately 7000 workers were employed in the mining and milling operations [5]. In addition, about 5000 people, who did not work in the production of asbestos, are documented to have lived in the nearby town, where they received environmental exposure to asbestos [6]. Former workers and residents were also frequently cigarette smokers, and increased risks of asbestos- and smoking-related diseases have been shown in these populations [6].
Cohorts of former workers and residents of Wittenoom have been followed up continuously since 1975, and since 1990 some of them have participated in an intervention program including supplemental vitamin A in an attempt to reduce the occurrence of malignant mesothelioma and lung cancer [7,8]. The Program also included advice on diet, stopping smoking and physical exercise. As no beneficial effect was reported for beta-carotene after five years [7,8], and other studies showed increased risks of developing lung cancer in subjects taking beta-carotene supplements [9,10], all participants have been provided with retinol since 1997.
The aim of this analysis was to examine the relationships between lung function and plasma levels of retinol, carotene and vitamin E, after controlling for potential confounders, in people exposed to crocidolite in Wittenoom who have participated in the Vitamin A Program. In analysing the relationships between plasma vitamin concentrations and lung function, we are not intending to assess of the efficacy of the Program, which will be described in a separate report (Musk et al, in preparation). The predictors of the levels and the rate of change of lung function, as well as the assessment of the efficacy of the Program, have been described in separate reports ([11,33]).
Subjects
Cohorts of former workers and residents of Wittenoom have been followed up continuously since 1975. In 1989, all ex-residents or ex-workers that could be located were invited to participate in the Vitamin A Program, and the Program has commenced accepting in July 1990. Each participant attends a clinic annually to receive a year's supply of the vitamin supplement. Annual spirometric measurements and plasma vitamins determinations have also been taken since 1992. All participants who had at least one spirometric measurement and a blood test for plasma vitamin concentrations performed on the same day were included in the present study. Participants were followed from the time of their first test, until the last result available before 23 September 2002. People younger than 25 years old were excluded from the analysis, because growth in lung function continues until this age [12].
Methods
Plasma vitamin concentrations
On the same day as the spirometry was performed, a non-fasting blood sample was collected and the plasma concentrations of retinol, total carotene, alpha-carotene, beta-carotene and alpha-tocopherol (vitamin E) were determined by reverse-phase high performance liquid chromatography (HPLC) [13]. Total carotene was measured from July 1990 to May 1994, but subsequently, alpha- and beta-carotene were measured separately. For this analysis, after May 1994, total carotene was assumed to be the sum of alpha-carotene and beta-carotene. For measurements whose plasma values were undetectable by laboratory methods, a value midpoint between zero and the laboratory's minimum detectable value was used (42% in alpha-carotene measurements, and 19% in beta-carotene) [14,33].
Pulmonary Function Testing
FEV1 and FVC measurements were performed by trained technicians according to the guidelines of the American Thoracic Society (ATS) [15] and adjusted for body temperature and pressure-saturated with water vapour. The highest of at least three technically satisfactory attempts was selected. Only measurements in which the two highest attempts fulfilled the ATS criteria for reproducibility (an agreement within 5%) were included in the analysis.
Asbestos Exposure Assessment
Category of exposure (former worker or resident of Wittenoom), age at first exposure, cumulative asbestos exposure and radiographic asbestosis at first visit were the indicators of crocidolite exposure used in the present analysis. The three first indicators were obtained from the Vitamin A Program records.
Methods describing the assessment of cumulative asbestos exposure for workers have been described previously [16,33]: job histories were obtained from employment records, and fibre concentrations for all job categories were estimated from the results of a survey of airborne respirable fibres of crocidolite that was carried out at various work sites at Wittenoom in 1966, and particle counts performed by the Mines Department of Western Australia throughout the period of operation of the mining industry. Each subject's cumulative exposure was calculated by adding over all their different jobs the product of fibre concentration and the length of time in that job.
For ex-residents, the estimation of individual asbestos exposure levels has also been described in detail elsewhere [6]. Subjects not working directly with asbestos were assigned an intensity of exposure of 1 fibre/millilitre (f/ml) from 1943 to 1957, when a new mill was commissioned and the town was moved, and then 0.5 f/ml between 1958 and 1966, when the mining operation ceased. Since then, interpolation between periodic hygiene surveys using personal monitors assigned exposures from 0.5 f/ml in 1966 to 0.01 f/ml in 1992. Duration of residence was obtained from a questionnaire completed at the time of attending the Vitamin A Program. Individual cumulative exposure was calculated by combining the duration of residence and the intensity of exposure for each person. These calculations have been demonstrated to be internally valid on the basis of lung fibre determinations and dose-response characteristics for mesothelioma, lung cancer and asbestosis [6].
To assess radiographic asbestosis, without knowledge of exposure status, a panel of trained and experienced readers read the plain chest radiograph at the first visit according to the International Labour Office's (ILO) classification of the chest x-ray manifestations of pneumoconioses [17], utilizing the standard films provided by the ILO for side by side comparison. For the purpose of this study, radiographic asbestosis was defined as a profusion score of 1/0 or greater, as indicated in the ILO guidelines [17].
Smoking History Assessment
Smoking history was obtained from a questionnaire administered at entry to the Program. Information on smoking status (never-smokers, ex-smokers and current smokers) was evaluated as a categorical variable. The history over time of smoking was not available for this analysis.
Statistical Analysis
To assess the relationships between lung function and plasma vitamin concentrations, the dependent variables (FEV1 and FVC) were regressed on time, controlling for demographic characteristics (sex, age and height), asbestos exposure and tobacco smoking. Plasma levels of retinol, carotene and vitamin E (as quartiles) were included in the model, both separately and at the same time. A linear trend was also evaluated by including the raw values (continuos variable) of each plasma vitamin concentration. The main effect of each vitamin concentration was interpreted as the level of lung function at entry into the study, while the interaction of each vitamin concentration with time of follow up (in years) was interpreted as the annual change of lung function over time [18]. Specifically, levels of lung function at entry into the Program correspond to the values of lung function (or differences between groups) when time of follow up was zero. The analyses were performed initially for each vitamin separately, and then by including simultaneously the three vitamin values into the model.
The data were modelled by a general linear mixed-effects model, using SAS PROC MIXED [19]. Random-effects models are most appropriate for unbalanced longitudinal data or when the intervals between measurements for each subject are not equally spaced [18]. The model-building process was performed following the steps suggested by Verbeke and Molenberghs [18]. Initially, a model with all the explanatory variables and their biologically plausible interactions was applied in order to remove any systematic trends. In the second step, random effects were included in the model; the best model was selected according to a likelihood ratio test. Thirdly, several residual covariance structures were tested and the best was selected according to the Akaike Information Criterion (AIC). Finally, the model was simplified by deleting non-significant terms. Estimation was made by the restricted maximum likelihood method, and tests were performed using the 5% level of significance. Model validation was carried out by checking normal distribution of residuals.
Results
A total of 5,750 determinations from 1,378 subjects who had at least one test of both spirometry and plasma vitamin concentrations were analysed. The demographic characteristics and exposure histories at the first visit showed that former workers had lower levels of FEV1 and FVC than ex-residents (Table 1). Workers were more likely to be male and residents were female. Workers were exposed to higher cumulative amounts of asbestos (median 6.40 fibres/ml per year, interquartile range 1.92–26.01), and included a higher proportion of participants with radiographic asbestosis. Although ex-residents were exposed to asbestos for longer periods of time, their cumulative asbestos exposure was lower, as the intensity of exposure was much higher in the mining and milling processes. Ex-residents tended to be exposed to asbestos at a younger age than workers. In addition, workers included a higher proportion of ever-smokers. Initial levels of plasma vitamins were similar between workers and ex-residents. At the first visit, plasma retinol concentrations were not significantly correlated with plasma carotene concentrations (r = -0.006, p = 0.84); plasma concentrations of retinol and vitamin E were highly correlated (r = 0.33, p = <0.0001); and carotene and vitamin E concentrations tended to be correlated (r = 0.047, p = 0.08).
Table 1 Demographic characteristics of participants at first visit a
Residents Workers
Participants, n (%) 567 (41.1) 811 (58.9)
Male, n (%) 270 (47.6) 749 (92.4)
FEV1, litres a 2.9 (0.9) 2.7 (0.7)
FVC, litres a 3.7 (1.1) 3.6 (0.9)
Age, yr a 50.8 (12.5) 59.2 (7.7)
Height at baseline, cm a 168 (9.5) 171 (7.5)
Cumulative asbestos, f/ml-year a 6.9 (7.5) 24.5 (47.8)
Age at first asbestos exposure, yr a 13.9 (12.7) 24.8 (6.1)
Radiographic asbestosis, n (%) 8 (1.4) 143 (17.6)
Current-smokers, n (%) 114 (19.6) 182 (21.8)
Ex-smokers, n (%) 183 (31.5) 452 (54.2)
Never-smokers, n (%) 284 (48.9) 199 (23.9)
Plasma retinol, (μmol/L)a 2.70 (0.70) 2.70 (0.69)
Plasma carotene, (μmol/L)a 1.38 (1.57) 1.38 (1.52)
Plasma vitamin E, (μmol/L)a 38.13 (13.85) 35.43 (11.40)
amean (SD) range
Twenty-four percent of participants had one visit at which lung function and plasma vitamins were measured, 21% had two measurements, 17% had three, 14% had four, 10% had five, and 14% had more than 5 measurements during the period of observation. The Vitamin A Program initially gave preference to enrolling workers, who were mostly males, because of their greater risks of developing asbestos-related diseases. Therefore they had more spirometric measurements per person and a longer follow-up time (Table 2).
Table 2 Characteristics of the follow up of the study cohort a
Residents Workers
Total lung function and blood tests, n (%) 1942 (33.7) 3898 (66.2)
Measurements per person, a 3.43 (1.5) 4.7 (2.5)
Follow up, years, a 3.0 (1.7) 5.4 (3.3)
Months between measurements,a 14.0 (4.8) 16.5 (9.7)
amean (SD
After adjusting for confounders, plasma retinol concentrations (as a single vitamin in the model) at entry into the study were not associated with levels of lung function (Table 3), (p-value for linear trend = 0.36 for FEV1, p = 0.50 for FVC). For example, people in the third quartile (Q3) have, on average, 4.4 ml more of FEV1 and 6.6 ml more of FVC, compared to those in the first quartile (Q1). However, these differences were not statistically significant as the confidence intervals contain negative values. However, higher concentrations of plasma retinol were associated with a lower rate of decrease in lung function over time: the annual decline in people in the highest quartile of retinol concentration was 11.3 ml (95% CI = 4.8–17.7) of FEV1 and 18.6 ml (95% CI = 10.4–26.8) of FVC less steeper than those in the lowest quartile. The linear trends were significant for FEV1 and FVC (not shown in table 3).
Table 3 Relationships between plasma retinol concentrations and lung function levels and annual rate of change*
Quartiles FEV1 95%CI FVC 95%CI
Level at entry into program (ml) Q1 0 0
Q2 5.5 -15.5,26.6 -8.9 -35.8,17.9
Q3 4.4 -19.8,28.7 6.6 -24.4,37.6
Q4 2.6 -25.5,30.9 -16.8 -52.7,19.1
Annual change (ml/year) Q1 0 0 0
Q2 4.1 -1.8,10.1 10.3 2.7,17.9
Q3 7.1 1.0,13.4 11.1 3.2,19.1
Q4 11.3 4.8,17.7 18.6 2.7,10.3
*Adjusted for age, sex, height, asbestos exposure and smoking history
Quartile 1: Retinol concentrations lower than 2.5 μmol/L; Quartile 2: concentrations more than 2.5 and less than 3.0 μmol/L; quartile 3: concentrations more than 3.0 and less than 3.5; and quartile 4: concentrations equal or greater than 3.5 μmol/L.
Higher plasma carotene concentrations (as a single vitamin in the model) were associated with higher levels of lung function at entry into the study (Table 4). On the contrary, higher plasma carotene concentrations were associated with a steeper decline of lung function. The annual decline in people in the highest quartile was 9.3 ml (95% CI = 2.6–15.9) of FEV1, and 8.3 ml (95% CI = 0–16.5) of FVC less steeper than those in the lowest quartile (Table 4). Linear trends were significant for both levels and rates of change of lung function (not shown).
Table 4 Relationships between plasma carotene concentrations and lung function levels and annual rate of change*
Quartiles FEV1 95%CI FVC 95%CI
Level at entry into program (ml) Q1 0 0
Q2 21.4 -3.2,46.1 0.06 -31.4,31.6
Q3 47.8 19.1,76.4 30.6 -5.8,67.0
Q4 67.3 38.0,96.6 59.5 22.6,96.5
Annual change (ml/year) Q1 0 0 0 0
Q2 -1.1 -6.5,4.3 3.04 -3.8-9.9
Q3 -6.0 -12.3,0.3 -6.6 -14.6-1.5
Q4 -9.3 -16.0,-2.7 -8.3 -16.8-0.1
Adjusted for age, sex, height, asbestos exposure and smoking history.
Quartile 1: Carotene concentrations lower than 0.45 μmol/L; Quartile 2: concentrations more than 0.45 and less than 0.8 μmol/L; quartile 3: concentrations more than 0.8 and less than 1.3; and quartile 4: concentrations equal or greater than 1.3 μmol/L.
Higher plasma vitamin E concentrations (as a single vitamin in the model) were associated with higher levels of lung function at entry into the study (Table 4). However, higher plasma vitamin E concentrations were associated with a steeper decline of lung function. The decline in people in the highest quartile was 7.9 ml (95% CI = 14.3-1.6) per year in FEV1, and 13.9 ml (95% CI = 21.9-6.0) per year in FVC, compared to those in the lowest quartile (Table 4). Linear trends were significant for both levels and rates of change of lung function (not shown).
Including retinol, carotene and vitamin E plasma concentrations in the same model, and adjusting for potential confounders, higher plasma retinol concentrations were associated with higher levels of lung function, as well as a slower decline in the annual rate of change of lung function (Table 6). Plasma carotene concentrations were associated with higher levels of lung function at entry, but with a steeper decline in lung function over time. Plasma vitamin E concentrations were associated with lower levels of lung function at entry into the study, but not associated with changes over time.
Table 6 Relationships between lung function and retinol, carotene and vitamin E concentrations, when the three vitamins were simultaneously included in the model*
Plasma vitamin concentrations (μMol/L) FEV1 (ml) Estimate (SD) p-value FVC (ml) Estimate (SD) p-value
Retinol Level 13.9 (7.5) 0.06 16.6 (9.5) 0.08
Change 3.5 (1.5) 0.02 5.7 (1.9) 0.00
Carotene Level 14.6 (3.6) 0.00 13.1 (4.5) 0.00
Change -2.6 (1.0) 0.01 -3.5 (1.3) 0.01
Vitamin E Level -1.0 (0.4) 0.01 -2.0 (0.5) <0.01
Change 0.07 (0.09) 0.38 0.2 (0.1) 0.13
* Adjusted for age, sex, height, asbestos exposure and smoking history. Only results for linear trend are shown.
The analysis was also performed for alpha- and beta-carotene. As was reported for total carotene concentrations, both alpha- and beta-carotene concentrations were associated with higher levels of lung function, and inversely related to the change of lung function over time (not shown).
Discussion
After adjusting for potential confounders and the joint effect of the plasma vitamin levels, plasma retinol concentrations were associated with higher levels of lung function as well as a slower rate of decline over time; plasma carotene concentrations (or alpha and beta-carotene) were associated with higher levels of lung function, but with a steeper decline in lung function; plasma vitamin E concentrations were associated with lower levels of lung function at entry, but not associated with amount change of lung function over time. The observed tendency in the separate analysis for each vitamin (Tables 3 to 5) was similar to the results obtained when the three variables were jointly included in the model, except for the effect of plasma retinol concentrations, which became positively associated with levels of lung function at the entry in the Vitamin A Program. We have previously shown that mortality in subjects with asbestosis was inversley related to plasma concentrations of retinol and vitamin E (at first visit and during the follow up period), while carotene concentrations at the first visit were associated with lower mortality but not during the follow up period [33]
Table 5 Relationships between plasma vitamin E concentrations and lung function levels and annual rate of change*
Quartiles FEV1 95%CI FVC 95%CI
Level at entry into program (ml) Q1 0 0
Q2 23.5 -1.9, 49.0 33.5 1.0, 66.0
Q3 38.4 12.0, 64.8 67.0 33.4, 100.6
Q4 35.9 8.1, 63.7 82.3 47.1, 117.5
Annual change (ml/year) Q1 0 0
Q2 -6.8 -10.1, 0.9 -4.4 -11.4, 2.7
Q3 -4.6 -12.7, -1.0 -7.8 -15.3, -0.3
Q4 -7.9 -14.3, -1.6 -13.9 -21.9, -6.0
Adjusted for age, sex, height, asbestos exposure and smoking history.
Quartile 1: Vitamin E concentrations lower than 31 μmol/L; Quartile 2: concentrations more than 31 and less than 37 μmol/L; quartile 3: concentrations more than 37 and less than 45; and quartile 4: concentrations equal or greater than 45 μmol/L.
The positive effect of plasma retinol on lung function has been recognised in previous cross-sectional and prospective studies [3,4], but not in longitudinal studies with multiple measurements per participant. This positive effect is consistent with the multiple biologic functions of retinoids, including regulation of cell proliferation, cell differentiation and morphogenesis [20]. In contrast to carotenoids, retinol is not a potent free radical scavenger; therefore other mechanisms have been advocated for explaining its possible beneficial effects [2]. For example, retinoids have a variety of effects on the cell membrane, involving altered glycoprotein synthesis and changes in membrane receptors for various hormones, including those mediated by c-AMP. The actions on these receptors may influence cell-cell interactions, cell adhesion, and cell membrane permeability. In a recent animal model, systemic administration of all-trans-retinoic acid (ATRA) reversed manifestations of elastase-induced emphysema [21]. Similar results have now been replicated by Belloni and colleagues with both ATRA and 9-cis-retinoic acid [22]. This is a promising finding, although is not possible to draw a close parallel between an elastase-induced emphysema murine model and the pathogenesis of human chronic obstructive pulmonary disease.
The positive association between blood carotene concentrations and the levels of lung function has also been previously recognized in cross-sectional and prospective studies [3,23,24]. The biological plausibility for a protective effect of carotenoids on lung function is associated with the multiple biological actions of these micronutrients, in particular their role in scavenging oxidised products [2]. It has been postulated that antioxidant defences play a critical role in the respiratory tract where high exposure to endogenous and environmental oxidants occurs, which may impair ventilatory function by inducing inflammation and destruction of the alveolar walls [25]. However, our results showed that plasma carotene concentrations were associated with a steeper decline of lung function. This disappointing effect mirrors the results obtained in this Vitamin A Program and other intervention studies, in which higher-than-expected rates of lung cancer were found in people taking beta-carotene [9,10].
Several observational studies have consistently showed that individuals who eat more fruits and vegetables tend to have lower risk of developing various cancers and other chronic diseases [26]. Several studies have found positive associations between pulmonary function and dietary intake and blood levels of both carotenoids and retinoids [2]. From the many potential nutrients in fruits and vegetables, beta-carotene was credited as the most important nutrient likely to be responsible for this beneficial effect, especially following the influential paper from Peto and colleagues [27]. However, results from at least 8 trials have shown that beta-carotene supplementation has essentially no significant role in decreasing cancer risk [28]. No previous studies have been reported on relationships between supplemental beta-carotene and lung function.
Two alternative explanations have been proposed to explain the conflicting results between observational and intervention studies: first, that beta-carotene is not the factor responsible for the beneficial effect of fruit and vegetable intake on cancer [28]; second, that the supplemental synthetic all-trans-beta-carotene does not reproduce the effect of the natural counterpart [29].
It is possible that many of the studies suggesting that dietary beta-carotene is protective were confounded [28]. Observational studies do not permit causal relationships to be established. Hundreds of carotenoids and thousands of nutrients exist in foods. Researchers are only able to examine and study a few hundred of them, and the synergy among these numerous nutrients may play a larger role that any single compound [28]. Although the reasons remain unclear, current wisdom indicates no role for synthetic beta-carotene supplementation for preventing cancer or improving lung function.
Despite being considered a powerful anti-oxidant, we did not find a consistent effect of Vitamin E on lung function. Whether the inverse relationships reported in the univariate analysis of this study reflects real relations peculiar to the subjects studied, or are the result of sampling variation or chance may be clarified by future studies.
Due to self-selection into the Vitamin A Program this population may not be representative of the whole cohort exposed to crocidolite at Wittenoom. People who participated in the Program were younger and had higher cumulative asbestos exposure [7]. The results from this study, may not apply to subjects with lower exposure to crocidolite or older age groups.
An additional difficulty in interpreting the results of our study may be that we examined vitamin levels in a population in which the vitamin levels were artificially increased and did not adjust for their assigned dose. Although the arm of the study to which participants had originally been assigned was known, we had no precise way of measuring their actual intake of vitamins, nor their dietary intake of fruit and vegetables. However, epidemiological theory [30] postulates a hierarchy of exposure measurement from available dose in the general environment (assigned vitamins), to administered dose (number of tablets taken), to absorbed dose (dose taken up from the gastrointestinal tract), to active dose at different target sites. The closer that the measurement is to the active dose, the more likely it is that a study will disclose any real relationship. Since we were able to measure the serum level of vitamins, adjusting for the assigned dose does not make theoretical sense, and would possibly artifactually decrease any relationship observed.
One problem in longitudinal studies is the occurrence of missing data due to withdrawals, which may result in follow-up bias [31]. Ninety subjects (6.3% of all participants) dropped out of the study, and 142 participants who during the follow up period. The people who died were not counted as dropouts as they had complete data [18]. People who abandoned the Program had lower levels of FEV1, FVC and FEV1 /FVC, compared to those who remained. We have assumed a 'missing at random pattern', implying that any withdrawal from the study did not depend on the level of lung function itself, although it may depend on some covariates. With a missing at random pattern, methods based on likelihood will produce valid inferences, if the model specification is correct [32]. Applying an available case analysis, using PROC MIXED of SAS, the reported relationships may be interpreted as the response conditional on the subject remaining in the study [18].
The present study has the advantage of having both cross-sectional and longitudinal components in the analysis of the data, which make efficient use of all the available information and explores hypotheses unable to be tested with cross-sectional approaches alone [33]. However, the accurate evaluation of tobacco smoke exposure in this report, as in most others, has some limitations because it was based on self-reporting. This report analyses the smoking status at the beginning of the follow up, and does not include the analysis of changes in smoking habits over time.
This longitudinal study demonstrates a beneficial relationship between plasma levels of retinol and level and rate of decline of lung function as measure by FEV1 and FVC, while showing no such beneficial effect for plasma levels of beta-carotene.
Ethical Approval
All subjects gave their informed consent and the study was approved by the Human Research Ethics Committee of the University of Western Australia and the Clinical Drug Trials Committee of the Sir Charles Gairdner Hospital, Nedlands, Western Australia.
Acknowledgements
The Vitamin A program has been funded by the Health Department of Western Australia and the Worker's Compensation and Rehabilitation Commission of Western Australia. The follow up of the Wittenoom cohort has been funded by the National Health and Medical Research Council (NHMRC) of Australia. We are grateful to the participants of the Program and the Asbestos Diseases Society of Western Australia for their ongoing support. We thank Lynne Watts, Meralyn Pearce, Diane Jacoby and Naomi Hammond for data collection, Ms J Sleith for data entry and Mr Philip Etherington for database management.
==== Refs
Rom WN Accelerated loss of lung function and alveolitis in a longitudinal study of non-smoking individuals with occupational exposure to asbestos American Journal of Industrial Medicine 1992 21 835 844 1621691
Schunemann HJ Freudenheim JL Grant BJ Epidemiologic evidence linking antioxidant vitamins to pulmonary function and airway obstruction Epidemiol Rev 2001 23 248 267 12192736
Chuwers P Barnhart S Blanc P Brodkin CA Cullen M Kelly T Keogh J Omenn GS Williams J Balmes JR The protective effect of beta-carotene and retinol on ventilatory function in an asbestos-exposed cohort American Journal of Respiratory and Critical Care Medicine 1997 155 1066 1071 9116988
Grievink L Smit HA Veer P Brunekreef B Kromhout D Plasma concentrations of the antioxidants beta-carotene and alpha-tocopherol in relation to lung function European Journal of Clinical Nutrition 1999 53 813 817 10556989 10.1038/sj.ejcn.1600854
Musk AW de Klerk NH Eccles JL Hobbs MS Armstrong BK Layman L McNulty JC Wittenoom, Western Australia: A modern industrial disaster American Journal of Industrial Medicine 1992 21 735 747 1609818
Hansen J de Klerk NH Musk AW Hobbs MS Environmental exposure to crocidolite and mesothelioma: exposure-response relationships American Journal of Respiratory and Critical Care Medicine 1998 157 69 75 9445280
Musk AW de Klerk NH Ambrosini GL Eccles JL Hansen J Olsen NJ Watts VL Lund HG Pang SC Beilby J Hobbs MS Vitamin A and cancer prevention I: observations in workers previously exposed to asbestos at Wittenoom, Western Australia International Journal of Cancer 1998 75 355 361 10.1002/(SICI)1097-0215(19980130)75:3<355::AID-IJC5>3.0.CO;2-1
de Klerk NH Musk AW Ambrosini GL Eccles JL Hansen J Olsen N Watts VL Lund HG Pang SC Beilby J Hobbs MS Vitamin A and cancer prevention II: comparison of the effects of retinol and beta-carotene International Journal of Cancer 1998 75 362 367 10.1002/(SICI)1097-0215(19980130)75:3<362::AID-IJC6>3.0.CO;2-0
Omenn GS Goodman GE Thornquist MD Balmes J Cullen MR Glass A Keogh JP Meyskens FL Valanis B Williams JH Barnhart S Hammar S Effects of a combination of beta carotene and vitamin A on lung cancer and cardiovascular disease. New England Journal of Medicine 1996 334 1150 1155 8602180 10.1056/NEJM199605023341802
ATBCCPSG The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers. The Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group. New England Journal of Medicine 1994 330 1029 1035 8127329 10.1056/NEJM199404143301501
Alfonso HS Fritschi L de Klerk NH Olsen N Sleith J Musk AW Effects of asbestos and smoking on the levels and rates of change of lung function in a crocidolite exposed cohort in Western Australia Thorax 2004 59 1052 1056 15563704 10.1136/thx.2004.022806
Quanjer PH Tammeling GJ Cotes JE Pedersen OF Peslin R Yernault JC Lung volumes and forced ventilatory flows. Report Working Party Standardization of Lung Function Tests, European Community for Steel and Coal. Official Statement of the European Respiratory Society European Respiratory Journal 1993 16 5 40 8499054
Milne DB Botnen J Retinol, alpha-tocopherol, lycopene, and alpha- and beta-carotene simultaneously determined in plasma by isocratic liquid chromatography Clinical Chemistry 1986 32 874 876 3084131
Clarke JU Evaluation of Censored Data Methods to Allow Statistical Comparisons Among Very Small Samples with Below Detection Limit Observations Environmental Science & Technology 1998 32 177 183 10.1021/es970521v
American Thoracic Society Standardization of spirometry--1987 update. Statement of the American Thoracic Society American Review of Respiratory Disease 1987 136 1285 1298 3674589
de Klerk NH Musk AW Armstrong BK Hobbs MS Smoking, exposure to crocidolite, and the incidence of lung cancer and asbestosis British Journal of Industrial Medicine 1991 48 412 417 1648376
International Labour Organization International classification of radiographs of pneumoconiosis. Guidelines for the use International Labor Organization (ILO). Geneva: International Labor Office. 1980
Verbeke G Molenberghs G Linear mixed models for longitudinal analysis 2000 New York, Springer-Verlag
SAS/STAT User's guide, version 8.2 1999 Cary, NC: SAS Institute Inc.,
Yoshimi T Takahashi Y Takahashi S Miura T Changes in lung-specific molecular expression during differentiation of hamster embryonic M3E3/C3 cell line Biochemistry and Cell Biology 2000 78 659 666 11206576 10.1139/bcb-78-6-659
Massaro GD Massaro D Retinoic acid treatment partially rescues failed septation in rats and in mice American Journal of Physiology 2000 278 L955 60 10781425
Belloni PN Garvin L Mao CP Bailey-Healy I Leaffer D Effects of all-trans-retinoic acid in promoting alveolar repair Chest 2000 117 235S 41S 10843926 10.1378/chest.117.5_suppl_1.235S
Schunemann HJ Grant BJ Freudenheim JL Muti P Browne RW Drake JA Klocke RA Trevisan M The relation of serum levels of antioxidant vitamins C and E, retinol and carotenoids with pulmonary function in the general population. American Journal of Respiratory and Critical Care Medicine 2001 163 1246 1255 11316666
Grievink L de Waart FG Schouten EG Kok FJ Serum carotenoids, alpha-tocopherol, and lung function among Dutch elderly American Journal of Respiratory and Critical Care Medicine 2000 161 790 795 10712323
MacNee W Rahman I Is oxidative stress central to the pathogenesis of chronic obstructive pulmonary disease? Trends Mol Med 2001 7 55 62 11286755 10.1016/S1471-4914(01)01912-8
Steinmetz KA Potter JD Vegetables, fruit, and cancer prevention: a review Journal of the American Dietetic Association 1996 96 1027 1039 8841165 10.1016/S0002-8223(96)00273-8
Peto R Doll R Buckley JD Sporn MB Can dietary beta-carotene materially reduce human cancer rates? Nature 1981 290 201 208 7010181 10.1038/290201a0
Moyad MA Results and lessons from clinical trials using dietary supplements for cancer: direct and indirect investigations Seminars in Urologic Oncology 2001 19 232 246 11769876
Young AJ Lowe GM Antioxidant and prooxidant properties of carotenoids Archives of Biochemistry and Biophysics 2001 385 20 27 11361018 10.1006/abbi.2000.2149
Armstrong BK White E Saracci R Principles of Exposure Measurement in Epidemiology 1992 Oxford, New York, Oxford University Press 229 233
Touloumi G Pocock SJ Babiker AG Darbyshire JH Estimation and comparison of rates of change in longitudinal studies with informative drop-outs Statistics in Medicine 1999 18 1215 1233 10363341 10.1002/(SICI)1097-0258(19990530)18:10<1215::AID-SIM118>3.3.CO;2-Y
Littell RC Rubin DB Statistical analysis with missing data 1987 New York, John Wiley
Alfonso HS Fritschi L de Klerk NH Ambrosini G Beilby J Oilsen N Musk AW Plasma concentrations of retinol, carotene, and vitamin E and mortality in subjects with asbestosis in a cohort exposed to crocidolite in Wittenoom, Western Australia J Occup Environ Med 2005 47 573 579 15951717 10.1097/01.jom.0000165750.54747.27
| 15888203 | PMC1156941 | CC BY | 2021-01-04 16:39:31 | no | Nutr J. 2005 May 11; 4:16 | utf-8 | Nutr J | 2,005 | 10.1186/1475-2891-4-16 | oa_comm |
==== Front
Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-171589007610.1186/1475-2891-4-17ResearchRelationship between maternal obesity and infant feeding-interactions Rising Russell [email protected] Fima [email protected] EMTAC Inc., 514 Santander Ave, #5, Coral Gables, FL 33134 USA2 Sansum Diabetes Research Institute, 2219 Bath St., Santa Barbara, CA 93105 USA2005 12 5 2005 4 17 17 16 2 2005 12 5 2005 Copyright © 2005 Rising and Lifshitz; licensee BioMed Central Ltd.2005Rising and Lifshitz; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
There are no data regarding the relationship between maternal adiposity and interaction and feeding of infants and possible contribution to childhood obesity. In this study we determined the relationship between maternal body weight and composition and infant feeding patterns and maternal-infant interaction during 24-hour metabolic rate measurements in the Enhanced Metabolic Testing Activity Chamber (EMTAC).
Methods
The amount of time four obese (BMI = 33.5 ± 5.3 kg/m2) and three normal weight (BMI = 23.1 ± 0.6 kg/m2) biological mothers, spent feeding and interacting with their infants, along with what they ingested, was recorded during 24-hour metabolic rate measurements in the EMTAC. The seven infants were 4.9 ± 0.7 months, 69 ± 3 cm, 7.5 ± 0.8 kg, 26 ± 3 % fat and 29 ± 25 percentile for weight for length. Energy and macronutrient intake (kcal/kg) were assessed. Maternal body composition was determined by air displacement plethysmorgraphy and that of the infants by skin-fold thicknesses. Pearson correlations and independent t-tests were utilized for statistical analysis (p < 0.05).
Results
Infants born to obese biological mothers consumed more energy (87.6 ± 18.9 vs. 68.1 ± 17.3) and energy as carbohydrate (25 ± 6 vs.16 ± 3; p < 0.05) than their normal weight counterparts. Most of the increased intake was due to complementary feedings. Twenty-four hour infant energy intake increased with both greater maternal body weight (r = 0.73;p < 0.06) and percent body fat. Furthermore, obese biological mothers spent less total time interacting (570 ± 13 vs. 381 ± 30 minutes) and feeding (298 ± 32 vs.176 ± 22 minutes) (p < 0.05) their infants than their normal weight counterparts. Twenty-four hour interaction time negatively correlated with both maternal body weight (r = -0.98; p < 0.01) and percent body fat (r = -0.92; p < 0.01). Moreover, infants of obese mothers slept more (783 ± 38 vs. 682 ± 32 minutes; p < 0.05) than their normal weight counterparts. However, there were no differences in total 24-hour energy expenditure, resting and sleeping metabolic rates (kcal/kg) for infants born to obese and normal weight biological mothers.
Conclusion
Greater maternal body weight and percent body fat were associated with greater infant energy intakes. These infants were fed less frequently and consumed more carbohydrates in a shorter period of time as compared to infants from normal weight biological mothers. These variations in feeding patterns may predispose certain infants to obesity.
==== Body
Introduction
Childhood obesity is now recognized as a national health epidemic, doubling among children 6–11 years old between the last (1976–1980) and the present (1999–2000) National Health and Nutrition Examination Surveys. Obesity in early childhood leads to more severe adult obesity [1]. Overweight and obese children are facing adult onset disease type risk factors such as cardiovascular heart disease, elevated blood pressure and type II diabetes [2]. Presently, type II diabetes represents 8–45% of all new cases of diabetes diagnosed in children and adolescents [2]. This epidemic increase in childhood obesity can only lead to a lower quality and duration of life and increased health care costs [3,4].
Many studies have identified the influence of genetics and environmental factors on potential causes of childhood obesity [5,6] and [7]. Some of these include having obese biological parents [8] and biological mothers being of a low social economic status [9,10]. There are other factors in early infancy that may also contribute to obesity in childhood. For example, being born small or large for gestational age [11], breast feeding less than three months [12], exposed to early introduction of complementary foods [13] and/or excess fruit juice consumption [14]. There may be many other potential factors that may contribute to childhood obesity but as yet undiscovered.
The relationship between maternal body composition and infant and maternal interaction and feedings have not been studied. Obese biological mothers may interact in such a way as to possibly contribute to greater energy intake in their infants earlier in life. Recently we published the results of 24-hour metabolic measurements in four-to-six month old infants in the Enhanced Metabolic Testing Activity Chamber (EMTAC) [15]. This study allowed collection of data to assess the relationship between maternal body composition and infant feeding and interaction patterns.
Methods
Subjects
The data for this analysis were derived from a previous study of 24-hour metabolic rate and physical activity in infants. Data from seven infants were complete in regards to metabolic rate, physical activity and body composition in both the infants and biological mothers. Moreover, complete data in regards to complementary food or formula intake were available. The physical characteristics of the biological mothers are shown in Table 1 and that of their infants in Table 2. Biological mothers were classified as obese (BMI >30 kg/m2) or of normal weight (BMI < 24.9 kg/m2) [16]. The mothers and their infants were recruited from the Outpatient Clinic of Miami Children's Hospital in Miami. A complete explanation regarding the purpose, procedure, risks and benefits of the study and informed consent was obtained from the biological mother of each infant at the time of the metabolic study. The study was approved by the Institute Review Board of Miami Children's Hospital.
Table 1 Physical characteristics of normal and obese biological mothers
Parameter for biological mothers Normal weight Obese
Number of subjects 3 4
Body weight (kg) 57.8 ± 2.0 85.6 ± 12.0*
Height (cm) 158 ± 3.8 152 ± 5.1
Age (y) 27.7 ± 10.6 28.8 ± 3.2
BMI (weight;kg/height2;m) 23.1 ± 0.6 33.5 ± 5.3*
Body fat (%) 35 ± 2 44 ± 5*
* p <0.05
Table 2 Physical characteristics of infants born to normal weight and obese biological mothers
Parameter for infants Biological Mothers
Normal weight Obese
Body weight (kg) 7.4 ± 0.6 7.4 ± 1.1
Length (cm) 69 ± 6 69 ± 0.9
Age (months) 4.7 ± 0.1 5.1 ± 1.0
BMI (weight;kg/height2;m) 15.6 ± 1.4 15.8 ± 1.9
Body fat (%) 26 ± 2 26 ± 3
Weight for length (percentile) 27 ± 23 31 ± 30
Weight for length Z-score -0.1 ± 0.9 0.07 ± 1.20
Weight for age (percentile) 72 ± 20 73 ± 17
Weight for age Z-score -0.02 ± 1.20 0.02 ± 1.00
Length for age (percentile) 73 ± 23 89 ± 9
Length for age Z-score -0.4 ± 0.6 0.39 ± 0.28
Maternal Anthropometry
Maternal height was measured using an Ayrton Model S100 hospital grade stadiometer (QuickMedical Inc., Snoqualmie WA) and weight was determined on a digital balance (LMI Inc., Concord, CA). Thereafter, body composition was measured by air displacement plethysmography using the BodPod Body Composition System (LMI Inc., Concord CA). The principle of the method is similar to hydrostatic weighting except that body volume is obtained by air displacement instead of water. The subjects sat for two 50 second testing sessions in a 450-liter chamber. A moving diaphragm determined the difference in air pressure between where the subject sat in the front chamber and a rear reference chamber. The pressure difference, along with the subject's body weight, was used to calculate body volume. From these results body fat was calculated using the Siri equation [17].
Infant Anthropometry
On the day of the study supine length (crown to heel) was measured in duplicate with a horizontal stadiometer (Perspective Enterprises, Kalamazoo, MI) and body weight was the average of two measurements obtained with an infant scale (Cardinal Detecto, Webb City, MO). Skin-fold thicknesses were the mean of two measures at each of five sites (biceps, triceps, sub scapular, flank and quadriceps) on the right side of the body using a Lange skin-fold caliper (Beta Technology, Cambridge, MD) according to a standard procedure [18]. Body fat and fat-free mass were calculated by appropriate equations [19].
Maternal interaction and infant feeding patterns
Energy and macronutrient intakes were determined from the formula and infant food manufacturer's proximate analysis for the nutrient components [20]. The amount of each food consumed by the infant during the 24-hour period was recorded. The actual amount of formula consumed by the infant was determined using calibrated infant feeding bottles and the amount of complementary feeding was also determined. The recording of energy and macronutrient intake started one-hour prior to and ended one-hour before the conclusion of the 24-hour metabolic testing period. The one-hour off-set was necessary to include or exclude energy consumed before and after the metabolic testing period which compensates for variations in feedings and the intestinal transient time in this age group of infants [15].
Only the biological mothers were present and cared for, fed and interacted with infants during the entire 24-hour testing period. During metabolic testing, biological mothers continued to feed their infants at their discretion in the same manner as before the study. They brought the milk formula and complementary foods they selected to the laboratory. The biological mothers were advised not to alter their infant feeding practices for the duration of the study. All infants born to normal weight biological mothers and one from an obese biological mother were only formula fed prior to and during metabolic measurements. Three out of the four infants born to obese biological mothers were receiving complementary foods beginning just after four months of age and were fed such during the study. The complementary foods included rice cereal, mixed vegetables, Beachnut® apple sauce and dessert fruit pudding. Six of the infants were fed Carnation Good Start® with iron while one was fed Carnation Alsoy®.
One member of the research team in charge of the 24-hour metabolic study was always directly involved with the testing procedure and was in close proximity to the infant and available to parents to address questions and concerns during the 24-hour period. The investigators acted as observers and recorded infant feedings, amount and type of formula fed along with consumption of complementary foods. Upon completion of the 24-hour metabolic measurement, macronutrient intake was determined from the amount of formula or complementary food fed utilizing the manufacturer's proximate analysis. This methodology of determining nutrient intake is valid and has been utilized in several previous studies in infants [21,22] and adults [20,23]. Investigators also recorded the interaction time and type of interaction and the observed periods of infant sleep. Furthermore, any other contact with the infant during the entire 24-hour testing period was also recorded.
The mean amount of time infants spent feeding was calculated by taking the sum of the time of all feeding periods and dividing by the number of feedings over the 24-hour testing period. The mean time between feedings during the day was calculated by taking the sum of the time intervals between each feeding session from 9:30 AM -11:30 PM and dividing by the number of feedings over the same period. The day and night periods were chosen to conform to the standards of previous 24-hour metabolic studies [24,25]. The mean time between feedings for the entire testing period was calculated in a similar manner but included all feedings over the course of the 24-hour metabolic measurement.
Interaction time was calculated by taking the sum of all data summary periods were the biological mother was interacting with her infant. Each data summary period was a five minute mean of continuous measurements of energy expenditure and physical activity index. The 24-hour testing period consisted of 288 five minute data summary periods. Interaction recorded included feeding, holding and cuddling the infant as well as diaper changes. In a similar manner the total amount of interaction time by biological mothers one hour prior to feeding was calculated by taking the sum of all summary periods one hour prior to each feeding episode. Finally, the total time infants spent engaged in feeding was calculated by taking the sum of all summary periods were the infant was being fed [15]. Feeding included all formula and complementary foods.
Measurements of metabolic rate and physical activity
Biological mothers were given instruction on how to interact with their infants while in the EMTAC and was allowed time to practice using the hand access ports prior to metabolic testing (Figure 1). Once all of the instruction and instrument calibrations were completed each infant was placed in the EMTAC for 24-hours from 9:30 AM till 9:29 AM the following day for continuous measurements of energy expenditure (EE; kcal/min) and physical activity (PA; oscillations in weight/min/kg body weight) as described previously [15]. Any supplies such as diapers, formula, complementary infant foods or toys were placed in the EMTAC in hanging bags before the start of the test. The mother of the infant being studied was provided lodging within the laboratory during the entire testing period. There were no restrictions in regards to room lighting, feeding or interaction of the infant or with any of the activities of the family during the entire testing procedure.
Figure 1 Photos showing the interaction between the biological mother and her infant utilizing the hand access ports of the EMTAC.
Energy expenditure (kcal/min) was continuously calculated during metabolic testing according to the method of Jequier [25] and summarized every five minutes as described previously [26]. At the conclusion of each metabolic test, all metabolic data were corrected for parental interaction, prior to the calculation of resting (RMR; kcal/kg/day) and sleeping metabolic rates (SMR; kcal/kg/d) as described previously [15]. All metabolic results were expressed as kcal/kg body weight/day.
The small number of infants in the study was due to the difficulties in recruiting biological mothers willing to stay in or around the laboratory for almost 30 hours. This time was necessary to instruct the biological mothers on how to interact and feed their infants using the hand access ports of the EMTAC and allow practice sessions prior to the start of the 24-hour metabolic measurement. Furthermore, body size and composition measurements had to be obtained for both the infants and biological mothers prior to the metabolic test. Moreover, the EMTAC had to be prepared and calibrated prior to the start of the 24-hour metabolic measurement.
Statistical Analysis
Person correlations were used to determine the relationship between the mother's anthropometry and interaction time and energy intake. Independent t-tests were used to determine differences in all metabolic and feeding parameters studied between obese and non-obese biological mothers. Significance (p < 0.05) was determined at the five percent level of probability.
Results
Height and age were similar among the normal weight and obese biological mothers. The latter had a greater body weight, BMI and percent body fat (Table 1) in comparison to their normal weight counterparts (p < 0.05). Furthermore no differences existed in regards to body weight, body composition and growth performance between infants of normal and obese biological mothers (Table 2). The similarity of the infants in terms of growth performance at the time of the study was further verified by the lack of significant differences in Z-scores obtained for weight for length, weight for age and length for age percentiles. None of the growth assessment parameters of the infants were more than 0.4 standard deviations from the mean.
Infants born to obese biological mothers consumed more energy, and energy as carbohydrate, than their normal weight counterparts (Table 3). Three, out of the four infants born to obese biological mothers consumed complementary foods. The amount of energy consumed from complementary foods by these infants of obese biological mothers was 18.3 ± 2.5 kcal/kg. This was in addition to the energy intake of 69.1 ± 20.3 kcal/kg from formula for these same infants. However, energy intake from protein and fat, for both complementary feedings and formula, were similar among the two groups (Table 3). The amount of formula intake was also similar (90.1 ± 16.3 vs. 98.9 ± 35.4 ml/kg) between the infants born to obese and the normal weight biological mothers. There was a significant (p < 0.05) correlation between total energy intake and maternal body weight (r = -0.73; p < 0.06; Figure 2a).
Table 3 Nutrient intake and feeding profile for the infants born to normal weight and obese biological mothers
Parameters for infants Biological Mothers
Normal weight Obese
Energy intake from formula (kcal/kg) 68.1 ± 17.2 69.1 ± 20.3
Energy intake from complementary foods (kcal/kg) 0 18.3 ± 2.5
Total 24-h energy intake (kcal/kg body weight)1 68.1 ± 17.2 87.6 ± 18.9
Formula intake (ml/kg body weight) 90.1 ± 16.3 98.9 ± 35.3
Energy intake from CHO (kcal/kg body weight)2 16 ± 3 25 ± 6*
Energy intake from PRO (kcal/kg body weight)3 4 ± 1 6 ± 3
Energy intake from FAT (kcal/kg body weight)4 38 ± 10 38 ± 12
Energy intake per feed (kcal/kg body weight) 7.4 ± 1.2 12.9 ± 4.8
Number of feedings (#) 9.3 ± 2.1 7.3 ± 2.2
Duration of each feeding (min) 33 ± 6 26 ± 5
Average time between feedings over 24-hours (min) 127 ± 28 188 ± 39
Average time between feedings over the day (min) 88 ± 32 108 ± 43
Total 24-hr total feeding time (min) 298 ± 32 176 ± 22*
Total 24-hr interaction time (min) 570 ± 13 381 ± 58*
Total interaction time one hour prior to feedings (min) 157 ± 29 68 ± 34*
Number of infants who had night time feedings 3 3
Total sleeping time (min) 682 ± 32 783 ± 38*
* = p < 0.05
1 24-hr EI = Twenty four hour energy intake expressed per kg body weight
2 CHO = Carbohydrate expressed per kg body weight
3 PRO = Protein expressed per kg body weight
4 FAT = Fat expressed per kg body weight
Figure 2 Correlation between 24-hour energy intake (kcal/kg/d) with both maternal body weight (kg; Figure 1a top plot) and body fat (%; Figure 2b bottom plot) for the seven infants in this study.
Obese biological mothers spent less time interacting and feeding their infants over the course of the 24-hour testing period (Table 3). There was a negative correlation between total 24-hour interaction time and both maternal body weight (r = 0.98; p < 0.01, Figure 3a) and body fat (r = 0.92; p < 0.01, Figure 3b). Moreover, overweight biological mothers spent less time interacting with their infants one hour prior to feeding (Table 3). The pattern of maternal interaction over the course of the 24-h testing period is shown in Figure 4. Normal weight biological mothers interacted with their infants more over the course of the 24-hour testing period than the obese biological mothers (Figure 4). When considering just the hour prior to each feeding, normal weight biological mothers interacted with their infants more than their obese counterparts (Table 3). The increased interaction was more significant (p < 0.05) during the day time hours (9:30 AM till 11:30 PM) while during the evening hours (11:31 PM till 5:30 AM) the difference in interaction time between normal and obese biological mothers was less significant (p < 0.10; Figure 4). Finally, infants of overweight biological mothers spent more time sleeping (Table 3) than their normal weight counterparts.
Figure 3 Correlation between interaction time (minutes) with both maternal body weight (kg; Figure 1a top plot) and body fat (%; Figure 1b bottom plot) for the seven infants in this study.
Figure 4 Twenty-four hour interaction profile for infants born to normal (top) and obese (bottom) biological mothers. The Y-axis represents the mean amount of interaction time over each five minute period. There are 288 five minute summary periods during a 24-hour metabolic measurement. A Obese biological mothers spent less time (p < 0.05) interacting with their infants during the day (9:30 AM till 11:30 PM) than their normal weight counterparts. Similar differences occurred during the evening hours (11:31 PM till 5:30 AM) but the differences were not significant.
The number of feedings and the amount of time necessary for each meal tended to be less for the infants of obese biological mothers, however, these differences were not significant (Table 3). There was also a tendency for infants of obese biological mothers to consume more energy at each feeding in comparison to those of normal weight mothers. Furthermore, infants of obese biological mothers tended to allow more time in between feedings during the day and over the entire 24-hour testing period (Table 3).
There were no significant differences in any of the metabolic parameters between infants from obese and normal weight biological mothers (Table 4). However, there was a tendency for a higher respiratory quotient and a greater amount of physical activity in those infants born to obese biological mothers.
Table 4 Twenty-four hour metabolic rate and physical activity of infants born to normal weight and obese biological mothers
Parameters for infants Biological Mothers
Normal weight Obese
24-hr EE (kcal/kg body weight)1 76.0 ± 5.2 74.3 ± 4.3
RMR (kcal/kg body weight)2 65.0 ± 5.0 64.9 ± 3.4
SMR (kcal/kg body weight)3 63.0 ± 4.3 64.6 ± 4.1
24-hr PA (oscillations/min/kg body weight)4 2.9 ± 0.4 5.0 ± 2.4
RQ (VCO2/VO2)5 0.84 ± 0.03 0.89 ± 0.02
1 24-hr EE = Twenty four hour energy expenditure expressed per kg body weight
2 RMR = 24-hour resting metabolic rate expressed per kg body weight
3 SMR = 24-hour sleeping metabolic rate expressed per kg body weight
4 PA = physical activity as determined by the amount of oscillations in weight (g) derived from the balance within the EMTAC enclosure. Data includes periods of sleep
5 RQ = Respiratory quotient determined as the rate of carbon dioxide production/oxygen consumption in liters
Discussion
In this study we demonstrated through direct observation that infants born to obese biological mothers ingested more energy and a greater amount of that energy were derived from the carbohydrates present in complementary foods. Furthermore, increased maternal body weight and fatness were related to increased 24-hour energy intake. Moreover, increased body weight and fatness of the biological mothers resulted in a decline in the amount of time spent interacting with their infants during the 24-hour testing period. This included less time spent feeding the infant, and particularly a lower amount of interaction one hour prior to feedings.
The biological mothers were able to freely interact with their infants in a comfortable setting. However, none of the biological mothers reported being uncomfortable with their infant being in the EMTAC or staying at the laboratory during the metabolic testing. Furthermore, they did not report any difficulties in interacting or feeding their infant utilizing the hand access ports of the EMTAC. Nonetheless, it is possible that the unfamiliar surroundings of the metabolic laboratory might have contributed to some changes in infant feeding practices not present in their maternal home setups.
Excess energy consumption early in an infant's life of those born to obese mothers, possibly accelerated with complementary food intake, might set the stage for future childhood obesity. In this study, there were no significant differences in infant body weight or composition between four to six months of age among the two groups of infants studied. However, infants of obese biological mothers consumed an average of 19.7 kcal/kg body weight more than the infants born to normal weight mothers during this study. Assuming that approximately 4900 kilocalories are needed per kilogram of body weight gain [27], it would take the infants of obese biological mothers a considerable amount of time to become obese if they would continue ingesting this amount of excess calories each day. This might explain why investigators report that it takes up to two years before a noticeable gain of body fat is observed in young children [1,28]. In overfed adults 66% of body weight gain is fat while the remainder is fat-free mass [29]. It is possible that overfeeding infants over a long period of time causes excess body weight gain of similar composition. However, no studies to date have quantified the composition of body weight gain in overfed infants from the time of birth.
Early introduction of complementary foods might increase body weight gain. In one study, early introduction of complementary foods was associated with greater infant body weight gain [30]. Other investigators [31] reported that complementary foods introduced to infants between 9 and 16 weeks showed a slight increase in weight gain velocity (g/week) in comparison to those infants who were introduced to these foods after 25 weeks. Both of these studies [30,31] suggest that early introduction of complementary foods might increase body weight gain. Therefore, early introduction of complementary foods, coupled with the increased energy intake at each feeding in the infants from obese biological mothers, as seen in our study, might set the stage for future childhood obesity. However, we did not have data to ascertain why obese mothers started complementary feedings in their infants.
Infants may be introduced to complementary foods at different times. In the Feeding Infants and Toddlers Study (FITS), 71% of the parents reported introducing complementary foods to their infants between four and six months of age while the other 29% reported introducing these foods to their infants at less than four months [32]. In another study were the National Health and Nutrition Examination Survey (NHANES III) data were analyzed, less than 25% of the parents reported feeding complementary foods to their infants prior to four months of age [33]. Moreover, the parents in both of these studies [32,33] were similar in regards to social economic status, age and ethnic background. In our study we did not have data on why obese biological mothers were feeding complementary foods to their infants. Moreover, neither the FITS [32] nor the NHANES III [33] studies related infant feeding practices to maternal body composition.
There were no significant differences in any of the metabolic parameters measured such as 24-hour energy expenditure, resting and sleeping metabolic rates, respiratory quotient and the index of physical activity. Moreover, there were no differences in any of the growth parameters (weight for length, weight for age and length for age percentiles) between the two groups of infants at the time of the study. Our results are in agreement with Stunkard et al [28] who found no differences in growth parameters, body composition, total energy expenditure by doubly labeled water, sleeping metabolic rate or physical activity in infants born to obese (greater than the 66th percentile for BMI) or lean (<33rd percentile for BMI) biological parents throughout the first year of life. In contrast to our results and those of Stunkard [28], Roberts et al [8] found reduced total daily energy expenditure in infants born to obese parents as determined by the doubly-labeled water method. The reduction of total daily energy expenditure was due to less physical activity in these infants [8]. It is possible that less interaction between obese mothers and their infants [8] accounted for the reduction in physical activity.
Since infants gain approximately 5 g/kg of body weight per day at four months of age [34], it is possible that constant additional caloric intake of those infants born to obese biological mothers will be manifested as additional daily body weight gain later in life. All these data suggest that maternal influences on infant body composition may not appear initially as obvious physical differences during the first six months of life. It is possible that the differences detected among biological obese mothers and their infants could affect the body composition of their infant as they age.
There may be other factors beginning in infancy that may be associated with the eventual increase in adiposity in later life. For example, fewer, but larger feeds and a higher sucking pressure were associated with greater adiposity in toddlers at two years of age [13]. This is in partial agreement with our results were we found that obese biological mothers spent less time interacting and feeding their infants. Moreover, infants from obese biological mothers consumed more energy in less time at each feeding. It is possible that the infants were hungrier due to the longer time between feedings.
Another study reported that greater maternal BMI during the first trimester of pregnancy was related to a higher prevalence of obesity in children from two to four years old. This is equivalent to 1 out of 4 children of obese mothers becoming obese as opposed to only 1 out of 10 children from normal weight mothers [35]. Moreover, it was also reported that a greater maternal BMI was a modest predictor of their daughter's relative weight at five years of age [36]. All of these studies [13,28,36] relate possible maternal influences upon future obesity of their infants. However, none eluted to the actual difference in the care of infants from either normal or obese mothers such as found in our analysis.
The association between the physical characteristics of biological mothers and their infants has not been ascertained. Six studies found no relationship between maternal BMI and infant's body weight after the first year of life [35,37-41] while two found such a relationship [42,43]. None of these studies accurately measured maternal body composition and did not include direct observation of the interaction dynamics between mothers and their infants. This was done in our study using air displacement plethysmography for maternal body composition and direct observation of food intake and feeding patterns including the number and length of infant feedings and the amount consumed at each meal during the entire 24-hour period. Additionally the type and length of maternal interaction with their infants revealed significant differences between normal weight and obese biological mothers which may not have been elucidated by other means. Though there were a small number of infants studied the results suggest that differences do exist on how mothers interact with their infants, depending on their body composition.
Conclusion
This study provided some new insight as to possible influences, beginning in infancy, as to the possible causes of childhood obesity. We have found additional factors that may contribute to future childhood obesity. The ability to conduct 24-hour metabolic rate measurements and direct accurate recordings of the biological mother's interaction with their infants elucidated specific differences in the care of infants that were related to maternal body weight and adiposity.
Abbreviations
EMTAC = Enhanced metabolic testing activity chamber
ANOVA = Analysis of Variance
BMI = Body mass index
EE = Energy expenditure
EI = Energy intake
RMR = Resting metabolic rate
SMR = Sleeping metabolic rate
PA = Physical activity index
RQ = Respiratory quotient
CHO = Carbohydrates
PRO = Protein
FITS = Feeding Infants and Toddlers Study
NHANES = Nutrition Health and Nutrition Examination Survey
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Dr. Russell Rising has contributed to the design of the experiment and conducted the data analysis. Furthermore, he either participated in some of the actual data acquisition or supervised pediatric research fellows in this regard. He also assisted in the preparation of the small grants necessary for funding of this project. Finally, he also assisted in the writing and editing of this manuscript.
Dr. Fima Lifshitz directed the research and contributed to the preparation of the manuscript and assisted with data analysis. He also generated some of the grant proposals necessary for the financial support of this study. Both authors were involved in the final writing of this manuscript.
Acknowledgements
This work was supported in part by NIH grant (#1R43HD/DK38180-01A2) and by Pediatric Sunshine Academics.
==== Refs
Dietz WH Overweight in childhood and adolescence N Engl J Med 2004 350 9 14702422 10.1056/NEJMp048008
Bhargava SK Sachdev HS Fall CHD Osmond C Lakshmy R Barker DJP Biswas SKD Ramji S Prabhakaran D Reddy KS Relation of serial changes in childhood body-mass index to impaired glucose tolerance in young adulthood N Engl J Med 2004 350 865 875 14985484 10.1056/NEJMoa035698
Wang G Dietz WH Economic burden of obesity in youths aged 6 to 17 years: 1979–1999 Pediatrics 2002 109 E81 100 11986487 10.1542/peds.109.5.e81
Gortmaker SL Must A Perrin JM Sobol AM Dietz WH Social and economic consequences of overweight in adolescence and young adulthood N Eng J Med 1993 329 1008 1012 10.1056/NEJM199309303291406
Poulsen P Vaag A The impact of genes and pre-and postnatal environment on the metabolic syndrome. Evidence from twin studies Panminerva Med 2003 45 109 115 12855935
Hanisch D Dittmar M Hohler T Alt KW Contribution of genetic and environmental factors to variation in body compartments-a twin study in adults Anthropol Anz 2004 62 51 60 15109035
Speakman JR Obesity: the integrated roles of environment and genetics J Nutr 2004 134 2090S 2105S 15284410
Roberts SB Savage J Coward WA Chew B Lucas A Energy expenditure and intake in infants born to lean and overweight mothers N Engl J Med 1988 318 461 466 3340127
Parson TJ Power C Logan S Summerbell CD Childhood predictors of adult obesity: a systematic review Int J Obes Relat Metab Disord 1999 23 S1 S107
Drewnowski A Specter SE Poverty and obesity: the role of energy density and energy costs Am J Clin Nutr 2004 79 6 16 14684391
Hediger ML Overpeck MD McGlynn A Kuezmarski RJ Maurer KR Davis WW Growth and fatness at three to six years of age of children born small-or- large-for-gestational age Pediatrics 1999 104 E33 10469816 10.1542/peds.104.3.e33
Gillman MW Rifas-Shiman SL Camargo CA jrBerkey CS Frazier AL Rockett HR Field AE Colditz GA Risk of overweight among adolescents who were breastfed as infants JAMA 2001 16 2461 2467 11368698 10.1001/jama.285.19.2461
Agras WS Kraemer HC Berkowitz RI Hammer LD Influence of early feeding style on adiposity at 6 years of age J Pediatr 1990 116 805 809 2329430
Dennison BA Fruit juice consumption by infants and children: a review J Am Coll Nutr 1996 15 4S 11S 8892177
Rising R Duro D Cedillo M Valois S Lifshitz F Daily metabolic rate in healthy infants J Peds 2003 143 180 185 10.1067/S0022-3476(03)00362-7
Calle EE Thun MJ Petrelli JM Rodriguez C Heath CW jr BMI and mortality in prospective cohort of U.S. adults N Eng J Med 1999 341 1097 1105 10.1056/NEJM199910073411501
Dempster P Aitkens S A new air displacement method for the determination of human body composition Med Sci Sports Exerc 1995 27 1692 1967 8614327
Heyward VH Stolarczyk LM Vivian H Skin-fold method Applied Body Composition Assessment 1996 Champaign, III: Human Kinetics 21 43
De Bruin NC Van Velthoven KA Stijnen T Juttmann RE Degenhart HJ Visser HK Body fat and fat-free mass in infants: new and classic anthropometric indexes and prediction equations compared with total-body electrical conductivity Am J Clin Nutr 1995 61 1196 1205
Rising R Alger S Boyce V Seagle H Ferraro R Fontvieille AM Ravussin E Food intake measured by an automated food selection system: relationship with energy expenditure Am J Clin Nutr 1992 55 343 349 1734670
Duro D Rising R Cedillo M Lifshitz F Association between colic history in the first three months of life and carbohydrate malabsorption from fruit juices in infancy Ped 2002 109 797 805 10.1542/peds.109.5.797
Cole C Rising R Lifshitz F Are there consequences of incomplete carbohydrate absorption from fruit juice consumption in infants ? Arch Pediatr Adolesc Med 1999 153 1098 1102 10520620
Rising R Larson DE Ravussin E Do obese eat faster than lean subjects: Food intake studies in Pima Indian men Obes Res 1994 2 19 23 16353604
Ravussin E Lillioja S Anderson TE Christin L Bogardus C Determinants of 24-hour Energy expenditure in man: Methods and results using a respiratory chamber J Clin Invest 1986 78 1568 1578 3782471
Jequier E Long-term measurement of energy expenditure in man: direct and indirect calorimetry Recent advances in obesity research Libbey, London 1981 3 130 135
Cole C Rising R Mehta R Hakim A Choudhury S Sundaresh M Lifshitz F Comprehensive assessment of the components of energy expenditure in infants using a new infant respiratory chamber J Am Coll Nutr 1999 18 233 41 10376779
Reichman BL Chessex P Putet G Verellen GJE Smith JM Heim T Swyer PR Partition of energy metabolism and energy cost of growth in the very low-birth-weight infant Ped 1982 69 446 451
Stunkard AJ Berkowitz RI Stallings VA Schoeller DA Energy intake, not energy output, is a determinant of body size in infants Am J Clin Nutr 1999 69 524 530 10075340
Tremblay A Despres JP Theriault G Fournier G Bouchard C Overfeeding and energy expenditure in humans Am J Clin Nutr 1992 56 857 862 1415004
Baker JL Michaelsen KF Rasmussen KM Sorensen TI Maternal prepregnant body mass index, duration of breastfeeding, and timing of complementary food introduction are associated with infant weight gain Am J Clin Nutr 2004 80 1579 1588 15585772
WHO Working group on the growth reference protocol and the WHO ask force on methods for the national regulation of fertility Growth of healthy infants and the timing, type and frequency of complementary foods Am J Clin Nutr 2002 76 620 627 12198009
Briefel RR Reidy K Karwe V Devaney B Feeding infants and toddlers study: Improvements needed in meeting infants feeding recommendations J Am Diet Assoc 2004 104 S31 37 14702015 10.1016/j.jada.2003.10.020
Hediger ML Overpeck MD Ruan WJ Troendle JF Early infant feeding and growth status of US-born infants and children aged 4–71 mo: analysis from the third National Health and Nutrition Examination Survey, 1988–1994 Am J Clin Nutr 2000 72 159 167 10871575
DeBruin NC Degenhart HJ Gal S Westerterp KR Stijnen T Visser HKA Energy utilization and growth in breast-fed and formula-fed infants measured prospectively during the first year of life Am J Clin Nutr 1998 67 885 896 9583846
Edwards LE Dickes WF Alton IR Hakanson EY Pregnancy in the massively obese: course, outcome and prognosis of the infant Am J Obstet Gynecol 1978 131 479 483 677188
Birch LL Fisher JO Mothers' child-feeding practices influence daughters' eating and weight Am J Clin Nutr 2000 71 1054 1061 10799366
Cardon L Turner L, Cardon L, Hewitt J Genetic influences on body mass index in early childhood Behavior genetic approaches in behavioral medicine 1995 New York: Plenum Press 133 143
Whitaker RC Wright JA Pepe MS Seidd RD Dietz WH Predicting obesity in young adulthood from childhood and parental obesity N Engl J Med 1997 337 869 873 9302300 10.1056/NEJM199709253371301
Morgan J Parental weight and its relation to infant feeding patterns and infant obesity Int J Nurs Stud 1986 23 255 264 3637163 10.1016/0020-7489(86)90023-4
Kramer MS Barr RG Leduc DG Boisjoly C McVey WL Pless JB Determents of body weight and adiposity in the first year of life J Pediatr 1985 106 10 14 3965671
Berkowitz RI Agras Korner AF Kraemer HC Zeanah CH Physical activity and adiposity: a longitudinal study from birth to childhood J Pediatr 1985 106 734 738 3998913
Rummler S Woit I Zur postnatalen Entwicklung von Kindern adiposer Mutter. (On the postnatal development of children of obese mothers) Z Geburtshilfe Perinnatal 1990 194 254 259
Hashimoto N Kawasaki T Kikuchi T Takahashi H Uchiyama M Influence of parental obesity on the physical constitution of preschool children in Japan Acta Paediatr Jap 1995 37 150 153 7793246
| 15890076 | PMC1156942 | CC BY | 2021-01-04 16:39:31 | no | Nutr J. 2005 May 12; 4:17 | utf-8 | Nutr J | 2,005 | 10.1186/1475-2891-4-17 | oa_comm |
==== Front
Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-191592705310.1186/1475-2891-4-19ReviewAfrican herbal medicines in the treatment of HIV: Hypoxis and Sutherlandia. An overview of evidence and pharmacology Mills Edward [email protected] Curtis [email protected] Dugald [email protected] Izzy [email protected] Department of Clinical Epidemiology & Biostatistics, McMaster University, 1200 Main Street West Hamilton, L8N 3Z5, Canada2 Division of Infectious Diseases, University of Ottawa, 501 Smyth Rd., Ottawa, K1H 8L6, Canada3 Department of Clinical Epidemiology, Canadian College of Naturopathic Medicine, 1255 Sheppard Ave. East, North York, M2K1M2, Canada4 Faculty of Pharmacy, Rhodes University, Grahamstown, South Africa2005 31 5 2005 4 19 19 18 2 2005 31 5 2005 Copyright © 2005 Mills et al; licensee BioMed Central Ltd.2005Mills et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
In Africa, herbal medicines are often used as primary treatment for HIV/AIDS and for HIV-related problems. In general, traditional medicines are not well researched, and are poorly regulated. We review the evidence and safety concerns related to the use of two specific African herbals, which are currently recommended by the Ministry of Health in South Africa and member states for use in HIV: African Potato and Sutherlandia. We review the pharmacology, toxicology and pharmacokinetics of these herbal medicines. Despite the popularity of their use and the support of Ministries of Health and NGOs in some African countries, no clinical trials of efficacy exist, and low-level evidence of harm identifies the potential for drug interactions with antiretroviral drugs. Efforts should be made by mainstream health professionals to provide validated information to traditional healers and patients on the judicious use of herbal remedies. This may reduce harm through failed expectations, pharmacologic adverse events including possible drug/herb interactions and unnecessary added therapeutic costs. Efforts should also be directed at evaluating the possible benefits of natural products in HIV/AIDS treatment.
==== Body
The use of traditional medicine and Natural Health Products is widespread among those living with HIV infection [1]. Many patients take a broad range of natural health products (NHPs) in addition to their conventional therapeutic products [2-4]. In Africa, traditional herbal medicines are often used as primary treatment for HIV/AIDS and for HIV-related problems including dermatological disorders, nausea, depression, insomnia, and weakness[2,5-8]. Some herbal and traditional medicines are not well-researched, poorly regulated, may contain adulterated products, and may produce adverse effects [8-13]. Notwithstanding these concerns, the use of traditional medicines by Africans living with HIV is believed to be widespread, although insufficiently documented [14-16].
Despite a paucity of evidence on effectiveness, and the possibility of harm, the Ministries of Health of several African nations currently promote traditional medicines for the treatment of HIV and associated symptoms [12,17]. In the case of South Africa, the Ministry of Health is actively promoting the use of traditional medicines with antiretroviral treatments[18].
Two principal African herbal compounds used for HIV/AIDS treatment in sub-Saharan Africa include Hypoxis hemerocallidea (common name: African potato), and Sutherlandia. These two herbal remedies are currently recommended by the South African Ministry of Health for HIV management [17]. The 14 member states of the South African Development Community (SADC) which includes Angola, Botswana, Democratic Republic of Congo, Lesotho, Malawi, Mauritius, Mozambique, Namibia, Seychelles, South Africa, Swaziland, Tanzania, Zambia, and Zimbabwe, also support their use [19]. Responding to the compelling need for evidence regarding traditional medicines, we reviewed the current evidence for the use of these herbal remedies in HIV care.
Methods
With the aid of an information specialist, we searched the following databases independently, in duplicate (from inception to December 2004):AltHealthWatch, AMED, CancerLit, CinAhl, Cochrane Controlled Trials Register (CENTRAL), MedLine, and EMBASE. In order to identify unpublished research, we searched Clinical trials.gov, National Research Register (UK) and the Meta-Register. Searches were not limited by language. We additionally searched bibliographies of identified reviews and contacted experts in the field. The following search terms were used, but not limited to: "Medicine, African Traditional", "Hypox*", "Sutherlandia", and "HIV."
Hypoxis hemerocallidea
Common names
Magic muthi, yellow stars, star lily, African potato (Eng.); sterretjie, Afrika-patat (Afr.); inkomfe, ilabatheka, sterblom, gifbol, lotsane, molikharatsa[20,21]
Hypoxis is a well-known genus of the family Hypoxidaceae. Easily recognizable by its bright yellow star-shaped flowers and strap-like leaves, it has a long history of medicinal use on the African continent. The South African primary health care community is currently using hypoxis as an immunostimulant for patients with HIV/AIDS. A daily dose of 2,400 mg of raw plant is purported to be therapeutically effective [22]. Within the genus, two species, H. hemerocallidea and H. colchicifolia are particularly popular both as African traditional remedies and for the preparation of herbal teas and tinctures.
Rootstocks of this plant have been used by Zulu traditional healers for centuries in the treatment of urinary infections, heart weakness, internal tumors, and nervous disorders [21]. Other unproven uses for this herb include benign prostatic hypertrophy, cancer and hyperglycemia [23-25]. The corms of H. hemerocallidea are being used for immune related illnesses such as the common cold, flu, arthritis, cancer and HIV/AIDS. There is some indirect evidence that sterols and sterolins, which are found in the root of Hypoxis, have the potential to enhance immunity [26-28]. The popular press in South Africa is promoting preparations of Hypoxis as an agent that can boost immunity in HIV/AIDS patients [29,30]. Multiple websites, popular magazines, and even the South African Ministry of Health have supported this assertion [29-32]. Irrespective of the evidence, many Africans claim benefit from eating the root of H. hemerocallidea [7,14].
Chemical constituents
An important constituent of the plant is a nor-lignan glycoside called hypoxoside, which once in the human gut, readily converts to the aglycone, rooperol, a biologically active compound that is purported to have medicinal properties [22,33]. The plant also contains various sterols (β-sitosterol, stigmasterol) and their glycosides (sterolins) such as β-sitosterol glycoside and stanols such as sitostanol also called stigmastanol, which have also been purported to have important biological activity [26,28].
Pharmacology and Pharmacokinetics
Hypoxoside
Hypoxoside is not absorbed intact into the blood stream. Once in the body, hypoxoside is converted into its aglycone, rooperol, a potent antioxidant [34]. This conversion is mediated by beta-glucosidase, an enzyme found predominantly in the gastrointestinal tract. This particular enzyme is released by rapidly dividing cancer cells.
Phase I biotransformation of both Hypoxoside and rooperol likely occurs via the P450 system and most likely by CYP 3A4 [35]. A multi-dosage trial found only diglucuronide, disulphate, and mixed glucuronide-sulphate metabolites of these two principal constituents in the serum of recipients. Elimination of the metabolites follows first order kinetics with half lives ranging from 20 hours for the two minor metabolites (i.e., diglucuronide and disulphate), to 50 hours for the major metabolite (i.e., the mixed glucuronide-sulphate)[22].
Our group recently reported Hypoxis' effect on the P-450 system CYP 3A4 enzyme, the drug transporter P-glycoprotein (P-gp), and the pregnane X receptor (PXR) [35]. Hypoxis inhibited up to 86% of the normal CYP 3A4 isoform activity. P-glycoprotein showed moderate activity from exposure to Hypoxis, showing 42–51% of the activity strength relative to the known P-gp inhibitor verapamil. Exposure to Hypoxis resulted in an almost 2-fold activation of the PXR (p < 0.05). This activation was dose-dependent. Whilst the concentrations used in the in vitro experiments were relatively high, the study nevertheless demonstrated that Hypoxis possesses the potential to interact with HIV drug metabolizing enzymes, which could subsequently lead to drug resistance, drug toxicity and/or treatment failure. It should be noted, however that this evidence is only from one in vitro model and may not translate to the same effect in vivo.
Toxicity
A Phase I trial in cancer patients failed to establish any clinical, hematological, or biochemical toxicities that could be ascribed to the ingestion of hypoxoside [24]. One recipient did experience an episode of anxiety, nausea, vomiting and diarrhea which was possibly hypoxoside related. The data and safety monitoring committee recently terminated a clinical trial of therapeutic effectiveness in HIV/AIDS patients citing apparent bone marrow suppression [36]. Supporters of this herbal medicine have disputed these inferences [37].
Hypoxoside, when infused in anaesthetized chacma baboons, had no effect on the cardiovascular system, whereas rooperol exerted moderate stimulation during drug administration. The cardiac output increased together with systemic and pulmonary arterial pressures and these changes were not accompanied by changes in heart rate, vascular resistances or in the filling pressures of the heart. These findings are suggestive of increased myocardial contractility, possibly related to rooperol's catechol structure. It is likely that these cardio stimulatory effects will prove to be clinically benign [22]. The molecular basis of rooperol toxicity still needs to be clarified. Biochemical studies have shown that rooperol is a potent inhibitor of leukotriene synthesis in polymorphonuclear leucocytes at a concentration of 1 μM or less [22].
Sutherlandia Frutescens subspecies Microphylla
Common names
Insiswa, Unwele, Mukakana, Phetola, Lerumo-lamadi, cancer bush, kankerbos, kankerbossie [38,39]
Background
The flowering shrub S. frutescens is a member of the Fabacea family. The recommended therapeutic dose of Sutherlandia in humans is 9 mg/kg/day[40]. Sutherlandia has been used in the treatment of cancer, tuberculosis, diabetes, chronic fatigue syndrome, influenza, rheumatoid arthritis, osteoarthritis, peptic ulcers, gastritis, reflux esophagitis, menopausal symptoms, anxiety, clinical depression and HIV infection [38,39]. The South African Ministry of Health has concluded that this product is safe based on primate safety studies.
However, scientific data relating to the mechanism whereby Sutherlandia acts on the immune system has not been comprehensively documented. Fernandes et al [41] recently described the antioxidant potential of Sutherlandia frutescens where extracts from hot water possessed superoxide as well as hydrogen peroxide scavenging activities which could account for anti-inflammatory properties. In a study by Tai et al, [42] ethanolic extracts were shown to have an anti-proliferative effect on several human tumor cell lines but did not show significant antioxidant activity.
Phyto Nova, of South Africa, is the principal distributor of both the powdered and encapsulated forms of this herb, and has attempted to evaluate the purported benefits of this remedy in HIV/AIDS treatment[38]. A definitive conclusion has not yet been reached. Despite the paucity of data, the South African Ministry of Health and member states currently recommend the use of this herbal remedy for HIV/AIDS treatment[17,40].
Constituents
The principal constituents of S. frutescens purported to be active include L-canavanine, GABA, and D-pinitol. L-canavanine is a non-protein amino acid that is the L-2-amino-4-guanidinooxy structural analogue of L-arginine. There is about 30–40 mg of L-canavanine per dry gram of the S. frutescens leaf[38]. D-pinitol is a type of sugar found in many types of legumes and is classified as a chiro-inositol. It is also known as 3-O-methyl-D-chiro-inositol, or 3-0-methyl-1,2,4 cis-3,5,6 trans-hexahydroxy-cyclohexanol. GABA (gabba-amino butyric acid) is both an amino acid and inhibitory neurotransmitter. It is found at levels of 14 mg per gram dry leaf of S. frutescens[38].
One of the chemical constituents of Sutherlandia, L-canavanine, is an arginine analogue. L-canavanine has been reported to have anti-viral activity against influenza and retroviruses, including HIV [43]. A US patent registered in 1988 claimed that 95% of HIV-infected lymphocytes were selectively destroyed in vitro. Unfortunately, no further studies of the effect of this herb on HIV have confirmed this claim. D-pinitol another important constituent of Sutherlandia has also been suggested for the treatment of wasting in cancer and AIDS patients although evidence is scant[44].
Pharmacokinetics and pharmacology
The pharmacokinetic properties of Sutherlandia have largely not been assessed[40]. We have demonstrated in vitro effects of Sutherlandia on CYP3A4, P-gp, and PXR [35]. Sutherlandia produced near complete inhibition of CYP3A4 (96%). P-gp activity was moderate under exposure of Sutherlandia, showing 19–31% of the activity strength relative to verapamil. A PXR assay demonstrated a greater than 2-fold activation with exposure to Sutherlandia which was dose-dependent (P < 0.01). Once again, in spite of the relatively high concentrations used in the in vitro experiments, these results tentatively suggest that human consumption of Sutherlandia could affect antiretroviral drug metabolism leading to bi-directional drug interactions and loss of therapeutic efficacy. In vivo human studies are required to determine if there is a clinically relevant drug/herb interaction and if so what the true extent of the interaction is.
Toxicity
Sutherlandia has a relatively long history of seemingly safe usage in Africa. Known side effects include occasional mild diarrhea, dry mouth, mild diuresis, and dizzyness in cachectic patients [38,39]. An extensive toxicology screening in a primate model using dosages up to 9 times greater than the recommended dose of 9 mg/kg/day did not identify clinical, hematological or physiologic toxicity with Sutherlandia [45].
L-canavanine may be associated with important toxicities including a systemic lupus erythematous syndrome [46]. The non-protein amino acid can be incorporated into protein in place of arginine and may, after long term usage, result in autoimmunity [47,48]. Rare reports of teratogenicity and induction of abortion exist [49].
Discussion
The widespread use of herbal compounds by Africans living with HIV/AIDS should be of concern to clinicians and policy makers. Clearly, patients will continue to access traditional healing systems as it is important to local cultural values and beliefs. Therefore, efforts should be made by mainstream health professionals to provide validated information to traditional healers and patients on the judicious use of herbal remedies. This may reduce harm through failed expectations, pharmacologic adverse events and unnecessary added therapeutic costs. Efforts should also be directed at evaluating the possible benefits of natural products in HIV treatment.
It is not unreasonable to suggest that some products may have therapeutic benefits as examples from history and the recent past have provided us with effective anti-malarials [50] and cancer treatments[51]. Indeed, some of the earliest forms of protease inhibitors were derived from natural products [52-54]. Efforts should be directed at determining the therapeutic efficacy of these remedies as well as the possibility of interactions through systematic research and clinical trials.
Several studies have highlighted key problems related to primary care delivery by traditional healers in Africa [8,55,56]. Key issues include hygiene, toxicity and financial cost. Traditional healers have been implicated in the spread of blood borne diseases including HIV and other infectious disease by the re-use of medical instruments and lack of hand washing [8,55,56]. Prescriptions to take toxic plants for HIV treatment have also resulted in severe adverse events, including death[56]. Recent policy efforts have recognized the substantial use of traditional medicines and several African nations have included traditional healers in educational campaigns in order to instruct them on safe and hygienic practices, condom distribution and knowledge dissemination[2,5,6,14,57].
Despite the relatively high concentrations of herbals used in our in vitro work, the results serve as a warning and suggest that biologically active constituents of these herbal remedies clearly may have an effect on HIV drug metabolism as a result of their inhibitory activity on enzymes and efflux drug transporter systems. These results highlight the need for in vivo investigations and circumspection when utilizing herbal drugs as routine care for HIV patients and underscores the need for clinical studies in humans to unveil any possible drug interaction of these herbal agents with antiretrovirals. Failure to do this may result in bi-directional drug interactions that may put patients at risk for treatment failure, viral resistance or drug toxicity.
Cultural values are an inherent part of healthcare and an important component of practicing evidence-based healthcare [58]. In the context of HIV treatment in Africa, patients often choose traditional healing systems as primary care. This, coupled with the difficulties in accessing antiretroviral treatment, justify further efforts to determine the scope of traditional medicine use, identify the negative consequences of this practice and evaluate the benefits of herbal remedies. In addition, it is important to understand the values of those providing mainstream healthcare and those practicing traditional medicine as their perspectives provide highly relevant social inferences and should be interpreted with an attempt to understand their cultural worldviews and practices.
In conclusion, given the Global Fund's recent announcement of funds to make anti-retroviral therapy widely available in Africa, and the South African Ministry of Health, along with member states and NGO's endorsement of the use of traditional African herbs such as Hypoxis and Sutherlandia as HIV/AIDS remedies [17], initiating policy on herbal medicines should be based on research evidence. Efforts are required to determine the safety, efficacy and pharmacological profile of the many herbal compounds used in Africa. Collaboration with traditional healers is justified to fully understand what remedies are in use for HIV and to educate those providing alternative medical services against unsafe practices.
==== Refs
Fairfield KM Eisenberg DM Davis RB Libman H Phillips RS Patterns of use, expenditures, and perceived efficacy of complementary and alternative therapies in HIV-infected patients Arch Intern Med 1998 158 2257 2264 9818806 10.1001/archinte.158.20.2257
Babb DA Pemba L Seatlanyane P Charalambous S Curchyard G Grant AD Use of traditional medicine in the era of antiretroviral therapy: experience from South Africa eJIAS 2004, MedGenMed 2004 6 B10640 [eJIAS. 2004 Jul 11;1(1):B10640].
Colebunders R Dreezen C Florence E Pelgrom Y Schrooten W The use of complementary and alternative medicine by persons with HIV infection in Europe Int J STD AIDS 2003 14 672 674 14596770 10.1258/095646203322387929
Duggan J Peterson WS Schutz M Khuder S Charkraborty J Use of complementary and alternative therapies in HIV-infected patients AIDS Patient Care STDS 2001 15 159 167 11313029 10.1089/108729101750123661
Fleming J Mozambican healers join government in fight against AIDS J Int Assoc Physicians AIDS Care 1995 1 32 11362312
Homsy J King R Balaba D Kabatesi D Traditional health practitioners are key to scaling up comprehensive care for HIV/AIDS in sub-Saharan Africa Aids 2004 18 1723 1725 15280784 10.1097/01.aids.0000131380.30479.16
McMillen H The adapting healer: pioneering through shifting epidemiological and sociocultural landscapes Soc Sci Med 2004 59 889 902 15186892 10.1016/j.socscimed.2003.12.008
Peters EJ Immananagha KK Essien OE Ekott JU Traditional healers' practices and the spread of HIV/AIDS in south eastern Nigeria Trop Doct 2004 34 79 82 15117130
Ernst E Heavy metals in traditional Indian remedies Eur J Clin Pharmacol 2002 57 891 896 11936709 10.1007/s00228-001-0400-y
Ernst E Toxic heavy metals and undeclared drugs in Asian herbal medicines Trends Pharmacol Sci 2002 23 136 139 11879681 10.1016/S0165-6147(00)01972-6
Ernst E Thompson Coon J Heavy metals in traditional Chinese medicines: a systematic review Clin Pharmacol Ther 2001 70 497 504 11753265
Morris K South Africa tests traditional medicines Lancet Infect Dis 2002 2 319 12144884 10.1016/S1473-3099(02)00302-X
Wittenstein B Rogers V Novelli V Clayton PT Kenny MB Peters MJ "African medicine" and Reye's syndrome Lancet 2004 363 860 15031031 10.1016/S0140-6736(04)15733-4
MacPhail CL Campbell CM Pitts M Operationalizing the dual use of traditional and Western health care for STIs and HIV XIV International AIDS Conference: 2002; Barcelona: MoPeB3209 2002
Sebit MB Chandiwana SK Latif AS Gomo E Acuda SW Makoni F Vushe J Quality of life evaluation in patients with HIV-I infection: the impact of traditional medicine in Zimbabwe Cent Afr J Med 2000 46 208 213 11317592
Zachariah R Nkhoma W Harries AD Arendt V Chantulo A Spielmann MP Mbereko MP Buhendwa L Health seeking and sexual behaviour in patients with sexually transmitted infections: the importance of traditional healers in Thyolo, Malawi Sex Transm Infect 2002 78 127 129 12081174 10.1136/sti.78.2.127
SADC SADC Ministerial Consultative meeting on Nutrition and HIV/AIDS Johannesburg 20 January 2002
ARV roll-out begins today Sunday Times (South Africa) March 1 2003
Giraldo R Brief report SOUTHERN AFRICAN DEVELOPMENT COMMUNITY (SADC) MEETING ON NUTRITION AND HIV/AIDS: 2002; Johannesburg 2002
Snijman D Hypoxidacea: cape plants – a conspectus of the cape flora of South Africa Strelitzia 2000 9 108 110
Singh Y Hypoxis: yellow stars of horticulture, folk remedies and conventional medicine Veld & Flora 1999 85 123 125
Albrecht CF Kruger PB Smit BJ Freestone M Gouws L Miller R van Jaarsveld PP The pharmacokinetic behaviour of hypoxoside taken orally by patients with lung cancer in a phase I trial S Afr Med J 1995 85 861 865 8545744
Wilt TJ Ishani A Rutks I MacDonald R Phytotherapy for benign prostatic hyperplasia Public Health Nutr 2000 3 459 472 11276294
Smit BJ Albrecht CF Liebenberg RW Kruger PB Freestone M Gouws L Theron E Bouic PJ Etsebeth S van Jaarsveld PP A phase I trial of hypoxoside as an oral prodrug for cancer therapy – absence of toxicity S Afr Med J 1995 85 865 870 8545745
Mahomed IM Ojewole JA Hypoglycemic effect of Hypoxis hemerocallidea corm (African potato) aqueous extract in rats Methods Find Exp Clin Pharmacol 2003 25 617 623 14671679 10.1358/mf.2003.25.8.778082
Bouic PJ Clark A Brittle W Lamprecht JH Freestone M Liebenberg RW Plant sterol/sterolin supplement use in a cohort of South African HIV-infected patients – effects on immunological and virological surrogate markers S Afr Med J 2001 91 848 850 11732455
Bouic PJ Clark A Lamprecht J Freestone M Pool EJ Liebenberg RW Kotze D van Jaarsveld PP The effects of B-sitosterol (BSS) and B-sitosterol glucoside (BSSG) mixture on selected immune parameters of marathon runners: inhibition of post marathon immune suppression and inflammation Int J Sports Med 1999 20 258 262 10376483
Bouic PJ Etsebeth S Liebenberg RW Albrecht CF Pegel K Van Jaarsveld PP beta-Sitosterol and beta-sitosterol glucoside stimulate human peripheral blood lymphocyte proliferation: implications for their use as an immunomodulatory vitamin combination Int J Immunopharmacol 1996 18 693 700 9172012 10.1016/S0192-0561(97)85551-8
Whitney JR Billoski TV Jones VR Billoski TV Evidence for Triceratops in Antarctica New Directions in Paleontology 1997 New York: Academic Press 24 27
"I read it in the paper" – Spud magic? THE SA RETAIL CHEMIST 1997, 1 AUGUST
Sengupta SSR DeVillis R Crouse Quinn S DeVillis B Ware WB Factors Affecting African-American Participation in AIDS Research JAIDS 2000 24 275 284 10969353
Aids conference starts
Albrecht CF Theron EJ Kruger PB Morphological characterisation of the cell-growth inhibitory activity of rooperol and pharmacokinetic aspects of hypoxoside as an oral prodrug for cancer therapy S Afr Med J 1995 85 853 860 8545743
Dietzsch E Albrecht CF Parker MI Effect of rooperol on collagen synthesis and cell growth IUBMB Life 1999 48 321 325 10690646
Mills EJ Foster BC van Heeswijk RP Wilson K E P Kosuge K Leonard B Kanfer I Impact of African herbal medicines on antiretroviral metabolism AIDS 2005 19 95 7 15627040 10.1097/00002030-200501030-00013
Terreblanche C HIV warning on African potato IOL 2003
Clarke L African potato gets bum rap as vital remedy Sunday Independent 2003
Gericke N Albrecht CF Van Wyk B Mayeng B Mutwa C Hutchings A Sutherlandia frutescens Australian Journal of Medical Herbalism 2001 13 9 15 10.1046/j.1442-2026.2001.00172.x
Dalvi S Sutherlandia for immune support Positive Health 2003 23 25
Seier JV Mdhuli M Dhansay MA Loza J Laubschner R Matsabisa G A toxicity study of Sutherlandia leaf powder (sutherlandia microphylla) consumption South African Ministry of Health Document 2002
Fernandes AC Cromarty AD Albrecht C van Rensburg CE The antioxidant potential of Sutherlandia frutescens J Ethnopharmacol 2004 95 1 5 15374599 10.1016/j.jep.2004.05.024
Tai J Cheung S Chan E Hasman D In vitro culture studies of Sutherlandia frutescens on human tumor cell lines J Ethnopharmacol 2004 93 9 19 15182898 10.1016/j.jep.2004.02.028
Green M Method of treating viral infections with amino acid analogue US Patent 5110,600 Jan 25 1988.
Oslund R Sherman W Pinotol and derivatives thereof for the treatment of metabolic disorders. US Patent. 5,882,896 March 4, 1996.
A toxicity study of Sutherlandia leaf powder (sutherlandia microphylla) consumption
Capasso R Izzo AA Pinto L Bifulco T Vitobello C Mascolo N Phytotherapy and quality of herbal medicines Fitoterapia 2000 71 S58 65 10930714 10.1016/S0367-326X(00)00173-8
Prete PE Effects of L-canavanine on immune function in normal and autoimmune mice: disordered B-cell function by a dietary amino acid in the immunoregulation of autoimmune disease Can J Physiol Pharmacol 1985 63 843 854 3876146
Prete PE The mechanism of action of L-canavanine in inducing autoimmune phenomena Arthritis Rheum 1985 28 1198 1200 4052135
Van Wyk B Gericke N People's plants – a guide to the useful plants in South Africa Pretoria: Briza 2000
Adjuik M Babiker A Garner P Olliaro P Taylor W White N Artesunate combinations for treatment of malaria: meta-analysis Lancet 2004 363 9 17 14723987 10.1016/S0140-6736(03)15162-8
Slichenmyer WJ Von Hoff DD Taxol: a new and effective anti-cancer drug Anticancer Drugs 1991 2 519 530 1687206
Egberg DC Potter RH Honold GR The semiautomated determination of trypsin inhibitors in textured soy protein J Agric Food Chem 1975 23 603 605 1171133 10.1021/jf60199a042
Egeblad K Effects of soy bean trypsin inhibitor on fibrin clot lysis Thromb Diath Haemorrh 1966 15 542 553 6011258
Rabek V Mansfield V Gel filtration of protease inhibitors from potatoes Experientia 1963 19 151 152 13972839
Miles SH Ololo H Traditional surgeons in sub-Saharan Africa: images from south Sudan Int J STD AIDS 2003 14 505 508 12935374 10.1258/095646203767869057
Munk K Traditional healers, traditional hospitals and HIV / AIDS: a case study in KwaZulu-Natal AIDS Anal Afr 1997 7 10 12 12348320
Sidley P South Africa to regulate healers BMJ 2004 329 758 15459034 10.1136/bmj.329.7469.758-b
Haynes RB Devereaux PJ Guyatt GH Physicians' and patients' choices in evidence based practice BMJ 2002 324 1350 12052789 10.1136/bmj.324.7350.1350
| 15927053 | PMC1156943 | CC BY | 2021-01-04 16:39:31 | no | Nutr J. 2005 May 31; 4:19 | utf-8 | Nutr J | 2,005 | 10.1186/1475-2891-4-19 | oa_comm |
==== Front
Popul Health MetrPopulation Health Metrics1478-7954BioMed Central London 1478-7954-3-41587174010.1186/1478-7954-3-4ResearchPediatric appendicitis rupture rate: a national indicator of disparities in healthcare access Jablonski Kathleen A [email protected] Mark F [email protected] The Biostatistics Center, The George Washington University, 6110 Executive Boulevard, Suite 750, Rockville, Maryland 20852, USA2 Department of Prevention and Community Health, The George Washington University School of Public Health and Health Services, Washington, DC, USA3 Center for Health Services and Community Research, Children's National Medical Center, 111 Michigan Avenue, NW, Washington, DC 20010, USA2005 4 5 2005 3 4 4 21 2 2005 4 5 2005 Copyright © 2005 Jablonski and Guagliardo; licensee BioMed Central Ltd.2005Jablonski and Guagliardo; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The U.S. National Healthcare Disparities Report is a recent effort to measure and monitor racial and ethnic disparities in health and healthcare. The Report is a work in progress and includes few indicators specific to children. An indicator worthy of consideration is racial/ethnic differences in the rate of bad outcomes for pediatric acute appendicitis. Bad outcomes for this condition are indicative of poor access to healthcare, which is amenable to social and healthcare policy changes.
Methods
We analyzed the KID Inpatient Database, a nationally representative sample of pediatric hospitalization, to compare rates of appendicitis rupture between white, African American, Hispanic and Asian children. We ran weighted logistic regression models to obtain national estimates of relative odds of rupture rate for the four groups, adjusted for developmental, biological, socioeconomic, health services and hospital factors that might influence disease outcome.
Results
Rupture was a much more burdensome outcome than timely surgery and rupture avoidance. Rupture cases had 97% higher hospital charges and 175% longer hospital stays than non-rupture cases on average. These burdens disproportionately affected minority children, who had 24% – 38% higher odds of appendicitis rupture than white children, adjusting for age and gender. These differences were reduced, but remained significant after adjusting for other factors.
Conclusion
The racial/ethnic disparities in pediatric appendicitis outcome are large and are preventable with timely diagnosis and surgery for all children. Furthermore, estimating this disparity using the KID survey is a relatively straightforward process. Therefore pediatric appendicitis rupture rate is a good candidate for inclusion in the National Healthcare Disparities Report. As with most other health and healthcare disparities, efforts to reduce disparities in income, wealth and access to care will most likely improve the odds of favorable outcome for this condition as well.
==== Body
Background
The persistence of racial and ethnic disparities in health and healthcare is a major theme in American healthcare policy. The Institute of Medicine report, "Unequal Treatment: Confronting Racial and Ethnic Disparities in Health Care" [1], describes and underscores the problem, and Healthy People 2010 [2] has made disparity elimination one of its two overarching goals. In response to these concerns Congress directed the Agency for Healthcare Research and Quality (AHRQ) to prepare a National Healthcare Disparities Report [3,4]. to monitor progress toward alleviating disparities. Over 250 measures of healthcare quality and access were considered for the report. However, while many measures are relevant to children, only 16 are specific to children. Of these, only two concerned pediatric inpatient services. Therefore the National Report could be improved by the inclusions of additional markers of pediatric health and healthcare disparities.
Appendicitis outcome is a good candidate measure because it is the most common intraabdominal surgical procedure performed on children [5], and may be the most common of any surgery among adolescents [6]. Furthermore, it has no known links to behavioral or social risk factors, and has only one treatment option – appendectomy. Timely surgery within a few hours or days of symptom onset is necessary to prevent rupture and other complications, which are costly [7] can result in loss of internal organs, female infertility and even death [8].
Therefore it is disconcerting that racial and ethnic disparities in the rate of appendicitis rupture (AR) have recently been reported among children of California and New York [9]. Compared to white children, odds of rupture was elevated by as much as 47% for African American children, 45% for Hispanic children, and 116% for Asian American children, after adjustments for the biologic risk factors, age and gender. Disparities were ameliorated but still apparent after adjustments for income, insurance status, and hospital characteristics. Subsequently, Ponsky et al. [10] analyzed a large sample of pediatric appendectomy cases from free-standing U.S. children's hospitals. Although their data lacked an indicator for Hispanic ethnicity, they found that odds of rupture was 16% higher for African Americans and 66% higher for Asians compared to whites.
While groundbreaking, neither previous study used a nationally representative sample, which limits their usefulness for National Healthcare Disparities Report. However, an underutilized federal survey, AHRQ's Healthcare Cost and Utilization Project (HCUP) Kids' Inpatient Database (KID), can be used for the purpose. KID is a stratified national sampling of pediatric discharges. Its latest version, from 2000, contains 2,516,833 discharges from 2,784 hospitals, weighted to represent 7,291,032 pediatric discharges nationwide. Therefore, KID is uniquely qualified to answer questions about pediatric inpatient services, including emergency surgical services, at the national level. In this paper we will report the latest estimates for frequency of pediatric acute appendicitis, and AR rates among all U.S. children, and determine if the previously reported disparities in AR rates are apparent at the national level. If KID bears evidence of disparities in AR rates, then this information should be included in the National Report.
Methods
Data source
This was a retrospective cohort study of the Kid's Inpatient Database (KID) 2000, a member of AHRQ's Healthcare Cost and Utilization Project (HCUP) family of databases [11]. The database was specifically designed to allow researchers to track and analyze national trends in pediatric inpatient utilization, outcomes and quality. KID 2000 contains 2,516,833 abstracts representing a finite population of 7,291,039 children derived from the pediatric discharges of 2,784 hospitals in 27 states during the calendar year 2000. The sampling frame includes all pediatric discharges from community, non-rehabilitation hospitals in the HCUP State Inpatient Databases (SID) that could be matched to American Hospital Association survey data. It is from the latter that KID derives hospital characteristics for each discharge. Case weights were based on stratification on six hospital characteristics. Admission age was <21 years for all cases. The sampling procedure selected 80 percent of non-birth pediatric discharges from each hospital in the sampling frame. The sampling design and weighting details are described elsewhere [11].
Cases
We followed the methods of Guagliardo et al. [9] for identifying non-incidental pediatric appendicitis cases and cases with rupture or complications. Cases were limited to children 4 to 18 years old because acute appendicitis is rarely diagnosed in a timely manner for very young children, regardless of race/ethnicity or access to care considerations [12]. Appendicitis cases were defined as any discharge with any ICD-9 CM diagnosis code in the 540.X range. Cases were excluded if an appendectomy was incidental (ICD-9 CM procedure code 47.1X) or there were comorbid conditions likely to hinder the timely diagnosis of appendicitis, such as injuries to the GI tract, GI neoplasms, or inflammatory bowel disease. (A complete list is available from the authors upon request.)
Within this sample, cases with rupture or other complications were identified as having any of the following ICD-9 CM diagnosis codes: 540.0 (acute appendicitis with generalized peritonitis), 540.1 (acute appendicitis with peritoneal abscess), 567.2 (other suppurative peritonitis), 569.5 (abscess of intestine), 614.3 (acute parametritis and pelvic cellulitis), and 614.4 (chronic or unspecified parametritis and pelvic cellulitis). Guagliardo et al. [9] also classified appendicitis cases with codes 682.2 (other cellulitis and abscess-trunk) and 998.59 (other postoperative infection) as having complications. However we agree with the reviewers and editors of this journal that these conditions could arise post-operatively in both complicated and non-complicated appendicitis cases. Therefore we removed from all analyses the 73 cases with either 682.2 or 998.59 that did not otherwise qualify for inclusion. To remain consistent with previous reports we refer to all cases with rupture or other complications as "rupture cases", although technically all were not ruptured.
Finally, we excluded cases with race/ethnicity designations that were missing, Native American, or "Other", as there were too few of these cases for meaningful analysis. This left us with a final sample of 32,784 cases, representing a weighted count of 62,555 patients.
Variables
Main predictor
Race is coded in the KID 2000 database as white, black, Hispanic, Asian/Pacific Islander, Native American or other. However, the 27 participating states report race and ethnicity in many different ways. The final KID variable is actually a combined indicator of race and ethnicity, with Hispanic ethnicity taking precedence in final assignment. For example, if a state reports a child's race as white and ethnicity as Hispanic, KID 2000 "race" was coded as Hispanic. All but two KID 2000 states, Iowa and North Carolina, reported Hispanic ethnicity as a separate variable or as a category in a combined race/ethnicity variable.
Covariates
Age is an important risk factor for pediatric appendicitis rupture [9,13,14], and hence a necessary covariate, because it is more difficult to quickly diagnose appendicitis in younger children. The literature is inconsistent regarding male gender as a risk factor for rupture. Gadomski et al. [13] found boys to be at elevated risk for a Maryland sample, while Guagliardo et al. [9] and Ponsky et al. [10] did not detect gender differences in larger samples. We include gender as a covariate as a precaution and to further explore it as a risk factor.
ZIP code median annual income was used as a proxy for family income. While less relevant than direct measures of family income or wealth, ZIP code median annual income has been shown to be a useful proxy in other hospital utilization studies [15,16] Guagliardo et al. [9] found lower ZIP code income to be a risk factor for AR, independent of insurance type and other covariates. In KID 2000 this variable is coded in the four categories used for the HCUP National Inpatient Sample [17]: $0–$24,999, $25,000–$34,999, $35,000–$44,000; and ≥ $45,000.
Insurance type (i.e. expected primary payer) is a well-established indicator of barriers to healthcare for most medical conditions, including acute appendicitis and timely appendectomy [9,10,14,15]. Typically, privately insured patients receive the most timely and highest quality care, followed by publicly insured patients, with the uninsured having the most difficulty. We created an expected payer variable that mimics these three categories. KID 2000 codes its uniform primary payer variable as Medicare, Medicaid, private insurance, self-pay, no charge, and other. We grouped Medicare and Medicaid into a combined publicly insured category. We carefully explored the state-specific codes underlying the KID 2000 "other payer" category to see if there were cases in this residual category that could be reasonably assigned to public, private or uninsured categories. This was possible for a worthwhile number of cases. For example, in California several "other government" and indigent care programs designed for the needy were coded as "other payer" in KID 2000. We grouped those with our publicly insured cases. (The authors may be contacted for a complete list of recodes.) Some of the other remaining KID 2000 "other payer" cases were excluded because they could not be confidently assigned to one of our three target categories. However, we created our own "other" category to approximate an uninsured group. It consisted of self-pay, no-charge, charity, and otherwise uninsured cases. We recognize that a very small percentage of these cases could be from uninsured but non-needy families. Therefore, we labeled the group as "other" rather than uninsured.
Admission from an emergency department (ED) appears to be a protective factor against risk of rupture [9,14,15,18], possibly due to quicker diagnosis than for cases who delayed seeking care or were first seen by a primary care provider. We created an admission source variable with three categories – admitted from ED, referred to the hospital for inpatient treatment by any other healthcare provider (e.g. other hospital, within-hospital clinic, stand-alone clinic or HMO), and "other source". The latter includes self-referrals and other unspecified sources.
There is continued interest in the effect of hospital teaching status on outcomes [19]. Teaching hospitals are noted for better outcomes, including appendicitis outcomes [15], but lower patient satisfaction.([20]. Therefore we included hospital teaching status as a covariate.
There is also interest in the effect of hospital patient volume on appendicitis outcome. The idea is that greater experience diagnosing and treating a given condition will improve an institution's performance [21-23]. The number of pediatric discharges from KID 2000 hospitals ranged from 4–22,785. We divided the hospitals into volume quartiles, assigned the quartile value of the discharging hospital to each patient, and used the volume quartile assignment as a covariate in the logistic regressions.
Analysis
As noted in Guagliardo, et al. [9], most previous studies of AR [13-15,18] found little evidence of race/ethnicity disparities because their regression models overcontrolled for the factors that mediate disparities. Overcontrolling is a common methodological hazard [24]. In order to reveal disparities at different levels of control for mediating factors we developed four logistic regression models. The first model used race/ethnicity alone as a predictor of AR odds. The second model included the biological and developmental covariates, gender and age. This model is of interest because it adjusts for factors that are not amenable to health policy or social policy changes. The disparities revealed by this model should be targeted with policy changes. The third model adds the social and system factors, median ZIP code income, insurance type and admission source. The final and fullest model adds the hospital-level factors, pediatric discharge volume and hospital teaching status. While overcontrolled, we report the fullest model because it is more comparable to previously published studies, and because it bears interesting revelations about the covariates.
All analyses were performed with SAS 9.1.3 [25]. We tested for mulitcollinearity among our variables using the collinearity index (PROC REG with the COLLINOINT option) [26]. Frequencies and means were computed with the SAS SURVEYFREQ and SURVEYMEANS procedures using the appropriate survey weighting variables. Regressions were modeled with PROC SUVEYLOGISTIC, a procedure for weighted logistic regression, taking into account the sampling design and sample discharge weights using the methods outlined in Houchens and Elixhauser [27]. We specified the hospital cluster and survey stratum in the CLUSTER and STRATUM statements. These methods assured unbiased variance estimates.
Results
There were 40,762 acute appendicitis cases in KID 2000. Our exclusion of cases with problematic race/ethnicity diagnosis codes limited our analyses to 32,784 cases. Weighted, these cases represented approximately 62,555 children hospitalized for acute appendicitis in 2000. All table values are weighted to represent national estimates.
Table 1 compares the number and proportion of appendiceal rupture (AR) and non-rupture outcomes for all study variables. All variables except gender showed statistically significant variations in rates of AR among groups (P < 0.05). Higher AR rates were noted for children who were minorities, younger, from poorer ZIP codes, lacking private insurance, referred from somewhere other than the ED, discharged from a teaching hospital, and discharged from a high-volume hospital.
Table 1 National estimates of ruptured and unruptured appendicitis cases for children 4–18 years old. Obtained from the KID Year 2000 data set. (Weighted count of cases = 62,555).
Ruptures (%) Non-ruptures (%)
Race/ethnicity* white 12,056 (29%) 29,514 (71%)
black 1,474 (36%) 2,579 (63%)
Hispanic 5,539 (36%) 1,001 (64%)
Asian 499 (36%) 893 (64%)
Gender male 11,947 (31%) 26,247 (69%)
female 7,613 (31%) 16,726 (69%)
Age group* 4–8 5,387 (42%) 7,591 (58%)
9–11 4,189 (31%) 9,221 (69%)
12–14 5,075 (31%) 11,444 (69%)
15–18 4,918 (25%) 14,731 (75%)
ZIP code median annual household income* $0–$24,999 2,059 (38%) 3,399 (62%)
$25,000–$34,999 5,612 (33%) 11,573 (67%)
$35,000–$44,999 5,284 (30%) 12,097 (70%)
≥ $45,000 6,309 (29%) 15,198 (71%)
Expected primary payer* private 12,226 (29%) 29,866 (71%)
public 5,552 (37%) 9,351 (63%)
other 1,554 (33%) 3,179 (67%)
Admission source* emergency department 13,068 (30%) 30,273 (70%)
other healthcare referral 4,461 (35%) 8,281 (65%)
other 1,097 (32%) 2,193 (68%)
Hospital type* teaching 9,552 (35%) 18,114 (65%)
other 10,016 (29%) 24,836 (71%)
Pediatric discharge volume* 4–1,578 4,759 (28%) 12,474 (72%)
1,579–3,471 4,343 (28%) 11,213 (72%)
3,472–6,582 4,320 (32%) 9,568 (68%)
6,583–22,785 6,045 (38%) 9,733 (62%)
Means (95% CIs)
Length of stay* 5.5 (5.4–5.6) 2.0 (1.9–2.0)
Total charges* $17,905 ($17,473–$18,336) $9,076 ($8,956–$9,198)
*Chi-square p < 0.01.
Utilization measures were much higher for AR cases. Mean length of stay was 5.5 days for AR cases, or 175% higher than the 2.0 day mean for non-ruptured cases. Mean total charges were 97% higher for AR cases, at $17,905 versus $9,076.
Odds of AR are presented in Table 2. In the first model, of unadjusted odds by race/ethnicity categories, all the minority groups have 36%-40% higher odds of rupture compared to whites. These odds decrease but remain significant across the table for all three minority groups even as covariate adjustment factors are added. The second model includes the covariates, gender and age group. Gender is irrelevant to odds of rupture in this and all subsequent models. On the other hand age is meaningful in all models that include it. As expected, younger children are consistently at higher risk of rupture – between 27% and 105% higher depending on the age group and adjustment covariables used. The third model includes social and healthcare system factors. Private insurance and higher income are protective against AR in this as well as the final model. Children referred to the hospital from a non-emergency department healthcare setting had nearly 30% higher odds of AR than children admitted directly from the discharging hospital's ED. This admission source pattern holds for the final, fullest model, which includes hospital characteristics as covariates. In that model teaching and non-teaching hospital discharges are indistinguishable for odds of AR. Surprisingly, lower pediatric discharge volume appears to have been a protective factor against odds of AR. Compared to children treated at hospitals in the highest volume quartile, children discharged from hospitals in the three lower volume groups had 21%-28% better odds of avoiding AR.
Table 2 Odds ratios for appendiceal rupture among children in the United States.
Unadjusted race/ethnicity odds Race/ethnicity odds adjusted for biologic factors With additional adjustments for social and system factors With additional adjustments for hospital factors
Race/ethnicity white ref ref ref ref
black 1.40 (1.27–1.55) 1.38 (1.25–1.53) 1.27 (1.14–1.42) 1.23 (1.10–1.37)
Hispanic 1.36 (1.28–1.44) 1.24 (1.17–1.31) 1.14 (1.07–1.22) 1.07 (1.00–1.15)
Asian 1.37 (1.16–1.61) 1.32 (1.13–1.52) 1.29 (1.10–1.53) 1.24 (1.05–1.46)
Gender male 1.00 (0.95–1.06) 0.99 (0.94–1.05) 0.99 (0.94–1.05)
female ref ref ref
Age group 4–8 2.05 (1.90–2.21) 2.02 (1.87–2.18) 1.90 (1.76–2.06)
9–11 1.36 (1.25–1.47) 1.36 (1.25–1.47) 1.30 (1.20–1.41)
12–14 1.32 (1.23–1.42) 1.31 (1.22–1.42) 1.28 (1.18–1.38)
15–18 ref ref ref
ZIP code median annual household income $0–$24,999 1.15 (1.05–1.27) 1.13 (1.02–1.24)
$25,000–$34,999 1.06 (0.99–1.14) 1.11 (1.03–1.20)
$35,000–$44,999 1.00 (0.93–1.08) 1.04 (0.96–1.12)
≥ $45,000 ref ref
Expected primary payer private ref ref
public 1.10 (0.99–1.23) 1.09 (0.98–1.21)
other 0.87 (0.78–0.97) 0.87 (0.78–0.96)
Admission source emergency department ref ref
other healthcare referral 1.28 (1.19–1.38) 1.29 (1.20–1.39)
other 1.03 (0.93–1.14) 1.05 (0.95–1.16)
Hospital type teaching 1.06 (0.99–1.14)
other ref
Pediatric discharge volume 4–1,578 0.72 (0.65–0.80)
1,579–3,471 0.73 (0.66–0.80)
3,472–6,582 0.79 (0.73–0.86)
6,583–22,785 ref
Sample size unweighted 32,784 32,773 30,940 30,940
weighted 62,555 62,533 57,791 57,791
Discussion
Appendiceal rupture (AR) was much more burdensome than appendectomy without rupture. The mean total hospital charges were 97% higher for AR cases, while mean length of hospital stay was 175% longer. These additional burdens fell disproportionately on minority children. Results for the first regression model in Table 2 show unadjusted racial/ethnic disparities in acute appendicitis outcome. However, the second regression model is more relevant to the national goal of reducing disparities, because it includes adjustments for factors that are not amenable to changes in health policy, social policy or medical practice – patient age and gender. Disparities revealed in this model should be targeted for elimination. African American children have approximately 38% greater odds of appendicitis rupture (AR) compared to whites, and the relative odds are not much better for Hispanic or Asian children – 24% and 32% higher, respectively.
These national disparity estimates differ somewhat from findings reported for other samples. Among 1997 California pediatric discharges Guagliardo et al. [9] found that African American and white children had indistinguishable AR rates, while Hispanic and Asian children had higher odds of AR – 45% and 30%, respectively. They found a different pattern of disparities among 1995 New York discharges. There, white and Hispanic children were indistinguishable, while African American children had 47% higher odds and Asian children had 116% higher odds of AR. In a sample of discharges from children's hospitals Ponsky et al. [10] found that African American children had 13% higher odds and Asian children had 66% higher odds of AR compared to white children. However, they could not identify Hispanic children in their sample, and they used different adjustment variables.
In spite of the incomparability of the samples studied to date, a pattern is emerging. White children have never been found to be at greater odds of AR than any minority group, while minority children are usually found to have poorer appendicitis outcomes than white children. The key question remains. Why is there racial/ethnic disparity in acute appendicitis outcomes?
To explore this question it is helpful to view acute appendicitis as a "delay-sensitive" condition [28]. Once its clock starts then rupture, broader infection, bleeding and death are inevitable without surgery. Lacking evidence to the contrary, it is generally assumed that the disease progresses at the same average rate for all social groups. Therefore inter-group differences in average delay of key milestones in the disease course must account for the disparities. The milestones include first complaint of abdominal pain, parental recognition of urgency, initial seeking of professional care, performance of diagnostic procedures and/or referrals to other healthcare facilities, eventual correct diagnosis, and finally surgical intervention. Reductions in time between any of these milestones will reduce the chance of rupture. Research suggests that in the U.S. there is little or no delay between correct diagnosis and surgery [12]. Children get to the operating room quickly once the diagnosis is made. Therefore, the disparities we have discovered are probably due to longer average delays for minorities prior to correct diagnosis.
Factors that can delay care seeking and timely diagnosis include family health beliefs and economic condition, insurance coverage, physician quality and distance to healthcare provider. We expected that our fuller regression models would provide insight into the effects of some of these factors. Two of the covariates used are taken from major domains generally involved in the production of racial/ethnic disparities in health and healthcare – income [29] and insurance type [30]. As mediators of disparity such as these are added to regression models, apparent race/ethnicity differences should decrease and eventually disappear if all explanatory factors could be included. Yet our fullest model in Table 2 has not achieved this ideal. In it all minority groups still have higher odds of AR relative to white children. This could be because our covariables are imperfect representatives of their domains. For example, insurance type may be too coarse of a measure of insurance quality, e.g. privately insured white children were in better plans than privately insured minority children. It is also possible that administratively derived data sets such as KID 2000 do not contain proxies for important disparity-producing factors. Two prime examples are language and cultural differences between patients and their healthcare providers [31,32], and level of geographic availability of local care providers [33]. Including more and better covariables might have accounted for all race/ethnic disparities. Furthermore, the behavior of these covariables in regression models could point out socioeconomic, demographic and health services factors to be targeted to achieve better outcomes and less disparity. However, finding a fully explanatory model could not excuse the disparities revealed in our second model (Table 2). The disparities revealed in that model should remain as the national indicators of disparity and should be targeted for reduction [34].
The disparity differences found between California, New York and the current national sample suggest that local sociocultural factors are at play. Guagliardo et al. [9] suggested a link between odds of being foreign-born in the two states and odds of AR. They hypothesized that general degree of acculturation could be a major precipitator of disparities, acting through language barriers, preferences for traditional healing and concern among undocumented immigrants for engaging the healthcare system.
A definitive understanding of the causes of disparity in pediatric AR rates will require prospective, primary data collection, including family interviews and in depth case reviews with special attention to the timing of the aforementioned milestones to discover the social, economic, provider and health systems circumstances associated with delayed surgery. However, that the precise causes of disparities remain unclear should not detract from the major thrust of this report. The disparities are real, occur on a national scale, and are most likely due to socioeconomic, cultural and healthcare system factors.
While the covariates in the fuller models do not account for all of the disparities and are not the main focus of this study, their effects are nonetheless noteworthy. The negative effect of lower income is consistent with a large literature on income, wealth and health [35,36]. As found in previous studies [9,14,15,18], admission from the emergency department (ED) of the hospital that performed the appendectomy reduces the odds of AR. In contrast, patients who go to other healthcare settings before referral to the surgical facility tend to have poorer outcomes, ostensibly due to the additional delay. There are similar findings in the acute myocardial infarction and stroke literature [37]. Patients with chest pains or stroke symptoms should go immediately to the nearest ED. Yet it would be premature to recommend that all children with abdominal pain report immediately to the nearest ED. Abdominal pain has many causes [8], and such a recommendation would fly in the face of decades of efforts to reduce unnecessary ED utilization [38]. Careful and thorough research is required to developed optimal recommendations to reduce both AR rates and unnecessary ED utilization [38-40].
Teaching hospitals have a reputation for lower patient satisfaction but better medical outcomes than non-teaching hospitals [19,20,41,42]. Findings for AR have been mixed. Braveman et al. [15] reported better outcomes for adults discharged from teaching hospitals, while Guagliardo et al. [9] found poorer outcomes for children discharged from teaching hospitals. Here, adjusted odds of AR are estimated to be 6% higher for children discharged from teaching hospitals, although the difference is not statistically significant. Still, future AR studies should consider hospital teaching status as a covariable.
Of all the covariables, pediatric discharge volume yielded the most surprising results. Contrary to the "practice makes perfect" maxim [21], neither of the previous large-sample pediatric studies [9,10] found a relationship between volume and outcome. The current analysis actually showed the reverse effect. Hospitals in the highest volume group had significantly higher AR rates. This is difficult to explain. It is possible that high-volume hospitals have unfavorable staff-to-patient ratios, leading to additional delays in care. It might also be that these hospitals are located in underserved areas, and hence more of their patients must travel farther for service, which might delay diagnosis. We know of no published studies that test these hypotheses. It is interesting to note that Smink et al. [22] found significantly higher rates of negative appendectomy in low-volume hospitals. (A negative appendectomy is an unnecessary surgery resulting from an incorrect diagnosis.) Smink et al. analyzed the 1997 version of KID. Comparing their nationally representative results with ours, it appears that two kinds of error might be in effect. Low-volume facilities might have excessive rates of misdiagnosis and premature surgery, while high-volume facilities might have excessive rates of delayed diagnosis and delayed surgery. We are planning another study to explore these questions. Recommendations are not possible at this time.
Additional limitations
Nearly 20% of otherwise eligible KID cases were missing race/ethnicity or were in the American Indian or "other" categories. Thus they were not used in the analyses. Fortunately these cases did not differ from our sample in the rate of AR (30% versus 31%, p < 0.52), reducing the probability that our analyses are biased. However, the problem could become acute if the proportion of missing race/ethnicity cases increases in future releases of KID. This underscores the danger of initiatives to prevent collection of race/ethnicity data in healthcare administrative databases, such as California's Proposition 54 [43].
Hospital ownership (public versus private) is often considered as a covariable in administrative data studies. However, too many states in the KID 2000 system reported ownership with a combined "public or private" categorization, making the distinction impossible in the current analysis. We do not consider this a serious limitation, as a previous study showed ownership not to be a factor for risk of pediatric appendicitis rupture [9].
It is very common in the healthcare literature to use ZIP code median income as a proxy for individual income as we have done. However, there is always the risk of ecological fallacy – the bias that can result from using aggregate data in lieu of individual-level measurements [44]. Studies have shown that aggregate statistics from the census block group and census tract levels are useful proxies for individual-level measures [45]. However, ZIP code socioeconomic indicators are somewhat insensitive to geographic variation in health indicators [46]. While this might diminish the predictive power of our income indicator, we can think of no reason that it would be biased.
Conclusion
The first National Healthcare Disparities Report states that, "While consistency of measures from year to year is highly desirable, the measures selected for inclusion in the first NHDR represent a small subset of currently available measures and are expected to evolve as the field of health care measurement itself evolves." [4] This paper proposes a new measure for inclusion in future reports – rate of appendiceal rupture (AR). Minority children with acute appendicitis in the U.S. are 24%-38% more likely than white children to experience (AR) and its attendant complications and expenses. Because AR can be avoided with timely surgery, and because rate of progression of infection is not linked to cultural factors, these outcome disparities are probably due to differences in timely access to quality care. Some of the factors that play a role in AR may be amenable to healthcare and social policy changes. It would be worthwhile to attempt to reduce disparities associated with insurance differences, income differences, referral patterns and/or how families seek urgent care, and in-hospital practice differences. However, there are significant residual disparities not attributable to these factors. Further research is required to discover and address the causes.
Regardless of the causes, the disparities are real and significant. It should be a national goal to reduce or eliminate disparities in risk of AR. Analysis of KID data for pediatric acute appendicitis outcomes is relatively straightforward. Therefore, Congressional funding for the survey should continue, and KID-based disparities in AR rates should be included and tracked over time as an indicator of racial/ethnic disparities in the National Healthcare Disparities Report.
List of abbreviations
AHRQ – Agency for Healthcare Research and Quality
AR – appendicitis rupture, appendiceal rupture
HCUP – Healthcare Cost and Utilization Project
KID – Kids Inpatient Database
NHDR – National Healthcare Disparities Report
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Both authors contributed equally to writing the introduction, discussion and conclusions. KAJ obtained the data, performed all statistical analyses, and wrote the methods and results sections. MFG conceptualized the project and performed most of the literature search.
Acknowledgements
MFG's efforts were supported in part by Grant 1P20MD000165 from the NIH National Center for Minority Health and Health Disparities, "Washington-Baltimore Center To Improve Child Health Disparities" (Principal Investigator: Jill G. Joseph, MD, PhD). Our ICD-9 grouping scheme and case selection were taken from Guagliardo, et al. [9] We gratefully acknowledge the contribution of the physician co-authors on that paper, Stephen J. Teach and James M. Chamberlain, in developing the classification and selection schemes.
==== Refs
Smedley BD Stith AY Nelson AR Institute of Medicine (U.S.), Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care Unequal Treatment Confronting Racial and Ethnic Disparities in Health Care 2002 Washington, D.C: National Academy Press
U.S. Department of Health and Human Services Healthy People 2010 2000 Washington, DC: U.S. Dept. of Health and Human Services. For sale by the U.S. G.P.O., Supt. of Docs
Agency for Healthcare Research and Quality National Healthcare Disparities Report: Update on Current Status 2002 Rockville, MD, Agency for Healthcare Research and Quality
Agency for Healthcare Research and Quality National Healthcare Disparities Report – July 2003 Draft 2003 Rockville, MD, Agency for Healthcare Research and Quality
Pearl RH Hale DA Molloy M Schutt DC Jaques DP Pediatric appendectomy J Pediatr Surg 1995 30 173 8 discussion 178–81 7738734 10.1016/0022-3468(95)90556-1
To T Guttmann A Dick P Inpatient and Day Surgery Use by Children in Ontario 2001 Toronto, Institute for Clinical Evaluative Sciences
Pittman-Waller VA Myers JG Stewart RM Appendicitis: why so complicated? Analysis of 5755 consecutive appendectomies Am Surg 2000 66 548 54 10888130
Silen W Cope Z Cope's Early Diagnosis of the Acute Abdomen 1987 New York: Oxford University Press
Guagliardo MF Teach SJ Huang ZJ Chamberlain JM Joseph JG Racial and ethnic disparities in pediatric appendicitis rupture rate Acad Emerg Med 2003 10 1218 27 14597498 10.1197/S1069-6563(03)00492-5
Ponsky TA Huang ZJ Kittle K Hospital- and patient-level characteristics and the risk of appendiceal rupture and negative appendectomy in children JAMA 2004 292 1977 82 15507583 10.1001/jama.292.16.1977
Healthcare Cost and Utilization Project Design of the HCUP Kids' Inpatient Database (KID), 2000 2003 Rockville, MD, Agency for Healthcare Research and Quality
Nelson DS Bateman B Bolte RG Appendiceal perforation in children diagnosed in a pediatric emergency department Pediatr Emerg Care 2000 16 233 7 10966339 10.1097/00006565-200008000-00004
Gadomski A Jenkins P Ruptured appendicitis among children as an indicator of access to care Health Serv Res 2001 36 129 42 11324740
Bratton SL Haberkern CM Waldhausen JHT Acute appendicitis risks of complications: age and Medicaid insurance Pediatrics 2000 106 75 8 10878152 10.1542/peds.106.1.75
Braveman P Schaaf VM Egerter S Bennett T Schecter W Insurance-related differences in the risk of ruptured appendix N Engl J Med 1994 331 444 9 7880234 10.1056/NEJM199408183310706
Hofer TP Wolfe RA Tedeschi PJ McMahon LF Griffith JR Use of community versus individual socioeconomic data in predicting variation in hospital use Health Serv Res 1998 33 243 59 9618670
Agency for Healthcare Research and Quality Overview of the HCUP Nationwide Inpatient Sample (NIS) 2000 2002 Rockville, MD, Agency for Healthcare Research and Quality.
Buckley RG Distefan J Gubler KD Slymen D The risk of appendiceal rupture based on hospital admission source Acad Emerg Med 1999 6 596 601 10386676
Kassirer JP Hospitals, heal yourselves N Engl J Med 1999 340 309 10 9920957 10.1056/NEJM199901283400410
Massachusetts Health Quality Partnership Massachusetts Acute Care Hospital Statewide Patient Survey Project 1998 Boston, MA
Hughes RG Hunt SS Luft HS Effects of surgeon volume and hospital volume on quality of care in hospitals Med Care 1987 25 489 503 3695658
Smink DS Finkelstein JA Kleinman K Fishman SJ The effect of hospital volume of pediatric appendectomies on the misdiagnosis of appendicitis in children Pediatrics 2004 113 18 23 14702441 10.1542/peds.113.1.18
Bratton SL Haberkern CM Waldhausen JH Sawin RS Allison JW Intussusception: hospital size and risk of surgery Pediatrics 2001 107 299 303 11158462 10.1542/peds.107.2.299
Kleinbaum DG Kupper LL Morgenstern H Epidemiologic Research Principles and Quantitative Methods 1982 Belmont, Calif: Lifetime Learning Publications
SAS. 9.1.3. Cary, NC, SAS Institute 2004
Belsley DA Kuh E Welsch RE Regression Diagnostics: Identifying Influential Data and Sources of Collinearity 1980 New York: Wiley
Houchens R Elixhauser A Final Report on Calculating Nationwide Inpatient Sample (NIS) Variances, 2001 2003 Rockville, MD, U.S. Agency for Healthcare Research and Quality
Bickell NA Siu AL Why do delays in treatment occur? Lessons learned from ruptured appendicitis Health Serv Res 2001 36 1 5 11324738
Newacheck PW Poverty and childhood chronic illness Arch Pediatr Adolesc Med 1994 148 1143 9 7921114
Forrest CB Starfield B Entry into primary care and continuity: the effects of access Am J Public Health 1998 88 1330 6 9736872
Weinick RM Krauss NA Racial/ethnic differences in children's access to care Am J Public Health 2000 90 1771 4 11076248
Flores G Rabke-Verani J Pine W Sabharwal A The importance of cultural and linguistic issues in the emergency care of children Pediatr Emerg Care 2002 18 271 84 12187133 10.1097/00006565-200208000-00010
Guagliardo MF Ronzio CR Cheung I Chacko E Joseph JG Physician accessibility: An urban case study of pediatric primary care Health and Place 2004 10 273 83 15177201 10.1016/j.healthplace.2003.01.001
Guagliardo Mark F Spatial accessibility of primary care concepts, methods and challenges Proceedings of the Eighteenth Annual Primary Care Research Methods and Statistics Conference 2003 San Antonio, TX: University of Texas Health Science Center
Mechanic D Disadvantage, inequality, and social policy. Major initiatives intended to improve population health may also increase health disparities Health Aff (Millwood) 2002 21 48 59 11900186 10.1377/hlthaff.21.2.48
Fiscella K Williams DR Health disparities based on socioeconomic inequities: implications for urban health care Acad Med 2004 79 1139 47 15563647 10.1097/00001888-200412000-00004
Auble TE Are outpatient admission sources truly a risk factor for appendiceal rupture? Acad Emerg Med 1999 6 579 80 10386673
Hurley RE Freund DA Taylor DE Emergency room use and primary care case management: evidence from four Medicaid demonstration programs Am J Public Health 1989 79 843 6 2660603
Gadomski A Jenkins P Nichols M Impact of a Medicaid primary care provider and preventive care on pediatric hospitalization Pediatrics 1998 101 E1 9481020 10.1542/peds.101.3.e1
Ullman R Block JA Stratmann WC An emergency room's patients: their characteristics and utilization of hospital services Med Care 1975 13 1011 20 1105025
Taylor DH JrWhellan DJ Sloan FA Effects of admission to a teaching hospital on the cost and quality of care for Medicare beneficiaries N Engl J Med 1999 340 293 9 9920955 10.1056/NEJM199901283400408
Chen J Radford MJ Wang Y Marciniak TA Krumholz HM Do "America's Best Hospitals" perform better for acute myocardial infarction? N Engl J Med 1999 340 286 92 9920954 10.1056/NEJM199901283400407
Krieger N Data, "race," and politics: a commentary on the epidemiological significance of California's Proposition 54 J Epidemiol Community Health 2004 58 632 3 15252062 10.1136/jech.2003.018549
Piantadosi S Byar DP Green SB The ecological fallacy Am J Epidemiol 1988 127 893 904 3282433
Krieger N Overcoming the absence of socioeconomic data in medical records: validation and application of a census-based methodology Am J Public Health 1992 82 703 10 1566949
Krieger N Chen JT Waterman PD Soobader MJ Subramanian SV Carson R Geocoding and Monitoring of US Socioeconomic Inequalities in Mortality and Cancer Incidence: Does the Choice of Area-based Measure and Geographic Level Matter?: The Public Health Disparities Geocoding Project Am J Epidemiol 2002 156 471 82 12196317 10.1093/aje/kwf068
| 15871740 | PMC1156944 | CC BY | 2021-01-04 16:37:42 | no | Popul Health Metr. 2005 May 4; 3:4 | utf-8 | Popul Health Metr | 2,005 | 10.1186/1478-7954-3-4 | oa_comm |
==== Front
Popul Health MetrPopulation Health Metrics1478-7954BioMed Central London 1478-7954-3-51588820110.1186/1478-7954-3-5ResearchEstimating summary measures of health: a structured workbook approach Flanagan William [email protected] Jane [email protected] Petit Christel [email protected] Jean-Marie [email protected] Statistics Canada, 24A RH Coats, Ottawa, K1A 0T6, Canada2 Public Health Agency of Canada, A. L. 6502D, 130 Colonnade Rd., Ottawa, K1A 0K9, Canada2005 11 5 2005 3 5 5 20 9 2004 11 5 2005 Copyright © 2005 Flanagan et al; licensee BioMed Central Ltd.2005Flanagan et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Summary measures of health that combine mortality and morbidity into a single indicator are being estimated in the Canadian context for approximately 200 diseases and conditions. To manage the large amount of data and calculations for this many diseases, we have developed a structured workbook system with easy to use tools. We expect this system will be attractive to researchers from other countries or regions of Canada who are interested in estimating the health-adjusted life years (HALYs) lost to premature mortality and year-equivalents lost to reduced functioning, as well as population attributable fractions (PAFs) associated with risk factors. This paper describes the workbook system using cancers as an example, and includes the entire system as a free, downloadable package.
Methods
The workbook system was developed in Excel and runs on a personal computer. It is a database system that stores data on population structure, mortality, incidence, distributions of cases entering a multitude of health states, durations of time spent in health states, preference scores that weight for severity, life table estimates of life expectancies, and risk factor prevalence and relative risks. The tools are Excel files with embedded macro programs. The main tool generates workbooks that estimate HALY, one per disease, by copying data from the database into a pre-defined template. Other tools summarize the HALY results across diseases for easy analysis.
Results
The downloadable zip file contains the database files initialized with Canadian data for cancers, the tools, templates and workbooks that estimate PAF and a user guide. The workbooks that estimate HALY are generated from the system at a rate of approximately one minute per disease. The resulting workbooks are self-contained and can be used directly to explore the details of a particular disease. Results can be discounted at different rates through simple parameter modification.
Conclusion
The structured workbook approach offers researchers an efficient, easy to use, and easy to understand set of tools for estimating HALY and PAF summary measures for their country or region of interest.
==== Body
Background
Over the past century, advances in public health and population health have dramatically increased life expectancy. Canadians now live longer, but during these added years, they may be affected by disease or chronic conditions. For this reason, indicators used to monitor changes in population health and guide policy decisions need to include how health conditions affect the day-to-day functioning of Canadians over their lifetime.
Summary measures of health that include both mortality and morbidity are being estimated for Canada [1]. Building on prior burden of disease studies by the World Health Organization [2] and Australia [3] that estimated disability-adjusted life years (DALY), the Canadian study will estimate the health-adjusted life years (HALY) lost to premature mortality and reduced functioning for approximately 200 diseases. HALY is computationally identical to DALY; however, it reflects a shift in terminology away from disability towards the broader term health, following recommendations originating from the International Network on Health Expectancy [4].
Estimating summary measures of health requires a wide variety of data including: population counts; incidence and mortality rates; life expectancies; cause-specific and observed survival; distributions, durations, and preference scores across a multitude of health states; and risk factor data to estimate population attributable fractions (PAF). Disaggregating by age group and sex further explodes the quantity of data. To efficiently manage such a large amount of information, we developed a database system, with a set of easy-to-use tools to automatically generate the summary measure estimates.
The main tool generates workbooks, one per disease, that calculate HALY, by importing the data from the database into a generic template. This makes it easy to update the database and quickly regenerate the results. The template is highly structured, which makes the generated workbooks easy to understand and use. There are also tools that summarize the HALY results across diseases for easy analysis. Parameters, such as the rate at which to discount future events, the population of study, and the reference life table, can be specified in the tools to evaluate different scenarios. Furthermore, the generated workbooks are self-contained and can be used directly for specific analysis of a disease. Finally, the tools were built generically to incorporate any number of diseases. Overall, we expect this workbook system will be attractive to other researchers, since it streamlines the process of estimating HALY and PAF, and at the same time provides an organized framework to document the work.
This paper focuses on the system, i.e., tools, database, templates and workbooks, that was developed to estimate the population health impact of cancers in Canada in 2001. Workbooks were generated to estimate health-adjusted life years lost for 26 cancer types (see Table 1) and population attributable fractions for five-related risk factors: alcohol, obesity, lack of fruit and vegetable consumption, physical inactivity, and smoking. The entire application, including a user guide, is available for download in this article's companion zip file.
Table 1 Cancer sites by ICD-9 code
ICD-9 Cancer site
140–149 Oral cancer
150 Esophageal cancer
151 Stomach cancer
153–154,159.0 Colorectal cancer
155 Liver cancer
156 Gall bladder cancer
157 Pancreatic cancer
161 Laryngeal cancer
162 Lung cancer
170–171 Bone and connective tissue cancer
172 Melanoma
173 Non-melanoma skin cancer
174 Breast cancer
180 Cervical cancer
182,179 Uterine cancer
183 Ovarian cancer
185 Prostate cancer
188 Bladder cancer
189 Kidney cancer
191–192 Brain
193 Thyroid cancer
200,202 Non-Hodgkin's lymphoma
201 Hodgkin's disease
203 Multiple myeloma
204–208 Leukemia
All sites between 140–208 not listed above All other cancers
Methods
The tools, database, templates, and workbooks that estimate HALY and PAF were developed in Microsoft Excel (version 2002, service pack 2). The interactions between the principal components of the system are shown in the data flow diagram in Figure 1.
Figure 1 Flowchart of workbook system.
Tools
The tools (identified by ovals in Figure 1) contain imbedded Visual Basic macros that perform three principal functions: generate workbooks to estimate HALY for each of the 26 cancer sites; summarize the HALY results across cancer sites for comparative analysis; and extract HALY results to be attributed to risk factors using population attributable fractions. The tools are called "Builder", "Summary" and "Extract", respectively. Two additional tools discussed below but not shown in Figure 1, are the "Master" and "UpdatePAF" tools. Each tool contains a command button that launches the macro and each has a set of options to control the macro's actions.
Templates
Each tool uses a template. A template is simply a pre-defined structure that contains formulae and place-holders for data. For instance, the template for the HALY workbooks contains formulae and formatting to calculate the HALY, but it does not contain data. The Builder tool copies the data from the database into the template for a selected disease.
Links
Some of the data in the generated workbooks are linked to the source files (shown as dashed arrow lines in Figure 1) using a feature of Excel called "links". This means that the data are stored externally to the workbook, but are shown and used in the workbook. The advantage of this approach is that it allows users to quickly change the source of data to easily update the workbook. For instance, the workbooks that estimate HALY can link to any one of the three reference life tables included (or users can create their own life table) to automatically update results. There is complete flexibility in the tool to choose which files to maintain as links. By default, only the population and life table database files are maintained as links. The rest of the data are simply copied from the database to minimize complexity.
Database
The database is a collection of 17 files organized by the type of data and include: mortality rates; incidence rates; population counts; life expectancy estimates; stage distributions; observed and cause-specific survival; case-fatality estimates; duration and distribution of common cancer health states (diagnosis, treatment, remission, palliative and terminal care); preference scores used to weight for the severity of each health state; utilities that describe the starting health state of the population; risk factor prevalence and relative risk of disease from risk exposure. In addition, three sets of life expectancies have been included in the database: a Canadian multi-cohort life table (2001), a Canadian period life table (1995–1997) and a model life table used by the World Health Organization [5]. The workbooks that estimate HALY link to the multi-cohort life table by default. In general, the database files have been structured by age group, sex, disease, stage, and health state.
To illustrate the workbook system, the database has been populated with Canadian data (or data representative of Canada) for 26 cancer sites and five related risk factors. Cancers were classified by ICD-9 code because of data availability at the time of study. Updating to ICD-10 would not require any change in the structure of the workbook system, only in the data entered into it. The data sources are identified in each of the database files and repeated in the generated workbooks. It is beyond the scope of this paper to discuss the methods used to arrive at this set of data (details available upon request from the authors).
Workbooks to estimate HALY
Algorithm
Each of the 26 cancer sites were modeled according to a general progression algorithm like the one shown in Figure 2. Although the experience of living with cancer may vary from patient to patient, for practical reasons, we are limited to identifying typical pathways that affect most patients. In this model, cancer patients progress from diagnosis, through a treatment phase to remission and eventually death, from either the cancer or another cause. We model palliative and terminal care phases for cases that die of the cancer.
Figure 2 General cancer progression model from diagnosis to death.
To improve our cancer model, we divided cancer cases by stage at diagnosis (localized, regional and distant) and disease progression into a set of discrete health states. Treatment options comprise surgery (in-patient and out-patient), chemotherapy (mild, moderate and severe toxicity), hormonal therapy and radiotherapy (curative and palliative). The treatment distributions are entered into the database by type of treatment. For instance, consider the distribution by type of surgery in the workbook for thyroid cancer (Figure 3): 53.2% of patients diagnosed with localized thyroid cancer receive in-patient surgery, 22.2% receive out-patient surgery and the remaining 24.6% do not have surgery. The residual, i.e., the proportion that do not receive surgery in this case, is not explicitly recorded in the database, but is calculated instead. We refer to the period after diagnosis and treatment and prior to the palliative/terminal phase (or death from other cause) as remission.
Figure 3 Example of treatment distribution for thyroid cancer.
Formulae
The workbooks contain imbedded formulae for calculating the summary measures. HALY is a summary measure that includes both the impact of mortality and morbidity in a single indicator. The mortality component measures the years of life lost due to premature mortality (YLL); the morbidity component quantifies the year equivalents of reduced functioning from living with the disease (YERF). YERF is analogous to years of life lived with disability (YLD) used by the World Health Organization and their collaborators in their burden of disease study; thus, HALY = YLL+YERF is synonymous with DALY = YLL+YLD.
For each cancer site, the HALY, YLL and YERF are estimated by age group and sex according to the following formulae:
HALYa,s = YLLa,s + YERFa,s (Eq1)
YLLa,s = Ma,s * La,s (Eq2)
YERFa,s = Σg Σe [la,s,g,e* Da,s,g,e* Wg,e] (Eq3)
where a represents the age group, s represents the sex, g represents the stage at diagnosis, e represents the state of progression of the cancer.
The YLLs are calculated from the number of cases that die from the cancer (M) and the estimated years of remaining life at the age of death (L). The latter is estimated from survival in the general population, by age and sex, and comes from the life table. The death counts are calculated from the mortality rates and the population counts.
The YERFs are calculated by health state and stage at diagnosis. They are estimated from the number of cases entering the health state (I), the duration in the state (D) and the weight for severity of the health state (W). The number of cases entering the health state is derived from the cancer incidence rates, the population counts, the stage distribution and the estimated proportion that experience the health state. For example, the number of women aged 50–54 that receive radiotherapy for cure of localized breast cancer is the product of the number of women in this age group (1,060,244 in Canada in 2001), the incidence rate (229 per 100,000), the estimated proportion that are diagnosed with localized disease (63.6%) and the proportion of these that receive radiotherapy for cure (43.0%), which amounts to 663 cases (these numbers can be found in the breast cancer workbook). The duration of the health state is a direct input parameter, except for the remission/on-going care states, which are calculated as the residual of the overall survival duration less the duration spent in the diagnostic, treatment and palliative/terminal phases.
The weight for severity of the health state is expressed in terms of preference scores (u), as W = 1-u. This assumes full health prior to entering the health state. However, the workbooks allow the population to start in partial health (u1) and persist co-morbidly with the cancer state (u2). The co-morbidity rule for combining preference scores, u1 and u2, of two conditions, was defined as:
u1,2 = (1- k) * minimum (u1, u2) + k * (u1 * u2)
The value of the comorbidity coefficient k was estimated at 0.34, based on a best-fit analysis of Health Utility Index [6] scores for conditions reported in the Canadian Community Health Survey, 2000–01 (CCHS) [7] (details available from the authors). Since we are interested in the reduced functioning relative to the initial health state (u1), the weight for severity of the cancer health state is given by W = u1 - u1,2.
Discounting
The workbooks include a parameter for discounting the durations of health states that occur at some time T after diagnosis. When a discount rate, r > 0, is specified, the YLL and YERF are estimated according to the modified functional forms:
YLLa,s = Ma,s * (1-e-rLa,s)/r (Eq4)
YERFa,s = Σg Σe [la,s,g,e* (1-e-rDa,s,g,e)*e-rTa,s,g,e]/r * Wg,e] (Eq5)
Although the timing and order of treatment, which determines the value of T, may vary from case to case in practice, we assume that treatments occur separately in time and in the following order: diagnosis; surgery; chemotherapy or hormonal therapy; radiotherapy; remission; palliative care; terminal care; and death. The palliative and terminal phases only apply to cases dying of the cancer. The duration preceding them is estimated from the cause-specific survival duration.
We did not include age-weighting in the HALY formulae due to its controversial interpretation [8,9].
Structure
Each of the cancer workbooks is structured identically because they originate from the same template. Figure 4 shows a pictorial representation of the cancer workbooks. The HALY and YLL calculations are straightforward implementations of the formulas described above. The total YERF estimate is a sum across the three stages (localized, regional and distant). The worksheets for each stage are structured identically and implement the YERF formulae at the level of the health state. In addition, each of the 17 health state calculations are structured identically, and show the number of new cases entering the health state (I), the duration of the state (D), the time from diagnosis (T), the preference score (u) and the YERF estimate, by age group and sex.
Figure 4 Structure of workbooks for cancer.
Colour scheme
For ease of use, all data elements and parameters that can be modified in the workbooks are identified as green-filled cells. Blue-filled cells are used to highlight labels and violet-filled cells highlight the summary measures.
Workbooks to estimate PAF
The population attributable fraction (PAF) is an estimate of the proportion of disease in the general population that is due to a particular risk factor. For the study of cancers, workbooks have been developed to estimate the population attributable fraction for five risk factors: alcohol, obesity, lack of fruit and vegetable consumption, physical inactivity, and smoking.
Given the lag time between the exposure to tobacco and the incidence of cancer, and given that the prevalence of smoking has been declining, using current prevalence of smoking will likely produce an underestimation of the population attributable fraction of smoking. In order to quantify this potential bias, we developed three workbooks to estimate the impact of smoking: the first uses current (2001) prevalence of smoking, a second uses prevalence reported in 1991 and the third is an indirect method developed by Peto and Lopez[10].
Formulae
For a given risk factor, the PAF is estimated by age group (a), sex (s) and cancer (c) according to the formula:
PAFa,s,c = Σi [ Pea,s,i * (RRa,s,i,c -1) / (1 + Pea,s,i * (RRa,s,i,c -1)) ] (Eq6)
where Pe is the proportion of the population exposed to the risk factor, RR is the relative risk of developing or dying of cancer due to the exposure, and index i represents the risk category[11]. For instance, the risk categories for obesity are underweight, normal weight, overweight and obese (base on BMI values).
To obtain a more global view of the impact of a risk factor, we produced summary estimates showing the proportion of the total number of cancer deaths, HALY, YLL and YERF attributable to each risk factor by applying the PAF estimates of equation 6 to each of these outcomes. For instance, the impact on deaths for a particular risk factor is given by the formula:
PAFDeathss = [ Σc Σa PAFa,s,c *DEATHSa,s,c] / [Σc Σa DEATHSa,s,c ] (Eq7)
The outcomes are first extracted from the HALY workbooks for a specific discount rate, life table and population choice. They are stored in a separate file and maintained as a link to each of the PAF workbooks. This allows the summary PAF estimates to be easily updated for different parameter choices.
Results
The tools, database files, templates and workbooks that estimate PAF are all available for download in this article's companion zip file. The workbooks to estimate the HALY need to be generated from the Builder tool after download. After the HALY workbooks have been built for all cancers, the Summary and Extract tools can be used to summarize the HALY results for specific parameter choices, and UpdatePAF tool can be used to update the file links in the PAF workbooks. A higher level tool, the Master tool, has been included to automatically execute these four tasks with the push of one button.
The database is currently populated with cancer data for Canada to illustrate usage, but can be easily adapted for other diseases and updated with data for other countries or regions. To update the database, simply open the database file(s) in Excel and replace the data using standard editing techniques. When adding other diseases, the structure of the database files may be changed to accommodate the number and naming of stages and health states (refer to the user guide for more details).
Here is a brief description of each of the components of the system. More details can be found in the user guide.
Master Tool
The Master tool runs all of the individual tools – Builder, Summary, Extract, and UpdatePAF – to build a complete system. It allows four parameter choices that are applied throughout the system: the disease chapter for which to run the system (only Cancer is present in this example); the rate at which to discount future events; the comorbidity coefficient to be applied in the rule for combining utilities of two conditions; and the reference life table used to estimate remaining life expectancies. A snapshot of the interface is shown in Figure 5. The Tables & Figures worksheet is an advanced feature used to create publication-style tables and figures. It is discussed more thoroughly in the user guide.
Figure 5 Snapshot of the Master macro tool.
Builder Tool
The purpose of the Builder tool is to generate workbooks that estimate HALY by importing data from the database into a pre-defined template. This allows great flexibility to update the data and recreate the workbooks. The tool, shown in Figure 6, generates one workbook per cancer site. The command button that launches the macro program is at the top of the worksheet and the green-filled cells indicate options that can be specified. The workbooks can be created for the entire cancer chapter or for a specific cancer. Each of the database files required for input to the workbooks is listed by name. Links can be maintained to all of these database files; however it is recommended that links be maintained only for the population and life table files. Other options identify the template file, the location of the database and output folder.
Figure 6 Snapshot of the Builder macro tool.
Summary Tool
We designed the Summary tool (Figure 7) to aid in the analysis of the summary measures across all cancer sites. It copies the total HALY, YLL and YERF estimates from each of the cancer workbooks into a single output file. The user can choose parameter values for three discount rates, the life table, and the comorbidity coefficient; these parameters are updated in the HALY workbooks before copying the summary measures to the output file. The parameter choices are automatically reflected in the name of the generated summary file.
Figure 7 Snapshot of the Summary macro tool.
Extract Tool
The Extract tool (Figure 8) was built to facilitate the attribution of summary measures (deaths, HALY, YLL and YERF) to the various risk factors across all cancer sites. The tool simply copies the summary measures, by sex and age group, from each of the cancer workbooks into one file. The data can be extracted under different choices of discount rate, life table, population and comorbidity coefficient. The name of the output file reflects these choices and allows users to create several different extract files for analysis.
Figure 8 Snapshot of the Extract macro tool.
UpdatePAF Tool
The workbooks that estimate population attributable fractions link to the file generated by the Extract tool. The UpdatePAF tool was created to facilitate the update of this link across all eight of these workbooks. The name of the file to be linked is specified in the tool. As with the other tools, UpdatePAF is run automatically by the Master tool.
Workbooks to estimate HALY
Once created, the workbooks that estimate HALY can be used as stand-alone workbooks in which parameters or data can be changed. They can also be regenerated at any time with the Builder tool. Figure 9 is a snapshot of the HALY workbook for lung cancer. Each of the workbooks that estimate HALY contains 10 worksheets.
Figure 9 Snapshot of the workbook that estimates HALY for lung cancer.
The Instructions worksheet offers basic guidance on using the workbooks. It identifies the cancer by name and ICD-9 code. The choice of the reference population and the discount rate are specified in this sheet and applied in all subsequent worksheets. The population counts are displayed in the HALY sheet.
The Algorithm worksheet contains the distribution of treatment and remission associated with the cancer, the preference scores for each of the cancer's health states, and utilities that describe the starting health state of the population. It also contains the comorbidity coefficient as a parameter that can be changed. Changing the values here automatically updates the YERF estimates.
The HALY worksheet calculates the health-adjusted life years lost as the sum of the YLL and YERF values. The population counts, chosen in the Instructions sheet, are displayed here. The mortality rates are input to the YLL worksheet. Population counts and life expectancy estimates are linked data elements used in the calculation of the YLL.
The YERF worksheet calculates the total YERF values by summing across stages. The incidence rates for the cancer are found in this sheet. They are combined with the population counts to generate incidence counts, which are then distributed by stage.
The YERF local, YERF regional, and YERF distant worksheets calculate the year-equivalents of reduced functioning by health state for each stage, respectively. The stage distribution and the various durations (cause-specific survival, observed survival, duration of treatments, and duration of palliative and terminal care) can be modified. The distribution of treatment and the preference scores are more easily modified in the Algorithm sheet.
The Sources worksheet lists every data element used in the workbook, its source and in which worksheet it is found.
The Notes worksheet highlights anything exceptional or noteworthy about the cancer.
Workbooks to estimate PAF
There are PAF workbooks for each of the five risk factors. They each contain nine worksheets as shown in Figure 10. The Info worksheet provides some basic information related to the calculation of the PAF; the Pe worksheet contains risk factor prevalence data, the RR worksheet contains the relative risk data for the risk factor and the PAF worksheet implements the formulae to estimate PAF by cancer site, age group and sex. In the remaining worksheets, the PAFs are applied to various summary measures, such as HALYs, to estimate how much may be attributable to the risk factor. These results are summarized in the Summary worksheet.
Figure 10 Snapshot of the PAF workbook for alcohol.
The workbook that calculates the PAF associated with smoking by the indirect method includes two additional worksheets with data on the number of lung cancer deaths in a reference population (American Cancer Society, CPS-II, 1984–1988) and in Canada. They are used to estimate the hypothetical proportion that would have to have been exposed to smoking to account for the lung cancer mortality observed in 2001.
HALY Summary Workbook
Figure 11 is a snapshot of the workbook that summarizes the HALY results. It shows the HALY, YLL and YERF estimates by sex and cancer site, rank ordered by the total HALY. It reports these values for three different discount rates as specified in the Summary tool at the time of creation. The life table and comorbidity coefficient choices are shown and reflected in the file name.
Figure 11 Snapshot of the HALY summary workbook.
PAF Summary Workbook
The PAF summary workbook (Figure 12) maintains links to each of the individual PAF workbooks for the risk factors and links to the extracted HALY file to obtain the total counts of deaths, HALY, YLL and YERF. It is important that all PAF workbooks link to the same extracted file (the UpdatePAF tool can be use to accomplish this).
Figure 12 Snapshot of the PAF summary workbook.
Discussion
The structured approach of the workbook system provides researchers and policy makers with an easy to use and easy to understand tool for estimating HALY and PAF summary measures. Developed for use on personal computers using Excel, it is widely accessible to all levels of researchers. The database can easily be updated with data for other regions or countries and the entire set of results quickly regenerated. Parameter choices in the tools and in the resulting workbooks offer great flexibility to create alternative scenarios. Counterfactuals, used to evaluate the impact of basic health policy interventions, can be created by modifying any of the data elements. For instance, the population attributable fraction associated with obesity could be re-evaluated by reducing the prevalence of obese individuals by an amount that might be achieved by an intervention strategy.
The workbook system has a number of limitations. First, we have not causally linked disease incidence to mortality. Instead, the mortality in 2001 is taken as a proxy for the mortality that would result from the incidence observed in 2001. However, a system that linked incidence to mortality would be more realistic, especially when looking at interventions that reduce incidence. Similarly, survival is not causally linked to treatment in the workbook model. Scenarios that alter treatment patterns would not impact survival time.
A second limitation is that the model of cancer progression does not include the treatment of local or distant recurrence. This means we have not incorporated the weight for severity of these conditions, which would occur during the period labeled remission/on-going care. The preference scores associated with distant cancer are lower than for other stages, so we would expect some underestimation of the HALY by the omission of distant recurrences. This is not expected to have much impact on the ranking of cancers.
Third, the workbook model assumes that cancer treatment follows a fixed sequential order: surgery, chemotherapy, radiotherapy. While the implications for individuals may be extremely important, from a population perspective, and more practically, from the perspective of data availability and model complexity, simplifying assumptions are required. Since the durations of these treatments are relatively short compared to the observed survival of cancer patients, we would expect them to have little impact on the overall outcomes and we would expect their order of occurrence to have even less impact. As a crude sensitivity analysis, we changed the order of occurrence of chemotherapy and radiotherapy in the workbook system, which led to a negligible impact on the estimate of morbidity: overall YERF changed by 0.00025%.
Fourth, clustering of risk factors cannot be easily modeled in the workbooks. This means that the proportion of cancer attributable to one risk factor may also be attributed to another risk factor, even though the two risk factors collectively contribute to the cancer. For instance, alcohol and smoking may be clustered risk factors with respect to death from laryngeal cancer, which may explain why our estimate of the total proportion of laryngeal cancer deaths exceeds 100%. In general, we expect we have overestimated the population attributable fractions across all cancers.
Finally, the data have been obtained at levels of disaggregation to represent the heterogeneity of the cancer population. However, in the case of the calculated duration of remission, it has not been sufficient to avoid logical inconsistencies. We found that in older age groups, it was possible to generate negative durations of remission, because the observed survival of people in this age group was less than the duration spent in diagnostic, treatment and terminal phases. The input data could be refined to avoid this, but as a rare occurrence with small impact, we opted to check for negative durations and set them to zero when they occur. This is done automatically by the imbedded formulae and requires no intervention by the user.
These limitations can be overcome through more advanced modeling techniques, such as microsimulation modeling. The Population Health Model, a continuous time, competing risk microsimulation model developed at Statistics Canada, is being adapted to implement all of the functionality of these workbooks. Our experience with both types of modeling suggests that it is not necessarily more difficult to develop the microsimulation model, although it is often considered less transparent. The benefit of developing both models is that the workbooks provide a benchmark against which the microsimulation results can be compared. Of course, we expect small differences, not only from random noise introduced by the stochastic nature of the microsimulation model, but also because it avoids the limitations outlined above.
The workbook system presented here focused on cancers. However, it was developed more generally, so that it can produce workbooks for other chronic diseases or injuries, once the data have been assembled. The main criterion is that the disease(s) can be decomposed into a series of health states from diagnosis to death. The number of health states is virtually unlimited and can include disease progression and sequelae, and the diseases can be staged at diagnosis or not. By changing a few labels in the database files to reflect health state names, an entire new system of HALY workbooks can be generated (refer to user guide for detailed steps). The underlying macro code has been built with this generalizability in mind.
As with any generalized system, exceptions may arise that do not fit within its structured framework. As more diseases are studied in the Canadian study, the workbook structure will be modified or expanded to address any exceptions that arise. This may take the form of minor modifications to the current structure, a separate structure to accommodate multiple disease exceptions that fall into a common framework, or a series of ad-hoc, stand-alone workbooks for unique exceptions. We expect that most diseases will fit within the structure described here. Future releases of the workbook system will be made through the Public Health Agency of Canada's website[1].
Conclusion
The structured workbook approach offers researchers an efficient, easy to use, and easy to understand set of tools for estimating HALY and PAF summary measures for their country or region of interest. The estimation of summary measures for cancers presented here highlights the functionality of the system; however, the tool is easily expanded to other diseases. The workbooks are transparent in their calculations, but are limited in their ability to model the impact of clustered risk factors and competing risks of disease. These limitations can be overcome by more advanced modeling techniques such as microsimulation.
Competing Interests
The author(s) declare that they have no competing interests.
Authors' contributions
WF participated in the conceptualization and testing of the workbook system, designed and developed the system, and drafted the manuscript. CLP and JBP participated in the conceptualization and testing of the workbook system. JMB provided the principal guidance in the conceptualization of the workbook system. All authors have read and approved the manuscript.
Supplementary Material
Additional File 1
PHI Workbook System. This 2.6 Mb zip file for download contains the Excel files that comprise the workbook system. The system estimates summary measures of health and is initialized with cancer data for Canada. When unzipped, the workbook system will occupy approximately 6 Mb of additional disk space and a folder structure will be created under C:\PHI, where you will find the user guide.
Click here for file
Acknowledgements
This work is part of the Population Health Impact of Disease in Canada (PHI) research program, a collaboration of Statistics Canada, Public Health Agency of Canada, and researchers from McGill University, the University of Ottawa, the University of Manitoba, the Institute for Clinical Evaluative Sciences (ICES) and the Montérégie Regional Board of Health and Social Services. The PHI is funded by Statistics Canada and Public Health Agency of Canada. The authors acknowledge Kathy White for her editorial input to this manuscript, Hélène Roberge and Serge Tanguay for their contribution to the assembly of the data, and Dr Bill Evans for review of the cancer model and for providing the cancer treatment algorithm.
==== Refs
Population Health Impact of Disease in Canada
Murray CJ Lopez AD Global health statistics Global Burden of Disease and Injury Series 1996 2 Harvard: Harvard School of Public Health
Mathers C Vos T Stevenson C The burden of disease and injury in Australia 1999 AIHW cat. no. PHE 17. Canberra: AIHW
Mathers CD Robine J-M Wilkins R Mathers C, McCallum J, Robine J-M Health expectancy indicators: recommendations for terminology Advances in health expectancies 1994 Canberra: Australian Institute of Health and Welfare 34 41
Coale A Guo G Revised regional model life tables at very low levels of mortality Popul Index Winter 1989 55 613 43
Furlong WJ Feeny DH Torrance GW Barr RD The Health Utilities Index (HUI) System for assessing health-related quality of life in clinical studies Ann Med 2001 33 375 84 11491197
Health Statistics Division, Statistics Canada Canadian Community Health Survey 2000–01 Statistics Canada survey 3226
Barendregt JJ Bonneux L Van der Maas PJ DALYs: the age-weights on balance Bull World Health Organ 1996 74 439 43 8823967
Barendregt JJ Robine JM, Jagger C, Mathers D Disability-adjusted life years (DALYs) and disability-adjusted life expectancy (DALE) Determining life expectancies 2003 Chichester (UK): Wiley 247 261
Peto R Lopez AD Boreham J Thun M Heath C Jr Mortality from tobacco in developed countries: indirect estimation from national vital statistics Lancet 1992 339 1268 78 1349675 10.1016/0140-6736(92)91600-D
Jekel JF Katz DL Elmore JG Epidemiology, Biostatistics, and Preventive Medicine 2001 second Philadelphia (PA): WB Saunders Company
| 15888201 | PMC1156945 | CC BY | 2021-01-04 16:37:42 | no | Popul Health Metr. 2005 May 11; 3:5 | utf-8 | Popul Health Metr | 2,005 | 10.1186/1478-7954-3-5 | oa_comm |
==== Front
Proteome SciProteome Science1477-5956BioMed Central London 1477-5956-3-31590452610.1186/1477-5956-3-3ResearchTarget selection of soluble protein complexes for structural proteomics studies Shen Weiping [email protected] Steven [email protected] Bonny [email protected] Kush [email protected] Frederic F [email protected] Department of Molecular Biology and Biochemistry, Simon Fraser University, 8888 University Drive, Burnaby, British Columbia, Canada, V5A 1S62005 18 5 2005 3 3 3 24 12 2004 18 5 2005 Copyright © 2005 Shen et al; licensee BioMed Central Ltd.2005Shen et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Protein expression in E. coli is the most commonly used system to produce protein for structural studies, because it is fast and inexpensive and can produce large quantity of proteins. However, when proteins from other species such as mammalian are produced in this system, problems of protein expression and solubility arise [1]. Structural genomics project are currently investigating proteomics pipelines that would produce sufficient quantities of recombinant proteins for structural studies of protein complexes. To investigate how the E. coli protein expression system could be used for this purpose, we purified apoptotic binary protein complexes formed between members of the Caspase Associated Recruitment Domain (CARD) family.
Results
A combinatorial approach to the generation of protein complexes was performed between members of the CARD domain protein family that have the ability to form hetero-dimers between each other. In our method, each gene coding for a specific protein partner is cloned in pET-28b (Novagen) and PGEX2T (Amersham) expression vectors. All combinations of protein complexes are then obtained by reconstituting complexes from purified components in native conditions, after denaturation-renaturation or co-expression. Our study applied to 14 soluble CARD domain proteins revealed that co-expression studies perform better than native and denaturation-renaturation methods. In this study, we confirm existing interactions obtained in vivoin mammalian cells and also predict new interactions.
Conclusion
The simplicity of this screening method could be easily scaled up to identify soluble protein complexes for structural genomic projects. This study reports informative statistics on the solubility of human protein complexes expressed in E.coli belonging to the human CARD protein family.
==== Body
Background
The aim of structural proteomics is to experimentally derive the 3-dimensional structure of all proteins in genomes through determination of the 3D structure of sufficient members in a protein family such that the remaining structures can be predicted accurately through computational approaches [2]. The success rates in studies so far, mainly applied to prokaryotes, are still low. Effectively, 1 to 5% of structures are solved experimentally from a few thousand of Open Reading Frame (ORF) tested per study. The reasons for this limitation are primarily due to the fact that many proteins are insoluble when expressed in a heterologous system or when purified and concentrated at the levels necessary for crystallization studies. The success rate is even lower when applied to eukaryotes, where for example, it was shown recently that in C. elegans only 20% of the ORF cloned produced soluble recombinant protein suitable for structural studies. This rate is significantly lower than that of structural genomics studying prokaryotics ORF [3]. In these studies, a tremendous amount of effort has been spent to improve protein solubility by switching the expression system, changing the fusion tags, replacing the expression host, generating a new structural variant by modifying the protein through genetic engineering including mutagenesis [4] and computational approaches, and finally by performing crystallization studies on a protein orthologs to the initial protein candidate [2,5]. However, full automation of these approaches needs to be improved and new technologies are constantly being implemented in the pipelines to increase our success rates in structure determination. An even more challenging and important task for structural proteomics studies is to determine the 3D structure of all the protein complexes of an organism. Most proteins do not work as a monomer, but interact with other proteins to perform their functions in the form of stable or transient complexes. In addition, the knowledge of these interactions is of fundamental importance, since the genome complexity of an organism is not simply related to its number of genes but rather is more directly related to the complexity of its protein-protein interaction networks. Full automation processes are still difficult because the proteomics pipeline to obtain sufficient quantities of soluble protein complexes has to allow for the possibility of different purifications scenarios for each protein complex. Furthermore, protein complexes are generally difficult to reconstitute from individual protein components and usually it is almost impossible to purify sufficient quantities of in vivo protein complexes for structural studies. Finally, interactions between proteins in cells are effectively dynamic with a wide range of affinities. The half-life of protein complexes is quite diversified from stable complexes to transient interactions and can be dependent on post-translation modifications. The post-translational modification may not be reproduced during in vitro experiments or are inherently not suitable for crystallization studies since highly stable systems are necessary for successful crystallographic studies [6-8].
Three steps are needed to obtain sufficient protein complex in vitro to perform structural studies of protein complexes from individual components. (i) Produce soluble protein, (ii) make a soluble complex, and (iii) generate high-concentration and highly purified protein complexes to tackle the crystallization project. In this study, we focus on the first step, that of obtaining soluble protein complexes. A few structural genomics initiatives are currently trying to develop proteomics pipelines that includes the production in E.coli of recombinant protein complexes [9]. We tried to identify an ideal protein family that could be used for this purpose. It should be a small soluble domain that can perform simple protein complexes and from which many protein complexes have been identified in vivo. As a result we choose the human CARD protein family. This protein family belongs to the death domain super-family consisting of the Death Domain (DD), Death Effecter Domain (DED) and the Pyrin, AIM (Absent-in-melanoma), ASC, Apoptosis-associated speck-like protein containing a caspase recruitment domain CARD, and Death-Domain (DD)-like (PAAD/DAPIN/PYRIN) domain subfamilies. The CARD domain family contains 35 members that are involved in apoptosis and share a domain of 85 amino-acids folded into a 6 helix bundle. The domain is a recruitment domain that has the ability to form hetero-dimers between members of the same family providing a structural framework to complete the apoptotic cascade leading to caspase activation and cell death. To date, two different structures of death domain complex have been solved by x-ray crystallography represented by the structure of the human CARD domains of Apoptotic Protein Activating Factor-1 (APAF-1) with caspase-9[10], and the death domain of Pelle with Tube in Drosophillia[11]. Structural analysis of these two different protein complexes revealed different modes of protein-protein recognition while similar domains are involved in protein-protein interaction.
In this work, the first step of a structural proteomic project to obtain soluble protein complexes from the CARD domain is reported. From 25 CARD proteins cloned and expressed in E.coli, 14 different CARD soluble proteins were obtained. Protein purification of CARD recombinant proteins were performed and binary protein complexes were reconstituted systematically from purified individual components under native, denaturation-renaturation conditions and by co-expressing two different CARD proteins simultaneously in E. coli to identify stable protein-protein interactions.
Results
Expression and solubility
To study the interaction between CARD-containing proteins, 14 CARD recombinant proteins were subjected to three different binding assays. First, all recombinant clones were characterized for their expression and solubility in E. coli. Comparison of the total cell lysate and soluble protein fractions revealed that most of the CARD proteins were expressed at medium or high levels (2–5 mg/liter). Only pET constructs (that add a poly-histidine to the N-terminus of the expressed protein of interest) containing the CARD domains of the cellular homolog of the equine herpesvirus-2 E10 gene containing an amino-terminal caspase recruitment domain (BCL10) and the apoptotic protein containing a CARD domain (CLAN) (pET-28b-BCL10 and pET-28b-CLAN) were producing a recombinant protein expressed at low level (less than 1 mg/ml). Protein expression was induced at two different temperatures (37°C and 30°C) and two different IPTG concentrations (1 mM and 0.1 mM) to find the optimal solubilization conditions.
Three pET-28b-constructs of the, CARD only protein (COP), NucleOLar protein 3 (NOL3), and the Apoptosis-associated Speck-like protein containing a CARD (ASC) (pET-28b-COP, pET-28b-NOL3, and pET-28b-ASC) produced in E.coli recombinant proteins were insoluble. Although some improvement of solubility was seen after induction at 30°C, we classified them as "not soluble" because of low yield (below 1 mg/L of culture). These insoluble proteins were purified from E. coli under denaturing conditions and then renatured back to the native state. The status of refolding was confirmed by circular dichroism (CD) experiments (data not shown).
Protein purification and direct native binding assay
The 14 cloned, CARD-containing proteins were purified on a larger scale and their optimal solubilization conditions were determined empirically. If insufficient protein purity (<90%) was obtained after his-tag affinity purification, some of the proteins were further subjected to ion-exchange chromatography. A total of 91 samples were initially set up for direct binding under native conditions. A quick way to observe any binding was to run these samples on a continuous non-denaturing polyacrylamide gel, along with the individual proteins as a control. Any shift in the electrophoretic mobility relative to that of the control lanes would indicate positive binding. This approach resulted in the identification of one direct binding and 3 indirect bindings (See additional file 1, Figure 1). Only one sample, run on a non-denaturating polyacrylamide gel from a mixture of Apaf-1 and caspase-9, showed a shifted band relative to Apaf-1 and caspase-9 control bands. In 3 other samples of the apoptotic protein containing a nucleotide binding domain and a CARD domain (NAC) and ASC, CARD-9 and BCL-10, ASC and BCL-10, we observed a decrease in the intensity of the stained bands for the protein mixture sample lane corresponding to the individual protein, but we did not detect any shifted band for the protein complex. We suggest that in these assays binding occurred because an equal amount of protein was loaded in the control lanes as in the protein mixture sample lanes, but the total charge of the protein complex is such that the complex cannot migrate into the gel during the electrophoresis. Consequently, the protein complex is not detected by the staining procedure. In addition, the two protein-protein interactions between NAC and ASC, CARD-9 and BCL-10 were confirmed by other investigators [12,13] while the binding between ASC and BCL-10 has not yet been determined experimentally. There may also be additional bindings that were not detected because of the limited sensitivity of Coomassie blue staining.
Figure 1 Native binding assay between Apaf-1 and Caspase-9 detected by PAGE. Binding reactions were performed using an equal amount of putative interacting histidine-tagged native recombinant proteins and incubation overnight in the binding buffer containing 50 mM Tris, 100 mM NaCl, pH 8.0. The binding was analyzed by a continuous Poly-Acrylamide Gel Electrophoresis (PAGE) (10%) followed by Coomassie Blue staining.
Induced denaturation-renaturation binding assay
To identify any additional bindings, we decided to denature the protein mixture first, and then refold both proteins in order to obtain the protein complex. This approach resulted in the identification of a total of 5 bindings including the one observed from the native binding assay with Apaf-1 and caspase-9 (See additional file 2). Non- denaturating polyacrylamide gel electrophoresis revealed bindings within the following proteins: The Apoptotic Protein Activating Factor-1 (Apaf-1) and caspase-9, NAC and Apaf-1, the Tumor Up-regulated CARD-containing Antagonist of caspase-Nine (TUCAN) and caspase-9, NAC and CLAN, and CARD-9 and CLAN (Figure 2). NAC and Apaf-1 have been previously shown to bind to each other by immunoprecipitation [14]. Bindings were also reported for TUCAN and caspase-9 [15], and NAC and CLAN [16], which is consistent with our findings. In addition, one additional binding observed between CLAN and CARD-9 was newly identified.
Figure 2 Denaturation-renaturation binding assay between NAC and APAF-1, TUCAN and CASPASE-9 detected by PAGE. Continuous PAGE analyses of CARD-containing proteins. An equal amount of proteins were incubated in a dialysis bag and dialyzed sequentially in a denaturation buffer, renaturation buffer and binding buffer (see material and methods). The protein complex formation was visualized by non-denaturating PAGE.
Co-expression analysis
In this study, the interaction of 14 CARD-containing proteins was investigated by co-expression in E. coli using a two-vector expression system with two different antibiotic resistances. The chosen expression vectors were pET-28b and pGEX-2T (that add a glutathione S-Transferase at the N-terminus of the expressed protein) with kanamycin and ampicillin resistances, respectively. Since only the protein expressed from pET-28b carries an N-terminal histidine-tag, retention of the other protein on a Ni-NTA affinity column depends on its interaction with the pET-28b expressed protein. In order to prevent any possible false positive bindings, pGEX-2T proteins were expressed alone and were also subjected to Ni-NTA affinity chromatography. In addition, each CARD protein was cloned in both vectors to perform the assay in both directions. A total of 156 co-expressions were performed and analyzed by SDS-PAGE. We obtained 12 positive bindings (See additional file 3). Among them, NAC, CARD-9, and the Receptor-Interacting Protein (RIP)-Associated protein with a Death domain (RAIDD) could form homodimers, regardless of which vector was used in the pull down assay (Figure 3). All bindings observed in the previous two assays were also detected. The previously reported binding of CLAN and ASC [17] was also confirmed. However, there were some reported bindings that we did not obtain in this study including: The binding of CLAN and BCL-10 [16] CARD-9 and BCL-10 [13], TUCAN and the caspase-1 inhibitor (ICEBERG) [18], TUCAN and COP [18], and COP and RICK [19]. In addition, some new bindings were observed between CARD-9 and RAIDD, CARD-9 and CLAN, RAIDD and NAC, ASC and ICEBERG
Figure 3 Immunoprecipitation of putative CARD binary complexes obtained from co-expression assay. Protein lysates obtained from the co-expressed E. coli putative CARD complexes were immunoprecipitated with anti-6XHis or anti-GST antibody. The precipitated protein samples (lane 3) were loaded onto a SDS-PAGE along with the individually-expressed protein lysates (lane 1 and 3). The protein samples from the gel were transferred onto Hybond-ECL (Amersham Biosciences). The immunoprecipitated protein complexes were detected with anti-GST and anti-6XHis antibody.
In order to determine whether these in vitro bindings from co-expression studies also occur in mammalian system, equal amounts of E. coli recombinant proteins were incubated in rabbit reticulocyte lysate (Figure 4), followed by immuno-precipitation and western blotting. As expected, we confirmed protein complex formation in the mammalian system among the newly identified protein complexes by co-expression. One interesting observation is that the co-expression of the pET-28b-ASC and pGEX-2T-ICEBERG construct, leads to a soluble protein complex, while the pET-28b-ASC when expressed individually produced a recombinant protein considered insoluble.
Figure 4 Binary complex formation of CARD protein complex in rabbit reticulocyte lysates. Equal amounts of in vitro expressed proteins were incubated in rabbit reticulocyte lysates at 4°C for 1 hour. After incubation, protein and complexes were immunoprecipitated with anti-6XHis antibody. The precipitated samples (lane 2) together with the individually-expressed proteins (lane 1 and 3) were detected by western blot using anti-GST antibody.
Discussion
In order to develop a low cost, simple and reliable system, which may be used to generate in vitro sufficient quantity of protein complexes for protein structure determination, we compared three in vitro binding assay systems: native binding assay, denaturation-renaturation binding assay, and co-expression binding assay. While these three different systems can be easily scaled up to generate large quantities of protein complex at very low cost, their accuracy in reproducing known CARD-CARD interactions described in the previous literature were quite different (Figure 5). According to previous studies, seven CARD protein complexes were reported within our 14 CARD domain members ([12-17,20]). Co-expression assay reproduced five of them with the highest accuracy of 71.4%. Induced denaturation-renaturation binding assay was able to generate four complexes from seven published complexes (57%). Native binding assay detected three complexes from the seven identified complexes with an accuracy of 42.9%. In addition, these three in vitro assays also identified a total of four new CARD protein complexes. Co-expression studies identified all four new complexes and the other two methods identified one complex each. Our results from rabbit reticulocyte lysate incubation study indicated that CARD protein complexes obtained between RAIDD and NAC, ASC and ICEBERG could be formed in rabbit reticulocyte mammalian system. A protein-protein interaction detected between CARD9 and RAIDD was also identified, but the identified proteins had different sizes compared to the E.coli recombinant proteins not incubated in the rabbit reticulocyte lysate. These data might suggest that the differences observed in the electrophoretic mobility of the interacting proteins could be accounted for by aggregation, post-translational modification or assisted folding. Such events would occur during their incubation in the mammalian lysate. It is not surprising that the CARD domain of ASC and ICEBERG could interact with each other in our assay since both CARD domains have been shown by others to bind to a common protein procaspase-1. Effectively, recent studies [21] demonstrated that ICEBERG binds to procaspase-1 through CARD- CARD interaction, which prevented procaspase-1 association and activation with the NF-kappaB-activating and cell death-inducing kinase RIP2. As a result, ICEBERG inhibits the processing of procaspase-1 essential for the activation of caspase-1. In addition, ASC also interacts with the CARD domain of procaspase-1[22]. Functional outcome of this interaction is dependent on the expression level of ASC in cells, since ASC enhances caspase-1 secretion into the cell culture at low concentration, but suppresses it at high concentration. Further studies will need to be performed to assess the functional significance of the interaction between ASC and ICEBERG. It is also quite striking that the newly identified protein complex obtained between NAC [14] and RAIDD [23] are among two proteins targets of the stress-induced apoptosis pathway, but previously shown to be involved at different point of the mitochondrial intrinsic pathway. RAIDD participates in the activation of caspase-2 by forming a multiprotein complex named P53-Induced protein with a Death Domain protein complex (PIDDosome) [24] that contain PIDD, RAIDD, and caspase-2. On the contrary NAC increases the processing of caspase-9 by binding to Apaf-1 in a multiprotein complex named the apoptosome consisting of apaf-1 and procaspase-9. So far, there is no evidence or publications to indicate a protein interaction between the CARD domain of RAIDD and NAC. Further functional studies need to be performed to assess the significance of this protein complex in apoptosis. Co-incubation between CARD9 and CLAN in a mammalian system failed to demonstrate interaction between the CARD domains of these two proteins. The in vitro protein complex formed between the CARD domains of CARD9 and CLAN in the co-expression assay could be an artificial or a newly identified complex. Further investigation, such as co-expression of these two proteins in mammalian system, should be performed to confirm these results.
Figure 5 Venn diagram representing the overlap between the 3 binding assays tested. N, represent native, CO, co-expression and DN-RN DeNaturation-ReNaturation assay.
In this study, co-expression binding assay performed the best compared to the other 2 approaches tested. It was also the simplest assay system since it did not require prior purification of the binding partners before complex formation. In addition, we observed that some proteins were not soluble when expressed alone, but could form a soluble protein complex in co-expression experiments. For example, the CARD domain of ASC protein is barely soluble when expressed alone in E. coli, but its complex with the CARD domain of ICEBERG in co-expression assay is totally soluble.
Both, native binding assay and induced denaturation-renaturation binding assay were more sensitive to sample preparation. For the native assay, among most of the protein pairs tested the binding was not detected even when both proteins were well folded and stable. This could indicate that the binding conditions needs to be optimized for each complex or that assisted folding and post-translational modifications are necessary for optimal binding. For the induced denaturation – renaturation binding assay, since the co-incubated pairs of proteins are denaturated then renaturated in the same conditions to form complexes, those conditions should be such that some potential protein complex could not be formed between some of the pairs tested in this study [25,26]. This may be the reason why three out of seven published CARD-CARD complexes could not be reproduced in this assay.
Conclusion
Our experimental results indicate that the co-expression system is recommended as the first choice for structural study of protein-protein complexes. This binding assay system is simple, inexpensive and can be easily automated for structural proteomics studies.
Methods
Cloning
EST clones encoding CARD domain were searched in the NCBI database . 14 EST spanning the CARD domain were found and used as templates for PCR amplification. The CARD domains were amplified by PCR and cloned into the expression vector pET or pGEX to obtain the following CARD-containing proteins: NAC (residues 1373–1473), TUCAN (residues 341–431), Apaf-1 (residues 1–108), caspase-9 (residues 1–111), CARD9 (residues 14–100), BCL10 (residues 1–104), RAIDD (residues 9–95), COP (residues 1–98), NOL3 (residues 1–97), CLAN (residues 1–87), RICK (residues 417–540), ICEBERG (residues 1–90), CARD10 (residues 31–117), and ASC (residues 92–195). The individual CARD domains (except RICK) were cloned into the Nde1-BamH1 and BamH1-EcoR1 restriction sites of pET-28b (Novagen) and pGEX-2T vectors, respectively. Since RICK had an internal BamH1 site in its sequence we cloned it into Nde1-EcoR1 sites of the pET28b vector.
Protein expression, solubility, and purification
pET-28b clones were transformed into E.coli BL21(DE3) (Novagen) and grown on petri dishes containing 50 μg/ml kanamycin. Cultures were grown from single colonies in 25 ml LB medium supplemented with kanamycin (40 μg/ml) at 37°C. His-tagged proteins were induced at an OD600 of 0.6 by the addition of IPTG to a final concentration of 1mM. After an additional 4 hours of growth at 30°C or 37°C, 2x1 mL aliquots of the culture were centrifuged to obtain cell pellets (5 min at 13000 rpm). One of the cell pellets were re-suspended in denaturing buffer and kept as the whole cell fraction. The other fractions were re-suspended in native lysis buffer (50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole), sonicated and centrifuged (5 min at 13000 rpm) in Eppendorf tubes. The resulting supernatant representing the soluble protein fraction was compared against the whole cell fraction by SDS-PAGE to determine the size, and the relative expression and solubility levels of each protein.
Binding assay – native condition
For this assay the 14 pET-28b constructs were transformed into E. coli BL21 (DE3). Transformed cloned were inoculated into a 50 ml LB for overnight preculture. The non-saturated bacterial cultures were further diluted into 1 Liter of LB medium for protein expression. The cultures were grown at 37°C with shaking until an A600 of 0.6 was reached. Protein expression was induced by the addition of IPTG to a final concentration of 1 mM and the induction was allowed to take place for an additional 4 hours. The induced bacteria were collected by centrifugation (20 min at 4000 rpm). Purification was carried out by either native or denaturing conditions depending on the solubility status of the individual proteins. For purification under native conditions, the cells were resuspended in 15 ml 50 mM NaH2PO4, 300 mM NaCl, 10 mM imidazole, 5 mM β-mercaptoethanol, and 0.1 mM PMSF at pH 8.0. The cells were lysed with a french pressure cell and centrifuged for 30 min at 4°C at 18,000 rpm. Ni-NTA superflow (Qiagen) was used in the His-tag purification of the induced protein. 1 ml 50% Ni-NTA slurry was used to pack the chromatography column. The purification of 6xHis-tagged protein was performed in this column using a 10 to 250 mM imidazole gradient. To facilitate the purification process, the columns were centrifuged for 1 min at 4°C at 800 rpm for the washing and elution steps. The purified proteins were then dialyzed against binding buffer containing 50 mM Tris, and 100 mM NaCl at pH 8.0. For purification under denaturing conditions, the cells were re-suspended in 15 ml of 100 mM NaH2PO4, 10 mM Tris-HCl, 8 M urea, pH 8.0. After loading the lysate onto the column the renaturation was performed using a linear 6 M – 1 M urea gradient in 50 mM Tris, 100 mM NaCl, 5% glycerol, 1 mM EGTA, 0.1 mM PMSF, 2 mM MgCl2, 14.4 mM β-mercaptoethanol, pH 8.0. Elution of the bound proteins was performed using a pH gradient from pH 8.0 to pH 4.5. After elution of the re-natured proteins the protein solution were dialyzed against a binding buffer containing 50 mM Tris, 100 mM NaCl, pH 8.0. An equal amount of the purified proteins (91 combinations) were incubated overnight at 4°C and the binding was analyzed by continuous non-denaturing polyacrylamide gel electrophoresis (10% PAGE) followed by Coomassie blue staining.
Binding assay – induced denaturation-renaturation
The purification of the 14 recombinant proteins were performed as described previously. An equal amount of the purified proteins (91 combinations) were dialyzed in a micro dialysis bag (100 μL of total volume used) against 8 M urea, 1 mM EGTA, 0.1 mM PMSF, 50 mM Tris, 100 mM NaCl, 14.4 mM β-mercaptoethanol, pH 8.0. The denatured proteins were then renaturated by dialyzing against a renaturation buffer containing 2 M urea, 1 mM EGTA, 0.1 mM PMSF, 50 mM Tris, 100 mM NaCl, 2 mM MgCl2, 5% glycerol, 14.4 mM β-mercaptoethanol. The protein complexes were detected by continuous non-denaturing polyacrylamide gel electrophoresis (10%) followed by Coomassie blue staining.
Binding assay – co-expression in E. coli
A mixture of 10 ng of pET-28b and pGEX-2T constructs (156 combinations) were transformed into E. coli BL21(DE3) and grown on petri plates containing 40 μg/ml kanamycin with 50 μg/ml ampicillin. Cells from 5 ml overnight cultures grown at 37°C were used to inoculate 100 ml LB medium containing 40 μg/ml kanamycin and 50 μg/ml ampicillin. Protein expression was induced at an OD600 of 0.6 with the addition of 1 mM of IPTG. After an additional 4 hours of growth at 37°C, the cells were collected by centrifugation (10 min at 4000 rpm). The induced cells were then re-suspended in 50 mM Tris (pH 8.0), 100 mM NaCl, 5 mM β-mercaptoethanol, 0.1 mM PMSF and lysis were performed with a french pressure cell. Cell lysate were centrifuged 30 min at 4°C at 25,000 g. Ni-NTA superflow (Qiagen) affinity gel was used for the His-tag purification of the complex. Elution of the proteins was performed using a 20 to 250 mM imidazole gradient. After washing extensively, the bound proteins were eluted in 500 μL aliquots. The Ni-NTA purified co-proteins and protein complexes were loaded on SDS-PAGE and stained with silver nitrate.
Binding assay in rabbit reticulocyte lysate
In order to verify if the positive in vitro interactions obtained by co-expressing both CARD protein partners in E. coli were also able to form protein complex in a mammalian system, equal amounts of the putative interacting proteins were incubated in 50 μL of Rabbit Reticulocyte Lysate (Promega) at 4°C for 1 hour.
The co-expressed proteins or protein complexes and the binding products in Rabbit Reticulocyte Lysate (Promega) were immunoprecipitated with anti-6xHis or anti-GST antibody. Protein-protein interactions were detected by SDS-PAGE, western blotting and anti-GST or anti-6xHIS.
List of abbreviations
CARD: Caspase Associated Recruitment Domain, ORF: Open Reading Frame, PAAD/DAPIN/PYRIN: Pyrin, AIM (Absent-in-melanoma), ASC, Apoptosis-associated speck-like protein containing a caspase recruitment domain CARD, and Death-Domain (DD)-like/ Domain, ApoPtosis, INterferon response/Pyrin, ORF: Open Reading Frame, CD: Circular Dichroism, APAF-1: Apoptotic Protein Activating Factor-1, ICEBERG: caspase-1 inhibitor, ASC: Apoptosis-associated Speck-like protein containing a CARD, NAC: apoptotic protein containing a nucleotide binding domain and a CARD domain, RAIDD: receptor-interacting protein (RIP)-Associated ICH-1/CED-3-homologous protein with a Death domain, DD: Death domain, DED: Death effector domain, PIDD: P53-Induced protein with a Death Domain, TUCAN: Tumor Up-regulated CARD-containing Antagonist of caspase-Nine, BCl10: cellular homolog of the equine herpesvirus-2 E10 gene containing an amino-terminal caspase recruitment domain, COP: CARD Only Protein, NOL3: NucleOLar protein 3 or Apoptosis Repressor with CARD domain (ARC), CLAN: apoptotic protein containing a CARD domain, a Leucine rich repeat domain (LRR) and a NACHT domain (initially found in the genes NAip, Ciita, Het-e and Tp1), RICK: RIP-like interacting CLARP kinase, IPTG: Isopropyl-βD-thiogalactopyranoside, LB: Luria Broth medium, PMSF: phenylmethylsulfonyl fluoride SDS-PAGE: Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis.
Authors' contributions
WS, performed immunoprecipitation, some co-expression assays and western blotting studies and wrote part of the manuscript, SY performed all native, renaturated and co-expression binding assays, BT was involved in cloning and immunoprecipitation experiments, FP provided guidance during this project and wrote the manuscript
Supplementary Material
Additional File 1
Native binding assays of CARD proteins complexes (ppt).
Click here for file
Additional File 2
Denaturation-renaturation binding assays of CARD protein complexes (ppt).
Click here for file
Additional File 3
Co-expression binding assays of CARD protein complexes (ppt).
Click here for file
Acknowledgements
This work was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC), British Columbia Advanced System Institute (BC ASI), the MSFHR/CIHR training program in bioinformatics and the Canadian Foundation for Innovation.
==== Refs
Mayer MR Dailey TA Baucom CM Supernak JL Grady MC Hawk HE Dailey HA Expression of human proteins at the Southeast Collaboratory for Structural Genomics J Struct Funct Genomics 2004 5 159 165 15263854 10.1023/B:JSFG.0000029202.77832.34
Terwilliger TC Structures and technology for biologists Nat Struct Mol Biol 2004 11 296 297 15048099 10.1038/nsmb0404-296
Finley JB Qiu SH Luan CH Luo M Structural genomics for Caenorhabditis elegans: high throughput protein expression analysis Protein Expr Purif 2004 34 49 55 14766299 10.1016/j.pep.2003.11.026
Yang JK Park MS Waldo GS Suh SW Directed evolution approach to a structural genomics project: Rv2002 from Mycobacterium tuberculosis Proc Natl Acad Sci U S A 2003 100 455 460 12524453 10.1073/pnas.0137017100
Hewitt L McDonnell JM Screening and optimizing protein production in E. coli Methods Mol Biol 2004 278 1 16 15317987
Garcia RA Pantazatos D Villarreal FJ Hydrogen/deuterium exchange mass spectrometry for investigating protein-ligand interactions Assay Drug Dev Technol 2004 2 81 91 15090213 10.1089/154065804322966342
Lanman J Prevelige PEJ High-sensitivity mass spectrometry for imaging subunit interactions: hydrogen/deuterium exchange Curr Opin Struct Biol 2004 14 181 188 15093832 10.1016/j.sbi.2004.03.006
Pantazatos D Kim JS Klock HE Stevens RC Wilson IA Lesley SA Woods VLJ Rapid refinement of crystallographic protein construct definition employing enhanced hydrogen/deuterium exchange MS Proc Natl Acad Sci U S A 2004 101 751 756 14715906 10.1073/pnas.0307204101
Alexandrov A Vignali M LaCount DJ Quartley E De Vries C De Rosa D Babulski J Mitchell SF Schoenfeld LW Fields S Hol WG Dumont ME Phizicky EM Grayhack EJ A Facile Method for High-throughput Co-expression of Protein Pairs Mol Cell Proteomics 2004 3 934 938 15240823 10.1074/mcp.T400008-MCP200
Qin H Srinivasula SM Wu G Fernandes-Alnemri T Alnemri ES Shi Y Structural basis of procaspase-9 recruitment by the apoptotic protease-activating factor 1 Nature 1999 399 549 557 10376594 10.1038/21124
Xiao T Towb P Wasserman SA Sprang SR Three-dimensional structure of a complex between the death domains of Pelle and Tube Cell 1999 99 545 555 10589682 10.1016/S0092-8674(00)81542-1
Martinon F Burns K Tschopp J The inflammasome: a molecular platform triggering activation of inflammatory caspases and processing of proIL-beta Mol Cell 2002 10 417 426 12191486 10.1016/S1097-2765(02)00599-3
Bertin J Guo Y Wang L Srinivasula SM Jacobson MD Poyet JL Merriam S Du MQ Dyer MJ Robison KE DiStefano PS Alnemri ES CARD9 is a novel caspase recruitment domain-containing protein that interacts with BCL10/CLAP and activates NF-kappa B J Biol Chem 2000 275 41082 41086 11053425 10.1074/jbc.C000726200
Chu ZL Pio F Xie Z Welsh K Krajewska M Krajewski S Godzik A Reed JC A novel enhancer of the Apaf1 apoptosome involved in cytochrome c-dependent caspase activation and apoptosis J Biol Chem 2001 276 9239 9245 11113115 10.1074/jbc.M006309200
Pathan N Marusawa H Krajewska M Matsuzawa S Kim H Okada K Torii S Kitada S Krajewski S Welsh K Pio F Godzik A Reed JC TUCAN, an antiapoptotic caspase-associated recruitment domain family protein overexpressed in cancer J Biol Chem 2001 276 32220 32229 11408476 10.1074/jbc.M100433200
Damiano JS Stehlik C Pio F Godzik A Reed JC CLAN, a novel human CED-4-like gene Genomics 2001 75 77 83 11472070 10.1006/geno.2001.6579
Geddes BJ Wang L Huang WJ Lavellee M Manji GA Brown M Jurman M Cao J Morgenstern J Merriam S Glucksmann MA DiStefano PS Bertin J Human CARD12 is a novel CED4/Apaf-1 family member that induces apoptosis Biochem Biophys Res Commun 2001 284 77 82 11374873 10.1006/bbrc.2001.4928
Razmara M Srinivasula SM Wang L Poyet JL Geddes BJ DiStefano PS Bertin J Alnemri ES CARD-8 protein, a new CARD family member that regulates caspase-1 activation and apoptosis J Biol Chem 2002 277 13952 13958 11821383 10.1074/jbc.M107811200
Lee SH Stehlik C Reed JC Cop, a caspase recruitment domain-containing protein and inhibitor of caspase-1 activation processing J Biol Chem 2001 276 34495 34500 11432859 10.1074/jbc.M101415200
Zhou P Chou J Olea RS Yuan J Wagner G Solution structure of Apaf-1 CARD and its interaction with caspase-9 CARD: a structural basis for specific adaptor/caspase interaction Proc Natl Acad Sci U S A 1999 96 11265 11270 10500165 10.1073/pnas.96.20.11265
Humke EW Shriver SK Starovasnik MA Fairbrother WJ Dixit VM ICEBERG: a novel inhibitor of interleukin-1beta generation Cell 2000 103 99 111 11051551 10.1016/S0092-8674(00)00108-2
Stehlik C Lee SH Dorfleutner A Stassinopoulos A Sagara J Reed JC Apoptosis-associated speck-like protein containing a caspase recruitment domain is a regulator of procaspase-1 activation J Immunol 2003 171 6154 6163 14634131
Chou JJ Matsuo H Duan H Wagner G Solution structure of the RAIDD CARD and model for CARD/CARD interaction in caspase-2 and caspase-9 recruitment Cell 1998 94 171 180 9695946 10.1016/S0092-8674(00)81417-8
Tinel A Tschopp J The PIDDosome, a protein complex implicated in activation of caspase-2 in response to genotoxic stress Science 2004 304 843 846 15073321 10.1126/science.1095432
Vincentelli R Canaan S Campanacci V Valencia C Maurin D Frassinetti F Scappucini-Calvo L Bourne Y Cambillau C Bignon C High-throughput automated refolding screening of inclusion bodies Protein Sci 2004 13 2782 2792 15388864 10.1110/ps.04806004
Tresaugues L Collinet B Minard P Henckes G Aufrere R Blondeau K Liger D Zhou CZ Janin J Van Tilbeurgh H Quevillon-Cheruel S Refolding strategies from inclusion bodies in a structural genomics project J Struct Funct Genomics 2004 5 195 204 15263835 10.1023/B:JSFG.0000029017.46332.e3
| 15904526 | PMC1156946 | CC BY | 2021-01-04 16:37:15 | no | Proteome Sci. 2005 May 18; 3:3 | utf-8 | Proteome Sci | 2,005 | 10.1186/1477-5956-3-3 | oa_comm |
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-181588820610.1186/1477-7827-3-18ReviewAquaporins in development – a review Liu Huishu [email protected] E Marelyn [email protected] Guangzhou Obstetric and Gynecology Institute, Second Municipal Hospital of Guangzhou, Guangzhou Medical College, Guangzhou, PR China2 Department of Physiology, Monash University, Clayton, Victoria, 3800, Australia2005 11 5 2005 3 18 18 24 3 2005 11 5 2005 Copyright © 2005 Liu and Wintour; licensee BioMed Central Ltd.2005Liu and Wintour; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Water homeostasis during fetal development is of crucial physiologic importance. It depends upon maternal fetal fluid exchange at the placenta and fetal membranes, and some exchange between fetus and amniotic fluid can occur across the skin before full keratinization. Lungs only grow and develop normally with fluid secretion, and there is evidence that cerebral spinal fluid formation is important in normal brain development. The aquaporins are a growing family of molecular water channels, the ontogeny of which is starting to be explored. One question that is of particular importance is how well does the rodent (mouse, rat) fetus serve as a model for long-gestation mammals such as sheep and human? This is particularly important for organs such as the lung and the kidney, whose development before birth is very much less in rodents than in the long-gestation species.
==== Body
Introduction
There are, at present, eleven known members of the mammalian aquaporin gene family, which encode proteins which function as membrane channels, for water alone (AQP0,1,2,4,5,8,10), or for water plus small molecules, mostly glycerol and urea (AQP 3, 7, 9), or nitrate (AQP 6) [1-5]. In some cases the aquaporin is constitutively present in the cell membrane (e.g. AQP1,3 in red cell membrane, AQP1 kidney). However, in other cases the aquaporin resides in intracellular vesicles, and is trafficked to the membrane upon appropriate stimulation e.g. AQP 2 in collecting duct cells, after vasopressin exposure [6]; AQP1 in cholangiocytes with secretin stimulation [7]; AQP 8 in hepatocytes, after glucagon treatment [8,9]; aquaporin 5 in rat parotid, with muscarinic stimulation [10]. These aquaporins subserve the rapid transport of fluid across epithelial and endothelial cells, but are also found in other tissue types, such as muscle and nerve cells. In general the water channels are 'open' but there is some evidence that 'closure' can be induced by a specific treatment.
During development there are some unique fluid compartments (amniotic, allantoic fluids, lung liquid) and the functions of some organs, such as the kidney, differ from the function in the adult, as discussed below. Although some insights into the developmental roles of aquaporins might be obtained from the study of mice with deletions of various aquaporin genes, this is complicated by the facts that either much of normal organ development occurs postnatally in the rodent, rather than prenatally in the human (e.g.kidney), chick (brain) or the ontogeny of aquaporins differs significantly in the rodent organ (e.g.lung) from that in long-gestation species such the sheep [11]. In addition, the fetus contains a higher percentage of water than does the adult, and organs such as the brain are more vulnerable to excess water loss which might occur in the premature neonate, due either to immaturity of the skin permeability barrier, or to immaturity of the water-retaining functions of the kidney. The role of aquaporins in fluid balance during fetal development is beginning to be explored.
Placenta and fetal fluid compartments
Amniotic fluid surrounds the developing fetus and is essential for normal morphological development. Inputs into amniotic fluid include the dilute fetal urine and the isotonic lung liquid, and pathways of exit of fluid include fetal swallowing, and transmembrane fluxes [12,13]. Thus abnormalities of amniotic fluid volume (oligo- and polyhydramnios) can result from abnormalities in fetal renal function, and oligohydramnios can be corrected, to some extent, by increase in maternal hydration [14,15]. Under normal circumstances the fetal fluid osmolality follows that of the mother, and fluid exchange occurs across the placenta, as well as across the amnion/chorion [16].
Before implantation the conceptus develops into a blastocyst, composed of the inner cell mass, and a fluid filled cavity surrounded by trophoectoderm epithelium. In the mouse aquaporins 3, 8, and 9 have been found to be expressed at this time, AQP3 and AQP8 being predominantly in the basolateral membranes of the trophoectoderm, and AQP9 in the apical membrane [17]. The trophoectoderm gives rise to the placenta and chorion; aquaporins 1, 3, 8 and 9 are water channel genes previously reported to be in the placenta and/or chorion of the human and sheep [18-21]. AQP1 has also been reported to be in the chick chorioallantoic membrane [22]. AQP1 is in the vasculature and AQP3 and 9 are in the apical membranes of human and ovine term placenta and chorion. The polarity of the AQP 8 has not yet been determined [18-20]. Recently we reported that AQP8 mRNA was also found in the ovine placenta [23].
From 45 d gestation (term is ~150 days), AQP3, functioning both as a water and urea channel, and expressed in the trophoblast epithelial cells, is the major AQP, which increases throughout gestation, and is quantitatively the most highly expressed AQP gene in the ovine placenta. The permeability of the ovine placenta to urea increases markedly after ~100 days of gestation, coordinately with a sharp increase of AQP3 expression in the placenta at this time.
Similarly, AQP8, which is expressed in the trophoblast epithelial cells and membrane epithelial cell [24], is also present at significant levels from 45 d gestation.
In sheep, the placenta ceases growth close to mid-gestation, despite the dramatic increase in the fetal weight during the last half of gestation [25]. To maintain fetal growth, there is a requirement for increased fluid transfer to the conceptus. The presence of substantial expression of water channel proteins in the placenta correlates well with the placental transfer of fluid. It was not possible to compare expression at the protein level as large quantities of AQP1 and AQP3 protein, in the maternal red cell membranes present in the haemophagous zone of the ovine placenta [26]. Thus comparison at the mRNA level is the only feasible one that can be made.
Kidney function in the fetus
The fetal metanephric kidney produces a relatively large volume of dilute urine, essential for the maintenance of amniotic and (in some species) allantoic fluid volumes. In the most common animal model (sheep) used for the study of fetal renal function it has been shown that the volume of urine production is 0.3 l/kg/d compared with 0.02 l/kg/d in the adult sheep. This occurs in spite of a glomerular filtration rate which is approximately one third of adult values, and is due to both a decrease in total sodium reabsorption (95 % in the fetus vs 99% in the adult) and to absence of significant concentration of the urine. In the unstressed ovine fetus the urine osmolality is always less than 200 mosmoles.kg water, and may be as low as 60 [27].
Aquaporins in development – kidney
In the adult kidney the bulk of the filtrate (81%) is reabsorbed in the proximal tubule and descending limb of the loop of Henle, where AQP 1 is expressed. AQP1 is also expressed in the nonfenestrated descending vasa recta which are thought to be important for the establishment of the hypertonic environment of the medulla. In the mouse with the AQP1 gene deleted there is a lowered capacity to maximally concentrate urine [28]. However, the major concentration of urine depends on the presence of aquaporin 2 in the apical membranes of the principal cells of the collecting duct. This water channel protein resides in sub -membraneous vesicles in the absence of action of circulating vasopressin. Under the stimulus of increased vasopressin second messenger systems are activated which result in the phosphorylation of the vesicular AQP2 and transport and insertion into the apical membranes. Without this water channel it is impossible to reabsorb water in the medulla, even when an adequate osmotic gradient exists [2]. In many situations in which polyuria/concentrating defect occurs (potassium deficiency, lithium levels greater than 0.3 mmol/l, hypercalcemia, low protein diet among others) it can be linked to low levels of AQP2 [2]. The water absorbed via AQP2 in the apical membrane leaves the cell via aquaporins 3 and 4 which are constitutively expressed in the basolateral membranes of these cells [2]. In mice lacking expression of the AQP1 gene there is polyuria, and failure to be able to concentrate urine normally [28], and a similar urinary concentrating defect is seen in the rare humans who lack AQP1 [29]. A milder urinary concentrating defect is seen in transgenic mice lacking AQP4 [30]. This maybe because AQP3 is colocalised with AQP4 on the basolateral membrames of collecting duct principal cells, but when AQP3 is deleted a poyuria with a severe concentrating defect occurs [31]. AQP3 levels are regulated to some extent by vasopressin, as are those of AQP2, but are also regulated by aldosterone and the cystic fibrosis transmembrane conducting factor (CFTR) [2,32-34].
Metanephric kidney development varies in different species, being complete before birth in human and sheep, but not until substantially after birth in pigs, mice and rats. The ontogeny of some renal aquaporins has been examined in rats, sheep and humans. In the rat there is very little mRNA for AQP1 detected by Northern blotting or RNase protection, in the kidney, until a few days before birth [35,36]. However, there is some protein detected, by immunohistochemistry in the capillaries at the nephrogenic zone-medullary border by day 16. From day 17 the arcuate arteries are labeled, and, indeed the descending vasa recta are strongly labeled as they develop fully until 21 days post partum [37]. In contrast, in the sheep and human kidneys, AQP 1 mRNA, and protein are detected before mid-gestation (12 /40 weeks, human; 41/150 days, sheep) though the levels are just below 50% of adult levels even at term [38,39]. Levels of expression can be increased by both glucocorticoid and angiotensin II treatment of the fetus, both probably due to maturation of the kidney and longer proximal tubules which develop with treatment [39]. Adult levels are achieved after 15 months in the human, or 6 weeks in the sheep.
Aquaporin 2 (AQP2) is low at birth in the rat, but plateaus by 4 weeks post-partum [40]. Later studies showed it was present by Day 18 of fetal life and started increasing by day 3 post-natally [41]. In the sheep, at the beginning of the last third of gestation (100/150 d) the level of AQP2 mRNA is 17% of the adult, and near term it is still only ~40% of the adult [42]. This correlates with reduced sensitivity of the fetal kidney to infused arginine vasopressin – at 100 days the plasma AVP concentration has to be raised to 16 pg/ml to achieve negative free water clearance, whereas close to term a level of 2 pg.ml is effective [43]. This is still a much higher level than required in the adult sheep, and so the fetal kidney resembles that of a subject with nephrogenic diabetes insidipus, due to inadequate expression of AQP2. The human fetal kidney also has a low level of AQP2 during the last half of gestation, and premature neonates produce dilute urine for many weeks [38,44]. AQP2 protein does appear in the urine [45], and there is a low level in the urine of premature neonates [46]. However the concentration of AQP2 protein in the urine of premature neonates did not correlate well with changes in urine osmolality, suggesting that it did not serve as a good marker of AVP function in the human premature neonate [47]. Fetal renal AQP2 levels can be increased by angiotensin II infusion, which is a real up-regulation of gene expression, and similar to the up-regulation of vasopressin V2 receptor seen with angiotensin II infusion in adult rats [48].
There has been one study of AQP3, in fetal kidneys, suggesting it is there by day 18 in the rat [41]. The level of AQP4 protein labeling is very weak in the rat kidney 3 days after birth [36].
The low level of AQP2 expression, however, seems to be the major factor in allowing the production of a large volume of hypotonic urine to be formed, and this is essential for the maintenance of adequate volumes of amniotic fluid.
Lung liquid
During fetal life, the future airways of the lung are filled with a liquid that plays a crucial role in the growth and development of the lungs by maintaining them in an expanded state. Lung liquid is secreted across the pulmonary epithelium into the lung lumen due to the osmotic gradient established by the net movement of Cl- in the same direction. It is not known exactly when lung liquid secretion begins, but fluid is present by mid-gestation in fetal sheep and is secreted at 2–4 ml/kg/h between 120 days of gestation and term (~150 d). Fetal lung liquid exits the lungs via the trachea, whereby it is either swallowed (approximately 50%) or passes directly into the amniotic sac, where it contributes to amniotic fluid volume [49].
If the fetal trachea is obstructed, which prevents the outward flow of lung liquid, the fetal lung expands with accumulated liquid. This is a potent stimulus for fetal lung growth and also greatly reduces the proportion of type-II alveolar epithelial cells (AECs). Lung liquid drainage on the other hand, deflates the lung, causes lung growth to cease, but increases the proportion of type II AECs, possibly via type-I to type-II cell differentiation [50]. As a result it is now widely recognized that the degree to which the fetal lungs are expanded by lung liquid, determines the growth and structural development of the lung, as well as the differentiated state of type-I and type-II AECs [49]. Despite the importance that lung liquid plays in the development of the lung, the factors controlling the movement of liquid across the pulmonary epithelium have not been fully explored. Furthermore, the effective clearance of lung liquid at birth is vital to allow the entry of air into the lungs with the onset of respiratory gas exchange. This process is largely dependent on the capability of the epithelium to reabsorb large quantities of water.
Aquaporins in development – lung
At least four AQPs (AQP 1, 3, 4 and 5) are expressed in the lungs of various species, including humans, rats, mice and rabbits, although some discrepancies exist in the specific sites of distribution of these proteins. (Table 1 near here) In all species described so far (human, rat, mouse), AQP1 is expressed in the apical and basolateral membrane of the microvascular endothelium and decreased pulmonary vascular permeability has been shown in AQP1-null humans [3]. AQP3 is expressed in the basolateral membrane of basal cells of the tracheal epithelium and in submucosal gland cell membranes in rodents, but is also found in bronchioles (apical membrane) and type-II alveolar epithelial cells of adult humans [51]. AQP4 is present in the basolateral membrane of columnar cells in bronchi and trachea of rats but is also found in type-I AECs in humans. AQP5 is expressed in the apical membrane of type-I AECs and the apical plasma membranes of the secretory epithelium in upper airway and salivary glands [3]; it has also been detected in type-II AECs in mice [52]. These data are summarized in Table 2.
Table 1 Species variations in Aquaporin Distribution in Lung
Species Sheep Human Rat Mouse
Bronchus AQP1,3,4,5 AQP1,3,4,5 AQP1,3,4,5 AQP1,3,4,5
Bronchioles AQP1,3,4 AQP1,3 ? ?
Alveoli AQP1,5 AQP1,3,4,5 AQP1,3 AQP1,3
Table 2 Aquaporins in lung cell types
Bronchus
Superficial Epithelium AQP5 (Apical), AQP4 (Basolateral)
Basal Cells AQP3
Submucosal Glands AQP5 (Apical), AQP3,4 (Basolateral)
Bronchioles
Pseudostratified AQP3 (Apical), AQP4 (Basolateral)
Alveolar Cells
Type I AQP5 (Apical), AQP4 (Human only--?)
Type II AQP5 (Mouse only, apical)
AQP3 (Human only, basolateral)
Ontogeny of lung AQPs
In mice very low levels of AQP5 mRNA were detected before birth [53,54]. The ontogeny of the AQPs has also been described throughout development in rats, but only AQP1 and a small amount of AQP4 were detected before birth [55-58]. Furthermore, little is known of the physiological factors controlling AQP1 mRNA expression before birth, although its expression (and protein levels) is increased in the lungs of fetal and neonatal rats following treatment with synthetic glucocorticoids [55,58]. In one study [58], but not in another [55], AQP4 was increased by corticosteroids. In the same study [58], β-adrenergic agents also increased AQP4. Although AQP5 protein was almost undetectable in lung tissue homogenates at E21 and PN1, a strong signal was detected at PN2 [55], indicating that the accumulation of AQP5 protein in the rat lung is predominantly postnatal. Indeed, AQP5 protein levels in lung tissue increased twenty-fold to PN14 and then increased a further ten-fold from PN14 to adult. In contrast to AQP1, AQP5 is not influenced by corticosteroids in rats, which is consistent with the finding that AQP5 protein predominantly accumulates in the lung postnatally. Similarly, AQP3 protein levels were undetectable in fetal lung tissue and then were only detected in the trachea of postnatal animals well after the time of birth. AQP4 protein seemed to be present transiently at PN2 in peripheral lung membranes and only appeared by PN12 in the trachea of rats
In a recent study we have shown that the mRNAs for at least four AQPs (1, 3, 4 and 5), as well as their respective proteins, are present in the ovine fetal lung well before birth [11]. For AQP1 and AQP5, the level of mRNA expression in the fetal lung exceeded that of the adult lung. Furthermore, we have shown that cortisol infusions significantly up-regulated the expression of AQPs 1 and 5, whereas increases in fetal lung expansion, induced by tracheal obstruction (TO), significantly decreased AQP5 mRNA levels in fetal lung tissue. Although AQP5 protein levels did not appear to decrease with TO, measurable changes in AQP5 levels in whole lung tissue is likely to be complicated by the localisation of this protein to multiple cell types within the lung. These findings indicate that factors known to regulate fetal lung growth and maturation as well as fluid secretion, also regulate the expression of AQPs 1 and 5. This suggests that there are physiological roles for some lung aquaporins before birth.
In conclusion, we have shown that the lung of a long-gestation species, such as sheep, expresses both the mRNA and protein of the four typical lung AQPs, beginning well before the expected time of birth. Furthermore, we found that the expression of some, particularly AQP5, is altered by factors known to regulate fetal lung growth and development and parallel changes in fetal lung liquid secretion rates in different animal models. Our findings suggest that gene knock-out studies in mice, in which there is little lung expression of AQPs in fetal life, might not give a realistic picture of the role of AQPs during fetal life in long-gestation species. We predict that these AQPs are also expressed well before birth in the human fetal lung and are also differentially regulated by factors known to influence fetal lung development. As lung liquid is secreted, at least in part, into amniotic fluid, the lung aquaporins are also then implicated in amniotic fluid regulation.
Skin
The skin of the adult 70 kg man normally contains about 7 l of fluid, about 50% of which is interstitial [59]. The fluid is stored in the dermis associated with hyaluronic acid, glycosaminoglycans and proteoglycans, and helps to determine the turgor, distensibility and elasticity of the skin. The major barrier to water loss from the skin is the superficial stratum corneum – flattened dead corneocytes [60]. Below this are the keratinocytes, which express the gene for aquaporin 3, particularly in the basal and intermediate layers [61-63]. Aquaporin 3 is a membrane protein which increases the permeability to water, urea and glycerol. When the gene is deleted in the mouse the skin has decreased hydration but grossly normal morphology [62]. The reduction in skin elasticity, as well as the delay in recovery of barrier function after tape stripping, were thought to be related to the deficiency in glycerol transport which occurred in the AQP3 deficient mice [64]. This was further supported by the reversal of these deficits by glycerol replacement [65].
Aquaporins in development – skin
In the human fetus there is a double layer of epidermal cells by 4 weeks; the stratum corneum begins to develop by 24 weeks, and is generally well developed by 34 weeks. [60]. Barrier function, which is conferred by the stratum corneum, of cornified cells and extracellular lipid, can be measured by transepidermal water loss (TEWL), and generally forms late in gestation in mice, rats, rabbits and humans [66,67]. Amniotic fluid, particularly early in pregnancy, is very similar in composition to fetal extracellular fluid, and it is quite likely that here is fairly free exchange across the fetal skin, particularly in the first half of gestation [68]. Even in species such as the sheep, which develop substantial wool covering in the last third of gestation, there is substantial exchange of fluid and electrolyte across the skin until relatively late in development [69]. There is also substantial expression of AQP3 in mid-gestation ovine fetal skin. Preterm infants are at risk of dehydration because of very large TEWL [70]. In fetal rats the TEWL is high at day E18, and there are higher levels of AQP3 mRNA in the fetal than in the adult skin [71].
Aquaporins in the heart – changes with intrauterine growth retardation (IUGR)
Aquaporin 1 mRNA was found in rat heart [72,73]. Most of the AQP1 expression was thought to be in the blood vessels, although the there was a substantial amount in a sub-sarcolemnal caveolar membrane in the rat heart, and changes in the osmotic environment caused reversible changes in the membrane localization of AQP1 [74]. Recently it was found that the human heart contained both AQP1 and AQP4, but not AQP8 [75]. AQP1 co-localised with vinculum, a t-tubule component, and caveolin 3, whereas AQP4 was found in the nuclear membrane of human cardiac myocytes.
Caveolin-3 is a marker for the caveolae – the specialised areas of cell membrane in which a number of receptors cluster [76]. Some of these receptors are known to play a role in the proliferation of cardiac myocytes in the embryonic and early post-natal life [77-80].
Based on studies in isolated rabbit hearts, it was concluded that water permeabilitity values were much lower than expected if a functioning aquaporin were present [81]. In a more recent study of the osmotic transient responses of isolated adult rabbit hearts [82] it was estimated that 28% of the transcapillary water flux going to form lymph was through aquaporin channels in the capillaries, but they did not make any histological studies on the cardiocytes. It would have been very interesting to have had immunohistochemistry for AQP1, at least, on these hearts.
During development AQP1 was found in the endocardium of the sheep fetal heart at a very early stage [83]. Later in gestation one report suggested that total cardiac AQP1 levels reflected predominantly vascular sites, and that the total amount could be increased by fetal anemia [84].
Using RNase protection assay only AQP1 (but not AQPs 2,3,4,5) was detected in rat heart [72], however with RT-PCR some AQP8 mRNA was detected in mouse heart [85]. AQP 1 was reported to be present in fetal rat hearts from day E14 with lower level present in the myocardium than in the endothelial cushions, primordial valves, and septa [35]. Cardiac expression of AQP1 decreased, but did not disappear, after birth [35].
In a recent study we showed that the small hearts of late gestation growth-retarded ovine fetuses had significantly reduced expression of AQPs 1,3,4 but not AQP8 [86]. It was not possible to ascertain the different contributions of cardiac muscle and blood vessels to this reduced expression. In the fetal sheep heart at mid-gestation, all the myocytes are uninucleated and can divide, but by 135 days more than 50% of the myocytes are binucleated, and terminally differentiated [87]. When growth retardation occurs in the fetal heart, we postulated that it might have occurred by 'slowing down' of cell division resulting in a greater proportion of uninucleated cells in late gestation. In order to see whether AQP1 can be a marker of cardiac myocyte differentiation, we measured the AQP1 mRNA concentration in hearts from fetuses in which cardiac myocyte counts had been performed previously. Our results show that the level of AQP1 mRNA expression did not change significantly at any point during gestation, suggesting that it could not be used as a marker of cardiac myocyte differentiation. Thus the heart is different from vascular smooth muscle.
In conclusion, we demonstrated that the AQP1/3/4/8 are present in the late gestational fetal heart. The low-dose dexamethasone treatment, administered early in gestation, down regulated the expression of AQP1/3/4 in the late gestation fetal heart. In most studies of experimentally induced fetal growth retardation some organs, eg the brain and adrenal gland are 'spared', but others, such as the heart, are reduced in size in proportion to the overall decrease in body size [88]. There are a number of genes which have been implicated in cardiac myocyte growth, including mineralocorticoid and glucocorticoid receptors, angiotensin II receptors, and local cardiac angiotensinogen [89-96]. However, the mRNA for none of these was affected in the hearts of the IUGR fetuses.
There is evidence in the literature suggesting that fetuses suffering from severe intrauterine growth retardation (IUGR) show a progressive impairment of cardiac function, as demonstrated by reduced peak velocities at outflow tracts, decreased cardiac output and abnormal venous flow patterns [95-98]. Furthermore, in growth-retarded human fetuses the ventricular ejection force was equally decreased in both ventricles [95]. Studies in the adult offspring of rats subjected to prenatal protein restriction, which caused IUGR, demonstrated higher incidence of cardiac arrhythmias and raised diastolic blood pressure [97].
The exact function for AQP1 in cardiac muscle is unknown. As it is a pure water channel one would suspect that it could regulate the rate at which cells might swell in osmotic stress, such as encountered in myocardial ischemia [98]. Such osmotic swelling is predicted to shorten the action potential, thus modulating the excitability of the heart. It is known that cell swelling inhibits the action of some antiarrhythmic drugs [98]. AQP 4 is well established as a component of skeletal fast-twitch fibres [99] and the level of AQP4 is decreased by muscle denervation [100]. In mice which are dystrophic due to dystrophin gene knock-out (mdx mice) AQP4 mRNA levels remain the same as controls, but the protein levels decrease by 90% [101]. However, in patients with Duchenne muscular dystrophy both the mRNA and protein of AQP4 are reduced in myofibers [102]. Taken together it is attractive to propose that AQP's play a role in the cardiac myocyte contraction allowing therefore normal cardiac function.
Brain-central, nervous system, eye, ear-fluid compartments
In the adult brain fluid balance is critical, as the inflexible bony skull does not permit big variations in total brain volume without risking severe damage. The extracellular fluid of the brain is specialized as cerebrospinal fluid, with a composition different from that of normal extracellular fluid, as a result of the development of the 'Blood-brain barrier'. There is now increasing evidence that cerebrospinal fluid plays an important part in the correct development of the brain [103,104]. Specialised fluid compartments are also vital to the normal functioning of the sensory organs – the eye and the ear [105,106]. In the eye fluid movements are important for the regulation of intraocular pressure, the maintenance of transparency of the lens, and retinal signal transduction [106]. The fluids of the inner ear, endolymph and perilymph, have at least two roles – to transduce the signal to the cochlear and vestibular hair cells, and to participate in the ionic exchanges between fluid and hair cells [106]. The endolymph is a potassium-rich extracellular fluid, whereas the perilymph has a composition closer to that of extracellular fluid [107]. It is well-known that vestibular functions can be altered by a number of peptide e.g arginine vasopressin, and steroid hormones [108-110], which act by changing composition, and maybe the volume, of the endolymph.
A number of aquaporins have been found in the central nervous system – AQPs 1,4,5,9 [111,112]. AQP1 is found on the apical membrane of the epithelial cells of the choroid plexus. AQP 4,5, and 9 are found on glia/astrocytes particularly in the region of subpial vessels and near the ventricles. Of these it seems that AQP4 provides the principal route for water transport in astrocytes [113]. Glial cells are indispensable for regulating ionic homeostasis, particularly in aspirating the excess extracellular potassium which occurs after neural excitation [107]. It is of interest that in the specialized glial Muller cells of the eye, there is a close correlation between concentrations of the potassium channel, Kir4.1 and AQP4 levels [114], and retinal function is mildly impaired in mice lacking AQP4 [115]. The absence of AQP4, in the brain, paradoxically, in the genetically-engineered 'knock-out' mouse, reduces the swelling seen with hyponatremia [116]. The distribution of AQP4 protein is disrupted in the dystrophin-deficient mdx mouse, in which a 60% reduction occurs in the amount of AQP4 in the perivascular glial processes, which are swollen and contain debris [101,104]. In these mice the there is a marked reduction in the amount of AQP4 in the astroglial feet surrounding capillaries, and at the glia-limitans, and a significant delay in the in the development of brain edema induced by systemic hyponatremia [117]. The protein, alpha syntrophin, is associated with the dystrophin, and also important for the anchoring of the AQP4 in the cell membrane [118]. In mice lacking the alpha-syntrophin gene the there is also a marked loss of AQP4 from perivascular and subpial membranes, but no decrease in other membrane domains, and brain edema was attenuated when transient ischemia was induced [119]. All of this evidence suggests that any inhibitor of AQP4 expression may have therapeutic benefits in the treatment of brain edema [111,112].
The ontogeny of AQP4 in the cerebellum coincides with the development of the blood brain barrier in rat and chick. [120,121]. In the rat brain there is no AQP4 before birth [122] and only 2% of the adult level one week after birth. The level doubles in the next week, and reaches 63% of adult levels by nine weeks. In contrast, the chick brain, has a much better level of AQP4 at birth and a more mature blood-brain barrier [121]. This has not yet been studied in the human, but one would expect that the very premature baby would have little barrier protection.
In the ear of the adult rat mRNA for aquaporins 1,2,3,4,5,6 have been found [109], whereas AQP7 and AQP9 were also detected in the adult mouse, but at relatively low levels [122]. Aqp1 is strongly expressed in the non-epithelial stria vascularis [123] and can be up-regulated, in a dose-dependent fashion, by intra-tympanic injections of dexamethasone [109]. AQP1 was detected at the earliest day studied, E14, in mice but in much lower concentrations than those found in the adult ear [122].
AQP2 mRNA, at 10% of the levels found in kidney, is found in rat and mouse ear [124]. It is in structures bordering the endolymph – Reissner's Membrane, Organ of Corti, sulcus cells, and spiral limbus. Treatment of rats with arginine vasopressin caused a doubling of AQP2 mRNA in the cochlea and endolymphatic sac [125,126], and the authors suggested that overexpression of AQP2 might be involved in the formation of endolymphatic hydrops. During development of the ear in the mouse AQP2 was expressed diffusely in the early otocyst at embryonic days 12,13 but the expression became more restricted by days 15–18 [127].
Quantitatively the most important aquaporin expressed in the ear is AQP4, and it is expressed in Hensen's cells and inner sulcus cells and Claudius cells, which are all supporting cells of the Organ of Corti [128]. In the vestibular end organs it was in the cristae and maculae. It also occurred in the central part of the cochlear and vestibular nerves. In mice lacking AQP4 expression there is a moderate impairment of hearing [129], but no conduction abnormality was detected in neural signals [130]. AQP4 was detected by E14 in the developing mouse ear, and the level was increased ~100 fold during after birth and continued to increase through post-natal day 15 and even further in the adult [122].
AQP3 was found by one group [122] in the spiral ligament of the mouse cochlea, near where the basilar membrane anchors, and in cells bordering the inner spiral tunnel. In the vestibular system it was in sub-epithelial fibrocytes in the saccule, but not in the utricle. There was a moderate increase in AQP3 from day E14 to adult.
All these results in rodents are tantalizing, and it will be very interesting to see the ontogeny of brain and sensory organ aquaporins in the primate/human. It is expected that significant expression of these water channels will be seen well before birth, as is the case for the lung in long-gestation species [11].
Conclusion
Much information on the role of various members of the mammalian aquaporin family of water channels has been gained in the relatively short time since Peter Agre and his colleagues described the Channel-forming integral membrane protein of the red blood cell of 28 kD (CHIP28), [1], and justifiably earned the 2003 Nobel Prize for Chemistry. Some exciting new studies are suggesting that AQP1 may have roles hitherto unsuspected – evidence has been obtained supporting a role for AQP1 in angiogenesis, particularly in wound healing, organ regeneration and possibly in tumour spread [131]. The limited information that exists on the ontogeny of these proteins in various organs and tissues suggests that there are many more important findings to be made on their roles in the development of the embryo and fetus.
==== Refs
King LS Kozono D Agre P From structure to disease: the evolving tale of aquaporin biology. Nature Rev: Molecular, Cell Biol 2004 5 687 698 12096044 10.1038/nrm1469
Nielsen S Frokiaer J Marples D Kwon TH Agre P Knepper MA Aquaporins in the kidney: from molecules to medicine Physiol Rev 2002 82 205 244 11773613
King LS Yasui M Aquaporins and disease: lessons from mice to humans Trends Endocrinol Metab 2002 13 355 360 12217493 10.1016/S1043-2760(02)00665-3
Verkman AS Physiological importance of aquaporin water channels Ann Med 2002 34 192 200 12173689
Ikeda M Beitz E Kozono D Guggino WB Agre P Yasui M Characterization of aquaporin-6 as a nitrate channel in mammalian cells. Requirement of pore-lining residue threonine 63 J Biol Chem 2002 277 39873 39879 12177001 10.1074/jbc.M207008200
Knepper MA Inoue T Regulation of aquaporin-2 water channel trafficking by vasopressin Curr Opin Cell Biol 1997 9 560 564 9261056 10.1016/S0955-0674(97)80034-8
Marinelli RA Pham L Agre P LaRusso NF Secretin promotes osmotic water transport in rat cholangiocytes by increasing aquaporin-1 water channels in plasma membrane. Evidence for a secretin-induced vesicular translocation of aquaporin-1 J Biol Chem 1997 272 12984 12988 9148905 10.1074/jbc.272.20.12984
Garcia F Kierbel A Larocca MC Gradilone SA Splinter P LaRusso NF Marinelli RA The water channel aquaporin-8 is mainly intracellular in rat hepatocytes, and its plasma membrane insertion is stimulated by cyclic AMP J Biol Chem 2001 276 12147 12152 11278499 10.1074/jbc.M009403200
Gradilone SA Garcia F Huebert RC Tietz PS Larocca MC Kierbel A Carreras FI Larusso NF Marinelli RA Glucagon induces the plasma membrane insertion of functional aquaporin-8 water channels in isolated rat hepatocytes Hepatology 2003 37 1435 1441 12774023 10.1053/jhep.2003.50241
Ishikawa Y Skowronski MT Inoue N Ishida H alpha(1)-adrenoceptor-induced trafficking of aquaporin-5 to the apical plasma membrane of rat parotid cells Biochem Biophys Res Commun 1999 265 94 100 10548496 10.1006/bbrc.1999.1630
Liu H Hooper SB Armugam A Dawson N Ferraro T Jeyaseelan K Thiel A Koukoulas I Wintour EM Aquaporin gene expression and regulation in the ovine fetal lung J Physiol 2003 551 503 514 12819302 10.1113/jphysiol.2003.044875
Ross MG Brace RA National Institute of Child Health and Development Conference summary: amniotic fluid biology--basic and clinical aspects J Matern Fetal Med 2001 10 2 19 11332413
Daneshmand SS Cheung CY Brace RA Regulation of amniotic fluid volume by intramembranous absorption in sheep: role of passive permeability and vascular endothelial growth factor Am J Obstet Gynecol 2003 188 786 793 12634658 10.1067/mob.2003.160
Magann EF Doherty DA Chauhan SP Barrilleaux SP Verity LA Martin JNJ Effect of maternal hydration on amniotic fluid volume Obstet Gynecol 2003 101 1261 1265 12798534 10.1016/S0029-7844(03)00344-2
Malhotra B Deka D Duration of the increase in amniotic fluid index (AFI) after acute maternal hydration Arch Gynecol Obstet 2003
Wintour EM Shandley L Effects of fetal fluid balance on amniotic fluid volume Semin Perinatol 1993 17 158 172 8378800
Barcroft LC Offenberg H Thomsen P Watson AJ Aquaporin proteins in murine trophectoderm mediate transepithelial water movements during cavitation Dev Biol 2003 256 342 354 12679107 10.1016/S0012-1606(02)00127-6
Johnston H Koukoulas I Jeyaseelan K Armugam A Earnest L Baird R Dawson N Ferraro T Wintour EM Ontogeny of aquaporins 1 and 3 in ovine placenta and fetal membranes Placenta 2000 21 88 99 10692256 10.1053/plac.1999.0445
Damiano A Zotta E Goldstein J Reisin I Ibarra C Water channel proteins AQP3 and AQP9 are present in syncytiotrophoblast of human term placenta Placenta 2001 22 776 781 11597198 10.1053/plac.2001.0717
Wang S Kallichanda N Song W Ramirez BA Ross MG Expression of aquaporin-8 in human placenta and chorioamniotic membranes: evidence of molecular mechanism for intramembranous amniotic fluid resorption Am J Obstet Gynecol 2001 185 1226 1231 11717661 10.1067/mob.2001.117971
Mann SE Ricke EA Yang BA Verkman AS Taylor RN Expression and localization of aquaporin 1 and 3 in human fetal membranes Am J Obstet Gynecol 2002 187 902 907 12388974 10.1067/mob.2002.127168
Ribatti D Frigeri A Nico B Nicchia GP De Giorgis M Roncali L Svelto M Aquaporin-1 expression in the chick embryo chorioallantoic membrane Anat Rec 2002 268 85 89 12221713 10.1002/ar.10123
Liu H Koukoulas I Ross MC Wang S Wintour EM Quantitative comparison of placental expression of three aquaporin genes. Placenta 2003 In Press
Wang S Su Y Ramirez BA Ross MC Cloning and expression of aquaporin 8 water channel in ovine and human chorioamniotic membranes: molecular mechanism of intramembraneous pathway for amniotic fluid reabsorption [abstract]. J Soc Gyneol Investig 2000 7 183A
Reynolds LP Redmer DA Angiogenesis in the placenta Biol Reprod 2001 64 1033 1040 11259247
Baird R Wintour EM Placentae with haemophagous zones and water channel proteins; a cautionary tale Placenta 2000 21 587 588 10940210 10.1053/plac.1999.0502
Wintour EM Moritz KM Comparative aspects of fetal renal development Equine Vet J Suppl 1997 51 58 9355802
Vallon V Verkman AS Schnermann J Luminal hypotonicity in proximal tubules of aquaporin-1-knockout mice Am J Physiol Renal Physiol 2000 278 F1030 3 10836992
King LS Choi M Fernandez PC Cartron JP Agre P Defective urinary-concentrating ability due to a complete deficiency of aquaporin-1 N Engl J Med 2001 345 175 179 11463012 10.1056/NEJM200107193450304
Chou CL Ma T Yang B Knepper MA Verkman AS Fourfold reduction of water permeability in inner medullary collecting duct of aquaporin-4 knockout mice Am J Physiol 1998 274 C549 54 9486146
Ma T Song Y Yang B Gillespie A Carlson EJ Epstein CJ Verkman AS Nephrogenic diabetes insipidus in mice lacking aquaporin-3 water channels Proc Natl Acad Sci U S A 2000 97 4386 4391 10737773 10.1073/pnas.080499597
Schrier RW Cadnapaphornchai MA Renal aquaporin water channels: from molecules to human disease Prog Biophys Mol Biol 2003 81 117 131 12565698 10.1016/S0079-6107(02)00049-4
Kwon TH Nielsen J Masilamani S Hager H Knepper MA Frokiaer J Nielsen S Regulation of collecting duct AQP3 expression: response to mineralocorticoid Am J Physiol Renal Physiol 2002 283 F1403 21 12388415
Schreiber R Nitschke R Greger R Kunzelmann K The cystic fibrosis transmembrane conductance regulator activates aquaporin 3 in airway epithelial cells J Biol Chem 1999 274 11811 11816 10206998 10.1074/jbc.274.17.11811
Bondy C Chin E Smith BL Preston GM Agre P Developmental gene expression and tissue distribution of the CHIP28 water-channel protein Proc Natl Acad Sci U S A 1993 90 4500 4504 8506291
Yamamoto T Sasaki S Fushimi K Ishibashi K Yaoita E Kawasaki K Fujinaka H Marumo F Kihara I Expression of AQP family in rat kidneys during development and maturation Am J Physiol 1997 272 F198 204 9124396
Kim J Kim WY Han KH Knepper MA Nielsen S Madsen KM Developmental expression of aquaporin 1 in the rat renal vasculature Am J Physiol 1999 276 F498 509 10198408
Devuyst O Burrow CR Smith BL Agre P Knepper MA Wilson PD Expression of aquaporins-1 and -2 during nephrogenesis and in autosomal dominant polycystic kidney disease Am J Physiol 1996 271 F169 83 8760258
Wintour EM Earnest L Alcorn D Butkus A Shandley L Jeyaseelan K Ovine AQP1: cDNA cloning, ontogeny, and control of renal gene expression Pediatr Nephrol 1998 12 545 553 9761352 10.1007/s004670050502
Yasui M Marples D Belusa R Eklof AC Celsi G Nielsen S Aperia A Development of urinary concentrating capacity: role of aquaporin-2 Am J Physiol 1996 271 F461 8 8770180
Baum MA Ruddy MK Hosselet CA Harris HW The perinatal expression of aquaporin-2 and aquaporin-3 in developing kidney Pediatr Res 1998 43 783 790 9621988
Butkus A Earnest L Jeyaseelan K Moritz K Johnston H Tenis N Wintour EM Ovine aquaporin-2: cDNA cloning, ontogeny and control of renal gene expression Pediatr Nephrol 1999 13 379 390 10412857 10.1007/s004670050627
Wintour EM Congiu M Hardy KJ Hennessy DP Regulation of urine osmolality in fetal sheep Q J Exp Physiol 1982 67 427 435 7111663
Baum M Quigley R Satlin L Maturational changes in renal tubular transport Curr Opin Nephrol Hypertens 2003 12 521 526 12920400
Rai T Sekine K Kanno K Hata K Miura M Mizushima A Marumo F Sasaki S Urinary excretion of aquaporin-2 water channel protein in human and rat J Am Soc Nephrol 1997 8 1357 1362 9294826
Tsukahara H Hata I Sekine K Miura M Hata K Fujii Y Mayumi M Urinary excretion of aquaporin-2 in term and preterm infants Early Hum Dev 1998 51 31 37 9570029 10.1016/S0378-3782(97)00072-8
Nyul Z Vajda Z Vida G Sulyok E Frokiaer J Nielsen S Urinary aquaporin-2 excretion in preterm and full-term neonates Biol Neonate 2002 82 17 21 12119536 10.1159/000064147
Wong NL Tsui JK Angiotensin II upregulates the expression of vasopressin V2 mRNA in the inner medullary collecting duct of the rat Metabolism 2003 52 290 295 12647265 10.1053/meta.2003.50047
Harding R Hooper SB Rodeck CH and Whittle MJ Lung development and maturation. Fetal medicine: Basic science and clinical practice 1999 , Churchill Livingstone, London 181 196
Flecknoe SJ Wallace MJ Harding R Hooper SB Determination of alveolar epithelial cell phenotypes in fetal sheep: evidence for the involvement of basal lung expansion J Physiol 2002 542 245 253 12096066 10.1113/jphysiol.2001.014274
Kreda SM Gynn MC Fenstermacher DA Boucher RC Gabriel SE Expression and localization of epithelial aquaporins in the adult human lung Am J Respir Cell Mol Biol 2001 24 224 234 11245621
Krane CM Fortner CN Hand AR McGraw DW Lorenz JN Wert SE Towne JE Paul RJ Whitsett JA Menon AG Aquaporin 5-deficient mouse lungs are hyperresponsive to cholinergic stimulation Proc Natl Acad Sci U S A 2001 98 14114 14119 11707570 10.1073/pnas.231273398
Liu C Morrisey EE Whitsett JA GATA-6 is required for maturation of the lung in late gestation Am J Physiol Lung Cell Mol Physiol 2002 283 L468 75 12114210
Torday JS Rehan VK Testing for fetal lung maturation: a biochemical "window" to the developing fetus Clin Lab Med 2003 23 361 383 12848449 10.1016/S0272-2712(03)00030-1
King LS Nielsen S Agre P Aquaporins in complex tissues. I. Developmental patterns in respiratory and glandular tissues of rat Am J Physiol 1997 273 C1541 8 9374639
Ruddy MK Drazen JM Pitkanen OM Rafii B O'Brodovich HM Harris HW Modulation of aquaporin 4 and the amiloride-inhibitable sodium channel in perinatal rat lung epithelial cells Am J Physiol 1998 274 L1066 72 9609747
Umenishi F Carter EP Yang B Oliver B Matthay MA Verkman AS Sharp increase in rat lung water channel expression in the perinatal period Am J Respir Cell Mol Biol 1996 15 673 679 8918374
Yasui M Serlachius E Lofgren M Belusa R Nielsen S Aperia A Perinatal changes in expression of aquaporin-4 and other water and ion transporters in rat lung J Physiol 1997 505 3 11 9409467 10.1111/j.1469-7793.1997.003bc.x
Eisenbeiss C Welzel J Eichler W Klotz K Influence of body water distribution on skin thickness: measurements using high-frequency ultrasound Br J Dermatol 2001 144 947 951 11359379 10.1046/j.1365-2133.2001.04180.x
Cartlidge P The epidermal barrier Semin Neonatol 2000 5 273 280 11032710 10.1053/siny.2000.0013
Matsuzaki T Suzuki T Koyama H Tanaka S Takata K Water channel protein AQP3 is present in epithelia exposed to the environment of possible water loss J Histochem Cytochem 1999 47 1275 1286 10490456
Ma T Hara M Sougrat R Verbavatz JM Verkman AS Impaired stratum corneum hydration in mice lacking epidermal water channel aquaporin-3 J Biol Chem 2002 277 17147 17153 11880378 10.1074/jbc.M200925200
Sougrat R Morand M Gondran C Barre P Gobin R Bonte F Dumas M Verbavatz JM Functional expression of AQP3 in human skin epidermis and reconstructed epidermis J Invest Dermatol 2002 118 678 685 11918716 10.1046/j.1523-1747.2002.01710.x
Hara M Ma T Verkman AS Selectively reduced glycerol in skin of aquaporin-3-deficient mice may account for impaired skin hydration, elasticity, and barrier recovery J Biol Chem 2002 277 46616 46621 12270942 10.1074/jbc.M209003200
Hara M Verkman AS Glycerol replacement corrects defective skin hydration, elasticity, and barrier function in aquaporin-3-deficient mice Proc Natl Acad Sci U S A 2003 100 7360 7365 12771381 10.1073/pnas.1230416100
Hardman MJ Sisi P Banbury DN Byrne C Patterned acquisition of skin barrier function during development Development 1998 125 1541 1552 9502735
Hardman MJ Moore L Ferguson MW Byrne C Barrier formation in the human fetus is patterned J Invest Dermatol 1999 113 1106 1113 10594759 10.1046/j.1523-1747.1999.00800.x
Seeds AE Current concepts of amniotic fluid dynamics Am J Obstet Gynecol 1980 138 575 586 7425022
Bor NM Karpuzoglu T Hamzadi T Edguer E Kis M Role of fetal skin in circulation of amniotic fluid Arch Int Physiol Biochim 1970 78 69 78 4100886
Agren J Sjors G Sedin G Transepidermal water loss in infants born at 24 and 25 weeks of gestation Acta Paediatr 1998 87 1185 1190 9846922 10.1080/080352598750031194
Agren J Zelenin S Hakansson M Eklof AC Aperia A Nejsum LN Nielsen S Sedin G Transepidermal water loss in developing rats: role of aquaporins in the immature skin Pediatr Res 2003 53 558 565 12612219 10.1203/01.PDR.0000055777.25933.98
Umenishi F Verkman AS Gropper MA Quantitative analysis of aquaporin mRNA expression in rat tissues by RNase protection assay DNA Cell Biol 1996 15 475 480 8672243
Hasegawa H Lian SC Finkbeiner WE Verkman AS Extrarenal tissue distribution of CHIP28 water channels by in situ hybridization and antibody staining Am J Physiol 1994 266 C893 903 7513954
Page E Winterfield J Goings G Bastawrous A Upshaw-Earley J Water channel proteins in rat cardiac myocyte caveolae: osmolarity-dependent reversible internalization Am J Physiol 1998 274 H1988 2000 9841525
Au C Cooper ST Lo HP Yang N Wintour EM North KN Winlaw DS Novel expression of aquaporins in the human heart. J Mol Cell Cardiol 2004 36 655 662 15135660 10.1016/j.yjmcc.2004.01.009
Parton RG Caveolae and caveolins Curr Opin Cell Biol 1996 8 542 548 8791446 10.1016/S0955-0674(96)80033-0
Bloch W Addicks K Hescheler J Fleischmann BK Nitric oxide synthase expression and function in embryonic and adult cardiomyocytes Microsc Res Tech 2001 55 259 269 11748864 10.1002/jemt.1175
Galbiati F Engelman JA Volonte D Zhang XL Minetti C Li M Hou HJ Kneitz B Edelmann W Lisanti MP Caveolin-3 null mice show a loss of caveolae, changes in the microdomain distribution of the dystrophin-glycoprotein complex, and t-tubule abnormalities J Biol Chem 2001 276 21425 21433 11259414 10.1074/jbc.M100828200
Rybin VO Xu X Lisanti MP Steinberg SF Differential targeting of beta -adrenergic receptor subtypes and adenylyl cyclase to cardiomyocyte caveolae. A mechanism to functionally regulate the cAMP signaling pathway J Biol Chem 2000 275 41447 41457 11006286 10.1074/jbc.M006951200
Tseng YT Kopel R Stabila JP McGonnigal BG Nguyen TT Gruppuso PA Padbury JF Beta-adrenergic receptors (betaAR) regulate cardiomyocyte proliferation during early postnatal life FASEB J 2001 15 1921 1926 11532972 10.1096/fj.01-0151com
Suleymanian MA Baumgarten CM Osmotic gradient-induced water permeation across the sarcolemma of rabbit ventricular myocytes J Gen Physiol 1996 107 503 514 8722563 10.1085/jgp.107.4.503
Kellen MR Bassingthwaighte JB Transient transcapillary exchange of water driven by osmotic forces in the heart Am J Physiol Heart Circ Physiol 2003 285 H1317 31 12738617
Butkus A Alcorn D Earnest L Moritz K Giles M Wintour EM Expression of aquaporin-1 (AQP1) in the adult and developing sheep kidney Biol Cell 1997 89 313 320 9468602 10.1016/S0248-4900(97)83384-9
Jonker S Davis LE van der Bilt JD Hadder B Hohimer AR Giraud GD Thornburg KL Anaemia stimulates aquaporin 1 expression in the fetal sheep heart Exp Physiol 2003 88 691 698 14603367 10.1113/eph8802626
Ma T Yang B Verkman AS Cloning of a novel water and urea-permeable aquaporin from mouse expressed strongly in colon, placenta, liver, and heart Biochem Biophys Res Commun 1997 240 324 328 9388476 10.1006/bbrc.1997.7664
Wintour EM Liu H Dodic M Moritz K Differential renal and cardiac gene expression in ovine fetuses programmed to become hypertersive adults by early glucocorticoid treatment. Proceedings of the 12th International Congress of Endocrinolgy 2004 501 505 Medimond Italy
Burrell JH Boyn AM Kumarasamy V Hsieh A Head SI Lumbers ER Growth and maturation of cardiac myocytes in fetal sheep in the second half of gestation Anat Rec 2003 274A 952 961 10.1002/ar.a.10110
Robinson JS Moore VM Owens JA McMillen IC Origins of fetal growth restriction Eur J Obstet Gynecol Reprod Biol 2000 92 13 19 10986429 10.1016/S0301-2115(00)00421-8
Sundgren NC Giraud GD Stork PJ Maylie JG Thornburg KL Angiotensin II stimulates hyperplasia but not hypertrophy in immature ovine cardiomyocytes J Physiol 2003 548 881 891 12626668
Firsov D Muller OG Aldosterone action in the heart Pflugers Arch 2003 446 328 333 12733077
Burrell JH Hegarty BD McMullen JR Lumbers ER Effects of gestation on ovine fetal and maternal angiotensin receptor subtypes in the heart and major blood vessels Exp Physiol 2001 86 71 82 11429622 10.1113/eph8602075
Everett AD Fisher A Tufro-McReddie A Harris M Developmental regulation of angiotensin type 1 and 2 receptor gene expression and heart growth J Mol Cell Cardiol 1997 29 141 148 9040029 10.1006/jmcc.1996.0259
Lamparter S Sun Y Weber KT Angiotensin II receptor blockade during gestation attenuates collagen formation in the developing rat heart Cardiovasc Res 1999 43 165 172 10536701 10.1016/S0008-6363(99)00111-X
Price RL Carver W Simpson DG Fu L Zhao J Borg TK Terracio L The effects of angiotensin II and specific angiotensin receptor blockers on embryonic cardiac development and looping patterns Dev Biol 1997 192 572 584 9441690 10.1006/dbio.1997.8764
Rizzo G Capponi A Rinaldo D Arduini D Romanini C Ventricular ejection force in growth-retarded fetuses Ultrasound Obstet Gynecol 1995 5 247 255 7600206 10.1046/j.1469-0705.1995.05040247.x
Yum MK Park EY Kim CR Hwang JH Alterations in irregular and fractal heart rate behavior in growth restricted fetuses Eur J Obstet Gynecol Reprod Biol 2001 94 51 58 11134826 10.1016/S0301-2115(00)00314-6
Hu XW Levy A Hart EJ Nolan LA Dalton GR Levi AJ Intra-uterine growth retardation results in increased cardiac arrhythmias and raised diastolic blood pressure in adult rats Cardiovasc Res 2000 48 233 243 11054470 10.1016/S0008-6363(00)00167-X
Machlitt A Wauer RR Chaoui R Longitudinal observation of deterioration of Doppler parameters, computerized cardiotocogram and clinical course in a fetus with growth restriction J Perinatal Med 2001 29 71 76 10.1515/JPM.2001.010
Frigeri A Nicchia GP Verbavatz JM Valenti G Svelto M Expression of aquaporin-4 in fast-twitch fibers of mammalian skeletal muscle J Clin Invest 1998 102 695 703 9710437
Jimi T Wakayama Y Murahashi M Shibuya S Inoue M Hara H Matsuzaki Y Uemura N Aquaporin 4: lack of mRNA expression in the rat regenerating muscle fiber under denervation Neurosci Lett 2000 291 93 96 10978582 10.1016/S0304-3940(00)01382-3
Frigeri A Nicchia GP Nico B Quondamatteo F Herken R Roncali L Svelto M Aquaporin-4 deficiency in skeletal muscle and brain of dystrophic mdx mice FASEB J 2001 15 90 98 11149896 10.1096/fj.00-0260com
Wakayama Y Jimi T Inoue M Kojima H Murahashi M Kumagai T Yamashita S Hara H Shibuya S Reduced aquaporin 4 expression in the muscle plasma membrane of patients with Duchenne muscular dystrophy Arch Neurol 2002 59 431 437 11890849 10.1001/archneur.59.3.431
Miyan JA Nabiyouni M Zendah M Development of the brain: a vital role for cerebrospinal fluid Can J Physiol Pharmacol 2003 81 317 328 12769224 10.1139/y03-027
Nico B Frigeri A Nicchia GP Corsi P Ribatti D Quondamatteo F Herken R Girolamo F Marzullo A Svelto M Roncali L Severe alterations of endothelial and glial cells in the blood-brain barrier of dystrophic mdx mice Glia 2003 42 235 251 12673830 10.1002/glia.10216
Verkman AS Role of aquaporin water channels in eye function Exp Eye Res 2003 76 137 143 12565800 10.1016/S0014-4835(02)00303-2
Ferrary E Sterkers O Mechanisms of endolymph secretion Kidney Int Suppl 1998 65 S98 103 9551441
Horio Y Potassium channels of glial cells: distribution and function Jpn J Pharmacol 2001 87 1 6 11676192 10.1254/jjp.87.1
Seemungal BM Gresty MA Bronstein AM The endocrine system, vertigo and balance Curr Opin Neurol 2001 14 27 34 11176214 10.1097/00019052-200102000-00005
Fukushima M Kitahara T Uno Y Fuse Y Doi K Kubo T Effects of intratympanic injection of steroids on changes in rat inner ear aquaporin expression Acta Otolaryngol 2002 122 600 606 12403121 10.1080/000164802320396268
Kumagami H Loewenheim H Beitz E Wild K Schwartz H Yamashita K Schultz J Paysan J Zenner HP Ruppersberg JP The effect of anti-diuretic hormone on the endolymphatic sac of the inner ear Pflugers Arch 1998 436 970 975 9799415 10.1007/s004240050731
Papadopoulos MC Krishna S Verkman AS Aquaporin water channels and brain edema Mt Sinai J Med 2002 69 242 248 12357265
Badaut J Lasbennes F Magistretti PJ Regli L Aquaporins in brain: distribution, physiology, and pathophysiology J Cereb Blood Flow Metab 2002 22 367 378 11919508 10.1097/00004647-200204000-00001
Solenov E Watanabe H Manley GT Verkman AS Seven-fold reduced osmotic water permeability in primary astrocyte cultures from aquaporin-4 deficient mice measured by a calcein quenching method Am J Physiol Cell Physiol 2003 2003
Nagelhus EA Horio Y Inanobe A Fujita A Haug FM Nielsen S Kurachi Y Ottersen OP Immunogold evidence suggests that coupling of K+ siphoning and water transport in rat retinal Muller cells is mediated by a coenrichment of Kir4.1 and AQP4 in specific membrane domains Glia 1999 26 47 54 10088671 10.1002/(SICI)1098-1136(199903)26:1<47::AID-GLIA5>3.0.CO;2-5
Li J Patil RV Verkman AS Mildly abnormal retinal function in transgenic mice without Muller cell aquaporin-4 water channels Invest Ophthalmol Vis Sci 2002 43 573 579 11818406
Manley GT Fujimura M Ma T Noshita N Filiz F Bollen AW Chan P Verkman AS Aquaporin-4 deletion in mice reduces brain edema after acute water intoxication and ischemic stroke Nat Med 2000 6 159 163 10655103 10.1038/72256
Vajda Z Pedersen M Fuchtbauer EM Wertz K Stodkilde-Jorgensen H Sulyok E Doczi T Neely JD Agre P Frokiaer J Nielsen S Delayed onset of brain edema and mislocalization of aquaporin-4 in dystrophin-null transgenic mice Proc Natl Acad Sci U S A 2002 99 13131 13136 12232046 10.1073/pnas.192457099
Neely JD Amiry-Moghaddam M Ottersen OP Froehner SC Agre P Adams ME Syntrophin-dependent expression and localization of Aquaporin-4 water channel protein Proc Natl Acad Sci U S A 2001 98 14108 14113 11717465 10.1073/pnas.241508198
Amiry-Moghaddam M Otsuka T Hurn PD Traystman RJ Haug FM Froehner SC Adams ME Neely JD Agre P Ottersen OP Bhardwaj A An alpha-syntrophin-dependent pool of AQP4 in astroglial end-feet confers bidirectional water flow between blood and brain Proc Natl Acad Sci U S A 2003 100 2106 2111 12578959 10.1073/pnas.0437946100
Wen H Nagelhus EA Amiry-Moghaddam M Agre P Ottersen OP Nielsen S Ontogeny of water transport in rat brain: postnatal expression of the aquaporin-4 water channel Eur J Neurosci 1999 11 935 945 10103087 10.1046/j.1460-9568.1999.00502.x
Nico B Frigeri A Nicchia GP Quondamatteo F Herken R Errede M Ribatti D Svelto M Roncali L Role of aquaporin-4 water channel in the development and integrity of the blood-brain barrier J Cell Sci 2001 114 1297 1307 11256996
Huang D Chen P Chen S Nagura M Lim DJ Lin X Expression patterns of aquaporins in the inner ear: evidence for concerted actions of multiple types of aquaporins to facilitate water transport in the cochlea Hear Res 2002 165 85 95 12031518 10.1016/S0378-5955(02)00288-5
Sawada S Takeda T Kitano H Takeuchi S Okada T Ando M Suzuki M Kakigi A Aquaporin-1 (AQP1) is expressed in the stria vascularis of rat cochlea Hear Res 2003 181 15 19 12855358 10.1016/S0378-5955(03)00131-X
Mhatre AN Jero J Chiappini I Bolasco G Barbara M Lalwani AK Aquaporin-2 expression in the mammalian cochlea and investigation of its role in Meniere's disease Hear Res 2002 170 59 69 12208541 10.1016/S0378-5955(02)00452-5
Sawada S Takeda T Kitano H Takeuchi S Kakigi A Azuma H Aquaporin-2 regulation by vasopressin in the rat inner ear Neuroreport 2002 13 1127 1129 12151754 10.1097/00001756-200207020-00011
Takeda T Sawada S Takeda S Kitano H Suzuki M Kakigi A Takeuchi S The effects of V2 antagonist (OPC-31260) on endolymphatic hydrops Hear Res 2003 182 9 18 12948596 10.1016/S0378-5955(03)00135-7
Merves M Bobbitt B Parker K Kishore BK Choo D Developmental expression of aquaporin 2 in the mouse inner ear Laryngoscope 2000 110 1925 1930 11081612 10.1097/00005537-200011000-00030
Takumi Y Nagelhus EA Eidet J Matsubara A Usami S Shinkawa H Nielsen S Ottersen OP Select types of supporting cell in the inner ear express aquaporin-4 water channel protein Eur J Neurosci 1998 10 3584 3595 9875338 10.1046/j.1460-9568.1998.00360.x
Li J Verkman AS Impaired hearing in mice lacking aquaporin-4 water channels J Biol Chem 2001 276 31233 31237 11406631 10.1074/jbc.M104368200
Mhatre AN Stern RE Li J Lalwani AK Aquaporin 4 expression in the mammalian inner ear and its role in hearing Biochem Biophys Res Commun 2002 297 987 996 12359252 10.1016/S0006-291X(02)02296-9
Saadoun S Papadopoulos MC Harachikuma M Verkman AS Impairment of angiogenesis and cell migration by targeted aquaporin-1 gene disruption Nature 2005 434 786 792 15815633 10.1038/nature03460
| 15888206 | PMC1156947 | CC BY | 2021-01-04 16:37:13 | no | Reprod Biol Endocrinol. 2005 May 11; 3:18 | utf-8 | Reprod Biol Endocrinol | 2,005 | 10.1186/1477-7827-3-18 | oa_comm |
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-191589289610.1186/1477-7827-3-19ResearchCytoskeleton reorganization mediates alpha6beta1 integrin-associated actions of laminin on proliferation and survival, but not on steroidogenesis of ovine granulosa cells Bellego Frédérique Le [email protected] Stéphane [email protected] Claudine [email protected] Danielle [email protected] Physiologie de la Reproduction et des Comportements, UMR 6175 INRA-CNRS-Université de Tours-Haras Nationaux, INRA 37380 Nouzilly, France2005 16 5 2005 3 19 19 30 3 2005 16 5 2005 Copyright © 2005 Bellego et al; licensee BioMed Central Ltd.2005Bellego et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Laminin (LN) is one of the most abundant extracellular matrix components of the basal lamina and granulosa cell layers of ovarian follicles. Culture of ovine granulosa cells (GC) on LN substratum induces cell spreading, enhances cell survival and proliferation, and promotes luteinization. Previous investigations have shown that these effects are mostly mediated by the alpha6beta1 integrin, but its signalization pathways have not been investigated. This study aimed to assess the importance of the cytoskeleton in the alpha6beta1 integrin-mediated actions of laminin on survival, proliferation and steroidogenesis of ovine GC.
Methods
The relationships between morphology and functions of ovine GC cultured on substrata containing LN or/and RGD peptides were investigated. The effects of (1) cytochalasin D, an actin cytoskeleton-disrupting drug, (2) a specific function-blocking antibody raised against alpha6 integrin subunit (anti-alpha6 IgG), and (3) an inhibitor of the ERK1/2 signalization pathway (PD98059) were assessed for GC shape, pyknosis and proliferation rates, oestradiol and progesterone secretions.
Results
Cytoskeleton disruption by cytochalasin D induced cell rounding, inhibited proliferation, promoted pyknosis, inhibited progesterone secretion and enhanced oestradiol secretion by GC cultured on LN. When GC were cultured on various substrata containing LN and/or RGD peptides in the presence or absence of anti-alpha6 IgG, both the existence of close correlations between the percentage of round cells, and the GC proliferation rate (r = -0.87) and pyknotic rate (r = 0.76) were established, but no relationship was found between cell shape and steroidogenesis. Inhibition of the ERK1/2 signalization pathway by PD98059 had no effect on GC shape, proliferation or pyknotic rates. However, it dramatically reduced progesterone secretion, expression of cytochrome P450 cholesterol side-chain cleavage and 3beta-hydroxysteroid deshydrogenase enzymes, and enhanced oestradiol secretion, thereby reproducing all the effects of the anti-alpha6 IgG on steroidogenesis of GC cultured on LN.
Conclusion
LN may participate in the paracrine control of follicular development through different mechanisms. It could enhance proliferation and survival of GC through its alpha6beta1 integrin-mediated actions on cytoskeleton. In contrast, its stimulating action on GC luteinization could be partly mediated by the ERK1/2 pathway, irrespective of cell shape.
==== Body
Background
Follicular development is under the control of both gonadotropins and numerous paracrine factors that are critically involved in determining the fate of follicles, atresia or ovulation. From the primordial to the preovulatory follicular stage, the outer layer of granulosa cells (GC) lays on a basal lamina that separates them from the theca layers and interstitial ovarian tissue [1]. This basal lamina, consisting of extracellular matrix (ECM) components such as laminin (LN), fibronectin, collagens and various glycoproteins and proteoglycans, is subjected to intense remodeling during follicular development and atresia, changing its composition from the primordial to the preovulatory or atretic stages [2]. For example, the basal lamina becomes less collagenous and more laminin-rich during follicular development [3,4]. In antral follicles, laminin and other ECM components are also present within the multilayered wall of GC [5,6], particularly in basal lamina-like material deposited as aggregates between the GC layers, recently called focimatrix (for focal intra-epithelial matrix) [7]. These observations indicate that ECM components contribute to the microenvironment of GC, but their specific roles in follicular development have not yet been established.
LN is one of the most abundant ECM components of the basal lamina [2,4,5,8-12] and, as stated above, it is also present within the granulosa layers of antral follicles. In sheep, LN levels increase considerably in the granulosa of antral follicles during the follicular and preovulatory phases of the cycle [6]. In vitro experiments have shown that LN improves GC survival (rat: [13]; sheep: [14]) and stimulates the proliferation of GC from small antral follicles [14]. In GC from large antral and preovulatory follicles, LN increases progesterone secretion (rat: [13,15]; pig: [16]; sheep:[6]) and decreases estradiol secretion [6], suggesting that it might promote luteinization. Overall, these results suggest that LN has an important regulatory effect on GC functions throughout the terminal development of antral follicles.
Among the different integrins that can bind LN and mediate its action in various cell types, α6β1 and α6β4 have the particular feature of being highly specific LN receptors [17]. The α6 integrin subunit has been shown to be greatly expressed in GC of different animal species (human: [18]; marmoset: [19,20]; pig: [21]; mouse: [22]; sheep: [6]). In sheep, GC of healthy antral follicles express high levels of α6β1 integrin, and when a function-blocking antibody raised against the α6-integrin subunit is added to the medium of GC cultured on LN, their survival, proliferation and steroidogenesis are dramatically altered [6]. These results suggest that α6β1 integrin mediates most LN actions on GC, but the mechanisms involved in α6β1 integrin-mediated functional changes in GC are unknown.
From previous observations, addition of the antibody raised against the α6-integrin subunit in the GC culture medium impairs cell-spreading on LN substratum and induces the formation of clusters of rounded cells [6]. It can be hypothesized that changes in cell shape might be responsible for all or part of the functional changes observed in survival, proliferation and steroidogenesis of GC. It has been established in various cell models that integrin binding to ECM components promotes changes in the mechanical tension of the cytoskeleton and thereby induces multiple signaling pathways [23,24]. The cytoskeleton consists of actin microfilaments, microtubules and intermediate filaments which connect to form a three-dimensional network that runs from the plasma membrane, and particularly from integrins, to the chromosomes in the nucleus [25]. The importance of the cytoskeleton in mediating steroidogenesis in response to gonadotropins has been suggested [26-33]. It is likely that cell morphology influences cell polarization and organelle organization through the cytoskeleton, thereby controlling steroid production and secretion [28,30,34], but the possible role of integrin-mediated cytoskeleton changes has not yet been established in regulating GC steroidogenesis, survival and proliferation.
This study aimed to assess the importance of the cytoskeleton in the α6β1 integrin-mediated actions of LN on survival, proliferation and steroidogenesis of ovine GC. For this purpose, different experiments were performed to investigate the existence of coupling between cell shape and function when α6β1 integrin activity was altered. Firstly, the action of cytochalasin D, an actin cytoskeleton-disrupting drug, was studied on both the shape and functions of GC cultured on LN substratum. Secondly, the action of a function-blocking antibody raised against α6 integrin subunit was studied on both the shape and functions of GC cultured on substrata containing different ratios of LN and RGD peptides that specifically recognize the α5/ α8/ αv/ αIIb but not the α6 integrin subfamilies [35]. Lastly, the consequences of inhibiting the ERK1/2 (Extracellular signal-Related kinase) signalization pathway, that has been shown to transduce some of the α6β1 integrin-mediated effects of LN [36-39], were studied on both GC shape and functions.
Methods
Reagents and chemicals
Fluorogestone acetate sponges used to synchronize estrous cycles were obtained from Intervet Pharma (Angers, France). Porcine FSH (pFSH) from pituitary extract (pFSH activity = 1.15 × activity NIH pFSH-P1) used for animal injections was obtained from Dr. Y. Combarnous (Nouzilly, France). B2 medium for cell cultures was prepared according to Menezo [40]. Rat monoclonal antibody GoH3 raised against human α6 integrin subunit (anti-α6 IgG) for use in cell cultures was purchased from Serotec (Oxford, England). For western immunoblotting, rabbit polyclonal antibody raised against cytochrome P450 cholesterol side-chain cleavage (anti-P450scc) was purchased from Chemicon (Euromedex, Mundolsheim, France), rabbit polyclonal antibody raised against 3β-hydroxysteroid deshydrogenase (anti-3βHSD) was a gift of Dr. V. Luu-The (Quebec, Canada) and rabbit polyclonal antibody raised against ERK1 and ERK2 (anti-ERK1/2) was purchased from Santa Cruz (Le Perray-en-Yveline, France). Rabbit polyclonal antibody raised against the phosphorylated forms of ERK1 and ERK2 (anti-P-ERK1/2) was purchased from Calbiochem (Meudon, France). Anti-rabbit IgG antibody coupled to horseradish peroxidase was purchased from Interchim (Montluçon, France). The following reagents were purchased from Sigma (L'Isle d'Abeau Chesnes, France): McCoy's 5a medium with bicarbonate, penicillin/streptomycin, bovine serum albumin (BSA tissue culture grade) used for culture medium, transferrin, selenium, bovine insulin, androstenedione, LN from EHS tumor (mainly LN-1: [41]), cytochalasin D, FITC-conjugated phalloidin, PD98059, Igepal, phenylmethylsulfonyl fluoride (PMSF), leupeptin, aprotinin, sodium fluoride, sodium pyrophosphate and sodium orthovanadate. Hepes, L-glutamine, fungizone and trypsin were purchased from GIBCO BRL (Cergy-Pontoise, France). RGD peptides (arginin – glycin – aspartic acid sequence) were obtained from Interchim (Montluçon, France). Sterile 96-well plates (Nunclon Delta) were obtained from Nunc (Naperville, IL, USA) and plastic tissue-culture chamber slidesfrom Poly-Labo (Strasbourg, France). [3H]thymidine (specific activity 6.7 Ci/nmol) was obtained from Dupont De Nemours (Les Ulis, France) and NTB2 emulsion for autoradiography from Integra Bioscience (Cergy-Pontoise, France). For protein assay, the BC Assay protein kit was obtained from Uptima Interchim (Montluçon, France). Immobilon P membranes for western blots were obtained from Millipore Corporation (Bedford, MA, USA). ECL (enhanced chemiluminescence) reagents were obtained from Amersham Pharmacia Biotech (Orsay, France).
Animals
All procedures were approved by the Agricultural and Scientific Research agencies (approval number A 37801) and conducted in accordance with the guidelines for Care and Use of Agricultural Animals in Agricultural Research and Teaching. Experimental research was performed with the approval of the regional ethics committee of the Région Centre (Tours, France). During the reproductive season, adult Romanov ewes were treated with intravaginal sponges impregnated with progestagen (fluorogestone acetate, 40 mg) for 15 days to mimic a luteal phase. GC were collected from animals slaughtered in the luteal phase of the following estrous cycle (10 days after sponge removal), after treatment with intramuscular injections of 6 IU and 5 IU pFSH administered 24 h and 12 h prior to slaughter respectively.
Isolation of GC
For each culture experiment, immediately after slaughter, ovaries from 3 ewes were immersed for 15 min in isotonic solution containing amphotericin (50 mg/ml) and antibiotics (2 million UI/ml penicillin and 2 g/l streptomycin). The ovaries were placed in B2 medium, and follicles larger than 1 mm in diameter were dissected within 1 hour of slaughter. A total number of 50–70 small (1–3 mm in diameter) and 10–15 large (> 4 mm in diameter) were dissected from these pooled ovaries. Follicular fluid from large follicles (> 4 mm) was aspirated with a 26-gauge needle. Each follicle was then slit open in B2 medium, and GC were removed by gently scraping the interior surface of the follicle with a platinum loop. GC suspensions were pooled by follicle size (small: 1–3 mm, or large: > 4 mm). The two resulting cell suspensions were centrifuged at 300 × g for 7 min and re-suspended in culture medium (McCoy's 5a containing bicarbonate supplemented with 20 mmol/l Hepes, 100 kUI/l penicillin, 0.1 g/l streptomycin, 3 mmol/l L-glutamine, 0.1% BSA (w/v), 100 μg/l insulin, 0.1 μmol/l androstenedione, 5 mg/l transferrin, 20 μg/l selenium). The total number of cells per suspension was estimated by counting an aliquot of each suspension using a hemocytometer under a phase-contrast microscope. The number varied between 10 × 106 and 20 × 106 cells per suspension. Cell viability, determined after vital staining with trypan blue dye (0.125%, final concentration) varied between 60 and 80%.
GC culture
GC culture was performed according to Campbell's method [42]. GC from small and large follicles were cultured in 96-well tissue-culture plates or in tissue-culture chamber slides coated with LN (5 μg/cm2 in distilled water), unless specified. In the experiments using culture substrata containing LN and/or RGD peptides, different mixes were prepared using the LN solution described above and an RGD peptide solution (1.67 μg/cm2 in PBS) to obtain different ratios of LN and RGD peptides (100%, 43%, 25%, 18%, 14%, 0% w/w % of LN in the mix used for coating). The concentrations of the peptides had been determined in preliminary experiments for their morphological effects on GC cultured in the presence or absence of anti-α6 IgG (0.5 μg/ml). The different substrata were prepared 72 h before use and allowed to dry at room temperature.
GC suspensions from small and large follicles were seeded at 105 viable cells/well and cultured at 37°C in a humidified atmosphere with 5% CO2, in serum-free culture medium (see isolation of GC). The effect of cytochalasin D, an inhibitor of actin polymerization, was tested at different concentrations in the 0.05 – 5 μg/ml range. The effect of PD98059, an inhibitor of the ERK1/2 activation pathway, was tested in the 1 – 30 μM range. The effects of anti-α6 IgG (0.5 μg/ml) were always compared with the effects of the inhibitors within the same culture experiment. Each condition was tested in triplicate in each GC culture from small and large follicles. Culture media were partially replaced (175 μl out of 250 μl) every 48 h and the spent medium was stored at -20°C until assay. At 144 h of culture, cells were detached with trypsin and counted as described above (see isolation of GC), or prepared for western immunoblotting. For studies of ERK1/2 phosporylation, GC were cultured on LN or plastic, with and without inhibitors, for 24 h before western immunoblotting. In preliminary experiments, this culture time was shown to be needed for cell plating on substratum [14] and allowed detection of early effects of LN on ERK1/2 phosporylation.
Determination of thymidine labeling index and pyknotic index
GC proliferation and survival were assessed by measuring the thymidine labeling index (percentage of labeled cells) and the pyknotic index (percentage of pyknotic cells) after 48 h of culture. This has been shown to allow the be the optimal culture time for the study of the effects of various factors, particularly ECM components, on both proliferation and survival of cultured ovine GC [6,14,43].
To determine the thymidine labeling index, cells were washed with B2 medium without thymine, then incubated with [3H]thymidine (0.25 μCi/ml) at 37°C for 2 h. After 2 washes with B2 medium (with thymine), cells were detached with 1% trypsin, pelleted and fixed in 3% glutaraldehyde for 1.5 h at room temperature. Cells were then smeared onto histological slides by cytocentrifugation. Smears were stained with Feulgen, dipped in NTB2 emulsion, air-dried, and exposed for autoradiography for 6 days at 4°C. The thymidine labeling index was estimated by counting the number of labeled and unlabeled cells in 20 different microscopic fields (100X objective).
To determine the pyknotic index, cells were detached by trypsin, fixed in glutaraldehyde, cytocentrifuged, and smears were stained with Feulgen as described above. The pyknotic index was estimated by counting the number of pyknotic cells, i.e. cells with condensed or fragmented nuclear chromatin [44], and non-pyknotic cells in 20 different microscopic fields (100X objective). Previous results have established that the presence of pyknotic cells is associated with DNA fragmentation characteristic of apoptotic process in cultured GC [14].
For both the thymidine labeling index and the pyknotic index, calculations were made on 500–1000 cells per slide.
Estradiol-17β and progesterone radioimmunoassays
The concentrations of estradiol and progesterone in the culture medium of GC from large follicles were measured after 144 h of culture. This has been shown to be the optimal culture time for the study of the effects of ECM components on steroidogenesis of cultured ovine GC [6,14]. The radioimmunoassay protocol previously described [45-47] was adapted to measure steroids in cell culture media directly. The estradiol detection limit was 1.5 pg/tube (7.5 pg/well) and the intra- and inter-assay coefficients of variation were less than 7% and 9% respectively. The progesterone detection limit was 12 pg/tube (60 pg/well), and the intra- and inter-assay coefficients of variation were less than 10% and 11% respectively. Results were expressed as the amount of steroids secreted between 96 h and 144 h of culture per 50,000 cells recovered at the end of the culture period.
Western immunoblotting
GC whole extracts were obtained by resuspension in 100 μl cell lysis buffer [150 mM NaCl, 50 mM Tris-HCl (pH 7.5), 1% Igepal, 0.5% sodium deoxycholate, 0.1% SDS] containing several protease inhibitors (10 mM PMSF, 1 μg/ml leupeptin, 1 μg/ml aprotinin) and phosphatase inhibitors (100 mM sodium fluoride, 10 mM sodium pyrophosphate, 2 mM sodium orthovanadate) at 4 °C for 20 min. Cell lysates were centrifuged at 20,000 × g for 20 min, and the protein concentration in the supernatants was determined by the BC Assay protein kit following the manufacturer's protocol. Aliquots (5 to10 μg, corresponding to 5 × 104 to 105 GC) were subjected to 10% SDS-PAGE under reducing conditions, then the proteins were electrophoretically transferred from the gels onto Immobilon P membranes. These membranes were incubated for 1 h at room temperature with 20 mM Tris-buffered saline (TBS, pH 7.6), containing 3% BSA and 0.1% Tween-20 to saturate nonspecific sites. They were then incubated for 1 h at room temperature with anti-P-ERK1/2, or anti-ERK1/2, or anti-P450scc, or anti-3βHSD (final dilutions 1:1000) in TBS containing 1% BSA and 0.1% Tween-20. After washing in TBS containing 0.1% Tween-20, the membranes were incubated for 1 h at room temperature with horseradish peroxidase-conjugated anti-rabbit IgG (final dilution 1:10,000) in TBS containing 0.01% Tween-20, and the signal was visualized using ECL system followed by autoradiography. The autoradiograms were quantified using a videodensitometer (VDS-CL, Amersham Pharmacia Biotech).
Actin staining and fluorescence microscopy
After culture in chamber slides, GC were fixed for 15 min with 4% paraformaldehyde in PBS. Cells were then washed with PBS and left in 0.1 M glycine in PBS for 15 min. After an additional wash, the cells were permabilized with 0.2% Triton X-100 (w/v) in PBS containing 1% BSA for 10 min, and nonspecific binding sites were blocked in 2% BSA. Cells were then treated for 30 min with 0.5 μM FITC-conjugated phalloidin. All the above incubations were performed at room temperature. After washing, cells were mounted in Mowiol and were studied under fluorescence microscopy. The percentage of the different morphological states of GC (spread cells, spindle-shaped cells, round cells) was established by counting 1000 to 1500 cells per culture well.
Statistical analysis
All experimental data are presented as the mean ± SEM. Data were fitted to sigmoidal dose – response curves or Gaussian distributions with GraphPrad PRISM software (San Diego, CA, USA). The effects of increasing doses of inhibitors (cytochalasin D or PD98059) on cell numbers, percentages of round cells, thymidine labeling index and pyknotic index were analyzed using a one-way ANOVA for repeated measures followed by Tukey-Kramer tests. For a given substratum, the effects of the addition of anti-α6 IgG on cell numbers, percentages of round cells, thymidine labeling index and pyknotic index were analyzed using a paired t test. The effects of anti-α6 IgG or inhibitors (cytochalasin D or PD98059) on GC steroidogenesis were analyzed using a two-way ANOVA in order to assess the effects of treatment as well as of culture resulting from variations between both animals and the quality of the ovarian follicles dissected for each culture. Results from western immunoblotting analysis were analyzed by a paired t test. Comparisons with p > 0.05 were not considered significant.
Results and discussion
Consequences of α6β1 integrin activation by LN on GC shape and functions
Previous results have shown that LN, used as a culture substratum of ovine GC, induces cell spreading, enhances proliferation and survival rates, stimulates progesterone and reduces estradiol secretion by GC and that these effects are the consequence of α6β1 integrin activation by LN [6,14]. Accordingly, addition of anti-α6 IgG to the medium of GC cultured on LN induced dramatic cell rounding and important functional changes such as a decrease in proliferation rate and an increase in pyknotic rate of GC, as well as a decrease in progesterone and an increase in estradiol secretion (p < 0.01 for all parameters, Fig. 1). These effects were highly specific to the presence of LN in the culture substratum [6].
Figure 1 Effect of anti-α6 IgG on morphology and functions of GC cultured on LN substratum. GC were cultured up to 144 h on LN with (b and solid bars in c, d, e and f) or without (a and empty bars in c, d, e and f) anti-α6 IgG (0.5 μg/ml) in culture medium. (a) and (b): representative microscopical fields of cultured GC throughout the culture period; (c): proliferation rates of GC from small follicles, assessed by thymidine labelling at 48 h of culture; (d): pyknotic rates of GC from large follicles at 48 h of culture; (e): progesterone (P4) secretion by GC from large follicles between 96 h and 144 h of culture; (f): estradiol (E2) secretion by GC from large follicles between 96 h and 144 h of culture. Data represent mean ± SEM of 5 independent experiments. * : p < 0.01, with vs. without anti-α6 IgG.
Effects of cytoskeleton disruption by cytochalasin D on GC cultured on LN substratum
To assess whether all or part of these functional changes might be the consequence of cell rounding, the action of cytochalasin D, an inhibitor of actin polymerization, was studied on GC cultured on LN substratum. Addition of cytochalasin D to the culture medium of GC from small and large follicles reduced the formation of actin stress fibers and impeded cell spreading on LN, inducing the formation of spindle-shaped cells at doses higher than 0.1 μg/ml, and of round cells at doses higher than 0.5 μg/ml (p < 0.001 for GC from both small and large follicles, Fig. 2 a, b and 2 c). In GC from both small and large follicles, cytochalasin D treatment induced a clear dose-dependent decrease in cell numbers (p < 0.001 for GC from both small and large follicles, Fig. 3 a and 3 b). In cultures of GC from small follicles with a high proliferative activity, this decrease in cell number was associated with a dose-dependent decrease in cell proliferation rate (p < 0.001, Fig. 3 c). In cultures of GC from both small and large follicles, cytochalasin D also induced an increase in pyknotic rate at the 5 μg/ml dose (p < 0.05, Fig. 3 d). In cultures of GC from large follicles with a high steroidogenic activity, cytochalasin D induced a dose-dependent increase (p < 0.001) and decrease (p < 0.01) in secretion of estradiol and progesterone respectively (Fig. 4 a and 4 b). In accordance with the drop in progesterone secretion, a clear decrease in expression of the key steroidogenic enzymes P450scc (p < 0.05) and 3βHSD (p < 0.01) was also observed in cytochalasin-treated GC (Fig. 4 c and 4 d). All these effects of cytochalasin D on GC functions mimicked the actions of the anti-α6 IgG on GC cultured on LN substratum (Fig. 2, 3 and 4). Overall, cell rounding was associated with dramatic functional changes in GC. However, changes in proliferation rate and estradiol secretion were observed at cytochalasin D doses lower than 0.1 μg/ml, which did not induce observable changes in cell morphology.
Figure 2 Effect of cytochalasin D on morphology of GC cultured on LN substratum. GC from small and large follicles were cultured for 48 h on LN with or without (control) cytochalasin D at different concentrations (between 0.05 and 5 μg/ml) in culture medium, and then actin was stained with FITC-conjugated phalloidin. (a): representative microscopical fields of cultured GC; GC spread on LN in control or in presence of low concentrations of cytochalasin D; most cells adopted a spindle-shaped morphology at 0.5 μg/ml of cychochalasin D, then rounded up at higher doses. (b) and (c): percentages of unspread cells, i.e. spindle-shaped (dashed line) and round (solid line) GC from small (b) and large (c) follicles; empty bars: percentages of round cells in control; solid bars: percentages of round cells with anti-α6 IgG (0.5 μg/ml) in culture medium. Data represent mean ± SEM of 3 independent experiments. In each graph, different letters indicate significant differences (p < 0.001); *** : p < 0.001, with vs. without anti-α6 IgG.
Figure 3 Effect of cytochalasin D on cell numbers, proliferation and pyknotic rates of GC cultured on LN substratum. GC from small (a, c) and large follicles (b, d) were cultured for 144 h on LN as described in legend of Figure 2. (a) and (b): numbers of cells at 144 h of culture; (c): proliferation rates of GC from small follicles at 48 h of culture; (d): pyknotic rates of GC from large follicles at 48 h of culture. Empty bars: control; solid bars: with anti-α6 IgG (0.5 μg/ml) in culture medium. Data represent mean ± SEM of 6 independent experiments. In each graph, different letters indicate significant differences (p < 0.05); * : p < 0.05, *** : p < 0.001, with vs. without anti-α6 IgG.
Figure 4 Effect of cytochalasin D on steroidogenesis of GC from large follicles cultured on LN substratum. GC from large follicles were cultured for 144 h on LN as described in legend of Figure 2. (a) and (b): estradiol (E2) and progesterone (P4) secretions between 96 h and 144 h of culture; data are expressed as percentages of control (100%, empty bars, corresponding to 0.40 ± 0.17 ng/50 × 105 cells/48 h for E2 and 700 ± 176 ng/50 × 105 cells/48 h for P4) and represent mean ± SEM of 5 independent experiments; solid bars: with anti-α6 IgG (0.5 μg/ml) in culture medium; in each graph, different letters indicate significant differences (p < 0.05); *** : p < 0.001, compared to control. (c) and (d): expression of P450scc and 3βHSD enzymes in GC at 144 h of culture, in control or in the presence of anti-α6 IgG or cytochalasin D (0.5 μg/ml) in culture medium; results show representative western immunoblotting experiments performed on 5 μg of GC extracts; data correspond to quantification of autoradiograms in arbitrary units (control mean = 100) and represent mean ± SEM of 5 independent experiments; * : p < 0.05, ** : p < 0.01, treated vs. control.
From our results, cytochalasin D impeded ovine GC spreading on LN and induced cell rounding. These changes in cell shape were associated with a decrease in proliferation rate, an increase in cell death, a decrease in progesterone secretion and an increase in estradiol secretion. Previous investigations using inhibitors of actin microfilament polymerization, such as cytochalasin B or D, have shown the importance of actin cytoskeleton in GC functions. In GC, cytochalasin decreases actin and tropomyosin synthesis [29], causes the disorganization of microfilaments, intermediary filaments and microtubules [48] and induces cell rounding. These effects have been reported to mimick to some extent the effects of gonadotropins on the cytoskeleton [28,32,34]. To our knowledge, our study is the first report of the effects of cytochalasin on GC proliferation rate and cell death. In various cell models, such as kidney epithelial cells [49], hepatocytes [50] or carcinoma cells [51], cytochalasin has been shown to induce cell detachment from the substratum, to promote apoptosis and to inhibit cell proliferation. However, cytochalasin-induced changes in steroidogenesis are still a subject of controversy. Cytochalasin has been shown either to enhance (rat: [28,33,52]) or inhibit (hamster: [53], pig: [54], sheep: present study) progesterone secretion by GC and to inhibit it by luteal cells (cow: [55], pig: [56]). From the few data available concerning estradiol secretion, cytochalasin B is observed to block aromatase activity of hamsters GC [53] but to enhance it in immature rat Sertoli cells [57]. Whether these differences are the consequence of differences in cell types, animal species, or culture conditions is unknown. In any case, the functional effects of a disrupting drug such as cytochalasin must be considered with caution owing to its overall actions on cell metabolism and ion membrane exchanges. Moreover, experiments with cytochalasin have not allowed the specific role of LN-induced cytoskeleton changes on GC functions to be assessed. In our experiments, the observed effects of cytochalasin were likely to be the consequence of both cytoskeleton disruption and loss of interaction between cells and LN accompanying their detachment from the substratum.
Effects of specific inactivation of the α6β1 integrin signalization pathway on GC cultured on LN and/ or RGD peptide substratum
To further investigate the existence of a relationship between GC shape and function, an experiment was designed with the purpose of inactivating specifically the α6β1 integrin signalization pathway without changing GC shape. When cultured on substrata containing different ratios of LN and RGD peptides (varying between 100% and 0% LN), GC spread on all these different substrata (Fig. 5a and 5b). Addition of anti-α6 IgG in culture medium induced significant cell rounding only when the substratum contained at least 43% LN for cultures of GC from small follicles (p < 0.01), and 14% LN for cultures of GC from large follicles (p < 0.05) (Fig. 5 c and 5 d). In cultures of GC from small follicles, cell rounding induced by addition of anti-α6 IgG was associated with a decrease in cell number (p < 0.01 for cells cultured on a substratum containing at least 43% LN), a decrease in proliferation rate (p < 0.01 for cells cultured on a substratum containing at least 43% LN) and an increase in pyknotic rate (p < 0.05, only for cells cultured on 100% LN) (Fig. 6 a, b and 6 c). In cultures of GC from large follicles, cell rounding induced by addition of anti-α6 IgG was associated with a decrease in cell number (p < 0.01), an increase in pyknotic rate (p < 0.05), an increase in estradiol secretion (p < 0.001) and a decrease in progesterone secretion (p < 0.05), for cells cultured on a substratum containing at least 14% LN (Fig. 7 a, b, c and 7 d). Overall analysis of results of this experiment showed the existence of significant correlations between the percentage of round cells and the proliferation rate of GC from small follicles (r = -0.87, p < 0.001) on the one hand (Fig. 8a), and the pyknotic rate of GC from small follicles (r = 0.76, p < 0.05) and large follicles (r = 0.75, p < 0.05) on the other (Fig. 8b). In contrast, no significant correlation was found between the percentage of round cells and the amount of estradiol or progesterone secreted by GC from large follicles (Fig. 8c). These results support the existence of uncoupling of GC shape and steroidogenesis and suggest that mechanisms independent of cytoskeleton reorganization are also involved in α6β1 integrin-mediated LN actions on GC.
Figure 5 Effect of anti-α6 IgG on morphology of GC cultured on substrata containing different ratios of LN and RGD peptides. GC were cultured for 48 h on substrata containing percentages of LN varying between 100% (0% RGD peptides) and 0% (100% RGD peptides) in the mix, with or without anti-α6 IgG (0.5 μg/ml) in culture medium, and then actin was stained with FITC-conjugated phalloidin. (a) and (b): representative microscopic fields of cultured GC from small (a) and large (b) follicles. (c) and (d): percentages of round GC from small (c) and large (d) follicles; empty bars: control; solid bars: with anti-α6 IgG in culture medium; data represent mean ± SEM of 5 independent experiments; * : p < 0.05, ** : p < 0.01, *** : p < 0.001, with vs. without anti-α6 IgG.
Figure 6 Effect of anti-α6 IgG on cell numbers, proliferation and pyknotic rates of GC from small follicles cultured on substrata containing different ratios of LN and RGD peptides. GC from small follicles were cultured for 144 h as described in legend of Figure 5. (a): numbers of cells at 144 h of culture; (b): proliferation rates of GC at 48 h of culture; (c): pyknotic rates of GC at 48 h of culture. Empty bars: control; solid bars: with anti-α6 IgG in culture medium. Data represent mean ± SEM of 6 independent experiments. * : p < 0.05, ** : p < 0.01, *** : p < 0.001, with vs. without anti-α6 IgG.
Figure 7 Effect of anti-α6 IgG on cell numbers, pyknotic rates and steroidogenesis of GC from large follicles cultured on substrata containing different ratios of LN and RGD peptides. GC from large follicles were cultured for 144 h as described in legend of Figure 5. (a): numbers of cells at 144 h of culture; (b): pyknotic rates of GC at 48 h of culture. (c) and (d): estradiol and progesterone secretions between 96 h and 144 h of culture. Empty bars: control; solid bars: with anti-α6 IgG in culture medium. Data represent mean ± SEM of 6 to 8 independent experiments. * : p < 0.05, ** : p < 0.01, *** : p < 0.001, with vs. without anti-α6 IgG.
Figure 8 Relationships between percentages of round cells and proliferation rates, pyknotic rates or steroidogenesis of GC cultured on substrata containing different ratios of LN and RGD peptides. GC were cultured for 144 h as described in legend of Figure 5. (a): proliferation rates of GC from small follicles at 48 h of culture. (b): pyknotic rates of GC from small (solid circles) and large (empty circles) follicles at 48 h of culture. (c): steroid secretion by GC from large follicles between 96 h and 144 h; solid triangles: estradiol secretion, expressed in pg/50 × 103 cells/48 h, solid squares: progesterone secretion, expressed in ng/50 × 103 cells/48 h. Each data point is the mean of 7 independent experiments.
When GC spread on LN, cell proliferation and survival are clearly enhanced [13,14]. From our results, both proliferation and survival of GC cultured on LN, unlike steroidogenesis, were closely associated with cell shape and cytoskeleton organization. In other cell models, it has been established that cell spreading on ECM and associated changes in the actin cytoskeleton are necessary for progression through G1 and entry into the S phase of the cycle [50], and that signals elicited by the cytoskeleton act independently of ERK1/2 signals to drive the cell cycle machinery through the G1/S boundary and, hence to promote cell proliferation [58]. Interestingly, recent work has shown that the cytoskeleton orients the cell metabolic and signal transduction machinery, and that mechanical distortion of cells and the cytoskeleton through surface integrin receptors can switch cells between distinct gene programs (e.g. proliferation, differentiation and apoptosis) [59-61]. In ovarian follicles, GC lying directly on the basal lamina containing ECM adopt a columnar shape, whereas GC located in the follicular wall near the antral cavity are generally round [62-64]. These different cell morphologies are associated with different distributions of intracellular actin [65]. It can be hypothesized that these different cell shapes, resulting from cell interactions with ECM through integrin receptors, can directly influence GC functions. In particular, they could be responsible for establishing the gradient of survival and proliferation existing between the basal and the antral zone of the granulosa wall in ovarian follicles [64,66-68].
Effects of specific inactivation of the ERK1/2 signalization pathway on GC cultured on LN
To further study the mechanisms of action of LN on GC, we tested the importance of the ERK1/2 signalization pathway, which has been shown to transduce some of the α6β1 integrin-mediated effects of LN [36-39], in regulating both GC shape and functions. At 24 h of culture, phosphorylation of ERK1 and ERK2 was higher in cells cultured on LN than on the plastic substratum, indicating that LN might activate the ERK1/2 pathway in GC (p < 0.01, Fig. 9). Addition of PD98059, a specific inhibitor of the ERK1/2 activation pathway, inhibited ERK1/2 phosphorylation in GC cultured on LN substratum (p < 0.01) as expected, but the anti-α6 IgG and cytochalasin D had no effect (Fig. 9). The absence of effect of the function-blocking antibody raised against α6 subunit on ERK1/2 phosphorylation in GC cultured on LN was unexpected. It could be explained by the existence of other subunit integrins such as α3 and αv that are also able to bind LN [41] and have been shown to be present on GC ([19,69] and our unpublished observations). Whether these integrins participate in activation of the ERK1/2 signalization pathway when GC are attached to LN substratum remains to be studied. Alternatively, the ERK1/2 signalization pathway might respond poorly to α6β1 integrin activation, or the western immunoblotting method might not be sensitive enough to detect small quantitative changes in ERK1/2 phosphorylation.
Figure 9 Expression and phosphorylation of ERK1/2 in GC cultured on LN or on plastic substratum. GC were cultured for 24 h on plastic (Pl) or on LN in the presence or absence of anti-α6 IgG (0.5 μg/ml), PD98059 (30 μM) or cytochalasin D (0.5 μg/ml) in culture medium. (a): representative western immunoblotting experiment performed on 10 μg of GC extracts. P-ERK1/2: phosphorylated ERK1/2. (b): ratio P-ERK1/2 / ERK1/2, obtained by quantification of autoradiograms in arbitrary units ; data are expressed as percentages of LN condition (100%, empty bar) and represent mean ± SEM of 6 independent experiments. **: p < 0.01, vs. LN.
In cultures of GC from both small and large follicles, PD98059 induced no significant change in GC spreading on LN (Fig. 10 a and 10 b) or in the pyknotic rate of GC, and had only a weak inhibiting effect on GC proliferation rate, acting at the highest dose of 30 μM (p < 0.05, Fig 10 c and 10 d). In contrast, PD98059 strongly increased estradiol secretion (p < 0.001) and decreased progesterone secretion (p < 0.001) as well as P450scc (p < 0.05) and 3βHSD (p < 0.05) expression in GC from large follicles cultured on LN (Fig. 10 e, f, g and 10 h). These effects were not observed when GC were cultured on plastic (data not shown). Overall, these results suggest that the ERK1/2 signalization pathway might be involved in LN actions on GC steroidogenesis, independently of LN actions on cell shape. In contrast, GC survival and proliferation are probably closely associated with cell shape and cytoskeleton organization.
Figure 10 Effect of PD98059 on morphology, proliferation rates, pyknotic rates and steroidogenesis of GC cultured on LN. GC from small (a, c) and large (b, d, e, f, g, h) follicles were cultured for 144 h on LN in the presence or absence (control) of PD98059 at different concentrations (between 1 and 30 μM) in culture medium, and with or without anti-α6 IgG (0.5 μg/ml). (a, b): percentages of round cells at 48 h of culture; *** : p < 0.001, with anti-α6 IgG vs. control and PD98059-treated cells (30 μM). (c, d): proliferation rates and pyknotic rates of GC at 48 h of culture; empty bars: control; solid bars: with anti-α6 IgG in culture medium; in each graph, different letters indicate significant differences (p < 0.05); ** : p < 0.01, with vs. without anti-α6 IgG. (e, f): estradiol and progesterone secretions between 96 h and 144 h of culture; solid bars: with anti-α6 IgG in culture medium; data are expressed as percentages of control (100%, empty bars); in each graph, different letters indicate significant differences (p < 0.05) ; *** : p < 0.001 compared to control. (g, h): expression of P450scc (g) and 3βHSD (h) enzymes in GC at 144 h of culture, in control or in the presence of anti-α6 IgG or PD98059 (30 μM) in culture medium; results show representative western immunoblotting experiments performed on 5 μg of GC extracts; data correspond to quantification of autoradiograms in arbitrary units (control mean = 100); * : p < 0.05, ** : p < 0.01, treated vs. control. Data of the Figure represent mean ± SEM of 4 to 8 independent experiments.
In previous investigations, the ERK1/2 signalization pathway has been shown to be involved in the regulation of GC steroidogenesis, but its activation has either inhibiting or stimulating effects depending on the stimulus [70]. For example, ERK1/2 activation by gonadotropins inhibits Star protein expression and progesterone secretion, thereby contributing to desensitization mechanisms of hormonal action [71,72]. Likewise, prostaglandin F2alpha- or adenosine triphosphate-induced ERK1/2 activation inhibits hCG-stimulated progesterone secretion [73,74], whereas IGF-I- or melatonin-induced ERK1/2 activation increases it [75,76] and IGF-I-induced ERK1/2 activation enhances LDL receptor expression [77]. Interestingly, ERK1/2 have been shown to be intracellular signaling molecules that differentially regulate FSH-induced progesterone and estradiol synthesis in GC [78]. Thus, the divergent regulation of LN-induced progesterone and estradiol secretion by PD98059 observed in ovine GC supports the hypothesis that ERK1/2 is an important signalization pathway in the regulation of steroidogenesis by LN in GC.
Conclusion
The results of this study emphasize the role of cytoskeleton and cell shape in controlling GC proliferation and survival. In GC, as in many other cell types, mechanotransduction processes strongly control these basic cell functions [60]. From our results, LN participates in this control by its α6β1 integrin-mediated actions on GC cytoskeleton. In contrast, steroidogenesis is a specific function of fully differentiated GC which is under the control of various signalization pathways of which the cytoskeleton would be of minor importance compared to the cAMP or ERK1/2 pathways. It is suggested that changes in cytoskeleton and cell shape induced by actions of gonadotropins, growth factors or ECM components would accompany rather than cause changes in steroidogenesis. Interestingly, the ERK1/2 pathway could play an important role in mediating the action of LN on GC luteinization.
Authors' contributions
FLB, SF, CP and DM carried out the culture experiments together. FLB carried out the steroid assays and western immunoblotting experiments, participated in actin staining experiments, statistical analysis and interpretation of data, and participated in drafting the manuscript. SF participated in steroid assays and contributed to the conception and design of the study, the interpretation of data and the draft of the manuscript. CP carried out the thymidine labeling index and pyknotic index determination experiments and participated in actin staining experiments and interpretation of data. DM conceived the study, participated in its design and coordination, statistical analysis and interpretation of data and drafted the manuscript. FLB, SF and DM read and approved the final manuscript. CP died during preparation of the manuscript but her contribution was such that she deserves to be acknowledged as an author. This study is dedicated to her memory.
Acknowledgements
The authors thank F. Dupont and all the members of the ovine experimental unit for animal management and collaboration in the experimental design. This work was supported in part by a fund from the French organization "La Ligue contre le cancer". FLB was supported by a French fellowship, funded jointly by INRA and the "Région Centre".
==== Refs
Luck MR The gonadal extracellular matrix Oxford Reviews in Reproductive Biology 1994 16 33 85
Rodgers RJ van Wezel IL Irving-Rodgers HF Lavranos TC Irvine CM Krupa M Roles of extracellular matrix in follicular development J Reprod Fertil Suppl 1999 54 343 352 10692866
Rodgers RJ Irving-Rodgers HF Russell DL Extracellular matrix of the developing ovarian follicle Reproduction 2003 126 415 424 14525524 10.1530/rep.0.1260415
Wordinger RJ Rudick VL Rudick MJ Immunohistochemical localization of laminin within the mouse J Exp Zool 1983 228 141 143 6363605 10.1002/jez.1402280116
Huet C Monget P Pisselet C Monniaux D Changes in extracellular matrix components and steroidogenic enzymes during growth and atresia of antral ovarian follicles in the sheep Biol Reprod 1997 56 1025 1034 9096887
Le Bellego F Pisselet C Huet C Monget P Monniaux D Laminin-alpha6beta1 integrin interaction enhances survival and proliferation and modulates steroidogenesis of ovine granulosa cells J Endocrinol 2002 172 45 59 11786373 10.1677/joe.0.1720045
Irving-Rodgers HF Harland ML Rodgers RJ A novel basal lamina matrix of the stratified epithelium of the ovarian follicle Matrix Biol 2004 23 207 217 15296935 10.1016/j.matbio.2004.05.008
Leardkamolkarn V Abrahamson DR Immunoelectron microscopic localization of laminin in rat ovarian follicles Anat Rec 1992 233 41 52 1605377 10.1002/ar.1092330107
Alexopoulos E Shahid J Ongley HZ Richardson MC Luteinized human granulosa cells are associated with endogenous basement membrane-like components in culture Mol Hum Reprod 2000 6 324 330 10729314 10.1093/molehr/6.4.324
Lee VH Britt JH Dunbar BS Localization of laminin proteins during early follicular development in pig and rabbit ovaries J Reprod Fertil 1996 108 115 122 8958837
van Wezel IL Rodgers HF Rodgers RJ Differential localization of laminin chains in bovine follicles J Reprod Fertil 1998 112 267 278 9640266
Zhao Y Luck MR Gene expression and protein distribution of collagen, fibronectin and laminin in bovine follicles and corpora lutea J Reprod Fertil 1995 104 115 123 7636792
Aharoni D Meiri I Atzmon R Vlodavsky I Amsterdam A Differential effect of components of the extracellular matrix on differentiation and apoptosis Curr Biol 1997 7 43 51 8999995 10.1016/S0960-9822(06)00026-1
Huet C Pisselet C Mandon-Pepin B Monget P Monniaux D Extracellular matrix regulates ovine granulosa cell survival, proliferation and steroidogenesis: relationships between cell shape and function J Endocrinol 2001 169 347 360 11312151 10.1677/joe.0.1690347
Aten RF Kolodecik TR Behrman HR A cell adhesion receptor antiserum abolishes, whereas laminin and fibronectin glycoprotein components of extracellular matrix promote, luteinization of cultured rat granulosa cells Endocrinology 1995 136 1753 1758 7895687 10.1210/en.136.4.1753
Sites CK Kessel B LaBarbera AR Adhesion proteins increase cellular attachment, follicle-stimulating hormone receptors, and progesterone production in cultured porcine granulosa cells Proc Soc Exp Biol Med 1996 212 78 83 8618955
Sonnenberg A Modderman PW Hogervorst F Laminin receptor on platelets is the integrin VLA-6 Nature 1988 336 487 489 2973567 10.1038/336487a0
Honda T Fujiwara H Ueda M Maeda M Mori T Integrin alpha 6 is a differentiation antigen of human granulosa cells J Clin Endocrinol Metab 1995 80 2899 2905 7559872 10.1210/jc.80.10.2899
Giebel J de Souza P Rune GM Expression of integrins in marmoset (Callithrix jacchus) ovary during folliculogenesis Tissue Cell 1996 28 379 385 8760853
Giebel J Rune GM Relationship between expression of integrins and granulosa cell apoptosis in ovarian follicles of the marmoset (Callithrix jacchus) Tissue Cell 1997 29 525 531 9364802
Fujiwara H Ueda M Takakura K Mori T Maeda M A porcine homolog of human integrin alpha 6 is a differentiation antigen of granulosa cells Biol Reprod 1995 53 407 417 7492694
Nakamura K Fujiwara H Higuchi T Honda T Nakayama T Kataoka N Fujita K Ueda M Maeda M Mori T Integrin alpha6 is involved in follicular growth in mice Biochem Biophys Res Commun 1997 235 524 528 9207189 10.1006/bbrc.1997.6825
Kumar CC Signaling by integrin receptors Oncogene 1998 17 1365 1373 9779984 10.1038/sj.onc.1202172
Vuori K Integrin signaling: tyrosine phosphorylation events in focal adhesions J Membr Biol 1998 165 191 199 9767673 10.1007/s002329900433
De Loof A Vanden J Janssen I Hormones and the cytoskeleton of animals and plants Int Rev Cytol 1996 166 1 58 8881772
Amsterdam A Rotmensch S Structure-function relationships during granulosa cell differentiation Endocr Rev 1987 8 309 337 2820706
Amsterdam A Rotmensch S Ben-Ze'ev A Structure-function relationships in the differentiating granulosa cell Prog Clin Biol Res 1989 296 121 130 2662203
Ben-Ze'ev A Amsterdam A In vitro regulation of granulosa cell differentiation. Involvement of cytoskeletal protein expression J Biol Chem 1987 262 5366 5376 3104332
Ben-Ze'ev A Amsterdam A Regulation of cytoskeletal protein organization and expression in human granulosa cells in response to gonadotropin treatment Endocrinology 1989 124 1033 1041 2492216
Ben-Ze'ev A Kohen F Amsterdam A Gonadotropin-induced differentiation of granulosa cells is associated with the co-ordinated regulation of cytoskeletal proteins involved in cell-contact formation Differentiation 1987 34 222 235 2828140
Carnegie JA Tsang BK The cytoskeleton and rat granulosa cell steroidogenesis: possible involvement of microtubules and microfilaments Biol Reprod 1988 38 100 108 3365461
Sasson R Rimon E Dantes A Cohen T Shinder V Land-Bracha A Amsterdam A Gonadotrophin-induced gene regulation in human granulosa cells obtained from IVF patients. Modulation of steroidogenic genes, cytoskeletal genes and genes coding for apoptotic signalling and protein kinases Mol Hum Reprod 2004 10 299 311 15026540 10.1093/molehr/gah041
Tsang BK Li M Carnegie JA Microfilaments and FSH stimulation of rat granulosa cell steroidogenesis in vitro J Reprod Fertil 1988 83 263 271 3135389
Lawrence TS Ginzberg RD Gilula NB Beers WH Hormonally induced cell shape changes in cultured rat ovarian granulosa cells J Cell Biol 1979 80 21 36 217881 10.1083/jcb.80.1.21
Zhu X Evans JP Analysis of the roles of RGD-binding integrins, alpha(4)/alpha(9) integrins, alpha(6) integrins, and CD9 in the interaction of the fertilin beta (ADAM2) disintegrin domain with the mouse egg membrane Biol Reprod 2002 66 1193 1202 11906941
Chen CC Chen N Lau LF The angiogenic factors Cyr61 and connective tissue growth factor induce adhesive signaling in primary human skin fibroblasts J Biol Chem 2001 276 10443 10452 11120741 10.1074/jbc.M008087200
Ferletta M Kikkawa Y Yu H Talts JF Durbeej M Sonnenberg A Timpl R Campbell KP Ekblom P Genersch E Opposing roles of integrin alpha6Abeta1 and dystroglycan in laminin-mediated extracellular signal-regulated kinase activation Mol Biol Cell 2003 14 2088 2103 12802077 10.1091/mbc.E03-01-0852
Walker JL Zhang L Zhou J Woolkalis MJ Menko AS Role for alpha 6 integrin during lens development: Evidence for signaling through IGF-1R and ERK Dev Dyn 2002 223 273 284 11836791 10.1002/dvdy.10050
Wei J Shaw LM Mercurio AM Regulation of mitogen-activated protein kinase activation by the cytoplasmic domain of the alpha6 integrin subunit J Biol Chem 1998 273 5903 5907 9488728 10.1074/jbc.273.10.5903
Menezo Y [Synthetic medium for gamete survival and maturation and for culture of fertilized eggs] C R Acad Sci Hebd Seances Acad Sci D 1976 282 1967 1970 821667
Mercurio AM Laminin receptors: achieving specificity through cooperation Trends Cell Biol 1995 5 419 423 14732046 10.1016/S0962-8924(00)89100-X
Campbell BK Scaramuzzi RJ Webb R Induction and maintenance of oestradiol and immunoreactive inhibin production with FSH by ovine granulosa cells cultured in serum-free media J Reprod Fertil 1996 106 7 16 8667349
Monniaux D Pisselet C Control of proliferation and differentiation of ovine granulosa cells by insulin-like growth factor-I and follicle-stimulating hormone in vitro Biol Reprod 1992 46 109 119 1547308
Aharoni D Dantes A Oren M Amsterdam A cAMP-mediated signals as determinants for apoptosis in primary granulosa cells Exp Cell Res 1995 218 271 282 7537693 10.1006/excr.1995.1156
Saumande J Radioimmunoassay of estradiol-17 beta in unextracted ewe plasma Steroids 1981 38 425 437 7314159 10.1016/0039-128X(81)90077-5
Saumande J Culture of bovine granulosa cells in a chemically defined serum-free medium: the effect of insulin and fibronectin on the response to FSH J Steroid Biochem Mol Biol 1991 38 189 196 1900720 10.1016/0960-0760(91)90125-O
Saumande J Tamboura D Chupin D Changes in milk and plasma concentrations of progesterone in cows after treatment to induce superovulation and their relationships with number of ovulations and of embryos collected Theriogenology 1985 23 719 731 16726042 10.1016/0093-691X(85)90147-5
Sutovsky P Flechon JE Pavlok A F-actin is involved in control of bovine cumulus expansion Mol Reprod Dev 1995 41 521 529 7576620 10.1002/mrd.1080410415
Anderson RJ Ray CJ Popoff MR Evidence for Rho protein regulation of renal tubular epithelial cell function Kidney Int 2000 58 1996 2006 11044220 10.1111/j.1523-1755.2000.00372.x
Rescan C Coutant A Talarmin H Theret N Glaise D Guguen-Guillouzo C Baffet G Mechanism in the sequential control of cell morphology and S phase entry by epidermal growth factor involves distinct MEK/ERK activations Mol Biol Cell 2001 12 725 738 11251083
Pidgeon GP Tang K Rice RL Zacharek A Li L Taylor JD Honn KV Overexpression of leukocyte-type 12-lipoxygenase promotes W256 tumor cell survival by enhancing alphavbeta5 expression Int J Cancer 2003 105 459 471 12712435 10.1002/ijc.11134
Carnegie JA Byard R Dardick I Tsang BK Culture of granulosa cells in collagen gels: the influence of cell shape on steroidogenesis Biol Reprod 1988 38 881 890 3401543
Silavin SL Javitt NB Strauss JF Reevaluation of the effects of cytochalasins on steroidogenesis: studies on hamster granulosa cells Endocrinology 1984 115 1511 1516 6592092
Danisova A Kolena J Involvement of the cytoskeleton components in the secretion of luteinization stimulator by porcine granulosa cells Endocr Regul 1991 25 207 210 1813033
Gwynne A Condon WA Effects of cytochalasin B, colchicine, and vinblastine on progesterone synthesis and secretion by bovine luteal tissue in vitro J Reprod Fertil 1982 65 151 156 7200520
Gregoraszczuk EL Stoklosowa S The effect of microtubule and microfilament-disrupting drugs on prolactin-stimulated progesterone synthesis and secretion by cultured porcine theca cells Acta Histochem 1997 99 207 215 9248578
Meroni SB Schteingart HF Pellizzari EH Cigorraga SB Possible involvement of microfilaments in the regulation of Sertoli cell aromatase activity Mol Cell Endocrinol 1995 112 69 75 7589787 10.1016/0303-7207(95)03587-W
Huang S Ingber DE A discrete cell cycle checkpoint in late G(1) that is cytoskeleton-dependent and MAP kinase (Erk)-independent Exp Cell Res 2002 275 255 264 11969294 10.1006/excr.2002.5504
Han B Bai XH Lodyga M Xu J Yang BB Keshavjee S Post M Liu M Conversion of mechanical force into biochemical signaling J Biol Chem 2004 279 54793 54801 15485829 10.1074/jbc.M406880200
Ingber DE Tensegrity II. How structural networks influence cellular information processing networks J Cell Sci 2003 116 1397 1408 12640025 10.1242/jcs.00360
Mammoto A Huang S Moore K Oh P Ingber DE Role of RhoA, mDia, and ROCK in cell shape-dependent control of the Skp2-p27kip1 pathway and the G1/S transition J Biol Chem 2004 279 26323 26330 15096506 10.1074/jbc.M402725200
Chang SC Anderson W Lewis JC Ryan RJ Kang YK The porcine ovarian follicle. II. Electron microscopic study of surface features of granulosa cells at different stages of development Biol Reprod 1977 16 349 357 191113
Lipner H Cross NL Morphology of the membrana granulosa of the ovarian follicle Endocrinology 1968 82 638 641 5641554
Rodgers RJ Irving-Rodgers HF van Wezel IL Krupa M Lavranos TC Dynamics of the membrana granulosa during expansion of the ovarian follicular antrum Mol Cell Endocrinol 2001 171 41 48 11165009 10.1016/S0303-7207(00)00430-5
Cran DG Musk L The distribution of actin in sheep ovaries J Exp Zool 1985 235 375 380 4056698 10.1002/jez.1402350308
Fricke PM Ford JJ Reynolds LP Redmer DA Growth and cellular proliferation of antral follicles throughout the follicular phase of the estrous cycle in Meishan gilts Biol Reprod 1996 54 879 887 8924509
Marion GB Gier HT Choudary JB Micromorphology of the bovine ovarian follicular system J Anim Sci 1968 27 451 465 5646345
Monniaux D Mandon-Pepin B Monget P Follicular atresia, a programmed wastage M S-Medecine Sciences 1999 15 157 166
Yamada S Fujiwara H Honda T Higuchi T Nakayama T Inoue T Maeda M Fujii S Human granulosa cells express integrin alpha2 and collagen type IV: possible involvement of collagen type IV in granulosa cell luteinization Mol Hum Reprod 1999 5 607 617 10381815 10.1093/molehr/5.7.607
Dewi DA Abayasekara DR Wheeler-Jones CP Requirement for ERK1/2 activation in the regulation of progesterone production in human granulosa-lutein cells is stimulus specific Endocrinology 2002 143 877 888 11861509 10.1210/en.143.3.877
Seger R Hanoch T Rosenberg R Dantes A Merz WE Strauss JF Amsterdam A The ERK signaling cascade inhibits gonadotropin-stimulated steroidogenesis J Biol Chem 2001 276 13957 13964 11278318
Tajima K Dantes A Yao Z Sorokina K Kotsuji F Seger R Amsterdam A Down-regulation of steroidogenic response to gonadotropins in human and rat preovulatory granulosa cells involves mitogen-activated protein kinase activation and modulation of DAX-1 and steroidogenic factor-1 J Clin Endocrinol Metab 2003 88 2288 2299 12727988 10.1210/jc.2002-020913
Tai CJ Kang SK Choi KC Tzeng CR Leung PC Role of mitogen-activated protein kinase in prostaglandin f(2alpha) action in human granulosa-luteal cells J Clin Endocrinol Metab 2001 86 375 380 11232027 10.1210/jc.86.1.375
Tai CJ Kang SK Tzeng CR Leung PC Adenosine triphosphate activates mitogen-activated protein kinase in human granulosa-luteal cells Endocrinology 2001 142 1554 1560 11250936 10.1210/en.142.4.1554
Seto-Young D Zajac J Liu HC Rosenwaks Z Poretsky L The role of mitogen-activated protein kinase in insulin and insulin-like growth factor I (IGF-I) signaling cascades for progesterone and IGF-binding protein-1 production in human granulosa cells J Clin Endocrinol Metab 2003 88 3385 3391 12843192 10.1210/jc.2002-021965
Woo MM Tai CJ Kang SK Nathwani PS Pang SF Leung PC Direct action of melatonin in human granulosa-luteal cells J Clin Endocrinol Metab 2001 86 4789 4797 11600542 10.1210/jc.86.10.4789
Sekar N Veldhuis JD Concerted transcriptional activation of the low density lipoprotein receptor gene by insulin and luteinizing hormone in cultured porcine granulosa-luteal cells: possible convergence of protein kinase a, phosphatidylinositol 3-kinase, and mitogen-activated protein kinase signaling pathways Endocrinology 2001 142 2921 2928 11416012 10.1210/en.142.7.2921
Moore RK Otsuka F Shimasaki S Role of ERK1/2 in the differential synthesis of progesterone and estradiol by granulosa cells Biochem Biophys Res Commun 2001 289 796 800 11735115 10.1006/bbrc.2001.6052
| 15892896 | PMC1156948 | CC BY | 2021-01-04 16:37:13 | no | Reprod Biol Endocrinol. 2005 May 16; 3:19 | utf-8 | Reprod Biol Endocrinol | 2,005 | 10.1186/1477-7827-3-19 | oa_comm |
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-221594147910.1186/1477-7827-3-22ResearchTransforming growth factor beta-1 decreases the yield of the second meiotic division of rat pachytene spermatocytes in vitro Damestoy Anne [email protected] Marie-Hélène [email protected] Michèle [email protected] Odile [email protected] Philippe [email protected] INSERM U418; INRA UMR1245; Université Claude-Bernard Lyon 1, 29 rue sœur Bouvier, 69322 Lyon cedex 05, France2 Centre commun de Cytométrie en Flux, Faculté de Médecine, Université Jean Monnet, 42023 St Etienne, France2005 7 6 2005 3 22 22 29 3 2005 7 6 2005 Copyright © 2005 Damestoy et al; licensee BioMed Central Ltd.2005Damestoy et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
TGF beta and its receptors are present in both germ cells and somatic cells of the male gonad. However, knock-out strategies for studying spermatogenesis regulation by TGF beta have been disappointing since TGF beta-or TGF beta receptor-null mice do not survive longer than a few weeks.
Methods
In the present study, we addressed the role of TGF beta-1 on the completion of meiosis by rat pachytene spermatocytes (PS) cocultured with Sertoli cells. Identification and counting of meiotic cells were performed by cytology and cytometry.
Results
Under our culture conditions, some PS differentiated into round spermatids (RS). When TGF beta-1 was added to the culture medium, neither the number of PS or of secondary spermatocytes nor the half-life of RS was modified by the factor. By contrast, the number of RS and the amount of TP1 mRNA were lower in TGF beta-1-treated cultures than in control cultures. Very few metaphase I cells were ever observed both in control and TGF beta-1-treated wells. Higher numbers of metaphase II were present and their number was enhanced by TGF beta-1 treatment. A TGF beta-like bioactivity was detected in control culture media, the concentration of which increased with the time of culture.
Conclusion
These results indicate that TGF beta-1 did not change greatly, if any, the yield of the first meiotic division but likely enhanced a bottleneck at the level of metaphase II. Taken together, our results suggest strongly that TGF beta participates in an auto/paracrine pathway of regulation of the meiotic differentiation of rat spermatocytes.
==== Body
Background
Multiplication, differentiation and survival or death of testicular germ cells are tightly regulated processes. Over the last decades it has become obvious that, in addition to the regulation exerted by the pituitary hormones (mainly FSH and LH) [1], spermatogenesis is under the control of a complex network of factors originating from both the somatic cells and the germ cells of the testis [2,3]. Moreover, it is becoming clear that hormones and intratesticular factors may compensate at least in part, for the absence of some hormones or factors, including FSH [4-6] and androgen [7-10] or luteinizing hormone [11] receptors. Thus, it is likely that synergism and/or redundancy between regulatory molecules is a characteristic of the spermatogenic process. Since most of the growth factors, cytokines and neurotrophins produced within the testis are widely expressed in the organism, the attempts to understand their role in spermatogenesis by knock-out strategies have been often disappointing. Transforming growth factor (TGFβ) is an example of such molecules. TGFβ1, TGFβ2 and TGFβ3 are expressed in the male gonad and their receptors are present in the rat testis in both somatic cells and germ cells [12-16]. TGFβ- or TGFβ receptor-null mice have been created [17-23]. However, as these mice do not survive longer than a few weeks, their usefulness for studying spermatogenesis is limited. Hence, use of culture systems associating spermatogenic cells and testicular somatic cells might be a valuable alternative to study the possible involvement of intratesticular factors such as the TGFβs on some step(s) of spermatogenesis. We [24,25] and others [26-28] have demonstrated that meiosis can proceed in vitro when mammalian spermatogenic cells are cocultured with Sertoli cells. In rodents, the kinetic of the meiotic process is similar in vivo and during the first week of culture [24,26,29] and round spermatids developed in vitro can produce normal offspring [30]. Such culture systems have allowed to study some cellular aspects of the meiotic process [31] or of its regulation [32]. Hence, in the present work, we addressed the role of TGFβ1 on the completion of meiosis by rat pachytene spermatocytes cultured together with Sertoli cells.
Methods
Isolation and co-culture of rat Sertoli cells and pachytene spermatocytes or round spermatids
Co-cultures of 21-day old rat Sertoli cells and adult rat pachytene spermatocytes (PS) or round spermatids (RS) were performed in bicameral chambers as described by Weiss et al. [24]. More than 90% of the cells in the PS fraction were 4C cells and more than 80% in the RS fraction were 1C cells; moreover, we found previously that 13% of elutriated PS were early PS (stages XIV-IV), 61% were middle PS (stages V-IX) and 26% were late spermatocytes (stages X-XIII) [32]. In some experiments, adult rats were injected with 5-bromo-2'-deoxyuridine (BrdU, Sigma-Aldrich, St Quentin Fallavier, France) (50 mg/kg) 14 days before they were killed, in order to label PS/diplotene spermatocytes of stages VII-XIII [32]. Under these conditions, BrdU-labelled stages XI, XII, and XIII (the only stages containing spermatocytes which may have time to differentiate into RS over a 3-day culture period [24,26,29]) represented 18% of the total number of stages containing BrdU-labelled PS/diplotene spermatocytes. These procedures were approved by the Scientific Research Agency (approval n° 69306) and conducted in accordance with the guidelines for care and use of laboratory animals.
Cells were cultured in Dulbecco's Modified Eagle's Medium (F12/DMEM) supplemented with antibiotics, NaHCO3 (13.4 mM), insulin (10 μg/ml), transferrin (10 μg/ml), vitamin C (10-4M), vitamin E (10 μg/ml), testosterone (10-7M), retinoic acid (3.3 × 10-7 M), retinol (3.3 × 10-7 M), pyruvate (1 mM) (all products from Sigma-Aldrich, St Quentin Fallavier, France), and ovine NIH FSH-20 (1 ng/ml) obtained through NIADDK at 33°C in an humidified atmosphere of 95% air : 5% CO2. In TGFβ1-treated wells, human recombinant TGFβ1 (10 ng/ml) (R&D Systems, Lille, France) was added in both the apical and the basal compartments of the bicameral chamber. Only basal media (with or without 10 ng/ml TGFβ1) were renewed every two days; TGFβ1 was added to the apical media at the same interval.
Immunolabeling of cells for flow cytometry
At selected days of culture, cells were detached from culture wells by trypsinization. After counting and determination of cell viability by Trypan blue exclusion, cells were fixed with ice cold 70% ethanol.
Immunolabeling of cultured cells and flow cytometric analyses were performed as described by Vigier et al. [32] and Godet et al. [33]. After washing with phosphate-buffered saline (PBS), fixed cells were resuspended in permeabilizing buffer (0.25% Triton X-100/1% BSA/PBS) for 20 minutes on ice. The cells were then exposed to an anti-vimentin antibody (clone V9 – Dako SA, Trappes, France) at a dilution of 1:500 in blocking buffer (5% fetal calf serum/1% BSA/PBS) for 3 h at 4°C. After 3 washes in 1% BSA/PBS, the cells were incubated with fluorescein (FITC)-conjugated rabbit anti-mouse immunoglobulins (Dako SA, Trappes, France) diluted 1:60 in blocking buffer for 1 hour at 4°C. Before analysis, Hoechst 33342 (Sigma-Aldrich, St Quentin Fallavier, France) was added to the labeled cell suspension at a final concentration of 20 μg/ml.
Flow cytometric and computer analyses
After immunolabeling, cells were analysed using a FacsStar plus Cytometer (Becton Dickinson, Le Pont de Claix, France) equipped with a 50-mW argon laser tuned to 448 nm and an INNOVA 300 ion multilined/UV laser tuned to UV. Emission of fluorescence was measured with a DF 530/30 filter for FITC, and a DF 42/44 filter for Hoechst 33342. Data acquisition and analysis were performed with CellQuest software (Becton Dickinson, Le Pont de Claix, France). Four data parameters were acquired and stored in list mode files: linear forward light scatter (FSC) and linear side angle light scatter (SSC) which roughly represent cell size and cellular granularity, respectively; logarithmic FITC (vimentin) to detect immunolabeling, and linear Hoechst to measure the DNA content of the different cell populations. Contaminating events such as debris and clumped cells were eliminated from the analysis. Each acquisition was performed on 50 000 cells negative for vimentin and allowed to identify 1C, 2C and 4C germ cells. Hence, the percentage of each category of germ cells was multiplied by the total number of cells per well in order to obtain the absolute number 1C, 2C and 4C cells.
RNA extraction
Total RNA from cultured cells was extracted according to Chomczynski and Sacchi [34], using Trizol Reagent (Invitrogen, Cergy-Pontoise, France). The purity and the integrity of the RNA were checked spectroscopically and by gel electrophoresis respectively.
Reverse transcription-polymerase chain reaction (RT-PCR)
Sequences corresponding to TP1 mRNA [EMBL: X07284] or p21 mRNA [EMBL: NM080782] and 16S rRNA [EMBL: XM341815] were amplified by RT-PCR as follows. Single strand cDNAs were obtained from reverse transcription (RT) of 1 μg of total RNA in a final volume of 10 μl containing 100 IU of Moloney murine leukaemia virus reverse transcriptase (MMLVRT), 1.25 mM dNTPs, 0.4 μg of oligo-dT, 10 mM DTT and 15 IU of RNase Guard (all products from Invitrogen, Cergy-Pontoise, France). RT was performed for 1 hour at 37°C. A sample in which RNA was replaced by sterile water was used as a negative control.
TP1/16S
coamplification reactions were carried out on 80 ng of reverse-transcribed RNA in a final volume of 30 μl containing 1,5 IU of Taq DNA Polymerase (Roche Diagnostic, Mannheim, Germany), 2 μM dNTPs, 1 μCi [33P]dATP (Amersham Biosciences, Orsay, France) and 12.5 pmol of each of the sequence-specific primers for TP1 and 16S (Sigma-Aldrich, St Quentin Fallavier, France) (Table 1). Amplification was performed in 5 sequential cycles (16, 18, 19, 20 and 22 cycles) at 94°C for 1 minute, 58°C for 1 minute and 72°C for 1 minute.
Table 1 Sequences of the primers used to amplify mRNAs for TP1, p21 and 16S RNA
RNA Primers Product size(nt)
TP1 [EMBL: X07284]* 5'26 – CCAGCCGCAAACTAAAGACTCATGG – 3' 175
5'200 – AGCTCATTGCCGCATTACAAGTGGG – 3'
p21 [EMBL: NM080782]* 5'107 – GACCTGTTCCACACAGGAGCAAAGT – 3' 145
5'251 – GTCTCAGTGGCGAAGTCAAAGTTCC – 3'
16S [EMBL: XM341815]* 5'138 – TGGGCTCATCAAGGTGAATGG – 3' 268
5'405 – GGTCCGATCGTACTGGATGAGGATA – 3'
nt : nucleotides
* gene bank accession numbers.
p21/16S
coamplification reactions were carried out on 80 ng of reverse-transcribed RNA in a final volume of 30 μl containing 1,5 IU of Taq DNA Polymerase (Roche Diagnostics), 3 μM dNTPs, 1.5 μCi [33P]dATP (Amersham Biosciences) and 12.5 pmol of the sequence-specific primers for p21 (Sigma-Aldrich) (Table 1). In a first step, amplification was performed for 6 cycles at 94°C for 1 minute, 58°C for 1 minute and 72°C for 1 minute. Then, 12.5 pmol of the sequence-specific primers for 16S were added and amplification was continued and performed in 5 sequential cycles (20, 22, 24, 26 28 cycles).
The amplified products were electrophoresed in parallel with size markers on 3% Metaphor Agarose gels (Tebu, Le Perray-en-Yvelines, France). The bands were cut from the gels and melted in distilled water at 100°C. Radioactivity was measured using a liquid scintillation counter.
TGFβ bioassay
TGFβ concentrations in apical culture media were determined using mink lung epithelial cell (CCl-64), according to Danielpour et al. [35]. Briefly, CCl-64 cells were maintained in 75 cm2 flasks in F-12 DMEM medium supplemented with 10% foetal calf serum. In 0.32-cm2 wells (96-well plate), aliquots (5–10 μl) of conditioned medium heated (or not) at 80°C for 5 minutes were added in 100 μl assay medium supplemented with 5% foetal calf serum. After 1 hour at 37°C, CCl-64 cells in logarithmic growth phase were trypsinised, washed once with assay medium, and plated at 3.104 cells/100 μl in wells. CCl-64 cells were cultured overnight, followed by pulse labelling with [3H]thymidine (40–60 Ci/mmol; Amersham Biosciences, Orsay, France) at a final concentration of 1 μCi/ml for 4 hours at 37°C. The cells were fixed three times with 250 μl methanol-acetic acid (3:1, vol/vol) and were washed twice with 80% methanol. 3H-labeled DNA was extracted by 250 μl 0.4% deoxycholate 0.5 N NaOH. Radioactivity was measured by a liquid scintillation counter. In some cases TGFβ bioactivity of culture media was determined in the presence of a TGFβ1/2/3 blocking antibody (clone 1D11, R&D Systems, Lille, France).
Immunocytochemical studies on cultured cells
At selected days of culture, cells were fixed directly in the well with Bouin's fixative for 20 minutes at room temperature.
1-Immunocytochemical reaction against phosphorylated Histone H3 (pH3)
After washing with PBS, fixed cells were permeabilized with 0.03% TritonX-100 in PBS for 30 minutes. The immunocytochemical reaction was performed with the Dako LSAB kit (Dako SA, Trappes, France) as follows. Cells were incubated with 3% H2O2 for 5 minutes, rinsed with PBS, then incubated with a monoclonal mouse anti-pH3 antibody (Upstate Biotechnology, Euromedex, Mundolsheim, France) at 1:200 dilution in Dako's antibody diluent for 3 hours at room temperature. After three washes with PBS, the staining reaction was performed using the avidin-biotin-peroxydase complex and amino-3-ethyl-9-carbazole (AEC) as a chromogen.
2-BrdU immunochemical reaction
The same protocol as described above was performed, but denaturation of DNA was required before the immunoreaction. Cells were incubated for 5 minutes with NaOH 0,07N diluted in alcohol:water (v:v) and then dipped in PBS for 10 minutes. Cells were incubated with a monoclonal mouse anti-bromodeoxyuridine antibody (Dako SA, Trappes, France) diluted 1:100 overnight at 4°C. After three washes with PBS, the staining reaction was performed using avidin-biotin-peroxidase complex and AEC as a chromogen.
The membrane of the insert was cut off and mounted in Dako aqueous mounting medium.
The morphologic identification and counting of the different types of germ cells was performed as previously described [32].
Statistical analysis
ANOVA was used to compare between data from more than two groups. Paired Student's t test was used to assess statistical differences between control and TGFβ1-treated cells.
Results
Effects of TGFβ1 on the number and viability of total germ cells and somatic cells
When about 3.105 purified PS were cocultured with about 3.105 Sertoli cells, both the number of total cells (somatic cells plus germ cells) and their viability decreased slightly during the culture period, as expected [24]. For both parameters, the decreases were similar in control and TGFβ1-treated cultures. On day 7, the percentages of cells, compared with day 1, were 79 ± 8 and 79 ± 10% (m ± SEM, n = 7) for controls and treated cells; the viabilities were 71 ± 4 and 69 ± 3%, respectively. Every day they were studied, the percentages of germinal cells and of somatic cells were similar in control and TGFβ1-treated cultures but the percentage of germ cells decreased during the culture, whereas that of somatic cells increased 1.3 fold (data not shown). This indicates that germinal cells were lost at a similar rate irrespective the culture conditions, whereas the number of somatic cells remained constant during the culture period.
Effects of TGFβ1 on the number of germ cells of different ploidy and on the amount of TP1 mRNA
The number of seeded PS (4C cells) decreased 2.5 fold between day 1 and day 7 in control and TGFβ1-treated cultures (both p < 0.01) and no difference was observed between these two conditions on any studied day (Fig. 1). Under our culture conditions, some of the PS differentiated into RS (1C cells) [24,31,32]. However, when TGFβ1 was added to the culture medium, the number of in vitro formed RS was significantly lower on days 5 and 7 than in control cultures (p < 0.01 and p < 0.05 respectively). By contrast, no significant difference was ever observed in the population composed of secondary spermatocytes and of doublets of RS (2C cells) (Fig. 1). In addition, the amount of TP1 mRNAs (specific to the haploid state), assessed by the TP1/16S ratio on day 14, of TGFβ1-treated cultures was lower than that of controls (p < 0.05) (Fig. 2).
Figure 1 Changes in the number of germ cells of different ploidy in cocultures of Sertoli cells and PS (4C cells) maintained in the absence (control) or presence of 10 ng/ml TGFβ1 (TGFβ1). Results are the mean ± SEM of seven experiments. ** p < 0.01 vs control; * p < 0.05 vs control.
Figure 2 Effect of TGFβ1 on the TP1 mRNA/16S rRNA ratio on day 14 of cocultures of Sertoli cells and PS. PS were cocultured with Sertoli cells for 14 days either in the absence (control) or presence of 10 ng/ml TGFβ1. On day 14 of coculture, total RNA was extracted and TP1 mRNA and 16S rRNA were coamplified by RT-PCR, as described in the Materials and Methods section, and their ratio were calculated. Each point is the mean ± SEM of four experiments. * p < 0.05 vs control.
Effects of TGFβ1 on the survival of RS cultured with Sertoli cells
Taken together, these results suggest that TGFβ1 inhibited the differentiation of PS into RS and/or affected the survival of RS. Therefore, in the next series of experiments, purified RS were cocultured with Sertoli cells for 5 days either in the absence or presence of TGFβ1 and the number of RS was determined daily between days 2 and 5. As expected [24,32], there was a steady decrease in the number of RS during the culture period. The regression line fitting the number of remaining RS on each day allowed the calculation of the half-life of RS which was identical in the absence or presence of TGFβ1 (Table 2).
Table 2 Half-lives of RS cocultured with Sertoli cells in the absence or presence of TGFβ1
Treatment Half-life (days)
Control 1.9 ± 0.2
TGFβ1 1.9 ± 0.2
Elutriated RS were cocultured with Sertoli cells for 5 days in the absence (control) or presence of 10 ng/ml TGFβ1 and the number of remaining RS was assessed by flow cytometry daily between day 2 and day 5 allowing the determination of their half-life in culture. Each value is the mean ± SEM of three different experiments.
Effects of TGFβ1 on the number of BrdU-labeled meiotic germ cells and the number of meiotic metaphases
Hence, the above results indicate that most likely, TGFβ1 was able to inhibit, at least partly, some step(s) in the differentiating pathway of PS into RS. In an attempt to identify this step, in the next experiments, BrdU-labeled PS/diplotene spermatocytes of stages VII-XIII were cocultured with Sertoli cells in the absence or presence of TGFβ1 ; then, the number of BrdU-labeled PS, BrdU-labeled secondary spermatocytes (SII) and that of BrdU-labeled RS were assessed by microscopic examination. The data presented in Table 3 show that TGFβ1 treatment modified neither the number of PS nor the number of SII on any studied day, but decreased the number of RS on day 3 of culture. In addition, the number of metaphases seemed to increase a little on day 3 in the presence of TGFβ1.
Table 3 Numbers of BrdU-labeled PS, of BrdU-labeled secondary spermatocytes (SII) and BrdU-labeled RS on day 1 and day 3 of coculture of BrdU-labeled PS/diplotene spermatocytes of stages VII-XIII and Sertoli cells in the absence or presence of TGFβ1
Number of PS/mm Number of SII/mm Number of RS/mm Number of metaphases
Days of culture - + - + - + - +
1 (n = 3) 67.6 ± 14.5 NS 63.9 ± 17.4 1.3 ± 0.3 NS 1.7 ± 0.3 0.1 ± 0.1 NS 0.1 ± 0.1 0 NS 0
3 (n = 7) 42.7 ± 5.1 NS 33.4 ± 6.1 4.1 ± 0.8 NS 3.8 ± 0. 9 5.7 ± 1.3 p < 0.05 4.2 ± 1.3 1.3 ± 0.2 NS 1.8 ± 0.3
BrdU-labeled PS/diplotene spermatocytes were cocultured for 3 days in the absence (-) or presence (+) of 10 ng/ml TGFβ1. On days 1 and 3, the numbers of BrdU-labeled PS, BrdU-labeled SII, BrdU-labeled RS and BrdU-labeled metaphases were determined by microscopic examination of the culture wells and expressed as the number of each cell type per mm of well and on the width of the microscope field (objective × 100). About 500 BrdU-labeled cells were counted in each condition and in each experiment.
Each value is the mean ± SEM ; n : number of experiments.
NS : not significant.
Therefore in the next experiments, cultured cells were stained with a Ser 10 phosphorylated histone H3 antibody in order to label cells in the division phase [36,37] and the numbers of metaphase I and metaphase II were recorded at the level of the whole germ cell population, not only BrdU-labelled cells (Fig. 3 and Table 4). Very few meiotic metaphase I cells were ever observed both in control and in TGFβ1-treated cultures precluding meaningful statistical analysis. By contrast, higher numbers of metaphases II were present on any studied day. Furthermore, the number of metaphase II was significantly enhanced over control by TGFβ1 on days 3 and 5 of the experiment (p < 001 and p < 0.05 respectively).
Table 4 Numbers of metaphase I and metaphase II on different days of coculture of PS and Sertoli cellsin the absence or presence of TGFβ1
Number of metaphases I/mm Number of metaphases II/mm
Days of coculture Control TGFβ1 Control TGFβ1
1 0.09 ± 0.02 NT 0.03 ± 0.01 1.8 ± 0.3 NS 1.8 ± 0.3
3 0.15 ± 0.04 NT 0.05 ± 0.03 2.3 ± 0.3 p < 0.01 3.1 ± 0.4
5 0.04 ± 0.02 NT 0.03 ± 0.01 0.7 ± 0.1 p < 0.05 0.9 ± 0.2
PS were cocultured for 5 days in the absence (control) or presence of 10 ng/ml TGFβ1. On days 1, 3 and 5, the numbers of metaphase I and of metaphase II were determined by microscopic examination of the culture wells and expressed as the number of each type per mm of well and on the width of the microscope field (objective × 100). 400 to 1700 metaphase II were counted in each condition, whereas the number of metaphase I which could be counted ranged between 15 and 75 precluding meaningful statistical analysis.
Each value is the mean ± SEM of 5 different experiments.
NS : not significant ; NT : not tested.
Figure 3 Staining of metaphase I (left) and metaphase II (right) by a Ser 10 phosphorylated histone H3 antibody in cocultures of PS and Sertoli cells (day one of coculture; bars represent 10 μm).
Effects of TGFβ1 on the expression of p21 mRNA
In an attempt to explore the mechanism of this inhibitory effect of TGFβ1 on the cell cycle of the meiotic cells, in the following experiments the amount of mRNA encoding for the cyclin-dependent kinase inhibitor p21waf1/cip1 was assessed on days 1, 2, and 3 of cocultures performed in the absence or presence of TGFβ1. The p21/16S ratio was similar in control and TGFβ1-treated cultures on any tested day (data not shown).
Changes in the concentration of TGFβ bioactivity during coculture of PS with Sertoli cells
Secretion of TGFβ1 by Sertoli cell-germ cell cocultures has been demonstrated by western blotting [38]. Therefore in the following experiments we studied whether a TGFβ-like bioactivity was released under our culture conditions. No TGFβ-like bioactivity could be detected in the media of the apical compartments of the culture chambers prior heat-treatment (data not shown). By contrast significant amounts were observed following heat activation; the concentration increased almost linearly with the time of culture from 0.8 ± 0.1 on day 3, up to 4.0 ± 0.2 ng/ml on day 15 (m ± SEM, n = 3). This bioactivity was completely blocked by a TGFβ neutralizing antibody directed against TGFβ1,2,3 (data not shown).
Discussion
To our knowledge, there has been no previous study on the effect of TGFβ on the meiotic differentiation of male germ cells. Yet it has been shown that germ cells synthesize TGFβ and are targets for the peptide (see [39] for review). However, TGFβ-null mice die in pre-or neonatal stages, making it virtually impossible to investigate the significance of TGFβ in the late stages of spermatogenesis [17-23]; hence we have studied the effect of TGFβ on meiotic differentiation of spermatogenic cells by using a system of culture of rat PS previously established in our laboratory [24,31,32]. No effect of TGFβ1 was observed on the number of remaining PS or Sertoli cells on any day under our culture conditions. Likewise TGFβ1 did not modify the half-life of seeded RS. Hence, it seems reasonable to conclude that the lower number of RS observed in TGFβ1-treated wells was due to an inhibitory effect of this factor on the transformation of PS into RS. This assumption appears substantiated by the lower content in TP1 mRNAs (specific to the haploid state) of TGFβ1-treated cultures of 14 days. Screening of TP1 mRNA content was performed on day 14 for two reasons : i) the total number of RS in TGFβ1-treated cultures was significantly lower than in controls only from day 5 onward (Fig. 1). ii) We showed previously that expression of TP1 appears to be important only from steps 5–6 of spermiogenesis, that is, 4–5 days after completion of meiosis [40]. However, the possibility that TGFβ1 might decrease the transcription of TP1 and/or the half-life of its mRNA cannot be excluded.
An effect of TGFβ1 was observed only beyond one day of culture, which is the time necessary to the more advanced elutriated spermatocytes (stage XIII) to differentiate into secondary spermatocytes under our culture conditions [31], and affected the 1C-cell population but not the 2C-cell population. Moreover, cytological monitoring of the fate of BrdU-labeled PS showed that the numbers of SII were similar in the absence or presence of TGFβ1, whereas the number of BrdU-labeled metaphase seemed enhanced by this factor. This latter point was confirmed, at the level of the whole germ cell population, by use of a Ser 10 phosphorylated histone H3 antibody which allowed us to show that the number of second metaphases was enhanced in the presence of TGFβ1. Taken together, these results suggest that TGFβ1 did not change greatly, if any, the yield of the first meiotic division, but likely enhanced a bottleneck resulting in an increase in the number of metaphase II.
Progression of cells through the cell cycle requires assembly and activation of cyclin-CDK complexes [41]. CDK activity can be regulated by changes in cyclin or CDK levels, by post-translational modifications of the CDK subunit and by formation of a complex with a set of proteins called CKI [42]. P21waf1/cip1 belongs to the CKI family and binds to CDK2-cyclin A/E complexes [43] which are active during the G1 phase, S phase and/or the G2/M transition. Moreover, CDK2 and cyclin A1 appear essential for meiosis [44-46]. Several studies have shown that TGFβ1 induces p21waf1/cip1 expression [47,48]. In the testis, Beumer et al. [49] have suggested that p21waf1/cip1 could be important during the meiotic phase of spermatocytes as it is present in PS of stage V up to step 5 RS. As the work stands, we did not observe an increase in the amount of p21waf1/cip1 mRNA in TGFβ1-treated cultures. However, additional mechanisms appear able to mediate the cell cycle arrest caused by TGFβ [43,50-52].
In vivo, TGFβ1 is present mainly in spermatocytes and RS [13,53] and rat testicular germ cells and Sertoli cells release different types of bioactive TGFβs in vitro ( [12,38,53] and present results).
Immunohistochemistry studies have localized both type I (TβRI) and type II (TβRII) transducing receptors for TGFβ to the seminiferous epithelium in the rat testis [12,15,16]. Moreover, Smad2 and Smad3 that mediate intracellular signaling of the TGFβ superfamily from the cell surface to the nucleus are present in meiotic cells [54,55] and TGFβ can regulate gene expression in these cells [56]. The amounts of TGFβ-like bioactivity measured in our culture media fit quite well with those reported in cell culture supernatants of different testicular cell types by Haagmans et al. [53]. Moreover, these concentrations are quite in the range of the TGFβ concentrations required to regulate Cyp19 gene expression in isolated rat PS and RS [56] or to perturb the inter-Sertoli tight junction permeability barrier in vitro [57]. Furthermore, the work of Haagmans et al. [53] indicates that latent TGFβ released by primary cultures of germ cells and Sertoli cells can be converted into a bioactive form in the culture medium. Taken together, these results strongly suggest that TGFβ can be involved in an auto/paracrine pathway of regulation of the meiotic differentiation of rat spermatocytes. In experiments not reported here, we tried to inhibit the meiotic action of "endogenous" TGFβ with the anti-TGFβ antibody we used in the above bioassay ; however, the acid wash required to remove the antibody before counting the germ cells by either method precluded accurate analyses.
It might be argued that the effects of TGFβ1 observed in the present study are relatively minor. However, it must be underlined that their amplitude is quite similar to the effects of TGFβ on [3H]thymidine incorporation by rat seminiferous tubule segments [58] or on the reduction of the number of rat gonocytes [59]. Moreover, it must be emphasized that the rather high standard errors of means, both in control and TGFβ1-treated cultures, were much more due to variations between the different experiments, as occurs often in vitro, than to variations of the effect of TGFβ1 treatments. Indeed, for instance, of the seven experiments reported in Fig. 1 or in Table 3 (day 3), six or seven exhibited a lower number of RS in TGFβ1-treated wells vs controls. Therefore it is not surprising that statistical analysis resulted in significant differences. In addition, it is clear that synergism and/or redundancy between local regulatory factors are a characteristic of the spermatogenic process ([2,32] and unpublished results). Hence it is not unexpected that many regulatory molecules, produced within the testis, may have somewhat limited effect when tested in the presence of the other factors regulating, in the same sense, the same step of spermatogenesis. As for the effects of TGFβ1 observed in the present work, two points should be emphasized: i)it is most likely that the amplitude of the effects of exogenous (added) TGFβ1 were limited by the TGFβ produced by our germ cell-Sertoli cell cocultures (see above); ii) in other studies not reported here, we observed that NGF, which is produced by PS and RS both in vivo [60, 61] and under our culture conditions (M.H. Perrard, unpublished results) is also able to negatively regulate the meiotic divisions of rat PS. Thus, it is very likely, that the presence of endogenous NGF also limited the effects of TGFβ1 observed in the present studies. Indeed, such a synergism/redundancy appears to be a major problem when exploring the local regulations of spermatogenesis which makes the knowledge of this topic far from being complete. Additional studies are now required to understand the mechanism of this action of TGFβ1 and why it is likely to be on the second meiotic division.
Conclusion
These in vitro results, together with previous studies showing the presence of TGFβ and its receptors in both the germ cells and the somatic cells of the male gonad, suggest strongly that TGFβ1 participates in an auto/paracrine pathway of regulation of the meiotic differentiation of rat spermatocytes.
Authors' contributions
AD participated in the design of the study, performed cultures and immunocytochemical and flow cytometry analyses and carried out PCR experiments. MHP participated in the design of the study, in the cultures and performed immunocytochemical studies. MV participated in the design of the study, in the cultures and in the PCR experiments, and performed TGFβ bioassays. OS carried out flow cytometry analyses. PD designed and coordinated the experiments, participated in the cultures, performed statistical analyses and drafted the manuscript. All authors read and approved the manuscript.
Acknowledgements
This work was supported by Institut National de la Santé et de la Recherche Médicale, Institut National de la Recherche Agronomique and Université Claude-Bernard Lyon I. AD was supported by the Ministère de l'Education Nationale, de l'Enseignement Supérieur et de la Recherche. We thank T. O'Neill for English revision and J. Bois for excellent secretarial assistance.
==== Refs
McLachlan RI O'Donnell L Meachem SJ Stanton PG de Kretser DM Pratis K Robertson DM Identification of specific sites of hormonal regulation in spermatogenesis in rats, monkeys, and man Recent Prog Horm Res 2002 57 149 179 12017541 10.1210/rp.57.1.149
Mauduit C Benahmed M Hamamah S and Mieusset R Growth factor in the testis : development and function Research in Male Gametes : Production and Quality 1996 Paris, INSERM 3 45
Parvinen M Ventela S Local regulation of spermatogenesis: a living cell approach Hum Fertil 1999 2 138 142
Kumar TR Wang Y Lu N Matzuk MM Follicle stimulating hormone is required for ovarian follicle maturation but not male fertility Nat Genet 1997 15 201 204 9020850 10.1038/ng0297-201
Dierich A Sairam MR Monaco L Fimia GM Gansmuller A LeMeur M Sassone-Corsi P Impairing follicle-stimulating hormone (FSH) signaling in vivo: targeted disruption of the FSH receptor leads to aberrant gametogenesis and hormonal imbalance Proc Natl Acad Sci U S A 1998 95 13612 13617 9811848 10.1073/pnas.95.23.13612
Abel MH Wootton AN Wilkins V Huhtaniemi I Knight PG Charlton HM The effect of a null mutation in the follicle-stimulating hormone receptor gene on mouse reproduction Endocrinology 2000 141 1795 1803 10803590 10.1210/en.141.5.1795
Yeh S Tsai MY Xu Q Mu XM Lardy H Huang KE Lin H Yeh SD Altuwaijri S Zhou X Xing L Boyce BF Hung MC Zhang S Gan L Chang C Generation and characterization of androgen receptor knockout (ARKO) mice: an in vivo model for the study of androgen functions in selective tissues Proc Natl Acad Sci U S A 2002 99 13498 13503 12370412 10.1073/pnas.212474399
Chang C Chen YT Yeh SD Xu Q Wang RS Guillou F Lardy H Yeh S Infertility with defective spermatogenesis and hypotestosteronemia in male mice lacking the androgen receptor in Sertoli cells Proc Natl Acad Sci U S A 2004 101 6876 6881 15107499 10.1073/pnas.0307306101
De Gendt K Swinnen JV Saunders PT Schoonjans L Dewerchin M Devos A Tan K Atanassova N Claessens F Lecureuil C Heyns W Carmeliet P Guillou F Sharpe RM Verhoeven G A Sertoli cell-selective knockout of the androgen receptor causes spermatogenic arrest in meiosis Proc Natl Acad Sci U S A 2004 101 1327 1332 14745012 10.1073/pnas.0308114100
Holdcraft RW Braun RE Androgen receptor function is required in Sertoli cells for the terminal differentiation of haploid spermatids Development 2004 131 459 467 14701682 10.1242/dev.00957
Zhang FP Pakarainen T Poutanen M Toppari J Huhtaniemi I The low gonadotropin-independent constitutive production of testicular testosterone is sufficient to maintain spermatogenesis Proc Natl Acad Sci U S A 2003 100 13692 13697 14585929 10.1073/pnas.2232815100
Mullaney BP Skinner MK Transforming growth factor-beta (beta 1, beta 2, and beta 3) gene expression and action during pubertal development of the seminiferous tubule: potential role at the onset of spermatogenesis Mol Endocrinol 1993 7 67 76 8446109 10.1210/me.7.1.67
Teerds KJ Dorrington JH Localization of transforming growth factor beta 1 and beta 2 during testicular development in the rat Biol Reprod 1993 48 40 45 8418916
Cupp AS Kim G Skinner MK Expression and action of transforming growth factor beta (TGFbeta1, TGFbeta2, and TGFbeta3) during embryonic rat testis development Biol Reprod 1999 60 1304 1313 10330085
Olaso R Pairault C Habert R Expression of type I and II receptors for transforming growth factor beta in the adult rat testis Histochem Cell Biol 1998 110 613 618 9860260 10.1007/s004180050324
Caussanel V Tabone E Hendrick JC Dacheux F Benahmed M Cellular distribution of transforming growth factor betas 1, 2, and 3 and their types I and II receptors during postnatal development and spermatogenesis in the boar testis Biol Reprod 1997 56 357 367 9116134
Shull MM Ormsby I Kier AB Pawlowski S Diebold RJ Yin M Allen R Sidman C Proetzel G Calvin D Annunziata N Doetschman T Targeted disruption of the mouse transforming growth factor-beta 1 gene results in multifocal inflammatory disease Nature 1992 359 693 699 1436033 10.1038/359693a0
Kulkarni AB Huh CG Becker D Geiser A Lyght M Flanders KC Roberts AB Sporn MB Ward JM Karlsson S Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death Proc Natl Acad Sci U S A 1993 90 770 774 8421714
Kaartinen V Voncken JW Shuler C Warburton D Bu D Heisterkamp N Groffen J Abnormal lung development and cleft palate in mice lacking TGF-beta 3 indicates defects of epithelial-mesenchymal interaction Nat Genet 1995 11 415 421 7493022 10.1038/ng1295-415
Sanford LP Ormsby I Gittenberger-de Groot AC Sariola H Friedman R Boivin GP Cardell EL Doetschman T TGFbeta2 knockout mice have multiple developmental defects that are non-overlapping with other TGFbeta knockout phenotypes Development 1997 124 2659 2670 9217007
Oshima M Oshima H Taketo MM TGF-beta receptor type II deficiency results in defects of yolk sac hematopoiesis and vasculogenesis Dev Biol 1996 179 297 302 8873772 10.1006/dbio.1996.0259
Stenvers KL Tursky ML Harder KW Kountouri N Amatayakul-Chantler S Grail D Small C Weinberg RA Sizeland AM Zhu HJ Heart and liver defects and reduced transforming growth factor beta2 sensitivity in transforming growth factor beta type III receptor-deficient embryos Mol Cell Biol 2003 23 4371 4385 12773577 10.1128/MCB.23.12.4371-4385.2003
Leveen P Larsson J Ehinger M Cilio CM Sundler M Sjostrand LJ Holmdahl R Karlsson S Induced disruption of the transforming growth factor beta type II receptor gene in mice causes a lethal inflammatory disorder that is transplantable Blood 2002 100 560 568 12091349 10.1182/blood.V100.2.560
Weiss M Vigier M Hue D Perrard-Sapori MH Marret C Avallet O Durand P Pre- and postmeiotic expression of male germ cell-specific genes throughout 2-week cocultures of rat germinal and Sertoli cells Biol Reprod 1997 57 68 76 9209082
Staub C Hue D Nicolle JC Perrard-Sapori MH Segretain D Durand P The whole meiotic process can occur in vitro in untransformed rat spermatogenic cells Exp Cell Res 2000 260 85 95 11010813 10.1006/excr.2000.4998
Parvinen M Wright WW Phillips DM Mather JP Musto NA Bardin CW Spermatogenesis in vitro: completion of meiosis and early spermiogenesis Endocrinology 1983 112 1150 1152 6822207
Lee DR Kaproth MT Parks JE In vitro production of haploid germ cells from fresh or frozen-thawed testicular cells of neonatal bulls Biol Reprod 2001 65 873 878 11514353
Sousa M Cremades N Alves C Silva J Barros A Developmental potential of human spermatogenic cells co-cultured with Sertoli cells Hum Reprod 2002 17 161 172 11756382 10.1093/humrep/17.1.161
Perrard MH Hue D Staub C Le Vern Y Kerboeuf D Durand P Development of the meiotic step in testes of pubertal rats: comparison between the in vivo situation and under in vitro conditions Mol Reprod Dev 2003 65 86 95 12658637 10.1002/mrd.10259
Marh J Tres LL Yamazaki Y Yanagimachi R Kierszenbaum AL Mouse round spermatids developed in vitro from preexisting spermatocytes can produce normal offspring by nuclear injection into in vivo-developed mature oocytes Biol Reprod 2003 69 169 176 12620938 10.1095/biolreprod.102.015099
Godet M Damestoy A Mouradian S Rudkin BB Durand P Key role for cyclin-dependent kinases in the first and second meiotic divisions of rat spermatocytes Biol Reprod 2004 70 1147 1152 14695906 10.1095/biolreprod.103.023705
Vigier M Weiss M Perrard MH Godet M Durand P The effects of FSH and of testosterone on the completion of meiosis and the very early steps of spermiogenesis of the rat: an in vitro study J Mol Endocrinol 2004 33 729 742 15591031
Godet M Thomas A Rudkin BB Durand P Developmental changes in cyclin B1 and cyclin-dependent kinase 1 (CDK1) levels in the different populations of spermatogenic cells of the post-natal rat testis Eur J Cell Biol 2000 79 816 823 11139145
Chomczynski P Sacchi N Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction Anal Biochem 1987 162 156 159 2440339 10.1016/0003-2697(87)90021-2
Danielpour D Dart LL Flanders KC Roberts AB Sporn MB Immunodetection and quantitation of the two forms of transforming growth factor-beta (TGF-beta 1 and TGF-beta 2) secreted by cells in culture J Cell Physiol 1989 138 79 86 2910889 10.1002/jcp.1041380112
Cobb J Miyaike M Kikuchi A Handel MA Meiotic events at the centromeric heterochromatin: histone H3 phosphorylation, topoisomerase II alpha localization and chromosome condensation Chromosoma 1999 108 412 425 10654080 10.1007/s004120050393
Manzanero S Arana P Puertas MJ Houben A The chromosomal distribution of phosphorylated histone H3 differs between plants and animals at meiosis Chromosoma 2000 109 308 317 11007489
Avallet O Gomez E Vigier M Jegou B Saez JM Sertoli cell-germ cell interactions and TGF beta 1 expression and secretion in vitro Biochem Biophys Res Commun 1997 238 905 909 9325190 10.1006/bbrc.1997.7275
Lui WY Lee WM Cheng CY TGF-betas: their role in testicular function and Sertoli cell tight junction dynamics Int J Androl 2003 26 147 160 12755993 10.1046/j.1365-2605.2003.00410.x
Marret C Avallet O Perrard-Sapori MH Durand P Localization and quantitative expression of mRNAs encoding the testis-specific histone TH2B, the phosphoprotein p19, the transition proteins 1 and 2 during pubertal development and throughout the spermatogenic cycle of the rat Mol Reprod Dev 1998 51 22 35 9712314 10.1002/(SICI)1098-2795(199809)51:1<22::AID-MRD3>3.0.CO;2-Y
Morgan DO Cyclin-dependent kinases: engines, clocks, and microprocessors Annu Rev Cell Dev Biol 1997 13 261 291 9442875 10.1146/annurev.cellbio.13.1.261
Vidal A Koff A Cell-cycle inhibitors: three families united by a common cause Gene 2000 247 1 15 10773440 10.1016/S0378-1119(00)00092-5
Gong J Ammanamanchi S Ko TC Brattain MG Transforming growth factor beta 1 increases the stability of p21/WAF1/CIP1 protein and inhibits CDK2 kinase activity in human colon carcinoma FET cells Cancer Res 2003 63 3340 3346 12810668
Ortega S Prieto I Odajima J Martin A Dubus P Sotillo R Barbero JL Malumbres M Barbacid M Cyclin-dependent kinase 2 is essential for meiosis but not for mitotic cell division in mice. Nat Genet 2003 35 25 31 12923533 10.1038/ng1232
Berthet C Aleem E Coppola V Tessarollo L Kaldis P Cdk2 knockout mice are viable Curr Biol 2003 13 1775 1785 14561402 10.1016/j.cub.2003.09.024
Liu D Matzuk MM Sung WK Guo Q Wang P Wolgemuth DJ Cyclin A1 is required for meiosis in the male mouse Nat Genet 1998 20 377 380 9843212 10.1038/3855
Datto MB Li Y Panus JF Howe DJ Xiong Y Wang XF Transforming growth factor beta induces the cyclin-dependent kinase inhibitor p21 through a p53-independent mechanism Proc Natl Acad Sci U S A 1995 92 5545 5549 7777546
Li CY Suardet L Little JB Potential role of WAF1/Cip1/p21 as a mediator of TGF-beta cytoinhibitory effect J Biol Chem 1995 270 :4971 4974 7890601 10.1074/jbc.270.10.4971
Beumer TL Roepers-Gajadien HL Gademan LS Rutgers DH de Rooij DG P21(Cip1/WAF1) expression in the mouse testis before and after X irradiation Mol Reprod Dev 1997 47 240 247 9170103 10.1002/(SICI)1098-2795(199707)47:3<240::AID-MRD2>3.0.CO;2-L
Geng Y Weinberg RA Transforming growth factor beta effects on expression of G1 cyclins and cyclin-dependent protein kinases Proc Natl Acad Sci U S A 1993 90 10315 10319 7694291
Liu JH Wei S Burnette PK Gamero AM Hutton M Djeu JY Functional association of TGF-beta receptor II with cyclin B Oncogene 1999 18 269 275 9926943 10.1038/sj.onc.1202263
Nagahara H Ezhevsky SA Vocero-Akbani AM Kaldis P Solomon MJ Dowdy SF Transforming growth factor beta targeted inactivation of cyclin E:cyclin-dependent kinase 2 (Cdk2) complexes by inhibition of Cdk2 activating kinase activity Proc Natl Acad Sci U S A 1999 96 14961 14966 10611320 10.1073/pnas.96.26.14961
Haagmans BL Hoogerbrugge JW Themmen AP Teerds KJ Rat testicular germ cells and Sertoli cells release different types of bioactive transforming growth factor beta in vitro Reprod Biol Endocrinol 2003 1 3 12646048 10.1186/1477-7827-1-3
Hu J Zhang YQ Liu XP Wang RA Jin Y Xu RJ Expression and localization of Smad1, Smad2 and Smad4 proteins in rat testis during postnatal development Asian J Androl 2003 5 51 55 12647004
Xu J Beyer AR Walker WH McGee EA Developmental and stage-specific expression of Smad2 and Smad3 in rat testis J Androl 2003 24 192 200 12634305
Bourguiba S Chater S Delalande C Benahmed M Carreau S Regulation of aromatase gene expression in purified germ cells of adult male rats: effects of transforming growth factor beta, tumor necrosis factor alpha, and cyclic adenosine 3',5'-monosphosphate Biol Reprod 2003 69 592 601 12700195 10.1095/biolreprod.102.013961
Lui WY Lee WM Cheng CY Transforming growth factor-beta3 perturbs the inter-Sertoli tight junction permeability barrier in vitro possibly mediated via its effects on occludin, zonula occludens-1, and claudin-11 Endocrinology 2001 142 1865 1877 11316752 10.1210/en.142.5.1865
Hakovirta H Kaipia A Soder O Parvinen M Effects of activin-A, inhibin-A, and transforming growth factor-beta 1 on stage-specific deoxyribonucleic acid synthesis during rat seminiferous epithelial cycle Endocrinology 1993 133 1664 1668 8404607 10.1210/en.133.4.1664
Olaso R Pairault C Boulogne B Durand P Habert R Transforming growth factor beta1 and beta2 reduce the number of gonocytes by increasing apoptosis Endocrinology 1998 139 733 740 9449647 10.1210/en.139.2.733
| 15941479 | PMC1156949 | CC BY | 2021-01-04 16:37:13 | no | Reprod Biol Endocrinol. 2005 Jun 7; 3:22 | utf-8 | Reprod Biol Endocrinol | 2,005 | 10.1186/1477-7827-3-22 | oa_comm |
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-321594388510.1186/1742-4690-2-32ResearchAPOBEC3G targets human T-cell leukemia virus type 1 Sasada Amane [email protected] Akifumi [email protected] Kotaro [email protected] Masayuki [email protected] Aierkin [email protected] Masakatsu [email protected] Kazunori [email protected] Yuetsu [email protected] Takashi [email protected] Department of Hematology and Oncology, Graduate School of Medicine, Kyoto University, 54 Shogoin-Kawaracho, Sakyo-ku, Kyoto 606-8507, Japan2 Department of Immunology, Graduate School and Faculty of Medicine, University of the Ryukyus, Uehara 207, Nishihara-cho, Nakagami-gun, Okinawa 903-0215, Japan2005 19 5 2005 2 32 32 21 4 2005 19 5 2005 Copyright © 2005 Sasada et al; licensee BioMed Central Ltd.2005Sasada et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Apolipoprotein B mRNA-editing enzyme-catalytic polypeptide-like 3G (APOBEC3G) is a host cellular protein with a broad antiviral activity. It inhibits infectivitiy of a wide variety of retroviruses by deaminating deoxycytidine (dC) into deoxyuridine (dU) in newly synthesized minus strand DNA, resulting in G-to-A hypermutation of the viral plus strand DNA. To clarify the mechanism of its function, we have examined the antiviral activity of APOBEC3G on human T-cell leukemia virus type 1 (HTLV-1), the first identified human retrovirus.
Results
In this study, we have demonstrated that overexpressed as well as endogenous APOBEC3G were incorporated into HTLV-1 virions and that APOBEC3G inhibited the infection of HTLV-1. Interestingly, several inactive mutants of APOBEC3G also inhibited HTLV-1 and no G-to-A hypermutation was induced by APOBEC3G in HTLV-1 genome. Furthermore, we introduced the human immunodeficiency virus type 1 (HIV-1) vif gene into HTLV-1 producing cell line, MT-2, to antagonize APOBEC3G by reducing its intracellular expression and virion incorporation, which resulted in upregulation of the infectivity of produced viruses.
Conclusion
APOBEC3G is incorporated into HTLV-1 virions and inhibits the infection of HTLV-1 without exerting its cytidine deaminase activity. These results suggest that APOBEC3G might act on HTLV-1 through different mechanisms from that on HIV-1 and contribute to the unique features of HTLV-1 infection and transmission.
==== Body
Background
APOBEC3G, also known as CEM15 [1], is a host cellular protein which has a broad antiviral activity on a wide variety of retroviruses including HIV-1, other lentiviruses, and murine leukemia virus (MLV) [2-4]. The protein belongs to the Apobec superfamily of cytidine deaminases [5] and inhibits the infectivity of these viruses by being packaged into virions. During reverse transcription, it deaminates deoxycytidine (dC) into deoxyuridine (dU) in newly synthesized minus strand DNA, resulting in either G-to-A hypermutation of the viral plus strand DNA or degradation of dU-rich reverse transcripts [3,6-8], though several resent studies suggest cytidine deaminase adtivity is essential but not a sole determinant for antiviral activity of APOBEC3G. [7]. Most lentiviruses express an accessory protein called virion infectivity factor (Vif) which blocks the antiviral function of APOBEC3G by preventing its packaging into virions. Vif binds to APOBEC3G and induces its ubiquitination and subsequent degradation by the proteasome [9-13]. It has also been reported that APOBEC3G inhibits the replication of hepatitis B virus (HBV) without inducing G-to-A hypermutation [14]. This suggests that APOBEC3G has a broad antiviral activity not only on retroviruses but also on other viruses through different mechanisms from that on retroviruses.
HTLV-1 is a member of retroviruses which is the etiologic agent of adult T-cell leukemia(ATL) [15] and HTLV-1 associated myelopathy/tropical spastic paraparesis (HAM/TSP) [16]. HTLV-1 has a unique feature of its infectivity and transmission, that is, cell-to-cell contacts are necessary for HTLV-1 transmission, because HTLV-1-infected lymphocytes produce very few cell-free virions, of which, only 1 in 105to 106 is infectious [17]. The fact that infusion of fresh frozen plasma from the seropositive individuals did not cause the transmission also supports the notion that living infected cells are essential for the transmission in vivo [18,19]. Furthermore, the genetic diversity of HTLV-1 is much lower than that of other retroviruses such as HIV-1, although the most frequent mutations in HTLV-1 are also G-to-A transitions [20]. In addition to gag, pol, and env genes, HTLV-1 genome has four open reading frame (ORF) regions at its 3' end, which encode regulatory proteins including Rex and Tax. Although the functions of other encoded proteins such as p12, p13, and p30 have been under investigation [21,22], any counterparts of HIV-1 Vif have not been identified in HTLV-1. These findings suggest the involvement of APOBEC3G in the characteristic infectious and genetic features of HTLV-1 and lead us to investigate this possibility.
In this report, we have investigated the antiviral activity of APOBEC3G on HTLV-1. We examined the packaging of APOBEC3G into HTLV-1 virions, induction of mutations in the viral genome, and regulation of the viral infectivity. Our finding would be a clue to understand the unique infectious mechanism of HTLV-1.
Results
APOBEC3G was incorporated into HTLV-1 virions
We first examined the incorporation of APOBEC3G into HTLV-1 virions. We transfected HEK293T cells with an infectious molecular clone of HTLV-1 (K30) and infectious molecular clones of HIV-1 with or without vif (pNL43-Luc or pNL43/Δvif-Luc, respectively) with or without an expression vector for HA-APOBEC3G and performed Western blotting to detect APOBEC3G in producer cells and produced virions. Incorporation of APOBEC3G was clearly detected in HTLV-1 virions produced from cells cotransfected with HTLV-1 K30 and APOBEC3G expression vector (Fig. 1A, lane 2). Expression of APOBEC3G and its incorporation into HIV-1 were reduced by expression of Vif as reported previously (Fig. 1A, lane 4) [3,4,7,8]. Packaging of APOBEC3G into virions was also confirmed by Western blotting of HTLV-1 K30 virions purified by sucrose density equilibrium gradients method (Fig. 1B). APOBEC3G were detected and colocalized with HTLV-1 Gag (p19) proteins (lanes 4, 5), indicating the incorporation of APOBEC3G into HTLV-1 virion. APOBEC3G mutants and murine APOBEC3G (muAPOBEC3G) were also detected in HTLV-1 virions (Fig. 1C). Since we detected the incorporation of overexpressed APOBEC3G into HTLV-1 virions, we next examined the incorporation of endogenous APOBEC3G into HTLV-1 virions using an HTLV-1 producing cell line, MT-2, which expressed endogenous APOBEC3G (Fig. 1D, lane 1, upper panel). We also detected the incorporation of endogenous APOBEC3G in HTLV-1 virions produced from MT-2 cells (Fig. 1D, lane 1, lower panel). An abundant cytoplasmic protein, β-tubulin, was not detected in MT-2 virion, which excluded the possibility of contamination of the MT-2 virion preparations by cytoplasmic proteins (Fig. 1D lane 2). These indicate that APOBEC3G cannot be excluded from HTLV-1 virions.
Figure 1 Incorporation of APOBEC3G into HTLV-1 virions. (A) Overexpressed APOBEC3G was incorporated into HTLV-1 virions. HEK293T cells were cotransfected with K30, pNL43-Luc (WT), or pNL43/Δvif-Luc (ΔVif) with or without an expression vector for HA-APOBEC3G. Western blotting was performed to detect HA-APOBEC3G in HEK293T cells and produced virions with anti-HA mAb. APOBEC3G was expressed in producer cells and efficiently incorporated into produced virions (lane 2). Expression of APOBEC3G and its incorporation into HIV-1 virions were reduced by expression of Vif as described previously (lane 4). Western blotting with anti-p19 and anti-p24 mAbs showed that similar amounts of virions were produced from each transfection (bottom panel). (B) Incorporation of APOBEC3G was confirmed in HTLV-1 virions purified by sucrose density equilibrium gradient analysis. HTLV-1 K30 virions were purified by sucrose density equilibrium gradient analysis. Gradient fractions were collected and used for analyzing incorporation of APOBEC3G into virions. APOBEC3G were detected and colocalized with HTLV-1 Gag (p19) proteins (lanes 4, 5). (C) APOBEC3G, its mutants, and muAPOBEC3G were incorporated into HTLV-1 virions. Expression vectors for HA-APOBEC3G, its mutants, or HA-muAPOBEC3G were cotransfected with K30 into HEK293T cells and APOBEC3G was detected with anti-HA mAb. HA-APOBEC3G, its mutants, and HA-muAPOBEC3G were all incorporated into virions. A3G and muA3G indicate human and murine APOBEC3G, respectively. E67Q, E259Q, and E67Q/E259Q were inactive mutants of human APOBEC3G that have a point mutation in N-terminal active site, C-terminal active site, and both, respectively, as described previously [7]. (D) Endogenous APOBEC3G was also incorporated into HTLV-1 virions. Western blotting with anti-APOBEC3G Ab revealed expression of endogenous APOBEC3G in MT-2 cells (lane 1, upper panel) and its incorporation into produced virions (lane 1, lower panel). No cytoplasmic proteins were detected with anti-β-tubulin mAb in MT-2 virions (lane 2, lower panel).
HTLV-1 infectivity was inhibited by APOBEC3G
We next examined whether APOBEC3G packaged into HTLV-1 virions deteriorated the infectivity of the virus. For this purpose, we employed the PCR-based infectivity assay as previously described [23] with modification because of very low infectivity of HTLV-1 virions. In brief, we prepared viruses from HEK293T cells transfected with K30 and expression vectors for APOBEC3G or its mutants and challenged these viruses to target SupT1 cells. Infectivity was determined by measuring HTLV-1 proviral DNA load in target cells with real-time quantitative polymerase chain reaction (RQ-PCR) [24]. To exclude the possibility that the residual viral DNA in the supernatant was detected by PCR method, we treated viruses with DNase before assay and prepared heat-inactivated virus as a negative control. Infectivity of K30 was suppressed almost to the level of that of heat-inactivated virus when expressed with APOBEC3G, its mutants, and muAPOBEC3G (Fig. 2 and data not shown). Interestingly, all the APOBEC3G inactive mutants also lowered the infectivity, suggesting that the enzymatic activity of APOBEC3G was dispensable for the antiviral activity on HTLV-1 and that APOBEC3G might act on HTLV-1 through different mechanisms.
Figure 2 Inhibition of HTLV-1 infection by APOBEC3G. APOBEC3G as well as its mutants inhibited the infectivity of HTLV-1. Infectivity of HTLV-1 was measured as described in Materials and Methods. HTLV-1 proviral DNA load in target SupT1 cells was suppressed by APOBEC3G and its mutants to the level of that of heat-inactivted virus. Six independent experiments gave similar results and the data was presented as the mean of these values. Values are presented as infectivity ratio relative to K30 virus without expression of APOBEC3G.
APOBEC3G did not induce G-to-A hypermutation in HTLV-1 genome
To confirm the above hypothesis, we examined whether APOBEC3G induces G-to-A hypermutation in HTLV-1 DNA. p12 region was amplified from target cell DNA and sequenced. We detected a few G-to-A mutations in HTLV-1 K30 genome integrated into target cell DNA in the presence of APOBEC3G (Fig. 3C), but not in the absence of APOBEC3G (Fig. 3D). These G-to-A mutations were only seen with expression of APOBEC3G and mostly occurred in the context of GpG sequence which is the preferred substrate for APOBEC3G, suggesting that these mutations were induced by APOBEC3G, although the frequency is very low as seen with HBV [14]. In contrast, G-to-A hypermutation was induced in HIV-1ΔVif DNA by APOBEC3G (Fig. 3A) as previously reported [3,6-8]. Accordingly, this again suggests the former notion that hypermutation may not be necessary for the antiviral activity of APOBEC3G on HTLV-1.
Figure 3 No G-to-A hypermutation in HTLV-1 genome was induced by APOBEC3G. Mutations in HTLV-1 and HIV-1 ΔVif viruses were detected by sequencing p12 and Env regions, respectively. G-to-A hypermutation was induced by APOBEC3G in HIV-1 ΔVif DNA, but not in HTLV-1 DNA. We detected very few G-to-A mutations in HTLV-1 K30 genome with expression of APOBEC3G (C), but not without expression of APOBEC3G (D), whereas G-to-A hypermutation was induced in HIV-1ΔVif DNA by APOBEC3G (A). We also detected a very few G-to-A mutations in MT-2/Mock virus DNA (E) as well as MT-2/Vif virus DNA (F). G-to-A mutations are shown in red, while other mutations are denoted in black. The numbers before the sequence indicate the number of each clone, while those in parentheses indicate the total number of clones sequenced. WT indicates no mutations in this region.
HIV-1 Vif reverses the infectivity of HTLV-1 suppressed by endogenous APOBEC3G
Finally, we examined the antiviral activity of endogenous APOBEC3G. First, we confirmed the function of endogenous APOBEC3G in MT-2 cells by infection with HIV-1 wild type (WT) and ΔVif virions. WT virus could replicate in MT-2 cells, but ΔVif virus not (data not shown), indicating that endogenous APOBEC3G in MT-2 cells may be able to function as an anti-HIV-1 factor or that there may exist other APOBEC3 protein members sensitive to Vif. Based on this result, we performed an infectivity assay using HTLV-1 virions produced from MT-2 cells. Since we found that endogenous APOBEC3G was incorporated into HTLV-1 virions produced from MT-2 cells (Fig. 1D), we introduced HIV-1 Vif into MT-2 cells to see whether Vif can upregulate the infectivity of HTLV-1 virions produced from MT-2 cells by blocking the virion incorporation of APOBEC3G. MT-2/Mock and MT-2/Vif cell lines were established for this purpose using retrovirus vectors. We confirmed that Vif reduced expression of APOBEC3G in MT-2/Vif cells as well as its incorporation into produced virions (Fig. 4A). Unfortunately, expression of Vif was not enough to totally suppress the expression of APOBEC3G in MT-2/Vif cells and there were some levels of virion incorporation of APOBEC3G left. In order to affirm the inhibitory activity of HIV-1 Vif against APOBEC3G, we performed an infectivity assay using virions produced from these cell lines. The infectivity of viruses produced from MT-2/Vif cells was more than 4 times higher than that from MT-2/Mock cells (Fig. 4B). The infectivity assay on target cells after 10 days of culture also showed similar results (data not shown), suggesting that the possible detection of residual viral DNA in the culture was unlikely. These results indicate that endogenous APOBEC3G incorporated into HTLV-1 virions is functional and suppresses the infectivity of HTLV-1, which can be overcome by HIV-1 Vif. We also examined whether these proviruses have G-to-A hypermutation when integrated into the infected target cell DNA and again found very few G-to-A mutations in both viruses (Fig. 3E and 3F), suggesting that G-to-A hypermutation was not necessary for the inhibition of virus infectivity.
Figure 4 HIV-1 Vif reduced the incorporation of APOBEC3G into HTLV-1 virions, resulting in the upregulation of the infectivity. (A) Expression of APOBEC3G in MT-2 cells and its incorporation into produced virions were reduced by HIV-1 Vif. Expression level of APOBEC3G was reduced in MT-2/Vif cells (lane 2, middle panel) as compared to MT-2/Mock cells (lane 1, middle panel). Incorporation of APOBEC3G into produced virions was also reduced in virions produced from MT-2/Vif cells (lane 2, bottom panel). Expression of Vif protein in MT-2/Vif cells was detected with anti-Vif mAb (top panel). (B) HIV Vif upregulated the infectivity of HTLV-1 produced from MT-2 cells. Infectivity of HTLV-1 virus produced from MT-2 cells was determined as described in Materials and Methods. Infectivity of viruses produced from MT-2/Vif cells was more than 4 times higher than that from MT-2/Mock cells. Four independent experiments gave similar results and the data was presented as the mean of these values. Values are presented as infectivity ratio relative to viruses from MT-2/Mock cells.
Discussion
In this study, we have demonstrated that APOBEC3G has an antiviral activity on HTLV-1. APOBEC3G was efficiently incorporated into HTLV-1 virions and inhibited the infectivity of HTLV-1 without inducing G-to-A hypermutation. First, we showed that APOBEC3G, overexpressed or endogenous, was efficiently incorporated into HTLV-1 virions. Our finding suggests that HTLV-1 cannot exclude this protein from visions unlike HIV-1 [2-4,6-8]. Previous reports have shown that some accessory proteins encoded in open reading frames of HTLV-1 genome could enhance the infectivity of the virus. For example, deletion or mutants of p12 led to impaired infectivity of HTLV-1 both in vivo and in vitro [21,25]. We could not fully exclude the possibility that both K30 and the provirus in MT-2 cells possess mutations in some of these accessory genes so that these viruses could not exclude APOBEC3G from virions, although the possibility is quite low. Whether p12 potentially overcomes APOBEC3G has not been clarified and further investigations are necessary.
Second, we also showed that APOBEC3G inhibited the infection of HTLV-1. Because of low infectivity of cell-free HTLV-1 virions, we could not detect p19 production in the supernatant of infection culture (data not shown). Instead, we performed an infectivity assay as described previously with modification [23], in which RQ-PCR methods enabled us to quantify HTLV-1 genome integrated into target cells and measure the infectivity of cell free virions of HTLV-1, which was very low [24]. Using this method, we demonstrated that APOBEC3G suppressed the infectivity of HTLV-1. Interestingly, not only APOBEC3G but also its inactive mutants inhibited the infectivity of HTLV-1. Taken together with the data that APOBEC3G doesn't induce G-to-A hypermutation in HTLV-1 genome, these results indicate that the enzymatic activity is dispensable for the anti-HTLV-1 activity of APOBEC3G and that it may inhibit HTLV-1 through different mechanisms. In contrast, we previously reported that point mutants of C-terminal active site of APOBEC3G (E259Q, E67Q/E259Q) abrogated its antiviral activity on HIV-1, indicating that the enzymatic activity is essential for anti-HIV-1 activity of APOBEC3G [7]. Furthermore, some groups recently reported that APOBEC3G acts as an antiviral factor on HBV through several mechanisms [14,26]. One is induction of G-to-A mutations in cell type dependent manner, and the other is interference with pregenomic HBV RNA packaging without inducing G-to-A hypermutation. The reason why APOBEC3G inhibits HTLV-1 without inducing G-to-A hypermutation as seen with other retroviruses, even though it is a member of retroviruses, remains unclear. In order to elucidate the precise mechanisms of the antiviral activity of APOBEC3G on HTLV-1, further studies, such as its effects on translation of viral proteins, packaging of viral genome, and budding of virions, other than its cytidine deaminase activity, should be performed in the future.
To confirm the notion above, we prepared MT-2/Vif cells to block incorporation of endogenous APOBEC3G into HTLV-1 virions. Expression of Vif in MT-2 cells reduced the expression of APOBEC3G and its incorporation into virions. In the presence of Vif, APOBEC3G in MT-2 cells seemed to be ubiquitinated and degraded by the proteasome, because we detected two bands of APOBEC3G in MT-2/Vif cells by immunoblotting, of which the upper band might indicate mono-ubiquitinated APOBEC3G, while the faded lower band indicate the intact APOBEC3G remained (Fig. 4A, lanes 1 and 2, middle panel). Interestingly, we demonstrated that viruses released from MT-2/Vif cells recovered their infectivity which had been suppressed in MT-2/Mock cells. Then, we sequenced integrated HTLV-1 genome in target cells infected with viruses produced from MT-2/Vif and MT-2/Mock cells, and detected no G-to-A hypermutation (Fig. 3E and 3F). We hereby propose that the presence of functional endogenous APOBEC3G in virions from MT-2 cells inhibited the infectivity of the virus and that it might be linked to very low infectious titers of cell free HTLV-1 viruses. Taken together, our findings suggest that APOBEC3G might contribute to the unique features of HTLV-1 transmission, such as low infectivity of the virions [17] with very low genetic diversity [20].
During the preparation of this manuscript, Navarro et al. reported that HTLV-1 is relatively resistant to the antiviral effect of encapsidated APOBEC3G [27]. In that paper, they have shown that AOBEC3G is incorporated into HTLV-1 virion and suppresses the infectivity of HTLV-1, although the antiviral activity on HTLV-1 is very weak. We speculate that this discrepancy between their study and ours may originate from different assay systems to measure the infectivity of HTLV-1. They used a luciferase reporter HTLV-1 molecular clone in their study. However, luciferase activity was very low (below 10,000 cps) as compared to that of HIV-1 (more than 20 million cps). Taken together with our data that we could not detect the elevation of p19 levels in the supernatant of infection culture, we suspect that after integration the transcription level of viral gene is very low, resulting in low levels of luciferase activity and p19 production. In such a situation, luciferase reporter system might be inappropriate for evaluation of the infectivity of HTLV-1. Furthermore, in our study, we have shown that APOBEC3G inhibits HTLV-1 infection without exerting its cytidine deaminase activity, suggesting that APOBEC3G might act on HTLV-1 through different mechanisms from that on HIV-1. We believe that this is the first detailed report on the anti-HTLV-1 function of APOBEC3G and first description of possible involvement of other mechanisms than inducing G-to-A hypermutation in anti-HTLV-1 activity.
Finally, our findings have also broadened the spectrum of antiviral activity of APOBEC3G and further studies on the mechanisms of the antiviral activity of APOBEC3G on HTLV-1 will provide us with new insights into the function of this molecule as an antiviral innate immunity.
Conclusion
APOBEC3G is incorporated into HTLV-1 virions and inhibits the infection of HTLV-1 without exerting its cytidine deaminase activity. This suggests that APOBEC3G might act on HTLV-1 through different mechanisms from that on HIV-1 and contribute to the unique features of HTLV-1 infection and transmission.
Materials and methods
Expression vectors and molecular clones
Expression vectors for hemagglutinin (HA)-tagged human APOBEC3G (APOBEC3G), its point mutants (E67Q, E259Q, and E67Q/E259Q), and murine APOBEC3G (muAPOBEC3G) were described previously [4,7]. pNL43-Luc and pNL43/Δvif-Luc were also constructed as previously described [7]. HTLV-1 K30 was a kind gift from Dr. Thomas Kindt through the AIDS Research and Reference Reagent Program [28]. The vif gene was amplified by PCR method from pNL43 and cloned into pDON-AI (Takara Bio Inc., Otsu, Japan) to construct a retrovirus vector, pDON/Vif.
Cell lines
HEK293T cells were maintained in Dulbecco's modified Eagle's medium (Invitrogen, Carlsbad, California) containing 10% fetal calf serum, penicillin, streptomycin, and glutamine (Invitrogen). SupT1 cells and MT-2 cells were maintained in RPMI 1640 (Sigma, St. Louis, Missouri) containing 10% fetal calf serum, penicillin, streptomycin, and glutamine. MT-2/Mock and MT-2/Vif cells were established by transduction of retrovirus vectors (pDON-AI and pDON/Vif, respectively) and selection with Neomycin (Nacalai tesque, Kyoto, Japan).
Expression of APOBEC3G in producer cells and its incorporation into visions
Western blotting was performed to detect expression of APOBEC3G, its mutants, and muAPOBEC3G in producer cells, and their incorporation into virions as described previously [4]. In brief, expression vectors for HA-APOBEC3G, its mutants, or HA-muAPOBEC3G were cotransfected with K30, pNL43-Luc, or pNL43/Δvif-Luc into HEK293T cells. Two days after transfection, viruses in the supernatant were collected and ultracentrifuged with Beckman TL-100s ultracentrifuge at 60,000 × g for 10min and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) together with whole cell lysates of producer HEK293T cells. To detect HA-tagged proteins, they were immunoblotted with anti-HA monoclonal antibody (mAb) (12CA5) (F. Hoffmann-La Roche Ltd., Basel, Switzerland). Virus production was confirmed by immunoblotting with the following antibodies; GIN-7(anti-p19 mAb)[29] for HTLV-1 and anti-p24 mAb (ZeptoMetrix Corporation, Buffalo, New York) for HIV-1. To detect endogenous APOBEC3G in MT-2 cells and its incorporation into virions, whole cell lysates of MT-2 cells and precipitated virions were subjected to immunoblotting with anti-APOBEC3G antibody (a kind gift from Dr. Warner C. Greene, Gladstone Institute of Virology and Immunology, University of California, San Francisco). Vif expression in MT-2/Vif cells was detected with anti-Vif mAb (#319) (a kind gift from Dr. Michael H. Malim through the AIDS Research and Reference Reagent Program) (18). Cytoplasmic proteins were detected with anti-β-tubulin mAb (D-10)(Santa Cruz Biotechnology, Santa Cruz, California). Samples applied to Western blotting were equalized according to p19 antigen levels for HTLV-1 and p24 antigen levels for HIV-1.
Purification of HTLV-1 virions by sucrose density equilibrium gradients and analysis of APOBEC3G packaging
To confirm the incorporation of APOBEC3G into virion, HTLV-1 K30 virions were purified by sucrose density equilibrium gradients as previously reported with slight modifications [30]. Briefly, HTLV-1 K30 virions were prepared as described above and pelleted by ultracentrifugation, then resuspended in 150μl of PBS. They were laid on top of the sucrose gradient, prepared in PBS ranging from 10 to 60%, and centrifuged for 13 h at 20,000 rpm in an SW-41Ti rotor (Beckman, Palo Alto, California). Gradient fractions were collected from the top of the gradient. These samples were used for analyzing protein profiles of the virion by Western blotting. They were subjected to immunoblotting with anti-HA mAb (12CA5) and GIN-7 for detection of HA-APOBEC3G and p19, respectively.
Assessment of HTLV-1 infectivity
Infectivity of HTLV-1 was detected as previously reported with slight modifications [23]. In brief, expression vectors for HA-APOBEC3G, its mutants, or HA-muAPOBEC3G were cotransfected with K30 into HEK293T cells. Viruses in the supernatants were collected 2 days after transfection, then treated with DNase (80 U/ml) (Roche Diagnostics GmbH, Germany) at 37°C for 1 h and filtrated through a 0.45-μm-pore-size filter. Viruses from MT-2 cells were also collected and treated in the same way. We also used noninfectious HTLV-1 as a negative control that had been heat inactivated at 56°C for 1 h. Virus titers were measured with an enzyme-linked immunosorbent assay kit for the p19 antigen (RETRO-TEK, ZeptoMetrix Corporation). SupT1 cells were challenged with viruses whose amounts were equalized according to p19 antigen levels, and washed five times after incubation at 37°C for 8 h. These target cells were cultivated for 2 to 10 days and total cellular DNA was extracted with DNA Mini kit (Quiagen, Valencia, California). HTLV-1 proviral DNA loads were measured by RQ-PCR as described previously [24].
Detection of mutations in the viral DNA
Mutations in HTLV-1 DNA were detected by sequencing p12 region of HTLV-1 integrated into target cells [4]. Preparation of total cellular DNA of target cells infected with HTLV-1 is described above [23]. The p12 region of HTLV-1 was amplified with the following primer pairs:op-32.1(ATAGTCGACCTGTTTCGCCTTCTCAGCCC) and op-32.3(TATCTCGAGGAAGCTGTGCTTGACGG). The PCR products were cloned into pT7-Blue (Novagen, Darmstadt, Germany) and the inserts of individual clones were sequenced. Mutations in HIV-1 NL43 Env region were also detected as previously described [7].
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AS designed research, performed research, contributed vital new reagents, analyzed data, and wrote the paper. AT-K designed research, performed research, contributed vital new reagents, analyzed data, wrote the paper, and organized research. KS performed a part of research. MK performed a part of research. AA performed a part of research. MH performed a part of research and contributed vital new analytical tools. KI contributed vital new analytical tools and analyzed data. YT contributed vital new reagents. TU analyzed data, drafted the paper, and organized research.
Acknowledgements
The following reagents were obtained through the AIDS Research and Reference Reagent Program, Divirion of AIDS, NIDS, NIH: HTLV-1 K30 DNA from Dr. Thomas Kindt, anti-HIV-1 Vif mAb (#319) from Dr. Michael H. Malim. We also thank Dr. Warner C. Greene for providing us with the anti-APOBEC3G Ab.
==== Refs
Sheehy AM Gaddis NC Choi JD Malim MH Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein Nature 2002 418 646 650 12167863 10.1038/nature00939
Harris RS Bishop KN Sheehy AM Craig HM Petersen-Mahrt SK Watt IN Neuberger MS Malim MH DNA deamination mediates innate immunity to retroviral infection Cell 2003 113 803 809 12809610 10.1016/S0092-8674(03)00423-9
Mangeat B Turelli P Caron G Friedli M Perrin L Trono D Broad antiretroviral defence by human APOBEC3G through lethal editing of nascent reverse transcripts Nature 2003 424 99 103 12808466 10.1038/nature01709
Kobayashi M Takaori-Kondo A Shindo K Abudu A Fukunaga K Uchiyama T APOBEC3G Targets Specific Virus Species J Virol 2004 78 8238 8244 15254195 10.1128/JVI.78.15.8238-8244.2004
Jarmuz A Chester A Bayliss J Gisbourne J Dunham I Scott J Navaratnam N An anthropoid-specific locus of orphan C to U RNA-editing enzymes on chromosome 22 Genomics 2002 79 285 296 11863358 10.1006/geno.2002.6718
Lecossier D Bouchonnet F Clavel F Hance AJ Hypermutation of HIV-1 DNA in the absence of the Vif protein Science 2003 300 1112 12750511 10.1126/science.1083338
Shindo K Takaori-Kondo A Kobayashi M Abudu A Fukunaga K Uchiyama T The enzymatic activity of CEM15/Apobec-3G is essential for the regulation of the infectivity of HIV-1 virion but not a sole determinant of its antiviral activity J Biol Chem 2003 278 44412 44416 12970355 10.1074/jbc.C300376200
Zhang H Yang B Pomerantz RJ Zhang C Arunachalam SC Gao L The cytidine deaminase CEM15 induces hypermutation in newly synthesized HIV-1 DNA Nature 2003 424 94 98 12808465 10.1038/nature01707
Yu X Yu Y Liu B Luo K Kong W Mao P Yu XF Induction of APOBEC3G ubiquitination and degradation by an HIV-1 Vif-Cul5-SCF complex Science 2003 302 1056 1060 14564014 10.1126/science.1089591
Sheehy AM Gaddis NC Malim MH The antiretroviral enzyme APOBEC3G is degraded by the proteasome in response to HIV-1 Vif Nat Med 2003 9 1404 1407 14528300 10.1038/nm945
Marin M Rose KM Kozak SL Kabat D HIV-1 Vif protein binds the editing enzyme APOBEC3G and induces its degradation Nat Med 2003 9 1398 1403 14528301 10.1038/nm946
Stopak K de Noronha C Yonemoto W Greene WC HIV-1 Vif blocks the antiviral activity of APOBEC3G by impairing both its translation and intracellular stability Mol Cell 2003 12 591 601 14527406 10.1016/S1097-2765(03)00353-8
Kobayashi M Takaori-Kondo A Miyauchi Y Iwai K Uchiyama T Ubiquitination of APOBEC3G by an HIV-1 Vif-Cullin5-Elongin B-Elongin C Complex Is Essential for Vif Function J Biol Chem 2005 280 18573 18578 15781449 10.1074/jbc.C500082200
Turelli P Mangeat B Jost S Vianin S Trono D Inhibition of Hepatitis B Virus Replication by APOBEC3G Science 2004 303 1829 15031497 10.1126/science.1092066
Uchiyama T Yodoi J Sagawa K Takatsuki K Uchino H Adult T-cell leukemia: clinical and hematologic features of 16 cases Blood 1977 50 481 492 301762
Uchiyama T Human T cell leukemia virus type I (HTLV-I) and human diseases Annu Rev Immunol 1997 15 15 37 9143680 10.1146/annurev.immunol.15.1.15
Igakura T Stinchcombe JC Goon PK Taylor GP Weber JN Griffiths GM Tanaka Y Osame M Bangham CR Spread of HTLV-I between lymphocytes by virus-induced polarization of the cytoskeleton Science 2003 299 1713 1716 12589003 10.1126/science.1080115
Derse D Hill SA Lloyd PA Chung H Morse BA Examining human T-lymphotropic virus type 1 infection and replication by cell-free infection with recombinant virus vectors J Virol 2001 75 8461 8468 11507191 10.1128/JVI.75.18.8461-8468.2001
Matsuoka M Human T-cell leukemia virus type I and adult T-cell leukemia Oncogene 2003 22 5131 5140 12910250 10.1038/sj.onc.1206551
Mansky LM In vivo analysis of human T-cell leukemia virus type 1 reverse transcription accuracy J Virol 2000 74 9525 9531 11000222 10.1128/JVI.74.20.9525-9531.2000
Albrecht B Collins ND Burniston MT Nisbet JW Ratner L Green PL Lairmore MD Human T-lymphotropic virus type 1 open reading frame I p12(I) is required for efficient viral infectivity in primary lymphocytes J Virol 2000 74 9828 9835 11024109 10.1128/JVI.74.21.9828-9835.2000
Franchini G Fukumoto R Fullen JR T-cell control by human T-cell leukemia/lymphoma virus type 1 Int J Hematol 2003 78 280 296 14686485
Fan N Gavalchin J Paul B Wells KH Lane MJ Poiesz BJ Infection of peripheral blood mononuclear cells and cell lines by cell-free human T-cell lymphoma/leukemia virus type I J Clin Microbiol 1992 30 905 910 1572977
Hishizawa M Imada K Ishikawa T Uchiyama T Kinetics of proviral DNA load, soluble interleukin-2 receptor level and tax expression in patients with adult T-cell leukemia receiving allogeneic stem cell transplantation Leukemia 2004 18 167 169 14603340 10.1038/sj.leu.2403204
Collins ND Newbound GC Albrecht B Beard JL Ratner L Lairmore MD Selective ablation of human T-cell lymphotropic virus type 1 p12I reduces viral infectivity in vivo Blood 1998 91 4701 4707 9616168
Rosler C Kock J Malim MH Blum HE von Weizsacker F Comment on "Inhibition of Hepatitis B Virus Replication by APOBEC3G" Science 2004 305 1403a 15353783 10.1126/science.1100464
Navarro F Bollman B Chen H Konig R Yu Q Chiles K Landau NR Complementary function of the two catalytic domains of APOBEC3G Virology 2005 333 374 386 15721369 10.1016/j.virol.2005.01.011
Zhao TM Robinson MA Bowers FS Kindt TJ Characterization of an infectious molecular clone of human T-cell leukemia virus type I J Virol 1995 69 2024 2030 7884847
Tanaka Y Lee B Inoi T Tozawa H Yamamoto N Hinuma Y Antigens related to three core proteins of HTLV-I (p24, p19 and p15) and their intracellular localizations, as defined by monoclonal antibodies Int J Cancer 1986 37 35 42 3000953
Yonezawa A Hori T Takaori-Kondo A Morita R Uchiyama T Replacement of the V3 region of gp120 with SDF-1 preserves the infectivity of T-cell line-tropic human immunodeficiency virus type 1 J Virol 2001 75 4258 4267 11287575 10.1128/JVI.75.9.4258-4267.2001
| 15943885 | PMC1156950 | CC BY | 2021-01-04 16:36:40 | no | Retrovirology. 2005 May 19; 2:32 | utf-8 | Retrovirology | 2,005 | 10.1186/1742-4690-2-32 | oa_comm |
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-341591068310.1186/1742-4690-2-34ResearchQuantification of HTLV-I proviral load in experimentally infected rabbits Zhao Tong-Mao [email protected] Bishop [email protected] David L [email protected] R Mark [email protected] Thomas J [email protected] Molecular and Cellular Immunogenetics Section, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bldg #50, Room 5515, 50 South Drive, Bethesda, MD 20892, USA2 Molecular Pathology Unit, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bldg #37, Room 2002, 37 Convent Drive, Bethesda, MD 20892, USA2005 23 5 2005 2 34 34 12 4 2005 23 5 2005 Copyright © 2005 Zhao et al; licensee BioMed Central Ltd.2005Zhao et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Levels of proviral load in HTLV-1 infected patients correlate with clinical outcome and are reasonably prognostic. Adaptation of proviral load measurement techniques is examined here for use in an experimental rabbit model of HTLV-1 infection. Initial efforts sought to correlate proviral load with route and dose of inoculation and with clinical outcome in this model. These methods contribute to our continuing goal of using the model to test treatments that alleviate virus infection.
Results
A real-time PCR assay was used to measure proviral load in blood and tissue samples from a series of rabbits infected using HTLV-1 inocula prepared as either cell-free virus particles, infected cells or blood, or by naked DNA injection. Proviral loads from asymptomatically infected rabbits showed levels corresponding to those reported for human patients with clinically silent HTLV-1 infections. Proviral load was comparably increased in 50% of experimentally infected rabbits that developed either spontaneous benign or malignant tumors while infected. Similarly elevated provirus was found in organs of rabbits with experimentally induced acute leukemia/lymphoma-like disease. Levels of provirus in organs taken at necropsy varied widely suggesting that reservoirs of infections exist in non-lymphoid organs not traditionally thought to be targets for HTLV-1.
Conclusion
Proviral load measurement is a valuable enhancement to the rabbit model for HTLV-1 infection providing a metric to monitor clinical status of the infected animals as well as a means for the testing of treatment to combat infection. In some cases proviral load in blood did not reflect organ proviral levels, revealing a limitation of this method for monitoring health status of HTLV-1 infected individuals.
==== Body
Background
HTLV-I was the first human retrovirus discovered and was isolated from cell lines derived from patients with cutaneous T cell lymphoma or adult T cell leukemia (ATL) [1,2]. Later it was found that a variety of human diseases are causally associated with HTLV-I infection, including tropical spastic paraparesis (TSP) and myelopathy/tropical spastic paraparesis (HAM/TSP) [3,4].
Previous studies of infected human subjects suggest that high proviral load is associated with increased tendency to develop HTLV-I-associated HAM/TSP, while ATL is associated with extremely high levels of provirus [5-8]. High proviral load was also found in HTLV-I infected patients with seborrheic dermatitis and severe anemia [9] and patients with rheumatoid arthritis or connective tissue disease [10]. The role of HTLV-I proviral load in the development of diseases was studied in asymptomatic carriers [11], and blood donors [12,13]. Proviral load measurement was also used to evaluate the risk of mother-to-child transmission of HTLV-I by breast-feeding [14], study the mortality in HIV-2 and HTLV-I coinfected subjects [15], monitor disease activity in HAM/TSP patients [16], count HTLV-I infected cells in healthy carriers and ATL patients [17], monitor patients following administration of interferon-α [18,19] or green tea extract powder [20], determine the genetic susceptibility to HTLV-I associated diseases [21-23] as well as determine the influence of cytokines [24,25].
Rabbit experimental infection has proven to be an excellent model of human HTLV-I infection [26-31]. Research findings made in rabbits have shed light on transmission modes, and outcomes in the infected rabbits reflect the global diversity of clinical manifestations that occur in HTLV-1 associated diseases, including a variety of cancers, immunologic diseases, and neurologic disorders [3,4]. As is the case for human beings, the majority of HTLV-1 infections in rabbits are chronic asymptomatic infections [28,29]. Data relating proviral load and disease status for the rabbit infection model would greatly enhance the utility of this experimental system and would allow further comparison to human infection. In addition the flexibility afforded by the rabbit model can allow examination of modes possible for transmission of HTLV-I infection.
In this paper we report adaptation of techniques [32] to measure HTLV-I proviral load in PBMC and organs of experimentally infected rabbits. Comparisons were made among rabbits that were inoculated either with cell-free virus, whole blood from HTLV-I infected rabbits, or with an HTLV-I cloned naked DNA. A cohort of infected rabbits monitored for as long as 2.5 yrs produced several examples of rabbits with proviral levels exceeding those established for asymptomatically infected rabbits; examination of these revealed clinical abnormalities including nephroblastoma and uterine tumors.
Results
Cell-free HTLV-I mediates in vivo infectivity in rabbit models
An HTLV-I producing cell line BH24 was derived from rabbit BH24 inoculated with HTLV-I molecular clone K30p naked DNA. HTLV-I env protein gp46 was detected on the surface of BH24 cells and HTLV-I virions isolated from BH24 cell line have normal size and density (Figure 1). Cell free HTLV-I prepared from BH24 cell culture was injected intravenously into rabbit TO11 and rabbit TO12 was given whole blood from BH24. After infection was established rabbit BH42 received blood from TO11 (Figure 1). After two weeks post inoculation all three rabbits including TO11, TO12 and BH42 produced HTLV-I antibodies, HTLV-I provirus was detected in their PBMC, and HTLV-I gag p19 protein was detected in PBMC culture supernatants (Figure 2). These data indicated that cell-free HTLV-I can mediate infectivity in rabbits as does infected blood. In order to determine whether the HTLV-I mutated during the course of infection and transfer, provirus from the rabbits (BH24, TO11, TO12, and BH42) was subjected to sequence analysis at three time points: 8, 12 and 20 months post inoculation. Based on previously observed sequence differences in HTLV-1 regions selected from LTR, gag, pol, env and rex genes were analyzed. For each isolate 4,486 bases were compared and no differences from the original K30 clone were detected for the period of observation. These data gave confidence that proviral load studies may be conducted with little concern for effects of mutations on primer recognition of the provirus.
Figure 1 Source and characterization of HTLV-I virions used in this study. (A) Schematic representation of source and route of HTLV-I exposure. Rabbit BH24 was inoculated with HTLV-I clone K30p naked DNA and an HTLV-I producing cell line BH24 was derived from rabbit BH24 PBMC. Rabbit BH27 was inoculated with plasmid vector pSV2 DNA as negative control. Rabbit TO11 was infected with cell free virus prepared from BH24 cell line supernatant. Rabbits TO13, TO12 and BH42 received whole blood from rabbits BH27, BH24 and TO11, respectively. (B) Analysis of virus particles produced by cell line BH24. Fluorescence-activated cell analysis of cell line BH24 was carried out using antibodies directed against HTLV-I gp46 protein. Goat anti-mouse Ig labeled with fluorescein isothiocyanate was used as the second reagent. The figure (below) shows electro micrographs of particles isolated from supernatant of BH24. The scale bars represent approximately 100 nm. The virions concentration determined by electro micrographs was 2 × 1010 per ml of BH24 cell culture supernatant. The density of particles was 1.16 g / ml measuring by ultracentrifugation on a 20% to 60% sucrose gradient. (Data not shown)
Figure 2 Cell-free HTLV-I particles mediate in vivo infectivity. (A) HTLV-I provirus was detected in rabbit PBMC; (B) HTLV-I antibodies in rabbit sera were detected using a western blot assay (Genelabs Techologies, Singapore). A goat anti-rabbit IgG conjugated with alkaline phosphatase (Santa Cruz Biotechnology, Santa Cruz, CA) was used for rabbit samples instead of goat anti-human IgG conjugate provided by the kit. Mo., month post inoculation; +, positive control serum; -, negative control serum; rgp46, HTLV-I envelope recombinant protein; gp46, HTLV-I env protein; p19 and p24, HTLV-I gag proteins; GD21 specific HTLV-I and HTLV-II epitope recombinant envelop protein. (C) HTLV-I gag p19 protein was detected in the culture supernatants of rabbit PBMC taken at one month post inoculation. (D) Schematic representation showing the regions sequenced. The target regions were amplified by PCR and purified PCR products served as templates for direct sequencing. Stable transmissions of HTLV-I sequence fragments in rabbit BH24, TO11, TO12 and BH42 were observed. No mutation was detected in the analyzed LTR, gag, pol, env, and rex regions for the period of observation up to 20 months. The red arrows indicate the primers used to amplify an env fragment in real-time QPCR assay.
Proviral load in PBMC of HTLV-I infected asymptomatic rabbits
The proviral load was determined using a real time PCR-based QPCR assay, in which HTLV-I env gene was selected as an amplification target. To determine the sensitivity for this assay, scalar dilutions of K30p clone DNA ranging from 1 to one billion (109) copies were analyzed. The results indicate that a positive signal was consistently detected at HTLV-I DNA concentrations above 1 copy per ng of DNA. Based on these results the limitation of this assay was considered to be 1 copy of HTLV-1 proviral DNA per ng of genomic DNA.
Fifty-seven rabbits infected by different routes and using different sources of HTLV-I were monitored for proviral load over a period of 75 weeks at two to four weekly intervals (Figure 3). The highest average proviral load was observed in PBMC from rabbits inoculated with HTLV-I infected whole blood and values peaked at 30 weeks post inoculation. Rabbits injected with HTLV-I naked DNA produced lower levels of provirus and did not reach maximum levels until later than rabbits in the other groups. Provirus loads were intermediate in rabbits injected with cell-free virus and reached maximum levels early as did those given infectious blood. The proviral load peaked around 30 weeks post infection in blood of all rabbits and decreased after that time.
Figure 3 HTLV-I proviral load in PBMCs of asymptomatic rabbits infected with HTLV-I by different routes. Rabbits inoculated with: (1) whole blood (open squares, n = 29), (2) cell free virus (solid diamonds, n = 19), and (3) naked K30p DNA (open circles, n = 5). Proviral load is present as mean and standard errors (error bars). Probability for statistically significant: (1) vs. (2), P = 0.05; (1) vs. (3), P < 0.001; (2) vs. (3), P < 0.01.
Average proviral load measured for HTLV-I infected asymptomatic rabbits was compared with reported data for human samples (Table 1). When all data are converted to the same units, that is, copies of provirus per nanogram of DNA, a close similarity in levels of provirus is seen between the experimentally infected rabbits and the asymptomatic human subjects.
Table 1 Comparison of proviral load in HTLV-I infected asymptomatic rabbits and human samples
Subjects Number tested Copies/ng DNA* Ref.
Rabbits inoculated with
Naked K30p DNA 5 1.5
BH24 cell-free virus 19 2.8
Whole blood 29 3.7
pSV2 plasmid DNA 1 0
Human Asymptomatic carriers 200 1.8 [5]
15 1.0† [36]
83 2.7† [20]
120 2.4 [12]
Human HAM/TSP 202 12.0 [5]
15 8.8† [36]
9 19.9 [19]
Human ATLL 4 47.4† [37]
* The mean or median (†) of proviral loads are present as HTLV-I copy number / ng of genomic DNA prepared from PBMC. The rabbits samples were collected after 2 months post inoculation. All published data were converted to this format on the basis of one ng of genomic DNA corresponding to approximately 150 cells.
In the course of collecting proviral load data one rabbit (TO9) showed an unexpected increase from about 3 copies/ng in a sample taken at 4 months to more than 10 copies/ng at 8 months post infection with cell-free virus. Examination of the rabbit revealed an enlarged kidney, which upon necropsy, was a nephroblastoma. Tissues selected from rabbit TO9 and tested for provirus revealed elevated levels in the thymus, spleen and the tumor dissected from the kidney (Figure 4). Proviral load in the non neoplastic portion of the kidney was 4 times higher than the rabbit's own blood lymphocytes and 10 times the average blood value for all the rabbits.
Figure 4 Proviral load in rabbit TO9 that developed a renal nephroblastoma. Rabbit TO9 was inoculated with cell-free HTLV-I prepared from HTLV-I-producing rabbit cell line RH/K30. The rabbit was necropsied at 8 months post inoculation due to renomegaly. (A) HTLV-I proviral load was determined in PBMC, selected organs, and in both neoplastic (tumor) and non-neoplastic (kidney) regions of the kidney (B) Rabbit kidney, gross photograph of nephroblastoma.
HTLV-I provirus load in rabbit organs and tumor during early and chronic phase of infection
In order to determine proviral loads in major organs from animals infected by different protocols, samples taken at necropsy were analyzed. Table 2 shows the distributions of HTLV-I provirus in rabbit organs. In general the levels in the organs tested were lower than those of the PBMC taken at the same time. However, there are sporadic instances of high proviral load in certain samples, (for example, the thymus, skin and heart samples from BH19 and the spinal cord of T4-9) but no consistent pattern emerged from this analysis. Rabbit BH76 exhibiting a typical PBMC proviral load had an increased level of provirus in its uterus (Table 2). The proviral load within the benign uterine endometrial tumor collected at necropsy was greater than adjacent nontumorous uterine endometrium, by contrast. A somewhat similar relationship among proviral loads in blood, tumor and nonneoplastic adjacent endometrium was observed in rabbit BH25 with a uterine adenocarcinoma (Table 2).
Table 2 Distribution of proviral load in PBMC and organs of HTLV-I infected rabbits*
Inoculated with
Naked DNA Cell-free virus Whole blood
ID (Mo.)† BH19(30) BH21(30) BH25(30) T4-9(11) BH69(19) BH76(17) BH89 (18)
PBMC 7.5 3.9 1.4 6.7 1.3 2.2 8.7
Thymus 22.6 0.4 0.6 0.1 0.3 0.5 1.0
Liver 1.3 0.3 0.3 0.3 0.4 1.8 0.8
Spleen 1.6 0.4 0.1 0.3 0.3 1.7 ND
Skin 23.0 0.3 0.2 4.1 0.2 2.3 0.5
Heart 12.7 2.5 0.3 12.0 0.3 3.1 1.9
Lung 1.8 0.8 0.8 ND 0.7 1.5 2.0
Kidney 3.4 0.5 1.6 0.2 0.5 0.9 1.4
Uterus 3.2 3.0 0.3 ND 1.2 5.0 0.5
Spinal cord 1.7 0.2 0.9 50.5 0.4 0.3 1.2
Brain 0.1 0.3 ND 6.9 ND ND ND
Tumor ‡ NA NA 1.8 1.9 NA 11.0 NA
* Proviral load was reported as copy number per ng of genomic DNA. ND, not determined.
† Rabbit identification number (month post inoculation).
‡ Rabbits T4-9 had a uterine leiomyoma, BH76 had benign endometrial dysplasia, and BH25 developed a uterine adenocarcinoma. NA, not applicable.
Rabbit T4-9 also harbored a uterine neoplasm, however the lesion was not within the endometrium, but resided in the tunica muscularis and was classified as a uterine leiomyoma. This rabbit was found to have an elevated level of PBMC provirus, while the proviral load in the tumor was 1.9 copies per ng of DNA. Interestingly this rabbit's spinal cord showed an unusually high proviral load, which may have been what the blood value reflected.
The proviral load measurement also provided a means to track inoculated rabbits infected with HTLV-1 cell lines known to cause an ATLL-like disease. In the rabbits receiving a high dose of RH/K34 cells known to result in experimental ATLL, proviral loads observed in the lung, kidney and thymus were well above the range established for asymptomatically infected rabbits (Table 3). One of the two rabbits sampled had high values for liver and spleen while the same organs of the other were negative for provirus. The thymus, lung, and kidney of both rabbits had high levels of provirus, consistent with data obtained by histologic studies of organs from these rabbits with experimental ATLL [30].
Table 3 Proviral load in rabbits inoculated with RH/34 cells*
Cells inoculated
2 × 106 2 × 108
ID BH121 BH120 BH101
PBMC 0.5 1.0 ND
Thymus 1.6 88.3 26.3
Live 2.5 207.0 0.5
Spleen 0.0 25.7 0.4
Skin 0.1 6.9 4.8
Heart 0.3 7.3 24.5
Lung 5.1 221.6 217.1
Kidney 0.7 213.0 99.6
* Proviral load was reported as copy number per ng of genomic DNA. The samples were collected at 96 hours post inoculation. ND, not determined.
Discussion
The present study describes infection of the rabbit with HTLV-1 by several different modes and compares the results of infection. In addition the data show that virus sampled at various time post infection retained the sequence of the original HTLV-1 clone indicating that variations in response to infection cannot be attributed to virus mutation. The data here show reproducible in vivo infectivity of rabbits using naked DNA, cell-free virus, infected cell lines or whole blood obtained from HTLV-1 infected rabbits. As previously reported most infections were asymptomatic although certain rabbits monitored for extended periods did develop tumors. An exception to the asymptomatic infection involved rabbits challenged with high doses of the infected cell line RH/K34 [30]; these rabbits succumbed to an aggressive leukemia-like disease within several days.
The assay used to measure provirus load in human patients was adapted for use in the rabbit model and levels of HTLV-1 in blood and parenchymal organs were measured for rabbits infected using different inocula. The rabbits inoculated with either infected whole blood or with cell-free virus showed similar levels of proviral load. In these animals the provirus quickly assumed maximum values and stayed high for about 30 weeks then dropped somewhat to levels that were maintained throughout the period of observation which was up to 70 weeks. A different pattern of provirus load was seen in rabbits infected with naked HTLV-1 DNA clone. Provirus load levels rose slowly in these animals and after reaching maximum levels around 30 weeks began to decline. Average provirus load in the rabbits infected with DNA reached values approximately half those infected with blood or cell free virus.
Comparison of the provirus load values observed for the rabbits were compared to those reported for human subjects infected with HTLV-1. A close correlation with levels in asymptomatic infected humans was seen. Levels in rabbits infected using whole infected blood were slightly higher on average than the human average values but in general the levels suggest that control of infection is similar in the two species.
In several incidents the provirus level rose in an unexpected manner in infected rabbits. One of these, rabbit T09, showed an increase in blood level of provirus to over 10 copies per nanogram of DNA which is about 4 times normal value. Physical examination and subsequent radiograph of the rabbit revealed an enlarged kidney which upon necropsy was shown to harbor a large tumor. The tumor was a nephroblastoma and DNA from it had about 20 copies of provirus per ng. Examination of DNA from lymphoid tissue and the kidney tissue indicated high levels of provirus. In all cases these were considerably greater than the blood levels of provirus. The kidney levels were higher than those of the tumor.
Several other rabbits in the study developed signs that warranted examination and these animals were sacrificed and their organs examined and provirus load determined. For most organs the level of provirus was at the limit of detection and could be dismissed as negative or due to slight amount of contamination by blood. Exceptions to this were seen and point to unusual consequences of infection. For example BH19 had high blood levels and its skin was shown to harbor exceptionally high provirus load. It is tempting to speculate that this animal was enroute to developing cutaneous signs of infection as has been seen in the rabbit model [31]. Rabbit T4-9 had a provirus load of 6.7 copies per ng at sacrifice and examination revealed spinal cord and brain with high provirus load. A neurologic consequence of this infection may be predicted in patients with HAM [38,39]. Of three rabbits with uterine tumors one had high levels of provirus in the tumor tissue whereas two others did not. In the cohort of infected rabbits four developed tumors (3 uterine and 1 kidney) and of these 2 had elevated blood levels of provirus. While this number of events is too low to draw a conclusion about correlation between tumor development and proviral load it is interesting that examination of every rabbit with elevated blood provirus revealed either organ infection or development of a tumor.
In this study, we successfully used a quantitative assay to measure the proviral load of HTLV-1 in PBMC and organs from several cohorts of infected rabbits. Validation of this adapted methodology strengthens the utility of this model for the study of human patients with chronic HTLV-1 infections. Proviral load measurements were made in rabbits infected by different methods; proviral loads from this series of animals infected by different methods documented levels that appeared to stratify according to source of inoculum. Such findings suggest potential of this model for study of HTLV-1 transmission and its relationship to differences in infectious load. In addition to monitoring rabbits that were asymptomatic carriers, proviral load was determined in a subset of rabbits with ATLL-like disease. Data suggested proviral load varied according to tissue compartment, to severity of leukemic infiltration of organs, and to original inoculum dose. If substantiated in larger studies, assay for proviral loads in tissue compartments may reveal additional insight into pathogenesis of lesions in ATLL [12]. Additionally, preclinical therapeutic strategies and drug efficacy designed to combat retroviral infections can be monitored in this system with greater confidence by measuring proviral load status as a response to treatment.
Materials and methods
Animals
The female New Zealand White rabbits were used in this study. Six rabbits were given four 100 μg intramuscular injections of HTLV-I clone K30p naked DNA [33] at biweekly intervals [34]. Twenty-one rabbits were tested for cell-free virus infectivity by intravenous inoculation with 1 to 3 ml of virus preparation containing 1 to 5 × 1012 copies of viral RNA. A total of 30 rabbits received 3.0 ml of whole blood obtained from HTLV-I infected rabbits. Three rabbits were inoculated with rabbit cell line RH/K34, which induces lethal leukemia-like disease in rabbit in high dose inoculation. Infection in rabbits was monitored by the presence of anti-HTLV-1 antibody, virus production in PBMC culture, and detection of viral sequences in PBMC and organs as previously described [34]. The health status of all rabbits on study was monitored by physical examination at time of blood drawing.
Cell lines
The RH/K30 and RH/K34 cell lines were derived by infection of rabbit peripheral blood mononuclear cells using human HTLV-I infected cell line, MT-2. The BH24 cell line was derived from a rabbit inoculated with an infectious HTLV-I molecular clone K30p naked DNA. BH24 cell line is available for research purposes from the AIDS Research and Reference Reagent Program (McKesson BioServices, Germantown, MD)
Preparation of cell-free HTLV-I
Cell-free viruses were prepared from the culture supernatant of HTLV-I producing cell lines. Cells and debris were removed from supernatants by centrifugation at 800 g for 10 min, and then passed through a 0.22 mm filter (Millipore Corporation, Bedford, MA). The filtrates were concentrated to15 to 20 fold through a centrifugal filter device with 100 NMWL membrane (Millipore Corporation, Bedford, MA). The virus stock preparation was stored at -80°C until use. The virus quantitation was measured by a real-time QRT-PCR assay. Thawed virus preparations lost binding activity within several hours unless kept at 4°C [35]. HTLV-I gag p19 protein was determined by a commercial ELISA test (Cellular Product Inc. Buffalo, NY).
Preparation of genomic DNA
The PBMC genomic DNAs were isolated from EDTA-treated blood samples using Wizard Genomic DNA Purification Kit (Promega Corporation, Madison, WI). The organ DNAs were prepared using DNeasy Tissue Kit (Qiagen, Hilden, Germany).
Quantification of HTL V-I proviral load
Two sequence-specific primers that detect HTLV-I env region were used to amplify a 185 bp fragment. The sequences of HTLV-I env primers are: 5'-ATC CAC TTG GCA CGT CCT ATA-3' (nt 5890–5910, GenBank accession no. L03561) and 5'-GCA GGA TGA GGG AGT TAT GTC-3' (nt 6054–6074). The dual-labeled fluorescent probe was FAM -5'-CTT TAC CCA TCG TTA GCG CTT CCA GCC CCC-3'-BHQ1 (nt 5954–5983). Rabbit beta-globin DNA quantitation was performed in parallel on all samples in order to determine the amount of cellular DNA present and was used as an endogenous reference to normalize variations due to differences in the PBMC count or DNA extraction. A 187 bp fragment of the rabbit beta-globin gene was amplified by forward primer 5'-GGT ATC CTT TTT ACA GCA CAA C-3' (nt 372–393, GenBank accession no.V00882) and reverse primer 5'-CAG GTC CCC AAA GGA CTC G-3' (nt 531–549) in a real-time PCR assay. The fluorogenic probe used to detect rabbit beta-globin gene was 5'Quasar 670 - CCT GGG CTG TTT TCA TTT TCT CAG G - BHO-2, 3' (nt 471–495). Both the primers and probes were synthesized by a commercial company (Biosearch Technologies, Inc., Novato, CA)
HTLV-I env and rabbit beta-globin gene fragments were amplified separately with an Mx3000P Real-Time PCR System (Stratagene, La Jolla, Calif.) in 50 μl reaction mixture consisting of 10 μl of DNA sample, 25 μl of Brilliant QPCR Master Mix (containing PCR buffer, SureStart Taq DNA polymerase) (Stratagene, La Jolla, Calif.), 10 pmol of each primer, and 5 pmol of TaqMan probe. Thermal cycling conditions were as follows: 95°C for 10 min, and 45 cycles of 95°C for 30 s, 55°C for 1 min, and 72°C for 30 s. Each sample was analyzed in duplicate, and HTLV-I proviral load was calculated at the copy number of each per ng of genomic DNA.
List of abbreviations
ATLL, adult T-cell leukaemia/lymphoma
HAM/TSP, HTLV-I -associated myelopathy/ tropical spastic paraparesis
PBMC, peripheral blood mononuclear cells
QPCR, quantitative polymerase chain reaction
QRT-PCR, quantitative reverse transcription-polymerase chain reaction
Competing interests
The author(s) declare that they have no competing interests.
Acknowledgements
The assistance of Charles Davis with animal experiments and Matthew Starost with pathology are gratefully acknowledged.
==== Refs
Gallo RC The discovery of the first human retrovirus: HTLV-1 and HTLV-2 Retrovirology 2005 2 16 15743528 10.1186/1742-4690-2-17
Takatsuki K Discovery of adult T-cell leukemia Retrovirology 2005 2 17 15743526 10.1186/1742-4690-2-16
Uchiyama T Human T cell leukemia virus type I (HTLV-I) and human diseases Annu Rev Immunol 1997 15 15 37 9143680 10.1146/annurev.immunol.15.1.15
Nagai M Osame M Human T-cell lymphotropic virus type I and neurological diseases J Neurovirol 2003 9 228 235 12707853
Nagai M Usuku K Matsumoto W Kodama D Takenouchi N Moritoyo T Hashiguchi S Ichinose M Bangham CR Izumo S Osame M Analysis of HTLV-I proviral load in 202 HAM/TSP patients and 243 asymptomatic HTLV-I carriers: high proviral load strongly predisposes to HAM/TSP J Neurovirol 1998 4 586 593 10065900
Nagai M Yamano Y Brennan MB Mora CA Jacobson S Increased HTLV-I proviral load and preferential expansion of HTLV-I Tax-specific CD8+ T cells in cerebrospinal fluid from patients with HAM/TSP Ann Neurol 2001 50 807 812 11761481 10.1002/ana.10065
Matsuzaki T Nakagawa M Nagai M Usuku K Higuchi I Arimura K Kubota H Izumo S Akiba S Osame M HTLV-I proviral load correlates with progression of motor disability in HAM/TSP: analysis of 239 HAM/TSP patients including 64 patients followed up for 10 years J Neurovirol 2001 7 228 234 11517397 10.1080/13550280152403272
Yamano Y Nagai M Brennan M Mora CA Soldan SS Tomaru U Takenouchi N Izumo S Osame M Jacobson S Correlation of human T-cell lymphotropic virus type 1 (HTLV-1) mRNA with proviral DNA load, virus-specific CD8 (+) T cells, and disease severity in HTLV-1-associated myelopathy (HAM/TSP) Blood 2002 99 88 94 11756157 10.1182/blood.V99.1.88
Maloney EM Nagai M Hisada M Soldan SS Goebel PB Carrington M Sawada T Brennan MB Cranston B Hanchard B Jacobson S Prediagnostic human T lymphotropic virus type I provirus loads were highest in Jamaican children who developed seborrheic dermatitis and severe anemia J Infect Dis 2004 189 41 45 14702151
Yakova M Lezin A Dantin F Lagathu G Olindo S Jean-Baptiste G Arfi S Cesaire R Increased proviral load in HTLV-1-infected patients with rheumatoid arthritis or connective tissue disease Retrovirology 2005 2 4 15686595 10.1186/1742-4690-2-4
Okayama A Stuver S Matsuoka M Ishizaki J Tanaka G Kubuki Y Mueller N Hsieh CC Tachibana N Tsubouchi H Role of HTLV-1 proviral DNA load and clonality in the development of adult T-cell leukemia/lymphoma in asymptomatic carriers Int J Cancer 2004 110 621 625 15122598 10.1002/ijc.20144
Lee TH Chafets DM Busch MP Murphy EL Quantitation of HTLV-I and II proviral load using real-time quantitative PCR with SYBR Green chemistry J Clin Virol 2004 31 275 282 15494269 10.1016/j.jcv.2004.05.016
Murphy EL Lee TH Chafets D Nass CC Wang B Loughlin K Smith D HTLVOutcomes Study Investigators Higher human T lymphotropic virus (HTLV) provirus load is associated with HTLV-I versus HTLV-II, with HTLV-II subtype A versus B, and with male sex and a history of blood transfusion J Infect Dis 2004 190 504 510 15243924 10.1086/422398
Li HC Biggar RJ Miley WJ Maloney EM Cranston B Hanchard B Hisada M Provirus load in breast milk and risk of mother-to-child transmission of human T lymphotropic virus type I J Infect Dis 2004 190 1275 1278 15346338 10.1086/423941
Ariyoshi K Berry N Cham F Jaffar S Schim van der Loeff M Jobe O N'Gom PT Larsen O Andersson S Aaby P Whittle H Quantification of Human T-lymphotropic virus type I (HTLV-I) provirus load in a rural West African population: no enhancement of human immunodeficiency virus type 2 pathogenesis, but HTLV-I provirus load relates to mortality J Infect Dis 2003 188 1648 1651 14639534 10.1086/379780
Takenouchi N Yamano Y Usuku K Osame M Izumo S Usefulness of proviral load measurement for monitoring of disease activity in individual patients with human T-lymphotropic virus type I-associated myelopathy/tropical spastic paraparesis J Neurovirol 2003 9 29 35 12587066
Kamihira S Dateki N Sugahara K Hayashi T Harasawa H Minami S Hirakata Y Yamada Y Significance of HTLV-1 proviral load quantification by real-time PCR as a surrogate marker for HTLV-1-infected cell count Clin Lab Haematol 2003 25 111 117 12641615 10.1046/j.1365-2257.2003.00503.x
Saito M Nakagawa M Kaseda S Matsuzaki T Jonosono M Eiraku N Kubota R Takenouchi N Nagai M Furukawa Y Usuku K Izumo S Osame M Decreased human T lymphotropic virus type I (HTLV-I) provirus load and alteration in T cell phenotype after interferon-alpha therapy for HTLV-I-associated myelopathy/tropical spastic paraparesis J Infect Dis 2004 189 29 40 14702150
Feng J Misu T Fujihara K Misawa N Koyanagi Y Shiga Y Takeda A Sato S Takase S Kohnosu T Saito H Itoyama Y Th1/Th2 balance and HTLV-I proviral load in HAM/TSP patients treated with interferon-alpha J Neuroimmunol 2004 151 189 194 15145617 10.1016/j.jneuroim.2004.02.007
Sonoda J Koriyama C Yamamoto S Kozako T Li HC Lema C Yashiki S Fujiyoshi T Yoshinaga M Nagata Y Akiba S Takezaki T Yamada K Sonoda S HTLV-1 provirus load in peripheral blood lymphocytes of HTLV-1 carriers is diminished by green tea drinking Cancer Sci 2004 95 596 601 15245597
Jeffery KJ Usuku K Hall SE Matsumoto W Taylor GP Procter J Bunce M Ogg GS Welsh KI Weber JN Lloyd AL Nowak MA Nagai M Kodama D Izumo S Osame M Bangham CR HLA alleles determine human T-lymphotropic virus-I (HTLV-I) proviral load and the risk of HTLV-I-associated myelopathy Proc Natl Acad Sci U S A 1999 96 3848 3853 10097126 10.1073/pnas.96.7.3848
Vine AM Witkover AD Lloyd AL Jeffery KJ Siddiqui A Marshall SE Bunce M Eiraku N Izumo S Usuku K Osame M Bangham CR Polygenic control of human T lymphotropic virus type I (HTLV-I) provirus load and the risk of HTLV-I-associated myelopathy/tropical spastic paraparesis J Infect Dis 2002 186 932 939 12232833 10.1086/342953
Nitta T Tanaka M Sun B Hanai S Miwa M The genetic background as a determinant of human T-cell leukemia virus type 1 proviral load Biochem Biophys Res Commun 2003 309 161 165 12943677 10.1016/S0006-291X(03)01543-2
Nishimura M Maeda M Yasunaga J Kawakami H Kaji R Adachi A Uchiyama T Matsuoka M Influence of cytokine and mannose binding protein gene polymorphisms on human T-cell leukemia virus type I (HTLV-I) provirus load in HTLV-I asymptomatic carriers Hum Immunol 2003 64 453 457 12651071 10.1016/S0198-8859(02)00829-7
Sabouri AH Saito M Lloyd AL Vine AM Witkover AW Furukawa Y Izumo S Arimura K Marshall SE Usuku K Bangham CR Osame M Polymorphism in the interleukin-10 promoter affects both provirus load and the risk of human T lymphotropic virus type I-associated myelopathy/tropical spastic paraparesis J Infect Dis 2004 190 1279 1285 15346339 10.1086/423942
Akagi T Takeda I Oka T Ohtsuki Y Yano S Miyoshi I Experimental infection of rabbits with human T-cell leukemia virus type I Jpn J Cancer Res 1985 76 86 94 2984072
Taguchi H Sawada T Fukushima A Iwata J Ohtsuki Y Ueno H Miyoshi I Bilateral uveitis in a rabbit experimentally infected with human T-lymphotropic virus type I Lab Invest 1993 69 336 339 8377474
Kindt TJ Said WA Bowers FS Mahana W Zhao TM Simpson RM Passage of human T-cell leukemia virus type-1 during progression to cutaneous T-cell lymphoma results in myelopathic disease in an HTLV-1 infection model Microbes Infect 2000 2 1139 1146 11008104 10.1016/S1286-4579(00)01268-5
Simpson RM Zhao TM Schmidt Hubbard B Said W Kindt TJ Source and route of exposure influence infectivity of a molecular clone of human T cell leukemia virus type I AIDS Res Hum Retroviruses 1998 14 711 715 9618084
Simpson RM Zhao TM Hubbard BS Sawasdikosol S Kindt TJ Experimental acute adult T cell leukemia-lymphoma is associated with thymic atrophy in human T cell leukemia virus type I infection Lab Invest 1996 74 696 710 8600320
Simpson RM Leno M Hubbard BS Kindt TJ Cutaneous manifestations of human T cell leukemia virus type I infection in an experimental model J Infect Dis 1996 173 722 6 8627039
Dehee A Cesaire R Desire N Lezin A Bourdonne O Bera O Plumelle Y Smadja D Nicolas JC Quantitation of HTLV-I proviral load by a TaqMan real-time PCR assay J Virol Methods 2002 102 37 51 11879691 10.1016/S0166-0934(01)00445-1
Zhao TM Robinson MA Bowers FS Kindt TJ Characterization of an infectious molecular clone of human T-cell leukemia virus type I J Virol 1995 69 2024 2030 7884847
Zhao TM Robinson MA Bowers FS Kindt TJ Infectivity of chimeric human T-cell leukemia virus type I molecular clones assessed by naked DNA inoculation Proc Natl Acad Sci U S A 1996 93 6653 6658 8692873 10.1073/pnas.93.13.6653
Hague BF Zhao TM Kindt TJ Binding of HTLV-1 virions to T cells occurs by a temperature and calcium-dependent process and is blocked by certain type 2 adenosine receptor antagonists Virus Res 2003 93 31 39 12727340 10.1016/S0168-1702(03)00048-0
Cesaire R Dehee A Lezin A Desire N Bourdonne O Dantin F Bera O Smadja D Abel S Cabie A Sobesky G Nicolas JC Quantification of HTLV type I and HIV type I DNA load in coinfected patients: HIV type 1 infection does not alter HTLV type I proviral amount in the peripheral blood compartment AIDS Res Hum Retroviruses 2001 17 799 805 11429121 10.1089/088922201750251990
Dehee A Cesaire R Desire N Lezin A Bourdonne O Bera O Plumelle Y Smadja D Nicolas JC Quantitation of HTLV-I proviral load by a TaqMan real-time PCR assay J Virol Methods 2002 102 37 51 11879691 10.1016/S0166-0934(01)00445-1
Kira J Itoyama Y Koyanagi Y Tateishi J Kishikawa M Akizuki S Kobayashi I Toki N Sueishi K Sato H Presence of HTLV-I proviral DNA in central nervous system of patients with HTLV-I-associated myelopathy Ann Neurol 1992 31 39 45 1543349 10.1002/ana.410310108
Kubota R Umehara F Izumo S Ijichi S Matsumuro K Yashiki S Fujiyoshi T Sonoda S Osame M HTLV-I proviral DNA amount correlates with infiltrating CD4+ lymphocytes in the spinal cord from patients with HTLV-I-associated myelopathy J Neuroimmunol 1994 53 23 9 7914211 10.1016/0165-5728(94)90060-4
| 15910683 | PMC1156951 | CC BY | 2021-01-04 16:36:40 | no | Retrovirology. 2005 May 23; 2:34 | utf-8 | Retrovirology | 2,005 | 10.1186/1742-4690-2-34 | oa_comm |
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-351591670710.1186/1742-4690-2-35CommentaryRNA interference: more than a research tool in the vertebrates' adaptive immunity Mak Johnson [email protected] Virology Program, Macfarlane Burnet Institute for Medical Research and Public Health, Melbourne, Australia2 Department of Biochemistry and Molecular Biology, Monash University, Clayton, Australia2005 25 5 2005 2 35 35 13 5 2005 25 5 2005 Copyright © 2005 Mak; licensee BioMed Central Ltd.2005Mak; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
In recent years, RNA silencing, usage of small double stranded RNAs of ~21 – 25 base pairs to regulate gene expression, has emerged as a powerful research tool to dissect the role of unknown host cell factors in this 'post-genomic' era. While the molecular mechanism of RNA silencing has not been precisely defined, the revelation that small RNA molecules are equipped with this regulatory function has transformed our thinking on the role of RNA in many facets of biology, illustrating the complexity and the dynamic interplay of cellular regulation. As plants and invertebrates lack the protein-based adaptive immunity that are found in jawed vertebrates, the ability of RNA silencing to shut down gene expression in a sequence-specific manner offers an explanation of how these organisms counteract pathogen invasions into host cells. It has been proposed that this type of RNA-mediated defence mechanism is an ancient form of immunity to offset the transgene-, transposon- and virus-mediated attack. However, whether 1) RNA silencing is a natural immune response in vertebrates to suppress pathogen invasion; or 2) vertebrate cells have evolved to counteract invasion in a 'RNA silencing' independent manner remains to be determined. A number of recent reports have provided tantalizing clues to support the view that RNA silencing functions as a physiological response to regulate viral infection in vertebrate cells. Amongst these, two manuscripts that are published in recent issues of Science and Immunity, respectively, have provided some of the first direct evidences that RNA silencing is an important component of antiviral defence in vertebrate cells. In addition to demonstrating RNA silencing to be critical to vertebrate innate immunity, these studies also highlight the potential of utilising virus-infection systems as models to refine our understanding on the molecular determinants of RNA silencing in vertebrate cells.
RNA silencingsiRNAmiRNAHIVPFV-1vertebrateimmunityviral invasion
==== Body
RNA silencing was originally recognised as post-transcriptional gene silencing in plants (PTGS) [1], co-suppression in plants [2], or RNA-mediated virus resistance in plants [3]. It was subsequently understood that a similar mechanism (RNA interference) is also found in Caenorhabditis elegans [4] and fungi [5,6]. The generation of ~25 nucleotide RNA which pair to yield a ~19 base-pair double helix is the common denominator amongst these different systems. These RNAs are able to 'silence' the target mRNA through complementarity, which can leads to the degradation of the target mRNA or suppression of protein translation from the target mRNA.
When these small RNAs are generated from double-stranded RNA, they are known as small interfering RNAs (siRNAs); however, if these small RNAs are produced from within the cell as natural RNAs that fold into imperfect hairpin structures, they are referred to as microRNAs (miRNAs). siRNAs are produced by cleavage mediated through a ribonuclease-III (RNaseIII)-related enzyme known as Dicer, which gives rise to the siRNA duplex. This siRNA duplex is then unwound by RNA helicase and assembled into the RNA induced silencing complex (RISC). It is believed that the RISC then directs the siRNA to the target mRNA via sequence specificity which leads to cleavage of the target mRNA and silencing of the target gene. In contrast with siRNAs, miRNAs are first produced as ~70 nucleotides pri-microRNAs (pri-miRNA) in the nucleus, which are then cleaved by a RNaseIII-like enzyme known as Drosha to generate the pre-miRNA. The pre-miRNA is then transported into the cytoplasm aided by Exportin 5. Once the pre-miRNA has reached the cytoplasm, Dicer cleaves the pre-miRNA to generate the miRNA duplex. The miRNA duplex is subsequently unwound by RNA helicase and assembles with RISC. As seen with siRNA, some of the miRNA will be guided by RISC for cleavage and degradation of the target mRNA, however, most miRNA will suppress translation of the target mRNA by binding to the 3' untranslated region (3'UTR) of the target.
Because plants and invertebrates lack the protein-based adaptive immunity that is found in higher vertebrates, it is believed that the sequence-specific RNA silencing mechanism is important for the host cell to fight off invading-viruses, -pathogens and -nucleic acids [7]. Similar to the protein-based adaptive immune response in higher vertebrates, the RNA silencing in plant and C elegans can be spread to other uninfected cells in the organism to prevent further infection. This spreading of RNA silencing relies on an RNA-directed RNA polymerase (RdRP) to amplify the RNA silencing targeting sequences. Interestingly, no RdRP was identified in either Drosophila melanogaster or human genomes when a BLAST search was performed on the nearly completed genomes for these two species [8]. The lack of RdRP in Drosophila melanogaster and human genomes to amplify the siRNA signals as part of their immune responses could suggest that the existence of RNA silencing machineries in these species represents a 'molecular fossil' of an ancient innate immunity, and both Drosophila and vertebrates have since evolved to counter viral invasion through an RNA silencing independent mechanism.
Recent studies have provided a number of indirect and tantalising clues to support the participation of RNA silencing in viral infection of vertebrates. Using a heterologous system, it has been shown that some of the mammalian virus-encoded proteins, such as influenza viral protein NS1 and vaccinia viral protein E3L, have a negative regulatory role on RNA silencing in both plant and insect cells, providing circumstantial evidence to illustrate the potential involvement of RNA silencing of these mammalian viral proteins in their natural vertebrate target hosts [9-11]. Others have shown that the VA non-coding RNAs of adenovirus can down regulate RNA silencing in mammalian cells [12]. However, one must bear in mind that the double strand nature of the siRNAs can results in an interferon-mediated activation of the JAK/STAT (Janus kinase/signal transducer and activator of transcription) pathway and global up-regulation of interferon-stimulated genes. This process is regulated in part by the dsRNA-dependent protein kinase (PKR). As the disruption of adenoviral VA1 RNA significantly affects the level of adenovirus found in infected cells through a strong activation of PKR activity, making it difficult to isolate the precise contribution of the adenoviral VA1 non-coding RNA to the process RNA silencing in mammalian cells [12]. Similarly, using herpesviruses infection systems, it was found that a number of herpesviruses (such as Epstein-Barr virus, Kaposi sarcoma-associated virus, human cytomegalovirus and mouse gammaherpesvirus 68) encode an array of miRNA genes [13-15], however, the physiological functional significance of these miRNAs are yet to be validated [13-15].
Direct evidence of the importance of RNA silencing in vertebrates to control viral invasion has recently emerged from studies using human retroviruses. In the first study, primate foamy virus type 1 (PFV-1) was used to infect mammalian cells [16]. While Lecellier et al. were unable to identify viral derived small RNA that suppressed the propagation of PFV-1 in the host cell, they noted that PFV-1 infection promoted the non-specific accumulation of cellular derived miRNAs as a means to interfere with the miRNA regulatory pathway [16], a situation that has been previously described in plant virus infection. More specifically, they have reported the presence of a cellular derived miRNA that can effectively suppressed PFV-1 replication [16]. In contrast to the PFV-1 study, Bennasser et al. found that human immunodeficiency virus type 1 (HIV-1) contains a rare siRNA precursor within its genomes, which can be utilised by the host cell to regulate HIV-1 infection [17]. It is not excluded that similar to the PFV-1 system, other yet to be identified host cell sequence derived miRNA may also play roles in suppressing HIV-1 infection. One remarkable commonality between PFV-1 and HIV-1 is that both viruses have evolved to use their respective viral transcriptional factors (the PFV-1 Tas and the HIV-1 Tat protein, respectively) as suppressors of RNA silencing (SRS) to counteract this host cell immune response [16,17]. Using a transcriptional inactive HIV-1 Tat mutant, Bennasser et al. further demonstrated that the transcriptional activity of Tat is not essential for this SRS activity of HIV-1 Tat [17]. Currently, it remains unclear whether the transcriptional activity of PFV-1 Tas is also dispensable for its SRS function. Further study is required to unravel the precise mechanism by PFV-1 Tas and HIV-1 Tat indict the host cell's antiviral RNA silencing. It remains to be seen whether these two viruses share a similar mechanism in this type of innate immunity, the demonstration of RNA silencing in vertebrate cells clearly highlights the significance of this ancient immunity in higher eukaryotes [16,17]. This is further underscored by the rarity of siRNA sequence found within the HIV-1 genome and the lack of siRNA precursor sequence in PFV genome, implying that these two viruses have evolved under the selective pressure of RNA silencing and have attempted to alter their sequences to evade this antiviral selection. These observations also emphasise the potential to explore RNA silencing as means to suppress viral infection (such as HIV-1) in mammalian cells, although an effective strategy to deliver small interference RNAs into the target cells has to be developed. On a separate note, the recent studies have shown that viral infection of vertebrate cells can be used as an important tool to dissect the molecular basis for this fascinating but somewhat 'poorly defined' silencing process in mammals.
Competing interests
The author(s) declare that they have no competing interests.
Figure 1 Model of RNA silencing pathway. The biogenesis of RNA silencing transcripts can be derived from either the host cell nucleus mRNA pathway to yield miRNA or the cytoplasmic double strand RNA to yield siRNA. HIV-1 and PFV have evolved to use their transcriptional factor to counteract this ancient host cell immunity.
Acknowledgements
I thank the reviewers for their valuable comments. I am supported by a Pfizer Senior Research Fellowship, Australian NHMRC project grants, Australian Centre for HIV and Hepatitis Virology, and Monash University.
==== Refs
Napoli C Lemieux C Jorgensen R Introduction of a Chimeric Chalcone Synthase Gene into Petunia Results in Reversible Co-Suppression of Homologous Genes in trans Plant Cell 1990 2 279 289 12354959 10.1105/tpc.2.4.279
van der Krol AR Mur LA Beld M Mol JN Stuitje AR Flavonoid genes in petunia: addition of a limited number of gene copies may lead to a suppression of gene expression Plant Cell 1990 2 291 299 2152117 10.1105/tpc.2.4.291
Lindbo JA Dougherty WG Untranslatable transcripts of the tobacco etch virus coat protein gene sequence can interfere with tobacco etch virus replication in transgenic plants and protoplasts Virology 1992 189 725 733 1641986 10.1016/0042-6822(92)90595-G
Fire A Xu S Montgomery MK Kostas SA Driver SE Mello CC Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans Nature 1998 391 806 811 9486653 10.1038/35888
Cogoni C Irelan JT Schumacher M Schmidhauser TJ Selker EU Macino G Transgene silencing of the al-1 gene in vegetative cells of Neurospora is mediated by a cytoplasmic effector and does not depend on DNA-DNA interactions or DNA methylation Embo J 1996 15 3153 3163 8670816
Wu-Scharf D Jeong B Zhang C Cerutti H Transgene and transposon silencing in Chlamydomonas reinhardtii by a DEAH-box RNA helicase Science 2000 290 1159 1162 11073454 10.1126/science.290.5494.1159
Tijsterman M Ketting RF Plasterk RH The genetics of RNA silencing Annu Rev Genet 2002 36 489 519 12429701 10.1146/annurev.genet.36.043002.091619
Zamore PD Ancient pathways programmed by small RNAs Science 2002 296 1265 1269 12016303 10.1126/science.1072457
Li WX Li H Lu R Li F Dus M Atkinson P Brydon EW Johnson KL Garcia-Sastre A Ball LA Palese P Ding SW Interferon antagonist proteins of influenza and vaccinia viruses are suppressors of RNA silencing Proc Natl Acad Sci U S A 2004 101 1350 1355 14745017 10.1073/pnas.0308308100
Delgadillo MO Saenz P Salvador B Garcia JA Simon-Mateo C Human influenza virus NS1 protein enhances viral pathogenicity and acts as an RNA silencing suppressor in plants J Gen Virol 2004 85 993 999 15039541 10.1099/vir.0.19735-0
Bucher E Hemmes H de Haan P Goldbach R Prins M The influenza A virus NS1 protein binds small interfering RNAs and suppresses RNA silencing in plants J Gen Virol 2004 85 983 991 15039540 10.1099/vir.0.19734-0
Lu S Cullen BR Adenovirus VA1 noncoding RNA can inhibit small interfering RNA and MicroRNA biogenesis J Virol 2004 78 12868 12876 15542639 10.1128/JVI.78.23.12868-12876.2004
Pfeffer S Sewer A Lagos-Quintana M Sheridan R Sander C Grasser FA van Dyk LF Ho CK Shuman S Chien M Russo JJ Ju J Randall G Lindenbach BD Rice CM Simon V Ho DD Zavolan M Tuschl T Identification of microRNAs of the herpesvirus family Nat Methods 2005 2 269 276 15782219 10.1038/nmeth746
Pfeffer S Zavolan M Grasser FA Chien M Russo JJ Ju J John B Enright AJ Marks D Sander C Tuschl T Identification of virus-encoded microRNAs Science 2004 304 734 736 15118162 10.1126/science.1096781
Cai X Lu S Zhang Z Gonzalez CM Damania B Cullen BR Kaposi's sarcoma-associated herpesvirus expresses an array of viral microRNAs in latently infected cells Proc Natl Acad Sci U S A 2005 102 5570 5575 15800047 10.1073/pnas.0408192102
Lecellier CH Dunoyer P Arar K Lehmann-Che J Eyquem S Himber C Saib A Voinnet O A cellular microRNA mediates antiviral defense in human cells Science 2005 308 557 560 15845854 10.1126/science.1108784
Bennasser Y Le SY Benkirane M Jeang KT Evidence that HIV-1 encodes a siRNA and a suppressor of RNA silencing Immunity 2005
| 15916707 | PMC1156952 | CC BY | 2021-01-04 16:36:40 | no | Retrovirology. 2005 May 25; 2:35 | utf-8 | Retrovirology | 2,005 | 10.1186/1742-4690-2-35 | oa_comm |
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-371592153210.1186/1742-4690-2-37CommentaryAPOBEC3G & HTLV-1: Inhibition without deamination Strebel Klaus [email protected] Laboratory of Molecular Microbiology, Viral Biochemistry Section; National Institute of Allergy and Infectious Diseases, NIH; Building 4, Room 310; 4 Center Drive, MSC 0460; Bethesda, MD 20892-0460, USA2005 29 5 2005 2 37 37 18 5 2005 29 5 2005 Copyright © 2005 Strebel; licensee BioMed Central Ltd.2005Strebel; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
APOBEC3G is a cellular cytidine deaminase that was recently identified as the Vif-sensitive antiviral host factor responsible for the restriction of vif-defective HIV-1 in primary human cells and certain non-permissive T cell lines. Inhibition of HIV-1 replication is thought to be the result of APOBEC3G-induced hypermutation of the viral genome that occurs early during reverse transcription. Against this backdrop is a new report from the Uchiyama laboratory that proposes deaminase-independent restriction of HTLV-1 by APOBEC3G (Sasada et al. Retrovirology 2005, 2:32). These findings combined with recent reports of deaminase-independent inhibition of Hepatitis B virus as well as HIV-1 suggest that cytidine deaminase activity and antiviral activity may be separable functional properties of APOBEC3G.
==== Body
The identification of APOBEC3G (APO3G) in 2002 as the long elusive cellular target of the HIV viral infectivity factor (Vif) [1] has triggered an outburst of research activity that has produced in a short period of time a rather comprehensive working model for APO3G function (Fig. 1). Although the details of this model are changing almost daily, it is generally believed that APO3G does not interfere with virus production from infected cells but acts at a post-entry level to prevent the productive infection of new target cells. This model is consistent with more than a decade worth of biological and genetic data demonstrating that the Vif-sensitive host factor inhibits HIV-1 infectivity when expressed in the virus-producing cell but does not inhibit infection by HIV-1 when expressed in the target cell (for review see [2]). It is clear that the antiviral activity of APO3G requires its presence in progeny virions where it can cause hypermutation of the viral genome during the reverse transcription step early after infection. Hypermutation of the viral genome at that stage of the viral replication cycle is thought to either result in mutational inactivation of the viral genome – ensuing in the production of defective virions in the next replication cycle – or to trigger degradation of the viral genome in the target cell by activating a host DNA repair mechanism thus resulting in abortive infection (reviewed in [3]).
Figure 1 Inhibition of virus infectivity by APO3G. Cells restrictive for the replication of Vif-defective HIV express the cytidine deaminase APOBEC3G (APO3G). In the absence of Vif, APO3G is packaged into virus particles (1). Such virions are capable of penetrating a target cell and initiate minus-strand cDNA synthesis ((-)-cDNA). However, APO3G causes hypermutation of the viral (-)-cDNA resulting in the conversion of deoxycytidine to deoxyuridine (2). Deoxyuridine residues in the viral cDNA can be targeted by uracil-DNA glycosylase, which could lead to endonucleolytic cleavage by endonucleases present in the target cell (3). Alternatively, hypermutated cDNA enters the nucleus (4) and integrates into the host genome but results in the production of defective or aberrant viral proteins (5). This can lead to an impairment of virus assembly or result in the assembly of non-infectious viruses (6).
Primate lentiviruses such as HIV and SIV have adapted to APO3G with the help of the virus-encoded vif gene whose function is to prevent the packaging of APO3G into virus particles (reviewed in [3]). However, APO3G can target other viruses that do not encode a vif gene or a Vif-like activity. Indeed, APO3G was shown to inhibit the replication of Hepatis B virus (HBV) [4] and was found to affect the retrotransposition of endogenous retroviruses alike [5]. Inhibition of retrotransposition was correlated with G-to-A hypermutation of the endogenous retroviral genomes [5] and APO3G-induced hypermutation of HBV genomes was observed at least in one cell type [6]. However, APO3G-induced hypermutation of the HBV genome seemed to be rare and inhibition of HBV by APO3G appeared to be primarily due to the suppression of viral DNA synthesis through a deamination-independent mechanism [4]. The precise mechanism of such deamination-independent suppression of HBV by APO3G remains elusive. Also, the functional relevance of the APO3G-based restriction of HBV replication in vivo remains to be determined given the fact that expression of APO3G in human hepatocytes – the primary target for HBV – is very low.
Against this backdrop appeared a study by the Uchiyama laboratory investigating the potential antiviral activity of APO3G towards HTLV-1 (Sasada et al. Retrovirology 2005, 2:32). HTLV-1 differs from HIV-1 in that it produces only very low levels of cell-free infectious virions suggesting a mode of virus transmission that is dependent on close cell-to-cell contacts [7,8]. Interestingly, the genetic diversity of HTLV-1 is much lower than that of HIV-1 even though both viruses target primarily APO3G-expressing cells and despite the fact that HTLV-1 does not appear to encode a gene with Vif-like function. Similar resistance to APO3G was observed for MuLV, which replicates in APO3G-expressing murine cells without accumulation of hypermutations despite the fact that murine APO3G is packaged into MuLV virions [9]. These results suggest that packaging of APO3G into viral particles per se may not be sufficient to inhibit viral infectivity. Rather, it seems that APO3G has to be specifically packaged into the viral core in association with the viral RNA to exert its inhibitory activity [10]. In the new study, Sasada et al report that APO3G is efficiently packaged into HTLV-1 particles. This is true for endogenous and exogenously expressed APO3G alike. Interestingly, and consistent with the MuLV model, packaging of APO3G into HTLV-1 did not result in a significant accumulation of APO3G-induced hypermutations. In contrast to MuLV, however, Sasada et al note a profound effect of APO3G on the infectivity of HTLV-1 particles, which was reduced to almost background levels. Similar findings were recently reported by Navarro et al although the effects of APO3G on HTLV-1 infectivity in that study were found to be modest when compared to HIV-1 [11]. Surprisingly, Sasada et al found that variants carrying mutations in either the first or the second zinc finger domain of APO3G were capable of inhibiting HTLV-1 with similar efficiency than the wild type protein. This is in contrast to their previous finding that APO3G enzymatic activity was essential for anti-HIV-1 activity [12]. On the other hand, Newman et al recently demonstrated anti-HIV activity for deaminase-defective APO3G variants similar to the ones used in the current study [13]. Thus, while details have yet to be sorted out, there is an emerging picture of a multifunctional host factor that can exert antiviral activity by way of its inherent deaminase activity or through a deaminase-independent mechanism. One possible deaminase-independent mode of action would be interference with virus-maturation in analogy to the reported inhibition of Gag maturation by high levels of HIV-1 Vif [14]. Such a model seems particularly attractive as APO3G was found to interact with viral Gag precursor proteins [15].
==== Refs
Sheehy AM Gaddis NC Choi JD Malim MH Isolation of a human gene that inhibits HIV-1 infection and is suppressed by the viral Vif protein Nature 2002 418 646 650 12167863 10.1038/nature00939
Strebel K Virus-host interactions: role of HIV proteins Vif, Tat, and Rev AIDS 2003 17 S25 S34 15080177 10.1097/00002030-200317004-00003
Schrofelbauer B Yu Q Landau NR New insights into the role of Vif in HIV-1 replication AIDS Rev 2004 6 34 9 15168739
Turelli P Mangeat B Jost S Vianin S Trono D Inhibition of hepatitis B virus replication by APOBEC3G Science 2004 303 1829 15031497 10.1126/science.1092066
Esnault C Heidmann O Delebecque F Dewannieux M Ribet D Hance AJ Heidmann T Schwartz O APOBEC3G cytidine deaminase inhibits retrotransposition of endogenous retroviruses Nature 2005 433 430 433 15674295 10.1038/nature03238
Rosler C Kock J Malim MH Blum HE von Weizsacker F Comment on "Inhibition of hepatitis B virus replication by APOBEC3G" Science 2004 305 1403 15353783 10.1126/science.1100464
Derse D Hill SA Lloyd PA Chung Hk Morse BA Examining human T-lymphotropic virus type 1 infection and replication by cell-free infection with recombinant virus vectors J Virol 2001 75 8461 8468 11507191 10.1128/JVI.75.18.8461-8468.2001
Igakura T Stinchcombe JC Goon PK Taylor GP Weber JN Griffiths GM Tanaka Y Osame M Bangham CR Spread of HTLV-I between lymphocytes by virus-induced polarization of the cytoskeleton Science 2003 299 1713 1716 12589003 10.1126/science.1080115
Mariani R Chen D Schrofelbauer B Navarro F Konig R Bollman B Munk C Nymark-McMahon H Landau NR Species-Specific Exclusion of APOBEC3G from HIV-1 Virions by Vif Cell 2003 114 21 31 12859895 10.1016/S0092-8674(03)00515-4
Takeuchi H Kao S Miyagi E Khan MA Buckler-White A Plishka R Strebel K Production of infectious SIVagm from human cells requires functional inactivation but not viral exclusion of human APOBEC3G J Biol Chem 2005 280 375 382 15528183 10.1074/jbc.M413863200
Navarro F Bollman B Chen H Konig R Yu Q Chiles K Landau NR Complementary function of the two catalytic domains of APOBEC3G Virology 2005 333 374 386 15721369 10.1016/j.virol.2005.01.011
Shindo K Takaori-Kondo A Kobayashi M Abudu A Fukunaga K Uchiyama T The enzymatic activity of CEM15/Apobec-3G is essential for the regulation of the infectivity of HIV-1 virion but not a sole determinant of its antiviral activity J Biol Chem 2003 278 44412 6 12970355 10.1074/jbc.C300376200
Newman EN Holmes RK Craig HM Klein KC Lingappa JR Malim MH Sheehy AM Antiviral function of APOBEC3G can be dissociated from cytidine deaminase activity Curr Biol 2005 15 166 70 15668174 10.1016/j.cub.2004.12.068
Akari H Fujita M Kao S Khan MA Shehu-Xhilaga M Adachi A Strebel K High Level Expression of Human Immunodeficiency Virus Type-1 Vif Inhibits Viral Infectivity by Modulating Proteolytic Processing of the Gag Precursor at the p2/Nucleocapsid Processing Site J Biol Chem 2004 279 12355 62 14722068 10.1074/jbc.M312426200
Douaisi M Dussart S Courcoul M Bessou G Vigne R Decroly E HIV-1 and MLV Gag proteins are sufficient to recruit APOBEC3G into virus-like particles Biochem Biophys Res Commun 2004 321 566 573 15358144 10.1016/j.bbrc.2004.07.005
| 15921532 | PMC1156953 | CC BY | 2021-01-04 16:36:40 | no | Retrovirology. 2005 May 29; 2:37 | utf-8 | Retrovirology | 2,005 | 10.1186/1742-4690-2-37 | oa_comm |
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-421588820710.1186/1465-9921-6-42ResearchPersistence of lung inflammation and lung cytokines with high-resolution CT abnormalities during recovery from SARS Wang Chun-Hua [email protected] Chien-Ying [email protected] Yung-Liang [email protected] Chun-Liang [email protected] Kuo-Hsiung [email protected] Horng-Chyuan [email protected] Shu-Min [email protected] Tzou-Yien [email protected] Kian Fan [email protected] Han-Pin [email protected] Department of Thoracic Medicine II, Chang Gung Memorial Hospital, Taipei, Taiwan2 Department of Diagnostic Radiology, Chang Gung Memorial Hospital, Taipei, Taiwan3 Division of Pediatric Infectious Diseases, Chang Gung Children's Hospital, Taipei, Taiwan4 National Heart & Lung Institute, Imperial College & Royal Brompton Hospital, London, UK2005 11 5 2005 6 1 42 42 9 3 2005 11 5 2005 Copyright © 2005 Wang et al; licensee BioMed Central Ltd.2005Wang et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
During the acute phase of severe acute respiratory syndrome (SARS), mononuclear cells infiltration, alveolar cell desquamation and hyaline membrane formation have been described, together with dysregulation of plasma cytokine levels. Persistent high-resolution computed tomography (HRCT) abnormalities occur in SARS patients up to 40 days after recovery.
Methods
To determine further the time course of recovery of lung inflammation, we investigated the HRCT and inflammatory profiles, and coronavirus persistence in bronchoalveolar lavage fluid (BALF) of 12 patients at recovery at 60 and 90 days.
Results
At 60 days, compared to normal controls, SARS patients had increased cellularity of BALF with increased alveolar macrophages (AM) and CD8 cells. HRCT scores were increased and correlated with T-cell numbers and their subpopulations, and inversely with CD4/CD8 ratio. TNF-α, IL-6, IL-8, RANTES and MCP-1 levels were increased. Viral particles in AM were detected by electron microscopy in 7 of 12 SARS patients with high HRCT score. On day 90, HRCT scores improved significantly in 10 of 12 patients, with normalization of BALF cell counts in 6 of 12 patients with repeat bronchoscopy. Pulse steroid therapy and prolonged fever were two independent factors associated with delayed resolution of pneumonitis, in this non-randomized, retrospective analysis.
Conclusion
Resolution of pneumonitis is delayed in some patients during SARS recovery and may be associated with delayed clearance of coronavirus, Complete resolution may occur by 90 days or later.
SARSalveolar macrophagesT lymphocytecoronaviruscytokinesbronchoalveolar lavage
==== Body
Introduction
Severe acute respiratory syndrome (SARS) has affected more than 8 thousand patients in 22 countries causing 774 deaths between July 2002 and September 2003 [1]. SARS-associated Coronavirus (SARS-CoV) has been identified as the causative agent [2]. Typical clinical manifestations include fever, cough, dyspnea and rapid progression of pulmonary infiltration or consolidation [3]. The mean mortality rate is 9.6% [1], mostly attributed to hypoxemic respiratory failure. In the acute phase, typical pathological findings in the lungs include mononuclear cells infiltration, alveolar cell desquamation and hyaline membrane formation [4]. Those mononuclear cells may develop into multinucleated giant cells [4]. Proinflammatory cytokines released by alveolar macrophages may play a prominent role in the pathogenesis in SARS [5]. Marked elevation of inflammatory cytokines such as IL-1, IL-6 and IL-12, of the Th1 cytokine, IFN-γ, and of chemokines IL-8, monocyte chemoattractant protein-1 (MCP-1), and IFN-α-induced protein-10 (IP-10) have been reported [6]. High resolution Computed tomography (HRCT) findings at presentation include as unilateral or bilateral ground-glass opacities or focal unilateral or bilateral areas of consolidation [7-9]. Such residual abnormalities have been described also after discharge from hospital at 36.5 days and at 6-months [10,11]. However, limited information is available on recovery of inflammatory abnormalities during recovery from SARS, particularly at 60 days and beyond..
In the current study, we conducted a study to examine HRCT changes in patients who recovered from the acute phase of SARS at days 60 and 90, and measured the associated inflammatory profiles directly by examining bronchoalveolar lavage fluid (BALF). We also examined the presence of coronavirus in BALF. We found persistence of HRCT abnormalities and of lung inflammation at day 60, and determined retrospectively the potential influence of pulse corticosteroid therapy in this process.
Methods
Study subjects
Twelve (9 women and 3 men, aged 18 to 51 years) of 28 confirmed SARS patients who were treated in Chang Gung Memorial Hospital in Taiwan between April and May 2003 during the last epidemic of SARS in Taiwan, agreed to participate in this study. All the patients met the modified Centers for Disease Control and Prevention (CDC) case definition of SARS [12]. SARS was confirmed by either positive real-time polymerase chain reaction (PCR) assays or elevated serum anti-coronavirus antibody by ELISA or both. Nasopharyngeal-aspirate samples were obtained from all study patients to exclude common viruses including influenza viruses A and B, respiratory syncytial (RSV) virus, adenovirus, and parainfluenzavirus types 1, 2, and 3, using commercially-available immunofluorescence assays (IFA). Sputum and blood cultures were performed on all the cases to exclude bacterial or fungal infections. At 90 days, all the close contact relatives of the study SARS patients had their serum anti-coronavirus antibody measured by ELISA.
Nine non-smoking healthy volunteers (5 women and 4 men, aged 18 to 40 years) without evident current or past history of pulmonary diseases based on history as well as physical, chest radiographic and bronchoscopic examinations were selected as controls for this study. None of them had any upper respiratory tract infection within the last 6 weeks or was on antibiotics or other medications at the time of evaluation.
Study protocol
The study protocol was approved by Chang Gung Memorial Hospital Ethical Committee. Informed consent was obtained from all the subjects. Treatment of SARS patients on admission to our unit included broad spectrum antibiotics to target common pathogens causing community-acquired pneumonia, according to current recommendations [13,14]. These patients received variable therapy regimens, including oral ribavirin (1 g twice a day for 5–7 days), or intravenous immunoglobulin (IVIG, 1 g/kg body weight/day for 2 days), pulse steroid therapy (methylprednisolone 500 mg twice a day for 3 days and then prednisolone 1 mg/kg body weight/day for 5 days), or maintenance corticosteroid therapy (prednisolone 10 mg twice a day for more than 3 weeks). Pulse steroid therapy was administered within 3 days of the onset of fever in some patients, depending on the attending physicians' decision irrespective of severity of presentation. Some patients who did not receive pulse steroid therapy were given a short course of corticosteroid therapy (hydrocortisone 100 mg 3 times/day for 3 days) if there was rapid deterioration of pulmonary infiltration or hypoxemia. Maintenance steroid therapy (prednisolone 10 mg per day for one week) was given after pulse or short course corticosteroid therapy in all patients. Two patients were intubated with ventilator support because of hypoxemic respiratory failure.
Patients underwent HRCT and BAL on the 60th and 90th day after the onset of disease. HRCT was performed with 1- to 2-mm collimation sections reconstructed by the use of a high spatial frequency algorithm using a (General Electric Medical Systems, Milwaukee, WI). The HRCT protocol consisted of thin sections obtained at 10-mm through the chest in a supine position without using intravenous contrast medium.
Scoring of HRCT findings
The HRCT findings, as previously described [9], were categorized the predominant pattern as: normal attenuation; ground glass opacification (hazy areas of increased attenuation without obscuration of the underlying vessels); consolidation (homogeneous opacification of the parenchyma with obscuration of the underlying vessels); reticular pattern; mixed pattern (combination of consolidation, ground glass opacities and reticular opacities in the presence of architectural distortion); ground-glass attenuation with traction bronchiolectasis or bronchiectasis; and honeycomb pattern. The extent of involvement of each abnormality was assessed independently for each of three zones: upper (above the carina), middle (below carina up to the inferior pulmonary vein), and lower (below the inferior pulmonary vein). Each lung zone (total of 6 lung zones) was assigned a score, modified from previously described [9], based on the following: 0 when no involvement, 1 when <25% involvement, 2 when 25 <50% involvement; 3 when 50% <75% involvement and 4 when 75% involvement. Summation of scores provided overall lung involvement (maximal CT score 24). The grading of the patient's chest radiograph and HRCT was the consensus of two observers who were blind to clinical information of the patients.
Fibreoptic bronchoscopy and BAL
BAL was performed on all the study subjects using six aliquots (50 ml each) of 0.9% saline solution as described previously [15]. Briefly, sterile saline solution was introduced into the subsegmental bronchus of the most severely involved lobe. The BAL fluid was retrieved and centrifuged. The supernatant was stored at -70°C until analysis and the cell pellet was washed and resuspended at 106 cells per ml. The cell viability and differential cell counts were determined. Total RNA and DNA were extracted from nasopharyngeal aspirates and cells retrieved by BAL with the Viral RNA minikit and QIAmp DNA minikit (QIAGEN, Hilden, Germany). Reverse-transcriptase (RT) PCR was done for influenza A, adenovirus, human metapneumovirus, and SARS-CoV as- described previously [16].
Measurement of T cell subpopulations by flow cytometric analysis
BAL cells were simultaneously stained with fluorescein isothiocyanate or phycoerythrin-conjugated monoclonal antibodies (anti-IgG1, -IgG2a, -CD3, -CD4, -CD8, -CD19, -CD56) (Beckon Dickinson, Mountain View, CA) according to the manufacturer's protocol to identify the proportions of T lymphocytes, CD4, CD8 T cells, B cells and natural killer (NK) cells subpopulations respectively. The relative ratio of CD4 or CD8 in CD3-positive cells was assayed by a dual-color analysis. Data were acquired and analyzed using Becton Dickinson BD LYSYS II and Cytometric Bead Array (CBA) software (San Jose, CA).
Levels of cytokine and chemokine in BAL fluid
The levels of cytokines and chemokines in BAL fluid were assayed using Becton Dickinson (BD) Cytometric Bead Array™ [17] (CBA; BD Biosciences, San Jose, CA) according to manufacturer's instructions with an antibody (PharMoingen, San Diego, CA) against one of five cytokines (Human Chemokine Kit I: CXCL8/IL-8, CCL5/RANTES, CXCL9/MIG, CCL2/MCP-1, CXCL10/IP-10, BD Biosciences,) or of the six cytokines (Human Inflammation Kit: IL-8, IL-1β, IL-6, IL-10, TNF-α, IL-12, BD Biosciences). Commercially available ELISA (R&D Systems, Minneapolis) was used for measurement of the growth factors, TGF-β, IGF-1 and EGF.
Electron microscopic (EM) examination
We used a previously-described method for virus detection by electron microscopy [18]. Cells retrieved by BAL from SARS patients and normal subjects were centrifuged. The cell pellets were fixed, embedded and stained with 4% tannic acid and 0.5% uranyl acetate. The ultra-thin sections were cut from Epon-embedded blocks, stained with uranyl acetate and lead citrate, and examined using a transmission electron microscope (TEM) (H-500, Hitachi, Tokyo, Japan).
Statistical analysis
Data are expressed as mean ± SEM. The baseline characteristics, disease and laboratory variables between groups were compared using the two-tailed Student t-test and chi-square test, respectively. Spearman rank test was used to determine correlations between HRCT scores and T cell numbers, and their subpopulations, as well as CD4/CD8 ratio. Univariate analyses to determine the factors responsible for persistence of HRCT abnormalities were primarily used for selection of variables, based on a p value <0.05. The significant variables were entered into a stepwise logistic regression analysis to determine the net effect for each predictor while controlling of the others. A p-value <0.05 was considered as statistically significant. Analysis was performed using SPSS software version 10.0 (Chicago, IL, USA).
Results
Study subjects
28 patients with confirmed diagnosis of SARS were admitted during the study period. Sixteen patients received intubation and ventilatory support for respiratory failure. Three died of intractable hypoxemic respiratory failure. Twenty-five patients recovered subsequently and were discharged from the hospital. No patient relapsed with either fever or new pulmonary infiltrates after discharge from the hospital. Twenty of the 25 patients were randomly selected into the protocol and twelve agreed to participate. These patients complied with the protocol at 60 days but at 90 days, only 10 of 12 patients agreed to have a repeat HRCT, and 6 of 12 patients had a follow-up bronchoscopy.
Clinical manifestations
At 60 days, the commonest symptoms in SARS patients were general weakness (8 of 12 patients), exertional dyspnea (6 of 12 patients), joint pains (4 of 12 patients) and partial hair loss (11 of 12 patients). At 90 days, all the 12 SARS patients were well without any of the above-described symptoms. There was no detectable SARS-CoV antibody in the sera of close contact relatives of the study patients, even though SARS patients were not isolated after discharge from hospital from their close relatives.
HRCT score
At 60 days, 5 SARS patients were found with an HRCT abnormality of <10 % of the total lung field. In 3, the score was zero, in one ground glass attenuation was found in 7.5% of total lung field, and in another, there was consolidation in 1.7% of total lung field. The other 7 SARS patients had HRCT abnormality > 10% of each lung field (a mean of 37.5 ± 7.9% involvement of total lung field) (Table 1). The most prominent HRCT findings in these patients were ground-glass attenuation (80.8 ± 12.2% of total abnormality on HRCT) and consolidation (13.6 ± 10.9% of total abnormality on HRCT). Honeycombing and bronchiectatic changes were found in only 3 SARS patients with high HRCT score (5.5 ± 2.7% of total abnormality on HRCT). Seven of 11 patients were found normal on their follow-up HRCT at 90 days (Table 2; Figure 1). Two of the patients had persistently high HRCT scores (Table 1). One with very high HRCT score at 60 days refused a follow-up HRCT.
Table 1 Individual HRCT score at 60 and 90 days, and electron microscopic findings in patients with SARS
HRCT score Virus particle in AM by EM
60 days 90 days 60 days 90 days
Case 1 0 0 - N/D
Case 2 4 0 - -
Case 3 0 0 - N/D
Case 4 2 N/D - N/D
Case 5 3 0 - N/D
Case 6 9 0 + -
Case 7 12 3 + N/D
Case 8 11 0 + -
Case 9 12 7 + -
Case 10 13 2 + -
Case 11 15 12 + N/D
Case 12 24 N/D + -
Mean ± SE 8.8 ± 2.1 2.4 ± 1.3*
Abbreviation: HRCT, high resolution computed tomography; SARS, severe acute respiratory syndrome; AM, alveolar macrophage; EM, electron microscopy; N/D, not done.
p < 0.01 indicates a comparison of HRCT score between 60 days and 90 days in corresponding group.
Table 2 Univariate and multivariate analysis: predictors based on presence of virus particle and lung involvement in patients with SARS.
Factor Low HRCT score and Absence of virus particle
(n = 5)
no. (%) High HRCT score and Presence of virus particle
(n = 7)
no. (%) Univariate analysis Multivariate analysis
P value Odd ratio 95% confidence interval P value
Age, year 25.6 ± 4.2 34.9 ± 2.9 0.09 - - -
Female gender 5 (100%) 4 (57.1%) 0.09 1.75 0.92–3.32 -
Titer of Anti-CoV IgG (OD) * 0.8 ± 0.2 1.3 ± 0.1 0.04 - - 1.0
Days of fever 4.2 ± 0.5 11.0 ± 1.0 0.0003 - - 0.011
Positive PCR 2 (28.6%) 5 (71.4%) 0.276 3.75 0.33–42.47 -
Use of ribavirin 4 (57.1%) 6 (85.7%) 0.79 1.50 0.71–31.58 -
Use of IVIG 4 (57.1%) 6 (85.7%) 0.79 1.50 0.71–31.58 -
Pulse corticosteroid therapy 0 (0%) 4 (57.1%) 0.04 2.33 0.99–5.49 0.004
Maintenance corticosteroid therapy 0 (0%) 3 (42.9%) 0.09 1.75 0.92–3.32 -
Need for intubation 1 (14.3%) 1 (14.3%) 0.79 0.67 0.03–14.03 -
Abbreviation: HRCT, high resolution computed tomography; SARS, severe acute respiratory syndrome; IVIG, intravenous immunoglobulin; CoV, coronavirus; OD, optical density; PCR, polymerase chain reaction. Data are shown as mean ± SEM.
*The cut value of positive SARS infection is 0.12 OD.
Figure 1 Residual abnormality on HRCT of a SARS patient with high HRCT score at 60 days (A). HRCT became almost normal at 90 days (B).
Factors associated with residual HRCT abnormalities at day 60
The residual abnormality on HRCT at 60 days was related to the clinical course. Univariate analysis identified 3 factors associated with the residual abnormality on HRCT. There were a greater proportion of patients receiving pulse steroid therapy (4 of 7) in patients with high HRCT score (Table 2). In contrast, none of the patients with low HRCT score received pulse steroid therapy (Table 2). There was no significant difference in other therapy, including maintenance or short course corticosteroid therapy, IVIG or ribavirin, between patients with high HRCT score and those with low HRCT score (Table 2). Patients with high HRCT score had significantly longer course of fever and higher serum SARS-CoV antibody titer when compared to those in patients with low HRCT score (Table 2).
Inflammatory profiles of BAL fluid
At 60 days, compared to normal subjects, there was a significant increase in total cell counts in BAL fluid from SARS patients (Table 3) with a significant increase in alveolar macrophages (AM) and lymphocytes., The proportion of CD8+ T cells was increased to a greater extent than CD4+ T cells, leading to a significant decrease in CD4/CD8 ratio (Table 4). There was also a significant increase in the proportion of NK cells in SARS patients (Table 4). There was no significant difference in B lymphocytes between SARS patients with low or high HRCT scores and normal subjects. HRCT scores were highly correlated with the cell counts of total lymphocyte, CD4+ and CD8+ T cells, and inversely related to the CD4/CD8 ratio (Figure 2). At 90 days, the cellular profiles in BAL fluid of 6 SARS patients were significantly improved compared with those at 60 days, with near normalization (Tables 3, 4).
Table 3 Characteristics of bronchoalveolar lavage in normal subjects and patients with SARS
Normal subjects SARS patients
60 days 90 days
(n = 9) (n = 12) (n = 6)
Age (years) 24.1 ± 2.2 34.0 ± 2.7* 36.6 ± 3.9
Female gender 5 4 3
Cellularity (104 cells/ml) 9.6 ± 0.9 32.9 ± 9.0* 26.2 ± 9.1
Cell viability (%) 91.5 ± 4.3 90.4 ± 1.3 91.6 ± 1.8
AM (%) 93.2 ± 1.2 88.8 ± 1.2* 95.0 ± 0.6†
AM (104 cells/ml) 8.9 ± 0.8 29.0 ± 7.8* 25.1 ± 9.8
Lymphocytes (%) 5.9 ± 1.2 10.2 ± 1.2* 4.1 ± 0.5†
Lymphocytes (104 cells/ml) 0.6 ± 0.1 3.8 ± 1.2* 1.0 ± 0.2†
Neutrophils (%) 0.9 ± 0.2 0.7 ± 0.2 0.9 ± 0.6
Neutrophils (104 cells/ml) 0.1 ± 0.02 0.2 ± 0.1 0.2 ± 0.1
Eosinophils (%) 0.1 ± 0.1 0.3 ± 0.2 0.0 ± 0.0
Eosinophils (104 cells/ml) 0.01 ± 0.01 0.05 ± 0.04 0.0 ± 0.0
Abbreviation: AM, alveolar macrophages; HRCT, high resolution computed tomography; SARS, severe acute respiratory syndrome.
*p < 0.01 compared with normal subjects.
† p < 0.05 compared with SARS patients at 60 days.
Data are mean ± SEM.
Table 4 Lymphocyte subpopulations in bronchoalveolar lavage from normal subjects and patients with SARS
Normal subjects SARS patients
60 days 90 days
(n = 9) (n = 12) (n = 6)
Lymphocytes (103 cells/ml) 5.8 ± 1.4 39.2 ± 12.1* 9.7 ± 2.4†
CD3 cells (%) 39.7 ± 6.4 33.1 ± 6.7 37.8 ± 6.1
CD3 cells (103 cells/ml) 2.4 ± 0.5 16.3 ± 6.4* 3.3 ± 0.7
CD4 cells (%) 9.2 ± 2.6 8.7 ± 2.2 10.4 ± 4.3
CD4 cells (103 cells/ml) 1.2 ± 0.3 4.4 ± 2.0* 0.8 ± 0.3
CD8 cells (%) 6.6 ± 2.6 20.1 ± 5.5* 13.2 ± 3.3
CD8 cells (103 cells/ml) 0.7 ± 0.1 11.8 ± 4.7* 1.1 ± 0.2†
CD4/CD8 (ratio) 1.89 ± 0.22 0.62 ± 0.12* 0.73 ± 0.12†
B cells (%) 6.7 ± 1.2 3.2 ± 0.8 2.8 ± 0.6
B cells (103 cells/ml) 0.4 ± 0.1 1.4 ± 0.7 0.3 ± 0.1
NK cells (%) 1.8 ± 0.2 8.8 ± 2.6* 5.8 ± 2.1
NK cells (103 cells/ml) 0.1 ± 0.03 4.0 ± 2.4** 0.3 ± 0.1†
Abbreviation: HRCT, high resolution computed tomography; NK, natural killer.
*p < 0.05, ** p < 0.01 compared with normal subjects.
†p < 0.05 compared with SARS patients at 60 days.
Data are mean ± SEM.
Figure 2 Correlation of the cell counts of (A) total lymphocytes, (B) CD4 and (C) CD8 T cells, or the (D) CD4/CD8 ratio with HRCT scores in SARS patients. The analysis is made by Spearman rank test and the number and significance are indicated.
Cytokine and chemokine level in BAL fluid
At 60 days, SARS patients had a significantly higher level of chemokines, IL-8, MCP-1, and RANTES (Table 5), and of pro-inflammatory cytokines, TNF-α and IL-6. However, the growth factors, transforming growth factor-β (TGF-β), epidermal growth factor (EGF), insulin-like growth factor-1 (IGF-1), were not increased (Table 5).
Table 5 Cytokine and chemokine levels in bronchoalveolar lavage from normal subjects and SARS patients
Normal subjects (n = 9) SARS patients (n = 12)
CXCL10/IP-10 (pg/ml) 95.8 ± 25.7 133.1 ± 37.5
CXCL9/MIG (pg/ml) 20.2 ± 6.5 53.1 ± 14.1*
IL-8 (pg/ml) 1.5 ± 0.2 6.3 ± 1.0**
CCL2/MCP-1 (pg/ml) 2.4 ± 0.8 9.0 ± 1.2**
CCL5/RANTES (pg/ml) 1.0 ± 0.4 34.6 ± 9.3**
TNF-α (pg/ml) 0.004 ± 0.002 1.1 ± 0.3*
IL-1β (pg/ml) 0.00 ± 0.00 2.5 ± 1.8
IL-6 (pg/ml) 0.001 ± 0.001 1.7 ± 0.5**
IFN-γ (pg/ml) 0.0 ± 0.0 0.4 ± 0.3
IL-2 (pg/ml) 0.00 ± 0.00 0.4 ± 0.2
TGF-β (pg/ml) 9.6 ± 2.9 15.4 ± 4.6
IGF-1 (ng/ml) 0.06 ± 0.03 0.07 ± 0.05
EGF (pg/ml) 0.0 ± 0.0 0.0 ± 0.0
* p < 0.05, ** p < 0.01 compared with normal subjects.
Data are shown as mean ± SEM.
Virus detection
The 12 enrolled patients had serological evidence of recent infection with the SARS-CoV and in seven, viral RNA was detected in samples taken from nasopharyngeal aspirate or stool. However, viral RNA was not detectable in the stool or nasopharyngeal aspirate of any of the SARS patients at 60 days. Healthy controls had no evidence of SARS-CoV antibody or RNA in the serum or the respiratory tract. There were no detectable common viruses including influenza viruses A and B, RSV virus, adenovirus, human metapneumovirus, and parainfluenzavirus types 1, 2, and 3, using IFA for nasopharyngeal aspirates or using RT-PCR assay for cells retrieved by BAL at 60 or 90 days. Serological studies for Clamydia, Mycoplasma or Legionella were negative in all subjects.
At 60 days, EM examination of BAL fluid revealed many coronavirus-infected alveolar macrophages with intracellular viral particles in 7 of 12 patients (Figure 3; Table 1). These patients had the high HRCT scan scores. Coronavirus infected cells were not detected in any of SARS patients with low HRCT score or in normal subjects (Table 1). RT-PCR amplification of coronavirus nucleic acids was positive in 3 of 7 patients with high HRCT score, but in none of patients with low HRCT score or normal subjects. At 90 days, EM examination did not detect any coronavirus-infected cells in 6 SARS patients, in 5 of the 6, viral inclusions were found in AM at day 60 (Table 1). One patient with persistent high HRCT score (case 12) refused follow-up BAL study at 90 days.
Figure 3 Ultrastructural characteristics of a Coronavirus-Infected cell in BAL fluid from a SARS patient at 60 days, with several intracellular particles. The virions are indicated by the arrowheads in Panel A. Panel B shows the area indicated by the asterisk in Panel A at higher magnification. The bar in Panel A (500 nm) and Panel B (100 nm) is indicated.
Discussion
This study was performed during the last epidemic of SARS in Taiwan, and the number of patients recruited has been limited. The epidemic did not recur during 2004, and there have been no further cases of SARS in Taiwan, such that we were not able to increase the number of patients in this study. Despite the relatively low numbers, our observations indicate that there are persistent important inflammatory and radiological abnormalities in some patients who have recovered from acute SARS at 60 days after the illness. These changes were those of ground-glass or/and consolidation abnormalities which may be overlooked on examination of plain chest radiographs. The BAL fluid examination performed for the first time in recovering SARS patients confirmed the presence of an on-going active inflammatory process in most patients with increased macrophages, NK cells and T cells, and augmented levels of chemokines and pro-inflammatory cytokines. These inflammatory responses may be elicited by the persistent presence of coronavirus in alveolar macrophages, since the patients with the highest HRCT changes had coronaviruses present and there was no evidence of bacterial or other viral infection in these patients.
Most viral diseases are characterized by the development of a specific infiltration consisting predominantly of mononuclear leukocytes while neutrophils are absent [19]. The most striking features of alveolar inflammation in patients were increased numbers of alveolar macrophages, T lymphocytes and NK cells, with a striking decrease in CD4/CD8 ratio. The HRCT score was highly correlated with T lymphocyte numbers and their subpopulations, and was inversely related to CD4/CD8 ratio. CD8+ T cells may act as cytotoxic cells and are key effectors of virus clearance [20]. The concurrent increase in CD4+ T cells may promote the clonal expansion of virus-specific CD8+ T cells and is essential for maintaining continued CD8+ T cells surveillance and effector capacity [19].
Exposure of monocytes or macrophages to viruses causes the release of proinflammatory cytokines, such as TNF-α, IL-1, and IL-6, and chemokines [20-22] as well as members of the CC-chemokine subfamily such as MIP-lα, MCP-1, and RANTES which preferentially attract monocytes and lymphocytes [22]. The CXC-chemokines, such as IL-8 or GRO-α, are major neutrophil chemoattractants [23]. MIG/CXC chemokine ligand (CXCL) 9 and IP-10/CXCL10, both inducible by interferon-γ, are ELR-negative CXC chemokines and are potent chemoattractants for mononuclear cells, specifically activated T lymphocytes and NK cells [24]. In this report, we demonstrated elevated levels of TNF-α, IL-6, MCP-1, RANTES and IL-8 in BAL fluid in SARS patients compared to those of normal subjects. Increased secretion of TNF-α and IL-6 may be derived from virus-infected macrophages or from CD4+ or CD8+ T cells, and these cytokines may promote T-lymphocyte extravasation and macrophage activation [19], but such processes may not be sufficient on their own to recruit and activate mononuclear cells in virus-infected lungs. The increased levels of MCP-1 and RANTES in BAL fluid of all SARS patients may be responsible for the generation of mononuclear infiltrates observed after coronavirus infection. IP-10 and MIG, whose levels are also increase in SARS patients, recruit monocytes and macrophages, NK cells and activated, but not resting T lymphocytes [25,26].
Although there were increased levels of IL-8 in BAL fluid in SARS patients, the number of neutrophils in BALF were sparse. The absence of neutrophil infiltration in influenza A virus or respiratory syncytial viruses (RSV) infection is attributed to the suppression of neutrophil attracting CXC-chemokines or by induction of IL-10 [27]. However, IL-8 production can be induced by measles virus infection of fibroblasts [28] and by influenza A virus, RSV and rhinovirus in pulmonary epithelial cells or AM [28-31]. The reasons for the lack of neutrophil recruitment in response to elevated IL-8 levels in SARS patients are not known and this deserves further investigation.
Despite the presence of virus in AM at 60 days when patients had already been discharged from hospital, these patients were not infectious, because none of their close contact relatives developed any detectable SARS-CoV antibody in their sera. The HRCT and the clinical course until the 90th day of illness did not suggest any evidence of pulmonary fibrosis in SARS patients. This was in accord with the low level of cytokines and growth factors responsible for tissue repairing and fibrosis [32], such as IL-1β, TGF-β, IGF-1, and EGF detected in BAL fluid. However, evidence of fibrosis on HRCT has been obtained on HRCT scans particularly in patients with very severe disease during the acute phase of SARS [33].
The use of corticosteroids together with ribavirin has been reported to confer clinical benefit, although randomized clinical trials to support its clinical efficacy are not available. Pulse steroid therapy was reported to lead to less oxygen requirement, better radiographic outcome, and less likelihood to require rescue pulse steroid therapy than their counterparts [34]. However, corticosteroids have been shown to increase mortality of pneumovirus-infected mice by accelerating replication of virus [35]. Pulse steroid therapy was also reported to be associated with residual abnormality on HRCT in SARS patients after discharge from hospital [10]. In the current retrospective, non-randomized series, pulse steroid therapy appeared to be associated with delayed resolution of pneumonitis. We have planned a prospective future study on investigating the effect of pulse steroid therapy in case of future outbreaks of SARS. This is because this issue is extremely important in outcome from SARS.
In conclusion, a proportion of recovered SARS patients have delayed resolution of pneumonitis and delayed clearance of coronavirus in the alveolar space at day 60. This was associated with persistent inflammatory response characterized by macrophages, T cells particularly CD8+ T cells, and NK cells, and by increase in cytokines and chemokines. This host inflammatory response against SARS-CoV infection may contribute to persistent HRCT abnormalities during recovery phase of SARS. On the other hand, we found no evidence of pulmonary fibrosis in SARS patients during recovery. At day 90, many of the abnormalities have disappeared. Patients recovering from SARS need to be followed up for at least 3 months after the infection.
Acknowledgements
This project was supported by CMRP-32012S from Chang Gung Memorial Hospital.
==== Refs
Cumulative number of reported probable cases of severe acute respiratory syndrome (SARS) Geneva: World Health Organization 2003 Accessed September 26, 2003
Ksiazek TG Erdman D Goldsmith CS Zaki SR Peret T Emery S Tong S Urbani C Comer JA Lim W Rollin PE Dowell SF Ling AE Humphrey CD Shieh WJ Guarner J Paddock CD Rota P Fields B DeRisi J Yang JY Cox N Hughes JM LeDuc JW Bellini WJ Anderson LJ SARS Working Group A Novel coronavirus associated with severe acute respiratory syndrome N Engl J Med 2003 348 1953 1966 12690092 10.1056/NEJMoa030781
Poutanen SM Low DE Henry B Finkelstein S Rose D Green K Tellier R Draker R Adachi D Ayers M Chan AK Skowronski DM Salit I Simor AE Slutsky AS Doyle PW Krajden M Petric M Brunham RC McGeer AJ National Microbiology Laboratory, Canada; Canadian Severe Acute Respiratory Syndrome Study Team Identification of severe acute respiratory syndrome in Canada N Engl J Med 2003 348 1993 2003 10.1056/NEJMoa030634
Tse GMK To KF Chan PKS Lo AWI Ng KC Wu A Lee N Wong HC Mak SM Chan KF Hui DSC Sung JJY Ng HK Pulmonary pathological features in coronavirus associated severe acute respiratory syndrome (SARS) J Clin Pathol 2004 57 260 265 14990596 10.1136/jcp.2003.013276
Nicholls JM Poon LL Lee KC Ng WF Lai ST Leung CY Chu CM Hui PK Mak KL Lim W Yan KW Chan KH Tsang NC Guan Y Yuen KY Peiris JS Lung pathology of fatal severe acute respiratory syndrome Lancet 2003 361 1773 1778 12781536 10.1016/S0140-6736(03)13413-7
Wong CK Lam CWK Wu AKL Ip WK Lee NLS Chan HIS Lit LCW Hui DSC Chan MHM Chung SSC Sung IJY Plasma inflammatory cytokines and chemokines in severe acute respiratory syndrome Clin Exp Immunol 2004 136 95 103 15030519 10.1111/j.1365-2249.2004.02415.x
Tsang KW Ho PL Ooi GC Yee WK Wang T Chan-Yeung Lam WK Seto WH Yam LY Cheung TM Wong PC Lam B Ip MS Chan J Yuen KY Lai KN A cluster of cases of severe acute respiratory syndrome in Hong Kong N Engl J Med 2003 348 1977 1985 12671062 10.1056/NEJMoa030666
Muller NL Ooi GC Khong PL Zhou LJ Tsang KWT Nicolaou S High-resolution CT findings of severe acute respiratory syndrome at presentation and after admission AJR 2004 182 39 44 14684509
Ooi GC Khong PL Muller NL Yiu WC Zhou LJ Ho JCM Lam B Nicolaou S Tsang KW Severe acute respiratory syndrome: temporal lung changes at thin-section CT in 30 patients Radiology 2004 230 839 844
Antonio GE Wong KT Hui DS Wu A Lee N Yuen EH Leung CB Rainer TH Cameron P Chung SS Sung JJ Ahuja AT Thin-section CT in patients with severe acute respiratory syndrome following hospital discharge: preliminary experience Radiology 2003 228 810 815 12805557
Ng CK Chan JW Kwan TL To TS Chan YH Ng FY Mok TY Six month radiological and physiological outcomes in severe acute respiratory syndrome (SARS) survivors Thorax 2004 59
Centers for Disease Control and Prevention Severe acute respiratory syndrome (SARS) updated interim US case definition 2003 Accessed April 22, 2003
Bartlett JG Dowell SF Mandell LA File TM JrMusher DM Fine MJ Practice guidelines for the management of community-acquired pneumonia in adults Clin Infect Dis 2000 31 347 382 10987697 10.1086/313954
Guidelines for the management of adults with community-acquired pneumonia: diagnosis, assessment of severity, antimicrobial therapy, and prevention Am J Respir Crit Care Med 2001 163 1730 1754 11401897
Wang CH Liu CY Lin HC Yu CT Chung KF Kuo HP Increased exhaled nitric oxide in active pulmonary tuberculosis due to inducible nitric oxide synthase upregulation in alveolar macrophages Eur Respir J 1998 11 809 815 9623681 10.1183/09031936.98.11040809
Drosten C Gottig S Schilling S Asper M Panning M Schmitz H Gunther S Rapid detection and quantification of RNA of Ebola and Marburg viruses, Lassa virus, Crimean-Congo hemorrhagic fever virus, Rift Valley fever virus, dengue virus, and yellow fever virus by real-time reverse transcription-PCR J Clin Microbiol 2002 40 2323 2330 12089242 10.1128/JCM.40.7.2323-2330.2002
Aoe K Hiraki A Murakami T Murakami K Makihata K Takao K Eda R Maeda T Sugi K Darzynkiewicz Z Takeyama H Relative abundance and patterns of correlation among six cytokines in pleural fluid measured by cytometric bead array Int J Mol Med 2003 12 193 198 12851716
Wu S Ko YS Teng MS Ko YL Hsu LA Hsueh C Chou YY Liew CC Lee YS Adriamycin-induced cardiomyocyte and endothelial cell apoptosis: in vitro and in vivo studies J Mol Cell Cardiol 2002 34 1595 1607 12505058 10.1006/jmcc.2002.2110
Doherty PC Topham DJ Tripp RA Cardin RD Brooks JW Stevenson PG Effector CD4+ and CD8+ T-cell mechanisms in the control of respiratory virus infections Immunol Rev 1997 159 105 117 9416506
Nain M Hinder F Gong JH Schmidt A Bender A Sprenger H Gemsa D Tumor necrosis factor-α production of influenza A virus-infected macrophages and potentiating effect of lipopolysaccharides J Immunol 1990 145 1921 1928 2391423
Gong JH Sprenger H Hinder F Bender A Schmidt A Horch S Nain M Gemsa D Influenza A virus infection of macrophages: enhanced tumor necrosis factor-alpha (TNF-alpha) gene expression and lipopolysaccharide-triggered TNF-alpha release J Immunol 1991 147 3507 3513 1940351
Henke A Mohr C Sprenger H Graebner C Stelzner A Nain M Gemsa D Coxsackievirus B3-induced production of tumor necrosis factor-α, IL-1, and IL-6 in human monocytes J Immunol 1992 148 2270 2277 1312105
Oppenheim JJ Zachariae COC Mukaida N Matsushima K Properties of the novel proinflammatory supergene "intercrine" cytokine family Annu Rev Immunol 1991 9 617 648 1910690
Baggiolini M Dewald B Moser B Interleukin-8 and related chemotactic cytokines – CXC and CC chemokines Adv lmmunol 1994 55 97 179
Loetscher M Gerber B Loetscher P Jones SA Piali L Clark-Lewis I Baggiolini M Moser B Chemokine receptor specific for IP-10 and Mig: structure, function and expression in activated T lymphocytes J Exp Med 1996 184 963 969 9064356 10.1084/jem.184.3.963
Arai K Liu ZX Lane T Dennert G IP-10 and MIG facilitate accumulation of T cells in the virus-infected liver Cell Immunol 2002 219 48 56 12473267 10.1016/S0008-8749(02)00584-1
Panuska JR Merolla R Rebert NA Hoffmann SP Tsivitse P Cirino NM Silverman RH Rankin JA Respiratory syncytial virus induces interleukin-10 by human alveolar macrophages. Suppression of early cytokine production and implications for incomplete immunity J Clin Invest 1995 96 2445 2453 7593633
Van Damme J Decock B Conings R Falkenburg JH Opdenakker G Billiau A The chemotactic activity for granulocytes produced by virally infected fibroblasts is identical to monocyte-derived interleukin 8 Eur J Immunol 1989 19 1189 1194 2668011
Choi AMK Jacoby DB Influenza virus A infection induces interleukin-8 gene expression in human airway epithelial cells FEBS Lett 1992 309 327 329 1516705 10.1016/0014-5793(92)80799-M
Becker S Quay J Soukoup J Cytokine (tumor necrosis factor, IL-6, and IL-8) production by respiratory syncytial virus-infected human alveolar macrophages J Immunol 1991 147 4307 4312 1753101
Subauste MC Jacoby DB Richards SM Proud D Infection of a human respiratory epithelial cell line with rhinovims. Induction of cytokine release and modulation of susceptibility to infection by cytokine exposure J Clin Invest 1995 96 549 557 7615827
Kelly M Kolb M Bonniaud P Gauldie J Re-evaluation of fibrogenic cytokines in lung fibrosis Curr Pharm Des 2003 9 39 49 12570673 10.2174/1381612033392341
Muller NL Ooi GC Khong PL Zhou LJ Tsang KW Nicolaou S High-resolution CT findings of severe acute respiratory syndrome at presentation and after admission Am J Roentgenol 2004 182 39 44 14684509
Ho JC Ooi GC Mok TY Chan JW Hung I Lam B Wong PC Li PC Ho PL Lam WK Ng CK Ip MS Lai KN Chan-Yeung M Tsang KW High dose versus non-pulse corticosteroid regimens in severe acute respiratory syndrome Am J Respir Crit Care Med 2003 168 1449 1456 12947028 10.1164/rccm.200306-766OC
Domachowske JB Bonville CA Ali-Ahmad D Rosenberg HF Glucocorticoid administration accelerate mortality of pneumovirus-infected mice J Infect Dis 2001 184 1518 1523 11740726 10.1086/324664
| 15888207 | PMC1156954 | CC BY | 2021-01-04 16:23:25 | no | Respir Res. 2005 May 11; 6(1):42 | utf-8 | Respir Res | 2,005 | 10.1186/1465-9921-6-42 | oa_comm |
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-431590448510.1186/1465-9921-6-43ResearchThe effect of refurbishing a UK steel plant on PM10 metal composition and ability to induce inflammation Hutchison Gary R [email protected] David M [email protected] Leon R [email protected] Mathew R [email protected] Ken [email protected] Robert L [email protected] Michelle [email protected] Andy [email protected] Vicki [email protected] Biomedicine Research Group, Napier University, Edinburgh EH10 5DT, UK2 School of Chemistry, University of Edinburgh, West Mains Road, Edinburgh, UK3 ELEGI & COLT Research Laboratory, Medical School, University of Edinburgh, UK4 Department of Health UK, Skipton House, 80 London Road, London SE1 6LH, UK5 Institute of Occupational Medicine, Research Park North, Riccarton, Edinburgh, EH14 4AP, Scotland, UK2005 18 5 2005 6 1 43 43 23 12 2004 18 5 2005 Copyright © 2005 Hutchison et al; licensee BioMed Central Ltd.2005Hutchison et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In the year 2000 Corus closed its steel plant operations in Redcar, NE of England temporarily for refurbishment of its blast furnace. This study investigates the impact of the closure on the chemical composition and biological activity of PM10 collected in the vicinity of the steel plant.
Methods
The metal content of PM10 samples collected before during and after the closure was measured by ICP-MS in order to ascertain whether there was any significant alteration in PM10 composition during the steel plant closure. Biological activity was assessed by instillation of 24 hr PM10 samples into male Wistar rats for 18 hr (n = 6). Inflammation was identified by the cellular and biochemical profile of the bronchoalveolar lavage fluid. Metal chelation of PM10 samples was conducted using Chelex beads prior to treatment of macrophage cell line, J774, in vitro and assessment of pro-inflammatory cytokine expression.
Results
The total metal content of PM10 collected before and during the closure period were similar, but on reopening of the steel plant there was a significant 3-fold increase (p < 0.05) compared with the closure and pre-closure samples. Wind direction prior to the closure was predominantly from the north, compared to south westerly during the closure and re-opened periods. Of metals analysed, iron was most abundant in the total and acid extract, while zinc was the most prevalent metal in the water-soluble fraction. Elevated markers of inflammation included a significant increase (p < 0.01) in neutrophil cell numbers in the bronchoalveolar lavage of rats instilled with PM10 collected during the reopened period, as well as significant increases in albumin (p < 0.05). Extracts of PM10 from the pre-closure and closure periods did not induce any significant alterations in inflammation or lung damage. The soluble and insoluble extractable PM10 components washed from the reopened period both induced a significant increase in neutrophil cell number (p < 0.05) when compared to the control, and these increases when added together approximately equalled the inflammation induced by the whole sample. PM10 from the re-opened period stimulated J774 macrophages to generate TNF-α protein and this was significantly prevented by chelating the metal content of the PM10 prior to addition to the cells.
Conclusion
PM10-induced inflammation in the rat lung was related to the concentration of metals in the PM10 samples tested, and activity was found in both the soluble and insoluble fractions of the particulate pollutant.
==== Body
Introduction
Elevated levels of ambient respirable particulate matter (PM10) are associated with increased morbidity and mortality, especially in susceptible individuals [1]. The composition of PM10 is variable and complex, which makes identification of the toxic material all the harder, although a variety of components have been proposed to induce inflammation leading to adverse health effects [2].
In 2000 the steel plant located at the Teesside works in Redcar, UK closed temporarily for a major repair programme to its blast furnace. During this period all steel making and casting operations at Lackenby and ore sintering at Redcar ceased (figure 1). The Department for Environment, Food and Rural Affairs (Defra) and the Devolved Administrations took advantage of this refurbishment to investigate the effect that closing the plant would have on locally produced PM10.
Figure 1 Map of Redcar and surrounding industrial sites. AUN TEOM collection site in proximity to the blast furnace.
This is the first study of its kind in the UK, but is similar in concept to that of the Utah study by Pope [1] Pope et al.,[3] reported that, during the closure of a steel mill in the Utah valley, a reduction in PM10 mass, and changes in its composition were associated with decreases in morbidity and mortality of the local population. The Utah scenario was a landmark study as it is unusual for an environmental intervention study to take place where the major source of the pollution is closed off and switched on again, allowing researchers to examine clearly the effects of air pollution. The temporary closure of the Utah valley steel mill provided researchers with the unique opportunity to demonstrate a correlation between changes in PM10 composition and observed health outcomes. Alterations were observed in PM10 composition and mass during the closure period [4]. Changes in total mass did not account for all of the variation in the biological effects of PM10 in the Utah valley between the closure of the steel mill, during its shutdown and following its reopening [4]. The hypothesis put forward suggested that the metal component of the PM10 was the predominant factor in driving inflammation. Workers at the US Environmental Protection Agency (EPA) showed the importance of the metal content of the Utah valley PM10 in relation to its toxicity and pro-inflammatory potential, by carrying out a range of human [4], animal [5] and in vitro studies [6,7]. Further analysis of Utah PM10 metal content showed iron (Fe), copper (Cu) and zinc (Zn) to be abundant during the active periods of the steel mill, but to be substantially reduced during closure. Such transition metals can act as initiators of inflammation and cytotoxicity via oxidative mechanisms, such as redox cycling. It has also been hypothesised that the allergen or endotoxin content of the PM10may have a role with respect to effects on health. None of these hypotheses have yet been proven, but the case for the role of transition metals has been emphasised through research into the Utah episode.
The current study aimed to investigate whether closure of a UK steel plant blast furnace would also impact upon the metal content of PM10 and whether this change in composition would alter the biological potency of this pollutant.
Methods
All materials were obtained from Sigma (Poole, U.K.) unless otherwise stated.
PM10 sample collection
PM10 samples were collected by Redcar and Cleveland Council, in collaboration with Casella Stanger using a Tapered Element Oscillating Microbalance (TEOM) with Automated Cartridge Collection Unit (ACCU). The flow rate was 16.7 l/min equivalent to the human lung ventilation rate. This means that over 24 hours 24048 l of air were sampled, and each filter was used to collect PM10 for 6–8 days. The sampling location was the Redcar Automated Urban Network (AUN) site, to the east of the steel plant blast furnace in a highly populated area (figure 1). Samples were collected from 21/06/00 until 15/12/00, during which time the steel plant closed operations on the week commencing 26/07/00 and reopened 28/09/00. At this location, the Corus Teesside works is the major industrial source of PM10 (table 1). To conduct compositional and toxicological analysis, PM10 filters were randomly selected from each of the 3 periods.
Table 1 Environment agency PM10 emissions data collected from year 2000 within the Redcar area (* Corus operations effected by blast furnace relining shown in map figure 1) [8].
REDCAR PM10 RELEASE FROM LOCAL INDUSTRY
OPERATOR NAME SITE ADDRESS TOTAL RELEASED (tonnes)
CORUS UK LTD TEESSIDE IRON WORKS (Blast furnace)* 1496
CORUS UK LTD STEEL HOUSE* 55
CORUS UK LTD TEESSIDE COKE WORKS* 33
CORUS UK LTD TEESSIDE TECHNOLOGY CENTR <1
WILTON POWER STATION <1
HECKETT MULTISERV BRITISH STEEL PLC (SR) LTD TEESSIDE WORKS <1
HECKETT MULTISERV (SR) LTD TEESSIDE WORKS <1
HECKETT MULTISERV (UK) LTD SURFACE DRESSING <1
HUNTSMAN POLYURETHANES (UK) LTD ANILINE PLANT <1
Wind rose construction
The wind speed and direction data obtained from Redcar and Cleveland Council and the Meteorological Office allowed the construction of wind roses for the town of Redcar centred at the AUN site. Four roses were constructed to examine the effect, if any, of wind speed and direction on particulate matter: (a) before the closure, (b) during the closure (c) on reopening of the plant and (d) the entire sampling period.
Chemical compositional analysis
A schematic of the extraction methodology is shown in figure 2 and followed that reported in detail by Heal [9]. The water extractable component of the PM10 samples was obtained by sonicating one filter that had been used to sample PM10 for 6–8 days in 6–8 ml of 18 Mohm water (i.e. 1 ml/24 hrs of PM10) at room temperature for 1 hr to generate suspension of dissolved and insoluble substances. Blank filters were also extracted using the same procedure for comparison. The PM10 components remaining on the filter were extracted by subsequent acid digestion using 2.8:1 HCl: HNO3 and sequentially heated and evaporated to dryness over 24 hrs. Both the aqueous extract and the acid extract samples were re-suspended in 2% HNO3 for analysis. The metal composition of the aqueous and acid extract PM10 samples was determined by inductively coupled plasma mass spectrometry (ICP-MS) to quantify the trace metal content of PM10. The elements measured were iron, zinc, copper, manganese, cobalt, nickel, chromium, vanadium, titanium, lead, arsenic and cadmium. Total metals as reported here refer to the arithmetical sum of the concentrations of these measured metals. The samples analysed for metal content are described in Table 2, these samples were also used for instillation into rats.
Figure 2 Diagram detailing the methods used to prepare samples to examine composition and toxicity of Redcar PM10.
Table 2 PM10 samples analysed for metal content and then subsequently instilled into rats. PM10 was collected using a TEOM ACCU with a flow rate of 16.7 l/min.
Filter Dates Period of collection (days) Mass of PM10 collected onto filter (μg) PM10 collected per 24 hours (μg) Maximum PM10 dose instilled (μg)
21/6–29/6 8 2893 360 180
29/6–6/7 7 2186 312 156
26/7–3/8 8 3741 466 233
1/9–7/9 6 3071 510 255
5/10–12/10 7 2011 286 143
26/10–2/11 7 1569 224 112
Intratracheal instillation of aqueous extracts of Redcar PM10
Male Wistar rats (Charles River UK LtD Manson Road Kent) were housed under standard conditions (Rats were kept between 20–22°C 4 per cage in a 12 hour light 12 hour dark cycle cages, bottles and food were changed and washed weekly). Rats weighed between 250 and 300 g at time of use (approximately 3 months old). Three rats were used for each treatment group and there were four treatment groups in total. Ethical approval for this project was obtained via the University Ethics committee.
Group one consisted of animals exposed to saline only (control), group 2 were treated with pre-closure PM10 extracts, group 3 with the closure extracts and group 4 with extracts collected on reopening of the steel plant. Each rat received the same aqueous extract of PM10 used in the metals analysis described in table 2. Saline was added to extracts prior to instillation to ensure the treatment was at physiological salt concentration. The PM10 dose given was not equalised for mass, but was the equivalent of a 24-hour PM10 exposure (table 2). It should be stressed that the values provided in table 2 are estimates based upon the flow rate of the sampler, and the ambient PM10 concentrations reported at the AUN site during the periods of collection for each filter. On this basis, the maximum PM10 dose instilled assumes a 100% efficiency for the recovery of PM10 from the filter. However, this is not the case. It was not possible to determine the efficiency of recovery by spectrophotometry due to the low turbidity of the samples recovered. Furthermore, it was not possible to reweigh the filters after extraction since the filters were digested by acid to extract the remaining metal on the filter.
The experiment was subsequently repeated after dividing the aqueous PM10 extract into soluble and insoluble extractable PM10 components. These samples were prepared from the aqueous PM10 washed from filters using water as described previously (1 ml water per 24 hours of PM10 collection), this extract was then separated into the soluble and insoluble fractions by centrifugation (12000 g). The insoluble pellet was resuspended in water (again 1 ml per 24 hours of PM10 collection). Both samples were treated with saline to generate a physiological salt concentration before subsequently instilling 0.5 ml into each male Wistar rat.
Rats were anaesthetised with halothane and then instilled intratracheally with 500 μl of treatments. As previously described 24 hr PM10 was extracted into 1 ml of water, however the exact concentrations of PM10 dose are unknown, as turbidomitry could not be carried out due to the low particle concentration and clarity of samples. The figures in table 2 represent the quantity of PM10 collected on each filter per 24 hr, but since recovery from the filter is less than 100% and each animal receives 500 μl, these figures are far greater than the dose administered. At 18 hrs following instillation the rats were euthanised by intraperitoneal injection of Euthatal and the lungs surgically removed. Eight ml of saline was injected into the lungs through a cannula and the lobes were massaged for 2 minutes to remove migratory cells and lung lining fluid. This primary bronchoalveolar lavage (BAL) fluid, removed from the lungs was kept separated from three further lavages, 8 ml each, which were pooled to form a secondary lavage. The primary lavage was kept separate from the secondary lavage in order to minimise dilution of constituents. After centrifugation (900 g for 2 minutes) the cells were re-suspended in 1 ml sterile saline and the cells from the primary and secondary lavage samples were pooled. A total cell count was determined, followed by cytospot preparations. These were stained with Diff Quick (Lamb) before determination of differential cell counts.
BAL biochemical analysis
The primary BAL from each rat was analysed for markers of cellular and tissue damage including lactate dehydrogenase (LDH) activity, [10,11] total protein [12] and albumin protein (bromocresyl green) levels. The pro-inflammatory cytokine proteins, tumour necrosis factor α (TNFα) and macrophage inflammatory protein 2 (MIP2) was also measured by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's guidelines, Biosource UK Cytosets™.
Assessment of pro-inflammatory cytokine mRNA expression in BAL cells using Multiprimer PCR
The BAL cells recovered from the control and treated animals were centrifuged (900 g, 2 min) and the pellet washed with phosphate buffered saline (PBS) before addition of 200 μl of Tri-reagent to the cells. The mixture was incubated for 10 minutes at 4°C, and stored at -80°C until required.
The mRNA purification and synthesis of cDNA was carried out following protocols provided with the Biosource Cytoxpress kit™. The human inflammatory cytokine Multiprimer PCR kit from Biosource was used to assess the mRNA expression of 6 cytokines (TNFα, transforming growth factor beta (TGFβ), MIP2, interleukin 6 (IL6), interleukin 1 beta (IL1β), granulocyte macrophage colony stimulating factor (GM-CSF) and 1 housekeeper gene (glyceraldehyde 3-phosphate dehydrogenase, GAPDH) according to the manufacturers guidelines.
The PCR products were detected and quantified by electrophoresis using a 1.5% agarose gel, in a horizontal Biorad GT system. The gels were stained with ethidium bromide and PCR products were detected using a UV transilluminator. Images were taken under UV conditions using a Synygene camera and the intensities of PCR product bands were quantified using Syngene software and expressed as a percentage of the house keeping gene (GAPDH) and then as a percentage of the negative control.
The effect of removal of Redcar PM10 metals via chelation experiments
The murine macrophage cell line, J774.1A was cultured in RPMI 1640 medium containing 10% heat inactivated foetal bovine serum (FBS), 1% L-glutamine, 0.06 U/ml penicillin, 30 mg/ml streptomycin, (all obtained from Life Technologies). The cells were grown and sub-cultured under standard conditions. Cells were removed from flasks using sterile cell scrapers (SLS, UK). J774.1A macrophage cells were treated with samples of Redcar PM10 for 4 hrs. Along side these treatments cells were treated with Redcar PM10 samples that had under gone chelation to remove metals. This was carried out by suspending particles in RPMI-1640 containing 50 mg/ml chelex beads and mixed on a rotating wheel for 4 hrs at room temperature. After incubation, samples were centrifuged at 12000 g (5 min) to pellet the chelex beads. The resultant suspensions were applied to J774.A1 cells and incubated at 37°C for 4 hrs. Cell culture supernatants were subsequently analysed for TNFα protein via ELISA (Biosource UK Cytosets™).
Statistical analysis
Experiments were conducted, at minimum, in triplicate and the data shown in each figure represents the mean of three separate experiments ± the standard error of the mean (S.E.M) unless other wise stated. Statistical significance was determined using One Way Analysis of Variance (ANOVA) with Tukey's pair wise comparison (Minitab Version 13). * p < 0.05 is denoted as being significant, with ***p < 0.001 representing high significance.
Results
Redcar PM10, wind speed and direction before, during and after blast furnace closure
The PM10 mass collected per 24 was greater during the closure period than in the preclosure or postclosure periods (Table 2). There is no information available to explain this observation, however coarse particulate emissions may have been increased during refurbishment and repair of the blast furnace lining.
Wind roses provided a visual aid when considering the effects of direction and speed. Although they can be constructed to display any period of time, the wind roses (Figure 3) prepared for the Redcar area refer to before (1/6/00 – 25/7/00), during (26/7/00–28/9/00) and after (29/9/00–31/12/00) the closure of the Corus blast furnace. A wind rose representing the whole period (June – December 2000) was also constructed.
Figure 3 Wind rose illustrating speed (knots) and direction of the wind every 15 minutes y-axis represents the number of 15 minute occurrences with the x-axis's representing direction in degrees. (a) Sampling before the closure of the blast furnace (1/6/00 – 25/6/00). (b) During the closure of the blast furnace (26/7/00 – 28/9/00). (c) Sampling after the blast furnace reopened (29/9/00–31/12/00) and (d) the total sampling period (1/6/00–31/12/00).
The wind rose constructed for the three weeks between 1/6/00 and 25/7/00 covering the pre-closure sampling period (Figure 3a) showed that during this period the majority of wind came from a north-to-north easterly direction (approx. 0–30°), however a smaller proportion was also directed from the south west (approx. 210°). The wind speed coming from the north was generally below 6 knots and predominantly less than 3 knots, with some faster episodes of 7–10 knots and 11–16 knots. South-westerly winds did reach speeds of 11–16 knots, but most ranged from between 7–10 knots with some as low as 4–6 knots.
The wind rose constructed for the closure period (Figure 3b) indicates the majority of the wind came from the SW (approx. 210°). Wind does however come from the N to NW direction, although this is minimal when compared with the volume coming from the SW. The speed of SW winds ranged from less than 3 knots to 16 knots, but the wind speeds generally occurred between 4–10 knots, although slower speeds did take place more westerly (<3 knots).
For the sample period after the blast furnace reopened the wind rose (Figure 3c) indicates that the wind came solely from the SW and that the range of speeds recorded was from less than 3 knots to 21 knots, all of equal prominence.
The wind rose of Figure 3d covers all three time points discussed previously summarising wind speed and direction for the whole sampling period. The chart indicates that the majority of the wind came from the SW with speeds ranging from less than 3 knots to 21 knots; the most commonly recorded wind speeds fell within 4 – 16 knots. A relatively small fraction came from the N to NE direction at a wind speed predominantly less than 3 knots.
PM10 atmospheric concentrations from sampling periods in years prior to, during and after the closure
Table 3 lists the maximum and the minimum 24 hour PM10 concentrations observed throughout the sampling period for 2000 and for the same period during 1999 and 2001. The lowest maximum and minimum mean daily PM10 concentrations occurred during the year the steel plant closed (46 μgm-3 and 4 μgm-3 respectively). The year before and after the closure of the plant saw maximum mean daily PM10 concentrations, exceeding the EU and UK 24 hour ambient concentration limit values of 50 μg/m3, that should not be exceeded more than 3 times in one year (Table 3).
Table 3 The daily mean PM10 concentrations (μgm-3) during the sampling period in 2000 the same periods in 1999 and 2001 for the Redcar and Cleveland area. (Data obtained from NETCEN).
Maximum PM10 Concentration μgm-3 Minimum PM10 Concentration μgm-3
Year Value Date Value Date
1999 50 06/09/99 6 27/09/99
2000 46 11/09/00 4 18/09/00
2001 52 11/12/01 5 12/08/01
Redcar PM10 metals analysis
The metal content of 7-day PM10 samples collected before, during and after the short-term closure of the Corus steel plant in Redcar was determined by ICP-MS. The PM10 samples were subjected to both aqueous and acid extraction sequentially as described in the methods. The combined results for both the aqueous and acid extractions were summed to give the total metal content of the PM10 samples. There was a significant increase in the total and acid extractable metal content of the PM10samples collected after the plant reopened when compared to that collected during the closure period (Figure 4). The aqueous extractable metal content did not differ significantly between the open and closed periods, although changes in specific transition metals did occur as, described below.
Figure 4 The measured metal content of 7 day PM10 samples collected before, during and after closure (* p < 0.05 compared to closure period). Extracts were made into ultra pure water (aqueous extract) followed by digestion of the remaining filter in HCl:HNO3(acid extract). Measurements were conducted by ICP-MS and values are the mean of 2 samples ± SEM.
Figure 5a shows the aqueous extractable transition metal components of the same PM10 samples described above. The soluble iron content was considerably lower than the total iron content, indicating a substantial proportion of iron was insoluble. Furthermore the soluble iron content of PM10 did not significantly alter between the open and closed periods of collection. In contrast, soluble zinc, which occurs at notable levels in all samples, increased dramatically on reopening of the plant (1.86 ng/μg PM10 compared with 0.26 ng/μg PM10 during closure). In addition, both copper and manganese increased significantly on reopening when compared to the closure period (0.33 ng/μg PM10 compared to 0.03 ng/μg PM10 and 0.7 ng/μg PM10 compared with 0.05 ng/μg PM10 respectively). Figure 5b shows data collected from the acid digest of the filter and PM10 not removed by the aqueous extraction and hence represents mainly the insoluble metal components of the PM10. Iron was the most abundant of all the acid extractable metals analysed and increased greatly on reopening of the steel plant (5.81 ng/μg PM10 compared with 0.69 ng/μg PM10). As observed in the aqueous extract both copper and manganese increased significantly in the acid extract on reopening when compared to the closure period (0.15 ng/μg PM10 compared to 0.01 ng/μg PM10 and 0.22 ng/μg PM10 compared to 0.02 ng/μg PM10).
Figure 5 Metal content of PM10 collected before, during and after the closure of the Redcar Corus steel plant. (a) Aqueous extractable (b) acid extractable metal content of PM10. Measurements were conducted by ICP-MS and values are of individual filter samples
Toxicology of Redcar PM10
Samples of the same aqueous extracts of PM10 analysed by ICP-MS were subsequently instilled into male Wistar rats. The aqueous PM10 extracts taken before and during the closure did not alter significantly the total number of lavage cells recovered (table 4) nor did the aqueous extracts induce any significant increase in neutrophil content (neutrophil number or % neutrophils) of BAL when compared to the saline control (Figure 6a and table 4). However PM10 extracts from the reopened period induced a significant increase in neutrophil cell number and percentage neutrophils when compared to animals treated with the extracts of PM10 from the closed period or the control animals (Figure 6a and table 4).
Figure 6 The mean neutrophil number in bronchoalveolar lavage from rat's 18 hr after exposure to either (a) total aqueous extracts or (b) the soluble or insoluble fractions of aqueous extracts of PM10 samples collected before, during and after the steel plant closure. Control animals were instilled with saline. Values represent the mean of (a) 6 and (b) 3 rats ± SEM (** p < 0.01 and * p < 0.05 to control, $ p < 0.05 closed to reopened
Table 4 The mean bronchoalveolar lavage total cell count and percentage neutrophil influx ± SEM from rat's 18 hr after exposure to either total aqueous extracts or the soluble or insoluble fractions of aqueous extracts of PM10 samples collected before, during and after the steel plant closure. Control animals were instilled with saline.
Treatment Total cell number × 106 ± SEM % Neutrophils ± SEM
Aqueous Soluble Insoluble Aqueous Soluble Insoluble
Control 2.83 ± 0.34 3.85 ± 0.59 3.85 ± 0.59 3 ± 0.9 4 ± 0.9 4 ± 1.0
Open 3.40 ± 0.27 3.02 ± 0.16 4.16 ± 0.13 8 ± 1.8 9 ± 0.9 10 ± 0.9
Closed 3.06 ± 0.29 3.16 ± 0.18 2.63 ± 0.46 8 ± 1.0 10 ± 0.9 7 ± 1.0
Reopened 3.71 ± 0.38 2.85 ± 0.74 3.12 ± 0.17 25 ± 4 13 ± 1.5 10 ± 1.5
The soluble and insoluble extractable PM10 components that were washed from filters in the aqueous extract were separated by centrifugation and subsequently instilled into male Wistar rats. The soluble PM10 fraction of the extracts taken before and during the closure did not induce any significant changes in the number or percentage of neutrophils in BAL when compared to the saline control (figure 6b and table 4). However the water soluble fraction of aqueous PM10 extracts from the reopened period induced a significant increase in neutrophil cell number (p < 0.05) when compared to the control (Figure 6b and table 4). The insoluble fraction of PM10 washed from the filter taken before and during the closure did not induce any significant inflammogenic effect when compared to the saline control (Figure 6b and table 4). However the insoluble components of PM10extracts from the reopened period induced a significant increase in neutrophil cell number (p < 0.05) when compared to both the control and (p < 0.05) closed period samples (Figure 6b). The neutrophil cell numbers counted in BAL after treatment with the soluble and insoluble extracts from the reopened periods were each approximately half those obtained on treatment with the whole sample from the reopened period. In fact, these values when added together equalled the neutrophil influx measured for the total aqueous extract. However, the neutrophil values obtained for the insoluble and soluble exracts did not add up to equal the neutrophil response observed for the total aqueous extract as the increase in neutrophil influx was not significant for these periods.
Treatment of the rats with whole aqueous extracts of PM10 from any collection period did not significantly increase BAL content of MIP2 or TNFα when compared with the saline control (table 5). However, the overall trend of results are similar to those observed for the neutrophil cell count and the PM10 metals content, that is an increase in neutrophil and metal levels were observed when the plant was reopened compared with the closure period.
Table 5 Biochemical analysis of primary bronchoalveolar lavage fluid. Markers of inflammation measured included the chemokine MIP-2 and the cytokine TNFα. Total protein content, albumin and LDH were all measured Values are the means of 3 observations ± SEM.
Protein ug/ml LDH U/ml MIP2 pg/ml TNFα pg/ml
Control 121.67 ±28.67 236.75 ± 55.42 22.52 ± 22.52 2.20 ± 2.20
Open 97.33 ± 13.03 229.63 ± 33.57 33.14 ± 12.96 0.00 ± 0.00
Closed 103.50 ± 13.41 233.17 ± 34.58 51.74 ± 16.17 2.07 ± 2.07
Reopened 114.83 ± 13.20 263.66 ± 34.33 80.61 ± 29.99 6.84 ± 3.17
Markers of lung damage including total protein and LDH did not increase in the BAL fluid of rats exposed to the whole aqueous extract of PM10 for 18 hrs when compared to the saline instilled rats. In contrast, the albumin content of BAL fluid increased significantly in rats instilled with PM10collected when the steel plant reopened compared to the control animals (Figure 7).
Figure 7 The mean concentration of albumin protein in BAL fluid from lungs exposed to aqueous extracts of PM10 and saline control (* p < 0.05 to control). Values represent the mean of 3 experiments ± SEM.
The mRNA expression of a range of pro-inflammatory cytokines (IL1β, IL6, MIP2, TNFα, TGFβ and GM-CSF) by BAL cells was analysed in response to exposure of rats to either saline (control) or aqueous extracts of PM10 by RT-PCR. The PM10 collected during any period of steel plant operation did not alter the mRNA expression levels of the cytokine TNFα, and the pro-fibrotic and inflammatory cytokine TGFβ when compared with the control (Figure 8). In contrast, mRNA expression of the cytokine IL1β by BAL cells did increase significantly in rats instilled with extracts of PM10 obtained on reopening when compared with the control. The mRNA expression of IL1β exhibits a similar trend to that observed for the metals analysis (Figure 4) and neutrophil influx (Figure 6). The mRNA for IL6, MIP2 and GM-CSF were not detectable in the BAL cell extracts from either control or treated animals.
Figure 8 The mRNA expression by rat bronchoalveolar lavage cells 18 hr following instillation of aqueous extracts of PM10 collected before, during and after the steel plant closure (* p < 0.05 to control). Values represent the mean of 6 rats ± SEM.
Chelation of Redcar PM10 metals
J774.A1 cells were treated for 4 hrs with Redcar PM10 samples taken from during the closure and on reopening of the plant. Cells were also treated with identical PM10 samples that previously underwent treatment with Chelex beads for 4 hrs to remove metals from samples.
Both the closed and reopened sample treatments stimulated the macrophage cells to increase TNFα protein release compared to control cells. Chelated closure PM10 sample treatments showed lower TNFα levels when compared with non-chelated treatments, almost returning to control levels (figure 9). Reopened chelated PM10 samples indicate a significantly lower TNFα production than non-chelated treatments, a significant decrease (p < 0.05) again almost returning to control levels (figure 9).
Figure 9 TNFα protein production in J774.A1 cells treated for 4 hrs with Redcar PM10sample during and after the closure of the Corus plant and identical samples which underwent 4 hrs of chelating treatment. *p < 0.05 when compared to control, $p < 0.05 when compared to untreated reopened PM10 samples. Values represent the mean of 3 experiments ± SEM.
Discussion
The PM10 sampling convention is mass-based and does not take account of composition. In this study the daily airborne mass of PM10 monitored by the AUN site before, during and after the steel plant closure did not alter significantly [13]. The samples studied had the greatest mass for the closure period. The ability of PM10 to induce inflammation in the rat lung when instilled varied greatly between the collection periods, with PM10 from the reopened period being most potent in causing inflammation despite the probability that this sample contained the lowest PM10 mass. The ability to cause pulmonary inflammation is generally considered to play an important role in the pulmonary and cardiovascular effects associated with increased PM exposure [14-16]. This study confirms the importance of composition in driving the pro-inflammatory effects of PM in this animal model, and by implication the adverse effects in exposed human populations. However, in this study it was not possible to determine the exact mass dose administered to each animal due to the low concentration of the samples employed. What is provided is an estimate of the maximum dose instilled per animal based upon the flow rate and the ambient PM10concentration. An identical extraction procedure was used for the PM10 samples collected from each sampling period, and this study assumes that the extraction efficiency is identical across all periods. Significant changes in particle composition however, could impact upon this process. However, an exact analysis of the metal content of each instilled sample was obtained, and as discussed below the metal content of the PM10 did impact directly on the potency of the sample instilled into each animal.
Data on the metal content of the PM10 showed that it was greatest in the 'reopened' sample, supporting the contention that this was indeed the factor that was responsible for the differences in ability to cause inflammation between the different PM samples. Study of the PM collected around a steel mill closure in Utah Valley, revealed a similar impact on composition. A greater inflammogenicity in rats exposed to the PM collected when the Utah plant was open was replicated in human lungs exposed by instillation [4,5].
The metal composition of PM10 samples varied, between the Redcar plant being operational and closed, in terms of both total metal content and concentrations of individual metals. The total metal content of PM10collected before the closure and during the closure was similar, but on reopening of the steel plant there was a 3-fold increase in total metal content of PM10 compared with the closure and pre-closure samples. This generally supports the hypothesis that, when the steel plant was closed, metal emissions were lower than during the operational period. The metal content of PM10 was relatively low before the closure of the plant but this may be a result of wind direction during the pre-closure sampling period, as the wind predominantly originated from the North Sea and hence not the Corus operations, in contrast to the closure and re-opened sampling periods. Although the change in metal content of the PM in relation to closure of the Corus plant may have occurred by chance and was a result of changes in emissions from other local sources, the emission inventory for this area shows such a remarkably predominant contribution from the Redcar steelworks that this seems unlikely [8].
The large increase in total metal on reopening of the steel plant was reflected in an increase in the acid extractable metals but not in the aqueous extractable metals, suggesting that the emissions from the steel plant contain metals which are predominantly water insoluble. However, within the total, acid and aqueous extracts of PM10, the individual metal contents varied significantly between the closed and reopened period. Of the twelve metals measured by ICP-MS, iron was the most abundant metal in both the total and acid-extractable fractions, in keeping with the nature of work being carried out at the plant. However in the aqueous extracts iron levels were small in comparison with other metals such as zinc and manganese, which again suggests that the iron particulate produced by the steel plant is predominantly insoluble. The metals which increased in the aqueous extract of PM10 on reopening of the steel plant included zinc, manganese and copper.
The same aqueous extracts analysed for metal content were also instilled into rats to determine the inflammogenic potency. The whole aqueous extractable Redcar PM10 samples collected prior to or during the steel plant closure, did not induce a significant increase in inflammation in the rat lung compared to the control. However whole PM10 collected on reopening of the plant, induced a significant increase in neutrophil influx into the lung. As discussed previously, the metals in PM10 that increased the most in the reopened period compared to the closed period were zinc, manganese and copper. Hence there was a clear relationship between metal content and inflammogenic potency of the PM10 samples. Metal chelation of the Redcar samples in vitro showed a significant decrease in pro-inflammatory protein production when compared to non-chelated treatments, re-affirming the suggested link between metal content and levels of inflammation.
When the soluble and insoluble sub-fractions of the PM10 collected during the reopened period were instilled, both samples induced neutrophil recruitment that was almost half that observed compared to the whole PM10 fraction from the same period. This would suggest that there are components in both the soluble and insoluble fraction that drive inflammation, and that they induced an additive effect in the rat lung. However, such an observation was not obvious for the pre-closure and closure period samples, possibly because none of these samples induced any significant increase in neutrophil influx compared to the control.
In addition to inducing inflammation, as indicated by cellular changes with in the lung, instillation with the whole aqueous extract of PM10 collected during the reopened period also induced a significant increase in mRNA expression of the pro-inflammatory cytokine IL1β in the BAL cells. In the same samples, the expression of TNFα and TGFβ mRNA expression did not change at the 18 hr time point investigated. Similarly, the chemokine MIP2 and pro-inflammatory cytokine TNFα protein content of BAL showed no statistical changes when compared with controls, but did exhibit similar trends to total metal content and neutrophil influx. The lack of significance was probably due to the dilution of BAL fluid during collection, as well as the use of a time point suitable for neutrophil influx but not TNFα measurement.
In agreement with previous work carried out in our laboratory, low doses of PM10 instilled into rat lungs did not induce indicators of gross cellular damage to lung cells (Lightbody et al., manuscript in preparation). However damage to the endothelium/epithelial barrier and hence an increase in the permeability of the vasculature was observed as indicated by an increase in the albumin content of the BAL fluid in rats exposed to PM10 extracts obtained on reopening of the steel plant. Hence, the PM10 samples that induced the greatest inflammation also induced the greatest lung damage and contained the highest metal content (especially for zinc, copper and manganese).
A number of recent studies have reported toxicity of metals and implicated them in the biological potency of ambient particulate matter [17-19]. Transition metal content of PM10 has been shown by Jimenez et al., [20] to activate transcription factors such as nuclear factor kappa B in epithelial cells, so linking transition metal content to the induction of inflammation. Wilson et al., [21] also showed that ultrafine carbon particles induced an inflammation in the rat lung that was potentiated by the addition of iron chloride, establishing an interaction between metals and particles in enhancing potency.
In conclusion, the PM10-induced inflammation was related to the concentration of metals in the PM10 samples tested and the reopening of the steel plant was associated with an increase in the metal content of the PM sufficient to change an average 180 μg/ml instilled dose from non-inflammogenic to inflammogenic. It is unlikely that all of the measured metals were biologically active in the Redcar PM10 samples, with some associations being correlative rather than causative. The potency of the biologically active metals is likely to vary, for example, Riley et al.[22] was able to rank metal toxicity to rat lung epithelial cells (V>Zn>Cu>Ni>Fe) by measuring TC50. Rice et al., [18] also showed that instillation of soluble metals into the rat lung induced inflammation. Copper was the most pro-inflammatory metal in their study followed by manganese and nickel, while vanadium, iron and zinc induced similar levels of inflammation. This study analysed the metal content of Redcar PM10 and related composition to toxicological effects observed. The effects of elemental carbon, organic compounds and biological components cannot be ignored, however no data on whether organic components of PM actually changed during the course of this study are available.
In conclusion, this study indicates that the operations of a local UK Steel plant impacts significantly on the metal content of PM10. Furthermore, this study confirms previous observations that the metal composition of PM10 is related to its ability to drive inflammation in the rat lung. This was confirmed by the ability of metal chelation to block the in vitro effects of the metal rich PM10 samples. Such observations are important when considering the potential impact on the health of susceptible people living within the vicinity of such industrial operations.
Acknowledgements
We would like to thank Mr Bob Cowell for his generous help in the collection of PM10 samples in Redcar.
This work was funded by Defra and the Devolved Administrators as well as Napier University.
==== Refs
Pope CA III Particulate pollution and health: a review of the Utah valley experience J Expo Anal Environ Epidemiol 1996 6 23 34 8777371
Stone V Wilson M Lightbody JH Donaldson K Investigating the potential for interaction between the components of PM10 Environ Health Prev Med 2003 7 6 246 253 10.1265/ehpm.2003.246
Pope CA IIISchwartz J Ransom MR Daily mortality and PM10 pollution in Utah Valley Arch Environ Health 1992 47 211 217 1596104
Ghio AJ Devlin RB Inflammatory lung injury after bronchial instillation of air pollution particles Am J Respir Crit Care Med 2001 164 704 708 11520740
Dye JA Lehmann JR McGee JK Acute pulmonary toxicity of particulate matter filter extracts in rats: coherence with epidemiologic studies in Utah Valley residents Environ Health Perspect 2001 109 395 403 11427389
Frampton MW Ghio AJ Samet JM Carson JL Carter JD Devlin RB Effects of aqueous extracts of PM10 filters from the Utah valley on human airway epithelial cells Am J Physiol 1999 277 L960 L967 10564181
Molinelli AR Madden MC McGee JK Stonehuerner JG Ghio AJ Effect of metal removal on the toxicity of airborne particulate matter from the Utah Valley Inhal Toxicol 2002 14 1069 1086 12396411 10.1080/08958370290084737
Environment Agency wed site 2004
Heal MR Hibbs LR Agius RM Beverland IJ Total and water-soluble trace metal content of urban background PM10, PM2.5 and black smoke in Edinburgh, UK Atmospheric Environment 2005 39 1417 1430 10.1016/j.atmosenv.2004.11.026
Brown DM Wilson MR MacNee W Size-dependent proinflammatory effects of ultrafine polystyrene particles: A role for surface area and oxidative stress in the enhanced activity of ultrafines Toxicol Appl Pharmacol 2001 175 191 199 11559017 10.1006/taap.2001.9240
Henderson RF Benson JM Hahn FF New approaches for the evaluation of pulmonary toxicity: Bronchoalveolar lavage fluid analysis Fundam Appl Toxicol 1985 5 451 458 3891479 10.1016/0272-0590(85)90092-2
Bradford MM A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding Anal Biochem 1976 72 248 254 942051
UK Air quality database 2004
Donaldson K MacNee W Potential mechanisms of adverse pulmonary and cardiovascular effects of particulate air pollution PM10 Int J Hyg Environ Health 2001 203 411 5 11556145
Donaldson K Stone V Borm PJ Oxidative stress and calcium signalling in the adverse effects of environmental particles PM10 Free Radical Biology and Medicine 2003 34 1369 1382 12757847 10.1016/S0891-5849(03)00150-3
Pope CA IIIBurnett RT Thurston GD Cardiovascular mortality and long-term exposure to particulate air pollution: epidemiological evidence of general pathophysiological pathways of disease 2003 109 71 77 14676145
Carter JD Ghio AJ Samet JM Devlin RB Cytokine production by human airway epithelial cells after exposure to an air pollution particle is metal-dependent Toxicol Appl Pharmacol 1997 146 80 188 10.1006/taap.1997.8254
Lambert AL Dong W Selgrade MK Gilmour IM Enhanced Allergic Sensitization by residual oil fly ash particles is mediated by soluble metal constituents Toxicol Appl Pharmacol 2000 165 84 93 10814556 10.1006/taap.2000.8932
Rice TM Clarke RW Godleski JJ Al-Mutairi E Jiang NF Hauser R Paulauskis JD Differential ability of transition metals to induce pulmonary inflammation Toxicol Appl Pharmacol 2001 177 1 46 53 11708899 10.1006/taap.2001.9287
Jimenez LA Thompson J Brown DA Rahman I Antonicelli F Duffin R Drost EM Hay RT Donaldson K MacNee W Activation of NF-kappaB by PM10 occurs via an iron-mediated mechanism in the absence of IkappaB degradation Toxicol Appl Pharmacol 2000 166 101 110 10896851 10.1006/taap.2000.8957
Wilson MR Lightbody JH Donaldson K Sales J Stone V Interactions between Ultrafine Particles and Transition Metals in Vivo and in Vitro Toxicol Appl Pharmacol 2002 184 172 179 12460745 10.1006/taap.2002.9501
Riley MR Boesewetter DE Kim AM Sirvent FP Effects of metals Cu, Fe, Ni, V, and Zn on rat lung epithelial cells Toxicology 2003 190 3 171 84 12927373 10.1016/S0300-483X(03)00162-8
| 15904485 | PMC1156955 | CC BY | 2021-01-04 16:23:26 | no | Respir Res. 2005 May 18; 6(1):43 | utf-8 | Respir Res | 2,005 | 10.1186/1465-9921-6-43 | oa_comm |
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-481592707810.1186/1465-9921-6-48ResearchActivation of the SPHK/S1P signalling pathway is coupled to muscarinic receptor-dependent regulation of peripheral airways Pfaff Melanie [email protected] Norbert [email protected] Sibel [email protected]ütz Werner [email protected] Yoshiko [email protected] Silke [email protected] Wolfgang [email protected] Jürgen [email protected] Rainer Viktor [email protected] Institute for Anatomy and Cell Biology Justus-Liebig-University Giessen, Germany2 Department of Cell Signaling, Graduate School of Medicine, Gifu University, Gifu, Japan3 Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive Kidney Diseases, Bethesda, Maryland 20892, USA2005 31 5 2005 6 1 48 48 12 11 2004 31 5 2005 Copyright © 2005 Pfaff et al; licensee BioMed Central Ltd.2005Pfaff et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In peripheral airways, acetylcholine induces contraction via activation of muscarinic M2-and M3-receptor subtypes (M2R and M3R). Cholinergic hypersensitivity is associated with chronic obstructive pulmonary disease and asthma, and therefore the identification of muscarinic signaling pathways are of great therapeutic interest. A pathway that has been shown to be activated via MR and to increase [Ca2+]i includes the activation of sphingosine kinases (SPHK) and the generation of the bioactive sphingolipid sphingosine 1-phosphate (S1P). Whether the SPHK/S1P signaling pathway is integrated in the muscarinic control of peripheral airways is not known.
Methods
To address this issue, we studied precision cut lung slices derived from FVB and M2R-KO and M3R-KO mice.
Results
In peripheral airways of FVB, wild-type, and MR-deficient mice, SPHK1 was mainly localized to smooth muscle. Muscarine induced a constriction in all investigated mouse strains which was reduced by inhibition of SPHK using D, L-threo-dihydrosphingosine (DHS) and N, N-dimethyl-sphingosine (DMS) but not by N-acetylsphingosine (N-AcS), a structurally related agent that does not affect SPHK function. The initial phase of constriction was nearly absent in peripheral airways of M3R-KO mice when SPHK was inhibited by DHS and DMS but was unaffected in M2R-KO mice. Quantitative RT-PCR revealed that the disruption of the M2R and M3R genes had no significant effect on the expression levels of the SPHK1-isoform in peripheral airways.
Conclusion
These results demonstrate that the SPHK/S1P signaling pathway contributes to cholinergic constriction of murine peripheral airways. In addition, our data strongly suggest that SPHK is activated via the M2R. Given the important role of muscarinic mechanisms in pulmonary disease, these findings should be of considerable therapeutic relevance.
==== Body
Background
Acetylcholine (ACh), released from parasympathetic nerve fibres, leads to bronchoconstriction via stimulation of muscarinic acetylcholine receptors (MRs) and subsequent increase in intracellular calcium levels [Ca2+]i. The MR-family consists of five molecularly distinct subtypes (M1–5R) that are coupled to heterotrimeric G-proteins [1,2]. Activation of the M2R and M4R subtypes generally decreases intracellular cAMP levels, whereas stimulation of the M1R, M3R and M5R subtypes leads to the activation of phospholipase Cβ (PLCβ) with subsequent generation of the second messenger inositol 1, 4, 5-trisphosphate (IP3) and an increase in [Ca2+]I [1,2]. Asthmatic patients show hypersensitivity to MR agonists, and consequently antimuscarinic agents are commonly used in treatment of upper and lower airway diseases [3].
Several studies suggest that the PLC-IP3-signalling pathway is not solely responsible for changes in [Ca2+]I. For example, activation of the cyclic ADP-ribose pathway abolishes ACh-induced Ca2+ oscillations in smooth muscle of porcine airways (White et al. 2003). Another alternative pathway that has been shown to increase [Ca2+]i involves the activation of sphingosine kinases (SPHK) and the subsequent generation of the bioactive sphingolipid sphingosine 1-phosphate (S1P) [4-6]. While S1P is well known as an important extracellular mediator of many biological pathways, including cell survival, angiogenesis and cell migration, recent studies indicate that S1P is also an intracellular second messenger which is coupled to changes in intracellular Ca2+ levels [4-6]. In the airways, S1P has been shown to stimulate airway smooth muscle proliferation and cytokine release [7]. When applied to cultured human tracheal myocytes, S1P also increases [Ca2+]i and evokes contractile responses [7,8]. Moreover, muscarinic activation of M2R-and M3R-transfected HEK293 cells stimulates S1P synthesis [4,5]. In the present study we therefore tested the hypothesis that activation of the SPHK/S1P signalling pathway may contribute to MR-dependent regulation of peripheral airway diameter. For all studies we used airways from wild-type (wt) and M2R and M3R mutant mice that were ~200 μm in diameter. It is well known that smaller airways are mainly responsible for airway resistance [9,10]. Moreover previous studies clearly demonstrated differences between larger and smaller airways concerning functional responses, receptor expression or ion conductance [11].
In murine tissues, S1P is synthesized after activation of two SPHK isoforms, SPHK1 and SPHK2. SPHK1 is highly expressed in adult lung [12-14], whereas the SPHK2 isoform shows much lower expression in lung tissue [15]. SPHK2 contains a nuclear localization sequence and has recently been identified as a nuclear protein capable of inhibiting DNA synthesis, whereas SPHK1 is mainly localized in the cytosol [14,16]. In the present study we therefore focused exclusively on the expression and potential functional role of the SPHK1 isoform in MR-mediated airway constriction.
The expression of the SPHK1 isoform was investigated by means of quantitative RT-PCR and immunohistochemistry using precision cut lung slices (PCLS) of murine lungs [17]. Using murine PCLS we also measured the constriction of small intraparenchymal airways in response to MR activation. Coupling of MRs to SPHK-activation was investigated by blocking SPHK with D, L-threo-dihydrosphingosine (DHS) or N, N-dimethyl-sphingosine (DMS, [18]). Moreover, the role of MR-dependent intracellular Ca2+ release on airway diameter was studied by inhibiting Ca2+-influx by La3+ and SKF 96365. Recent evidence indicates that MR agonists induce bronchoconstriction by activating a mixture of M2R and M3R subtypes, present on smooth muscle cells of extra-and intraparenchymal airways [19,20]. To study the potential roles of the M2R and M3R subtypes in SPHK1-dependent peripheral airway responses, we therefore also carried out functional studies with PCLS prepared from M2R-and M3R-deficient mice (M2R-KO, M3R-KO) and their corresponding wild type controls (M2R-wt, M3R-wt) [21,22].
Methods
Animals
Lungs were taken from 8-12 wk old FVB mice (Harlan-Winkelmann, Borchen, Germany), mice deficient in M2R or M3R (M2R-KO, M3R-KO) and their corresponding wild-type strains (M2R-wt, M3R-wt). The generation of M2R-KO-and M3R-KO mice has been described previously [21,22]. The M2R-KO mice and the M2R-wt mice had the following genetic background: 129J1 (50%) × CF1 (50%). The M3R-KO mice and the corresponding wild-type mice had the following genetic background: 129SvEv (50%) × CF1 (50%). The animals were killed by cervical dislocation. The mice were kept under specific pathogen free conditions until the experiments.
Quantitative RT-PCR
Real-time quantitative PCR (iCycler, Bio-Rad, München, Germany) was used to quantify levels of SPHK1 mRNA in PCLS of FVB, M2R-KO, M3R-KO, M2R-wt, and M3R-wt mice. Lung slices were transferred into lysis buffer (Qiagen, Heiden, Germany) and homogenized using a mixer mill with a frequency of 300 Hz (Qiagen). Total RNA was isolated according to the protocol recommended by the manufacturer (Rneasy kit, Qiagen). Contaminating DNA was removed using DNase (1 U/μg total RNA, Gibco-BRL, Life Technologies, Karlsruhe, Germany) in the presence of 20 mM Tris-HCl (pH 8.4), 2 mM MgCl2, 50 mM KCl for 15 min at 25°C. Equal amounts of RNA were reverse transcribed in the presence of 3 mM MgCl2, 75 mM KCl, 50 mM Tris-HCl (pH 8.3), 10 mM dithiothreitol, 0.5 mM dNTPs (Gibco-BRL) and 25 μg oligo (dT) (MWG Biotech, Ebersberg, Germany), with 200 U of Superscript RNase H- Reverse transcriptase (Gibco-BRL) for 50 min at 42°C. Gene specific PCR primers for mouse SPHK1 and β2-microglobulin (SPHK1, gi:22094104, fw TCCAGAAACCCCTGTGTAGC, rev GCTCCCTAGGGCCAGTAAAC product size 188 bp, β2-microglobulin gi:12861272 fw ATGGGAAGCCGAACATACTG, rev CAGTCTCAGTGGGGGTGAAT, product size 176 bp) were designed using Primer Express™ software (Applied Biosystems, Foster City, USA). All PCR-reactions were prepared in triplicate from four to eight animals using a ready-to-use kit according to the manufacturers protocol (QuantiTect™ SYBR Green PCR Kit, Qiagen). Primers specific for β-microglobulin were used for standardisation. The data were normalised by subtracting the threshold cycle (CT) levels between SPHK1 and β2-microglobulin. In each independent experiment qRT-PCR reactions were performed in triplicate.
Double-labelling immunofluorescence
Slices (220 μm thick) prepared for videomorphometry were fixed for 20 min in ice-cold acetone and washed repeatedly in 0.1 M phosphate buffer. Sections were covered for 1 h with blocking medium (50 % normal porcine serum in PBS) followed by overnight incubation with an antiserum directed against the SPHK1 isoform (1:400, [23]) in combination with a monoclonal FITC-labelled anti-α-smooth muscle actin antibody (1:500, clone 1A4, Sigma, Deisenhofen, Germany). The sections were then washed in PBS and covered for 1 h with Cy3-conjugated donkey anti-rabbit Ig antiserum (1:3000, Dianova, Hamburg, Germany). After incubation with the secondary antibody, the slides were washed in PBS and coverslipped in carbonate-buffered glycerol at pH 8.6. Omission of primary antisera or preabsorption of the SPHK1a antiserum with the corresponding synthetic peptide (20–100 μg antigen/ml diluted antiserum) abolished immunolabelling. The slides were evaluated by sequential confocal laser scanning microscopy (TCPSP, Leica, Bensheim, Germany) using the appropriate laser for Cy3 (excitation 543 nm) and FITC (excitation 488 nm).
Videomorphometry
PCLS were prepared using a slightly modified version of the protocol described by Martin et al. [24]. Briefly, the mice were killed by cervical dislocation and the lungs were perfused via the right ventricle with 37°C Krebs-Ringer buffer containing heparin (1000 I.U.), penicillin/streptomycin (1 %) and sodium nitroprusside (0.075 μM). The airways were filled via the cannulated trachea with agarose (low melting point agarose, 1.6 % in Krebs-Ringer buffer, Sigma, Deisenhofen, Germany). Subsequently, the lungs and heart were removed en bloc, placed in ice-cold HEPES-Ringer buffer and cut in 200–250 μm thick slices using a vibratome (VT1000S, Leica). Subsequently, the precision cut lung slices (PCLS) were incubated in minimal essential medium (MEM) at 37°C for 4–7 h. Experiments were performed in a lung slice superfusion chamber (Hugo Sachs Elektronik, March, Germany) mounted on an inverted microscope (Leica). Images were recorded using a CCD-camera (Stemmer Imaging, Puchheim, Germany) and analyzed using the Optimas 6.5 image analysis software (Stemmer). The slices were fixed in the chamber with nylon strings that were connected to a platinum ring. Viable airways of about 200 μm in diameter were examined and incubated in the slide chamber for 5 min in HEPES-Ringer buffer until the first image was acquired. The area of the airway lumen at the beginning of the experiment was defined as 100 % and bronchoconstriction or dilatation were expressed as relative decrease or increase of this area. Data from FVB mice and wt-strains were used only from those experiments where the reduction of luminal area in response to 10-6 M muscarine reached at least 25 %. Muscarine, [propoxy]-ethyl-1H-imidazole] (SKF96365), lanthanum chloride and N-acetylspingosine (N-AcS) were purchased from Sigma. D, L-threo-dihydrosphingosine (DHS) and N,N-dimethyl-sphingosine (DMS) were purchased from Biomol (Hamburg, Germany).
Statistical analysis
Data are presented as means ± standard error of the mean (SEM) of 5–10 slices obtained from five to nine animals. Matched pairs were evaluated by Wilcoxon's rank sum test. In the case of more than 2 non-matched groups, Mann-Whitney U-test for comparison between two groups was conducted only when statistically significant differences were reached by the global Kruskal-Wallis test that was performed first. Differences were considered as statistically significant when p < 0.05.
Results
The goal of the present study was to determine the potential involvement of SPHK1 activation in the MR-mediated constriction of small peripheral airways. To be able to measure the diameter of small murine airways, we used videomicroscopy of viable precision-cut lung slices (PCLS; [24,20]). Specifically, we analyzed the effects of SPHK1 blockade in FVB and M2R and M3R single-knockout mice (M2R-KO and M3R-KO mice, respectively) as well as the corresponding wild-type control animals (M2R-wt and M3R-wt mice, respectively).
qRT-PCR
We used qRT-PCR analysis to quantitate and compare SPHK1-mRNA levels between FVB mice, M2R-and M3R-deficient mice (M2R-KO, M3R-KO), and the corresponding wt control mice (M2R-wt, M3R-wt). Expression of β2-microglobulin served as an internal control. We found that inactivation of the M2R and M3R genes had no significant effect on the relative expression levels of SPHK1 compared to their corresponding wt controls (Mann-Whitney U-test, Fig. 1A). This analysis also showed that SPHK1 expression was significantly lower in M2R-wt and M2R-KO mice, as compared to FVB, M3R-wt, and M3R-KO-mice (p < 0.05, Mann-Whitney U-test, Fig. 1A), perhaps due to the fact that the M2R-KO/wt mice have a genetic background that is somewhat different from that of the other mice used (for details see "Materials and Methods").
Figure 1 A) Quantitative RT-PCR analysis of SPHK1 expression in PCLS from different mouse strains. The relative expression of SPHK1-mRNA in relation to the house-keeping gene β2 microglobulin is shown. Data are given as means ± S.E.M. of four independent experiments each carried out in triplicate. B) Indirect immunofluorescence studies examining SPHK1 expression in PCLS of FVB-mice. Immunoreactivity for SPHK1 was present in the wall of larger (a) and smaller (b) airways and in the media of pulmonary vessels (arrows in a, b). Consecutive sections (c, d) showing that preabsorption of the SPHK1 antiserum abolished immunolabelling (Pre, d). Bars = 50 μm.
Immunohistochemistry
We used single-and double-labeling immunohistochemistry to examine the distribution of SPHK1 in murine peripheral lung. Strong SPHK1-immunoreactivity was detected in the smooth muscle cells of intraparenchymal bronchi (Figs. 1B, 2) and pulmonary arteries (Fig. 1B) and veins as confirmed by double-staining of a marker of smooth muscle cells, α-smooth muscle actin. Preabsorption of the SPHK1 antiserum abolished immunolabelling (Fig. 1B). Immunoreactivity for SPHK1 was absent in bronchial epithelium (Figs. 1B, 2). SPHK1 showed a granular cytoplasmatic localisation but was also found in smooth muscle membranes (Fig. 2). The pattern of SPHK1 immunoreactivity was very similar in PCLS of FVB, M2R-KO/wt, and M3R-KO/wt mice..
Figure 2 Expression of SPHK1 in PCLS from different mouse strains studied via confocal laser scanning microscopy. Representative confocal laser scanning micrographs of double labeling immunohistochemistry demonstrates the restriction of SPHK1-immunoreactivity to smooth muscle cells, identified by immunoreactivity for the marker protein α-smooth muscle actin (α-sma). SPHK1a-immunoreactivity was present in peripheral airway smooth muscle of knock-out animals (M2R-KO, a-d) and wild-type (M2R-wt, e-h, FVB, i-l). Granular immunoreactivity could be detected in the cytoplasm and in the membrane of smooth muscle cells. Bars = 50 μm
Videomorphometry
FVB mice In PCLS preparations from FVB mice, the luminal area of peripheral bronchi rapidly decreased within the first 2 minutes after application of 10-6 M muscarine, followed by a sustained decrease in presence of the agonist (Tab. 1, Fig. 3). Application of the SPHK inhibitors DMS and DHS alone for ten minutes prior to coadministration of 10-6 M muscarine had no significant effect on bronchial luminal diameter (Figs. 3A, B, 5A–D). Coadministration of 10-6 M muscarine with DHS (10-4 M-10-10 M) significantly reduced both the initial and the sustained phase of the muscarine induced bronchoconstriction (Tab. 1, Fig. 3B). Similarly, coadministration of 10-6 M muscarine with 10-6-10-10 M DMS reduced both phases of the muscarine response. However 10-4 M DHS had no effect (Tab. 1). Coadministration of 10-6 M muscarine with the structurally related sphingolipid N-acetylsphingosine (10-6-10-10 M) which has no effect on SPHK1 function [5] did not alter the muscarine induced constriction (Tab. 1).
Table 1 Effects of N-AcS, DHS and DMS on muscarine-induced reductions in luminal airway area in FVB mice. The number of experiments (lungs/slices), mean airway diameter in μm, and the luminal airway area determined 1 min (1) and 15 min (2) after stimulation with muscarine (Mus) and 1 min (3) and 15 min (4) after repeated stimulation with muscarine or after repeated stimulation with muscarine in combination with N-AcS, DHS or DMS are shown. In all experiments, the muscarine concentration was 10-6 M. Data are given as means ± S.E.M. Matched pairs (1 vs. 3; 2 vs. 4) were evaluated by Wilcoxon's rank sum test. Differences were considered as statistically significant when p < 0.05 (n.s., not significant).
Lungs /slices Diameter [μm] Area [%] 1 Area [%] 2 Area [%] 3 Area [%] 4 p (1 vs. 3) p (2 vs. 4)
Mus Mus n.s. n.s.
4/4 206 ± 7 31 ± 6 38 ± 4 38 ± 7 28 ± 3
Mus Mus/N-AcS 10-6 M n.s. n.s.
6/5 182 ± 7 47 ± 8 42 ± 8 50 ± 8 42 ± 9
Mus Mus/N-AcS 10-8 M n.s. n.s.
5/5 218 ± 5 36 ± 1 28 ± 5 38 ± 6 26 ± 6
Mus Mus/N-AcS 10-10 M n.s. n.s.
4/4 212 ± 22 21 ± 8 16 ± 5 26 ± 6 30 ± 5
Mus Mus/DHS 10-4 M n.s. n.s.
4/6 194 ± 9 42 ± 5 37 ± 7 61 ± 8 45 ± 6
Mus Mus/DHS 10-6 M p < 0.05 p < 0.05
7/7 197 ± 14 45 ± 11 31 ± 7 70 ± 13 50 ± 5
Mus Mus/DHS 10-8 M p < 0.05 p < 0.05
7/9 181 ± 6 39 ± 7 29 ± 6 50 ± 9 47 ± 7
Mus Mus/DHS 10-10 M p < 0.05 p < 0.05
4/8 198 ± 6 34 ± 3 24 ± 3 42 ± 4 40 ± 4
Mus Mus/DMS 10-4 M p < 0.05 p < 0.05
4/7 180 ± 7 30 ± 7 25 ± 6 45 ± 10 41 ± 7
Mus Mus/DMS 10-6 M p < 0.05 p < 0.05
4/7 201 ± 11 48 ± 7 35 ± 8 70 ± 6 54 ± 9
Mus Mus/DMS 10-8 M p < 0.05 p < 0.05
4/7 206 ± 11 31 ± 5 25 ± 4 54 ± 8 50 ± 5
Mus Mus/DMS 10-10 M p < 0.05 p < 0.05
4/6 196 ± 15 43 ± 6 23 ± 4 71 ± 11 52 ± 7
Figure 3 Muscarine-induced reductions in luminal area of peripheral bronchi from FVB mice. (A) Luminal area of peripheral bronchi after application of 10-6 M muscarine (Mus) as recorded by videomorphometry. Data (means ± S.E.M) were expressed as % luminal area. Bronchi from FVB mice (4 slices from 6 lungs) responded to 10-6 M muscarine until wash out (Wash). Time points 1–4 were chosen as indicators for initial and sustained constriction. 1 = luminal airway area 1 min after muscarine application, 2 = luminal airway area 15 min after muscarine application, 3 = luminal airway area 1 min after repeated muscarine application, 4 = luminal airway area 15 min after repeated muscarine application. Effect of DHS (B) and DMS (C) on muscarine-mediated reductions in luminal area of peripheral bronchi from FVB mice. The luminal area of peripheral bronchi (expressed in %) was recorded by videomorphometry after application of 10-6 M muscarine (Mus) alone and after application of 10-6 M muscarine together with (B) 10-6 M DHS (diamonds, 7 slices from 4 lungs) and 10-10 M DHS (triangles, 8 slices from 4 lungs) or (C) 10-6 M DMS (diamonds, 9 slices from 7 lungs) and 10-10 M DMS (triangles, 6 slices from 4 lungs). Bronchoconstriction responses were significantly reduced in the presence of DHS and DMS. The inhibition was more pronounced following coadministration of DMS. Data are given as means ± S.E.M.
Inhibition of Ca2+ entry by La3+ (10-6 M) considerably reduced the initial and strongly inhibited the sustained constriction of peripheral airways following the administration of 10-6 M muscarine (Tab. 2, Fig. 4A, E, G). Combination of 10-8 M DHS or 10-8 M DMS with La3+ (10-6 M) almost completely abolished the initial and the sustained phase of the muscarine-dependent bronchoconstriction (Tab. 2, Fig. 4C, E, G). The combination of DHS or DMS with La3+ further significantly reduced both phases compared to La3+ alone (Wilcoxon's rank sum test p < 0.05)
Table 2 Effects of SKF96365 and La3+ alone and in combination with DHS or DMS on the muscarine-induced reductions in luminal airway area in FVB mice Effects of the Ca2+-entry inhibitors La3+ (10-6 M) and SKF 96365 (10-5 M) alone and in combination with DHS or DMS (both 10-6 M) on the muscarine induced reduction in luminal airway area (expressed in %). The number of experiments (lungs/slices), mean airway diameter in μm, and the luminal airway area determined 1 min (1) and 15 min (2) after stimulation with muscarine and 1 min (3) and 15 min (4) and after stimulation with muscarine (Mus) in combination with La3+ (Mus/La3+) or SKF96365 (Mus/SKF), or after stimulation with muscarine and La3+ or SKF96365 in combination with DHS or DMS are shown. In all experiments, the muscarine concentration was 10-6 M. Data are given as means ± S.E.M. Matched pairs (1 vs. 3; 2 vs. 4) were evaluated by Wilcoxon's rank sum test. Differences were considered as statistically significant when p < 0.05.
Lungs /slices Diameter [μm] Area [%] 1 Area [%] 2 Area [%] 3 Area [%] 4 p (1 vs. 3) p (2 vs. 4)
4/9 212 ± 3 Mus Mus/SKF p < 0.05 p < 0.05
60 ± 7 49 ± 7 76 ± 7 86 ± 3
4/8 193 ± 12 Mus Mus/SKF/DHS p < 0.05 p < 0.05
51 ± 5 43 ± 6 82 ± 6 86 ± 3
5/6 207 ± 11 Mus Mus/SKF/DMS p < 0.05 p < 0.05
65 ± 3 53 ± 6 88 ± 2 85 ± 8
4/7 180 ± 8 Mus Mus/La3+ p < 0.05 p < 0.05
39 ± 8 49 ± 12 65 ± 7 89 ± 2
4/7 211 ± 12 Mus Mus/La3+/DHS p < 0.05 p < 0.05
40 ± 4 33 ± 3 85 ± 2 89 ± 2
4/5 210 ± 15 Mus Mus/La3+/DMS p < 0.05 p < 0.05
45 ± 9 38 ± 9 94 ± 2 95 ± 1
Figure 4 Effect of various treatments on muscarine-mediated reductions in luminal area of peripheral bronchi from FVB mice. The luminal area of peripheral bronchi (expressed in %) was recorded by videomorphometry after application of 10-6 M muscarine alone or after application of 10-6 M muscarine in the presence of (A) 10-6 M La3+ (7 slices from 4 lungs), (B) SKF 96365 (SKF 10-5 M, 9 slices from 4 lungs), (C) 10-6 M La3+ in combination with 10-6 M DHS (7 slices from 4 lungs), or (D) 10-5 M SKF 96365 in combination with 10-6 M DHS (8 slices from 4 lungs). Data are given as means ± S.E.M.E-F) Summary of effects of various treatments on muscarine-mediated reductions in luminal area of peripheral bronchi from FVB mice. Luminal area (expressed in %) determined 1 min (initial contraction, E) and 15 min (sustained contraction, G) after application of 10-6 M muscarine alone or of muscarine in combination with either La3+ (both 10-6 M) or La3+ in combination with DHS or DMS (both 10-8 M). Luminal area (expressed in %) determined 1 min (initial contraction, F) and 15 min (sustained contraction, H) after application of muscarine alone or of muscarine in combination with either SKF 96365 (10-5 M) or SKF 96365 in combination with 10-8 M DHS or DMS. Asterisks indicate p < 0.05 for the comparison with application of muscarine alone (E-F). Daggers indicate p < 0.05 for the comparison with application of muscarine in combination with La3+ (E, G). Data are given as means ± S.E.M.
Blockade of Ca2+-entry by SKF96365 (10-5 M), which inhibits G-protein activated and voltage gated Ca2+-channels, greatly reduced both the initial and the sustained phase of muscarine-induced bronchoconstriction (Tab. 3, Fig. 4B, F, H). Combination of SKF96365 (10-5 M) with 10-8 M DMS or 10-8 M DHS exerted no further reduction of luminal airway area (Wilcoxon's rank sum test, Tab. 2, Fig. 4F, H).
M2R-KO/wt mice Muscarine (10-6 M) stimulation of PCLS preparations from M2R-KO mice resulted in an initial constriction response that was followed by a slight relaxation (4–8 % relaxation of luminal airway area within 3–15 min, 28 slices/27 lungs, Tab. 3, Fig. 5A, B). In contrast, in PCLS preparations from the corresponding wt mice (M2R-wt), the initial bronchoconstriction was followed by a sustained constriction response. In PCLS of M2R-KO and M2R-wt mice, DHS and DMS (10-8 M each) significantly inhibited the initial and sustained phase of muscarine induced constriction (Tab. 3, Fig. 5A, B). The initial constriction was inhibited by about 54 % (luminal area of M2R-wt: 48 ± 11.5 %; M2R-KO: 61.5 ± 9.5 %, Tab. 3, Fig. 5A, B) and the sustained phase by about 50 % (luminal area of M2R-wt: 34 ± 12 %; M2R-KO: 67.5 ± 15 %, Tab. 3, Fig. 5A, B). The degree of inhibition of the initial phase was comparable between M2R-wt and M2R-KO mice.
Table 3 Effects of DHS and DMS on muscarine-induced reductions in luminal airway area in PCLS from M2 R-KO and theircorresponding wild-type mice. The number of experiments (lungs/slices), mean airway diameter in μm, and the muscarine-induced constriction in luminal airway area (expressed in %) determined after 1 min (1) and 15 min (2) after stimulation with muscarine and 1 min (3) and 15 min (4) after stimulation with muscarine in combination with DHS or DMS (both 10-6 M or 10-8 M) are shown. In all experiments, the muscarine concentration was 10-6 M. The initial phase (1 vs 3) and the sustained phase (2 vs 4) of constriction were reduced in the presence of DHS or DMS. Data are given as means ± S.E.M.
M2R-wt Lungs /slices Diameter [μm] Area [%] 1 Area [%] 2 Area [%] 3 Area [%] 4 p (1 vs. 3) p (2 vs. 4)
6/7 208 ± 9 Mus Mus/DHS 10-6 M p ≤ 0.05 p ≤ 0.05
35 ± 9 32 ± 9 77 ± 11 52 ± 11
7/8 213 ± 8 Mus Mus/DHS 10-8 M n.s. p = 0.078 p ≤ 0.05
46 ± 6 45 ± 12 61 ± 9 60 ± 9
5/5 202 ± 17 Mus Mus/DMS 10-6 M p ≤ 0.05 p ≤ 0.05
40 ± 10 42 ± 9 72 ± 9 66 ± 10
7/7 217 ± 8 Mus Mus/DMS 10-8 M p ≤ 0.05 p ≤ 0.05
54 ± 10 41 ± 9 69 ± 10 60 ± 10
M2R-KO
6/6 210 ± 8 Mus Mus/DHS 10-6 M p ≤ 0.05 p ≤ 0.05
50 ± 11 54 ± 11 79 ± 7 82 ± 7
7/7 194 ± 9 Mus Mus/DHS 10-8 M p ≤ 0.05 p ≤ 0.05
44 ± 10 48 ± 8 74 ± 8 74 ± 7
7/8 218 ± 8 Mus Mus/DMS 10-6 M p ≤ 0.05 p ≤ 0.05
56 ± 8 62 ± 7 83 ± 6 79 ± 6
7/7 189 ± 9 Mus Mus/DMS 10-8 M p ≤ 0.05 p ≤ 0.05
43 ± 13 51 ± 11 71 ± 9 73 ± 9
Figure 5 Muscarine-mediated changes in luminal area of peripheral bronchi from wild-type and MR deficient mice. Changes in luminal area of mouse peripheral airways were recorded by videomorphometry after cumulative application of different concentrations of muscarine. (A) Videomorphometric images of precision cut lung slices before (∅), 1 min after stimulation with muscarine (10-6 M), after 15 min wash out (Wash) and 1 min after stimulation with muscarine in presence of DMS (10-8 M). The bronchi from wild-type (M3R-wt) and M3R-KO mice constricted in response to 10-6 M muscarine. In M3R-wt mice, the constriction was slightly reduced in presence of DMS but abolished in M3R-KO mice. Bar = 200 μm. (B-D) Effect of DHS and DMS on muscarine-mediated reductions in luminal area of murine peripheral bronchi in the absence of the M2R or M3R subtypes. (B-C) Luminal area of peripheral airways from M2R-KO (triangles) and M2R-wt mice (squares)(expressed in %
in response to application of 10-6 M muscarine (Mus) alone or after application of muscarine (10-6 M) in combination with DHS (B) or DMS (C) (both 10-8 M). Both DHS and DMS significantly inhibited the muscarine-induced constriction in airways of both M2R-KO and M2R-wt mice. (D, E) Luminal area of peripheral airways (expressed in %) from M3R-KO (triangles) and M3R-wt (squares) mice in response to application of 10-6 M muscarine alone or after application of muscarine (10-6 M) in combination with DHS (D) or DMS (E) (both 10-8 M). Both DHS and DMS significantly inhibited the muscarine-induced constriction in airways of both M3R-KO and M3R-wt mice. Data are given as means ± S.E.M.
M3R-KO/wt mice As in all other experiments, the area of the airway lumen at the beginning of the experiment was defined as 100 %. Application of 10-6 M muscarine constricted peripheral airways in wt-mice by about 66 % (28 slices/25 lungs), whereas in M3R-KO mice the constriction was reduced by about 57 % (29 slices/28 lungs, Tab. 4 Figs. 5C, D, 6). In wt-bronchi, inhibition of SPHK by DHS or DMS (10-8 M each) significantly reduced the initial phase of constriction by about 50 % (58 ± 12 %, Tab. 4, Figs. 5C, D, 6). Strikingly, treatment of PCLS from M3R-KO mice with DHS (10-8 M) or DMS (10-6 and 10-8 M) almost completely abolished the initial constriction response (94 ± 4 %, Tab. 4, Figs. 5, 6). Inhibition of SPHK by DHS (10-8 M) or DMS (10-6and 10-8 M) significantly reduced the sustained phase of constriction to a comparable degree in M3R-wt and M3R-KO mice (Tab. 4, Figs. 5C, D, 6).
Table 4 Effects of DHS and DMS on muscarine-induced reductions in luminal airway area in PCLS from M3 R-KO and their corresponding wild-type mice. The number of experiments (lungs/slices), mean airway diameter in μm, and the muscarine-induced constriction in luminal airway area (expressed in %) determined after 1 min (1) and 15 min (2) after stimulation with muscarine and 1 min (3) and 15 min (4) after stimulation with muscarine in combination with DHS or DMS (both 10-6 M or 10-8 M) are shown. In all experiments, the muscarine concentration was 10-6 M. The initial phase (1 vs 3) and the sustained phase (2 vs 4) of constriction were reduced in presence of DHS or DMS. Data are given as means ± S.E.M.
M3R-wt Lungs /slices Diameter [μm] Area [%] 1 Area [%] 2 Area [%] 3 Area [%] 4 p (1 vs. 3) p (2 vs. 4)
6/6 221 ± 10 Mus Mus/DHS 10-6 M n.s. p ≤ 0.05
53 ± 7 47 ± 8 75 ± 7 69 ± 11
8/9 204 ± 10 Mus Mus/DHS 10-8 M p ≤ 0.05 p ≤ 0.01
40 ± 6 30 ± 6 64 ± 9 55 ± 6
6/7 203 ± 6 Mus Mus/DMS 10-6 M p ≤ 0.05 p ≤ 0.05
41 ± 10 37 ± 8 64 ± 10 56 ± 8
5/6 197 ± 8 Mus Mus/DMS 10-8 M p ≤ 0.05 p ≤ 0.05
42 ± 5 41 ± 10 71 ± 7 69 ± 9
M3R-KO
4/4 205 ± 18 Mus Mus/DHS 10-6 M n.s. n.s.
69 ± 2 51 ± 4 77 ± 8 62 ± 12
8/9 188 ± 9 Mus Mus/DHS 10-8 M p ≤ 0.05 p ≤ 0.05
75 ± 6 67 ± 6 96 ± 2 88 ± 4
7/7 202 ± 12 Mus Mus/DMS 10-6 M p ≤ 0.05 p ≤ 0.05
75 ± 4 67 ± 6 97 ± 1 82 ± 4
9/9 196 ± 7 Mus Mus/DMS 10-8 M p ≤ 0.01 p ≤ 0.05
65 ± 5 56 ± 6 93 ± 2 81 ± 3
Discussion
MR signaling pathways play a key role in the regulation of airway resistance which is determined largely by the diameter of smaller, intrapulmonary airways [25]. A better understanding of the different pathways underlying MR activation in the intrapulmonary airways is of considerable clinical relevance. In the present study, we examined the hypothesis that S1P might be involved in the muscarinic control of peripheral airways. To address this question, we used the PCLS model which has been shown to maintain the integrity of all components of the peripheral lung including viable peripheral airways [20].
Using double labelling immunohistochemistry we demonstrated for the first time that the SPHK1 protein is highly expressed in the cytosol of murine airway smooth muscle cells with virtually no immunoreactivity in non-smooth muscle cells. On the other hand, human and murine lung tissue showed a high expression and activity of SPHK which was not restricted to smooth muscle alone [13,14]. The virtual restriction of SPHK1-immununoreactivity to smooth muscle could be due to the presence of SPHK isoforms other than SPHK1 in murine lung [14]. At the subcellular level, SPHK1 showed a cytoplasmic localisation but was also found in airway smooth muscle membranes, in agreement with functional and immunoprecipitation studies of mouse lung membranes [14].
In the present study, the use of membrane-permeable SPHK inhibitors, DHS and DMS [18], convincingly demonstrated that MR-mediated constriction of peripheral airways involves the activation of SPHK. In PCLS preparations from wt-mice, both SPHK inhibitors significantly reduced the fast initial constriction response (induced by 10-6 M muscarine) which is mainly dependent on Ca2+-release from intracellular stores [26]. This effect was observed with the lowest concentrations of the inhibitors but was absent when DHS and DMS were given alone for ten minutes prior to the application of DHS and DMS in combination with muscarine. This inhibition of muscarine induced bronchoconstriction was also absent when we used N-acetylsphingosine (N-AcS), an agent that is structurally related to DHS and DMS but that does not affect SPHK function [5]. These data suggest that DHS and DMS did not exert their inhibitory effects through nonspecific actions. Accordingly, the muscarine-induced initial bronchoconstriction was partly inhibited but persisted in presence of the blockers of Ca2+-influx, La3+ or SKF 96365 [27,28], but was almost abolished when La3+ was applied in combination with DHS or DMS. The sustained phase of muscarine-induced bronchoconstriction was also significantly inhibited by DHS and DMS in FVB mice. Interestingly, the responses that were inhibited by DHS/DMS were SKF96365-but not La3+-sensitive, since DHS/DMS significantly increased the La3+-mediated inhibition of the muscarine-mediated constriction, but had no effect on the SKF96365-mediated inhibition of the constriction.
MR-mediated constriction of peripheral airways has been shown to be mediated by a mixture of M2R and M3R subtypes [20]. It is well known that the M3R stimulates IP3-dependent intracellular Ca2+-release ([29,2]. However, like S1P [30], stimulation of M3R can also activate RhoA-dependent signalling pathways leading to increased myofilament sensitivity of smooth muscle [31,32]. On the other hand, activation of M2Rs in airway smooth muscle has been shown to increase the sensitivity of myofilaments to Ca2+ and to inhibit noradrenaline-induced increases in intracellular cAMP [33,34]. In HEK-293 cells, DLS and DMS markedly reduced both M2R-and M3R-mediated increases in [Ca2+]i [5].
In the present study, we therefore also examined which of these two receptor subtypes (M2R or M3R) is involved in the SPHK1-dependent constriction of peripheral airways. To address this issue, we carried out a series of functional experiments using PCLS preparations from M2R-and M3R-deficient mice (M2R-KO, M3R-KO) and their corresponding wild-type controls (M2R-wt, M3R-wt). To rule out potential differences in SPHK1 expression between wt and KO animals we initially performed a set of quantitative RT-PCR studies. These studies showed that inactivation of the M2R and M3R genes had no significant effect on the relative expression levels of SPHK1 compared to the corresponding wt controls. Moreover, confocal laser scanning microscopic studies showed that the localization and distribution of SPHK1 protein were similar in all mouse strains. In agreement with the results of a previous study [20], the peripheral airways of M3R-KO mice showed an about 50 % reduction in the magnitude of the initial muscarine induced constriction response, as compared to the corresponding response obtained with preparations from M3R-wt mice. We previously demonstrated that the bronchoconstrictor response remaining in the M3R-KO mice is exclusively mediated by M2Rs [20].
Changes in [Ca2+]i are the main trigger in the initiation of MR-mediated bronchoconstriction. The initial phase of constriction of intraparenchymal airways is known to be mediated partly via release of Ca2+ from intracellular stores, as shown by the presence of this phase under blockade of ion influx by La3+ or SKF96365. The considerable reduction of the initial constriction by La3+ or SKF96365 further indicates that influx of Ca2+ is also part of this phase of muscarine-mediated constriction. We defined the bronchoconstriction in response to muscarine as 100 % and analyzed the inhibition under various experimental conditions. Blockade of SPHK1 by either DHS or DMS almost completely abolished the initial phase of constriction in bronchi from M3R-KO mice (Fig. 5 Tab. 2). In contrast, DHS or DMS only partially reduced this early response in bronchi from M2R-KO and wt-mice (Fig. 5). This observation strongly suggests that stimulation of M2Rs mediates activation of SPHK1, which eventually triggers the release of Ca2+ from intracellular stores leading to bronchoconstriction. The role of M2R in airway smooth muscle contraction appears multi-functional in that the receptor can modulate the function of smooth muscle by activation of multiple signalling pathways including tyrosine kinase activation and stimulation/inhibition of ion channels [35,32]. Phosphorylation of myosin light chain and activation of PKC play important roles in the maintenance of smooth muscle contractions [36,37]. Since S1P has been shown to stimulate myosin phosphorylation [8,38], it is likely that this response is also mediated by the M2R subtype.
Our findings strongly suggest that SPHK1 activation is part of the signalling response to M2R stimulation in peripheral mouse airways. These airways resemble in their structure and composition of cell types human distal airways [39]. Like human distal airways, murine peripheral airways about 200 μm in diameter that are terminal bronchioles lack submucosal glands, cartilage and contain smooth muscle and Clara cells, in addition to ciliated cells [39]. In studies using monkey and rat peripheral airways, the ACh analogue methacholine (MCh) induced similar responses in large and small mammals [40]. It is therefore likely that the distal airways of man and mice, both of which are also MCh-sensitive [17], share similar MR signalling pathways. In addition, it has been shown that human airway smooth muscle cells constrict in response to S1P [7,8], suggesting that the MR-SPHK1-S1P signalling pathway is also present in human peripheral airways.
Conclusion
In conclusion, in this study we demonstrated the existence of a novel signalling pathway in the regulation of peripheral airways. We found that MR-mediated constriction of murine peripheral airways is mediated, in part, by activation of SPHK. Our data suggest that muscarinic activation of SPHK contributes to the initial and sustained phase of constriction. The SPHK activation in the initial phase is mainly mediated via the M2R subtype. These findings could be of relevance for the development of novel drugs useful for the treatment of chronic obstructive pulmonary disease and asthma. For example, one may speculate that altered S1P signalling may contribute to the hyperreactivity of peripheral airways under pathological conditions.
List of abbreviations
ACh acetylcholine
CT threshold cycle
MR muscarinic acetylcholine receptor
wt wild-type control
M2R-KO muscarinic acetylcholine receptor 2-deficient mice
M3R-KO muscarinic acetylcholine receptor 3-deficient mice
PLCβ phospholipase C beta
PCLS Precision cut lung slices
SKF96365 1-2-(4-methoxyphenyl)-2-[3-(4-methoxyphenyl) propoxy]-ethyl-1H-imidazole
S1P sphingosine 1-phosphate
SPHK sphingosine kinases
IP3 inositol 1, 4, 5-trisphosphate
DHS D, L-threo-dihydrosphingosine
DMS N, N-dimethyl-sphingosine
NAcS N-acetylsphingosine
MCh Methacholine
Authors' contributions
PM, NP, SA, WS, SW and RVH carried out the videomorphometric experiments and performed qRT-PCR and immunohistochemistry. JW developed the KO-mice and participated together with WK and RVH in writing and preparation of the manuscript and in the statistical analysis. YB was involved in the design of the study and the immunohistochemical investigations. The data presented in the manuscript are part of the doctoral thesis of PM, NP, SA, WS.
Acknowledgements
We thank Ms K. Michael for skilful technical assistance.
==== Refs
Wess J Blin N Mutschler E Bluml K Muscarinic acetylcholine receptors: structural basis of ligand binding and G protein coupling Life Sci 1995 56 11–12 915 922 10188793 10.1016/0024-3205(95)00028-5
Caulfield MP Birdsall NJ International Union of Pharmacology. XVII. Classification of muscarinic acetylcholine receptors Pharmacol Rev 1998 50 2 279 290 9647869
Jacoby DB Fryer AD Anticholinergic therapy for airway diseases Life Sci 2001 68 22–23 2565 2572 11392627 10.1016/S0024-3205(01)01053-0
Meyer zu Heringdorf D Lass H Alemany R Laser KT Neumann E Zhang C Schmidt M Rauen U Jakobs KH van Koppen CJ Sphingosine kinase-mediated Ca2+ signalling by G-protein-coupled receptors Embo J 1998 17 10 2830 2837 9582276 10.1093/emboj/17.10.2830
van Koppen CJ Meyer zu Heringdorf D Alemany R Jakobs KH Sphingosine kinase-mediated calcium signaling by muscarinic acetylcholine receptors Life Sci 2001 68 22–23 2535 2540 11392623 10.1016/S0024-3205(01)01049-9
Olivera A Spiegel S Sphingosine kinase: a mediator of vital cellular functions Prostaglandins Other Lipid Mediat 2001 64 1–4 123 134 11324702 10.1016/S0090-6980(01)00108-3
Ammit AJ Hastie AT Edsall LC Hoffman RK Amrani Y Krymskaya VP Kane SA Peters SP Penn RB Spiegel S Sphingosine 1-phosphate modulates human airway smooth muscle cell functions that promote inflammation and airway remodeling in asthma Faseb J 2001 15 7 1212 1214 11344091
Rosenfeldt HM Amrani Y Watterson KR Murthy KS Panettieri RA JrSpiegel S Sphingosine-1-phosphate stimulates contraction of human airway smooth muscle cells Faseb J 2003 17 13 1789 1799 14519658 10.1096/fj.02-0836com
Lacasse Y Brosseau L Milne S Martin S Wong E Guyatt GH Goldstein RS Pulmonary rehabilitation for chronic obstructive pulmonary disease Cochrane Database Syst Rev 2002 3 CD003793 12137716
Escolar JD Escolar MA Guzman J Roques M Morphological hysteresis of the small airways Histol Histopathol 2003 18 1 19 26 12507280
Wohlsen A Uhlig S Martin C Immediate allergic response in small airways Am J Respir Crit Care Med 2001 163 6 1462 1469 11371419
Kohama T Olivera A Edsall L Nagiec MM Dickson R Spiegel S Molecular cloning and functional characterization of murine sphingosine kinase J Biol Chem 1998 273 37 23722 23728 9726979 10.1074/jbc.273.37.23722
Melendez AJ Carlos-Dias E Gosink M Allen JM Takacs L Human sphingosine kinase: molecular cloning, functional characterization and tissue distribution Gene 2000 251 1 19 26 10863092 10.1016/S0378-1119(00)00205-5
Fukuda Y Kihara A Igarashi Y Distribution of sphingosine kinase activity in mouse tissues: contribution of SPHK1 Biochem Biophys Res Commun 2003 309 1 155 160 12943676 10.1016/S0006-291X(03)01551-1
Liu H Sugiura M Nava VE Edsall LC Kono K Poulton S Milstien S Kohama T Spiegel S Molecular cloning and functional characterization of a novel mammalian sphingosine kinase type 2 isoform J Biol Chem 2000 275 26 19513 19520 10751414 10.1074/jbc.M002759200
Young KW Willets JM Parkinson MJ Bartlett P Spiegel S Nahorski SR Challiss RA Ca2+/calmodulin-dependent translocation of sphingosine kinase: role in plasma membrane relocation but not activation Cell Calcium 2003 33 2 119 128 12531188 10.1016/S0143-4160(02)00205-1
Held HD Martin C Uhlig S Characterization of airway and vascular responses in murine lungs Br J Pharmacol 1999 126 5 1191 1199 10205008 10.1038/sj.bjp.0702394
Olivera A Spiegel S Sphingosine-1-phosphate as second messenger in cell proliferation induced by PDGF and FCS mitogens Nature 1993 365 6446 557 560 8413613 10.1038/365557a0
Coulson FR Fryer AD Muscarinic acetylcholine receptors and airway diseases Pharmacol Ther 2003 98 1 59 69 12667888 10.1016/S0163-7258(03)00004-4
Struckmann N Schwering S Wiegand S Gschnell A Yamada M Kummer W Wess J Haberberger RV Role of muscarinic receptor subtypes in the constriction of peripheral airways: studies on receptor-deficient mice Mol Pharmacol 2003 64 6 1444 1451 14645675 10.1124/mol.64.6.1444
Gomeza J Shannon H Kostenis E Felder C Zhang L Brodkin J Grinberg A Sheng H Wess J Pronounced pharmacologic deficits in M2 muscarinic acetylcholine receptor knockout mice Proc Natl Acad Sci U S A 1999 96 4 1692 1697 9990086 10.1073/pnas.96.4.1692
Yamada M Miyakawa T Duttaroy A Yamanaka A Moriguchi T Makita R Ogawa M Chou CJ Xia B Crawley JN Mice lacking the M3 muscarinic acetylcholine receptor are hypophagic and lean Nature 2001 410 6825 207 212 11242080 10.1038/35065604
Murate T Banno Y K TK Watanabe K Mori N Wada A Igarashi Y Takagi A Kojima T Asano H Cell type-specific localization of sphingosine kinase 1a in human tissues J Histochem Cytochem 2001 49 7 845 855 11410609
Martin C Uhlig S Ullrich V Videomicroscopy of methacholine-induced contraction of individual airways in precision-cut lung slices Eur Respir J 1996 9 12 2479 2487 8980957 10.1183/09031936.96.09122479
Martin RJ Therapeutic significance of distal airway inflammation in asthma J Allergy Clin Immunol 2002 109 2 Suppl S447 460 11842317 10.1067/mai.2002.121409
Hakonarson H Grunstein MM Regulation of second messengers associated with airway smooth muscle contraction and relaxation Am J Respir Crit Care Med 1998 158 5 Pt 3 S115 122 9817734
Murray RK Kotlikoff MI Receptor-activated calcium influx in human airway smooth muscle cells J Physiol 1991 435 123 144 1663158
Oonuma H Nakajima T Nagata T Iwasawa K Wang Y Hazama H Morita Y Yamamoto K Nagai R Omata M Endothelin-1 is a potent activator of nonselective cation currents in human bronchial smooth muscle cells Am J Respir Cell Mol Biol 2000 23 2 213 221 10919988
Wess J Molecular biology of muscarinic acetylcholine receptors Crit Rev Neurobiol 1996 10 1 69 99 8853955
Bolz SS Vogel L Sollinger D Derwand R Boer C Pitson SM Spiegel S Pohl U Sphingosine kinase modulates microvascular tone and myogenic responses through activation of RhoA/Rho kinase Circulation 2003 108 3 342 347 12847068 10.1161/01.CIR.0000080324.12530.0D
Murthy KS Zhou H Grider JR Brautigan DL Eto M Makhlouf GM Differential signalling by muscarinic receptors in smooth muscle: m2-mediated inactivation of myosin light chain kinase via Gi3, Cdc42/Rac1 and p21-activated kinase 1 pathway, and m3-mediated MLC20 (20 kDa regulatory light chain of myosin II) phosphorylation via Rho-associated kinase/myosin phosphatase targeting subunit 1 and protein kinase C/CPI-17 pathway Biochem J 2003 374 Pt 1 145 155 12733988 10.1042/BJ20021274
Ehlert FJ Contractile role of M2 and M3 muscarinic receptors in gastrointestinal, airway and urinary bladder smooth muscle Life Sci 2003 74 2-3 355 366 14607264 10.1016/j.lfs.2003.09.023
Chiba Y Sakai H Misawa M Characterization of muscarinic receptors in rat bronchial smooth muscle in vitro Res Commun Mol Pathol Pharmacol 1998 102 2 205 208 10100511
Janssen LJ Wattie J Lu-Chao H Tazzeo T Muscarinic excitation-contraction coupling mechanisms in tracheal and bronchial smooth muscles J Appl Physiol 2001 91 3 1142 1151 11509509
Togashi H Emala CW Hall IP Hirshman CA Carbachol-induced actin reorganization involves Gi activation of Rho in human airway smooth muscle cells Am J Physiol 1998 274 5 Pt 1 L803 809 9612296
Ochsner M Ca2+ transient, cell volume, and microviscosity of the plasma membrane in smooth muscle Biochem Pharmacol 1997 53 12 1765 1777 9256151 10.1016/S0006-2952(97)00033-6
Kai T Yoshimura H Jones KA Warner DO Relationship between force and regulatory myosin light chain phosphorylation in airway smooth muscle Am J Physiol Lung Cell Mol Physiol 2000 279 1 L52 58 10893202
Zhou H Murthy KS Distinctive G protein-dependent signaling in smooth muscle by sphingosine 1-phosphate receptors S1P1 and S1P2 Am J Physiol Cell Physiol 2004 286 5 C1130 1138 15075212 10.1152/ajpcell.00429.2003
Parent RA Comparative biology of the normal lung 1991 1 CRC Press
Kott KS Pinkerton KE Bric JM Plopper CG Avadhanam KP Joad JP Methacholine responsiveness of proximal and distal airways of monkeys and rats using videomicrometry J Appl Physiol 2002 92 3 989 996 11842031
| 15927078 | PMC1156956 | CC BY | 2021-01-04 16:23:25 | no | Respir Res. 2005 May 31; 6(1):48 | utf-8 | Respir Res | 2,005 | 10.1186/1465-9921-6-48 | oa_comm |
==== Front
Theor Biol Med ModelTheoretical Biology & Medical Modelling1742-4682BioMed Central London 1742-4682-2-211594904610.1186/1742-4682-2-21EditorialA new journal – "Theoretical Biology and Medical Modelling" Wheatley Denys N [email protected] BioMedES, Leggat House, Keithhall, Inverurie, Aberdeen AB51 0LX, UK2005 12 6 2005 2 21 21 25 5 2005 12 6 2005 Copyright © 2005 Wheatley; licensee BioMed Central Ltd.2005Wheatley; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Biology has a conceptual basis that allows one to build models and theorize across many life sciences, including medicine and medically-related disciplines. A dearth of good venues for publication has been perceived during a period when bioinformatics, systems analysis and biomathematics are burgeoning.
Steps have been taken to provide the sort of journal with a quick turnaround time for manuscripts which is online and freely accessible to all readers, whatever their persuasion or discipline. We have now been running for some time a journal which has had many good papers presented pre-launch, and a steady stream of papers thereafter. The value of this journal as a new venue has already been vindicated.
Within a short space of time, we have founded a state-of-the-art electronic journal freely accessible to all in a much sort-after interdisciplinary field that will be of benefit to the thinking life scientist, which must include medically qualified doctors as well as scientists who prefer to build their new hypotheses on basic principles and sound concepts underpinning biology. At the same time, these principles are not sacrosanct and require critical analysis. The journal promises to deliver many exciting ideas in the future.
==== Body
Introduction
Several stories have been told about theorists; here are two.
Two proponents of ideas from opposing schools of thought took advice from a third party. To A he said she ought to listen to the arguments propounded by B and not just dismiss them out of hand; to B he said he should listen to the arguments of A and not simply dismiss them. To them both, he said "I have been fair with you both, since in all probability you are both wrong!"
The other story shows up the facileness of some arguments (or their proponents): One day, Socrates was discussing Sod's law, also well known in the world of the Ancient Greeks, who like us all buttered their bread on one side. The law – also know as Murphy's law – states that if dropped, the piece invariably lands butter-side down. Next day one of his disciples came rushing to him, remarking that at home in the evening he had buttered a piece of bread, dropped it, and to his amazement it landed butter-side up! Socrates thought for a moment before severely rebuking him "you fool; you buttered the wrong side!"
How this new journal came about is an interesting story. Continued dissatisfaction with other theoretical journals, especially their slowness, and problems at Comments in Theoretical Biology, were largely the precipitating factors. I also had colleagues at the Torino Politechnico supporting me in the need for a vehicle to publish cancer modelling problems as part of developing a European network of excellence. This was reinforced by an editorial in Nature [1] starkly pointing out the lack of a reasonable focus for informatics on cancer and the need for a modeling forum, at least in Europe, to help oncologists, but equally many other scientists and doctors. Dr David Eagleman [2] made a similar highly relevant point about theoretical science. I suggested that the network of excellence we were striving for in our European Consortium ought to have a web presence, so why not have an online journal? That is why I started one. Having undertaken primarily the role of founding editor, my job was to get the journal up and running, and provide the editorial facilities. We have the assured services of Dr Agutter and Ms Angela Panther as managing editor and editorial officer, respectively, overseeing the smooth running of the journal on a daily basis.
Paul and I have published many papers in the field of theoretical biology and medicine, which are disseminated over a wide range of journals that do not allow interested parties to keep many of them easily in focus. Our experience in producing scientific papers as well as writing them comes from a knowledge of e-publications, having one of the first specialist (for which now read "independent") journals to be published by BioMed Central (BMC)3 online, viz. Cancer Cell International
The nature of the journal
Speculation within limits, theorizing, and philosophizing are part and parcel of biology and medicine, which are in turn part of life. Having established certain facts or observations, we can employ deductive or inductive reasoning depending on the number of such pieces of information – as Bacon suggested – to reach certain tenuous conclusions. From this position, we create hypotheses, the very best being those that can be tested, not the unapproachable ones that will remain for ever. Biology has a broad coverage, while medicine in contrast has a much narrower focus, because its concepts must be the very same ones as biology in general is built. These in turn have many of their groundings in chemical and particularly physical laws, but not all. The law of mass action cannot be applied under the conditions of synthesis of a new molecule of DNA. In addition to theorizing in medicine, we can also build models that have clinical relevance to help us understand such problems as the spread of diseases throughout the body.
The opening up of bioinformatics, the proliferation of databases on genes, proteins, metabolites and so on, have created a climate in which many scientists can puddle to their heart's content to make significant and meaningful correlations. New ideas spring from some of these, just as old ideas can find evidence suddenly becoming available that was needed years ago to support them. Interdisciplinary work is now highly fashionable; and mathematical biology needs to become a force within biology, not a peripheral, elitist activity. And so on; I rest my case.
The new journal on the web will consider high quality, peer-reviewed theoretical papers. It will also seek to provide new ideas that may be quite off the main-stream of biomedicine; after all, today's "crazy" notion has a not so infrequently had the habit of becoming tomorrow's received wisdom. My two colleagues, who are also Editors-in-Chief are Pier Paolo Delsanto in Turin and Hans-Peter Meinzer in Heidelberg. The editorial board includes many people well known to those who already publish in theoretical journals.
The advantages of using the web, if not already self-evident, will become clear to you all once you appreciate what a wonderfully interactive tool it affords. Debates on papers can be posted along with referees' criticisms, if they so choose. Commentaries can be put up online, much as already done in BioMed Central's Journal of Biology. But we also want theoretical work that is topical. Often editors see no need to move theoretical papers quickly through the publication process. Indeed, the only ones that get held up even longer are papers dealing with the history of biology, many of which often take years to complete the acceptance and publication processes, my last one taking very close to two years.
Many issues arise on which we are forced through incomplete knowledge to hypothesize about causes, such as the spread of SARS. When SARS arose, ideas needed to be discussed as critically as possible, and an online journal has almost a similar capacity to handle "copy" in the way that newspaper handles it, although there is a world of difference between the editorial role of "Tbiomed" (our shortened title and URL name) and one in a tabloid newspaper that is there to feed on drama, rumour and innuendo to sell their products rather than on sound facts and rational arguments. No one is selling anything in science, especially that which gets published and is accessible to everyone free of charge online. So a theoretical journal that is state- of-the-art emanating from BioMed Central offers exactly what the doctor ordered, PhD or MD. I hope you will make maximum use of it, and send papers that will stir the thoughts of others.
We accept that there is the old thorny problem of impact factors. My feeling is that no one scores heavily regarding grant proposals from publishing a paper with your latest and greatest hypothesis. The grant-awarding agencies provide funds for testing hypotheses, not formulating them. Therefore in most cases publishing in Tbiomed is unlikely to affect your street-credibility when it comes to grant proposals; indeed it can help because the hypothesis you want to test is out there online for all to read and explore. The other issue is that impact factor continues to relate to journal performance and not directly to the quality of the individual article. Because all new papers will soon be completely accessible via the web, the paper itself can be rated rather than the journal that holds it. An example of this is Faculty of 1000, published by BioMed Central, where experts can pick up and extol the virtues of an individual paper, no matter in which journal it was published.
A call for papers
Our policy will be liberal, but we will not publish substandard arguments, diatribes and purely rhetorical pieces. Peer review will be strict, but we also will give space to people with unconventional theories that have been presented with sound logical reasoning. So here is a call for papers:
So often we hear people say "...well, in theory, you may be right!"
– so why not submit your hypotheses to one of BMC's newest journals and find out what others think of them.
Advantages of publishing in Tbiomed
• Interactiveness
• Accessible to all
• No charges to authors from all BioMed Central subscribing member institutions
• Presentation of complex equations and figures simple and accurate
• Peer review criticisms can be published alongside articles
• Running commentaries can be made on published papers
• No restriction on length, although brevity of expression will be sought
• Colour can be used as extensively as required
• Is a journal that forms a focus and interface for bioinformatics and multidisciplinary articles
• Topical matters requiring theoretical consideration will be published quickly
• Juxtaposition of biology and medicine is intentional to spawn more and better interactiveness
• Articles can be accessed and disseminated freely, meaning no reprint costs
==== Refs
Gatenby RA Maini PK Cancer summed up Nature 2003 421 321only 12540881 10.1038/421321a
Eagleman DM Holcombe AO Improving science through online commentary Nature 2003 423 15 12721598 10.1038/423015a
| 15949046 | PMC1156957 | CC BY | 2021-01-04 16:39:26 | no | Theor Biol Med Model. 2005 Jun 12; 2:21 | utf-8 | Theor Biol Med Model | 2,005 | 10.1186/1742-4682-2-21 | oa_comm |
==== Front
J Transl MedJournal of Translational Medicine1479-5876BioMed Central London 1479-5876-3-221594387910.1186/1479-5876-3-22ResearchThe anti-tumor effect of Apo2L/TRAIL on patient pancreatic adenocarcinomas grown as xenografts in SCID mice Hylander Bonnie L [email protected] Rose [email protected] Remedios B [email protected] John F [email protected] Dilek [email protected] Jinrong [email protected] Elizabeth A [email protected] Department of Immunology, Roswell Park Cancer Institute, Buffalo, NY, 14263, USA2 Department of Pathology, Roswell Park Cancer Institute, Buffalo, NY, 14263, USA3 Department of Surgery, Roswell Park Cancer Institute, Buffalo, NY, 14263, USA4 Department of Experimental Therapeutics, Roswell Park Cancer Institute, Buffalo, NY, 14263, USA2005 19 5 2005 3 22 22 10 3 2005 19 5 2005 Copyright © 2005 Hylander et al; licensee BioMed Central Ltd.2005Hylander et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Apo2L/TRAIL has considerable promise for cancer therapy based on the fact that this member of the tumor necrosis factor family induces apoptosis in the majority of malignant cells, while normal cells are more resistant. Furthermore, in many cells, when Apo2L/TRAIL is combined with chemotherapy, the effect is synergistic. The majority of this work has been carried out using cell lines. Therefore, investigation of how patient tumors respond to Apo2L/TRAIL can validate and/or complement information obtained from cell lines and prove valuable in the design of future clinical trials.
Methods
We have investigated the Apo2L/TRAIL sensitivity of patient derived pancreatic tumors using a patient tumor xenograft/ SCID mouse model. Mice bearing engrafted tumors were treated with Apo2L/TRAIL, gemcitabine or a combination of both therapies.
Results
Patient tumors grown as xenografts exhibited a spectrum of sensitivity to Apo2L/TRAIL. Both Apo2L/TRAIL sensitive and resistant pancreatic tumors were found, as well as tumors that showed heterogeneity of response. Changes in apoptotic signaling molecules in a sensitive tumor were analyzed by Western blot following Apo2L/TRAIL treatment; loss of procaspase 8, Bid and procaspase 3 was observed and correlated with inhibition of tumor growth. However, in a tumor that was highly resistant to killing by Apo2L/TRAIL, although there was a partial loss of procaspase 8 and Bid in response to Apo2L/TRAIL treatment, loss of procaspase 3 was negligible. This resistant tumor also expressed a high level of the anti-apoptotic molecule Bcl-XL that, in comparison, was not detected in a sensitive tumor. Importantly, in the majority of these tumors, addition of gemcitabine to Apo2L/TRAIL resulted in a greater anti-tumor effect than either therapy used alone.
Conclusion
These data suggest that in a clinical setting we will see heterogeneity in the response of patients' tumors to Apo2L/TRAIL, including tumors that are highly sensitive as well as those that are resistant. While much more work is needed to understand the molecular basis for this heterogeneity, it is very encouraging, that Apo2L/TRAIL in combination with gemcitabine increased therapeutic efficacy in almost every case and therefore may be a highly effective strategy for controlling human pancreatic cancer validating and expanding upon what has been reported for cell lines.
==== Body
Introduction
The high mortality rate seen in patients with pancreatic cancer reflects both the difficulty in early detection and the lack of effective treatment to augment surgery [1,2] so that, following diagnosis, the average survival time of the majority of patients is between 4–5 months [3]. Within the last few years, the use of the deoxycytodine analog gemcitabine has been shown to result in improved clinical benefit, slightly longer mean survival time and has become the first line chemotherapy for pancreatic adenocarcinoma [4,5]. However, since the five-year survival rate has remained at 4%, many new approaches to the treatment of pancreatic adenocarcinoma are being investigated [5,6]. Several of these approaches focus on combination therapies in which gemicitabine is combined with a second cytotoxic agent (e.g. auristatin-PE, [7]), or a targeted biological therapy (e.g.; the anti-EGFR antibody C225, [8,9]; OSI-774, Tarceva, [10]).
In 1995, a new member of the tumor necrosis factor (TNF) family was independently identified by two different groups and named TRAIL (Tumor Necrosis Factor Related Apoptosis Inducing Ligand, [11]) and Apo2L (based on its homology to Fas/Apo1L [12]). This molecule induces apoptosis in a large number of human tumor cell lines, both in vitro and in vivo, while normal cells are not susceptible [11-15]. This is in contrast to other members of this family of ligands (i.e. TNF and FasL), which have marked toxicity when administered systemically (for further discussion see recent reviews by [16-18]). An important natural role for Apo2L/TRAIL in the immunosurveillance of tumors has been proposed based on its expression on several immune cells, including activated NK and T cells (see [19]for discussion). This natural role of Apo2L/TRAIL in anti-tumor activity provides further rationale for attempting to develop Apo2L/TRAIL as a therapeutic molecule. The original studies with Apo2L showed that it could act synergistically with the chemotherapeutic agents 5-FU and CPT-11 in animal studies using a colon tumor cell line [14,20]. There have since been numerous studies expanding these observations using a large number of cell lines of different tumor types with a variety of chemotherapies, both in vitro and in vivo. Among the types of solid tumors that have been studied are breast [21], lung [22], prostate [23], mesothelioma, [24], renal [25], ovarian [26], bladder [27], glioma [28] and pancreas [29]. However, there is a concern that these results might not be predictive of the response of actual patients' tumors. Therefore, an investigation of how patient tumors respond to Apo2L/TRAIL could validate and/or complement information obtained from cell lines and prove valuable in the design of future clinical trials.
Our group has previously investigated the efficacy of Apo2L/TRAIL and CPT-11 combination therapy on patient-derived colon tumors [30] using a SCID mouse xenograft model that our lab has developed [31-34]. The value of this model is that it enables evaluation of actual patient tumors that retain the heterogeneity and histological architecture of the original tumor. TRAIL exerted a significant anti-tumor effect on three different patient colon tumors grown as xenografts and this effect was significantly augmented by the addition of CPT-11 or 5-FU [30]. However, use of this model has also revealed the existence of patient-derived colon tumors which are resistant to Apo2L/TRAIL alone but are sensitive to the combination of Apo2L/TRAIL and CPT-11 (Kenji, manuscript in preparation) This suggested the possibility that a differential response to Apo2L/TRAIL may occur between patients and that additional research is critical for 1) appreciating the degree to which variability occurs between tumors, 2) developing strategies for using Apo2L/TRAIL in combination therapies and 3) determining methods for predicting ahead of time which patients will benefit from Apo2L/TRAIL.
It has previously been reported that pancreatic cell lines exhibit varying degrees of sensitivity to Apo2L/TRAIL and that some of these lines are extremely resistant [35,29,37]. It has also been reported that resistant cells can be sensitized to Apo2L/TRAIL (e.g. [29,38]). However, although the combination of Apo2L/TRAIL and gemcitabine in vitro has been investigated using pancreatic cell lines, there have been conflicting reports on whether this combination does [39] or does not [40] have a synergistic cytotoxic effect.
In this paper, we describe our experience in evaluating five different patient pancreatic tumors grown in SCID mice to Apo2L/TRAIL. The recombinant form of human Apo2L/TRAIL used shows low activity against the murine TRAIL receptor and therefore this model may not reveal any potential toxicities of Apo2L/TRAIL, however it does provide a relevant model for evaluating the sensitivity of patient tumors. Our data support the idea that some patients' tumors may exhibit significant sensitivity while others may be resistant. Still other patients' tumors may be heterogeneous and exhibit regions of both sensitivity and resistance. However, the combination of Apo2L/TRAIL and gemcitabine can enhance the anti-tumor effect against Apo2L/TRAIL sensitive tumors and, importantly, can overcome resistance to either single agent, and result in suppression of resistant tumors. Thus, these findings predict that patients' tumors will exhibit both sensitivity and resistance to Apo2L/TRAIL treatment and that it may be possible to develop approaches for overcoming this resistance by combining Apo2L/TRAIL and chemotherapy.
Materials and methods
Patient pancreatic tumor-SCID mouse model
Our use of the SCID mouse-patient tumor xenograft model has been previously described ([31,33,41,32,34,30]). For these studies, surgical specimens of patients' pancreatic tumors were received shortly after resection through the Tissue Procurement Facility (TP) of RPCI and cut into 2 mm × 2 mm pieces in tissue culture medium (RPMI 1640) under sterile conditions. SCID mice were then anesthetized by intraperitoneal injection of 0.4–0.5 ml Avertin (2.5 g 2,2,2-tribromoethanol dissolved in 5 ml 2-methyl-butanol/200 ml ddH2O) and individual tumor pieces were implanted subcutaneously in the abdominal wall of three mice (1st passage) and monitored for growth. The mice used in all experiments were 7–8 weeks old CB17 SCID mice with an average weight of 18–20 g. They were kept in sterile cages (4–5 mice per cage) and fed with autoclaved chow and water. Mice were maintained in air-conditioned and light controlled rooms (12 hr cycles). All procedures, injections and tumor measurements were carried out under a laminar flow hood using aseptic precautions. Tumor specimens that grew to a size of 1 cm (8–12 weeks) were retrieved and subsequently passaged into recipient mice (2nd passage) and were considered to have successfully engrafted when these tumors grew. Pathological diagnosis of patient specimens and evaluation of engrafted/ passaged tumors was performed in collaboration with a member of the Pathology Department at RPCI.
Experimental design
Five different pancreatic adenocarcinomas that successfully engrafted into SCID mice were selected for passage into groups of experimental mice. Tumors reached 4–5 mm in diameter in approximately 4–6 weeks and the mice were divided into experimental groups of similar tumor sizes. These tumors are referred to as Tumor #1 (TP#10791), Tumor #2 (TP#10978), Tumor #3 (TP#11424), Tumor #4 (TP#12424), and Tumor #5 (TP#11727).
Apo2L/TRAIL
Apo2L/TRAIL used in this investigation was prepared by Genentech, Inc. as described previously [14] and provided as a gift by Genentech and Amgen. A cycle of treatment with Apo2L/TRAIL consisted of daily intraperitoneal injection of 500 μg Apo2L/TRAIL /200 ul saline for 14 days. Mice received 2 such cycles separated by a 7–10 day rest period. Control mice received sterile saline. Tumor volume was calculated with the formula V= LD × (SD)2 / 2, where V is the tumor volume, LD is the longest tumor diameter and SD is the shortest tumor diameter. Data was graphed and the Students unpaired t-test was calculated using SigmaPlot. At various time points during and at the termination of an experiment, mice were sacrificed and pieces of tumor were fixed in formalin, snap-frozen in cryovials in liquid nitrogen, or both, for subsequent analysis. Sections of all tumor samples were processed for light microscopy by standard methods.
Chemotherapy
Gemcitabine (Gemzar, Eli Lilly obtained from McKesson, Buffalo, NY) was administered by intraperitoneal injection daily 5 days/wk in two two-week cycles with a rest interval of one week at doses between 1.0 – 2.5 mg/kg as indicated. Therefore mice received 7.5–13.5 mg/kg weekly, which is less than that routinely administered clinically to patients (25 mg/kg/week).
TUNEL assay
Apoptosis was evaluated by terminal deoxynucleotidyl transferase-mediated dUTP-nick end-labeling (TUNEL) staining (ApopTag, Intergen, Corp) according to the manufacturers instructions.
Western blotting
Cell and tissue lysates (lysis buffer containing 20 mM Tris pH 7.5, 120 mM NaCl, 100 mM NaF, 0.5% Nonionic P40, 200 μM Na3VO4, 50 mM β-Glycerophosphate, 10 mM NaPPi, 4 mM PMSF, 10 μg/ml Leupeptine, 2 mM Benzamidine, 10 μg/ml Aprotinin) were separated by SDS-PAGE and transferred to nitrocellulose membrane. Blots were immunostained by standard techniques: non-specific binding was blocked and membranes were incubated overnight at 4°C with primary antibody. Antibodies used were anti-caspase 8 (Oncogene # AM46), anti-Bcl-XL, (Cell Signaling #2762), anti-Bid (Cell Signaling #2002), anti-human mitochondria (Chemicon #Mab 1273; recognizes a 65 kd epitope on the membrane of intact human mitochondria) and anti-Caspase 3 (Imgenex IMG-144). This was followed by washing, incubation with peroxidase-conjugated secondary antibody and visualization of the bands by enhanced chemoluminescence (ECL) and exposure of the blots to Kodak BioMax film. Anti β-actin was used as a loading control.
Immunohistochemistry
Immunohistochemical evaluation of p53 (mouse monoclonal antibody DO-7, Novocasta) was performed on sections of formalin fixed, paraffin embedded tumors. Antigen retrieval was accomplished with DAKO Target Retrieval Solution using a Black and Decker steamer for 20 minutes followed by a 20 minute cooling period.
Results and discussion
Patient pancreatic adenocarcinomas engrafted into SCID mice maintain the histological features of the original tumor
In order to evaluate the sensitivity of patient pancreatic tumors to novel therapeutic agents, we have developed a patient tumor/SCID mouse xenograft model in which specimens of pancreatic adenocarcinomas obtained directly from surgeries performed at Roswell Park Cancer Institute are established as xenografts in SCID mice. Successful engraftment has been obtained in 33% of pancreatic tumors (18 of 53 tumors implanted over a six year period), including both adenocarcinomas and neuroendocrine tumors. The tumors used in this study were selected randomly from available patient tumor xenografts. Histological evaluation of these tumors demonstrated that engrafted tumors maintained a remarkable degree of similarity, in terms of the histological features seen, to the original tumors. The malignant cells of the xenografts formed glands resembling those in the patients' tumors and secretory material was often seen in the lumen of these glands (Figure 1). As in the patient tumors, these glands are separated and surrounded by stromal tissue.
Figure 1 The histological features of patient pancreatic adenocarcinomas (left hand panels: A, C, E, G) are maintained when these tumors are grown as xenografts (right hand panels: B, D, F, H) in SCID mice. The well differentiated glands (arrows) containing secretory material (double arrows) which are seen in surgical specimens, proliferate in xenografts of the same tumors. (Tumor #1: A, B; Tumor #2: C, D; Tumor #3: E, F; Tumor #4: G, H). Original magnification, ×10.
The growth of patient-derived pancreatic adenocarcinomas can be inhibited by Apo2L/TRAIL
In our initial experiment, we treated mice engrafted with Tumor #1 with two cycles of Apo2L/TRAIL following a 5 week schedule (500 μg/mouse- daily for 2 weeks, 1 week no treatment, daily for 2 weeks). At the end of the five weeks, the sizes of the tumors in the treated group were significantly smaller that those in the control group (Figure 2A). Histological analysis of the median tumor in each group showed that there were areas of substantial necrosis in the interior of these tumors. However, the periphery of an untreated tumor consisted of numerous glands of fairly uniform size containing copious amounts of secretory material (Figure 2B). In contrast, in the smaller, treated tumor (Figure 2C), the majority of the remaining glands were smaller, further apart and contained less secretory material. Also, in comparison to the untreated tumor, the treated tumor consisted of proportionately more connective tissue. Interestingly, in the periphery of the treated tumor, there were pools of secretory material that were not contained in glands, suggesting that the epithelial cells that formed these glands had been lost (Figure 2C). However, even though there was a marked anti-tumor effect of Apo2L/TRAIL on this tumor, the tumor was not totally eradicated and areas of viable tumor cells remained. Similar results were obtained with a second patient tumor (Tumor #2; Figure 3). The growth of this tumor was also significantly inhibited during the 35 day schedule of treatment with Apo2L/TRAIL. However, we observed that tumor growth resumed after cessation of treatment and these tumors grew at a rate similar to that of the untreated controls. Therefore, although Apo2L/TRAIL can significantly suppress tumor growth, it may not prove to be totally efficacious as a single agent and it is important to develop strategies to use it in combination with other therapeutic agents.
Figure 2 Apo2L/TRAIL significantly inhibited the growth of pancreatic adenocarcinoma #1. Mice bearing patient tumor xenografts were treated with Apo2L/TRAIL and the tumor volumes on the final day of treatment (day 35) were compared (7 mice/group, the mean is indicated by the horizontal line; p = 0.004). B, C: Comparison of the histology of the median tumors. B. In the periphery of an untreated tumor, there are numerous large glands (arrow) containing secretory product. C. The treated tumor has fewer glands (arrow) and a higher proportion of connective tissue. There are also large pools of residual secretory product (S) that are not surrounded by epithelial cells.
Figure 3 Growth of Tumor #2 is also significantly inhibited during treatment with Apo2L/TRAIL. A. Control; B. Apo2L/TRAIL treated. (4 mice/group; day 35, p = 0.016). Following cessation of treatment (arrows), growth of treated tumors progressed at a rate similar to that of the untreated tumors.
Combination therapy with Apo2L/TRAIL and gemcitabine resulted in enhanced inhibition of tumor growth over that of either single agent alone
We next carried out a series of experiments to determine whether the efficacy of Apo2L/TRAIL against patient derived pancreatic adenocarcinomas could be enhanced by using it in combination with gemicitabine. The results of an experiment using Tumor #2, demonstrated a dose dependent response of this tumor to gemcitabine and based on the fact that 1.5 mg/kg moderately, but incompletely, suppressed tumor growth (Figure 4), we chose this dose for future experiments in order to best observe any benefit of combination therapy. Subsequently, tumor bearing mice were treated with 500 μg of Apo2L/TRAIL alone (for a 35 day schedule as above), 1.5 mg/kg of gemcitabine alone (5x/week) or the two agents in combination. The responses of four different patient tumors to combination therapy were investigated and the results indicate interesting diversity in the responses of these tumors.
Figure 4 Evaluation of the response of Tumor #2 to increasing doses of gemcitabine. A. Control; B. Tumors in mice treated with 1.0 mg/kg gemcitabine grow at a rate comparable to that of untreated tumors; C. Tumors in mice treated with 1.5 mg/kg gemicitabine show growth inhibition to varying degrees (p = .045); D. Tumors in mice treated with 2.5 mg/kg gemcitabine alone show uniform suppression (p < .001).
One tumor, Tumor #3 was sensitive to both Apo2L/TRAIL and gemcitabine alone and underwent significant growth inhibition in response to either of these agents. Although growth suppression by combination treatment was not statistically different from the effects of either single treatment alone, the anti-tumor effect appears to be enhanced and these tumors underwent complete regression (Figure 5, A–D). This tumor was consistently sensitive to Apo2L/TRAIL in several passages (data not shown).
Figure 5 The effect of Apo2L/TRAIL in combination with gemcitabine on an Apo2L/TRAIL sensitive and resistant pancreatic adenocarcinoma. A-D. Tumor #3; E-H. Tumor #4. Groups of tumor bearing mice were treated with: A&E- Controls, B&F- Apo2L/TRAIL alone, C&G- gemcitabine alone, D&H- Apo2L/TRAIL and gemcitabine in combination. Tumor #3 is sensitive to Apo2L/TRAIL alone (B; p = 0.006) while Tumor #4 is resistant to Apo2L/TRAIL (F). This differential sensitivity is also seen in the response to gemcitabine alone (C, p = 0.02, and G). The efficacy of Apo2L/TRAIL in combination with gemcitabine appears to be enhanced in Tumor #3 (D) and is significantly greater than either treatment alone in Tumor #4 (H; p = 0.008).
In contrast to Tumor #3, Tumor #4 exhibited resistance to both Apo2L/TRAIL and gemcitabine when administered as single agents. However, the use of these two agents in combination was able to overcome this resistance and significant tumor growth inhibition was achieved (Fig 5, E–H ).
Tumor #2 showed variability in sensitivity to Apo2L/TRAIL. Although, when originally evaluated, Tumor #2 had demonstrated sensitivity to Apo2L/TRAIL alone, in a subsequent experiment, growth was not significantly suppressed by either Apo2L/TRAIL or gemcitabine alone (Figure 6, A–D). In this case, however, tumor growth was significantly inhibited by treatment with the agents in combination. During this experiment, one representative tumor was removed from each group on day 7 and prepared for histology (Figure 7, A–D). The normal histological appearance of this adenocarcinoma is shown in a control; this tumor consisted of elaborate glands containing copious secretory material and a moderate amount of stroma (Figure 7A). In comparison, both the Apo2L/TRAIL (Figure 7B) and gemcitabine (Figure 7C) treated tumors demonstrated evident histological changes with fewer, smaller glands. In marked contrast to the tumors treated with single agents alone, the tumor treated with the combination therapy consisted almost entirely of fibrotic connective tissue and contained only isolated foci of viable tumor cells (Figure 7D). Staining of these tumors with the TUNEL assay showed very few apoptotic cells in the untreated or gemicitabine alone treated tumors (Figure 7, E and 7 G). However, large numbers of apoptotic cells are present in both the tumors treated with Apo2L/TRAIL alone and Apo2L/TRAIL in combination with gemcitabine (Figure 7, F and 7 H). This experiment with Tumor #2 was repeated and although there was heterogeneity in the growth of tumors within each treatment group, the results were similar in that significant tumor growth inhibition was achieved by combination therapy (data not shown). The evaluation of Tumor #5 (Figure 6, E–H) yielded similar results in that growth of the tumors in the untreated group was variable and tumor growth was not significantly suppressed by either Apo2L/TRAIL or gemcitabine. However, combination therapy was able to significantly suppress growth of Tumor #5 (Figure 6H).
Figure 6 The anti-tumor effect of Apo2L/TRAIL and gemcitabine in combination is greater than that of either single agent alone. Mice were engrafted with two different patient tumors: A-D, Tumor #2; E-H, Tumor #5. Treatment: A, E. Control; B, F. Apo2L/TRAIL; C, G. gemcitabine; D, H. combination therapy. Although tumors showed variable degrees of sensitivity to either Apo2L/TRAIL or gemcitabine alone, the combination treatment resulted in significant inhibition of tumor growth (D, p = 0.035; H, p = 0.05).
Figure 7 The effect of Apo2L/TRAIL and gemcitabine alone and in combination (Tumor #2, day 5 of treatment). A. Many large glands are apparent in an untreated tumor. B, C. Tumors treated with Apo2L/TRAIL (B) or gemicitabine (C) contain fewer, smaller glands. D. Tumors treated with a combination of Apo2L/TRAIL and gemcitabine consist mainly of fibrotic tissue with foci of residual tumor cells (D, arrow). E-H. TUNEL assay. Few apoptotic cells are apparent in either the untreated (E) or gemcitabine treated (G) treated tumors. The Apo2L/TRAIL (F) and combination treated (H) tumors contain numerous cells undergoing apoptosis.
At this time, the basis for the cooperation between Apo2L/TRAIL and gemcitabine is not known. It is likely that knowledge about the expression of molecules such as p53 in these tumors will be important in understanding factors which affect sensitivity to chemotherapy. Although we have not sequenced p53 for mutations, we have evaluated p53 expression by immunohistochemistry in Tumors #2, 3, 4 and 5. Whereas p53 overexpression was not detected in Tumors #2, 3 and 4, heterogeneous overexpression was detected in Tumor #5 (data not shown). This suggests that the responses of these tumors to combination therapy may be independent of p53 status. It has been demonstrated that the response of tumors to gemcitabine can occur in a p53 independent manner in [42]. Future studies investigating the mechanism(s) of cooperation between Apo2L/TRAIL and gemicitabine will include analyses of the status of critical molecules such as p53.
Thus, these tumors demonstrated a heterogeneous range of responses to Apo2L/TRAIL. Tumor #3 is significantly sensitive to killing with Apo2L/TRAIL alone and this was consistently seen in subsequent experiments. Interestingly, Tumors #4 and 5 were resistant to Apo2L/TRAIL alone in the passage that was evaluated. It is informative that with Tumor #2, which was evaluated in several experiments, the response varied in different passages. Although the basis for this variability is unknown, it seems likely that this is the result of an inherent heterogeneity in the original tumor. Alternatively, this may indicate the response of this tumor to different factors in the tumor microenvironment. Interestingly, even when Tumor #2 was not inhibited by Apo2L/TRAIL alone, an increased amount of apoptosis was detected within the tumor early in the treatment. Although this variability is problematic, it is likely reflective of the situation that can be expected in the clinic and therefore, it is especially encouraging that combination therapy with gemcitabine shows such potential for complementing and enhancing the anti-tumor effect of Apo2L/TRAIL in the majority of these tumors.
Investigation of the basis for the difference in sensitivity to Apo2L/TRAIL
To further investigate the response of a sensitive and resistant tumor to Apo2L/TRAIL, we analyzed tumors from the experiments with Tumors #3 and #4 shown in Figure 5. Two mice were removed from the control and Apo2L/TRAIL treated groups on the second day of treatment (6 hours following treatment) and these tumors analyzed by Western Blot (Figure 8). We observed a distinct difference in the levels of several critical molecules in the apoptotic pathways in response to Apo2L/TRAIL. In Tumor #3 (Figure 8A), which is sensitive to Apo2L/TRAIL, greatly reduced levels of procaspase 8 and Bid are detected in the tumors following Apo2L/TRAIL treatment. Reduced levels of intact human mitochondria are also observed and there is a clear reduction in detectable procaspase 3. Samples of Tumor #4 were similarly examined. As can be seen in Figure 8B, tumors treated with Apo2L/TRAIL have only slightly reduced amounts of procaspase 8 and Bid and the levels of intact mitochondria and procaspase 3 are comparable to those seen in untreated tumors. These results suggest that Tumor #3 undergoes activation of procaspase 8 and 3 in response to Apo2L/TRAIL and that signaling through Bid cleavage results in recruitment of the mitochondrial pathway and subsequently, apoptosis. On the other hand there is only partial activation of procaspase 8 in the resistant Tumor #4 suggesting that there is a defect upstream of caspase 8 activation which results in failure to adequately activate caspase 8 and thereby initiate apoptotic signaling. One possible explanation for this reduced activation of caspase 8 could be a defect in Apo2L/TRAIL receptor expression on the surface of these cells and this remains to be determined.
Figure 8 The effect of Apo2L/TRAIL on levels of procaspase 8, Bid, a marker of intact mitochondria, and procaspase 3 in Tumor #3 (Apo2L/TRAIL sensitive) and Tumor #4 (Apo2L/TRAIL resistant). Two tumors were removed from each group shown in Figure 5 on day 2 of treatment. The lanes are: (+) - positive control for each antibody; C- two tumors from the control group; T- two tumors from the Apo2L/TRAIL treated group; A. The Apo2L/TRAIL sensitive Tumor #3. By the second day of treatment with Apo2L/TRAIL, the levels of procaspase 8, Bid, the mitochondrial marker and procaspase 3 (outlined) in these tumors are greatly reduced. B. The Apo2L/TRAIL resistant Tumor #4. In the tumor that was resistant to Apo2L/TRAIL treatment, although there is a slightly diminished amount of procaspase 8 and Bid present in these tumors following 2 days of treatment with Apo2L/TRAIL, the levels of the mitochondrial marker and procaspase 3 (outlined) remain comparable to the controls. (In A and B a representative actin loading control is shown for each tumor).
It has been previously reported that resistance of established pancreatic cell lines to Apo2L/TRAIL and chemotherapy is in part associated with levels of the anti-apoptotic molecule Bcl-XL [35,43]. Therefore the expression of Bcl-XL, a molecule that can block the loss of mitochondrial membrane potential, was investigated in these tumors. Bcl-XL expression was found to be high in several samples of the resistant Tumor #4, compared to the sensitive Tumor #3 in which Bcl-XL was undetectable in several passages (Figure 9). This differential expression of the anti-apoptotic molecule Bcl-XL in a tumor that is resistant to Apo2L/TRAIL is consistent with the idea that in Tumor #4, in which only little caspase 8 activation occurs, Bcl-XL may play a role in the resistance by inhibiting activation of the intrinsic apoptotic pathway by the low levels of cleaved Bid.
Figure 9 Comparison of Bcl-XL expression in a Tumor #4 that is resistant to Apo2L/TRAIL and Tumor #3 that is sensitive. Bcl-XL is detectable by Western blot in several passages of Tumor #4 (2nd passage, 5th passage) while Bcl-XL is not detectable in several passages of Tumor #3 (3rd passage, 8th passage, 11th passage).
Trauzold [37] characterized five pancreatic cell lines with regard to Apo2L/TRAIL sensitivity and concluded that although Bcl-XLwas differentially expressed in sensitive (Capan1, Colo357) vs. resistant (PancTul, Panc89, Panc1) cells and made a significant contribution to the observed resistance to Apo2L/TRAIL, it is not the only factor. These authors concluded that resistance arose from the combined effects of the downregulation of pro-apoptotic molecules (FADD, Bid) and the concurrent upregulation of anti-apoptotic molecules (Bcl-XL, FLIP or IAP). Their experiments support the idea that there is a balance of several pro- and anti-apoptotic factors in pancreatic cells that ultimately determines the efficacy of the apoptotic signal. Recently, Bai et al. have found that knock-down of Bcl-XL in pancreatic cells that predominantly overexpress it, results in increased sensitivity of these cells to TRAIL in combination with other anti-tumor drugs [44]. The possible role of Bcl-XL and other critical molecules, particularly those upstream of caspase 8, in resistance of patient pancreatic tumors to Apo2L/TRAIL needs to be more fully investigated. Additionally, the mechanism(s) of the interaction between Apo2L/TRAIL and gemcitabine needs to be determined.
Conclusion
Although much more work needs to be done, especially in evaluating a larger number of patients' tumors, these studies are important because they investigate for the first time the response of patient pancreatic tumors, grown as xenografts in SCID mice, to Apo2L/TRAIL. The results confirm previous work done with cell lines and support the idea that both the sensitivity and resistance to killing by Apo2L/TRAIL that has been observed in cell lines will be seen in patient tumors. Furthermore, these patients' tumors show heterogeneity of responsiveness both between tumors and within the same tumor that may be predictive of variability in a clinical setting. Importantly, it is encouraging that the combination of Apo2L/TRAIL with gemicitabine is able to enhance the antitumor efficacy and results in significant suppression of tumors that exhibit resistance to either one or both of these therapeutics. One potential benefit that remains to be explored is whether the enhanced efficacy achieved with combination therapy will allow lower doses of chemotherapy and/or Apo2L/TRAIL to be used, thus reducing the possibilities of toxicity and acquired resistance. The results of this study strongly support the further development of Apo2L/TRAIL as a therapeutic agent for the treatment of pancreatic adenocarcinoma.
Acknowledgements
This work used core facilities supported in part by Roswell Park Cancer Institute's National Cancer Institute-funded Cancer Center Support Grant CA 16056. The authors thank Jeanne Prendergast and Diane Thompson for their expert laboratory assistance. We also gratefully acknowledge the invaluable contributions of Nancy Reska, Harry Slocum and the Tissue Procurement Facility at RPCI.
==== Refs
Li D Xie K Wolff R Abbruzzese JL Pancreatic cancer Lancet 2004 363 1049 1057 15051286 10.1016/S0140-6736(04)15841-8
Chu QD Khushalani N Javle MM Douglass HOJ Gibbs JF Should adjuvant therapy remain the standard of care for patients with resected adenocarcinoma of the pancreas? Ann Surg Oncol 2003 10 539 545 12794020 10.1245/ASO.2003.06.015
Jemal A Murray T Samuels A Ghafoor A Ward E Thun MJ Cancer statistics, 2003 CA Cancer J Clin 2003 53 5 26 12568441
Kulke MH Recent developments in the pharmacological treatment of advanced pancreatic cancer Expert Opin Investig Drugs 2003 12 983 992 12783602 10.1517/13543784.12.6.983
Neoptolemos JP Cunningham D Friess H Bassi C Stocken DD Tait DM Dunn JA Dervenis C Lacaine F Hickey H Raraty MG Ghaneh P Buchler MW Adjuvant therapy in pancreatic cancer: historical and current perspectives Ann Oncol 2003 14 675 692 12702520 10.1093/annonc/mdg207
Rosenberg L Lipsett M Biotherapeutic approaches to pancreatic cancer Expert Opin Biol Ther 2003 3 319 337 12662145 10.1517/14712598.3.2.319
Mohammad RM Dugan MC Mohamed AN Almatchy VP Flake TM Dergham ST Shields AF Al-Katib AA Vaitkevicius VK Sarkar FH Establishment of a human pancreatic tumor xenograft model: potential application for preclinical evaluation of novel therapeutic agents Pancreas 1998 16 19 25 9436858
Xiong HQ Rosenberg A LoBuglio A Schmidt W Wolff RA Deutsch J Needle M Abbruzzese JL Li D Xie K Wolff R Cetuximab, a monoclonal antibody targeting the epidermal growth factor receptor, in combination with gemcitabine for advanced pancreatic cancer: a multicenter phase II Trial
Pancreatic cancer J Clin Oncol 2004 22 2610 2616 15226328 10.1200/JCO.2004.12.040
Bruns CJ Harbison MT Davis DW Portera CA Tsan R McConkey DJ Evans DB Abbruzzese JL Hicklin DJ Radinsky R Epidermal growth factor receptor blockade with C225 plus gemcitabine results in regression of human pancreatic carcinoma growing orthotopically in nude mice by antiangiogenic mechanisms Clin Cancer Res 2000 6 1936 1948 10815919
Ng SS Tsao MS Nicklee T Hedley DW Effects of the epidermal growth factor receptor inhibitor OSI-774, Tarceva, on downstream signaling pathways and apoptosis in human pancreatic adenocarcinoma Mol Cancer Ther 2002 1 777 783 12492110
Wiley SR Schooley K Smolak PJ Din WS Huang CP Nicholl JK Sutherland GR Smith TD Rauch C Smith CA Identification and characterization of a new member of the TNF family that induces apoptosis Immunity 1995 3 673 682 8777713 10.1016/1074-7613(95)90057-8
Pitti RM Marsters SA Ruppert S Donahue CJ Moore A Ashkenazi A Induction of apoptosis by Apo-2 ligand, a new member of the tumor necrosis factor cytokine family J Biol Chem 1996 271 12687 12690 8663110 10.1074/jbc.271.22.12687
Griffith TS Lynch DH TRAIL: a molecule with multiple receptors and control mechanisms Curr Opin Immunol 1998 10 559 563 9794836 10.1016/S0952-7915(98)80224-0
Ashkenazi A Pai RC Fong S Leung S Lawrence DA Marsters SA Blackie C Chang L McMurtrey AE Hebert A DeForge L Koumenis IL Lewis D Harris L Bussiere J Koeppen H Shahrokh Z Schwall RH Safety and antitumor activity of recombinant soluble Apo2 ligand J Clin Invest 1999 104 155 162 10411544
Walczak H Miller RE Ariail K Gliniak B Griffith TS Kubin M Chin W Jones J Woodward A Le T Smith C Smolak P Goodwin RG Rauch CT Schuh JC Lynch DH Tumoricidal activity of tumor necrosis factor-related apoptosis-inducing ligand in vivo Nat Med 1999 5 157 163 9930862 10.1038/5517
Almasan A Ashkenazi A Apo2L/TRAIL: apoptosis signaling, biology, and potential for cancer therapy Cytokine Growth Factor Rev 2003 14 337 348 12787570 10.1016/S1359-6101(03)00029-7
Ashkenazi A Dixit VM Apoptosis control by death and decoy receptors Curr Opin Cell Biol 1999 11 255 260 10209153 10.1016/S0955-0674(99)80034-9
Ashkenazi A Targeting death and decoy receptors of the tumour-necrosis factor superfamily Nat Rev Cancer 2002 2 420 430 12189384 10.1038/nrc821
Smyth MJ Takeda K Hayakawa Y Peschon JJ van den Brink MR Yagita H Nature's TRAIL-On a Path to Cancer Immunotherapy Immunity 2003 18 1 6 12530970 10.1016/S1074-7613(02)00502-2
Gliniak B Le T Tumor necrosis factor-related apoptosis-inducing ligand's antitumor activity in vivo is enhanced by the chemotherapeutic agent CPT-11 Cancer Res 1999 59 6153 6158 10626806
Keane MM Ettenberg SA Nau MM Russell EK Lipkowitz S Chemotherapy augments TRAIL-induced apoptosis in breast cell lines Cancer Res 1999 59 734 741 9973225
Frese S Brunner T Gugger M Uduehi A Schmid RA Enhancement of Apo2L/TRAIL (tumor necrosis factor-related apoptosis-inducing ligand)-induced apoptosis in non-small cell lung cancer cell lines by chemotherapeutic agents without correlation to the expression level of cellular protease caspase-8 inhibitory protein J Thorac Cardiovasc Surg 2002 123 168 174 11782771 10.1067/mtc.2002.119694
Nimmanapalli R Perkins CL Orlando M O'Bryan E Nguyen D Bhalla KN Pretreatment with paclitaxel enhances apo-2 ligand/tumor necrosis factor-related apoptosis-inducing ligand-induced apoptosis of prostate cancer cells by inducing death receptors 4 and 5 protein levels Cancer Res 2001 61 759 763 11212279
Liu W Bodle E Chen JY Gao M Rosen GD Broaddus VC Tumor necrosis factor-related apoptosis-inducing ligand and chemotherapy cooperate to induce apoptosis in mesothelioma cell lines Am J Respir Cell Mol Biol 2001 25 111 118 11472983
Dejosez M Ramp U Mahotka C Krieg A Walczak H Gabbert HE Gerharz CD Sensitivity to TRAIL/APO-2L-mediated apoptosis in human renal cell carcinomas and its enhancement by topotecan Cell Death Differ 2000 7 1127 1136 11139287 10.1038/sj.cdd.4400746
Cuello M Ettenberg SA Nau MM Lipkowitz S Synergistic induction of apoptosis by the combination of trail and chemotherapy in chemoresistant ovarian cancer cells Gynecol Oncol 2001 81 380 390 11371126 10.1006/gyno.2001.6194
Mizutani Y Nakao M Ogawa O Yoshida O Bonavida B Miki T Enhanced sensitivity of bladder cancer cells to tumor necrosis factor related apoptosis inducing ligand mediated apoptosis by cisplatin and carboplatin J Urol 2001 165 263 270 11125422 10.1097/00005392-200101000-00076
Nagane M Pan G Weddle JJ Dixit VM Cavenee WK Huang HJ Increased death receptor 5 expression by chemotherapeutic agents in human gliomas causes synergistic cytotoxicity with tumor necrosis factor-related apoptosis-inducing ligand in vitro and in vivo Cancer Res 2000 60 847 853 10706092
Matsuzaki H Schmied BM Ulrich A Standop J Schneider MB Batra SK Picha KS Pour PM Combination of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and actinomycin D induces apoptosis even in TRAIL-resistant human pancreatic cancer cells Clin Cancer Res 2001 7 407 414 11234897
Naka T Sugamura K Hylander BL Widmer MB Rustum YM Repasky EA Effects of tumor necrosis factor-related apoptosis-inducing ligand alone and in combination with chemotherapeutic agents on patients' colon tumors grown in SCID mice Cancer Res 2002 62 5800 5806 12384541
Sakakibara T Xu Y Bumpers HL Chen FA Bankert RB Arredondo MA Edge SB Repasky EA Growth and Metastasis of Surgical Specimens of Human Breast Carcinomas in SCID Mice Cancer J Sci Am 1996 2 291 9166547
Silver DF Hempling RE Piver MS Repasky EA Effects of IL-12 on human ovarian tumors engrafted into SCID mice Gynecol Oncol 1999 72 154 160 10021294 10.1006/gyno.1998.5239
Xu Y Silver DF Yang NP Oflazoglu E Hempling RE Piver MS Repasky EA Characterization of human ovarian carcinomas in a SCID mouse model Gynecol Oncol 1999 72 161 170 10021295 10.1006/gyno.1998.5238
Ghamande S Hylander BL Oflazoglu E Lele S Fanslow W Repasky EA Recombinant CD40 ligand therapy has significant antitumor effects on CD40-positive ovarian tumor xenografts grown in SCID mice and demonstrates an augmented effect with cisplatin Cancer Res 2001 61 7556 7562 11606394
Hinz S Trauzold A Boenicke L Sandberg C Beckmann S Bayer E Walczak H Kalthoff H Ungefroren H Bcl-XL protects pancreatic adenocarcinoma cells against CD95- and TRAIL-receptor-mediated apoptosis Oncogene 2000 19 5477 5486 11114725 10.1038/sj.onc.1203936
Satoh K Kaneko K Hirota M Masamune A Satoh A Shimosegawa T Tumor necrosis factor-related apoptosis-inducing ligand and its receptor expression and the pathway of apoptosis in human pancreatic cancer Pancreas 2001 23 251 258 11590320 10.1097/00006676-200110000-00005
Trauzold A Schmiedel S Roder C Tams C Christgen M Oestern S Arlt A Westphal S Kapischke M Ungefroren H Kalthoff H Multiple and synergistic deregulations of apoptosis-controlling genes in pancreatic carcinoma cells Br J Cancer 2003 89 1714 1721 14583775 10.1038/sj.bjc.6601330
Nozawa F Itami A Saruc M Kim M Standop J Picha KS Cowan KH Pour PM The combination of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL/Apo2L) and Genistein is effective in inhibiting pancreatic cancer growth Pancreas 2004 29 45 52 15211111 10.1097/00006676-200407000-00055
Xu ZW Kleeff J Friess H Buchler MW Solioz M Synergistic cytotoxic effect of TRAIL and gemcitabine in pancreatic cancer cells Anticancer Res 2003 23 251 258 12680221
Ibrahim SM Ringel J Schmidt C Ringel B Muller P Koczan D Thiesen HJ Lohr M Pancreatic adenocarcinoma cell lines show variable susceptibility to TRAIL-mediated cell death Pancreas 2001 23 72 79 11451151 10.1097/00006676-200107000-00011
Repasky EA Tims E Pritchard M Burd R Characterization of mild whole-body hyperthermia protocols using human breast, ovarian, and colon tumors grown in severe combined immunodeficient mice Infect Dis Obstet Gynecol 1999 7 91 97 10231015 10.1002/(SICI)1098-0997(1999)7:1/2<91::AID-IDOG16>3.0.CO;2-F
Robinson BW Ostruszka L Im MM Shewach DS Promising combination therapies with gemcitabine Semin Oncol 2004 31 2 12 15199526
Shi X Liu S Kleeff J Friess H Buchler MW Acquired resistance of pancreatic cancer cells towards 5-Fluorouracil and gemcitabine is associated with altered expression of apoptosis-regulating genes Oncology 2002 62 354 362 12138244 10.1159/000065068
Bai J Sui J Demirjian A Vollmer CMJ Marasco W Callery MP Predominant Bcl-XL knockdown disables antiapoptotic mechanisms: tumor necrosis factor-related apoptosis-inducing ligand-based triple chemotherapy overcomes chemoresistance in pancreatic cancer cells in vitro Cancer Res 2005 65 2344 2352 15781649
| 15943879 | PMC1156958 | CC BY | 2021-01-04 16:39:28 | no | J Transl Med. 2005 May 19; 3:22 | utf-8 | J Transl Med | 2,005 | 10.1186/1479-5876-3-22 | oa_comm |
==== Front
World J Surg OncolWorld Journal of Surgical Oncology1477-7819BioMed Central London 1477-7819-3-291591345410.1186/1477-7819-3-29Case ReportVentricular metastasis resulting in disseminated intravascular coagulation John Thomas [email protected] Ian D [email protected] Ludwig Institute Oncology Unit, Austin Health, Heidelberg, Victoria 3084, Australia2005 24 5 2005 3 29 29 1 3 2005 24 5 2005 Copyright © 2005 John and Davis; licensee BioMed Central Ltd.2005John and Davis; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Disseminated Intravascular Coagulation (DIC) complicates up to 7% of malignancies, the commonest solid organ association being adenocarcinoma. Transitional Cell Carcinoma (TCC) has rarely been associated with DIC.
Case presentation
A 74-year-old woman with TCC bladder and DIC was found to have a cardiac lesion suspicious for metastatic disease. The DIC improved with infusion of plasma and administration of Vitamin K, however the cardiac lesion was deemed inoperable and chemotherapy inappropriate; given the patients functional status. We postulate that direct activation of the coagulation cascade by the intraventricular metastasis probably triggered the coagulopathy in this patient.
Conclusion
Cardiac metastases should be considered in cancer patients with otherwise unexplained DIC. This may influence treatment choices.
==== Body
Background
DIC is characterised by the widespread activation of coagulation. This in turn results in intravascular formation of fibrin and ultimately thrombotic occlusion of small to medium sized vessels [1]. The commonest causes are sepsis and trauma. Malignancy is a well recognised cause of a prothrombotic state, with DIC occurring in up to 7% of solid organ tumours[2]. It is most frequently associated with adenocarcinomas such as pancreatic, breast and prostate cancer. Transitional cell carcinoma is rarely associated with DIC, with very few reports in the literature.
Case presentation
A 74 year old woman presented with a two day history of haematuria and increasing lethargy one month following palliative radiotherapy to the bladder for a T4 Grade III urothelial carcinoma. Clinical examination revealed large ecchymoses over both upper and lower limbs as well as oral mucosal bleeding without any evidence of petechiae. The patient was apyrexial and vital signs were all within normal limits. There were no other significant findings on cardiovascular, respiratory and abdominal examination.
Laboratory analysis revealed INR 2.7 (normal range: 0.9–1.3), APTT 73 sec (25–38) D-dimer 25 μg/mL (0–0.2), fibrinogen <0.05 g/L (1.5–4.0), platelets 25 × 109/L (150–400), consistent with disseminated intravascular coagulation (DIC). Computerised tomographic (CT) scan of the thorax and abdomen revealed a lesion within the right ventricle (Figure 1) without other evidence of metastatic disease within the pulmonary or hepatic parenchyma. The primary site appeared necrotic consistent with recent radiotherapy. Trans-oesophageal echocardiography showed that the lesion extended into the body of the right ventricle and significantly reduced ventricular volume (Figure 2). The edges of the lesion were fimbriated and mobile, consistent with a metastatic deposit. Although the platelet count and clotting improved with Vitamin K and plasma, the platelet count did not rise above 50. This was adequate in controlling her haematuria and further treatments were therefore not initiated. As the patient did not show any overt signs of thrombosis, we did not institute therapy with anticoagulants. As the patient's functional status was declining, we were unable to consider chemotherapeutic options. Palliative radiotherapy to the heart was considered but the patient deteriorated and died three months later. A request to perform an autopsy was refused by the patient's family.
Figure 1 Contrast CT thorax demonstrating filling defect and lesion in right anterior ventricular wall.
Figure 2 Trans oesophageal echo demonstrating right ventricular mass (arrows) RA: right atrium, LA: left atrium, LV: left ventricle.
Discussion
Although a hypercoagulable state exists in virtually all patients with malignancy, the incidence of overt DIC is much lower [3]. The aetiology of the activation of coagulability in malignancy is multifactorial and poorly understood. Molecular changes which may influence hypercoagulability include the expression of tissue factor (TF) as well as the proteases hepsin and cancer procoagulant by circulating tumour cells [4].
TF assembles with factor VIIa to activate factors IX and X thereby triggering thrombin formation. TF has been demonstrated to exist in normal vascular endothelial cells suggesting a role in initiating rapid activation of coagulation. It is rarely expressed in normal epithelial tissue but has been found to be expressed in malignant tissue such as breast cancer, whilst not being present in patients with benign fibrocystic disease [5]. The degree of tissue factor expression has been shown to correlate with metastatic potential and histological differentiation such that increased TF expression results an increased potential to metastasise and a more undifferentiated histological appearance [6,7]. Cancer procoagulant has been found in malignant and foetal tissue, but not in normally differentiated tissue[8,9]. Its only known property is the calcium and vitamin K dependent conversion of factor X into Xa, independent of the TF / factor VIIa complex. Hepsin, a protease found on hepatocyte surface membranes but also in several tumour cell lines, has been shown to initiate coagulation by activating factor VIIa independent of TF [10].
The heart is an under-recognised site of metastatic disease in patients with cancer. In one series, the incidence at autopsy was 1.25%. The commonest types of malignancy to spread to the heart in this series were lung and oesophageal cancers as well as lymphomas [11]. This reflects the direct proximity of these tumours to the heart, with direct invasion being the likely pathogenesis. Lung, breast, soft tissue carcinomas, renal cell carcinomas, lymphomas and melanoma are most frequently associated with cardiac involvement. This usually results in pericardial or epicardial disease, with endocardial metastases being less common. There are reports of cardiac metastases from metastatic TCC in the literature, however we could find only one other report of a patient with a cardiac metastases presenting with DIC [12] and this was associated with cervical carcinoma.
The mechanism of DIC in the described case is speculative but may relate to the activation of coagulation pathways directly by the interaction of tumour antigens with plasma [13]. The postulated mechanism of DIC in this case is the direct exposure of tumour cells within the right ventricle to circulating plasma, thus triggering direct activation of the coagulation cascade through mechanisms mentioned above. Although we are unable to prove causality, given that TCC is rarely associated with DIC and that this phenomenon only developed when the tumour metastasised to the patient's heart; this explanation appears more plausible than the assumption that the patients malignancy alone resulted in DIC.
Conclusion
This case demonstrates that cardiac metastases should be considered in all cancer patients with otherwise unexplained DIC and that this may influence treatment choices.
Conflict of interest
The author(s) declare that they have no competing interests.
Authors' contributions
TJ – principal author: preparation of manuscript, literature review, acquisition of images.
ID – revision of manuscript and figures.
All authors read and approved the final manuscript.
Acknowledgements
Permission was obtained through University of Melbourne Human Research and Ethics Committee for publication of this case report as due to death of the patient her consent could not be obtained.
==== Refs
Levi M Ten Cate H Disseminated intravascular coagulation N Engl J Med 1999 341 586 592 10451465 10.1056/NEJM199908193410807
Sallah S Wan JY Nguyen NP Hanrahan LR Sigounas G Disseminated intravascular coagulation in solid tumors: clinical and pathologic study Thromb Haemost 2001 86 828 833 11583315
Colman RW Rubin RN Disseminated intravascular coagulation due to malignancy Semin Oncol 1990 17 172 186 2183359
Sampson MT Kakkar AK Coagulation proteases and human cancer Biochem Soc Trans 2002 30 201 207 12023851 10.1042/BST0300201
Contrino J Hair G Kreutzer DL Rickles FR In situ detection of tissue factor in vascular endothelial cells: correlation with the malignant phenotype of human breast disease Nat Med 1996 2 209 215 8574967 10.1038/nm0296-209
Kakkar AK Chinswangwatanakul V Lemoine NR Tebbutt S Williamson RC Role of tissue factor expression on tumour cell invasion and growth of experimental pancreatic adenocarcinoma Br J Surg 1999 86 890 894 10417560 10.1046/j.1365-2168.1999.01153.x
Kakkar AK Lemoine NR Scully MF Tebbutt S Williamson RC Tissue factor expression correlates with histological grade in human pancreatic cancer Br J Surg 1995 82 1101 1104 7648165
Gordon SG Mielicki WP Cancer procoagulant: a factor X activator, tumor marker and growth factor from malignant tissue Blood Coagul Fibrinolysis 1997 8 73 86 9518049
Gordon SG Franks JJ Lewis B Cancer procoagulant A: a factor X activating procoagulant from malignant tissue Thromb Res 1975 6 127 137 234638 10.1016/0049-3848(75)90018-3
Kazama Y Hamamoto T Foster DC Kisiel W Hepsin, a putative membrane-associated serine protease, activates human factor VII and initiates a pathway of blood coagulation on the cell surface leading to thrombin formation J Biol Chem 1995 270 66 72 7814421 10.1074/jbc.270.1.66
Lam KY Dickens P Chan AC Tumors of the heart. A 20-year experience with a review of 12,485 consecutive autopsies Arch Pathol Lab Med 1993 117 1027 1031 8215825
Harvey RL Mychaskiw G Sachdev V Heath BJ Isolated cardiac metastasis of cervical carcinoma presenting as disseminated intravascular coagulopathy. A case report J Reprod Med 2000 45 603 606 10948477
Sutherland DE Weitz IC Liebman HA Thromboembolic complications of cancer: epidemiology, pathogenesis, diagnosis, and treatment Am J Hematol 2003 72 43 52 12508268 10.1002/ajh.10263
| 15913454 | PMC1156959 | CC BY | 2021-01-04 16:39:03 | no | World J Surg Oncol. 2005 May 24; 3:29 | utf-8 | World J Surg Oncol | 2,005 | 10.1186/1477-7819-3-29 | oa_comm |
==== Front
World J Surg OncolWorld Journal of Surgical Oncology1477-7819BioMed Central London 1477-7819-3-311592705910.1186/1477-7819-3-31Case ReportFallacies of preoperative lymphoscintigraphy in detecting sentinel node in breast cancer Pandey Manoj [email protected] Surya VS [email protected] R [email protected] Surgical Oncology, Jawaharlal Nehru Cancer Hospital and Research Centre, Bhopal, India2 Surgical Oncology, Institute Rotary Cancer Hospital, All India Institute of Medical Sciences, New Delhi 110 029, India3 Department of Nuclear Medicine, West fort Hi-tech Hospital Ltd, Punkunnam, Thrissure 680 002, India2005 31 5 2005 3 31 31 14 4 2005 31 5 2005 Copyright © 2005 Pandey et al; licensee BioMed Central Ltd.2005Pandey et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Preoperative lymphoscintigraphy is one of the three methods of evaluating sentinel nodes in patients with breast cancer; however, it has been reported to have a high false negative rate.
Case presentations
We report here two cases where the preoperative lymphoscintigraphy was found to be fallacious. A 44-year-old female with T2N0 breast cancer underwent preoperative lymphoscintigraphy with Tc99 sulfur colloid which failed to show any uptake in axilla or internal mammary chain. Intraoperative scintigraphy with blue dye and hand held gamma probe identified sentinel lymph node in axilla. Another patient with T2N0 lesion underwent preoperative lymphoscintigraphy which showed a sentinel lymph node in axilla and another in supraclevicular fossa. Intraoperative scintigraphy failed to show supraclevicular node however axillary node was correctly identified.
Conclusion
These two cases further strengthen the need to carry out triple test in identification of sentinel lymph node in patients with breast cancer. It also demonstrates the fallacies of preoperative lymphoscintigraphy.
==== Body
Background
Metastasis to the axillary lymph node is the single most important prognostic factor in breast cancer. The therapeutic decisions are based on the axillary status. However, in recent past sentinel lymph node identification and biopsy (SLNB) is fast emerging as an alternate to axillary dissection as it avoids the complications of axillary dissection like lymphedema, pain, numbness and limitations of shoulder movements [1,2]. SLNB has been found to be highly predictive of axillary lymph node status with false negative results of less than 5% [3-5]. A number of validity studies have been published however, the question of its oncological safety still awaits the results of randomized clinical trials [6,7].
The sentinel lymph node identification is usually carried out by preoperative localization using Tc99 colloid and gamma camera or by intraoperative localization using hand held gamma probe or by dye technique. Majority of the centers use a combination of techniques and it has been reported that the triple method using all of the above gives the best results [8-10].
Non-visualization of sentinel node at preoperative scintigraphy is a continued problem. Between 3 to 30% of the nodes are reported to be non visualized most of which are subsequently picked up on intraoperative scintigraphy [11-13]. Several factors like age, size of the breast, presence of metastasis, has been proposed to influence the non-visualization [11-17]. We report here two unusual cases of non-visualization or abnormal visualization during preoperative lymphoscintigraphy; both of these cases were subsequently identified and biopsied intraoperatively using a combination of blue dye and hand held gamma probe.
Case presentation
Case 1
A 44-year-old female presented with 2 months history of a progressive lump in right breast. She gave a past history of noncyclic mastalgia of two years duration. There was no other significant past history. Patient had undergone abdominal hysterectomy 4 years back for dysfunction uterine bleeding and was on hormone replacement therapy with estrogen alone for the same duration.
On examination there was 4 × 3 cm lump in upper outer quadrant of the right breast with no fixity to skin or underlying tissue. There were no significant axillary or supraclavicular nodes. Abdominal examination failed to show any organomegaly. Routine hematological, biochemical tests, chest roentgenogram, abdominal ultrasonogram and bone scans were normal. Fine needle aspiration cytology revealed Infiltrating duct carcinoma.
Patient was planned for sentinel lymph node biopsy follow by mastectomy. 8 ml of Tc99 sulfur colloid was injected around the tumor and immunoscintigraphy images (anterior and lateral view) were taken. These images failed to show any lymph node uptake either in axilla or else where (figure 1) at the time of surgery 4 ml of isosulfan blue was injected peritumorilly and 20 minutes later axilla was entered. Sentinel node detection was also carried out using a hand held gamma probe (Navigator®, Auto sutures). Sentinel lymph node was identified by combined technique in level I axilla lying just posterior to the primary tumor. On gross examination most of the node appeared to be replaced by tumor and only a part of it appeared normal this part was stained blue while rest of the node was white. Histopathology of primary tumors was infiltrating duct carcinoma, with involvement of skin. Sentinel lymph node showed metastatic deposit. Other lymph nodes in axillary dissection specimen were also positive.
Figure 1 Lymphoscientigraphic scan showing radioneucleotide uptake in primary tumor, no sentinel node is identified.
Case 2
A 42-year-old female presented with lump in left breast of 1-month duration. She was a known case of carcinoma breast and had undergone right modified radical mastectomy 7 years back followed by chest wall radiotherapy and 6 cycles of CMF chemotherapy.
On examination there was a 3 × 2 cm lump located in the retro areolar area of left breast. Scar of right mastectomy was seen. There were no palpable axillary nodes or supra clavicular nodes. Systemic examination was normal. Routine hematological and biochemical investigations chest x-ray, abdominal sonography and bone scans were normal. The treatment options were discuss with the patient and she wanted to conserve this breast and hence a wide excision of mass encompassing nipple areola complex, sentinel node biopsy followed by axillary clearance was planned.
On the morning of the surgery 4 ml of Tc99 sulfur colloid was injected peritumorily and scintigraphic images were taken 4 hours later. The scintigraphic image showed one sentinel node in axilla and other in supra clavicular fossa (figure 2). At surgery 4 ml of isosulfan blue was injected peritumorilly and sentinel node identification was carried out by combined method. On exploration of axilla the blue and hot sentinel node was identified and removed. However, hand held probe failed to pick a hot spot in supra clavicular fossa axillary dissection was completed.
Figure 2 Lymphoscintigraphic scan showing uptake in primary tumor with sentinel node in axilla and in left supra clavicular area.
Histopathology of the resected specimen showed infiltrative duct carcinoma margins of resection were negative. The sentinel node showed deposits from infiltrating duct carcinoma. Post operative period was uneventful and it is planned to start her on radiotherapy to residual breast with 4 cycles of anthracyclin based chemotherapy.
Discussion
The axillary dissection for axillary nodes from breast cancer is still the standard of care its routine use in node negative breast cancer has been questioned due to morbidity associated with the axillary dissection. SLNB has improved the morbidity in patients with node negative breast cancer, while providing the much needed prognostic information. Although these techniques have been successful, they are still evolving, and SLN biopsy is not yet considered the standard of care in breast cancer.
Preoperative visualization is one of the three methods commonly employed in detection of sentinel node. When radioactive colloid is used, a preoperative lymphoscintigram often is obtained to ease SLN identification further. This has been reported to have a false negative rate of 3–30% in various series [11-13]. Several factors like age, size of the breast, presence of metastasis, neoadjuvant chemotherapy, has been proposed to influence the non-visualization [11-19]. Other authors have found no significant predictor of non visualization [20]. The non-visualization in our first case was due to superimposition of locally advance tumor in the outer quadrant and hence separation was not achieved on scintigraphy, while the erroneous supra clavicular node in the other was probably due to the spillage of radioactive material at the time of injection, which was subsequently washed off during the part preparation and hence no signal was obtained at intraoperative gamma probe assisted SLNB.
Conclusion
These two cases demonstrate the importance of using the triple technique to maximize the identification of SLN and improve the sensitivity and specificity of SLNB and fallacies of preoperative lymphoscintigraphy. This also raises a question that should preoperative scintigraphy should be carried out in all the cases?
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MP: Conceived the idea, carried out the literature search, prepared the draft manuscript and edited it for publication.
SVSD: helped in preparation of the manuscript and edited the final version
RM: Carried out the preoperative lymphoscintigraphy and helped in preparation of the manuscript.
MP and SVSD carried out the intraoperative scintigraphy and managed the patients.
All authors read and approved the final manuscript.
Acknowledgements
Patient consent was obtained for publication of these case reports.
==== Refs
Taylor KO Morbidity associated with axillary surgery for breast cancer ANZ J Surg 2004 74 314 317 15144248 10.1111/j.1445-1433.2004.02992.x
Poole K Fallowfield LJ The psychological impact of post-operative arm morbidity following axillary surgery for breast cancer: a critical review Breast 2002 11 81 87 14965650 10.1054/brst.2001.0369
Veronesi U Paganelli G Galimberti V Viale G Zurrida S Bedoni M Costa A De CC Geraghty JG Luini A Sacchini V Veronesi P Sentinel-node biopsy to avoid axillary dissection in breast cancer with clinically negative lymph-nodes Lancet 1997 349 1864 1867 9217757 10.1016/S0140-6736(97)01004-0
Veronesi U Paganelli G Viale G Galimberti V Luini A Zurrida S Robertson C Sacchini V Veronesi P Orvieto E De CC Intra M Tosi G Scarpa D Sentinel lymph node biopsy and axillary dissection in breast cancer: results in a large series J Natl Cancer Inst 1999 91 368 373 10050871 10.1093/jnci/91.4.368
Veronesi U Galimberti V Zurrida S Pigatto F Veronesi P Robertson C Paganelli G Sciascia V Viale G Sentinel lymph node biopsy as an indicator for axillary dissection in early breast cancer Eur J Cancer 2001 37 454 458 11267853 10.1016/S0959-8049(00)00410-X
Mansel RE Goyal A European studies on breast lymphatic mapping Semin Oncol 2004 31 304 310 15190486
Veronesi U Paganelli G Viale G Luini A Zurrida S Galimberti V Intra M Veronesi P Robertson C Maisonneuve P Renne G De Cicco C De Lucia F Gennari R A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer N Engl J Med 2003 349 546 553 12904519 10.1056/NEJMoa012782
Meyer-Rochow GY Martin RC Harman CR Sentinel node biopsy in breast cancer: validation study and comparison of blue dye alone with triple modality localization ANZ J Surg 2003 73 815 818 14525573 10.1046/j.1445-2197.2003.02783.x
Rahusen FD Pijpers R van Diest PJ Bleichrodt RP Torrenga H Meijer S The implementation of the sentinel node biopsy as a routine procedure for patients with breast cancer Surgery 2000 128 6 12 10876178 10.1067/msy.2000.107229
Goyal A Newcombe RG Mansel RE Chetty U Ell P Fallowfield L Kissin M Sibbering M Role of routine preoperative lymphoscintigraphy in sentinel node biopsy for breast cancer Eur J Cancer 2005 41 238 243 15661548 10.1016/j.ejca.2004.05.008
Birdwell RL Smith KL Betts BJ Ikeda DM Strauss HW Jeffrey SS Breast cancer: variables affecting sentinel lymph node visualization at preoperative lymphoscintigraphy Radiology 2001 220 47 53 11425971
Burak WEJ Walker MJ Yee LD Kim JA Saha S Hinkle G Olsen JO Pozderac R Farrar WB Routine preoperative lymphoscintigraphy is not necessary prior to sentinel node biopsy for breast cancer Am J Surg 1999 177 445 449 10414690 10.1016/S0002-9610(99)00088-4
Wu CT Morita ET Treseler PA Esserman LJ Hwang ES Kuerer HM Santos CL Leong SP Failure to harvest sentinel lymph nodes identified by preoperative lymphoscintigraphy in breast cancer patients Breast J 2003 9 86 90 12603380 10.1046/j.1524-4741.2003.09205.x
Haigh PI Hansen NM Giuliano AE Edwards GK Ye W Glass EC Factors affecting sentinel node localization during preoperative breast lymphoscintigraphy J Nucl Med 2000 41 1682 1688 11037998
Kern KA Rosenberg RJ Preoperative lymphoscintigraphy during lymphatic mapping for breast cancer: improved sentinel node imaging using subareolar injection of technetium 99m sulfur colloid J Am Coll Surg 2000 191 479 489 11085727 10.1016/S1072-7515(00)00720-1
McMasters KM Wong SL Tuttle TM Carlson DJ Brown CM Dirk NR Glaser RL Vennekotter DJ Turk PS Tate PS Sardi A Edwards MJ Preoperative lymphoscintigraphy for breast cancer does not improve the ability to identify axillary sentinel lymph nodes Ann Surg 2000 231 724 731 10767794 10.1097/00000658-200005000-00013
Mudun A Aslay I Aygen M Muslumanoglu M Bozfakioglu Y Cantez S Can preoperative lymphoscintigraphy be used as a guide in treatment planning of breast cancer? Clin Nucl Med 2001 26 405 411 11317020 10.1097/00003072-200105000-00007
Kang SH Kim SK Kwon Y Kang HS Kang JH Ro J Lee ES Decreased identification rate of sentinel lymph node after neoadjuvant chemotherapy World J Surg 2004 28 1019 1024 15573258 10.1007/s00268-004-7367-7
Xing Y Cormier JN Kuerer HM Hunt KK Sentinel lymph node biopsy following neoadjuvant chemotherapy: review of the literature and recommendations for use in patient management Asian J Surg 2004 27 262 267 15564176
Rubello D Zavagno G Bozza F Lise M De Salvo GL Saladini G Mariani G Casara D Analysis of technical and clinical variables affecting sentinel node localization in patients with breast cancer after a single intradermal injection of 99mTc nanocolloidal albumin Nucl Med Commun 2004 25 1119 1124 15577591 10.1097/00006231-200411000-00009
| 15927059 | PMC1156960 | CC BY | 2021-01-04 16:39:03 | no | World J Surg Oncol. 2005 May 31; 3:31 | utf-8 | World J Surg Oncol | 2,005 | 10.1186/1477-7819-3-31 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1596677010.1371/journal.pbio.0030225Research ArticleBioinformatics/Computational BiologyGenetics/Genomics/Gene TherapySystems BiologySchizosaccharomymesThe Cell Cycle–Regulated Genes of Schizosaccharomyces pombe
The Cell Cycle-Regulated Genes of S. pombeOliva Anna
1
Rosebrock Adam
1
Ferrezuelo Francisco
1
Pyne Saumyadipta
2
Chen Haiying
1
Skiena Steve
2
Futcher Bruce [email protected]
1
Leatherwood Janet [email protected]
1
1 Department of Molecular Genetics and Microbiology, Stony Brook University, Stony Brook, New York, United States of America,2 Department of Computer Science, Stony Brook University, Stony Brook, New York, United States of AmericaSpellman Paul T. Academic EditorLawrence Berkeley LabUnited States of America7 2005 28 6 2005 28 6 2005 3 7 e22520 12 2004 21 4 2005 Copyright: © 2005 Oliva et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Transcriptional Waves in the Yeast Cell Cycle
Many genes are regulated as an innate part of the eukaryotic cell cycle, and a complex transcriptional network helps enable the cyclic behavior of dividing cells. This transcriptional network has been studied in Saccharomyces cerevisiae (budding yeast) and elsewhere. To provide more perspective on these regulatory mechanisms, we have used microarrays to measure gene expression through the cell cycle of Schizosaccharomyces pombe (fission yeast). The 750 genes with the most significant oscillations were identified and analyzed. There were two broad waves of cell cycle transcription, one in early/mid G2 phase, and the other near the G2/M transition. The early/mid G2 wave included many genes involved in ribosome biogenesis, possibly explaining the cell cycle oscillation in protein synthesis in S. pombe. The G2/M wave included at least three distinctly regulated clusters of genes: one large cluster including mitosis, mitotic exit, and cell separation functions, one small cluster dedicated to DNA replication, and another small cluster dedicated to cytokinesis and division. S. pombe cell cycle genes have relatively long, complex promoters containing groups of multiple DNA sequence motifs, often of two, three, or more different kinds. Many of the genes, transcription factors, and regulatory mechanisms are conserved between S. pombe and S. cerevisiae. Finally, we found preliminary evidence for a nearly genome-wide oscillation in gene expression: 2,000 or more genes undergo slight oscillations in expression as a function of the cell cycle, although whether this is adaptive, or incidental to other events in the cell, such as chromatin condensation, we do not know.
A comprehensive examination of gene expression throughout the cell cycle of fission yeast is compared with recent related studies to highlight robust transcriptional patterns.
==== Body
Introduction
The yeasts Schizosaccharomyces pombe and Saccharomyces cerevisiae are excellent organisms for the study of the cell division cycle. Both yeasts have many well-characterized cell division cycle (cdc) mutants [1–5], and both have a long history of genetic and molecular cell cycle studies. However, they diverged more than 1 billion years ago, and have many lifestyle differences.
In particular, the two yeasts have different cell cycles. S. pombe divides by fission, a symmetrical process in which a septum grows across the center of a long cylindrical cell, dividing the old cell into two equal new cells. Moreover, the main control point in the S. pombe cell cycle is a size control in G2, not in G1 as in S. cerevisiae and many other organisms. In S. pombe, when cells reach a critical size, the Cdc2 protein kinase is activated both by cyclin binding and also by Cdc25 phosphatase removal of the inhibitory phosphate from tyr15 of Cdc2, and this leads to mitosis. Once nuclear division has occurred, the cell moves quickly into S phase without an appreciable G1. Therefore S phase is largely completed by the time cytokinesis/cell separation occurs. Thus, when the cells are growing in good conditions, cells have a long G2, and most cell cycle–specific events are completed in a relatively small portion of the cell cycle encompassing M, G1, and S, with S occurring coincident with cytokinesis. When conditions are poor, a cryptic size control appears in G1 phase; that is, a G1 phase appears and becomes longer as growth rate becomes slower.
In contrast, S. cerevisiae divides by “budding,” an inherently asymmetrical process whereby a large mother cell generates a small daughter bud. Once born as a separate cell, the small daughter grows in volume through a long G1, and commits to division at a G1 event called “START.” START involves the activation of a pair of closely related transcription factors, MBF and SBF, and the induction of 100 or more genes. After START, DNA synthesis is initiated, and a bud forms. There is a short G2 phase, followed by mitosis and cytokinesis, and then cells enter the next G1. When cells are growing rapidly in good conditions, G1, S, G2, and M phases are of similar lengths, and so various cell cycle–specific events are distributed somewhat equally around the cycle. However, when cells are growing slowly in poor conditions, almost all the increased length of the cell cycle is accounted for by an increased G1, and most cell cycle–specific events occur over a relatively small percentage of the cell cycle, encompassing “START,” S phase, and mitosis.
Microarrays have been used to analyze gene expression in synchronized S. cerevisiae. There are at least 800 genes whose transcripts oscillate as a function of the cell cycle [6]. The cataloging of these transcripts has helped describe what happens in a cell cycle. In addition, because many of the oscillating genes are regulatory, the microarray analysis has helped us understand how the S. cerevisiae cycle is regulated. In view of the fact that S. pombe also has a well-studied cell cycle and because these two yeasts have both differences and similarities in the way they carry out a cell cycle, it is of interest to characterize oscillating transcripts inS. pombe also, to understand at a deeper level what is preserved and what changes across the cell cycles of these two model eukaryotes.
Recently, Rustici et al. [7] and Peng et al. [8] have published microarray analyses of S. pombe cell cycle genes. Our results are broadly similar to theirs, but as described below, each group finds a somewhat different set of genes. There is excellent agreement between the groups with respect to the most strongly regulated genes, but naturally there is less agreement for more weakly regulated genes. Here, we concentrate on the 750 genes that are most strongly regulated, but we believe that there may be a total of 2,000 or more genes that have at least weak cell cycle regulation. A large number of weakly to moderately oscillating genes peak in G2 phase, and these are highly enriched for functions in ribosome biogenesis. Our analysis of the cell cycle–regulated promoters shows them to be surprisingly complex, and shows clusters of multiple regulatory motifs similar to clusters of motifs found in the developmental genes of Drosophila. Although Rustici et al. [7] have pointed out several differences between the cell cycles of S. pombe and S. cerevisiae, we find that there are also striking similarities, suggesting deeply conserved mechanisms.
Results
Synchronous Cultures and Identification of Cell Cycle–Regulated Transcripts
Three synchronous cultures were studied, one generated by cdc25 block release, and two generated by elutriation. Each culture was sampled through three cell cycles, giving nine cell cycles of data. Synchrony and cell cycle position were assayed by scoring initiation of anaphase and septation microscopically (Figure 1). RNA was extracted, converted to cDNA, labeled, and hybridized to arrays, and then fluorescence was analyzed. Gene expression was assayed as a ratio of experimental cDNA to asynchronous control cDNA. In total, approximately 1.2 million data points were generated from cell synchrony experiments. Fourier analysis identified cyclic expression patterns. Monte Carlo simulations were used on shuffled expression-ratio data, and compared to the actual, cyclic data, to generate a p-value for the cyclicity of each gene. These p-values ranged from less than 10−16 for the most cyclic genes (i.e., the probability that the observed oscillation occurs by chance is less than 10−16), to 0.997 for the least cyclic gene of the 5,000 studied. We ranked all 5,000 genes by p-value, with the most significantly oscillating genes at the top. The amplitude of the oscillation is a major contributor to the p-value, so genes with higher amplitude oscillations tend to rank higher than genes with lower amplitudes.
Figure 1 Synchrony
(A) Samples from elutriation A were double stained with calcofluor (for septa) and DAPI (for nuclei). Cells were assayed for initiation of anaphase by scoring cells with two nuclei but no septum (binucleates, open circles). The cells were also scored for septation (filled squares).
(B) Cells from elutriation B were assayed for septation by phase contrast microscopy.
(C) Cells from the cdc25–22 block release were assayed for septation by phase contrast microscopy.
The list of all 5,000 genes ranked by p-value and other associated information such as time of peak expression is given in Table S1. The raw data have been deposited at ArrayExpress (http://www.ebi.ac.uk/arrayexpress/). The raw data, all figures and all tables, are available at: http://publications.redgreengene.com/oliva_plos_2005/.
The distribution of genes versus p-values is shown in Figure 2. There is no clear distinction between “cyclic” and “non-cyclic” genes. Rather, after the best 203 genes, there are simply more and more genes as one goes to poorer and poorer p-values.
Figure 2 Distribution of p-Values for Cell Cycle–Regulated Genes
The x-axis shows bins of p-values of the significance of cell cycle regulation. From the left, the bins are as follows: (1) Genes with p-values less than 10−16 (87 genes); (2) Genes with p-values between 10−15 and 10−16 (13 genes); (3) Genes with p-values between 10−14 and 10−15 (13 genes); (4) Genes with p-values between 10−13 and 10−14 (eight genes); etc. The number of genes in each bin is shown on the left y-axis (dark blue squares). Also shown (right y-axis, magenta diamonds) is the cumulative number of genes at each p-value or lower. Thus there are about 1,000 genes with a p-value of 10−3 or less.
Because the distribution of genes versus p-values continuously increases after gene 203, one must choose a somewhat arbitrary threshold for discussion of cell cycle–regulated genes. We have chosen to discuss the best 750 genes in our p-value list. This number is similar to the number of genes chosen by Peng et al. [8] and Rustici et al. [7] as being cell cycle regulated (747 and 407, respectively), and similar to the number of genes chosen for the yeast S. cerevisiae (800) [6], thus facilitating comparison of these gene sets. In the vicinity of the 750th gene (and even below), most genes display an oscillatory behavior to the eye, at least in one or two of the three experiments. Finally, the number 750 is obviously somewhat arbitrary, and indeed we have no basis for anything other than an arbitrary cutoff. Because the list of genes is ranked, other investigators may choose their own sets of oscillatory genes from ourp-value list (Table S1) by choosing any desired cutoff. For the top 750 genes, the false discovery rate is 0.00022, so on a statistical basis, less than one false positive is expected in the list of 750.
Although we will discuss primarily these 750 best genes, there are many more genes that appear to oscillate slightly. A total of 2,262 genes (nearly half the genes in the genome!) have a p-value less than 0.05, the usual statistical cutoff. Based on the false discovery rate, we would expect about 53 of these to be false positives, but even so, this leaves well over 2,000 genes with a slight but statistical oscillation.
Previously, 37 cell cycle–regulated genes have been reported inS. pombe; 29 of these (78%) are in our top 750. Of the eight that are not in our top 750, two are in the top 1,000. The remaining six (cdc19/mcm2, cmk1, dmf1/mid1, ppb1, uvi22/rrg1, and suc22) are also not in list of 407 of Rustici et al. [7], and three of these (cdc19/mcm2, cmk1, andppb1) are also not in the list of Peng et al. [8]. Thus, these genes are probably quite weakly regulated (except for suc22, for which there are two transcripts, one regulated and one not [9]).
The top 750 genes are shown in Figure 3 in order of time of expression (i.e., ordered by cell cycle phase). About halfway down this phasogram is an apparent discontinuity; this corresponds to the mid to late G2 trough, when there are relatively few cell cycle–regulated genes (see below).
Figure 3 Cell Cycle–Regulated Genes Ordered by Time of Peak Expression
(A) Expression data for the top 750 genes is shown, with genes ordered by time of peak expression. Every row represents a gene; every column represents an array from a time-course experiment. Red signifies up-regulation (i.e., an experiment/control ratio greater than one); green signifies down-regulation (i.e., an experiment/control ratio less than one). Black is a ratio close to one, and grey is missing data. Dynamic range is 16-fold from reddest red to greenest green. The time in hours since the beginning of the time course is shown in black numerals at the top of Figure 3. The peaks in septation index are marked with purple rectangles at the top and bottom of the figure. Genes from defined clusters are marked on the left by colored lines, according to the cluster color code shown at the bottom of the figure.
(B) As (A), but only the 514 genes found in our study but not found by Rustici et al. [7] are shown.
Rustici et al. [7] have recently compiled a list of 407 periodically-expressed S. pombe genes, and while our manuscript was in review, Peng et al. [8] identified 747 similar genes. A comparison of the three studies is shown in Figure 4. The total number of genes found to oscillate in at least one study is 1,373. Of these, 1,013 were unique to just one of the studies, whereas 360 were found in two or three studies, and 171 were found in all three.
Figure 4 Overlap between Cell Cycle Microarray Studies, by Number of Genes
A Venn diagram of the overlap between the three lists of cell cycle–regulated genes from this study, Rustici et al. [7], and Peng et al. [8]. The number of genes in each of the three lists is 750, 407, and 747, respectively. A few genes are not accounted for because of ambiguities in nomenclature.
Despite the fact that 1,013 genes were found in only one of the three studies, we believe that most of these 1,013 do indeed oscillate to some extent. There are two lines of evidence. First, most of the genes do display a clear oscillatory pattern to the eye, at least in one of the studies. For instance, Figure 3B shows the 516 genes found by us but not by Rustici et al. (63 of these were also found by Peng et al., but the remainder were unique to us). At least in part, we found these genes because our elutriated cells were more synchronous than those of Rustici et al. (compare our Figure 1 to Figure 1B in the supplemental data of Rustici et al.[7]), thus allowing detection of genes with moderate amplitudes.
The second line of evidence is that most of the 1,013 genes unique to one study also display some statistical oscillatory behavior in one or both of the other studies, even though this behavior is not strong enough to surpass the threshold for inclusion on the cell cycle list in those studies. This effect is shown in Figure 5. As might be expected, the top genes in our ranked list and the ranked list of Peng et al. [8] show excellent (approximately 85%) agreement with Rustici et al. [7]. The degree of agreement then drops as one proceeds down the ranked lists. But genes below rank 750 but above rank 2,500 in either list are much more likely to be in the list of Rustici et al. than are genes below rank 2,500. In other words, a gene unique to the study of Rustici et al. is likely to show some oscillatory behavior in the other two studies (i.e., be in the top half of the lists). Analogous comments apply to the genes unique to us, and genes unique to Peng et al. [8].
Figure 5 Overlap between Different Cell Cycle Microarray Studies, by Rank
(A) Our ranked list of cell cycle–regulated genes is divided into consecutive sets, or bins, of 50 genes. For each set of 50 genes, the number of genes in that set also found in the list of 407 cell cycle genes of Rustici et al. [7] is plotted on the left y-axis. For instance, of our best 50 genes, 44 (88%) are found in the list of of 407 genes of Rustici et al., and of our next-best 50 genes, 38 (76%) are also in their list. For the top 15 bins (750 genes), every bin of 50 genes is represented. Afterward, the number plotted represents an average over several bins. The cumulative number of genes in the list of 407 is plotted on the right y-axis.
(B) As (A), but the bins in our study are compared to the list of 747 genes of Peng et al. [8].
(C) As (A), but the ranked list of Peng et al. is divided into bins, and compared to the list of 407 of Rustici et al. Because Peng et al. ranked only their top 2,700 genes, the graph is truncated after gene 2,700, and the cumulative number of genes rises to only 325.
Before the publication of Peng et al. [8], we had compared our study to that of Rustici et al. [7] to look for discrepancies. We identified a total of 21 genes (11 from Rustici et al., ten from us) that appeared very strongly regulated in one study, but not at all regulated in the other. We have now checked these 21 genes against the results of Peng et al., and find that 17 of the 21 appear regulated in Peng et al., whereas four (three from us and one from Rustici et al.) do not appear regulated. Thus it seems that both we and Rustici et al. have been conservative in our identification of cell cycle–regulated genes and tend to get false negatives rather than false positives.
In summary, the three cell cycle lists together implicate about 1,300 genes, and our ranked p-value list does not become worse than a p-value of 0.05 until gene number 2,262. We believe that a very large number of S. pombe genes, 2,000 or more, have at least a weak cell cycle oscillation.
Two Genome-Wide Waves of Transcription
To examine the distribution of gene expression around the cycle, Fourier analysis was used to determine the time at which each gene's expression peaked (the “phase angle” of peak expression). For genes in the bottom half of the 5,000 gene rank list (i.e., genes that did not cycle appreciably), phase angles were largely determined by noise, but nevertheless would tend toward the peak of any weak cyclic behavior that may have existed. The number of genes peaking at each time in the cycle was plotted (Figure 6) for four groups of genes: the most-regulated 750 genes (Figure 6A), all genes (Figure 6B), the least regulated 4,000 genes (Figure 6C), and all genes after random shuffling of ratio data (Figure 6D). The peaks of septation and binucleates were also determined by Fourier analysis. (See Materials and Methods for information on red/green normalization.)
Figure 6 The Number of Genes Peaking during Each Portion of the Cell Cycle
The cell cycle was divided into 45 consecutive portions, or bins. If the cell cycle is considered as a circle of 360°, then each bin occupies 8°. Every gene was analyzed using a Fourier transform to determine the time of peak expression in elutriation A, from 0° to 360°. The number of genes peaking in each bin was summed and plotted (grey bars, background), with the number of genes in each bin shown on the y-axis. Genes in specific clusters are shown by colored bars (foreground). The Fourier transform calculation was similarly used to derive the time of peak septation and peak binucleate cells (from Figure 1) and these cell cycle landmarks are indicated.
(A) The top 750 cell cycle–regulated genes were analyzed for time of peak expression. Genes from different clusters are “stacked” when they occur in the same bin.
(B) All genes (∼5,000) were analyzed for time of peak expression, exactly as in (A).
(C) The bottom 4,000 genes were analyzed for time of peak expression. Data was extracted from arrays before the red/green normalization step, so that the bottom 4,000 genes would not be affected by the normalization and the cyclic expression of the top 750 genes.
(D) All genes (∼5,000) were analyzed for time of peak expression after genewise random shuffling of microarray observations. This randomization serves as a negative control for the Fourier calculation in parts (A), (B), and (C).
There were two striking findings. First, it appears that there are two broad waves of gene expression, one peaking in early to mid G2, and the second peaking in late G2/M, whereas there are troughs in mid to late G2, and in S. The early/mid G2 peak contains the Ribosome biogenesis cluster (see below) and associated genes, whereas the late G2/M peak contains the genes of the Cdc15, Cdc18, and Eng1 clusters (see below), which are important for M and S. Second, the two waves of gene expression were seen even in the 4,000 least-cyclic genes. As noted above, there is statistical evidence from p-values that 2,000 or more genes may oscillate slightly. The two waves of expression seen for the bottom 4,000 genes confirm that many of these genes do indeed oscillate. If the fluctuations in these 4,000 genes had simply been due to noise, then the peak phase angles would have been uniformly distributed from 0° to 360° (as confirmed by repeating the analysis on shuffled data; Figure 6D). Combined with the evidence of the p-values, this analysis suggests that many, or most (or possibly all!) of the least-cyclic 4,000 genes do in fact oscillate slightly, and that there are two nearly genome-wide peaks in gene expression. These peaks might represent periods when the cell is preparing for a high level of cell cycle–specific activity, or when transcription (of any kind) is activated on a genome-wide basis. Alternatively, one might focus on the troughs, which might represent periods with little cell cycle–specific activity, or periods when transcription is repressed on a genome-wide basis (see Discussion).
Cluster Analysis
To study the regulation of the cell cycle, we wished to find clusters of co-regulated genes potentially responding to the same transcription factor. However for this purpose it is not sufficient to find genes expressed at the same time, because such genes might be responding to different mechanisms of regulation. This is an acute problem in S. pombe, because mitosis, DNA synthesis, and cytokinesis all occur in a small window of the cell cycle under standard growth conditions.
Therefore our analysis included not only our three time courses of synchronous cells, but also eleven other array experiments that more directly addressed regulatory mechanisms. These experiments (see Materials and Methods) included small cells grown in poor nitrogen to induce a G1 phase; a cdc10-M17 block-release experiment, to separate S phase events from cytokinesis and septation events; an arrest at G1 (using cdc10-M17, encoding MBF transcription factor subunit); an arrest at S (using cdc22-M45, encoding ribonucleotide reductase); an arrest at late G2 (using cdc25–22, encoding the phosphatase that activates Cdc2); an arrest at M (using nuc2–663, encoding a subunit of the anaphase promoting complex); and finally, from the data of Rustici et al. [7], experiments using a constitutively active allele of cdc10
(cdc10-c4), null and over-expressor alleles of the forkhead transcription factor sep1, and null and overexpresser alleles of the transcription factor ace2.
Hierarchical clustering was used [10] because the underlying structure of a gene regulatory network is somewhat hierarchical. Thus, a hierarchy found by the clustering algorithm is often interpretable in terms of a hierarchical transcriptional network existing in the cell (see S. J. Gould's essay [11], “Linnaeus's Luck?”, for an illuminating discussion of this issue in a different context http://www.findarticles.com/p/articles/mi_m1134/is_7_109/ai_65132190).
The clustergram of 750 genes is shown in Figure 7. (Treeview files are available as Dataset File S1). We chose eight clusters for analysis and discussion, on the basis that the genes in these clusters are particularly tightly co-regulated. Most of the clusters are named for one representative gene. The clusters are, from the late G2/M wave, the Cdc15, Cdc18, and Eng1 clusters; from S and early G2, the telomeric, histone, and Wos2 clusters; and from the early to mid G2 wave, the Ribosome biogenesis and Cdc2 clusters.
Figure 7 Cluster Analysis of the Top 750 Cell Cycle–Regulated Genes
Gene expression data from all experiments were clustered by a hierarchical method (Eisen et al. [10]). Every row represents a gene; every column represents an array. Red signifies up-regulation (i.e., an experiment/control ratio greater than one); green signifies down-regulation (i.e., an experiment/control ratio less than one). Black is a ratio close to one, and grey is missing data. Dynamic range is 16-fold from reddest red to greenest green. The time in hours since the beginning of time-course experiments is shown in black numerals at the top of the Figure. Peaks in septation index are marked with purple rectangles at the top and bottom of the Figure. Clusters discussed in the text are marked with blocks of color. Data for the cdc10-C4 (asynchronous cells with the hyperactive allele cdc10-C4), ace2 OE (asynchronous cells over-expressing ace2), ace2Δ (asynchronous ace2Δ cells), sep1 OE (asynchronous cells over-expressing sep1), and sep1Δ (asynchronous sep1Δ cells) are taken from Rustici et al. [7]. cdc10 encodes a component of the MBF transcription factor; ace2 encodes the Ace2 transcription factor, and sep1 encodes a forkhead transcription factor. Other experiments are described in Materials and Methods.
If the genes in each cluster are truly co-regulated, then the promoters of these genes will be bound by the same transcription factor, and therefore the promoters should share a common DNA sequence motif corresponding to the transcription factor binding site. We searched for such motifs upstream of the genes in each cluster. We used three motif search programs: AlignAce, a Gibbs-sampling algorithm [12]; SPEXS, a word-count algorithm (http://www.egeen.ee/u/vilo/SPEXS/) [13,14], and MEME, an expectation-maximization algorithm (http://meme.sdsc.edu/meme/website/intro.html) [15,16]. In general, all three programs found the same motifs.
In the study of Rustici et al. [7], four clusters were found. It is difficult to compare the clusters of Rustici et al. with ours: The genes, experiments and clustering methods were different. However, in general, the clustering of Rustici et al. tended to produce fewer, larger clusters, and focused on time of expression as the main distinction between the clusters, whereas our method produced more, smaller clusters, and focused on regulatory mechanisms (as well as time of expression). Peng et al. [8], like us, used hierarchical clustering and found eight clusters, some of which are quite comparable to ours. However, again, we put more emphasis on regulatory mechanisms as opposed to time of expression, and this generated some different clusters.
The M Clusters
The wave of expression in the late G2 and M phases includes most of the strongly regulated genes. This wave contains three major clusters, which we call the Cdc15, Cdc18, and Eng1 clusters (Figure 8). Functionally, these clusters are important for mitosis and cell separation, DNA synthesis, and cytokinesis, respectively. Genes of the Cdc15 and Cdc18 clusters peak almost simultaneously with anaphase (see Figures 6 and S1), whereas the Eng1 genes peak slightly later.
Figure 8 M Phase Clusters
Clusters of apparently co-regulated genes were chosen from Figure 7. See legend to Figure 7 for further information.
(A) Cdc15 cluster, (B) Cdc18 cluster, and (C) Eng1 cluster.
The Cdc15 cluster (Figure 8A) is the largest of the three clusters and contains over 100 genes. These are involved in mitosis and mitotic exit, cytokinesis and septation, vesicle trafficking, cell wall remodeling, and other functions. Genes involved in mitosis and mitotic exit include the APC adaptor subunits srw1 and slt1, the prolyl isomerase pin1, spo12, the Cdk inhibitor rum1, five genes related to ubiquitination, four microtubule-related genes including kinesins klp5 and klp6, and four genes for chromosome segregation.
Cytokinesis/septation fuctions can be ascribed to at least 13 genes including the key SH3 domain gene cdc15 and its paralog imp2, and a third SH3 domain gene, pob1. Also present are the kinases fin1 and sid2 and phosphatase subunit par2, which regulate the septation initiation network. mob1, which interacts with sid2, is also cell cycle regulated with similar timing, but lies outside the cluster as defined here. Other members likely involved in cytokinesis include genes for the rho family member rho4, the putative rhoGEF rgf3, the septin spn2, and the myosin myo3.
Construction of the septum involves synthesis of plasma membrane and deposition of proteins into that membrane. The Cdc15 cluster is rich in proteins involved in these processes. The cluster includes gwt1, likely involved in GPI anchor synthesis, and SPAP27G11.01, SPCC306.05c, and SPBC2F12.05c, linked with sterol functions. SPAC227.06 (a predicted Rab interactor), psy1 and bet1 (SNAREs), and SPBC31F10.16 are likely to function in vesicle transport. The budding yeast homolog of SPBC31F10.16, CHS6, is important for movement of chitin synthase from the trans-Golgi network/endosome to the plasma membrane. Other genes encode cell surface glycoproteins, such as the gene mac1, which is localized at poles and septum and is important for cell separation.
Genes for cell wall metabolism include two chitin synthase homologs, a putative chitin synthase regulator, six putative sugar/starch hydrolases, and the MAP kinase pmk1.
Finally, diverse other functions are represented. There are at least five genes involved in transcription, most notably the transcription factor fkh2, which may be one of the regulators of the Cdc15 cluster [17] (see below). There are also multiple genes involved in mitochondrial functions and in glycosylation.
The three motif search programs all found the consensus motif
GTAAACAAA, easily recognizable as a binding site for a forkhead (FKH) transcription factor. Almost every gene in the cluster had such a motif. In S. cerevisiae, the main clusters of mitotic genes are also regulated (in part) by forkhead transcription factors. S. pombe has several forkhead transcription factors, but the two most likely to regulate the Cdc15 cluster are sep1 and/or fkh2. sep1 does not oscillate noticeably in our dataset, but it does have phenotypes that could be due to defects in the expression of genes of the Cdc15 cluster, and Rustici et al. [7] have shown defects in cell cycle expression in sep1 mutants. fkh2 does oscillate, and is a member of the Cdc15 cluster. The fkh2 promoter contains two sites each for Forkhead, Ace2, and Cdc10. Interestingly, peak expression of fkh2 precedes the peak of 94% of the other genes in the cluster, consistent with the idea that it might help regulate these other genes. No direct binding of either Sep1 or Fkh2 to any of these promoters has been demonstrated, and we believe it is still an open question which protein regulates this cluster. It is possible that both proteins contribute. Because forkhead transcription factors can both repress and activate, and because they are regulated both transcriptionally and post-transcriptionally, the regulatory mechanisms could be complex.
The motif search programs also found
CCAGCC (Ace2 binding sites) and
ACGCG (MBF/Cdc10 binding sites) in a substantial minority of the genes of the Cdc15 cluster. Many genes (e.g., fkh2 and pds5) had all three kinds of sites. MEME (but not the other programs) also found the motif (A/T)
TGACAAC. This is probably the same as the motif
CATG(A/T)
CAAC found by Rustici et al. [7] and named “New 1.” To minimize confusion, we will refer to our version of the motif as “New1v” (“v” for variant).
MEME also found the motif CC(T/A)CG(T/C)TCC, and this may be a variant of the motif (A/T)ACC(T/A)CGC(T/A) (“New 3”) found by Rustici et al. We will refer to our motif as “New 3v.” New 3v was found preferentially in front of genes for cell wall metabolism, such as hydrolases, glycoproteins, chitin synthases, and their regulators. Other functionally related genes are found in the Eng1 cluster (see below), where they appear to be regulated by Ace2. Interestingly, the consensus site for Ace2 (
CCAGCC) is reminiscent of the core of New 3v (
CCACGC), suggesting that an unknown Ace2-like factor could be involved.
We did not find the “PCB” consensus (
GCAAC(G/A)), previously implicated in the control of some of the genes of this cluster [18,19].
The Cdc18 cluster (Figure 8B) contains 18 genes involved in DNA replication. Included in the cluster arecdc18 (initiation of DNA synthesis),pol1 (DNA polymerase alpha), cdt1 (initiation of DNA synthesis),cig2 (S phase cyclin), mrc1 (S phase checkpoint), cdc22 (ribonucleotide reductase), cdt2 (DNA replication), smc3 (cohesin), and pif1 (DNA helicase). These genes are strongly regulated. Peak expression occurs at about the same time as that of the Cdc15 cluster, and is essentially simultaneous with the peak in binucleates (e.g., Figure S1).
The Cdc18 cluster has a very similar cluster in S. cerevisiae, called the CLN2 cluster. Both clusters contain genes involved in DNA replication, and both clusters appear to be regulated by the MBF transcription factor (see below). For the Cdc18 (pombe) and CLN2 (cerevisiae) clusters, many of the genes in the clusters are orthologs; e.g., mik1/SWE1, cig2/CLB5, mrc1/MRC1, cdc22/RNR1, andsmc3/SMC3. Thus the cell cycle clusters regulating DNA synthesis are very highly conserved, with the overall function of the clusters, the regulation of the clusters, and the genes in the clusters, all being quite similar from S. cerevisiae to S. pombe.
The three motif search programs found two motifs in the Cdc18 cluster:
ACGCG, and
ACGCG(A/T)
CGCG. The first of these is easily recognizable as the binding site for the MBF transcription factor (also known as DSC1) [20–22], whereas the second is a related motif that may be a tandem, double binding site for MBF, or for an MBF-like factor. Consistent with the idea that MBF is a major regulator of this cluster, the genes of the cluster are up-regulated by the cdc10-c4 mutation (see Figure 8B, and see Rustici et al. [7]) which creates a constitutively active form of MBF. Furthermore, six of these genes are known to be regulated by MBF (cig2, cdt1, cdt2, cdc18, cdc22, and mik1; GeneDB, Sanger Centre).
S. cerevisiae has two MBF-like transcription factors. One is itself called MBF and consists of the DNA-binding protein Mbp1 complexed with the modulatory protein Swi6. The second factor is called SBF and consists of a second DNA-binding protein, Swi4, complexed with Swi6. S. cerevisiae MBF and SBF, with their related but distinct DNA-binding proteins, bind to related but distinct motifs, and control the cell cycle expression of partially overlapping sets of genes [23,24]. In S. pombe, there is likewise one modulatory protein, Cdc10 (the ortholog of Swi6) and two DNA-binding proteins, Res1 and Res2 (possible orthologs of Mbp1 and Swi4) [20–22,25–27]. Some investigators believe that in S. pombe, there is a unique MBF transcription factor and that it contains Cdc10, Res1, and Res2 [25,26,28]. However, other investigators believe that the situation is similar to that found in S. cerevisiae and that there may be two MBF-like factors, one containing Cdc10 and Res1, and the other containing Cdc10 and Res2 [27,29]. Although our results do not speak directly to these models, the fact that we find two kinds of motifs is easier to interpret in terms of a model with two different but related forms of MBF.
The Eng1 cluster (Figure 8C) contains nine genes, and these are involved in cell separation. The genes are adg1 and adg2 (cell surface glycoproteins), adg3 (β-glucosidase), agn1 and eng1 (glycosyl hydrolases), cfh4 (chitin synthase regulatory factor), mid2 (an anillin needed for cell division and septin organization), ace2 (a cell cycle transcription factor), and SPCC306.11, a sequence orphan of unknown function. The genes are very strongly cell cycle regulated. Peak expression of most of the genes occurs slightly later than the genes of the Cdc15 and Cdc18 clusters. Motif searches showed that each gene of the cluster has at least one binding site for the Ace2 transcription factor (consensus
CCAGCC). In fact, eight of the nine gen-+-es contain multiple Ace2 binding sites. The exception is the ace2 gene itself, which contains only one Ace2 binding site, but multiple FKH binding sites. Interestingly, the ace2 gene is expressed earlier than the other genes of the cluster, consistent with the idea that it might regulate the other genes. The genes are up-regulated whenace2 is over-expressed, and down-regulated when ace2 is deleted (see Figure 8C; Rustici et al. [7]). Ace2 was previously shown to be a regulator of eng1 [30] and agn1 [31].
The Eng1 cluster has a recognizably similar functional cluster in S. cerevisiae, the SIC1 cluster [6]. This cluster also has many genes involved in cell separation (e.g., EGT2, an endoglucanase; CTS1, an endochitinase; YGL028c, a glucanase; DSE2, a glucanase; and CHS1, a chitin synthase), and the genes of the S. cerevisiae cluster are also regulated from Ace2 binding sites of the same consensus sequence (
CCAGC). However, there is only one gene that is clearly present in the cluster in both species, the glycosyl hydrolase eng1 in S. pombe, and its ortholog DSE4 in S. cerevisiae. Thus the overall function of the cluster (cell separation), the nature of many of the enzymes in the cluster (carbohydrate hydrolytic), and the mechanism of gene regulation (binding by Ace2) have been conserved, even though the individual genes in the cluster have been largely shuffled. It is easy to understand why the individual genes are different, because the two species have cell walls containing different carbohydrates (and so requiring different hydrolytic enzymes), and because the modes of cell separation are very different (fission vs. budding). In fact, given these differences, it is remarkable that the mode of regulation and the functional cluster seem to have been conserved.
The S/Early G2 Clusters
The relatively few genes that peak in late M, S, or early G2 fall into three small clusters: the telomeric cluster, the histone cluster, and the Wos2 cluster (Figure 9).
Figure 9 S/Early G2 Clusters
(A) Telomeric cluster, (B) Histone cluster, and (C) Wos2 cluster.
The telomeric cluster (Figure 9A) contains eight tightly clustered genes found near telomeres. Peak expression is in early S. Two of the genes are at telomere 1L; two at 1R; two at 2L, and two at 2R. Interestingly, S. cerevisiae also has a cluster containing only telomeric genes (the Y′ cluster), and the genes of that cluster also peak in late G1 or early S.
The histone cluster (Figure 9B) contains all nine histones of S. pombe. These are tightly co-regulated and strongly periodic, and form a very tight cluster. Presumably, peak expression of the histone genes marks the time of S phase. These genes are expressed about 30 min after the DNA synthesis genes of the Cdc18 cluster. Surprisingly, the histone cluster contains two non-histone genes, SPAC977.07c and SPAC1384.08c. These two genes are near telomeres and are homologs of each other, but have no known function. Possibly they are actually co-regulated with the genes of the telomeric cluster, which peak just before the histone cluster.
Motif searches showed that all the histone genes (but not the two telomeric genes) had the motif
GGGTTAGGGTT(T/G). A degenerate second copy was sometimes also present. This motif has been noted previously [32]. In addition, six of the histone genes (and both telomeric genes) had a motif similar to an MBF binding site, G(C/G)(T/G)
ACGCG.
In S. cerevisiae, the histone genes have at least three semi-redundant regulatory systems: First, they have the HIR gene system that represses histone expression outside of S [33,34]. Second, they have regulated mRNA stability, such that the messages are only stable during S [35]. Third, they have a system for gene induction during S. Recently, it has been suggested that this positive system relies on the SBF transcription factor, possibly in combination with a forkhead transcription factor [36]. The fact that an MBF motif is found in front of most of the S. pombe histone genes is consistent with the SBF motif found in front of most of the S. cerevisiae histones, and suggests that MBF may play a role, along with other mechanisms, in regulating histone expression in S. pombe.
The Wos2 cluster (Figure 9C) contains seven genes expressed in late S or very early G2. Expression of the genes in the cluster responds strongly to the two experiments that involve temperature shifts (cdc25–22 synchrony and cdc10-M17 block-release; note that control cDNA for simple cell cycle–arrest experiments was made from wild-type cells similarly shifted to high temperature). Motif searches found repeats of the sequence NGAAN, a typical heat shock response element. The cluster contains wos2, encoding a chaperone activator interacting with Hsp90; SPACUNK4.16c, important in trehalose synthesis (trehalose is a thermo-protectant); SPBC16D10.08c, encoding a chaperone similar to S. cerevisiae Hsp104; and SPBC4F6.17c, similar to S. cerevisiae
Hsp78, a mitochondrial chaperone.
The early to mid G2 genes: The Ribosome biogenesis and Cdc2 clusters
Although most of the strongly regulated genes peak near the G2/M transition, another large group of genes, 200 or more, peaks with a moderate amplitude at almost exactly the opposite side of the cell cycle, in early to mid G2 (Fig. 10). The expression of these genes does not respond to mutations in cdc10, ace2, or forkhead, and they are all strongly repressed at the nuc2 block. Near the center of this set of 200-plus genes is a sub-cluster of genes that is somewhat more tightly co-regulated than the rest. We have designated these the “Ribosomal biogenesis” cluster (Figure 10A). These genes includeSPAC1527.03 (RNA-binding protein, LA-related), SPAC57A7.06 (processome component, involved in rRNA processing), SPBC13G1.09 (bystin family protein, associated with U3 and U14 snoRNAs, involved in rRNA processing); SPCC16A11.02 (WD-repeat protein, processome component, involved in rRNA processing);SPAC23C4.17 (tRNA methyltransferase of the NOL1/NOP2/sun family involved in methylation of cytidine to 5-methyl-cytidine [m5C] at several positions in different tRNAs);SPBC11G11.03 (60S acidic ribosomal protein); rpl2403 (60S ribosomal protein L24–3); ker1 (interacts with RNA polymerase I); SPAC1093.05 (DEAD/DEAH box RNA helicase involved in rRNA processing); SPAC926.08c (Brix domain RNA-binding protein involved in ribosome biogenesis and assembly); and many others.
Figure 10 Early to Mid G2 Clusters
(A) Ribosome biogenesis cluster and (B) Cdc2 cluster.
Other genes in the ribosome biogenesis cluster are involved in nuclear/cytoplasmic import and export. These genes include: nup61 (nucleoporin with a RanBp-binding domain), kap123 (karyopherin), SPCC550.11 (RanBP7/importin-beta/Cse1p family, RanGTP-binding protein involved in mRNA export), and mep33 (mRNA export protein).
It is not clear why such genes would be cell cycle regulated. However, Mitchison and colleagues [37–41] have documented a cell cycle oscillation in the rate of growth and protein synthesis in S. pombe. In these studies, there seems to be an acceleration of protein synthesis, and a corresponding acceleration in cell growth rate, in mid G2. Furthermore, “NETO” (new end take off, the time when the new end begins to grow) occurs at about this time. The peak in expression of ribosome biogenesis genes we observe in early/mid G2 could lead to this slightly later peak of protein synthesis and growth rate. Sveiczer et al. [41] suggest that the acceleration in protein synthesis is the “sizer” that leads to commitment to division; in terms of our findings, the peak in transcription of the ribosome biosynthesis genes would be an important component of the sizer.
We have recently found that many S. cerevisiae ribosome biogenesis genes are also cell cycle regulated (Figure 11). Expression peaks in G1, and so this peak could be important for the cell sizer, which in S. cerevisiae is in late G1. These genes also show a minor expression peak in early G2. The oscillation of these genes is seen in an elutriation experiment done in ethanol medium, but not in block-release experiments done in glucose medium. The reason for these different, experiment-specific results is unclear, but on the basis of the literature we believe that the oscillation may be under the control of cyclic AMP, and this cyclic AMP signaling does not occur in media with high glucose [42,43].
Figure 11 Oscillation of Ribosome Biogenesis Genes in S. cerevisiae
One cell cycle of elutriation data is shown for 52 S. cerevisiae genes involved in ribosome biogenesis. The genes chosen for analysis were those listed [42,75] as involved in ribosome biogenesis. At the top of Figure 11 are three histone genes (HTA2, HHF1, and HHT10) known to peak in S, and three genes (CLN1, CLN2, and MCD1) known to peak in late G1. The raw data for this figure are taken from Spellman et al. [6]. In their experiment, cells were grown in ethanol medium, and then small G1 cells were isolated by elutriation and re-inoculated into ethanol medium. Samples were taken at intervals from 0 to 390 min, the duration of one cell cycle under these conditions.
Surprisingly, we found no DNA sequence motifs associated with the promoters of the genes in the ribosomal biogenesis cluster.
As one moves out from the center of the ribosome biogenesis cluster, one encounters many other genes peaking in G2 phase. These are of diverse function, but one interesting example is the pma1 gene, which encodes a proton pump. This pump is needed to maintain the proton gradient across the plasma membrane, affecting many processes, and so seems an unlikely candidate for a cell cycle–regulated gene. Nevertheless, it is cell cycle regulated both here and in S. cerevisiae [6]. The reason for the oscillation is unclear, but because Pma1 is an integral plasma membrane protein that must be inserted into the membrane at the time of synthesis, one possibility is that its synthesis matches the rate of plasma membrane production; in S. cerevisiae, this may reach a peak in G2, accounting for the peak in PMA1 transcription. A similar explanation could hold true in S. pombe.
Adjacent to the Ribosome biogenesis cluster is a cluster of 23 genes we call the Cdc2 cluster (see Figure 10B). Like the ribosome biogenesis genes, these oscillate moderately with a peak in G2. Their regulation is distinguished from the ribosome biogenesis genes by the fact that they are differentially regulated after heat stress. Motif search programs found heat shock motifs (NGAAN) associated with many of these genes. The Cdc2 cluster contains several interesting cell cycle genes, including cdc2 itself; SPBC1861.01c and abp2, which code for AT hook proteins thought to bind centromeric DNA and ARS DNA, respectively;res1, a key component of the MBF transcription factor; sds22, a protein phosphatase regulatory subunit known to be involved in the cell cycle; ash2, a member of the SET1 complex, and involved in lysine methylation of histone H3; alp1, a tubulin-specific chaperone; SPCC18.03, a putative transcriptional regulator; pkl1, a kinesin-like protein of the Kar3 family; and other genes. Other than the heat shock elements, no statistically significant motifs were associated with the promoters of these genes.
Characterization of Cell Cycle–Regulated Promoters
Each cluster was searched for DNA sequence motifs. The most significant motifs are summarized in Table 1. However, the presence of these motifs in the upstream regions of the genes of a cluster says little about promoter structure. To investigate promoter structure in more detail, we used a program called SpikeChart (S. Pyne, B. Futcher, and S. Skiena, unpublished data) that finds and displays motifs in DNA sequences. SpikeChart uses a weight matrix to define a consensus motif, and it shows each occurrence of a motif as a spike of varying height depending on that motif's match to the consensus. For instance, a motif that matches the consensus motif exactly would be given a spike height of ten, whereas a motif with one or more mismatches to the consensus would be given a lower score, depending on the number and nature of the mismatches. (Weight matrices and scoring functions are shown in Table S2). SpikeChart can score many different kinds of motifs simultaneously, and can show the position of all scored motifs, so it is well suited to finding groups of motifs, whether they be of the same kind or different kinds. Initially, because we did not know where regulatory motifs might occur, SpikeChart was used to examine the first 200 base pairs (bp) of the open reading frame in question, and 2,000 bp upstream of the start codon (regardless of whether this region included the next open reading frame or not).
Table 1 Clusters and Motifs
Groups of closely spaced, multiple motifs were usually visible, and these groups usually occurred in the upstream intergenic region (as opposed to within the open reading frame) (Figure 12). These groups of motifs were striking for their complexity. There were often four to ten motifs per group, and often the motifs were of several different kinds. The groups of motifs usually occupied about 400 bp of DNA. SpikeChart confirmed that the Cdc15 cluster was dominated by FKH motifs; the Cdc18 cluster was dominated by MBF and DBL10 motifs; and the Eng1 cluster was dominated by Ace2 motifs. However, SpikeChart showed that in addition to the predominant motif, the genes of all these clusters often had other motifs as well. In particular, for M phase genes, it was very common to have at least one FKH motif and at least one other kind of motif. There was a weak to moderate correlation between the number of motifs upstream of a gene and the amplitude of that gene's cell cycle oscillation (data not shown).
Figure 12 Distribution of Promoter Motifs
A total of 23 genes from the core of the Cdc15 cluster (cdc15), the 18 genes from the Cdc18 cluster (cdc18), the nine genes from the Eng1 cluster, plus one similarly regulated gene (SPBC83.18c; eng1) and 15 randomly chosen genes (Random) had their promoters examined for six sequence motifs using SpikeChart. For each gene, the DNA sequence examined was the 2,000 bp immediately upstream of the Start codon; the Start codon is at the right edge of the figure, and the upstream 2,000 bp extend to the left. The beginning of the next upstream open reading frame is indicated by a triangle; for instance, for the pof3 gene (Cdc18 cluster), the next open reading frame begins about 700 bp upstream of the pof3 Start codon. For the Cdc18 cluster gene ams2, all 2,000 bp are intergenic.
Consensus motifs are as follows: Dark Blue Fkh motif,
TGTAAACAAA; Purple Ace2 motif,
ACCAGCCT; Green MBF motif,
GACGCGTC; Black Dbl10 motif,
ACGCGACGCG; Light Blue/Aquamarine New 1v motif,
TGACAAC; Yellow New 3v motif, (A/T)ACC(A/T)CG(T/C)(A/T)(C/A)C.
Taller spikes indicate a better match to the consensus; the weight matrices, and the rules governing spike height, are given in Table S2. Dbl10 spikes (black) are obscured by overlapping MBF spikes (green), and so are hard to see. Only tall spikes (i.e., good matches to the consensus) are shown, so an acceptable motif may exist even when no spike is shown,
We did not notice any cases where the group of regulatory motifs was inside an open reading frame (either the downstream or upstream open reading frame).
In long (>1 kb) intergenic regions, the group of motifs usually occurred within 800 bp of the start codon, but this was not always true; a substantial minority of regulatory motif clusters occurred more than 800 bp upstream (but still within the intergenic region). Because the median S. pombe intergenic region is only 900 bp, we wondered whether the cell cycle genes might have unusually long promoters. We measured the length of upstream intergenic regions versus cell cycle rank in our list of all 5,000 genes. The most strongly regulated 200 genes had upstream intergenic regions of about 1,200-bp median length, versus a genome-wide median length of 900 bp. Thus, the more strongly cell cycle–regulated genes have longer than average upstream regions. We have noticed the same phenomenon with the cell cycle regulated genes of S. cerevisiae (S. Pyne, S. Skiena, and B. Futcher, unpublished data). The longer-than-average promoters found for cell cycle–regulated genes suggests that these promoters might be above average in complexity.
Discussion
How Many Cell Cycle–Regulated Genes Are There?
We have ranked S. pombe genes by the statistical significance of their oscillation, and we have discussed the most cyclic 750 genes. However, p-values (see Table S1) and other evidence (see Figure 6) suggest there are at least 2,000 genes with weak oscillations. This number fits well with the combined results of our study and the studies of Rustici et al. [7] and Peng et al. [8]. The three cell cycle lists of 750, 407, and 747 genes, respectively, implicate a total of 1,373 genes. Each study has uniquely identified some genes, but in general these are not just errors, because the vast majority of the uniquely identified genes show some cyclicity in one or both of the other studies even though they do not rise above the cutoff in those studies. Thus we feel the three groups of investigators are each fishing 400 to 750 genes out of a pool of about 2,000 detectably oscillating genes. The three groups are in excellent agreement with respect to the strongly oscillating genes, but then diverge with respect to the more weakly oscillating ones (see Figure 5).
However, at the same time, it seems unlikely that 2,000 genes would be directly involved in the cell cycle. There might be at least two kinds of reasons for the observed oscillations. First, an oscillation might be adaptive; i.e., there might be natural selection in favor of the oscillation. The DNA synthesis genes (e.g., cdc18, pol1, and cdc22) in the Cdc18 cluster are examples of genes in which it is easy to believe that the oscillation is adaptive. But second, some oscillations may be incidental. That is, there might be no selective advantage whatsoever to the oscillation, but instead the oscillation is a secondary or indirect effect. For example, chromatin condenses during mitosis. At least in multicellular eukaryotes, mitosis is associated with genome-wide repression of transcription. If there is a similar loss of transcription during mitosis in S. pombe, and if our microarray experiments are sufficiently sensitive, we will detect this decreased transcription as a cell cycle oscillation with a trough in mitosis for essentially all genes (preferentially the genes with a short mRNA half-life). But this cell cycle oscillation, though real, does not imply that the oscillation of any of these genes is beneficial; instead, it is a secondary consequence of mitotic repression and chromatin condensation, which presumably is beneficial. Incidental oscillation might also arise when two genes are adjacent to each other. One of the genes might oscillate for adaptive reasons, but the oscillation of this gene might carry over to adjacent genes, for which natural selection is perhaps indifferent to oscillation.
How can we distinguish adaptive from incidental oscillation? First, adaptive oscillations are likely to be large-amplitude oscillations, whereas incidental oscillations are likely to be small-amplitude oscillations. Our cutoff at 750 genes is a crude first screen to enrich for genes with adaptive oscillation. Second, one should consider the total oscillation of the gene's final activity. That is, the oscillation of a gene's transcript might be small. But if one finds that the same gene also has an oscillation in protein stability (e.g., because of regulated proteolysis), and also an oscillation in enzyme activity (e.g., because of phosphorylation), this suggests that the oscillation is adaptively significant. For example, in S. pombe, the cyclin transcripts oscillate only modestly, and yet the oscillation of the final product (Cdc2 protein kinase activity) is large. The modest oscillation of the transcript contributes in a significant, multiplicative way to the overall oscillation, and is undoubtedly adaptive. Third, one should consider co-regulated genes and the mode of regulation. If a gene is a member of a small cluster of genes, and the genes have related functions and are regulated by a specific cell cycle transcription factor, then the oscillation is almost certainly adaptive. But if the gene is co-regulated with hundreds of other genes all with very small oscillations, and there is no common function to the genes and no known cell cycle transcription factor, then the oscillation of the whole set of genes may be secondary to some effect such as chromatin condensation. Fourth, one should consider the chromosomal location. Genes adjacent to adaptively regulated genes could oscillate passively. In particular, genes in regions of special chromatin structures (e.g., near telomeres, centromeres, and silenced regions) could oscillate as a secondary consequence of cell cycle changes in the special chromatin structure.
In summary, we feel that a very large number of S. pombe genes, 2,000 or more, have at least very small cell cycle oscillations. But it is possible that in many cases this oscillation may be incidental and that only a smaller but unknown number oscillate for adaptive reasons. Sorting adaptive from incidental oscillations will require additional experiments.
Two Genome-Wide Waves of Transcription
There were two large waves of transcription, one peaking in early/mid G2, and the other peaking in late G2 or M (see Figure 6). The early/mid G2 wave contains hundreds of genes, including many genes for ribosome biogenesis. Interestingly, Mitchison and co-workers [41] have documented a cell cycle oscillation in protein synthesis, which peaks in mid G2, and may help trigger commitment to cell division. We believe that the early/mid G2 peak in ribosome biogenesis genes may lead to this slightly later peak in protein synthesis.
One property of these early/mid G2 genes is that they are deeply repressed at the nuc2 block in mitosis (see Figure 10). This is reminiscent of mitotic repression, a phenomenon observed in multicellular eukaryotes in which the majority of transcription (Pol I, Pol II, and Pol III) is repressed during mitosis [44–52]. Repression is especially well established for Pol I and Pol III polymerases, which are needed for transcription of ribosomal RNA and other RNAs required for protein synthesis. It is thought that highly active transcription may interfere with chromosome condensation, and so transcription is repressed to allow condensation.
A related observation is that in the 1970s and 1980s, metabolic labeling studies were done on synchronized cultures of S. pombe. These studies found “steps” of incorporation of labeled uridine into RNA (mostly ribosomal RNA) as a function of cell cycle phase. Around mitosis, incorporation was poor, then after mitosis, the rate of incorporation increased, and then flattened out again at the next mitosis, then increased, and so on. The interpretations of this step-like, cell cycle–regulated uridine incorporation were varied, and the subject disappeared from the literature without resolution [53–56].
Putting these observations together, we speculate that S. pombe, too, may have some degree of mitotic repression, perhaps important for chromosome condensation. Pol I accounts for the vast majority of the transcription in the cell. Mitotic repression of Pol I transcription of the ribosomal RNA genes would account for the pause in uridine incorporation seen in mitosis in the metabolic labeling studies. But if ribosomal RNA is not transcribed in M, and given that the components of the ribosome are tightly coordinated in their production, then genes for ribosomal proteins (as seen by Peng et al. [8]), and genes for ribosome biogenesis, might also be repressed in M. Repression in M would account for the oscillation of the ribosome biogenesis cluster and its repression at a nuc2 arrest. Finally, if the ribosome biogenesis genes cluster together because they are subject to mitotic repression, this might explain why the cluster does not contain any characteristic 5′ motifs: Mitotic repression might not work through a particular upstream site-specific transcription factor. Indeed, in S. cerevisiae, ribosome biogenesis transcripts are controlled in part at the level of mRNA stability [57]. Thus, we suggest that S. pombe may have a form of mitotic repression and that this repression in mitosis may account for the oscillation of the ribosome biogenesis genes and other genes peaking in early/mid G2 phase and troughing in M.
The second large wave of gene expression peaks in late G2 and in M. This wave includes the Cdc15 cluster (which has many genes for mitosis), the Cdc18 cluster (DNA replication), and the Eng1 cluster (cell separation). There are many important cell cycle events in M and S, and these two phases are close together in rapidly-growing S. pombe. The many genes peaking in late G2 and M may simply represent the cell's efforts to prepare for the many activities of M and S. It will be of interest to see what happens to the timing of the Cdc18 cluster (DNA synthesis genes) in slowly growing cells with a long G1: Will they still be transcribed in mitosis, or will they now be transcribed in late G1?
If mitotic repression does exist in S. pombe, how is it that the Cdc15, Cdc18, and Eng1 clusters peak in M phase? Baum et al. [58] have used nuclear run-on to show that cdc18 and some other members of the cdc18 cluster can be actively transcribed in mitosis at a time when histone H1 kinase activity is high and chromatin is presumably condensed. Our own results agree that essentially all the genes of the Cdc15, Cdc18, and Eng1 clusters are highly expressed at a nuc2 arrest, a time at which histone H1 kinase activity is high, and chromatin should be condensed. Our elutriation data suggest that in normal cells, the peak of expression of genes in the Cdc15 and the Cdc18 cluster is almost simultaneous with mitosis (see Figures 6 and 8). The genes of these clusters may be specialized for transcription in mitosis. Interestingly, the Cdc15 cluster genes have binding sites for a forkhead transcription factor. Forkhead transcription factors have a “winged-helix” fold, a structure they share with histone H1. Like histone H1, forkhead proteins may be capable of binding to linker DNA in between nucleosomes, and seem to be capable of binding even to chromatin that is relatively condensed [59–62]. That is, perhaps forkhead is an enabler of transcription for genes in condensed chromatin, and so is particularly suitable for driving expression of genes in mitosis. The Cdc18 cluster depends on the MBF factor, and the MBF/SBF/E2F family of DNA-binding proteins also has a winged helix fold [63]. Finally, the Ace2/Swi5 family of transcription factors has been associated with the recruitment of chromatin remodeling enzymes and histone acetylases [64]. Even in mammals, which clearly do have mitotic repression, there are mitotic genes strongly expressed during mitosis [65]. Interestingly, at least some of these genes are thought to be regulated by forkhead transcription factors [66].
The more moderately expressed genes in the G2/M wave (i.e., genes not in the Cdc15, Cdc18, or Eng1 clusters) tend to be expressed in late G2 rather than in M (see Figure 6C). Thus these genes may be subject to mitotic repression. Perhaps a large number of genes are expressed in late G2 because it the last chance to be expressed before M, a relatively inopportune time for transcription.
Comparison of Cell Cycle Genes in S. pombe and S. cerevisiae
Of our top 200 ranked cell cycle–regulated genes, 72 (36%) had S. cerevisiae homologs that cycled, 68 had S. cerevisiae homologs that did not cycle significantly, and 60 did not have clear S. cerevisiae homologs. (A detailed comparison of the top 200 S. pombe genes and their S. cerevisiae homologs is available as Table S3).
Genes involved in core cell cycle processes such as DNA synthesis and mitosis were especially likely to cycle in both organisms. On the other hand, genes involved in budding (in S. cerevisiae) or fission (in S. pombe), or in cell wall carbohydrate metabolism, generally did not cycle in both organisms for the obvious reasons that the mechanism of cell separation, and the nature of the carbohydrates in the cell wall, are not conserved between the two yeasts.
There are many individual cases where a process is cell cycle–regulated in both organisms, but either the level of regulation (i.e., transcriptional or post-transcriptional) or the identity of the gene regulated varies between the two yeasts. One example is the activity of the cdc2/Cdc28 protein kinase. In S. cerevisiae, most of the cyclins are very strongly regulated at the transcriptional level (e.g., CLN1,
CLN2, CLB5, CLB6, CLB1, and CLB2), but in S. pombe, the equivalent cyclins are only weakly or moderately regulated at the transcriptional level. Possibly compensating for this relatively weak transcriptional regulation, S. pombe has very strong post-translational regulation of Cdc2 kinase activity via Wee1/Mik1 inhibitory tyrosine phosphorylation of Cdc2, whereas the homologous system is relatively weak in S. cerevisiae. That is, both yeasts strongly regulate cdc2/Cdc28 activity through the cycle, but emphasize different mechanisms. A second example is provided by the gene products of dut1 (SPAC644.05c) and ung1. These proteins both work to exclude uracil from DNA, but by independent mechanisms. The Dut1 protein hydrolyses dUTP, whereas the Ung1 protein removes uracil from DNA by cleaving the glycosidic bond. In S. cerevisiae,
dut1 is very weakly cell cycle regulated, whereas ung1 is moderately regulated. In S. pombe,
dut1 (SPAC644.05c) is very strongly regulated, whereas ung1 appears not to be regulated at all. Thus both yeasts use cell cycle transcriptional control to exclude uracil from DNA, but the emphasis is on different genes.
Regulatory Networks and the Late G2 Bump
In S. cerevisiae, there is a regulatory network governing the transcription of cell cycle genes. This network is organized as a circular cascade, such that transcriptional and post-transcriptional changes occurring during one part of the cycle seem to promote changes in the next part of the cycle, and so on around a circle [67–69]. In principle, S. pombe must also have a circular cascade of some kind to make the cell cycle repeat. However, fewer cell cycle regulatory mechanisms have been described in S. pombe than in S. cerevisiae, and so the wiring of the putative cascade is still unclear. In particular, it is unclear how extensive a role is played by transcriptional control.
Moreover, in S. cerevisiae, genes displaying large-amplitude cell cycle changes are distributed throughout the cycle [6], consistent with the idea that transcriptional control contributes significantly to all phases of the cascade [67]. However, in S. pombe, most large-amplitude genes are expressed in a window near the G2/M transition, whereas genes of moderate and low amplitudes are distributed throughout the cycle. This concentration of large-amplitude genes near M may suggest that transcriptional control is most important for only some portions of the cascade.
Within the G2/M window of high-amplitude transcriptional regulation, one can discern what may be part of the regulatory wiring diagram. The transcription factor gene fkh2 peaks in the earliest part of the late G2 window. Over 100 other genes in this window, including fkh2 itself, have FKH binding sites, so the up-regulation of fkh2 may contribute to this large wave of gene expression.
One of the critical targets of the Fkh transcription factor may be the gene for the Ace2 transcription factor. The ace2 promoter has multiple sites for Fkh binding. The ace2 promoter also has one site for Ace2, so, like the fkh2 gene, ace2 may be autoregulatory. The Ace2 transcription factor then induces a cluster of genes involved in cell separation and cell wall metabolism. Interestingly, a forkhead transcription factor is involved in turning on the ACE2 gene in S. cerevisiae, so this particular part of the cell cycle wiring diagram appears to be conserved in the two species.
Three of the major cell cycle transcription factors in S. cerevisiae, MBF/SBF, Fkh, and Ace2/Swi5, have homologous cell cycle transcription factors in S. pombe. The major exception is Mcm1, a MADS-box transcription factor. In S. cerevisiae, there are two paralogs of this gene, MCM1, and ARG80. Mcm1 is a transcription factor for cell cycle genes and mating genes, whereas Arg80 controls various metabolic processes. The best S. pombe orthologs are Map1 and Mbx1 [19]. There was no noticeable enrichment of an Mcm1-like binding motif in front of any cluster of cell cycle–regulated genes; i.e., there was no evidence for a binding site for Map1 or Mbx1.
In multicellular animals, the major well-characterized cell cycle transcription factor(s) are those of the E2F/DP family [70,71]. These typically control a cluster of genes expressed in late G1, and the genes are involved in DNA replication and commitment to the cell cycle. Functionally, the genes controlled by E2F/DP in animals are similar to the genes controlled by MBF in the two yeasts. E2F and DP proteins are not very similar in sequence to the proteins found in MBF, but it is also true than various E2F and DP proteins are not very similar to each other, though they are clearly related. E2F and DP recognize binding sites with a
CGCG core, as does MBF. Furthermore, the DNA-binding domain of E2F/DP factors consists of a winged-helix fold [72], as do the DNA-binding domains of Swi4 and Mbp1 (components of S. cerevisiae SBF and MBF, respectively) [63,72]. Thus, despite the overall sequence dissimilarity, it is possible that MBF in the yeasts, and E2F/DP in animals, are cell cycle transcription factors that are related by descent and which have always controlled the cell cycle expression of genes involved in DNA replication.
Materials and Methods
Microarrays
Microarrays were made by spotting unmodified, double-stranded PCR products onto glass slides coated with aminopropylsilane (Erie Scientific). Spotting was done using a robot of the DeRisi design (http://cmgm.stanford.edu/pbrown/mguide/) and ArrayMaker2 software (http://derisilab.ucsf.edu/arraymaker.shtml). PCR primers were designed using Primer3 (Whitehead Institute, Cambridge, Massachusetts, United States) and a shell script. Primers were designed against approximately 5,000 open reading frames and RNAs (excluding pseudogenes) as annotated by the Sanger Centre (http://www.sanger.ac.uk/Projects/S_pombe/DNA_download.shtml). In general, PCR primer pairs were designed to give products 500 to 1,000 bp in length, because the yield of the PCR reaction decreased for products longer than 1,000 bp. When the PCR product was small compared to the length of the gene, it was usually chosen from the 3′ region of the gene, so as to maximize representation in poly dT-primed cDNA synthesis. PCR products were amplified from genomic S. pombe DNA, and so in some cases the final product included introns, but the design parameters maximize contiguous exonic sequence. A fuller description of the microarrays will be published elsewhere. A full description of the primer pairs, and hence the features on the microarrays, can be found at http://www.redgreengene.com.
Cell cycle synchronizations
Two methods of cell cycle synchronization were used, elutriation and a cdc25–22 block and release. Two independent elutriation experiments were carried out. For elutriations, 8 l of h-972 cells (wild-type) were grown in YES (autoclaved, elutriation B or filter-sterilized, elutriation A) to early log phase (OD600 = 0.4) at room temperature (25 °C). Cells were harvested by centrifugation, resuspended in approximately 100-ml YES, and sonicated, all at room temperature. For elutriation B, approximately half of the cell volume was reserved for the reference cDNA preparation. For elutriation A, the reference cDNA was prepared independently and the entire sample was used for elutriation. Cells were loaded into a Beckman elutriator rotor containing two 40-ml elutriation chambers connected in series. When two chambers are used in series, the bulk of the cells remain in the first chamber, but the smallest cells flow into the second chamber, and then, at higher pump speeds, some of these flow out of the elutriator for collection. This arrangement provides both high capacity and high resolution. The elutriator was used at 1,800 rpm at room temperature. After every increase in pump speed, a fraction of about 150 ml was collected, containing about 5 × 108 cells (elutriation B) or 3 × 109 cells (elutriation A). These were diluted to OD600 0.2–0.05 (greater dilution for samples harvested at late times) with conditioned (elutriation B) or fresh filter-sterilized (elutriation A) medium, and then sampled with time.
An entire three cell cycle time course was obtained from five elutriator fractions (elutriation B) or two fractions (elutriation A). We used adjacent fractions containing no (< 0.5%) septated cells; the elutriator fraction with the largest cells (i.e., the last fraction collected) was used first, then the elutriator fraction with the next largest cells, and finally the elutriator fraction with the smallest cells (i.e., the first fraction collected). In general, the fractions were “overlapped,” i.e., the last sample from one fraction and the first sample from the next fraction were collected at the same time. “Overlapped” fractions, though collected at the same time, were deemed to have been collected at slightly different times; the number of minutes by which overlapped fractions were offset was determined by the offset, in minutes, of the septation indicies for the two fractions. That is, for any pair of overlapped fractions, the smaller cells were deemed to have been collected earlier, by a time determined from the offset of the septation indicies of the two fractions. Note that elutration A used only two fractions, and so there was only one overlap. Samples were taken about 10 min. (elutriation A) or 15 min (elutriation B) apart; exact sampling times are given in the Treeview files 1, 2, and 3 (Dataset S1) and at http://www.redgreengene.com.
Cells (108 cells/sample) were harvested by centrifugation at 4 °C and washed with ice-cold water, snap frozen, and stored at −70 °C. For elutriation A, an equal volume of ice was added to the cell culture during harvest (harvest with ice). The reference sample for hybridizations was sonicated cells prior to elutriation (elutriation B), or h-972 grown to OD600 0.2 in filtered YES at 25 °C (elutriation A). Septation index was monitored by phase contrast microscopy of live cells during each experiment. In addition, frozen cell pellets were thawed and stained with DAPI and calcofluor to monitor anaphase (“binucleates”) and septation for elutriation A. Cells were scored as “binucleates” if two nuclei were visible, but there was no septum.
For the cdc25–22 block release, the prototrophic strain JLP1164 h+ cdc25–22 was grown in filtered YES at 25 °C to OD600 = 0.4 and then used to inoculate 4 × 500 ml filtered YES to an OD600 of 0.1 (flask 1), 0.08 (flask 2), 0.07 (flask 3), and 0.05 (flask 4). Cells were shifted to a water bath at 36.5 °C for 4 h to arrest them in G2 (time = 0 h) and then shifted back to 25 °C rapidly in an ice-water bath (26 °C was achieved in approximately 5 min; cultures did not cool below 25 °C). Samples were taken 10 min apart and harvested with ice as described above. The reference sample for hybridizations was JLP1164 h + cdc25–22 grown at 25 °C to OD600 = 0.2 in filtered YES. Septation index was monitored by phase contrast microscopy.
Other microarray experiments
To examine cells released synchronously from a cdc10 arrest, 8 l of strain JLP1166 h− cdc10-M17 was grown at 25 °C to OD600 = 0.5 in filtered YES, and then harvested and elutriated to obtain a fraction of G2 cells. These were diluted to 106 cells/ml, shifted to 36.5 °C for 3 h 15 min, rapidly cooled to 25 °C as described above (time = 0), and then sampled with time. Cells were harvested with ice. Samples were also collected and analyzed by flow cytometry to monitor DNA replication. The reference sample for hybridizations was JLP1166 h− cdc10-M17 grown to OD600 0.2 in YES at 25 °C.
To examine cells grown in low nitrogen, wild-type h-972 was grown in EMM lacking NH4 and supplemented with 20 mM phenylalanine (EMM-phe) to provide a limiting nitrogen source to expand the G1 window [73]. 8 l of cells were grown at 25 °C to OD600 = 0.4, collected by centrifugation at 4 °C and kept on ice and sonicated on ice. Approximately half of the total cell volume (125 ml, total 2 × 108 cells) was reserved for reference cDNA synthesis and the remainder was elutriated at 4 °C to fractionate the culture into 21 fractions ranging from small cells (50% G1) and then medium cells (G2) and finally to long, septated cells. Fractions were harvested immediately by centrifugation at 4 °C. Fraction assignments were confirmed by flow cytometry analysis and high-quality hybridizations were obtained with fractions 2, 3, 5, 7, 10, 13, and 16.
To examine cells arrested at the cdc10, cdc22, cdc25, and nuc2 block points, four strains carrying these cell cycle mutants (cdc22-M45, nuc2–663, cdc25–22, and cdc10-M17) and a wild-type reference control were grown to OD 0.05–0.08 in YES at 25 °C and shifted to 36.5 °C. After 4 h of arrest at this restrictive temperature, a sample was taken for microarray analysis. For each strain, the experiment was repeated with an independent single colony. Figures 6, 7, and 9 show results (in different columns) from both single colonies. Strains used were wild-type PR109 h− leu1–32 ura4-D18 (obtained from P. Russell), and the cell cycle mutants (F84) OM591 h− cdc22-M45 (P. Russell), (F58) PR580 h− leu1–32 nuc2–663 (P. Russell), JLP1165 h+ cdc25–22 (this study), and JLP1166 h− cdc10-M17 (this study).
Microarray hybridization and processing
Cell samples for RNA isolation were rapidly cooled by addition of an equal volume of ice (except for elutriation B in which samples were placed on ice) and then collected by centrifugation at 4,000 rpm. (3,300 × g) at 4 °C for 3 min. Pellets were washed twice in ice-cold dH20, frozen in liquid nitrogen, and stored at −70 °C. Total RNA was isolated using RiboPure Yeast (Ambion, Austin, Texas, United States) according to the manufacturer's instructions (elutriation A samples) or hot phenol essentially as described [7] (http://www.sanger.ac.uk/PostGenomics/S_pombe/docs/rnaextraction_website.pdf with slight modifications according to the detailed protocols at http://www.redgreengene.com). Isolated RNA was further purified by RNAeasy cleanup columns (Qiagen, Valencia, California, United States) and quantitated by absorption spectroscopy.
Microarray probes were prepared in two steps. First, cDNA was synthesized incorporating aminoallyl-dUTP (aadUTP). Purified aadUTP cDNA was then coupled with Cy3 or Cy5 fluorescent dyes according to protocols from the Institute for Genomic Research (http://www.tigr.org/tdb/microarray/protocolsTIGR.shtml) with slight changes (http://www.redgreengene.com) as follows: 20–25 μg of total RNA was used for cDNA synthesis with 4 μg of oligo-dT primer (not random hexamers), and reactions contained 300 μM aminoallyl-dUTP with 200 μM dTTP. RNA was destroyed using RNase instead of NaOH, and reactions were purified with a Qiagen PCR purification kit. Dye incorporation was determined by absorption spectra and was typically one fluor/20–30 nucleotides.
For hybridizations, cDNA with 50 pmol Cy3 plus reference cDNA with 50 pmol Cy5 was included in a 24 μl total hybridization solution (25% v/v formamide, 5× SSC, 0.1% SDS, and 100 μg/ml of sonicated salmon sperm DNA). Hybridizations were performed under 22 × 25 mm lifter cover slips (Erie Scientific, Portsmouth, New Hampshire, United States) at 50 °C in a humidified chamber for 16–20 h. Hybridized arrays were washed by gently shaking as follows: twice briefly with 2× SSC/0.1% SDS (50 °C), twice for 10 min with 2× SSC/0.1% SDS (50 °C), and four times briefly with 0.1× SSC at room temperature. Arrays were dried by centrifugation.
Arrays were scanned using an Axon 4000B scanner, controlled by GenePix Pro 5.1 software with a pixel size of 5 μm and two-pass sequential line averaging. Laser power was set to 100%, and PMT gains were subjectively adjusted during prescan to maximize effective dynamic range and to limit image saturation. Lossless image files were stored for later analysis.
Data extraction and storage
To extract data from microarray scans, previously stored image files representing all hybridizations were analyzed in parallel. Spot size, location, and quality were determined automatically by GenePix Pro algorithms. Dynamic spot resizing between 60 and 150 μm diameter was permitted based upon image examination and prior optimization. Misidentification of spot locations was corrected by manual adjustment of the map prior to automatic sizing and shifting. Only in cases of gross hybridization defect were spots/regions manually moved/resized or flags modified to “bad,” permitting consistent spot calling. Following spot location, parameters and values for each spot were calculated by GenePix Pro and exported. No normalization was applied within GenePix Pro.
Raw data and images exported from GenePix Pro were used to populate a local installation of the Longhorn Array Database (Peter Killion, University of Texas at Austin, http://www.longhornarraydatabase.org, an SQL database based upon the Stanford Microarray Database (http://yeastgenome.org, Stanford University). Initial data normalization was performed at the time of population. Briefly, spots were categorized as “pombe” or “other.” Pombe spots were further categorized into “normalization” (no bad, missing, absent, or not-found flags) or “non-normalization” (bad, missing, absent, or not-found flags). Only normalization spots were further considered for the normalization calculation. Finally, spots with greater than 5% saturation in either image channel were discarded from this group. The mean log2 ratio of the median net intensities (R
m, foreground pixel − median of the local background) was calculated. This “normalization factor” represented the distance from a red/green ratio of one, and was used as a scalar modifier for the ratios of all spots in the hybridization; i.e., though only spots meeting a stringent “good” criterion were used to determine the normalization value, this value was subsequently applied to all spots, good or not. During retrieval of data, several further criteria were used to ensure high-quality data in downstream analysis. Only spots with non-negative flag values were retrieved (not bad, missing, absent, or not-found), and only spots with a regression correlation of pixel ratios (a metric of internal spot consistency) greater than 0.6 were used. Spot values were averaged (mean) when multiple independent spots representing a single PCR product were present as internal controls or otherwise. When analyzing multiple hybridizations, such as during time-course analysis, iterative gene and array centering was performed. Briefly, within an array of genes × arrays, the mean log2 ratio of medians (R
m) was calculated and subtracted from each log2
R
m, first along the gene axis, then along the array axis, until subsequent iterations varied by < 0.001%.
Normalization
There are some special red/green normalization issues relevant to the genome-wide waves of expression (see Figure 6). In general, fluorescence from mRNA from synchronous cells was normalized to total fluorescence from mRNA from asynchronous cells. If all mRNAs were equally reduced in abundance during mitosis (e.g., because of genome-wide mitotic repression), this normalization would obscure the effect. However, despite normalization, any such repression would still be somewhat visible as a relative loss of unstable mRNAs versus stable mRNAs. We presume that if the trough of peaks of expression in M/G1/S seen in Figure 6 is indeed partially due to mitotic repression of transcription, then the individual genes troughing at this time are genes that produce unstable mRNAs, which are then noticeably repressed despite the red/green normalization because they are reduced relative to stable mRNAs. This argument suggests that stable mRNAs should tend to peak in the M/G1 interval. Indeed, some genes do peak at this time. These, however, would be fewer in number than the genes that fail to peak, because, in general, cells tend to express a small number of stable mRNAs to very high levels, comprising the bulk of mRNA, and a large number of unstable mRNAs to low levels [74].
A second normalization issue is that the oscillation of the strongly regulated genes would have an effect, via normalization, on the apparent expression of non-oscillating genes (i.e., genes that do oscillate would produce an artifactual, complementary oscillation in non-oscillating genes, via normalization). To side-step this artifact, pixel intensity data for the bottom 4,000 genes were extracted from the microarray data before the red/green normalization step, and then normalized and analyzed after extraction, so that the oscillation of the 1,000 most strongly cyclic genes would not interfere with normalization of the least cyclic genes.
Cluster analysis
For cluster analysis, array- and gene- centered log2
R
m data were hierarchically clustered along the gene axis by the agglomerative algorithm of Eisen et al. [10]. Data were visually presented using JavaTreeView (http://jtreeview.sf.net and http://jtreeview.sourceforge.net/manual.html). Separation of the total dendrogram into subordinate clusters was performed subjectively.
Motif analysis
To find DNA sequence motifs, nucleotides extending from −1 to the edge of the most 3′ proximal gene (stop or ATG, depending on orientation) with a maximum length of 12,000 bp were extracted genewise for each cluster and used as a target set for motif searching. Three different motif search programs were used: MEME, AlignAce, and SPEXS.
MEME (Multiple EM for Motif Elicitation) [15] was used to find motifs between five and nine nucleotides long present in any number of copies on either strand, weighted to find 1/3n to 3n total sites in the target set of n sequences. Parameters were set as follows: $ meme (sequence.name) –dna –minsites (n/3) –maxsites (n*3) –mod anr –minw 5 –maxw 9 –revcomp –nmotifs 10 –evt 0.1 –bfile (fifth order Markov model). (Other parameters were also tried in additional searches.) The top ten motifs exceeding an E-value of 0.1 were generated using a background set consisting of the fifth order Markov model representing possible nucleotide pentuplets in all S. pombe upstream regions.
AlignACE [12] uses a Gibbs sampling algorithm. Again, all S. pombe upstream regions were used as a background set.
SPEXS (Sequence Pattern EXhaustive Search) [14], a word-search enumeration algorithm, was also used. Relative frequency of 1- to 9-mers was calculated, and compared between the target set and all S. pombe upstream sequences.
Identification of oscillating transcripts
In general, identification of oscillating transcripts requires a method for finding oscillations in each experiment, and then a method for combining the results from different experiments. Here, we have used Fourier analysis to identify oscillating genes. A p-value for the hypothesis of oscillation was then established using Monte Carlo simulations on shuffled data. The p-values for different experiments were then combined using known statistical properties of the p-value. Finally, the p-values for each gene were ranked. Although we have ranked the p-values, these p-values are nevertheless closely correlated with the amplitude of the oscillation.
For each time series of observations in a single time course (e.g., a three cell cycle elutriation experiment), we calculated the Fourier sums A and B over the range of times, t, in the experiment:
Here, t is the time in minutes at which the sample was taken (where the beginning of sampling is zero time); T is the cell cycle period, i.e., the time in minutes required for a complete cell cycle; and ratio(t) is the ratio of experimental to control signal at time t.
We considered these two sums as a vector C = (A,B), and then calculated the magnitude of the vector, D
O = square root of (A
2 + B
2). This magnitude, D
O, is our basic Fourier measure of whether a transcript oscillates. Note that there is no need to calculate phase.
However, random noise would generate some value of D greater than zero, and genes whose transcripts are relatively variable in abundance could generate relatively large values of D, even if these variations had no connection to the cell cycle. Therefore, as a second step, we randomly shuffled the series of observations for each gene in question, and calculated a new magnitude, D
R, for the randomized series. This randomization was repeated 1,000 times, generating 1,000 values of D
R. These represent the distribution of D for each gene, given that gene's actual variance in gene expression. Finally, we compared the original value of D
O from the unshuffled data to the distribution of D from the shuffled data, found how many standard deviations D
O is from the mean of the distribution, and in this way calculated a z score for D
O.
This procedure was repeated for each gene and for each experiment. Thus, for each gene, there were three z scores, one per experiment (two elutriation experiments and the cdc25 block-release experiment). These three z scores were then combined by the method of Stouffer, yielding a single p-value for each gene. Genes were then ranked by p-value with the lowest p-value at the top of the list. In practice, a large amplitude of oscillation contributes tremendously to a low p-value, so the upper portion of the p-value list is almost exclusively occupied by genes with high-amplitude oscillations
Gene database
In general, we have used the information in the GeneDB database (http://www.genedb.org/genedb/pombe/index.jsp) to describe the various genes studied; when a fact is given in the text about some gene but no reference is given, the information comes from GeneDB. When the primary literature has been consulted directly, the reference for the primary literature is given.
Supporting Information
Dataset S1 Treeview Files
Datasets appropriate for Treeview (see below) are provided as a tar.gz file. Upon opening, this tar.gz file will create a folder containing the three supplementary dataset files S1, S2, and S3 (pombecellcycle.cdt, pombecellcycle.gtr, and pombecellcycle.jtv) along with an additional “treeview_configuration.txt” file detailing the configuration of Treeview to access Sanger GeneDB for S. pombe.
These cdt, gtr, and jtv files can then be used to view clusters of the top 750 genes in a convenient and searchable way using Treeview, an open-source cluster visualization package available for many different platforms (http://jtreeview.sourceforge.net). Launch Java Treeview and open file pombecellcycle.cdt. The two supporting files will be accessed automatically (given that they are in the same folder). The pombecellcycle.jtv file is not required and only provides configuration settings. Java Treeview can be instructed to link directly to GeneDB (or any other database) so that a user can quickly check information on any given gene in a cluster of interest. To configure a copy of Java Treeview to link to GeneDB, do the following. First, in “Settings” go to “gene URL presets” and change one of the presets to name GeneDB and template http://www.genedb.org/genedb/Search?organism=pombe&name=HEADER&isid=true and choose this as default. Second, go to Settings “URL settings” and select gene; check that your new template is selected and that the “UID” setting is chosen. Now when an individual gene in a cluster is selected, the GeneDB Web page for that gene should open automatically. Documentation for Java Treeview is available at http://jtreeview.sourceforge.net/manual.html. The cdt file contains log ratios for each gene at each timepoint, and the timepoint names also contain information on septation index (SI) for all three synchrony experiments and binucleates for elutriation A. The form of the header for timepoint names is: [experiment]_[time]min_[%]SI_[%]BN. SI indicates Septation Index (i.e., percent septation), assayed by calcofluor for “ElutA,” phase contrast for “ElutB,” and cdc25. “BN” indicates the binucleate “anaphase index” as measured by DAPI staining, applicable only to ElutA. These headers are displayed in the Zoom window of Java Treeview.
(210 KB ZIP).
Click here for additional data file.
Figure S1 Oscillation of cdc18
The oscillation of the cdc18 transcript through two cell cycles in elutriation A is plotted as a histogram (right y-axis). Also shown are the binucleate (blue triangles) and septation indices (cyan squares) (left y-axis).
(165 KB EPS).
Click here for additional data file.
Table S1 Cell Cycle Parameters of S. pombe Genes
Data is presented for all 4,988 genes analyzed. Column headings are as follows: “SUID” is the systematic name. “Common-name” is the common name. “Rank” is the rank in our p-value list, from smallest p-values (i.e., most significant genes) to largest p-values. A p-value of “0” means a p-value of less than 10−16. “Desc” is a one-line description of the gene from GeneDB. “Cluster” is the cluster to which the gene belongs, if any. “ElutA_Phase” is the phase angle of the gene calculated from elutriation A. Phase angles range from 0° to 360° (i.e., around a circle). The phase angles of landmark events are as follows: binucleates (anaphase) peak at a phase angle of 238°; septation peaks at a phase angle of 277°; and histone expression (S phase) peaks at a phase angle of about 312. Thus 0° is early in G2, but not the very beginning of G2. “ElutA_Fourier_component” is related to the amplitude of the gene's oscillation. It is the magnitude from the Fourier decomposition of the elutriation A data series. It is the magnitude of only one constituent waveform (the once-per-cell-cycle wave). It is in log2 space. “Combined_P” is the combined p-value obtained by using Stouffer's method to combine the z scores from elutriation A, elutriation B, and the cdc25 block release. “ElutA_Z,” ElutB_Z,” and “cdc25_Z” are the z scores for the elutriation A, elutriation B, and cdc25 block-release experiments, respectively. Z scores were calculated from Monte Carlo simulations (see Materials and Methods). “All names” are all other known synonyms for the gene in question other than the “SUID” and the “Common-name.” In some cells of the spreadsheet, the entry is “#N/A” or “X” or “Z.” These entries indicate that a result was not calculated because of excessive missing data.
(2.9 MB XLS).
Click here for additional data file.
Table S2 SpikeChart Weight Matrices
The weight matrices and spike height rules used by SpikeChart to generate Figure 10 are shown.
(24 KB DOC).
Click here for additional data file.
Table S3
S. cerevisiae Homologs of S. pombe Cell Cycle Genes
For the top 200 S. pombe cell cycle genes, the best homologs in S. cerevisiae (if any) are shown. If the S. cerevisiae homolog oscillates through the cell cycle, then the time of peak expression is shown in the “Sc peak” column; if the homolog is not known to oscillate, then this column is marked “ND.” Any transcription factors thought to regulate the S. cerevisiae homolog are noted. If there are more than two S. cerevisiae homologs, then all these additional homologs are combined in a single field in the right-most column.
(53 KB XLS).
Click here for additional data file.
We thank Rohan Fernandez for help in designing PCR primers for amplification of microarray target sequences, and Michael Sassen for help with PCR amplification. We thank Joe DeRisi for instruction and inspiration in the art of microarray manufacture, and also for providing ArrayMaker2 software. We thank the National Center for Research Resources for funding for S. pombe microarrays (NCRR grant P40RR01632004 to JL) and the National Institutes of Health for research funding (R01GM6481304 and R01GM3997816 to BF).
Competing interests. The authors have declared that no competing interests exist.
Author contributions. SS, BF, and JL conceived and designed the experiments. AO, AR, FF, and HC performed the experiments. AR, SP, SS, BF, and JL analyzed the data. FF, BF, and JL contributed reagents/materials/analysis tools. AR, BF, and JL wrote the paper.
Citation: Oliva A, Rosebrock A, Ferrezuelo F, Pyne S, Chen H, et al. (2005) The cell cycle–regulated genes of Schizosaccharomyces pombe. PLoS Biol 3(7): e225.
Abbreviations
bpbase pair
==== Refs
References
Nurse P Thuriaux P Nasmyth K Genetic control of the cell division cycle in the fission yeast Schizosaccharomyces pombe
Mol Gen Genet 1976 146 167 178 958201
Culotti J Hartwell LH Genetic control of the cell division cycle in yeast. 3. Seven genes controlling nuclear division Exp Cell Res 1971 67 389 401 5097524
Hartwell LH Culotti J Reid B Genetic control of the cell-division cycle in yeast. I. Detection of mutants Proc Nat Acad Sci U S A 1970 66 352 359
Hartwell LH Genetic control of the cell division cycle in yeast. II. Genes controlling DNA replication and its initiation J Mol Biol 1971 59 183 194 5283752
Hartwell LH Genetic control of the cell division cycle in yeast. IV. Genes controlling bud emergence and cytokinesis Exp Cell Res 1971 69 265 276 4950437
Spellman PT Sherlock G Zhang MQ Iyer VR Anders K Comprehensive identification of cell cycle-regulated genes of the yeast Saccharomyces cerevisiae by microarray hybridization Mol Biol Cell 1998 9 3273 3297 9843569
Rustici G Mata J Kivinen K Lio P Penkett CJ Periodic gene expression program of the fission yeast cell cycle Nat Genet 2004 36 809 817 15195092
Peng X Karuturi RK Miller LD Lin K Jia Y Identification of cell cycle-regulated genes in fission yeast Mol Biol Cell 2005 16 1026 1042 15616197
Harris P Kersey PJ McInerny CJ Fantes PA Cell cycle, DNA damage and heat shock regulate suc22+ expression in fission yeast Mol Gen Genet 1996 252 284 291 8842148
Eisen MB Spellman PT Brown PO Botstein D Cluster analysis and display of genome-wide expression patterns Proc Nat Acad Sci U S A 1998 95 14863 14868
Gould SJ Linnaeus's luck? Nat Hist 2000 109 18 25 66 76
Hughes JD Estep PW Tavazoie S Church GM Computational identification of cis-regulatory elements associated with groups of functionally related genes in Saccharomyces cerevisiae
J Mol Biol 2000 296 1205 1214 10698627
Brazma A Vilo J Gene expression data analysis Microbes Infect 2001 3 823 829 11580977
Vilo J Brazma A Jonassen I Robinson A Ukkonen E Mining for putative regulatory elements in the yeast genome using gene expression data Proc Int Conf Intell Syst Mol Biol 2000 8 384 394 10977099
Bailey TL Elkan C The value of prior knowledge in discovering motifs with MEME Proc Int Conf Intell Syst Mol Biol 1995 3 21 29 7584439
Bailey TL Gribskov M Methods and statistics for combining motif match scores J Comput Biol 1998 5 211 221 9672829
Bulmer R Pic-Taylor A Whitehall SK Martin KA Millar JB The forkhead transcription factor Fkh2 regulates the cell division cycle of Schizosaccharomyces pombe
Eukaryot Cell 2004 3 944 954 15302827
Anderson M Ng SS Marchesi V MacIver FH Stevens FE Plo1(+) regulates gene transcription at the M-G(1) interval during the fission yeast mitotic cell cycle EMBO J 2002 21 5745 5755 12411492
Buck V Ng SS Ruiz-Garcia AB Papadopoulou K Bhatti S Fkh2p and Sep1p regulate mitotic gene transcription in fission yeast J Cell Sci 2004 117 5623 5632 15509866
Tanaka K Okazaki K Okazaki N Ueda T Sugiyama A A new cdc gene required for S phase entry of Schizosaccharomyces pombe encodes a protein similar to the cdc 10+ and SWI4 gene products EMBO J 1992 11 4923 4932 1464317
Lowndes NF McInerny CJ Johnson AL Fantes PA Johnston LH Control of DNA synthesis genes in fission yeast by the cell-cycle gene cdc10+ Nature 1992 355 449 453 1734281
Reymond A Schmidt S Simanis V Mutations in the cdc10 start gene of Schizosaccharomyces pombe implicate the region of homology between cdc10 and SWI6 as important for p85cdc10 function Mol Gen Genet 1992 234 449 456 1406591
Koch C Moll T Neuberg M Ahorn H Nasmyth K A role for the transcription factors Mbp1 and Swi4 in progression from G1 to S phase Science 1993 261 1551 1557 8372350
Koch C Nasmyth K Cell cycle regulated transcription in yeast Curr Opin Cell Biol 1994 6 451 459 7917338
Ayte J Leis JF Herrera A Tang E Yang H The Schizosaccharomyces pombe MBF complex requires heterodimerization for entry into S phase Mol Cell Biol 1995 15 2589 2599 7739540
Whitehall S Stacey P Dawson K Jones N Cell cycle-regulated transcription in fission yeast: Cdc10-Res protein interactions during the cell cycle and domains required for regulated transcription Mol Biol Cell 1999 10 3705 3715 10564266
Tahara S Tanaka K Yuasa Y Okayama H Functional domains of rep2, a transcriptional activator subunit for Res2-Cdc10, controlling the cell cycle “start” Mol Biol Cell 1998 9 1577 1588 9614195
Zhu Y Takeda T Whitehall S Peat N Jones N Functional characterization of the fission yeast Start-specific transcription factor Res2 EMBO J 1997 16 1023 1034 9118941
Sturm S Okayama H Domains determining the functional distinction of the fission yeast cell cycle “start” molecules Res1 and Res2 Mol Biol Cell 1996 7 1967 1976 8970158
Martin-Cuadrado AB Duenas E Sipiczki M Vazquez de Aldana CR del Rey F The endo-beta-1,3-glucanase eng1p is required for dissolution of the primary septum during cell separation in Schizosaccharomyces pombe
J Cell Sci 2003 116 1689 1698 12665550
Dekker N Speijer D Grun CH van den Berg M de Haan A Role of the alpha-glucanase Agn1p in fission-yeast cell separation Mol Biol Cell 2004 15 3903 3914 15194814
Matsumoto S Yanagida M Histone gene organization of fission yeast: A common upstream sequence EMBO J 1985 4 3531 3538 4092687
Spector MS Osley MA The HIR4–1 mutation defines a new class of histone regulatory genes in Saccharomyces cerevisiae
Genetics 1993 135 25 34 8224824
Spector MS Raff A DeSilva H Lee K Osley MA Hir1p and Hir2p function as transcriptional corepressors to regulate histone gene transcription in the Saccharomyces cerevisiae cell cycle Mol Cell Biol 1997 17 545 552 9001207
Lycan DE Osley MA Hereford LM Role of transcriptional and posttranscriptional regulation in expression of histone genes in Saccharomyces cerevisiae
Mol Cell Biol 1987 7 614 621 3547077
Kato M Hata N Banerjee N Futcher B Zhang MQ Identifying combinatorial regulation of transcription factors and binding motifs Genome Biol 2004 5 R56 15287978
Mitchison JM Nurse P Growth in cell length in the fission yeast Schizosaccharomyces pombe
J Cell Sci 1985 75 357 376 4044680
Creanor J Mitchison JM Patterns of protein synthesis during the cell cycle of the fission yeast Schizosaccharomyces pombe
J Cell Sci 1982 58 263 285 7183688
Creanor J Mitchison JM Protein synthesis and its relation to the DNA-division cycle in the fission yeast Schizosaccharomyces pombe
J Cell Sci 1984 69 199 210 6490747
Mitchison JM Sveiczer A Novak B Length growth in fission yeast: Is growth exponential?—No Microbiology 1998 144 265 266 9556367
Sveiczer A Novak B Mitchison JM The size control of fission yeast revisited J Cell Sci 1996 109 2947 2957 9013342
Jorgensen P Rupes I Sharom JR Schneper L Broach JR A dynamic transcriptional network communicates growth potential to ribosome synthesis and critical cell size Genes Dev 2004 18 2491 2505 15466158
Muller D Exler S Aguilera-Vazquez L Guerrero-Martin E Reuss M Cyclic AMP mediates the cell cycle dynamics of energy metabolism in Saccharomyces cerevisiae
Yeast 2003 20 351 367 12627401
Johnson TC Holland JJ Ribonucleic acid and protein synthesis in mitotic HeLa cells J Cell Bio 1965 27 565 574 4287276
Gottesfeld JM Wolf VJ Dang T Forbes DJ Hartl P Mitotic repression of RNA polymerase III transcription in vitro mediated by phosphorylation of a TFIIIB component Science 1994 263 81 84 8272869
Gottesfeld JM Forbes DJ Mitotic repression of the transcriptional machinery Trends Biochem Sci 1997 22 197 202 9204705
Hartl P Gottesfeld J Forbes DJ Mitotic repression of transcription in vitro J Cell Bio 1993 120 613 624 8381119
Leresche A Wolf VJ Gottesfeld JM Repression of RNA polymerase II and III transcription during M phase of the cell cycle Exp Cell Res 1996 229 282 288 8986611
Hu P Samudre K Wu S Sun Y Hernandez N CK2 phosphorylation of Bdp1 executes cell cycle-specific RNA polymerase III transcription repression Mol Cell 2004 16 81 92 15469824
Klein J Grummt I Cell cycle-dependent regulation of RNA polymerase I transcription: The nucleolar transcription factor UBF is inactive in mitosis and early G1 Proc Nat Acad Sci U S A 1999 96 6096 6101
Spencer CA Kruhlak MJ Jenkins HL Sun X Bazett-Jones DP Mitotic transcription repression in vivo in the absence of nucleosomal chromatin condensation J Cell Biol 2000 150 13 26 10893252
Parsons GG Spencer CA Mitotic repression of RNA polymerase II transcription is accompanied by release of transcription elongation complexes Mol Cell Biol 1997 17 5791 5802 9315637
Elliott SG Coordination of growth with cell division: Regulation of synthesis of RNA during the cell cycle of the fission yeast Schizosaccharomyces pombe
Mol Gen Genet 1983 192 204 211 6580523
Elliott SG Regulation of the maximal rate of RNA synthesis in the fission yeast Schizosaccharomyces pombe
Mol Gen Genet 1983 192 212 217 6580524
Fraser RS Nurse P Novel cell cycle control of RNA synthesis in yeast Nature 1978 271 726 730 625340
Fraser RS Nurse P Altered patterns of ribonucleic acid synthesis during the cell cycle: A mechanism compensating for variation in gene concentration J Cell Sci 1979 35 25 40 370130
Grigull J Mnaimneh S Pootoolal J Robinson MD Hughes TR Genome-wide analysis of mRNA stability using transcription inhibitors and microarrays reveals posttranscriptional control of ribosome biogenesis factors Mol Cell Biol 2004 24 5534 5547 15169913
Baum B Nishitani H Yanow S Nurse P Cdc18 transcription and proteolysis couple S phase to passage through mitosis EMBO J 1998 17 5689 5698 9755169
Cirillo LA Lin FR Cuesta I Friedman D Jarnik M Opening of compacted chromatin by early developmental transcription factors HNF3 (FoxA) and
GATA-4
Mol Cell 2002 9 279 289 11864602
Cirillo LA Zaret KS An early developmental transcription factor complex that is more stable on nucleosome core particles than on free DNA Mol Cell 1999 4 961 969 10635321
Cirillo LA McPherson CE Bossard P Stevens K Cherian S Binding of the winged-helix transcription factor HNF3 to a linker histone site on the nucleosome EMBO J 1998 17 244 254 9427758
Carlsson P Mahlapuu M Forkhead transcription factors: Key players in development and metabolism Dev Biol 2002 250 1 23 12297093
Taylor IA McIntosh PB Pala P Treiber MK Howell S Characterization of the DNA-binding domains from the yeast cell-cycle transcription factors Mbp1 and Swi4 Biochemistry 2000 39 3943 3954 10747782
Cosma MP Tanaka T Nasmyth K Ordered recruitment of transcription and chromatin remodeling factors to a cell cycle- and developmentally regulated promoter Cell 1999 97 299 311 10319811
Whitfield ML Sherlock G Saldanha AJ Murray JI Ball CA Identification of genes periodically expressed in the human cell cycle and their expression in tumors Mol Biol Cell 2002 13 1977 2000 12058064
Alvarez B Martinez AC Burgering BM Carrera AC Forkhead transcription factors contribute to execution of the mitotic programme in mammals Nature 2001 413 744 747 11607034
Simon I Barnett J Hannett N Harbison CT Rinaldi NJ Serial regulation of transcriptional regulators in the yeast cell cycle Cell 2001 106 697 708 11572776
Futcher B Transcriptional regulatory networks and the yeast cell cycle Curr Opin Cell Biol 2002 14 676 683 12473339
Zhu G Spellman PT Volpe T Brown PO Botstein D Two yeast forkhead genes regulate the cell cycle and pseudohyphal growth Nature 2000 406 90 94 10894548
Attwooll C Denchi EL Helin K The E2F family: Specific functions and overlapping interests Embo J 2004 23 4709 4716 15538380
Bracken AP Ciro M Cocito A Helin K E2F target genes: Unraveling the biology Trends Biochem Sci 2004 29 409 417 15362224
Zheng N Fraenkel E Pabo CO Pavletich NP Structural basis of DNA recognition by the heterodimeric cell cycle transcription factor E2F-DP Genes Dev 1999 13 666 674 10090723
Carlson CR Grallert B Stokke T Boye E Regulation of the start of DNA replication in Schizosaccharomyces pombe
J Cell Sci 1999 112 939 946 10036243
Futcher B Latter GI Monardo P McLaughlin CS Garrels JI A sampling of the yeast proteome Mol Cell Biol 1999 19 7357 7368 10523624
Dragon F Gallagher JE Compagnone-Post PA Mitchell BM Porwancher KA A large nucleolar U3 ribonucleoprotein required for 18S ribosomal RNA biogenesis Nature 2002 417 967 970 12068309
| 15966770 | PMC1157095 | CC BY | 2021-01-05 08:28:15 | no | PLoS Biol. 2005 Jul 28; 3(7):e225 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030225 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1600001910.1371/journal.pbio.0030228Research ArticleBioinformatics/Computational BiologyEvolutionNoneThe Evolution of Connectivity in Metabolic Networks Evolution of Connectivity in Metabolic NetworksPfeiffer Thomas [email protected]
1
2
Soyer Orkun S
2
Bonhoeffer Sebastian
2
1 Computational Laboratory, ETH Zurich, Zurich, Switzerland,2 Ecology and Evolution, ETH Zurich, Zurich, SwitzerlandLevin Simon Academic EditorPrinceton UniversityUnited States of America7 2005 28 6 2005 28 6 2005 3 7 e2286 10 2004 26 4 2005 Copyright: © 2005 Pfeiffer et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Simple Rules Reproduce a Hub-Shaped Cell Metabolic Network
Processes in living cells are the result of interactions between biochemical compounds in highly complex biochemical networks. It is a major challenge in biology to understand causes and consequences of the specific design of these networks. A characteristic design feature of metabolic networks is the presence of hub metabolites such as ATP or NADH that are involved in a high number of reactions. To study the emergence of hub metabolites, we implemented computer simulations of a widely accepted scenario for the evolution of metabolic networks. Our simulations indicate that metabolic networks with a large number of highly specialized enzymes may evolve from a few multifunctional enzymes. During this process, enzymes duplicate and specialize, leading to a loss of biochemical reactions and intermediary metabolites. Complex features of metabolic networks such as the presence of hubs may result from selection of growth rate if essential biochemical mechanisms are considered. Specifically, our simulations indicate that group transfer reactions are essential for the emergence of hubs.
Computer simulations show how the complex organization of metabolic networks can arise from selection for a simple trait such as growth rate.
==== Body
Introduction
Metabolic networks belong to the best-studied biochemical networks. They consist of metabolites and of biochemical reactions transforming these metabolites into each other. In contrast to other biochemical networks such as signal transduction networks and genetic networks, complete network topologies can be easily obtained from annotated genomes. It has recently been reported that the connectivity of metabolic networks follows approximately a power law, i.e., the frequency, P(k), of metabolites participating in k reactions is given by P(k)∼k−γ, where γ is a constant coefficient [1–4]. This distribution implies that although most metabolites are involved in only few reactions, some metabolites serve as “hubs” and are involved in many reactions. These hubs are more frequent then would be expected, for example, in random networks. Interestingly, many hubs, such as ATP, NADH, glutamate, coenzyme A, and their derivates, serve as key compounds in the transfer of specific biochemical groups, such as phosphate groups, redox equivalents, amino groups, and acetyl groups.
Using computer simulations we here study whether hubs emerge in a widely accepted scenario for the evolution of metabolic networks. This scenario has originally been proposed by Kacser and Beeby [5] and considers an early stage in the evolution of life. It is based on the assumption that cells exist with (1) membranes that separate the interior of the cell from the environment, (2) a metabolism that allows cells to transform compounds taken up from the environment into those that are required for the formation of biomass, and (3) genes coding for enzymes that catalyze these biochemical reactions. Because of the low coding capacity and because it is unlikely that a large number of specialized enzymes emerge de novo, it is plausible to assume that these ancestral cells had only a few enzymes with broad specificities. The broad specificity allowed catalyzing all essential reactions at the cost of a low turnover for any single biochemical reaction. The ancestral cells are assumed to have been selected for growth rate and evolve by mutations affecting the kinetic properties of the enzymes and occasional gene duplications. This lead to the emergence of networks with highly specialized enzymes as observed in modern organism.
A number of alternative scenarios for the evolution of novel enzymes and metabolic pathways have been proposed [6]. One scenario for the evolution of metabolism in the early history of life has been proposed by Horowitz [7] and is referred to as “backwards evolution.” This scenario starts in an environment containing a large number of metabolites. Therefore only a low number of reactions are required to produce a useful product from one of these metabolites. As metabolites get depleted from the environment, longer pathways evolve in a reverse fashion, connecting the product to more distant metabolites present in the environment. A further scenario for the evolution of novel enzymes and metabolic pathways is referred to as “enzyme recruitment” [8]. This scenario is related to the scenario proposed by Kacser and Beeby. However, it does not apply to the initial emergence of metabolic networks early in the history of life, because it assumes that a core metabolic network with specialized enzymes is already present. Novel pathways emerge by recruiting enzymes from existing pathways. Analysis of distribution patterns of homologous enzymes in metabolic networks and of reactions catalyzed by different enzymes of the same family support enzyme recruitment rather than backwards evolution [6,9], but are also compatible with the Kacser and Beeby scenario. We therefore argue that this scenario represents a plausible mechanism for the evolution of metabolic networks in early organisms. It remains to be shown whether it indeed leads to the emergence of metabolic networks with connectivity distributions similar to those observed in nature. Based on the scenario of Kacser and Beeby, we therefore develop a model that allows simulating the evolution of metabolic networks and compare the resulting networks with natural networks.
Assumptions
We simulate the evolution of metabolic networks based on the following assumptions:
Because of the importance of group transfer in metabolic networks we assume that metabolic networks are group transfer networks (Figure 1). Metabolites consist of biochemical groups. Enzymes catalyze the transfer of groups between metabolites. Additionally there are transporters that catalyze the uptake or excretion of metabolites. Some of the metabolites are involved in the formation of biomass. The concentrations of these metabolites determine the rate of biomass formation. Biomass formation results in growth. The growth rate is given by the rate and the costs of biomass formation (i.e., the cost for the synthesis of enzymes and other compounds in the cell, see Materials and Methods). The dynamics of the metabolites is described by a set of ordinary differential equations for the metabolite concentrations and is based on the kinetic properties of the reactions and transport processes. Additionally, growth leads to the dilution of metabolites at a rate equal to the growth rate. The system has therefore no conservation relations, and metabolites that are not involved in a reaction or transport process have a steady-state concentration of zero. There are two ways in which a metabolite can have a non-zero steady state. Either it participates in at least two reactions—one of which produces the metabolite, while the other one consumes the metabolite–or a metabolite is produced by a single reaction and is diluted due to growth. Fitness is assumed to be proportional to the growth rate at steady state of the system. Thus fitness is a dynamic property of the entire network.
Figure 1 Structure of the Initial Network
(A) Metabolites are assumed to consist of N different biochemical groups. Metabolites can be denoted as binary strings, because each group can be either present or absent. In the example given, the metabolite carries groups 1, 3, 4, and 5 whereas all other groups are absent.
(B) Enzymes catalyze the transfer of specific biochemical groups between metabolites. For the group transfer we assume a ping-pong–like mechanism that consists of two half-reactions. In the first half-reaction, the group is transferred from a donor to the enzyme. Thereby the donor is transformed into its corresponding acceptor. In the second half-reaction, the enzyme transfers the group to another acceptor, thereby transforming it into its corresponding donor. We assume that a half-reaction follows linear reversible kinetics. This results in Michaelis-Menten–like kinetics for the transfer of a group from one metabolite to another. For the initial network we assume that enzymes are specific with regard to the group but unspecific with regard to donors and acceptors, i.e., they transfer a specific biochemical group from any donor to any acceptor.
(C) The initial network consists of 2N metabolites. A single half-reaction transforms the metabolites into one of the N neighbours, such that the initial network resembles an N-dimensional hypercube. Note, however, that for the transfer of a group, two edges (i.e., half-reactions) in the cube have to be coupled.
(D) To study the impact of group transfer on the connectivity distribution of the simulated networks, we performed simulations with monomolecular reaction networks, in which groups are added and removed, but not transferred, by the enzymes. The initial network has a similar structure. However, in contrast to the group transfer network, each reaction can be performed without coupling to another reaction.
We assume that initially all enzymes and all transporters are unspecific and allow small mutations in their kinetic properties. These mutations increase the rate constant of a single reaction while decreasing the rate constants of all other reactions of an enzyme or transporter. Alternatively they decrease the rate constant of a single reaction while increasing the rate constants of the other reactions. In addition to mutations affecting kinetic properties, we assume that there are gene duplications (or deletions) that increase (decrease) the dosage of enzymes and transporters. Changing the dosage of an enzyme has two opposing effects on fitness: An increased dosage increases the fluxes in the network, but it is also associated with increased metabolic costs of biomass formation. Duplications also enable enzymes to diverge and specialize on different reactions. Note that if a multifunctional enzyme has only a small impact on biomass formation, duplication can be detrimental provided that the metabolic costs of the additional copy are larger than the benefits in terms of increased growth rate. Consequently, enzymes and transporters do not necessarily specialize completely over the course of evolution. On the other hand, duplications can be beneficial if enzymes or transporters are already completely specialized. Therefore multiple identical copies of completely specialized enzymes and transporters can evolve provided they have a large impact on biomass formation.
In our simulations, we assume that there are seven different biochemical groups that characterize each metabolite. Because a metabolite can either carry one or no copies of each group, we have an initial network of 128 metabolites. We initiate the simulations with seven unspecific enzymes that catalyze the transfer of one of the biochemical groups from any donor to any acceptor, and with a single unspecific transporter that catalyzes the transport of all metabolites. Sixteen randomly chosen metabolites are involved in biomass formation. For the production of biomass, it is necessary to have at least one donor and acceptor of each group as external metabolites. We therefore assume that a minimal set of two metabolites, namely X0 and X127, is present in the environment. For reasons of computational tractability, we let the network evolve by fixation of the most beneficial mutations rather than random mutations. Duplications and deletions, because we assume that they are rare compared to mutations that alter the kinetic properties, are only considered if no further beneficial kinetic mutations are possible. A detailed description of the computer simulations is given in Materials and Methods.
Results
A sample simulation is shown in Figure 2. The simulation indicates that the scenario proposed by Kacser and Beeby [5] indeed leads to the evolution of metabolic networks with specialized enzymes (Figure 2A). The emerging network consists of 23 different reactions and seven transport processes that are catalyzed by fully specialized enzymes and transporters (Figure 2C). The enzymes and transporters are present in multiple identical copies. The network contains only a fraction of the metabolites present in the initial network (33 of 128). The remaining metabolites become unconnected as the reactions specialize, i.e., they do not serve as substrate or product of reactions present in the emerging network. Figure 2B shows the connectivity distribution of the metabolites in the emerging network. The connectivity includes participation in enzymatic reactions, transport processes, and biomass formation. Most metabolites have a connectivity of two. In this simulation there are no metabolites that participate in only one reaction or transport process. Some of the metabolites, such as X0, X16, X18, X126, and X127, are involved in a comparably large number of reactions. These metabolites “monopolize” the transfer of specific biochemical groups. One of the groups, for example, is always transferred from metabolite X127 to a number of different acceptors. Hence, it seems to be beneficial that at least some of the metabolites specialize in the transfer of specific groups by acting as the only donor or acceptor.
Figure 2 Sample Simulation for the Evolution of Metabolic Networks
(A) The initial network consists of 128 metabolites, seven unspecific enzymes (each of which transfers one of the seven biochemical groups that metabolites carry), and a single unspecific transporter. Within the course of evolution, the enzymes and transporter duplicate and increase in specificity (i.e., the number of half-reactions per enzyme and of metabolites per transporter decreases). The emerging network consists of 23 enzymatic reactions and seven transport processes. In the sample simulation, all enzymes and transporters in the emerging network are highly specific, i.e., the enzymes catalyze only two half-reactions, and the transporters transport single metabolites. The emerging network contains only 33 metabolites. The remaining metabolites are not involved in the emerging network.
(B) Connectivity distribution of the emerging group transfer network. Most metabolites are involved in only two reactions. However, a few metabolites are highly connected.
(C) Pathway scheme of the emerging group transfer network. The metabolites X0 and X127 are taken up from the environment, whereas metabolites X4, X22, X94, X95, and X111 are excreted into the environment (white boxes). The network eventually transforms metabolites X0 and X127 into those metabolites that are involved in biomass formation (grey boxes). Interestingly, metabolite X4 is excreted although it is involved in biomass formation. Note that some half-reactions evolve, such as the one from X127 to X126, and monopolize the transfer of a specific group (in this case the first group in the binary string). These metabolites are involved in many reactions and therefore have high connectivity. The group transfer reactions of these hubs are summarized in the first line of the pathway scheme. The emerging group transfer network is much more complex than the corresponding monomolecular reaction network (see Figure 2D) and even includes a cycle (X32 → X119 → X117 → X32), with the net reaction of X0 + X16 + X127 → X18 + X40 + X85).
(D) Pathway scheme of an emerging monomolecular reaction network. Although the network produces the same set of metabolites involved in biomass formation as the group transfer network shown in Figure 2C, it has a much simpler, tree-like structure. Additionally, the network has no obvious hubs.
To obtain more precise data for the connectivity distribution, we analysed 40 replicate simulations. The results of these simulations are similar to the sample simulation shown in Figure 2. Most of the enzymes and transporters of the emerging networks specialize completely, i.e., catalyze two half-reactions or the transport of a single metabolite. However, a small fraction of the enzymes in the emerging networks catalyze more than two half-reactions, i.e., are not fully specialized. These enzymes have only a weak impact on the rate of biomass formation and therefore cannot duplicate and specialize fully (see Assumptions). Most of the fully specialized enzymes and transporters are present in multiple identical copies. As in the sample simulation, the emerging networks contain only a small fraction of the metabolites present in the initial networks. Most of these metabolites are involved at least in two reactions or transport processes. Occasionally we observe metabolites that participate in only a single reaction (see Assumptions). Figure 3A shows the connectivity distribution over all emerging networks. Metabolites that are participating in fewer than two reactions are not included in the graph. The average maximum likelihood estimate [10] for the power-law coefficients of the individual networks is γ = 2.63 (±0.17). However, the overall connectivity distribution deviates strongly from a power law (Figure 3B). In particular, the overall distribution has more hubs than expected in the connectivity range between eight and 12, but has fewer hubs than expected with a connectivity of more than 12. This disagreement between the emerging networks and natural networks may arise because the networks emerging in our simulations are much smaller than the networks analyzed in the literature [2,3,11]. We therefore tested whether the simulated networks are consistent with subnetworks of similar size derived from real metabolic networks. These networks were obtained from the KEGG database [12]. The KEGG database displays several functional metabolic subnetworks such as glycolysis, the tricarboxylic acid cycle, or amino acid synthesis for a wide range of organisms. We analyzed 22 subnetworks fromEscherichia coli, each containing between 24 and 40 metabolites. The overall connectivity of these networks is shown in Figure 3C. Note, that in order to compare the connectivity distributions of the simulated and the natural networks, we treat the natural networks as independent networks, i.e., a metabolite that is present in two different networks is treated as two different metabolites. As in the simulated networks, we consider only metabolites that are involved in more than one reaction. A comparison of the connectivity distribution of the natural and the emerging networks indicates that both connectivity distributions are not significantly different (chi square test: p ≍ 0.085; see Materials and Methods). This result indicates that although the connectivity of the networks emerging in our simulation is not consistent with that of large metabolic networks described in the literature, it is similar to the connectivity observed in natural metabolic subnetworks of the same size. Note, however, that these subnetworks typically do not produce all compounds required for biomass formation and therefore have been selected for functions different from the networks emerging in our simulations.
Figure 3 Connectivity Distribution of Simulated and Natural Networks
(A) Connectivity distribution of 40 different simulated networks. As in the sample simulation, most metabolites are only involved in two or three reactions. On the other hand, the distribution shows that there are a number of hubs that are involved in up to 16 reactions.
(B) The double-logarithmic plot of the distribution reveals a considerable deviation from a power-law distribution.
(C) Connectivity distribution in natural metabolic networks of similar size. The networks are generated from the KEGG database for E. coli (see Materials and Methods). The connectivity distribution of the simulated networks is consistent with that of the natural networks (chi square test: p ≍ 0.085; see Materials and Methods).
(D) Simulations for monomolecular reaction networks. The initial network is equivalent to the group transfer networks (see Figure 1C). In contrast to the group transfer network, we assume that groups are not transferred, but are added or removed by the enzymes. Therefore a reaction can transform a donor into its corresponding acceptor without being coupled to another half-reaction. The connectivity distribution of the emerging networks clearly differs from the connectivity distribution in the bimolecular group transfer networks. In particular, hubs are comparatively rare. This indicates that group transfer plays a key role in the emergence of hubs.
To analyze whether group transfer plays a major role in the emergence of hubs, we performed additional simulations with monomolecular reaction networks. In contrast to the bimolecular group transfer networks described above, we assumed that groups are not transferred between metabolites, but are simply added or removed by enzymes. The topology of the resulting initial network is similar to the group transfer network (see Figure 1C and 1D). However, in contrast to group transfer networks, in monomolecular networks there are no half-reactions. For simulating the evolution of monomolecular networks, we used assumptions analogous to those used for the simulations of the bimolecular group transfer networks. An example for an emerging monomolecular network is shown in Figure 2D. Note that this network produces the same set of metabolites involved in biomass formation as the network shown in Figure 2C. However, the resulting monomolecular reaction network differs remarkably from the bimolecular reaction network. In contrast to the group transfer reactions, monomolecular reactions have only a single substrate and product. Therefore monomolecular networks can be displayed as graphs that typically have a tree-like structure (see Figure 2D). The connectivity distributions of the simulated monomolecular networks clearly differs from the simulated group transfer networks (Figure 3D). In particular, monomolecular networks have much fewer hubs.
Our simulation results indicate that group transfer plays a key role in the emergence of hubs. Within the evolution of group transfer networks, hubs emerge because some metabolites become a preferential donor or acceptor of a specific group, i.e., they have a particularly high group transfer potential. Therefore it is of advantage that an enzyme that transfers this specific group increases its specificity towards this donor (or acceptor). Once a preferential donor (or acceptor) emerges and the corresponding enzyme starts to specialize towards this metabolite, the remaining enzymes in the network are selected for further increasing the group transfer potential of this specific metabolite. Thereby a single metabolite may “monopolize” this group transfer reaction. It serves as the only donor (or acceptor) of a specific group towards a number of different acceptors (donors) in the network. As a consequence it participates in a number of different reactions. In the absence of group transfer reactions, i.e., in monomolecular reaction networks, this mechanism does not work. Therefore, hubs do not emerge in monomolecular reaction networks.
Discussion
In summary, our simulations demonstrate that metabolic networks may have emerged according to the scenario of Kacser and Beeby [5]. Starting from metabolic networks with multifunctional enzymes, networks emerge that consist of highly specialized enzymes. The key evolutionary mechanism in this scenario is the duplication of enzymes followed by the specialization of the copies for different biochemical reactions. The implication of this mechanism is that during the early evolution of metabolic networks, biochemical reactions and intermediate metabolites have been lost.
On the basis of our model, we expect that multifunctional enzymes can be maintained if they have only little impact in biomass formation and are therefore present as single copies. Alternatively, multifunctional enzymes can be maintained if there is no trade-off between the different functions. Recent results on the evolution of multifunctional enzymes indicate that this may at least be sometimes the case [13].
Although the overall connectivity of the networks emerging in our simulations differs clearly from a power-law distribution, it is consistent with the connectivity distribution derived from natural subnetworks of similar size. Our simulations indicate that the scenario of Kacser and Beeby leads to the emergence of hub metabolites. Additionally, our results predict that group transfer is essential for the emergence of hubs in these networks. This prediction is in line with the observation that most hubs in natural networks such as ATP, NADH, glutamate, and coenzyme A are key compounds in the transfer of biochemical groups. The high connectivity of these metabolites results mainly from transferase reactions rather than other reaction types (such as isomerase reactions).
In our simulations, hubs emerge as a consequence of selection for growth rate. In addition, along with a number of further publications given below, our simulations therefore support the view that hubs are not necessarily a result of selection for robustness as has been suggested previously [1,14]. First, it has been shown that the evolution of robustness against complete knock-outs of genes, as for example by duplicates, requires very specific conditions in terms of mutation rates, gene functions, and interactions [15]. Second, a recent study on robustness and enzyme indispensability in yeast metabolism indicates that the apparent dispensability of many enzymes is not due to network robustness, but to the fact that many enzymes are only required under specific environmental conditions [16]. Third, robustness against environmental changes is also unlikely to explain the connectivity distributions observed in natural networks. This is because power-law connectivity distributions have been observed in a wide range of organisms living in very different environments, including, for example, intercellular parasites that live in very stable environments [3]. Finally, no evolutionary scenarios have been presented that demonstrate that selection for increased robustness leads to the emergence of metabolic networks with power-law connectivity. Nevertheless, it has been shown that robustness of metabolic networks against partially deleterious mutations (such as a 50% reduction of gene dosage due to a defective allele in diploid organisms) may emerge as a “side product” of selection for growth rate [17,18]. It can therefore be expected that the networks emerging in our simulations are robust against slightly deleterious mutations.
Materials and Methods
Metabolites
Metabolites are assumed to consist of seven different biochemical groups. Each biochemical group is either present once or is absent. Thus there are 27 different metabolites that are characterized by a binary string of length 7, where “1” at position x denotes the presence of group x, whereas “0” denotes the absence of the group. Metabolites are associated with a random free energy that is taken from a uniform distribution between zero and one, and is required to specify thermodynamic properties of the biochemical reactions. For the production of biomass, it is necessary to have at least one donor and acceptor of each group as external metabolites. Thus a minimal set of two metabolites, namely X0 (characterized by the string 0–0–0–0–0–0–0) and X127 (characterized by the string 1–1–1–1–1–1–1), is assumed to be present in the environment. Sixteen randomly chosen metabolites are involved in biomass formation.
Enzymes
Enzymes catalyze the transfer of a specific biochemical group. We assume that groups are transferred by a “ping-pong mechanism” (see Figure 1): A donor of a group transfers the group to the appropriate enzyme and is thereby transformed into its corresponding acceptor. The enzyme then transfers the group to an acceptor, thereby transforming it into its corresponding donor. Thus an enzyme can be in two possible states, Ei
(1) and Ei
(0), with Ei
(1)
+ Ei
(0)
= Ei , where Ei is the total dosage of enzyme i,
Ei
(0) is the concentration of the enzyme without its group being bound to it, and Ei
(1) is the concentration of enzyme with its group being bound to it. The free energy difference between both states of an enzyme is assumed to be a random value taken from a uniform distribution ranging from zero to one. We further assume that in principle all metabolites that contain a specific biochemical group (i.e., half of the 27 metabolites) can serve as a donor for group transfer, whereas all metabolites that do not contain the group can in principle serve as an acceptor. We assume linear kinetics for the transfer of a group to an enzyme, given by v = kij (Ei
(0)
X
(ij)
(1)
−Ei
(1)
X
(ij)
(0)
/qij), where kij is the rate constant of the reaction j of an enzyme i,
X
(ij)
(1) is the concentration of the donor of the reaction, X
(ij)
(0) is the concentration of the corresponding acceptor, and qij is the equilibrium constant of the reaction resulting from the free energies of the reactants. For the transfer of a group from a donor to an acceptor, two half-reactions need to be coupled. This results in Michaelis-Menten–like kinetics and implies that functional enzymes need to maintain nonzero rate constants for at least two reactions. We assume that initially the enzymes are unspecific and transform groups from each donor to each acceptor with the same rate constant. The initial dosage of enzymes is Ei = 1.
Transporters
We further assume that transporters transport metabolites passively across the cell membrane. The rate of transport is given by v = Titij(Xj−Xjext), where Ti is the dosage of the transporter i,
tij is the rate constant of the transport of metabolite j,
Xj is the metabolite concentration in the cell, and Xjext is the metabolite concentration in the environment. We assume that in the initial network there is a single transporter that transports all metabolites with the same rate constant. The initial dosage of the transporter is T
0
= 1.
Biomass formation, growth, and network fitness
Biomass is formed by the condensation of specific metabolites. The rate of biomass formation follows linear kinetics given by the product vBM = kBMΠiXi over all metabolites Xi that are involved in biomass formation. The rate constant is set to kBM = 1 in all simulations. We assume that the formation of biomass leads to growth. The growth rate is given by W = 1/VdV/dt = vBM /(C
0
+CEE+CTT), where C
0 is the amount of biomass that is required for structural compounds (i.e., those compounds that are not directly involved in cellular metabolism), CE is the amount of biomass per enzyme, CT is the amount of biomass per transporter, E is the total dosage of enzymes, and T is the total dosage of transporters [5]. The parameters C
0, CE, and CT are set to 50, 1, and 1, respectively. Note that due to cell growth, metabolites are constantly diluted at a rate equal to the growth rate. The fitness of a network is assumed to be proportional to the steady-state growth rate.
Tradeoff between specificity and catalytic activity
In line with the scenario of Kascer and Beeby, we assume that reactions can either catalyze a large number of reactions but with low activity for each enzyme, or a lower number or reactions with improved catalytic activities. Specifically, we assume that the sum Σj kij
1/α and Σj tij
1/α over all rate constants kij or tij of an enzyme or transporter, i, respectively, is constant. For values of α > 1 increasing the rate constant for a single reaction has an over-proportional effect on all other rate constants. In our simulations we use Σj kij
1/α
= 1, Σj tij
1/α
= 1, and α
= 2. This implies, for example, that a transporter catalyzing the transport of a single metabolite has a four times higher rate constant for this reaction than a transporter that is specialized on the transport of two metabolites.
Mutations and the course of evolution
We assume that there are two types of mutations: (1) mutations that change the kinetic properties of an enzyme and (2) mutations that change the number of copies of an enzyme, i.e., gene deletions and duplications. For the first type of mutation we assume that the value of kij
1/α or tij
1/α
, respectively, for a single reaction is either increased or decreased by a small value of m = 0.01, while the rate constants of the other reactions are decreased, or increased appropriately. Gene deletions decrease the dosage of an enzyme. Duplications increase the dosage of an enzyme. In our simulations we calculate the effect of all possible mutations of the current network on the steady-state growth rate to obtain the mutant with maximal increase in fitness. This mutation is then assumed to become fixed and the resulting network is used to search for the next mutations. The steady states are calculated using the NAG library function c05nbc (see http://www.nag.co.uk/numeric/CL/CLdocumentation.asp for documentation) for root finding of systems of nonlinear equations. The solver may fail (or may obtain steady states with negative concentrations) if the initial approximation of the steady state is far away from the actual steady state of the system. In this case we integrated the system towards its steady state (using the NAG library function d02ejc) until the root finder succeeds. To calculate the effect of a mutation on the steady state, we solve the mutated system using the metabolite concentrations of the non-mutated system as initial approximation, because for small mutations, the steady state of the mutated systems can be expected to be close to the non-mutated system. Note, however, that in a system of nonlinear equations there may be multiple steady states, unstable steady states, and saddle points. We therefore test whether the steady state of the most beneficial mutation obtained by the above procedure (i.e., the mutation that is assumed to become fixed) is reached by integration from initial metabolite concentrations close to zero. This test failed in ten out of 50 simulations. These simulations are excluded from further analysis. Gene duplications and deletions are assumed to be rare compared to mutations affecting the catalytic properties of enzymes and transporters. We therefore consider these types of mutation only if none of the mutations affecting kinetic properties are beneficial. A simulation ends if there are no beneficial kinetic mutations nor gene duplications or deletions.
Analysis of E. coli metabolic subnetworks
We analyzed the connectivity in metabolic subnetworks of E. coli downloaded from the KEGG database (ftp.genome.ad.jp/pub/kegg/ pathways/eco/) on the basis of the annotation in the reaction database (ftp.genome.ad.jp/ pub/kegg/ligand/reaction). Metabolites participating in fewer than two reactions are excluded. For comparison with the networks emerging in our simulation, we included 22 subnetworks of similar size, i.e., with between 24 and 40 metabolites. To compare the connectivity distributions, we use Pearson's chi square test. Data from the tail of the distributions (k > 11) are put into a single class.
Simulation of monomolecular networks
For simulating the evolution of monomolecular networks, we assume that enzymes add or remove biochemical groups, thereby transforming a donor of a group into its corresponding acceptor and vice versa. In contrast to bimolecular reactions, it is not required to couple two half-reactions. Completely specialized enzymes catalyze a single reaction rather that two half-reactions. The rate of a reaction is given by v = Ei kij(X
(ij)
(1)
−X
(ij)
(0)
/qij), where kij is the rate constant of the reaction j of an enzyme i, Ei is the dosage of the enzyme, X
(ij)
(1) is the concentration of the donor, X
(ij)
(0) is the concentration of the corresponding acceptor, and qij is the equilibrium constant of the reaction resulting from the free energies of the reactants. All other assumptions are equal to those used in the simulations of the bimolecular group transfer networks.
The authors would like to thank A. Pohorille, M. Salathé, and A. Scherer for helpful discussion and the Gebert Rüf foundation for financial support.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. TP, OSS, and SB conceived and designed the experiments. TP performed the experiments. TP, OSS, and SB analyzed the data. TP wrote the paper.
Citation: Pfeiffer T, Soyer OS, Bonhoeffer S (2005) The evolution of connectivity in metabolic networks. PLoS Biol 3(7): e228.
==== Refs
References
Jeong H Tombor B Albert R Oltvai ZN Barabasi AL The large-scale organization of metabolic networks Nature 2000 407 651 654 11034217
Wagner A Fell DA The small world inside large metabolic networks Proc R Soc Lond B Biol Sci 2001 268 1803 1810
Ma H Zeng AP Reconstruction of metabolic networks from genome data and analysis of their global structure for various organisms Bioinformatics 2003 19 270 277 12538249
Arita M The metabolic world of Escherichia coli is not small Proc Natl Acad Sci U S A 2004 101 1543 1547 14757824
Kacser H Beeby R Evolution of catalytic proteins or on the origin of enzyme species by means of natural selection J Mol Evol 1984 20 38 51 6429341
Schmidt S Sunyaev S Bork P Dandekar T Metabolites: A helping hand for pathway evolution Trends Biochem Sci 2003 28 336 341 12826406
Horowitz NH On the evolution of biochemical synthesis Proc Natl Acad Sci U S A 1945 31 153 157 16578152
Jensen RA Enzyme recruitment in evolution of new function Annu Rev Microbiol 1976 30 409 425 791073
Copley RR Bork P Homology among (betaalpha)(8) barrels: Implications for the evolution of metabolic pathways J Mol Biol 2000 303 627 641 11054297
Goldstein ML Morris SA Yen GG Fitting to the power-law distribution Eur Phys J B 2004 41 255 258
Barabasi AL Albert R Emergence of scaling in random networks Science 1999 286 509 512 10521342
Kanehisa M Goto S Kawashima S Okuno Y Hattori M The KEGG resource for deciphering the genome Nucleic Acids Res 2004 32 D277 D280 (Database issue) 14681412
Aharoni A Gaidukov L Khersonsky O McQ Gould S Roodveldt C The ‘evolvability' of promiscuous protein functions Nat Genet 2005 37 73 76 15568024
Albert R Jeong H Barabasi AL Error and attack tolerance of complex networks Nature 2000 406 378 382 10935628
Nowak MA Boerlijst MC Cooke J Smith JM Evolution of genetic redundancy Nature 1997 388 167 171 9217155
Papp B Pal C Hurst LD Metabolic network analysis of the causes and evolution of enzyme dispensability in yeast Nature 2004 429 661 664 15190353
Kacser H Burns JA The molecular basis of dominance Genetics 1981 97 639 666 7297851
Hurst LD Randerson JP Dosage, deletions and dominance: Simple models of the evolution of gene expression J Theor Biol 2000 205 641 647 10931758
| 16000019 | PMC1157096 | CC BY | 2021-01-05 08:28:15 | no | PLoS Biol. 2005 Jul 28; 3(7):e228 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030228 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030243SynopsisBioinformatics/Computational BiologyGenetics/Genomics/Gene TherapySystems BiologySchizosaccharomymesTranscriptional Waves in the Yeast Cell Cycle Synopsis7 2005 28 6 2005 28 6 2005 3 7 e243Copyright: © 2005 Public Library of Science.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
The Cell Cycle-Regulated Genes of Schizosaccharomyces pombe
==== Body
Mobilizing an army to march into battle requires the increased activity of hundreds of people, from the quartermaster to the gunnery captain. The stately march of the cell cycle—from the first growth phase, through DNA synthesis, to the second growth phase, and on to mitosis and cell division—also demands increased activity, but of hundreds of genes, from histones to protein kinases. And just as the army must coordinate the shipment of C rations with the movement of its troops, so must the cell coordinate its genetic activities to ensure that raw materials and regulatory molecules are present where and when they are needed. In this issue, Janet Leatherwood, Bruce Futcher, and colleagues describe the waves of gene activity that accompany the phases of the cell cycle in the yeast Schizosaccharomyces pombe.
Using microarrays, the authors examined the expression level of 5,000 genes over the course of the cell cycle. They found that well over 2,000 of these genes undergo slight but observable and statistically meaningful oscillations. Of these, they chose to examine the top 750, an admittedly arbitrary cutoff that nonetheless highlights those whose expression levels rise and fall the most. They identified two broad waves of oscillation, one peaking in early to mid-G2 (the second growth phase) and the other late in G2 at the transition to mitosis. These two peaks were seen even in the 4,000 least cyclic genes, suggesting that many genes may be slightly upregulated not for adaptive purposes, but simply because some transcription factors inevitably go astray whenever there are lots of them around.
Such broad waves of upregulation are likely due to a simultaneous increase in the activity of multiple clusters of genes, each controlled by separate groups of transcription factors. A variety of cell culture manipulations allowed the researchers to identify eight clusters of genes, the activity of whose members was tightly co-regulated. (In this case, “cluster” refers not to genes physically grouped together on a chromosome, but to genes that are regulated similarly.) Scouring the promoters of these genes confirmed that each cluster was characterized by unique transcription factor binding sites. They also discovered that, as a group, these promoters tended to be longer than average, suggesting they may be more complex than those in non-oscillating genes.
The number of genes within each cluster ranged from only a few to over 100. The largest of them, the Cdc15 cluster, contains genes involved in mitosis, cytokinesis, and formation of the septum that separates the daughter cells, as well as genes for other functions. Other clusters regulate DNA replication, cell separation, synthesis of the histone proteins that act as spools on which DNA is wound, protein folding and stress response, ribosome biogenesis, and other aspects of the cell cycle.
Two other recent studies in S. pombe have found broadly similar patterns, and have identified 407 and 747 genes, respectively, as strong oscillators. There was a heartening degree of overlap, with 171 genes identified by all three studies, and 360 more found in two of the three. Even genes that made the cut in only one study were found likely to oscillate in the other two. Follow-up studies to further explore genes that are coordinated during the cell cycle march should help us to understand how the army of molecules exerts such fine control over both normal and abnormal cell growth and proliferation.
| 0 | PMC1157097 | CC BY | 2021-01-05 08:21:25 | no | PLoS Biol. 2005 Jul 28; 3(7):e243 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030243 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030261SynopsisBioinformatics/Computational BiologyEvolutionNoneSimple Rules Reproduce a Hub-Shaped Cell Metabolic Network Synopsis7 2005 28 6 2005 28 6 2005 3 7 e261Copyright: © 2005 Public Library of Science.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
The Evolution of Connectivity in Metabolic Networks
==== Body
A map of a cell's metabolic pathways looks like an airline route map on steroids, with hundreds of reactions forming a complex and interconnected network. And just as a few cities serve as destination hubs for many different flights, a few metabolites, such as ATP and NADH, form biochemical hubs within the metabolic network. These molecules are involved in far more reactions than the average, and thereby serve to couple otherwise unrelated reactions within the cell. How did this hub-shaped network arise? In this issue, Thomas Pfeiffer and colleagues employ a computer simulation of a simplified metabolic system to show that two key features in the evolution of a hub system are enzyme specialization and the transfer of chemical groups between metabolites.
The authors created “molecules” from all possible combinations of seven “groups.” They began the simulation with seven “enzymes” that catalyze the transfer of one group from one molecule to another. Initially, each enzyme was a generalist—it could take a group from any molecule and donate it to any other. This mirrors one plausible scheme for the actual evolution of cellular biochemistry. Over the course of the simulation, enzymes could mutate to preferentially increase their affinity for one substrate at the expense of others. The number of enzyme types could be increased by “gene duplications.” Other parameters allowed the simulated cell to take up and excrete metabolites, and to grow.
As the system evolved, enzymes proliferated and became more specialized, until the final mix included about two dozen enzymes, each of which catalyzed only one or two reactions. As a consequence, some metabolites fell out of use, and the final number of metabolites dropped from 128 to 33. While most took part in only two or three reactions, a few emerged as hubs, participating in eight or more separate reactions. While this mathematical distribution did not match that found in the metabolic network of a whole real cell, it did approximate that of similar-sized sub-cellular metabolic networks. The central importance of group transfer to this structure was brought out when the authors reran the simulation without group transfer. When reactions simply added or removed a group, without transferring it to another molecule, a much simpler network without hubs evolved instead.
For modeling purposes, metabolites can be denoted as binary strings of biochemical groups; enzymes catalyze the transfer of the groups
This is not the last word on biochemical evolution, but it does show how an initially generalist metabolism can evolve into the highly specialist system found in all existing cells. Further experiments that model known reactions more closely may be useful in elucidating more details of the evolutionary process that led to the emergence of the biochemical “hubbub” that characterizes life today.
| 0 | PMC1157098 | CC BY | 2021-01-05 08:21:24 | no | PLoS Biol. 2005 Jul 28; 3(7):e261 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030261 | oa_comm |
==== Front
BMC BiotechnolBMC Biotechnology1472-6750BioMed Central London 1472-6750-5-141591068510.1186/1472-6750-5-14Methodology ArticleSimultaneous imaging of GFP, CFP and collagen in tumors in vivo using multiphoton microscopy Sahai Erik [email protected] Jeffrey [email protected] Ulrike [email protected] Jeffrey E [email protected] Frank [email protected] John [email protected] Tumour Cell Biology Laboratory, Cancer Research UK London Research Institute, London WC2A 3PX, UK2 Department of Anatomy and Structural Biology, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA3 Department of Biology, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Cambridge, MA 02139, USA4 Analytical Imaging Facility, Albert Einstein College of Medicine, 1300 Morris Park Avenue, Bronx, NY 10461, USA2005 23 5 2005 5 14 14 18 2 2005 23 5 2005 Copyright © 2005 Sahai et al; licensee BioMed Central Ltd.2005Sahai et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The development of multiphoton laser scanning microscopy has greatly facilitated the imaging of living tissues. However, the use of genetically encoded fluorescent proteins to distinguish different cell types in living animals has not been described at single cell resolution using multiphoton microscopy.
Results
Here we describe a method for the simultaneous imaging, by multiphoton microscopy, of Green Fluorescent Protein, Cyan Fluorescent Protein and collagen in vivo in living tumors. This novel method enables: 1) the simultaneous visualization of overall cell shape and sub-cellular structures such as the plasma membrane or proteins of interest in cells inside living animals, 2) direct comparison of the behavior of single cells from different cell lines in the same microenvironment in vivo.
Conclusion
Using this multi-fluor, multiphoton technique, we demonstrate that motility and metastatic differences between carcinoma cells of differing metastatic potential can be imaged in the same animal simultaneously at sub-cellular resolution.
==== Body
Background
The ability to image cell movements and dynamic changes in sub-cellular structures in living mammals will significantly enhance our understanding of biology. Many applications have been developed to investigate how tumor cells act in vivo [1-3]. In recent years multi-photon laser scanning microscopy has demonstrated that it has both the resolution and tissue penetration to significantly improve the analysis of tumor cell behavior in vivo [4-6]. A key requirement for multiphoton microscopy is the need to fluorescently label cells, sub-cellular compartments or proteins of interest. Green Fluorescent Protein (GFP) has been widely used to label cells; however the use of a single fluorophore, though successful, has been limiting. Tumor development and cell migration from intravasation to metastatic growth has been studied in vivo in orthotopic models and transgenic mice [7] using GFP, but limited by the fact that only one cell type can be examined without introducing a dye from an external source. GFP and RFP have been used to study separate cell populations in tumors using conventional imaging methods [8-10] and multiple flours have been used in cells in vitro [11]. However, this combination of fluorophores are not compatible with multiphoton intravital imaging, since the high intensity pulsed infrared lasers commonly used for multiphoton microscopy produce light in the 720–980 nm range and are unable to excite RFP efficiently in deep tissue. While we and other groups have been able to use multiphoton laser scanning microscopy to image two or more chromophores in vivo, generally non genetically coded fluors such as Texas Red-labeled dextran or Hoechst are used [6,12-15] in conjunction with GFP or other fluorescent proteins which prevents the application of multiphoton microscopy from being applied to the study of multiple cell populations in vivo. Therefore, here we describe methods to image the genetically encoded fluors GFP and Cyan Fluorescent Protein (CFP) simultaneously in a living tissue. Previous work by our group has correlated patterns in gene expression in cells with differing metastatic potential with differences in cell motility and polarization in vivo [6,16]. Here we describe a method to compare the behavior of cancer cells in which expression of genes identified in these studies has been altered with the behavior of control cells in the same tumor micro-environment. We also describe a method for imaging two genetically encoded fluorophores in the same cell, thereby allowing imaging of sub-cellular compartments or proteins and the whole cell simultaneously.
Results and Discussion
To study two cell types in the same organ in vivo with differing fluorescent proteins, fluorescent pairs need to be chosen that can be excited equally at a common wavelength. We chose to simultaneously image GFP and CFP because their expression is well tolerated by most cell types and they are easily excited by standard Ti-sapphire lasers. Initial attempts to image RFP were not successful; both tetrameric and dimeric variants of dsRed formed aggregates which had deleterious effects on cell viability. The use of monomeric mRFP overcame these problems, however the power output of most Ti-Sapphire lasers drops off significantly at the wavelengths of light required to excite mRFP resulting in ineffective excitation of mRFP. While YFP has been used as part of a FRET pair in vivo [17], GFP was chosen over YFP because YFP is not effectively excited by wavelengths below 900 nm [18], YFP and GFP excitation closely mirror each other and are difficult to separate, and a large number of already available cell lines and transgenic animals use GFP.
Using a Biorad Radiance 2000 multiphoton microscope with an inverted Olympus IX70 connected to a Spectra Physics Tsunami Ti-Sapphire laser, we first determined the best 2-photon excitation wavelength by measuring the fluorescence emitted by GFP and CFP when excited by different wavelengths. The excitation of CFP was most effective around 850 nm and declined at longer wavelengths, whereas excitation of GFP was still increasing at 880 nm with a maximum at 960 nm (Figure 1a and data not shown). These data are in agreement with previously described excitation cross sections for GFP and CFP ( and [18]). In conclusion, wavelengths around 880 nm were most suitable for simultaneous imaging of GFP and CFP because 880 nm is very close to the optimal wavelength for exciting CFP and is also able to excite GFP.
Figure 1 2-photon excitation and emission properties of GFP and CFP expressed in cells. A) Stable cell lines expressing either CFP or GFP were imaged with the indicated excitation wavelength (in all cases the laser power was 0.95–1.0W). CFP fluorescence was captured using a 480/30 filter and non-descanned detection, GFP fluorescence was captured using a 515/30 filter and non-descanned detection. The amount of fluorescence captured using different excitation wavelengths is shown (relative to the maximum fluorescence). B) Stable cell lines expressing either CFP or GFP were imaged using an 880 nm excitation beam (in all cases the laser power was 0.95–1.0W). Fluorescence was captured using the indicated filters and non-descanned detection. The amount of GFP or CFP fluorescence captured using different filters is shown (relative to the maximum fluorescence).
Having established that both GFP and CFP can be effectively excited at 880 nm we then sought establish a simple method for separating the fluorescence emitted by the different proteins. We tested three filters for their ability to pass light emitted by cells expressing either GFP or CFP. Figure 1B shows that GFP fluorescence passed through the 515/30 nm filter (median wavelength/band width) but did not pass through the 450/80 nm and 480/30 nm filters. In contrast, CFP fluorescence passed efficiently through all three filters. Therefore, both GFP and CFP fluorescence will pass through a 515/30 filter whereas 450/80 and 480/30 filters will only pass CFP fluorescence. Subtraction using the 'Image Calculator' feature on Image J (available from ) of the CFP fluorescence captured using either a 450/80 or 480/30 filter from the mixture of GFP and CFP fluorescence passing through a 515/30 filter should yield the GFP fluorescence. To test this we tried to image either co-cultures of GFP and CFP expressing cells or cells co-expressing GFP and CFP constructs. To facilitate the subtraction we adjusted the gain settings during image acquisition to ensure that a CFP expressing cell produced the same pixel intensity when captured using either the 450/80, 480/30 or 515/30 filters. Figure 2 shows that subtraction of the image captured using the 450/80 filter from that captured the 515/30 filter leaves only the GFP expressing cells visible (Figure 2 left hand panels). Furthermore, if the same subtraction is performed on images of cells co-expressing GFP targeted to the plasma membrane (GFP-CAAX) and CFP in the cytoplasm. After subtraction of the CFP signal, the distribution of GFP signal is indistinguishable from GFP images of cells expressing only GFP-CAAX (Figure 2 middle panels and data not shown). GFP-CAAX also localized to some intracellular structures; it is possible that this may be the Golgi apparatus as Michaelson et al have reported that a similar construct localized to both the plasma membrane and the Golgi apparatus [19].
Figure 2 Simultaneous capture of GFP and CFP. Stable cell lines expressing either GFP or CFP were co-cultured (left panels) or a stable cell line expressing CFP and GFP-CAAX (membrane targeted – middle panels) or a cell line expressing CFP transiently transfected with GFP-ZYXIN (right panels) were analyzed. Top panels show the fluorescence captured using a 450/80 filter and non-descanned detection. Upper-mid panels show the fluorescence captured using a 515/30 filter and non-descanned detection. Lower-mid panels show the result of subtracting the 450/80 signal from the 515/30 signal. Bottom panels show a merge of the top panels (CFP signal in blue) and the lower-mid panels (GFP signal in green). GFP-CAAX, which generally targets to the membrane can be seen labeling of intracellular structures in the merge and GFP images (middle panel; arrow). All images were taken with a 40× objective and a final magnification of 500× for the right 2 columns and 1000× for the left. Scale bar = 25 μm for the right 2 columns and 10 μm for the left.
To test if this method can be used to monitor protein localization, we transiently transfected cells expressing CFP with a GFP-ZYXIN expression construct. Figure 2 (right hand panels) shows one such transfected cell amongst other non-transfected cells, subtraction of the 450/80 signal from the 515/30 signal clearly reveals the focal adhesion localization of GFP-ZYXIN in the transfected cell while the untransfected cells are not seen. These results demonstrate that GFP and CFP fluorescence can be successfully separated even when simultaneously excited using an 880 nm laser beam. Furthermore, specific sub-cellular compartments or proteins can be visualized in cells simultaneously expressing GFP and CFP.
We next investigated if the method described above would work in vivo, we generated tumors composed of either a mixture of GFP and CFP expressing cells or cells co-expressing membrane targeted GFP and cytoplasmic CFP by co-injecting 5 × 105 cells of each of the mixed population or 106 cells that have GFP-tagged proteins into the mammary fat pads of SCID mice. An additional potential structure that can be imaged in vivo using multiphoton microscopy is the collagen fiber matrix using second harmonic fluorescence [20]. When using an excitation wavelength of 880 nm the second harmonic of collagen would be predicted to have a wavelength of around 440 nm which is only slightly shorter than CFP fluorescence. We found that second harmonic light generated by collagen fibers passed through the 450/80 filter but not the 480/30 (Figure 3 left hand panels – fibers marked with arrowheads). It was also able to pass through a 450/40 filter which CFP fluorescence was not able to pass through. Thus by using 450/40, 480/30 and 515/30 filters followed by subtraction of the 480/30 signal from the 515/30 we can simultaneously image collagen, CFP and GFP in a living tissue. The left hand panels in Figure 3 show this method used on a tumor consisting of a mixture of GFP and CFP expressing cells. The right hand panels show that this method can also be used to simultaneously image a membrane targeted GFP and CFP expressed in the same cell. A limitation in the past with imaging GFP tagged proteins in vivo has been determining the outline of the cell. By labeling the cytoplasm of the cell with CFP, we can now determine protein localization within a cell in a living tissue. Tumors generated using GFP-expressing cells alone or CFP-expressing cells alone show no differences in tumor growth and metastatic ability when compared to parental MTLn3 cells ([1] and data not shown)
Figure 3 Simultaneous capture of GFP, CFP and collagen second harmonic fluorescence in a living tissue. The panels on the left-hand side show images of an experimentally generated mammary tumour creating by injecting a mixture of cells either expressing GFP or CFP. The panels on the right-hand side show images of an experimentally generated mammary tumour with cells expressing CFP to mark the entire cytoplasmic volume and GFP-CAAX (membrane targeted). 880 nm laser light was used to excite the all the samples and the fluorescence was captured with indicated filters using non-descanned detection. All images were taken with a 40× objective and a final magnification of 500×. Scale bar = 25 μm.
Past intravital imaging of cells in living tumors by our group was able to establish differences such as motility, protrusion, and, cell polarity and orientation [6,16] between non-metastatic and metastatic tumor cells. By comparing gene expression in tumors of differing metastatic ability we found that many genes regulating motility and invasion are dramatically changed in metastatic cells. Furthermore, using our in vivo invasion assay we demonstrated that the expression of these is further increased in the motile subset of metastatic tumor cells [6,21]. However, these studies were not able to determine if the altered metastatic potential of these tumors is due to non cell autonomous differences in the tumor environment or to cell autonomous changes in the behavior of individual tumor cells. To directly compare the behavior of cells with differing metastatic potential in the same tumor environment in vivo, we generated a tumor by injecting a mixture of cancer cells with low metastatic potential labeled with GFP and high metastatic potential labeled with CFP. These cells were derived from the same genetic background [21] and their metastatic potential was determined by counting lung metastases in animals with primary tumors derived from injection of these cells into the mammary gland as described previously [1,6,22]. Figure 4A shows a time-series in which a greater number of cells with high metastatic potential (shown in white) are seen moving compared to control (shown in green) along collagen fibers imaged by second harmonic generation (purple). A time lapse of this sequence can be found in supplemental data (Additional file 1). The number pixels for each cell type was calculated in each field and used to normalize for differences in cell number of either type of cell in any given field. These data are the first demonstration that cells with higher metastatic potential move more frequently, by about 4.5 fold (n = 12 animals), than low metastatic cells in the same tumor micro-environment.
Figure 4 Differences in motility and invasion can be determined by imaging cells with different metastatic potential in the same tumor. A) The panels show a time series of images of an experimentally generated mammary tumour with cells either expressing GFP (low metastatic in green) or CFP (high metastatic in white) with collagen fibres in purple. Arrows point to moving CFP cells and arrowheads to moving GFP cells. Supplementary movie shows this image sequence including intermediate frames (Additional file 1). Images were taken with a 20× objective and a final magnification of 250×. B) CFP-labeled cells (white) and control GFP labelled cells (green) are seen in the lung after extravasation and metastatic growth. Images were taken with a 20× objective and a final magnification of 250×. C) Left hand panel CFP cell (white) has a filopod (arrow) protruding into a field of GFP control cells (green). Right hand panel CFP cell (white) has a lamellapod (arrow) seen in the middle of a field of GFP control cells (green). Supplementary movie shows a lamellapod retracting over time (Additional file 2). Images were taken with a 60× objective and a final magnification of 1000×. D) The panels show a time series of high magnification images of an experimentally generated mammary tumor with cells either expressing GFP (in green) or CFP (in white) with collagen fibers in purple. Moving high metastatic cells are outlined in white and an orange arrow shows the path taken by both cells. Supplementary movie shows this image sequence including intermediate frames (Additional file 3). Images were taken with a 40× objective and a final magnification of 500×. Scale bar for A-E = 25 um.
Our multi-fluor technique can also be used to explore metastatic ability. While fluorescently labeled cell have been used to explore metastatic and extravasive ability of differing cell types [9,23,24], our technique allows for multiphoton imaging at single cell resolution within the primary tumor of two cell populations and determination of differences in metastatic potential within the same animal. By examining of the lungs of animals with tumors formed by the injection of a mixture of CFP and GFP cells, extravasated cells can be counted to determine differences in numbers of cells metastasizing and compare this score to cell behavior differences in the primary tumor. Figure 4B shows how both cell types can be visualized in the lungs.
To visualize dynamic structures involved in cell motility we used a 60× LUMPlan/IR 0.9NA dipping objective. The morphology of single cells labeled with either CFP or GFP was easily visible on a background of the opposing color. Cell structures that were not well defined in prior in vivo experiments using conventional microscopy [1] can now be easily discerned. Cells are seen with filopodia (Figure 4C, left) and lamellipodia (Figure 4C, right) which when visualized as time-lapse movies (Additional file 2) can be seen protruding and retracting. Morphological changes of moving cells can also be resolved. Analysis of these time-lapse movies revealed that motile cells moved with amoeboid characteristics with the edge moving at up to 8 μm/min and making rapid changes in direction (Figure 3E and Additional file 3). The speed of the highly metastatic cells were twice what was previously reported for tumor cells with average metastatic potential in vivo [22]. Furthermore, motile cells often followed the same path; such behavior has been inferred from the presence of chains of invading cells in fixed breast cancer samples [25] but this is the first direct observation of such behavior. These cells may follow the same path because it is physically favorable (i.e. lack of ECM blocking cell movement) or because of a chemotactic relay between the cells.
Conclusion
Recent advances in multiphoton imaging technologies have provided researchers with the tools to study cell-cell interactions and the microenvironment of living tissues [4,5,26,27]. Our multi-fluor method described here will enable the direct comparison of the behavior of control and experimental manipulated cells in the same biological environment using multiphoton microscopy. This will be invaluable for determining the effects of genetic manipulations on cell behavior in vivo. In addition, the simultaneous imaging of GFP and CFP within the same cell will greatly aid the study of sub-cellular compartments and protein localization in living tissues.
Methods
Plasmids and Cell lines
ECFP-N1 (Clontech), GFP-CAAX (AP4-CAAX described [28]), GFP-zyxin [29]. Standard retroviral and transfection methods were used to generate the following derivatives of the MTLn3 rat mammary carcinoma cell line: low metastatic MTLn3-pLXSN GFP, high metastatic MTLn3-ECFP N1,0 MTLn3-GFPCAAX + ECFP N1, MTLn3-GFPZYXIN + ECFP N1. Cells were maintained in αMEM + 5% Fetal Calf Serum. Pools of GFP and CFP expressing cells were isolated by FACS methods. For imaging, cells were grown in MatTek dishes and fixed with 4% formaldehyde in PBS.
Tumor analysis
To determine metastatic differences between the MTLn3 clones, a total of 106 cells were injected into the mammary fat-pad of groups of eight 5–6 week old female Balb/C SCID mice. After 4 1/2 weeks animals were euthanized and lung mets counted in H and E stained sections. To determine motility differences, cells were compared by using a Boyden Chamber chemotaxis assay. Both these experiments were performed as reported previously [6,22].
Imaging
106 MTLn3 cells were injected into the mammary fat-pad of either female 5–6 week old female Balb/C SCID mice for the comparison of cell motility and metastatic potential or 5–6 week old Fischer 344 rats to image the cell membranes. After 3–4 1/2 weeks animals were anaesthetized with isoflurane and small incision was made in the skin to expose the tumor [1,6,26]. The anaesthetic was maintained while the tumor was imaged using a Biorad Radiance 2000 multiphoton microscope with an inverted Olympus IX70, within a heated chamber maintained at 30°c, connected to a Spectra Physics Tsunami Ti-Sapphire laser. All images were collected using non de-scanned detectors. The objectives used were as follows: 20× Plan Apo 0.7NA (air), 40× LUMPlan/IR 0.8NA (water) and 60× LUMPlan/IR 0.9NA (water). The filters used were as follows: 450/40, 450/80, 480/30, 515/30 (all from Chroma). Time lapsed images were taken over the course of 30 mins and analyzed using ImageJ. Frequency and speed of motility were calculated as described previously [22].
Authors' contributions
ES and JW conceived of and designed the study, did all imaging, all mathematical and statistical analysis and drafted the manuscript. UP transfected and maintained all cells with GFP-tagged proteins. JS determined the metastatic potential of the cells used. FG aided in design of the study and drafting of the manuscript. JC participated in the study's design and coordination and helped to draft the manuscript.
Supplementary Material
Additional File 1
Movie for figure 4A (avi. Format): Time-lapsed movie of high metastatic CFP labeled cells (white) and low metastatic GFP labeled cells (green) crawling on collagen fibers (purple) in vivo. Objective used 20×.
Click here for file
Additional File 2
Movie for figure 4C (avi. Format): Time-lapsed movie of a CFP labeled cell with a ruffling lamella on a field of GFP labeled cells in a living tumor. Objective used 60×.
Click here for file
Additional File 3
Movie for figure 4E (avi. Format): Time-lapsed movie of high metastatic CFP labeled cells (white) and low metastatic GFP labeled cells (green) crawling on collagen fibers (purple) in vivo showing amoeboid characteristics and making rapid changes in direction. Objective used 40×.
Click here for file
Acknowledgements
We are grateful the members of the Analytical Imaging Facility for support and advice. This work was supported by NIH grant CA100324 (J.S. Condeelis) and NIH grant GM58801 (F. Gertler). E. Sahai was funded by a Unio Internationale Contra Cancrum Aventis Translational Cancer Research Fellowship. U. Philippar received a Ph.D. scholarship by the Boehringer Ingelheim Funds.
==== Refs
Farina KL Wyckoff JB Rivera J Lee H Segall JE Condeelis JS Jones JG Cell motility of tumor cells visualized in living intact primary tumors using green fluorescent protein Cancer Res 1998 58 2528 2532 9635573
Hoffman R Green fluorescent protein imaging of tumour growth, metastasis, and angiogenesis in mouse models Lancet Oncol 2002 3 546 556 12217792 10.1016/S1470-2045(02)00848-3
MacDonald IC Groom AC Chambers AF Cancer spread and micrometastasis development: quantitative approaches for in vivo models Bioessays 2002 24 885 893 12325121 10.1002/bies.10156
Zipfel WR Williams RM Webb WW Nonlinear magic: multiphoton microscopy in the biosciences Nat Biotechnol 2003 21 1369 1377 14595365 10.1038/nbt899
Jain RK Munn LL Fukumura D Dissecting tumour pathophysiology using intravital microscopy Nat Rev Cancer 2002 2 266 276 12001988 10.1038/nrc778
Wang W Wyckoff JB Frohlich VC Oleynikov Y Huttelmaier S Zavadil J Cermak L Bottinger EP Singer RH White JG Single cell behavior in metastatic primary mammary tumors correlated with gene expression patterns revealed by molecular profiling Cancer Res 2002 62 6278 6288 12414658
Ahmed F Wyckoff J Lin EY Wang W Wang Y Hennighausen L Miyazaki J Jones J Pollard JW Condeelis JS GFP expression in the mammary gland for imaging of mammary tumor cells in transgenic mice Cancer Res 2002 62 7166 7169 12499251
Yang M Li L Jiang P Moossa AR Penman S Hoffman RM Dual-color fluorescence imaging distinguishes tumor cells from induced host angiogenic vessels and stromal cells Proc Natl Acad Sci U S A 2003 100 14259 14262 14614130 10.1073/pnas.2436101100
Yamamoto N Yang M Jiang P Xu M Tsuchiya H Tomita K Moossa AR Hoffman RM Determination of clonality of metastasis by cell-specific color-coded fluorescent-protein imaging Cancer Res 2003 63 7785 7790 14633704
Glinskii AB Smith BA Jiang P Li XM Yang M Hoffman RM Glinsky GV Viable circulating metastatic cells produced in orthotopic but not ectopic prostate cancer models Cancer Res 2003 63 4239 4243 12874032
Lansford R Bearman G Fraser SE Resolution of multiple green fluorescent protein color variants and dyes using two-photon microscopy and imaging spectroscopy J Biomed Opt 2001 6 311 318 11516321 10.1117/1.1383780
Brown EB Campbell RB Tsuzuki Y Xu L Carmeliet P Fukumura D Jain RK In vivo measurement of gene expression, angiogenesis and physiological function in tumors using multiphoton laser scanning microscopy Nat Med 2001 7 864 868 11433354 10.1038/89997
Dunn KW Sandoval RM Kelly KJ Dagher PC Tanner GA Atkinson SJ Bacallao RL Molitoris BA Functional studies of the kidney of living animals using multicolor two-photon microscopy Am J Physiol Cell Physiol 2002 283 C905 916 12176747
Buehler C Kim KH Greuter U Schlumpf N So PT Single-photon counting multicolor multiphoton fluorescence microscope J Fluoresc 2005 15 41 51 15711876 10.1007/s10895-005-0212-z
Kahn E Vejux A Dumas D Montange T Frouin F Robert V Riedinger JM Stoltz JF Gambert P Todd-Pokropek A FRET multiphoton spectral imaging microscopy of 7-ketocholesterol and Nile Red in U937 monocytic cells loaded with 7-ketocholesterol Anal Quant Cytol Histol 2004 26 304 313 15678612
Condeelis J Segall JE Intravital imaging of cell movement in tumours Nat Rev Cancer 2003 3 921 930 14737122 10.1038/nrc1231
Stockholm D Bartoli M Sillon G Bourg N Davoust J Richard I Imaging calpain protease activity by multiphoton FRET in living mice J Mol Biol 2005 346 215 222 15663939 10.1016/j.jmb.2004.11.039
Blab GA M. LPH Cognet L Harms GS Schmidt T Two-photon excitation action cross-sections of the autofluorescent proteins Chemical Physical Letters 2001 350 71 77 10.1016/S0009-2614(01)01282-9
Michaelson D Silletti J Murphy G D'Eustachio P Rush M Philips MR Differential localization of Rho GTPases in live cells: regulation by hypervariable regions and RhoGDI binding J Cell Biol 2001 152 111 126 11149925 10.1083/jcb.152.1.111
Williams RM Zipfel WR Webb WW Interpreting second-harmonic generation images of collagen I fibrils Biophys J 2005 88 1377 1386 15533922 10.1529/biophysj.104.047308
Wang W Goswami S Lapidus K Wells AL Wyckoff JB Sahai E Singer RH Segall JE Condeelis JS Identification and testing of a gene expression signature of invasive carcinoma cells within primary mammary tumors Cancer Res 2004 64 8585 8594 15574765
Wyckoff JB Jones JG Condeelis JS Segall JE A critical step in metastasis: in vivo analysis of intravasation at the primary tumor Cancer Res 2000 60 2504 2511 10811132
Koop S Schmidt EE MacDonald IC Morris VL Khokha R Grattan M Leone J Chambers AF Groom AC Independence of metastatic ability and extravasation: metastatic ras-transformed and control fibroblasts extravasate equally well Proc Natl Acad Sci U S A 1996 93 11080 11084 8855312 10.1073/pnas.93.20.11080
Naumov GN Wilson SM MacDonald IC Schmidt EE Morris VL Groom AC Hoffman RM Chambers AF Cellular expression of green fluorescent protein, coupled with high-resolution in vivo videomicroscopy, to monitor steps in tumor metastasis J Cell Sci 1999 112 1835 1842 10341203
Pitts WC Rojas VA Gaffey MJ Rouse RV Esteban J Frierson HF Kempson RL Weiss LM Carcinomas with metaplasia and sarcomas of the breast Am J Clin Pathol 1991 95 623 632 1850949
Wyckoff J Segall J Condeelis J David L Spector, Robert D Goldman Single cell imaging in animals Live Cell Imaging: A Laboratory Manual 2004 22 Cold Spring Harbor Laboratory Press
Dunn KW Sandoval RM Molitoris BA Intravital imaging of the kidney using multiparameter multiphoton microscopy Nephron Exp Nephrol 2003 94 e7 11 12806182 10.1159/000070813
Bear JE Loureiro JJ Libova I Fassler R Wehland J Gertler FB Negative regulation of fibroblast motility by Ena/VASP proteins Cell 2000 101 717 728 10892743 10.1016/S0092-8674(00)80884-3
Rottner K Krause M Gimona M Small JV Wehland J Zyxin is not colocalized with vasodilator-stimulated phosphoprotein (VASP) at lamellipodial tips and exhibits different dynamics to vinculin, paxillin, and VASP in focal adhesions Mol Biol Cell 2001 12 3103 3113 11598195
| 15910685 | PMC1159039 | CC BY | 2021-01-04 16:38:25 | no | BMC Biotechnol. 2005 May 23; 5:14 | utf-8 | BMC Biotechnol | 2,005 | 10.1186/1472-6750-5-14 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1597480410.1371/journal.pbio.0030241Research ArticleEcologyEvolutionGenetics/Genomics/Gene TherapyMolecular Biology/Structural BiologyPaleontologyZoologyMammalsEvolution, Systematics, and Phylogeography of Pleistocene Horses in the New World: A Molecular Perspective Pleistocene Horses in the New WorldWeinstock Jaco [email protected]
1
Willerslev Eske
1
¤aSher Andrei
2
Tong Wenfei
3
Ho Simon Y.W
1
Rubenstein Dan
3
Storer John
4
Burns James
5
Martin Larry
6
Bravi Claudio
7
Prieto Alfredo
8
Froese Duane
9
Scott Eric
10
Xulong Lai
11
Cooper Alan [email protected]
1
¤b1 Ancient Biomolecules Centre, Department of Zoology, University of Oxford, United Kingdom,2 Institute of Ecology and Evolution, Russian Academy of Sciences, Moscow, Russia,3 Department of Ecology and Evolutionary Biology, Princeton University, United States of America,4 Government of the Yukon, Cultural Services Branch, Whitehorse, Canada,5 Quaternary Paleontology Program, Provincial Museum of Alberta, Edmonton, Canada,6 Natural History Museum, University of Kansas, Lawrence, Kansas, United States of America,7 Instituto Multidisciplinario de Biologia Celular (IMBICE), La Plata, Argentina,8 Instituto de la Patagonia, Universidad de Magallanes, Punta Arenas, Chile,9 Department of Earth and Atmospheric Science, University of Alberta, Canada,10 San Bernardino County Museum, Redlands, California, United States of America,11 China University of Geosciences, Wuhan, ChinaHillis David Academic EditorUniversity of TexasUnited States of America8 2005 28 6 2005 28 6 2005 3 8 e24114 2 2005 6 5 2005 Copyright: © 2005 Weinstock et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
New World Pleistocene Horses: Pruning the Equid Tree
The rich fossil record of horses has made them a classic example of evolutionary processes. However, while the overall picture of equid evolution is well known, the details are surprisingly poorly understood, especially for the later Pliocene and Pleistocene, c. 3 million to 0.01 million years (Ma) ago, and nowhere more so than in the Americas. There is no consensus on the number of equid species or even the number of lineages that existed in these continents. Likewise, the origin of the endemic South American genus Hippidion is unresolved, as is the phylogenetic position of the “stilt-legged” horses of North America. Using ancient DNA sequences, we show that, in contrast to current models based on morphology and a recent genetic study, Hippidion was phylogenetically close to the caballine (true) horses, with origins considerably more recent than the currently accepted date of c. 10 Ma. Furthermore, we show that stilt-legged horses, commonly regarded as Old World migrants related to the hemionid asses of Asia, were in fact an endemic North American lineage. Finally, our data suggest that there were fewer horse species in late Pleistocene North America than have been named on morphological grounds. Both caballine and stilt-legged lineages may each have comprised a single, wide-ranging species.
Ancient DNA simplifies the systematics of horses present in the New Word during the Pleistocene and reveals that the stilt- legged horse is native to the continent and not a Eurasian migrant.
==== Body
Introduction
Nowhere is the later evolution of horses more problematic than in the Americas, where more than 50 species of Pleistocene equids have been named, most of them during the 19th and early 20th centuries. While recent paleontological studies suggest that this number should be drastically revised [1,2], no consensus has been reached about the number of valid species or their phylogenetic relationships. This has limited an in-depth investigation of the biogeography, evolution, and extinction of American Pleistocene equids and their phylogenetic relationships with synchronous Eurasian forms. A particularly problematic example is the so-called North American or New World “stilt-legged” horses (NWSL) found in Mid- and Late Pleistocene deposits in the north and west of North America [1–3]. The NWSL are often found with a second equid form commonly regarded as closely related to the Eurasian caballines, a group that includes the domestic horse (Equus caballus) and the extant wild Przewalskii horse. The stilt-legged forms have been taxonomically assigned to a number of species, but all have gracile limbs similar to the extant Asian asses (hemionids—e.g., onager, E. hemionus; and kiang, E. kiang,
Figure 1), leading to suggestions that they are closely related or even result from a dispersal of Eurasian hemionids via the exposed Bering Strait during the last glacial period [4–9]. Alternatively, the NWSL have been regarded as North American endemics [10,11].
Figure 1 Metatarsal Shape and Size in Pleistocene and Extant Equids
(A) Bivariate plot showing metatarsal shape and size in extant and extinct horses. Modern asses: kiang (E. kiang) = light blue circles; onager (E. hemionus onager) = dark blue; kulan (E. hemionus kulan) = purple. Pleistocene equids: stilt-legged from Alaska and the Yukon = black; E. lambei (Alaska) = red; stilt-legged from Natural Trap Cave (Wyoming) = yellow; caballines from Natural Trap Cave = orange; caballines from Alberta = green; H. saldiasi from southern Patagonia = grey. “A” and “B” above/beside points of caballine horses denote the phylogenetic clade to which the specimens belong (Figure 5). Not all of the specimens were genetically analyzed.
(B) Metatarsal shape in Pleistocene horses. From left to right: H. saldiasi (southern Patagonia); E. lambei (Yukon); NWSL (Yukon). Scale bar is 10 cm.
Likewise, the systematic position and origins of the South American Pleistocene genus Hippidion need to be re-evaluated. This form appeared in South America approximately 2.5 million years (Ma) ago, i.e., shortly after the formation of the Isthmus of Panama enabled the movement of animals between the continents, an event known as the Great American Biotic Interchange [12]. Hippidion is considered to be a descendant of the pliohippines, a primitive group of Miocene horses that diverged from the ancestral lineage of equines (a category including all living and extinct members of the genus Equus, such as caballine horses, hemiones, asses, and zebras) prior to 10 Ma ago [13−15] (Figure 2). A recent study [16] presented genetic data from three southern Patagonian specimens morphologically identified as Hippidion. Unexpectedly, the sequences clustered inside the genus Equus, and it was concluded that the bones must be from E.
(Amerhippus), an extinct equine form that dispersed to South America more recently. It was suggested that the bones had been misidentified as Hippidion due to (assumed) morphological convergence, but the deep split between Hippidion and Equus lineages was not questioned. This view is intriguing, as elsewhere in South America where Hippidion and Equus were sympatric, e.g., central Argentina and southern Bolivia, each form shows well-defined morphological characters in the cranial and postcranial skeleton [17].
Figure 2 Diagrammatic Representation of North and South American Equid Taxonomy Superimposed on a Time Scale according to Paleontological (A, B) and Molecular Data (C)
(A) represents MacFadden's [13] view of hippidiform origins from a pliohippine, diversifying into two genera, Hippidion and Onohippidium, in North America during the Miocene. (B) shows Alberdi and Prado's [14,15] view of hippidiforms as descendants of a pliohippine during the Miocene; the single genus, Hippidion, originates only after dispersal into South America (they do not recognize the genus Onohippidium as valid [15]). (C) represents the results of the molecular data presented in the present study; it shows Hippidion as originating during the Pliocene, c. 3-3.5 Ma ago. The NWSL is possibly a sister species of Hippidion.
Results/Discussion
In order to investigate the phylogenetic position and evolutionary history of the South American hippidiforms and the North American stilt-legged horses, we analyzed a segment between 349 and 716 base pairs (bp) of the mitochondrial control region of fossil equid remains from a number of different localities in North and South America and Eurasia (Figure 3), ranging in date from c. 53,000 years ago to historical times. Mitochondrial DNA was extracted from cortical bone using ancient-DNA techniques and portions of both hypervariable regions (HVR I and II) amplified by PCR in a number of overlapping fragments. Phylogenetic relationships and divergence dates were inferred using maximum likelihood and Bayesian techniques (see Materials and Methods).
Figure 3 Provenance of Bone Samples Extracted and Sequenced for This Study
Details of each sample are given in Table S2.
Phylogenetic analysis using maximum likelihood (Figure 4) resolved Hippidion, NWSL, and caballines as three genetically divergent species within a monophyletic group. Hippidion and NWSL appear as sister taxa, although the support for this is not high. African zebras and asses and Asian asses (onager and kiang) form a basal polytomy. Bayesian analysis produced a similar topology.
Figure 4 Phylogeny of Recent and Pleistocene Equids
The maximum likelihood tree was constructed with two fragments of the mitochondrial control region (583 bp and 133 bp in the HVR1 and HVR2 regions, respectively). Bayesian analysis produced a similar topology. The general time-reversible substitution model was used in both techniques. Black numbers above/beside nodes are posterior probabilities and bootstrap values, respectively (only values > 50% are shown). White numbers on black background are divergence times as calculated from the molecular data. Numbers/letters in bold at the beginning of each specimen's name are sample numbers or GenBank accession numbers. Labels of prehistoric specimens are followed by their age, if available, in thousands of years. The outgroup (Rhinoceros and Ceratotherium) is not shown.
The close phylogenetic relationship between Hippidion and caballine horses is in direct contrast to current paleontological models of hippidiform origins. Nevertheless, we are confident that these sequences are those of Hippidion rather than the South American caballine form E. (Amerhippus), which dispersed into South America about 1 to 1.5 Ma later than Hippidion. Morphological studies have concluded that only one species of equid, H. saldiasi, is represented in the Late Pleistocene cave deposits of southern Patagonia [18] that generated all but one of our samples. Furthermore, our specimens, mostly phalanges, show not only the typical proportions of Hippidion—short and wide—but also morphological, size-independent characteristics typical of this taxon. The first phalanges, for instance, show two short tuberosities for the insertion of the trigonium phalangis; in Equus, by contrast, there is a single, long, V-shaped tuberosity [14,17,18]. Finally, both our sequences and those previously reported for three southern Patagonian specimens [16] are virtually identical, and it seems extremely unlikely that all equid DNA sequences from southern Patagonian horses have come from specimens of E. (Amerhippus) that have been misidentified as Hippidion (contra Orlando et al. [16]). Unfortunately, it was not possible to obtain DNA from any morphologically identified Amerihippus specimen to test this, as ancient-DNA survival appears relatively rare in northern South America.
Given the position of Hippidion in the tree (Figure 4), its origin postdates the more basal split at node A, estimated at 4.8 Ma (Table S1). Therefore, Hippidion should not be seen as a descendant from the Miocene pliohippines; instead, its origins appear to be much more recent, probably during the last stages of the Pliocene (c. 3 Ma), which is close to the time of the first fossil occurrence of this genus [12,14,17,18]. Furthermore, this later date is in opposition to the claim that hippidiforms (genera Hippidion and Onohippidium) were present in North America as early as the late Miocene (c. 7–8 Ma) [13,19].
Figure 4 also resolves the phylogenetic position of the NWSL as North American endemics rather than Eurasian migrants. Interestingly, specimens from both north (Alaska/Yukon) and south (Wyoming/Nevada) of the Pleistocene ice sheets clearly belong to the same taxon. The apparent large geographic distribution of this species raises the possibility that other Late Pleistocene NWSL currently described as different species—e.g., E. francisci, E. tau, E. quinni, E. cf. hemionus, E. (Asinus) cf. kiang, and others with similar limb proportions [1,2,20]—may in fact represent the same taxon. The origins of the NWSL probably lie south of the Pleistocene ice sheets, where Mid-Pleistocene remains (c. 0.5 Ma) with similar limb characteristics have been found [1,2] and Late Pleistocene specimens occur in high frequencies. North of the ice sheets, NWSL remains are relatively rare and appear to have had a restricted temporal distribution [21]. In spite of their presence in eastern Beringia (unglaciated Alaska/Yukon), the stilt-legged horses apparently failed to disperse through the Bering land-bridge into western Beringia (northeast Siberia). In addition, the extinctions of the NWSL north and south of the ice sheets clearly took place at quite different times. Populations in eastern Beringia disappeared around 31,000 years ago [21], while, as demonstrated by the AMS date from the Nevada specimen, NWSL persisted south of the ice sheets until at least 13,000 years ago before present (Table S2), close to the date of the other megafaunal extinctions [4,21].
The phylogenetic patterns and divergence dates in Figure 4 suggest that the morphological apomorphies developed relatively rapidly in both NWSL and Hippidion. The greatly retracted nasal notch and long nasal bones that define Hippidion [14,15] apparently originated between 4.4 Ma and the late Pliocene–early Pleistocene (c. 2–2.5 Ma), when Hippidion first appears in the fossil record [12,15]. The extreme divergence in distal limb proportions between NWSL and Hippidion would have occurred during the same interval and presumably reflects adaptations to particular environments [22]. The similar proportions of metapodials in NWSL and Eurasian hemiones are likely to be convergent cursorial adaptations to arid environments with hard soils and scarce vegetation.
A third significant discovery from Figure 4 is that all caballine horses from western Europe to eastern Beringia—including the domestic horse—are a single Holarctic species. A more detailed phylogeographic analysis was performed using only 349 bp of the HVR1 region in order to be able to include a larger number of sequences of Pleistocene, historic, and recent caballines (Figure 5). Little phylogeographic structure is apparent, probably reflecting the high degree of mobility and adaptability of this species. A recent study showed that caballines group into two major clades [23]. Figure 5 confirms this and shows that one clade (A) was broadly distributed from central Europe to North America north and south of the ice. The second clade (B), however, seems to have been restricted to North America. If present in the Old World at all, it probably disappeared there before horse domestication took place, around 5,000 years BP, as all domestic horses—only some of which are shown in the tree—cluster within clade A.
Figure 5 Maximum Likelihood Phylogenetic Tree of Recent, Historic, and Pleistocene Caballine Horses Based on 364 bp of the Mitochondrial Control Region (HVR1)
Posterior probabilities/bootstrap values are presented above or beside the main nodes only if either is > 50%. Labels of prehistoric specimens are followed by their age, if available, in thousands of years (k).
Size has been used as one of the main morphological criteria for defining species of Pleistocene equids [1], and the body size of the Late Pleistocene North American caballines we sampled did exhibit marked regional variability; e.g., horses in Alberta (Canada) were larger than their eastern Beringian and Wyoming counterparts (see Figure 1). The DNA evidence strongly suggests, however, that all of these large and small North American caballine samples belong to a single species (Figure 5 and Table S3). The presence of a morphologically variable caballine species widely distributed north and south of the Pleistocene ice sheets raises the tantalizing possibility that, in spite of the many taxa named on morphological grounds [1,2,4], most or even all North American caballines were members of the same species.
Thus, our results indicate that only two lineages—a caballine and a stilt-legged—may have been present in North America during the Late Pleistocene, each comprising perhaps only a single species with temporal and regional variation in body size and morphology. This model would greatly simplify the systematics of North American Pleistocene horses and could open the way to the analysis of morphological variation in terms of adaptation to different environments. The study of this variation, in combination with paleoclimatic and paleoenvironmental data, should dramatically improve our understanding of the biogeography, evolution, and extinction of horses in this continent.
Materials and Methods
DNA extraction, amplification, and sequencing
Details on the provenance and radiocarbon date (where available) of the specimens are given in Table S2. Samples (approximately 0.3–0.5 g) were removed from sections of cortical bone with a Dremel (Racine, Wisconsin, United States) tool, and the outermost layers removed by abrasion. These were then ground to a powder in a Braun (Melsungen, Germany) Mikrodismembrator and decalcified in 10–30 vol of 0.5 M EDTA (pH 8) overnight at room temperature. The sediment was collected by centrifugation and digested in 6 ml of extraction buffer containing 10 mM Tris-HCl (pH 8), 10 mM NaCl, 0.5 mg/ml proteinase K, 10 mg/ml dithiothreitol, 1% sodium dodecyl sulphate, and 0.001–0.01M N-phenacylthiazolium bromide. The extraction was incubated overnight at 45–50 °C. Following digestion, the extraction was added to an equal volume of Tris-saturated phenol and rotated constantly for 10 min before centrifugation at 6,000 rpm for an additional 10 min. The aqueous phase was then removed. This step was repeated twice, once with an equal volume of phenol and then with an equal volume of chloroform. DNA was desalted and concentrated to c. 100 μl using Centricon-30 (Amicon [Millipore, Billerica, Massachusetts, United States]) devices. DNA extractions were performed in batches of six to 14 samples, with one to two negative extraction controls (no bone powder) used for every run.
PCR amplifications were performed in a 25-μl reaction using 1 μl of extract, 1.25 U Platinum Taq Hi-Fidelity and 10X buffer (Invitrogen, Carlsbad, California, United States), 2 mg/ml BSA, 2 mM magnesium sulfate, 250 μM of each DNTP, and 1 μM of each primer. DNA was amplified in a MTR Tetrad under the following cycling conditions: 40 cycles of 45-s denaturation at 94 °C, 45-s annealing at 56–63 °C (depending on primer; Table S4), and 90-s extension at 68 °C with a final extension of 10 min at 68 °C. Negative extraction controls and negative PCR controls were used in each PCR reaction. A 583-bp contiguous segment of the HVR1 of the mitochondrial control region and a 133-bp segment of the HVR2 DNA were amplified using overlapping fragments (Table S4).
Sequencing reactions (10 μl) were performed on both strands using PRISM BigDye Terminator v3.1 Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, California, United States). After the removal of unincorporated dye terminators and primers through ethanol precipitation, PCR products were sequenced on an ABI 310 automated sequencer (PerkinElmer, Wellesley, California, United States) at the DNA-sequencing facility of the Department of Zoology, University of Oxford.
All procedures up until PCR amplification were performed in the specialist Henry Wellcome Ancient Biomolecules Centre, while amplification, characterization, and sequencing took place in the physically distant Department of Zoology to minimize contamination risk [24]. In order to evaluate template damage, cloning of six individual PCR products was performed (12 clones per product) using the Invitrogen Topo-TA cloning kit (see Table S2). Cloned products showed only limited amounts of sequence modification due to template damage [25,26]. In order to verify the authenticity of the sequences, four samples were independently extracted and sequenced at the ancient DNA laboratory at the University of Copenhagen. These included two stilt-legged equids (YG130.3 and YG109.7) and two caballine horses (IEM-202–128 and P94.1.415). Sequences were aligned manually using Se-Al v.2.0a11 [27].
Phylogenetic analyses used maximum likelihood and Bayesian Markov chain Monte Carlo (MCMC) methods in PAUP 4.0 and MrBayes 3.0 software packages, respectively [28, 29]. In the Bayesian analysis, the posterior probability distribution of trees was approximated by drawing a sample every 500 steps over 5,000,000 MCMC cycles, after discarding a burn-in of 500,000 cycles. In both likelihood and Bayesian MCMC analyses, a GTR+G+I model of nucleotide substitution was used. Two species of rhino, Ceratotherium simum and Rhinoceros unicornis, were used as outgroups.
Metrical data
Metrical data were obtained from stilt-legged and caballine equid metatarsals from Natural Trap Cave, Wyoming (University of Kansas, United States), from Alaska (Frick Collection, AMNH New York, United States), Alberta (Provincial Museum of Alberta, Canada), and specimen YG150.82 from Thistle Creek (Yukon, Canada). Measurements were taken to the nearest 0.1 mm, following Eisenmann and Bekouche [30]. Additional data for stilt-legged horses from the Yukon and for recent Eurasian asses were obtained from published sources [30,31]. Metrical data of H. saldiasi were gathered from published sources [32] and from a specimen from Mylodon Cave, Chile, in the collection of the Museo de la Plata (MLP-6–455).
Divergence date estimation
The ages of the most recent common ancestors (MRCAs) of Equus, stilt-legged horses, and caballine horses, along with the date of the caballine–hippidiform split, were estimated using Bayesian inference as implemented by the program BEAST [33,34]. This program can accommodate non-contemporaneous sequences. The date estimates were made under a general time-reversible model of nucleotide substitution. Rate variation among sites was modelled using a discrete gamma distribution with eight rate categories, and the proportion of invariant sites was estimated.
Monophyly was enforced for the clades only when the ages of MRCAs were being estimated; otherwise, no restrictions were placed on the topology. Consequently, the divergence dates produced by the program have been integrated over the various tree topologies sampled throughout the MCMC analysis, and weighted in proportion to their probability of occurrence. This implies that topological uncertainty has been factored into the divergence date estimates.
The program BEAST incorporates the ages of sequences derived from carbon dating into the estimation of divergence dates. However, because the ages of the sequences are considerably younger than the internal nodes of the tree, it is preferable to provide further calibration information in order to avoid the need for large extrapolations. Accordingly, an additional calibration point was provided by placing a normal prior, with mean 3.0 Ma and standard deviation 0.5 Ma, on the age of the MRCA of Hippidion and the stilt-legged horses. This calibration is based on the formation of the Panama isthmus, which has been estimated at 2.5 to 3.0 Ma old [12]. In order to accommodate the possibility that dispersal across continents occurred prior to the complete formation of the isthmus, we repeated the analysis with values of 4.0 Ma and 0.25 Ma for the mean and standard deviation, respectively. The results of these analyses were cross-validated by performing a third analysis, using a normal prior with mean 55 Ma and standard deviation 5 Ma, on the age of the MRCA of rhinoceroses and horses. In all three analyses, a rigid lower bound of 1.3 Ma was placed on the age of the caballine clade, based on fossil evidence [20].
The posterior distributions of the dates being estimated were approximated by sampling parameter values at every 250th cycle over 10,000,000 MCMC steps, after discarding 1,000,000 burn-in steps. The starting tree for the MCMC analysis was generated using a coalescent process. Convergence of the sampled parameters was checked using the program Tracer [35], which revealed that the effective sample sizes of the sampled parameters were well in excess of 100, the minimum recommended effective sample size (AJ Drummond, personal communication).
Supporting Information
Table S1 Estimated Divergence Times of Various Clades Using Bayesian Inference as Implemented by the Program BEAST
(59 KB DOC).
Click here for additional data file.
Table S2 Specimens Extracted for This Study
(99 KB DOC).
Click here for additional data file.
Table S3 Intra- and Inter-Clade HKY+G+I-Corrected Average Pairwise Distance in Equid Clades (Based on 349 bp of the HVR1 Control Region)
(21 KB DOC).
Click here for additional data file.
Table S4 Primers Used in the Amplification of mtDNA, Control Region
(24 KB DOC).
Click here for additional data file.
Accession Numbers
The GenBank (http://www.ncbi.nlm.nih.gov/Genbank) accession numbers for genes and gene products discussed in this paper are: C. simum (NC001808), R. unicornis (NC001779), and the segment between 349 and 716 bp of the mitochondrial control region of fossil equid remains from a number of different localities in North and South America and Eurasia (DQ007552–DQ007621).
We thank the following institutions for permission to sample specimens: University of Tübingen (N. Conard), State Museum of Natural History, Stuttgart (R. Ziegler), Landesdenkmalamt Baden-Württemberg (E. Stephan), Institute of Plant and Animal Ecology, Ekaterinberg (P. Kosintzev), Canadian Museum of Nature (C.R. Harington), Museo de La Plata (M. Reguero and S. Bargo), University of Fairbanks (R.D. Guthrie), Los Angeles County Museum (S. McLeod), American Museum of Natural History, New York (C. Norris), and Harbin (X. Gao). We thank T. Higham (Oxford Radiocarbon Accelerator Unit), and gratefully acknowledge the support of the NERC ORADS dating program. We also thank Tina B. Brand (University of Copenhagen), who performed the independent replications. Many thanks are also due to A. Lister (UCL), who critically read and commented on an earlier version of the manuscript. Financial support was provided by the Wellcome and Leverhulme Trusts (AC) and NERC (AC, JW).
Competing interests. The authors have declared that no competing interests exist.
Author contributions. JW, EW, DR, and AC conceived and designed the experiments. JW and WT performed the experiments. JW, AS, SYWH, JS, JB, and LM analyzed the data. AS, JB, LM, CB, AP, DF, ES, and LX contributed reagents/materials/analysis tools. JW, DR, and AC wrote the paper.
¤a Current address: Ancient DNA and Evolution Group, Niels Bohr Institute, University of Copenhagen, Denmark
¤b Current address: School of Earth and Environmental Sciences, University of Adelaide, Australia
Citation: Weinstock J, Willerslev E, Sher A, Tong W, Ho SYW, et al. (2005) Evolution, systematics, and phylogeography of Pleistocene horses in the new world: A molecular perspective. PLoS Biol 3(8): e241.
Abbreviations
bpbase pairs
HVRhypervariable region
Mamillion years
MCMCMarkov chain Monte Carlo
MRCAmost recent common ancestor
NWSLNew World “stilt-legged” horses.
==== Refs
References
Winans MC Prothero DR Schoch R A quantitative study of North American fossil species of the genus Equus
The evolution of perissodactyls 1989 Oxford Oxford Univ. Press 262 297
Azzaroli A The genus Equus in North America Palaeontographica Italica 1998 85 1 60
Guthrie RD Gerlach SC Murray MS Paleobehavior in Alaskan Pleistocene horses: Social structure, maturation rates, uses of the landscape, and mortality patterns People and wildlife in northern North America: Essays in honor of R. Dale Guthrie 2001 Oxford BAR 32 49
Kurtén B Anderson E Pleistocene mammals of North America 1980 New York Columbia University Press 442
Dalquest WW Phylogeny of American horses of Blancan and Pleistocene age Ann Zool Fenn 1978 15 191 199
Harington CR Pleistocene mammals from Lost Chicken Creek, Alaska Can J Earth Sci 1980 17 169 197
Harington CR Clulow FV Pleistocene mammals from Gold Run Creek, Yukon Territory Can J Earth Sci 1973 10 697 745
Uerpmann H-P Schiebler J Sedlmeier J Spycher H Altweltliche Faunenelemente im Spätpleistozän Alaskas und die frühe Besiedlungsgeschichte Nordamerikas Beiträge zur Archäozoologie, Anthropologie, Geologie und Paläontologie 1990 Basle Helbing and Lichtenhanhn 303 307
Willoughby DP The empire of Equus 1974 Cranbury (New Jersey) Barnes 475
Lundelius EL Stevens MS
Equus francisci Hay, a small stilt-legged horse. Middle Pleistocene of Texas J Paleontol 1970 44 148 153
Sher A Kuznetsova T Early to Middle Pleistocene equids from Beringida J Vert Paleontol 1996 16 65A
Webb SD Stehli FG Webb SD Late Cenozoic mammal dispersals between the Americas The great American biotic interchange 1985 New York and London Plenum 357 386
MacFadden BJ Pleistocene horses from Tarija, Bolivia, and the validity of the genus Onohippidium (Mammalia: Equidae) J Vert Paleontol 1997 17 199 218
Alberdi MT Prado JL Review of the genus Hippidion Owen, 1869 (Mammalia: Perissodactyla) from the Pleistocene of South America Zool J Linn Soc 1993 108 1 22
Alberdi MT Prado JL Comments on Pleistocene horses from Tarija, Bolivia, and the validity of the genus Onohippidium (Mammalia: Equidae), by B. J. MacFadden J Vert Paleontol 1998 18 669 72
Orlando L Eisenmann V Reynier F Sondaar P Hanni C Morphological convergence in Hippidion saldiasi and Equus
(Amerhippus) Southern-American equids elucidated by ancient DNA analysis J Mol Evol 2003 57 Supp 1 S29 S40 15008401
Alberdi MT Prado JL El registro de Hippidion Owen, 1869 y Equus (Amerhippus ) Hoffstetter, 1950 (Mammalia, Perissodactyla) en America del Sur Ameghiniana 1992 29 265 284
Alberdi MT Prieto A
Hippidion (Mammalia, Perissodactyla) de las cuevas de las provincias de magallanes y tierra del Fuego An Inst Patagonia 2000 28 147 171
MacFadden B Skinner M Diversification and biogeography of the one-toed horses Onohippidium and Hippidion
Yale Peabody Mus Nat Hist Postilla 1979 175 1 9
Eisenmann V von Koenigswald W Werdelin L Origins, dispersals, and migrations of Equus (Mammalia, Perissodactyla) Mammalian migration and dispersal events in the European Quaternary 1992 Frankfurt Courier Forschungsinstitut Senckenberg 161 170
Guthrie RD Rapid body size decline in Alaskan Pleistocene horses before extinction Nature 2003 426 169 171 14614503
Eisenmann V Karchoud A Analyses multidimensionelles de metapodes d'Equus
sensu lato (Mammalia, Perissodactyla) Bull Mus Natn Hist Nat Paris 1982 4 75 103
Vilá C Leonard J Götherström A Marklund S Sandberg K Widespread origins of domestic horse lineages Science 2001 291 474 477 11161199
Cooper A Poinar HN Ancient DNA: Do it right or not at all Science 2000 289 1139 10970224
Hansen AJ Willerslev E Wiuf C Mourier T Arctander P Statistical evidence for miscoding lesions in ancient DNA templates Mol Biol Evol 2001 18 262 265 11158385
Gilbert MTP Willerslev E Hansen AJ Barnes I Rudbeck L Distribution patterns of postmortem damage in human mitochondrial DNA Am J Hum Genet 2003 72 32 47 12489041
Rambaut A Rambaut A (1996) Se-Al: Sequence Alignment Editor [computer program] 1996 Available: http://evolve.zoo.ox.ac.uk/ . Accessed 25 May 2005
Swofford, DL PAUP*, version 4.0 [computer program] 2002 Sunderland (Massachusetts) Sinauer Associates
Huelsenbeck JP Ronquist F MRBAYES: Bayesian inference of phylogenetic trees Bioinformatics 2001 17 754 755 11524383
Eisenmann V Bekouche S Meadow RH Uerpmann H-P Identification and discrimination of metapodials from Pleistocene and modern Equus , wild and domestic Equids in the ancient world 1986 Wiesbaden Reichert Verlag pp 117 163
Harington CR Pleistocene mammals from Lost Chicken Creek, Alaska Can J Earth Sci 1980 17 169 197
Alberdi MT Miotti L Prado JL
Hippidion saldiasi Roth, 1899 (Equidae, Perissodactyla), at the Piedra Museo Site (Santa Cruz, Argentina): Its implication for the regional economy and environmental reconstruction J Arch Sci 2001 28 411 419
Drummond AJ Nicholls GK Rodrigo AG Solomon W Estimating mutation parameters, population history and genealogy simultaneously from temporally spaced sequence data Genetics 2002 161 1307 1320 12136032
Drummond AJ Rambaut A BEAST, version 1.2 [computer program] 2003 Available: http://evolve.zoo.ox.ac.uk/beast/ . Accessed 25 May 2005
Rambaut A Drummond AJ Tracer, version 1.1 [computer program] 2004 Available: http://evolve.zoo.ox.ac.uk/software.html?id=tracer . Accessed 25 May 2005
| 15974804 | PMC1159165 | CC BY | 2021-01-05 08:28:16 | no | PLoS Biol. 2005 Aug 28; 3(8):e241 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030241 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 1597480310.1371/journal.pbio.0030242Research ArticleCancer BiologyCell BiologyDevelopmentGenetics/Genomics/Gene TherapyHematologyOncologyMus (Mouse)The Ews-ERG Fusion Protein Can Initiate Neoplasia from Lineage-Committed Haematopoietic Cells Ews-ERG Fusion Causes Leukaemia in Invertor MiceCodrington Rosalind
1
Pannell Richard
1
Forster Alan
1
Drynan Lesley F
1
Daser Angelika
1
Lobato Nati
1
Metzler Markus
1
Rabbitts Terence H [email protected]
1
1MRC Laboratory of Molecular Biology, Cambridge, United KingdomMarshall Chris Academic EditorInstitute for Cancer ResearchUnited Kingdom8 2005 28 6 2005 28 6 2005 3 8 e24217 11 2004 9 5 2005 Copyright: © 2005 Codrington et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Cancer-Causing Genes Can Convert Even the Most Committed Cells
The EWS-ERG fusion protein is found in human sarcomas with the chromosomal translocation t(21;22)(q22;q12), where the translocation is considered to be an initiating event in sarcoma formation within uncommitted mesenchymal cells, probably long-lived progenitors capable of self renewal. The fusion protein may not therefore have an oncogenic capability beyond these progenitors. To assess whether EWS-ERG can be a tumour initiator in cells other than mesenchymal cells, we have analysed Ews-ERG fusion protein function in a cellular environment not typical of that found in human cancers, namely, committed lymphoid cells. We have used Ews-ERG invertor mice having an inverted ERG cDNA cassette flanked by loxP sites knocked in the Ews intron 8, crossed with mice expressing Cre recombinase under the control of the Rag1 gene to give conditional, lymphoid-specific expression of the fusion protein. Clonal T cell neoplasias arose in these mice. This conditional Ews gene fusion model of tumourigenesis shows that Ews-ERG can cause haematopoietic tumours and the precursor cells are committed cells. Thus, Ews-ERG can function in cells that do not have to be pluripotent progenitors or mesenchymal cells.
Using a mouse model, these authors study the potential of a known cancer-related gene to cause tumors in cells committed to different lineages.
==== Body
Introduction
Chromosomal translocations are characteristic of many human cancers, and are especially well studied in leukaemias, lymphomas, and sarcomas [1,2]. Epithelial tumours also carry chromosomal translocations, but while those in leukaemias, lymphomas, and sarcomas tend to be reciprocal translocations, those in carcinomas are often non-reciprocal translocations [3]. The main outcomes of the reciprocal translocations are either forced oncogene expression, as found in lymphoid malignancies, or gene fusion, found in both leukaemias and sarcomas [1,2]. Gene fusion occurs when the translocation breakpoints are within the introns of genes, such that the two translocated chromosomes have new exon organisation, leading to the formation of chimaeric mRNA species and, in turn, chimaeric proteins.
Some translocations must function within stem cells to initiate disease while others function in both stem cells and more committed cells to provide related or distinct functions (reviewed in [4]). Different tumour phenotypes can result from specific versions of related fusion proteins, for instance, the Philadelphia t(9;22) translocation (which yields the BCR-ABL [breakpoint cluster region gene–Abelson leukaemia oncogene] fusion) occurs in both myeloid- and lymphoid-lineage tumours, dictated by the position of the translocation junction within the BCR gene [5]. The biological consequences of the BCR-ABL fusion have been studied in relation to stem cell properties [6–8]. These studies show that the BCR-ABL fusion protein does not confer self-renewal potential on committed myeloid progenitors, whereas other translocation fusion proteins (e.g., MOZ-TIF2) do [7]. The MLL (mixed lineage leukaemia) gene has more than 30 known fusion partners [9,10], and retroviral transduction experiments with MLL-ENL fusion showed that the multipotent myeloid progenitors and committed myeloid progenitors could respond to an MLL fusion protein to become leukaemic, suggesting the existence of cancer stem cells distinct from multipotent stem cells of the tissue of origin [11]. While some translocations have these dual properties, others only function in committed cells such as those translocations mediated by the RAG-VDJ (recombinase activating gene and variable, joining, diversity regions) recombinase of the lymphoid lineage [1]. Thus, naturally occurring translocations, such as that of LMO2
(LIM domain only 2) in T cell acute leukaemia, only function in committed cells.
Such issues are not confined to haematopoietic malignancies. The EWS-FUS family, typically involved in sarcoma aetiology, exhibits characteristics of tumour initiators [12–15]. The EWS gene, at Chromosome 22 band q12, was first identified in Ewing's sarcoma by its association with FLI1 and ATF1 [16,17] and subsequently found in subsets of Ewing's sarcoma with either ERG or ETV1, and E1AF and FEV [17–21]. In addition, the EWS gene is involved in other sarcomas and with yet different partners, such as WT1 (Wilms tumour 1 gene) in desmoplastic small round cell sarcoma [22] or CHN in myxoid chondrosarcoma [23]. A further ramification of this infidelity is that the FUS gene, first identified by its fusion with the CHOP gene in t(12;16) of malignant myxoid liposarcoma and related to the EWS gene [24,25], has also been found in Ewing's sarcoma fused with the ERG gene, but also a similar FUS-ERG fusion has been described in some acute myeloid leukaemias [26–28]. Finally, the EWS gene can also be involved with the CHOP gene by chromosomal translocation in malignant myxoid liposarcoma [29], analogous to FUS-CHOP. The chromosomal translocations in Ewing's and other sarcomas are considered as primary initiating events. Thus the EWS fusions function in the cancer precursor cell and are required throughout tumour formation until the emergence of the overt cancer, because the chromosomal translocation is present in the cancer at the time of presentation. The fusion protein may therefore be instructive by imparting a phenotype on the cell by affecting a specific differentiation programme, which must be part of the genetic make-up of the cancer stem cell, since the chromosomal translocation is an initiating and persistent feature. The pre-existing chromosomal translocation creates a cellular environment allowing secondary mutations to arise in progeny cells, resulting in overt cancer. Conditional mouse models of the EWS- and FUS-associated chromosomal translocations should give insights into how these fusions influence the malignant phenotype. We have analysed whether EWS fusions function only in the domain of mesenchymal progenitors and whether they function in other cell lineages, specifically within committed cells, as has been demonstrated for MLL fusions [11].
We have developed mouse models mimicking chromosomal translocations by employing homologous recombination in embryonic stem (ES) cells to create mutant alleles in mice. These methods were either the knock-in model, which involved fusing the MLL gene partner AF9 into the mouse Mll gene, resulting in acute myeloid leukaemias in mice [30,31], or the translocator model [32,33] to give de novo chromosomal translocations, resulting in cell-specific leukaemias [34]. We have applied a new conditional chromosomal translocation model (designated the invertor model) [35] to the EWS-ERG gene fusion to gain insights into the functional significance of the fusion protein in tumourigenesis. Invertor mice are made by knocking a floxed cDNA cassette into the intron of a target gene downstream of the mouse exon equivalent to that usually involved in human chromosomal translocations [35]. The floxed cassette is introduced into the intron in a reverse transcriptional orientation and can be inverted by Cre recombinase to create a transcription unit capable of generating an appropriate fusion mRNA (see below and Figure 1A). We have sought to determine whether an Ews-ERG fusion can be oncogenic in a cell type not generally associated with human tumours (specifically in lymphoid cells) and also to determine whether Ews-ERG fusion can initiate leukaemia from committed cells. We found that clonal T cell lymphoma arose in the invertor model when Cre expression was controlled by the lymphocyte-specific Rag1 gene, showing that Ews-ERG is leukaemogenic in lymphoid cells. Our results show that Ews-ERG can function as an oncogene in committed cells of mice and suggest that EWS-ERG is able to contribute to neoplasia in a variety of cellular contexts in vivo.
Figure 1 Strategy for Generating Ews-ERG Invertor Gene by Homologous Recombination
(A) The method for making invertor mice is described in detail elsewhere [35]. In summary, an invertor cassette comprising a short intronic region with an acceptor splice site, human ERG cDNA sequence (shown in Figure S1), a polyA site, and the MC1neopA gene all flanked by loxP sites (depicted by black triangles) was knocked in, using homologous recombination, into Ews intron 8. The transcription orientation of the targeted ERG cDNA invertor cassette was opposite from that of the endogenous Ews gene after initial homologous recombination. Germ line carrier mice of this targeted allele were crossed with Cre-expressing mouse strains [36], and, as indicated in the right hand side of the diagram, the invertor cassette is turned around to create a transcription orientation identical with that of Ews. Thus, after transcription, a pre-mRNA is made with the donor splice site of Ews exon 7 adjacent to the acceptor site of the invertor cassette, allowing post-transcriptional fusion of Ews with ERG in an analogous format to that found in human sarcomas with t(21;22).
(B) Genomic sequence adjacent to the Ews exon 7 and the ERG invertor cassette (the derived amino acid sequence is shown in the single letter code) obtained from DNA of ES cells and of thymus after Cre-mediated inversion. The Ews intron 7/8 donor splice site is indicated by an arrow. The boxed sequence (
TCTAG/
CGAT) corresponds to the ligation of filled-in XbaI (Ews genomic) and ClaI (ERG invertor cassette) sites (for detailed description see [35]). The boxed XbaI site (
TCTAGA) corresponds to the position of cloning of the mouse Af4 intronic sequence (shaded) in the ERG invertor cassette. On the 3′ side of the fused XbaI-ClaI sites, there is a loxP site originating from the ERG invertor cassette (see Figure S1) followed by the Af4 intron 4 (shaded) upstream of the ERG cDNA sequence. The Af4 acceptor splice site is indicated by an arrow, and is followed by the ERG cDNA sequence. A NotI site used for cloning the amplified ERG sequence is boxed (see also Figure S1) (note this sequence is non-contiguous and the dots represent a gap; the full sequence appears in [35]).
Results
Conditional Inversion of the ERG Gene and Fusion with Ews
The use of homologous recombination to generate oncogene fusions was established with the creation of the Mll-AF9 mouse model [30]. We attempted a similar approach to make Ews fusions of the type found in specific human sarcomas and leukaemias (Figure S1 shows sequences of Chop,
ATF1, and Fli1 together with AF9 sequences used for the knock-in targeting clones). While we were able to obtain homologous recombination knock-in clones either with a vector carrying only a transfer cassette or with the Ews-AF9 control fusion, no clones were obtained with any of the Ews fusions equivalent to those naturally found in human cancers (Table S1). Furthermore, chimaeras could be made with the Ews cassette and the Ews-AF9 knock-in (Table S1). We concluded that Ews fusions of the type naturally associated with human cancers could be lethal in ES cells, obviating the generation of targeted cells.
We have recently developed a new conditional knock-in method based on inversion of a loxP-flanked cDNA cassette by Cre recombinase [35]. This Ews-ERG invertor model is diagrammatically shown in Figure 1A. In outline, homologous recombination in mouse ES cells was used to introduce a floxed ERG cDNA cassette downstream of Ews exon 7. The cassette comprised a short intronic sequence, an acceptor splice site, ERG cDNA, a polyA site, and a neomycin gene (to select homologous recombinant ES cells), placed in the opposite transcription orientation to the Ews gene. Expression of Cre recombinase inverts the cassette, bringing the acceptor splice site into the correct orientation with the Ews exon 7 donor splice site, to allow post-transcriptional production of Ews-ERG fusion mRNA (Figure 1B shows the post-inversion genomic sequence at the junction of Ews and ERG). Targeted ES cells were injected into blastocysts, and these yielded normal numbers of chimaeras, which gave rise to heterozygous carriers of the Ews-ERG invertor allele (invertor mice). In these invertor mice, Ews-ERG protein can only be made after Cre-mediated inversion and thus can be cell-specifically dependent on Cre expression.
The Ews-ERG Fusion Protein Causes T Cell Neoplasia in Invertor Mice
We investigated the ability of Ews-ERG fusion protein to contribute to leukaemogenesis by causing the inversion of the ERG-containing cassette in lymphoid cells using Rag1-Cre knock-in mice [36]. Out of the cohort of 29 mice carrying both Ews-ERG and Rag1-Cre genes, 25 mice (86%) developed leukaemia associated with thymoma within 500 d, whereas the Ews-ERG-only cohort (20 mice) did not display neoplasia (Figure 2). Blood smears at the time of sacrifice of Ews-ERG; Rag1-Cre mice typically showed elevated numbers of large leukocytes, with lymphoid morphology of different maturities, including lymphoblasts (Figure 2). In addition, bone marrow sections showed high levels of infiltrating lymphoblasts (Figure 2). The diseased mice all had large thymomas, usually associated with splenomegaly (Table 1) and lymphadenopathy, and the histology of spleens showed either complete loss of normal architecture (loss of demarcated white and red pulp), partial loss, or normal architecture (Figure S2). Similarly, the level of infiltration of leukaemic lymphocytes into liver and kidney varied, with some having large amounts of perivascular deposits and others marginal levels. All mice had thymomas with homogeneous presence of large leukaemic cells. These characteristics, together with evidence of clonal Tcrb gene rearrangements and expression of TCR-associated CD8 and CD4 (see below), permit diagnosis of large cell anaplastic T cell lymphoma according to the Bethesda proposals for lymphoid tumours in mice [37].
Figure 2 Incidence of Haematological Malignancy and Characteristics of Ews-ERG Invertor Mice
Cohorts of 29 Ews-ERG; Rag1-Cre mice and 20 Ews-ERG control mice were monitored over a period of approximately 17 mo. Mice were culled and a post-mortem conducted when signs of ill health were observed. Leukaemia/lymphoma was established by various criteria (see Materials and Methods). Top panel shows the survival curve of Ews-ERG; Rag1-Cre (+Cre) or Ews-ERG (−Cre) mice as a function of time (days). Bottom panel shows the histology of blood and bone marrow from leukaemic mice (M21 and M20, respectively). Blood smears were stained with May-Grünwald-Giemsa stain and photographed at 400× magnification, whilst bone marrow was photographed at 1,000×.
Table 1 Tumour Characteristics of Ews-ERG Invertor Mice Expressing Cre Recombinase from the Rag1 Gene
The Ews-ERG fusion appears crucial for malignancies in the invertor mice, as disease only arose in mice with both the Ews-ERG and Rag1-Cre genes (Figure 2) and comprised clonal tumours (Figure 3). The presence of Ews-ERG mRNA was confirmed in spleens of these mice, identical with that in Ews-ERG ES cells transfected with Cre expression plasmids (see Figure S1). In addition, the steady state orientation of the loxP-flanked ERG cassette was studied using filter hybridisation and we found that the ERG sequence had become inverted into the 5′ to 3′ orientation with respect to the Ews gene in each case of thymoma (Figure 3A; Table 1). The other tissues examined have little evidence of an inverted band. Finally, the presence of the Ews-ERG fusion protein was shown by Western blotting of thymoma proteins using antibodies binding to either Ews or ERG, which detected respectively the normal mouse Ews or Erg proteins and the Ews-ERG fusion molecule in thymoma T cells of an Ews-ERG; Rag1-Cre mouse (Figure 3D).
Figure 3 Clonality of T Cell Neoplasias inEws-ERG Invertor Mice
Genomic DNA was prepared from various tissues of invertor mice with thymomas. Genomic analysis was carried out using filter hybridisation to assess the presence of the inverted ERG cassette in the tumour cells (A), and to assess whether the lymphocytes involved in the tumours were clonal T (B) or B cells (C).
(A) Inversion hybridisation autoradiograph. DNA was prepared from the thymoma and other tissues of M18, cleaved with EcoRI and hybridised to the 5′ Ews probe (which detects a 5-kb targeted ERG cassette fragment or a 6.5-kb Cre-inverted fragment). If the ERG cassette is inverted by Cre activity, the size of the EcoRI fragment increases from the initial targeted gene size, as indicated in the maps below the figure. The data shown are for DNA extracted from spleen (spl), thymus (thy), liver (liv), kidney (kid) or tail (ES cell DNA is used as a control). The Cre-mediated inverted band (∼6.5 kb) is evident in thymus DNA (thymoma). The summary of the data from the cohort of invertor mice is in Table 2. Note that despite extensive infiltration of spleen detected by histology, we cannot see the inverted band by Southern blot; this presumably reflects regional clustering of neoplastic cells in spleen. Restriction fragment sizes are represented as germ line (GL), inverted allele of ERG cassette (inv) and initial targeted Ews allele (Tgt). The organisation of the targeted Ews allele is indicated underneath. The hybridisation autoradiograph shows the location of Ews exon 7 and the initial targeted orientation (bottom) or inverted orientation of ERG invertor cassette after Cre-mediated recombination (top). a, acceptor splice site.
(B) Autoradiograph showing rearrangement of T cell receptor β locus. A T cell receptor Jβ2 probe [40] (diagrammatically shown below the autoradiograph) detects a 5-kb germ line HindIII Jβ2 band whilst V-D-Jβ2 joins in T cells result in new HindIII-sized bands depending on the nature of the rearrangement. Each of 12 thymoma DNAs that were compared showed one or two Jβ2 alleles rearranged, signifying that these tumours were clonal T cells. In some thymoma samples, there is almost complete absence of the germ line band, indicating that the thymuses of these mice are solely comprised of malignant clonal T cells. DNA sequence analysis of the V-D-J junctions of mice M2, M5, M13, and M18 showed functional V-D-J joins (see Table 2).
(C) Autoradiograph showing rearrangement status of immunoglobulin heavy-chain genes. A Cμ intron probe [41] (diagrammatically shown below the autoradiograph) was used to hybridise a set of thymoma DNAs for the presence of Igh rearranged bands. Only two samples showed rearrangements.
(D) Detection of Ews-Erg fusion protein in thymoma cells. Single cell suspensions were made of T cells from a normal thymus (wt) or from the thymoma of M6, protein fractionated on 4%–20% acrylamide gel, and transferred to nylon membranes. Specific proteins were detected with anti-Ews or anti-ERG antibody.
Table 2 Sequences of Rearranged TCR Jβ Loci in Ews-ERG Thymoma DNA
Studies of progenitor gene expression have indicated the promiscuous expression of lymphoid markers precedes lineage commitment [38] and thus, the Rag1-Cre allele might be expressing in cells destined to become non-lymphoid. The Rag1-Cre knock-in allele was previously shown to be specific for lymphoid cells using a reporter assay dependent on Cre-mediated deletion of a loxP-flanked (floxed) segment of the Lmo2 gene [36]. We have sustained these observations using an additional reporter assay, namely the ROSA-loxSTOP-lacZ (ROSA26-R) allele [39] where β-galactosidase (βgal) expression is activated by deletion of a loxP-pA site. Haematopoietic populations from mice carrying both Rag1-Cre and ROSA26-R alleles were stained with fluorescent antibodies binding to various surface markers, and flow cytometry was carried out for co-detection of antibody- and βgal-derived fluorescence (Figure S3). βgal signal was found in thymus and spleen (almost all cells), and in bone marrow, a population of βgal+ cells exist but did not co-express with CD34, Sca1, Ckit, or Ter119 (or Mac1, not shown). This means that the Rag1-Cre is restricted to lymphoid cells, as previously shown [36].
The Rag1-Cre gene is expressed in both B and T cells, but only T cell tumours have arisen in the Ews-ERG; Rag1-Cre invertor line. The possible inversion of the Ews-ERG gene in B cells was investigated using RT-PCR analysis of expressed Ews-ERG fusion mRNA (Figure 4). RNA was prepared from whole spleen or thymus or from flow-sorted B220+ spleen cells (3,400 cells) or Thy1.2+ spleen cells (6,400 cells) and RT-PCR performed. Pax5 and CD3 primers [38] were used for specific detection of B cell and T cell transcripts, respectively. Pax5 transcripts were detected in cDNA made from spleen and B220+ sorted cells, and CD3 in the spleen, thymus and thy1.2+ sorted cells (Figure 4A and 4B). No evidence of CD3 expression was found in the sorted B220+ cells, or of Pax5 in the sorted Thy1.2+ cells, showing that the sorted cells were practically free of contaminating T or B cells, respectively. The presence of Ews-ERG fusion RNA was analysed with RT-PCR primers, yielding a product spanning the fusion junction that was detected with an internal junction probe. Ews-ERG RT-PCR product was detected in the unfractionated spleen and thymus sources, as well as in the purified, sorted B220+ and Thy1.2+ cells. Therefore, Cre-mediated inversion of the Ews-ERG gene occurs in both T and B cells.
Figure 4 B and T Cells Express the Ews-ERG Fusion RNA
A 96-d-old mouse with both Ews-ERG and Rag1-Cre alleles was used as a source of spleen and thymus cells. Single cell suspensions of spleen cells were labelled with anti-B220 or with anti-Thy1.2 and were purified using a MoFlo preparative flow cytometer. Estimated purities were achieved of greater than 95%. cDNA was prepared from RNA extracted from sorted cells or from aliquots of unsorted populations and RT-PCR (approximately 3,400 B220+ or 6,400 Thy1.2+ cell equivalents per PCR reaction) carried out with specific Pax5 (A), CD3 (B) or Ews-ERG (C) primers. PCR reaction products were fractionated on 1% agarose gels and either stained with ethidium bromide and photographed (A and B) or gel blotted and hybridised with an Ews-ERG probe (C)
Ews-ERG Induces Malignancies of Mature T Cells
The presence of thymomas in all afflicted Ews-ERG; Rag1-Cre mice suggested that the haematological malignancy in these mice comprised T cells. This was confirmed by FACS analysis of T cell surface markers and by determination of T cell or B cell receptor gene rearrangement status. The majority of normal mouse thymic T cells express both the CD4 and CD8 surface markers (double positive [DP] cells), as well as the pan-T cell marker Thy1 (CD90) (wt in Figure 5). The situation with the thymomas of invertor mice varied from mainly CD4+CD8+ DP cells (e.g., M2 in Figure 5) to mainly CD8+ single positive (SP) cells (e.g., M3 in Figure 5) to mainly double negative cells (e.g., M4 in Figure 5). Thirteen of the cohort of 25 leukaemic invertor mice were analysed for their thymic T cell profile by FACS (summarized in Table 1). While we found either CD8+ SP or CD4+CD8+ DP thymomas, none showed a CD4+ SP phenotype. Two of the thymomas had a mixed phenotype comprising mainly CD8+ SP cells, with a sub-population of DP cells (M3 and M6 in Table 1).
Figure 5 Analysis of Cell Surface Antigens of Thymomas from Ews-ERG Invertor Mice Using Flow Cytometry
Thymus tissue was resected from mice with thymoma and FACS analysis performed to determine T cell phenotype (summarised in Table 1). The data in the figure show representative flow diagrams of three tumour-bearing mice (M2, M3, and M4) compared with a wild-type C57BL/6 control (wt), using anti-Thy1 (y-axis) plus anti-B220 (x-axis) antibodies or using anti-CD8 (y-axis) plus anti-CD4 (x-axis) antibodies. The three thymomas show a range of CD4 and CD8 co-expression phenotypes characterizing the thymomas generated in the Ews-ERG; Rag1-Cre invertor mice..
The clonality of these T cell tumours was assessed by filter hybridisation of genomic thymoma DNA with a T cell receptor β probe (Tcrb) from the Jβ2 region [40] (diagrammatically shown in Figure 3B). The probe detects a 5-kb band in non-lymphoid DNA corresponding to the intact Tcrb gene; if V-D-J or D-J joins have occurred, new restriction fragments are created, giving rise to “rearranged” bands on the hybridisation autoradiograph. Figure 3B shows the rearrangement status of 12 of the Ews-ERG thymomas. All except one (M9) showed at least one rearranged Jβ2 allele and several have two rearranged alleles (e.g., M18). Variable amounts of residual germ line allele were present. The Tcrb rearrangement status of the cohort of Ews-ERG leukaemic mice is summarized in Table 1. In addition, the status of the immunoglobulin locus was examined with a probe from the region between JH and Cμ segments [41] (see Figure 3C). Single Igh allele rearrangements were observed in about half of the cohort (Table 1), including ones in which FACS analysis and Tcrb clonality hybridisations showed the tumours to be clonal T cell malignancies. The Igh rearrangements are therefore likely to be abortive rearrangements.
Confirmation that the Tcrb rearrangements observed by hybridisations were due to productive V-D-J joins was made by sequence analysis of the Tcrb genes in four of the cases. PCR amplification from thymoma DNA was carried out with pools of Vβ primers and a Jβ2 reverse primer (primer sequences from [42,43]) and sequences identified with the ImMunoGeneTics database [44,45]. Table 2 shows the V-D-J junctional sequences indicating the presence of productive T cell receptor β genes. In all cases, N-region diversity is present between the V-D and D-J junctions. In the thymoma DNA from M2, a non-productive join has also occurred, in addition to the productive one, resulting in an out-of-frame joint.
Discussion
EWS-ERG Can Mediate Haematopoietic or Mesenchymal Tumourigenesis
Human cancers of mesenchymal origin involve the EWS gene with various partner genes, such as in the chromosomal translocation t(21;22)(q22;q12) where EWS fuses with ERG to encode a novel EWS-ERG fusion protein [18]. In addition, EWS gene fusions with various partners are described in clear cell sarcoma, desmoplastic small round cell tumours, chondrosarcoma, and myxoid liposarcoma, suggesting that EWS fusion proteins are functional for various cell types of mesenchymal origin and that restrictions in humans are dictated largely by the cells in which the chromosomal translocations occur. We tested the hypothesis of cell type specificity by assessing the possible universality of EWS-ERG as an oncogene. Our experimental system was designed to invoke conditional, cell-specific expression of the Ews-ERG fusion to analyse the target cell in which it can function. Consistent EWS-ERG occurrence in human sarcomas suggests that it may only function in mesenchymal cells, where it may function in initiation of the cancer as well as in maintenance of the cancer stem cell (i.e., cells within the tumour that have self-renewing capacity and maintain the tumour). Our results show that lymphomas arise when Ews-ERG is aberrantly expressed in the committed cells of the lymphoid lineage.
The Ews-ERG invertor allele was activated in lymphocytes, using the Rag1-Cre knock-in mice [36], to determine if lymphoid malignancies would arise. We found that invertor Ews-ERG mice develop clonal T cell malignancies. The T cells were of varying phenotypes but the majority of cases expressed CD8 on the cell surface and had productive V-D-Jβ rearrangements, indicative of mature T cells. Since Cre-mediated inversion through loxP sites recreates the loxP sequence, there is potential for the cassette to flip back and forth with continued Cre expression. This had no obvious deleterious effect on tumourigenesis in the Ews-ERG invertors as judged by various criteria (tumour penetrance, consistent T cell phenotype, and pathology). Thus, the invertor allele of transformed cells may be fixed in the correct orientation for synthesis of Ews-ERG mRNA when Rag1-Cre is no longer effective. On the other hand, the lack of CD4+ SP thymomas is difficult to encompass with this explanation, suggesting that other biological mechanisms may be involved in this bias.
The Rag1-Cre mouse line used constitutively expresses Cre in lymphoid cells [36], and the resulting Ews-ERG fusion causes T lymphocyte tumours in mice, which is not known in the spectrum of EWS-associated human cancers. On the other hand, FUS-ERG has been found in leukaemias, albeit myeloid type [26,46]. This implies that the EWS-FUS fusion proteins could be effective in a spectrum of cell types, broader than normally seen in human malignancies (i.e., sarcomas and myeloid leukaemias). In some cells, it may be that EWS-FUS fusions are lethal and thus those cells acquiring a translocation would die; in others, the fusion protein may be tolerated and thus may become tumours. In this respect, the absence of B cell tumours in the Ews-ERG invertors is of interest as both B and T cells undergo inversion of the Ews-ERG cassette (see Figure 4) because Rag1-Cre is expressed in both cell types [36] (see Figure S3). The absence of B cell tumours may reflect toxicity of the fusion protein for B cells although this seems unlikely given that we can detect the fusion mRNA in selected B220+ B lymphocytes (see Figure 4). Alternatively, it may mean that Ews-ERG does not function in B cells or that the development of B cell tumours is inhibited by the T cell malignancies, which may arise earlier and then dominate the lymphoid compartment. The ability of Ews-ERG to specifically cause B cell tumours could be evaluated by using B-cell-specific Cre-expressing mouse lines such as CD19-Cre.
Ews-ERG Can Promote Tumourigenesis in Committed Cells
Our results show that the Ews-ERG fusion, normally restricted to sarcomas in humans, can initiate T lymphocyte tumours, if conditionally expressed in committed lymphoid cells. Leukaemogenesis in Ews-ERG invertor mice thus concurs with the hypothesis that EWS fusions can cause neoplasia arising in various cell types. Further, our data imply that the cellular context of EWS-associated chromosomal translocations in humans does not have to be a stem cell or even a multi-potent progenitor, as committed lymphocyte precursors expressing Rag recombinase genes are the leukaemic precursors in our invertor model. The same conclusions can be applied to FUS-associated chromosomal translocations that seem to be largely interchangeable with EWS, given the spectrum of cancers in which these are found and the relatedness of FUS- and EWS-coded proteins [24,25]. Thus, EWS or FUS chromosomal translocations probably arise more by virtue of accessibility of the genes to chromosomal translocations than by the precise cellular specificity of the resultant fusion proteins.
Materials and Methods
Generation of targeting constructs and homologous recombination
The Ews genomic targeting clone was constructed from γ phage DNA isolated from mouse 129 DNA [47]. The knock-in clones for potential Ews fusions with AF9, Fli1, Chop, and ATF1 were made using a transfer vector pC2A-neo [35], and the sequences of these cDNA cassettes are given in Figure S1. Knock-in was achieved by homologous recombination as described [47]. The ERG inversion knock-in cassette [35] was prepared using cDNA made from human colon carcinoma COLO320 mRNA (the sequence of the ERG segment, corresponding to the part of ERG found in EWS-ERG fusion [19] is shown Figure S1). A diagram of overall structure of the ERG inversion cassette is shown in Figure 1A, indicating the initial orientation of the ERG inversion cassette (3′ to 5′ with respect to Ews) and this was determined by restriction site mapping and genomic DNA sequencing. After identification of targeted ES cells, these were injected into C57BL/6 blastocysts, chimaeric mice were produced, and germ line transmission of the inversion knock-in allele was obtained. These mice were bred with mice expressing Cre from a Rag1 knock-in [36]. Specificity of expression of the Rag1-Cre allele has previously been described [36] and was sustained using the ROSA26-R reporter mice [39] (see Figure S3).
Molecular analyses
Genomic DNA was prepared from tissues using proteinase K digestion and phenol-chloroform extraction, and filter hybridisations were carried out with radiolabelled probes as described [48,49]. The probes used to detect antigen receptor gene rearrangements were a heavy-chain Ig μ intron enhancer probe [41] and a T cell receptor Jβ2 probe [40]. T cell receptor V-D-Jβ junction sequences were obtained by PCR amplification from thymoma DNA with pools of Vβ primers and a Jβ2 reverse primer (primer sequences from [42,43]), and the product was fractionated on agarose to determine the presence of a single amplified band. In turn, each band was eluted and the sequence obtained using the Jβ2N primer [43] and identified by comparison with the ImMunoGeneTics database [44,45].
RT-PCR was carried out on cDNA as described [34]. RNA was prepared using RNeasy (Quiagen, Valencia, California, United States) from cells that were labelled with FITC-conjugated antibodies and isolated by flow cytometry. Sequences have been described for Pax5 and CD3 RT-PCR primers [38], and the sequences of the Ews-ERG primers were 5′-
CCACAGGATGGTAACAAGCCTGC-3′ (Ews) and 5′-
CGAACTTGTAGGCGTAGC-3′ (ERG). The hybridisation probe was a 303-bp fragment of Ews-ERG residing within the RT-PCR product.
Analysis of leukaemia/lymphoma
A cohort of mice was established by inter-breeding the Ews-ERG invertor line with a Rag1-Cre line [36] to generate littermates with Ews-ERG + Rag1-Cre or Ews-ERG genotypes. Genotypes were determined using filter hybridisation of genomic DNA from a small tail biopsy. The health status of these mice was monitored, and if signs of ill health appeared, mice were sacrificed and a post-mortem was carried out. Tissue samples were removed for single cell preparation for determination of surface protein expression phenotype, for nucleic acid preparation, or for fixation in 10% formalin for histology. Blood smears were prepared and stained with May-Grünwald-Giemsa. For histology, fixed tissues were embedded in wax and 0.4-μ sections made, stained with haematoxylin and eosin after mounting on slides. FACS analysis was conducted using a FACSCalibur with fluorescent antibodies purchased from BD Biosciences (San Jose, California, United States). Data were analysed with CellQuest software (BD Biosciences). Western protein detection was carried out as described [36] using 18 μg of protein per lane and proteins fractionated on 4%–20% SDS-PAGE. The separated proteins were electro-transferred to PVDF nylon membranes (Millipore, Billerica, Massachusetts, United States) and specific proteins detected using anti-Ews antibody (Santa Cruz Biotechnology, Santa Cruz, California, United States; raised against the amino terminus of Ews) or anti-ERG antibody (Santa Cruz Biotechnology; raised against the carboxy terminus of ERG). Antibody bound to filter was detected using secondary antibodies and ECL as described by the manufacturer (Amersham Biosciences, Amersham, United Kingdom).
Supporting Information
Figure S1 Sequence of Mouse FliI RT-PCR Fragment
(111 KB PDF).
Click here for additional data file.
Figure S2 Histological Characteristics of Leukaemias in Ews-ERG Invertor Mice
(6.4 MB PDF).
Click here for additional data file.
Figure S3 ROSA26-R-βgal Reporter Assay for Expression of Rag1-Cre Allele
(354 KB PDF).
Click here for additional data file.
Table S1 Frequency of Targeted Clones and Chimaera Generation with Ews Knock-In ES Clone
(23 KB DOC).
Click here for additional data file.
This work was supported by the Medical Research Council. RC was the recipient of a Leukaemia Research Fund Gordon Pillar Studentship. NL was the recipient of a Kay Kendall fellowship and MM is the recipient of a fellowship from the German Research Foundation. We thank Dr. Philippe Soriano for providing the ROSA26-R reporter mice. We are very grateful to Dr. Barbara Bain for analysing the morphology of the abnormal lymphocytes. We thank Gareth King, Angela Middleton, and Claire Pearce for animal husbandry and Annette Lenton for the illustration work.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. THR conceived and designed the experiments. RC, RP, AF, LFD, AD, NL, and MM performed the experiments. RC, RP, AF, LFD, AD, NL, MM, and THR analysed the data. RP, AF, and LFD contributed reagents/materials/analysis tools. THR wrote the paper.
Citation: Codrington R, Pannell R, Forster A, Drynan LF, Daser A, et al. (2005) The Ews-ERG fusion protein can initiate neoplasia from lineage-committed haematopoietic cells. PLoS Biol 3(8): e242.
Abbreviations
β-galβ-galactosidase
DPdouble positive
ESembryonic stem cells
ROSA26-R
ROSA-loxSTOP-lacZ
SPsingle positive
==== Refs
References
Rabbitts TH Chromosomal translocations in human cancer Nature 1994 372 143 149 7969446
Look AT Oncogenic transcription factors in the human acute leukemias Science 1997 278 1059 1065 9353180
Mitelman F Johansson B Mertens F Mitelman Database of Chromosome Aberrations in Cancer [database] 2000 Bethesda (Maryland) National Cancer Institute Available: http://cgap.nci.nih.gov/Chromosomes/Mitelman . Accessed 19 May 2005
Huntly BJ Gilliland DG Leukaemia stem cells and the evolution of cancer-stem-cell research Nat Rev Cancer 2005 5 311 321 15803157
Gale RP Grosveld G Canaani E Goldman JM Chronic myelogenous leukemia: Biology and therapy Leukemia 1993 7 653 658 8464245
Jaiswal S Traver D Miyamoto T Akashi K Lagasse E Expression of BCR-ABL and BCL-2 in myeloid progenitors leads to myeloid leukemias Proc Natl Acad Sci U S A 2003 100 10002 10007 12890867
Huntly BJ Shigematsu H Deguchi K Lee BH Mizuno S MOZ-TIF2, but not BCR-ABL, confers properties of leukaemic stem cells to committed murine hematopoietic progenitors Cancer Cell 2004 6 587 596 15607963
Jamieson CH Weissman IL Passegue E Chronic versus acute myelogenous leukemia: A question of self-renewal Cancer Cell 2004 6 531 533 15607956
Ayton PM Cleary ML Molecular mechanisms of leukemogenesis mediated by MLL fusion proteins Oncogene Revs 2001 20 5695 5707
Daser A Rabbitts TH Extending the repertoire of the mixed lineage leukemia gene MLL in leukemogenesis Genes and Develop 2004 18 965 974 15132992
Cozzio A Passegue E Ayton PM Karsunky H Cleary ML Similar MLL-associated leukemias arising from self-renewing stem cells and short-lived myeloid progenitors Genes Dev 2003 17 3029 3035 14701873
Arvand A Denny CT Biology of EWS/ETS fusions in Ewing's family tumors Oncogene 2001 20 5747 5754 11607824
Bennicelli JL Barr FG Chromosomal translocations and sarcomas Curr Opin Oncol 2002 14 412 419 12130926
Sandberg AA Bridge JA Updates on the cytogenetics and molecular genetics of bone and soft tissue tumors: Chondrosarcoma and other cartilaginous neoplasms Cancer Genet Cytogenet 2003 143 1 31 12742153
Sandberg AA Genetics of chondrosarcoma and related tumors Curr Opin Oncol 2004 16 342 354 15187889
Delattre O Zucman J Plougastel B Desmaze C Melot T Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours Nature 1992 359 162 165 1522903
Zucman J Delattre O Desmaze C Epstein AL Stenman G EWS and ATF-1 gene fusion induced by t(12;22) translocation in malignant melanoma of soft parts Nat Genet 1993 4 341 345 8401579
Sorensen PHB Lessnick SL Lopez-Terrada D Liu XF Triche TJ A second Ewing's sarcoma translocation, t(21;22), fuses the EWS gene to another ETS-family transcription factor, ERG Nat Genet 1994 6 146 151 8162068
Jeon IS Davis JN Braun BS Sublett JE Roussel MF A variant Ewing's sarcoma translocation (7;22) fuses the EWS gene to the ETS gene ETV1 Oncogene 1995 10 1229 1234 7700648
Urano F Umezawa A Hong W Kikuchi H Hata J A novel chimera gene between EWS and E1A-F, encoding the adenovirus E1A enhancer-binding protein, in extraosseous Ewing's sarcoma Biochem Biophys Res Commun 1996 219 608 612 8605035
Peter M Couturier J Pacquement H Michon J Thomas G A new member of the ETS family fused to EWS in Ewing tumors Oncogene 1997 14 1159 1164 9121764
Kim J Lee K Pelletier J The desmoplastic small round cell tumor t(11;22) translocation produces EWS/Wt1 isoforms with differing oncogenic properties Oncogene 1998 16 1973 1979 9591781
Clark J Benjamin H Gill S Sidhar S Goodwin G Fusion of the EWS gene to CHN, a member of the steroid/thyroid receptor gene superfamily, in a human myxoid chondrosarcoma Oncogene 1996 12 229 235 8570200
Crozat A Aman P Mandahl N Ron D Fusion of CHOP to a novel RNA-binding protein in human myxoid liposarcoma Nature 1993 363 640 644 8510758
Rabbitts TH Forster A Larson R Nathan P Fusion of the dominant negative transcription regulator CHOP with a novel gene FUS by translocation t(12;16) in malignant liposarcoma Nat Genet 1993 4 175 180 7503811
Shimizu K Ichikawa H Tojo A Kaneko Y Maseki N An ets-related gene, ERG, is rearranged in human myeloid leukemia with t(16;21) chromosomal translocation Proc Natl Acad Sci U S A 1993 90 10280 10284 8234289
Ichikawa H Shimizu K Hayashi Y Ohki M An RNA-binding protein gene, TLS/FUS, is fused to ERG in human myeloid leukemia with t(16;21) chromosomal translocation Cancer Res 1994 54 2865 2868 8187069
Panagopoulos I Aman P Fioretos T Hoglund M Johannson B Fusion of the FUS gene in acute myeloid leukemia with t(16;21)(p11;q22) Genes Chromosomes Cancer 1994 11 256 262 7533529
Panagopoulos I Hoglund M Mertens F Mandhal N Mitelman F Fusion of the EWS and CHOP gene in myxoid liposarcoma Oncogene 1996 12 489 494 8637704
Corral J Lavenir I Impey H Warren AJ Forster A An Mll-Af9 fusion gene made by homologous recombination causes acute leukemia in chimeric mice: A method to create fusion oncogenes Cell 1996 85 853 861 8681380
Dobson CL Warren AJ Pannell R Forster A Lavenir I The Mll-AF9 gene fusion in mice controls myeloproliferation and specifies acute myeloid leukaemogenesis EMBO J 1999 18 3564 3574 10393173
Buchholz F Refaeli Y Trumpp A Bishop JM Inducible chromosomal translocation of AML1 and ETO genes through Cre/loxP-mediated recombination in the mouse EMBO Rep 2000 1 133 139 11265752
Collins EC Pannell R Simpson EM Forster A Rabbitts TH Inter-chromosomal recombination of Mll and Af9 genes mediated by cre-lox P in mouse development EMBO Rep 2000 1 127 132 11265751
Forster A Pannell R Drynan LF McCormack M Collins EC Engineering de novo reciprocal chromosomal translocations associated with Mll to replicate primary events of human cancer Cancer Cell 2003 3 449 458 12781363
Forster A Pannell R Drynan LF Codrington R Daser A The invertor knock-in conditional chromosomal translocation mimic Nat Methods 2005 2 27 30 15782166
McCormack MP Forster A Drynan LF Pannell R Rabbitts TH The LMO2 T-cell oncogene is activated via chromosomal translocations or retroviral insertion during gene therapy but has no mandatory role in normal T-cell development Mol Cell Biol 2003 23 9003 9013 14645513
Morse HC Anver MR Fredrickson TN Haines DC Harris AW Bethesda proposals for classification of lymphoid neoplasms in mice Blood 2002 100 246 258 12070034
Miyamoto T Iwasaki H Reizis B Ye M Graf T Myeloid or lymphoid promiscuity as a critical step in hematopoietic lineage commitment Dev Cell 2002 3 137 147 12110174
Soriano P Generalized lacZ expression with the ROSA26 Cre reporter strain Nat Genet 1999 21 70 71 9916792
Malissen M Minard K Mjolsness S Kronenberg M Goverman J Mouse T cell antigen receptor: Structure and organization of constant and joining gene segments encoding the β polypeptide Cell 1984 37 1101 1110 6611207
Neuberger MS Caskey HM Pettersson S Williams GT Surani MA Isotype exclusion and transgene down-regulation in immunoglobulin-λ transgenic mice Nature 1989 338 350 352 2493585
Aifantis I Buer J von Boehmer H Azogui O Essential role of the pre-T cell receptor in allelic exclusion of the T cell receptor beta locus Immunity 1997 7 601 607 9390684
Gartner F Alt FW Monroe R Chu M Sleckman BP Immature thymocytes employ distinct signaling pathways for allelic exclusion versus differentiation and expansion Immunity 1999 10 537 546 10367899
Lefranc MP IMGT, the international ImMunoGeneTics database: A high-quality information system for comparative immunogenetics and immunology Dev Comp Immunol 2002 26 697 705 12206833
Lefranc MP Pommie C Ruiz M Giudicelli V Foulquier E IMGT unique numbering for immunoglobulin and T cell receptor variable domains and Ig superfamily V-like domains Dev Comp Immunol 2003 27 55 77 12477501
Prasad DDK Ouchida M Lee L Rao VN Reddy ES TLS/FUS fusion domain of TLS/FUS-erg chimeric protein resulting from the t(16;21) chromosomal translocation in human myeloid leukemia functions as a transcriptional activation domain Oncogene 1994 9 3717 3729 7970732
Warren AJ Colledge WH Carlton MBL Evans MJ Smith AJH The oncogenic cysteine-rich LIM domain protein rbtn2 is essential for erythroid development Cell 1994 78 45 58 8033210
Feinberg AP Vogelstein BA A technique for radiolabelling DNA restriction endonuclease fragments to high specific activity Anal Biochem 1983 132 6 13 6312838
LeFranc MP Forster A Baer R Stinson MA Rabbitts TH Diversity and rearrangement of the human T cell rearranging γ genes: Nine germ-line variable genes belonging to two subgroups Cell 1986 45 237 246 2938743
| 15974803 | PMC1159166 | CC BY | 2021-01-05 08:21:24 | no | PLoS Biol. 2005 Aug 28; 3(8):e242 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030242 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030275SynopsisEcologyEvolutionGenetics/Genomics/Gene TherapyMolecular Biology/Structural BiologyPaleontologyZoologyMammalsNew World Pleistocene Horses: Pruning the Equid Tree Synopsis8 2005 28 6 2005 28 6 2005 3 8 e275Copyright: © 2005 Public Library of Science.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Evolution, Systematics, and Phylogeography of Pleistocene Horses in the New World: A Molecular Perspective
==== Body
Darwin's evolutionary theories got a big boost when Othniel Charles Marsh assembled his North American horse fossils in the 1870s. They showed the gradual (presumed linear) evolution of a small, multi-toed Hyracotherium (then known as Eohippus) into the larger, hoofed modern Equus. As new fossils appeared, it became clear that the equid tree—which includes modern horses, zebras, burros, and the like—shows a complex evolutionary pattern with many branches. By the early 1900s, over 50 equid species were named for the Pleistocene (1.8 million years [Ma] to 10,000 years ago) alone.
Since then, the validity, origin, and relationships of these classifications have been revised, especially for the species of the Pliocene (5 Ma to 1.8 Ma ago) and Pleistocene epochs. In a new paper, Jaco Weinstock, Alan Cooper, and their colleagues use ancient DNA to upset conventional models describing the origins of two extinct forms: the North American “stilt-legged” horses and the stocky South American genus, Hippidion.
New World “stilt-legged” horses often appear in fossil deposits along with a second equid form that shares morphological characteristics with Eurasian caballines—a group that includes the domestic horse and the nearly extinct Przewalskii horse of Mongolia. Because stilt-legged horses had builds similar to the onager and kiang, rare wild asses that live in Asia (Asian hemionids), it's been suggested that stilt-leggeds migrated from Asia across the Bering Strait. Hippidion shows up in South American fossils about 2.5 Ma ago. It has been considered a descendant of a primitive Miocene (23.8 Ma to 5.3 Ma ago) horse that diverged from the ancestral Equus lineage about 10 Ma ago. But a recent genetic analysis of Patagonian Hippidion specimens placed the fossils within an extinct Equus lineage that migrated to South America more recently—suggesting that the specimen had been misidentified as Hippidion.
To analyze the horse fossils, Weinstock et al. used a well-established genetic technique based on mitochondrial DNA (mtDNA). Mitochondria provide most of a cell's energy needs, but it is their genome that interests evolutionary biologists—specifically, a stretch of mtDNA sequence called the control region. This region mutates at a high rate, but the patterns of mutations remain stable over thousands of generations, providing a tool for inferring evolutionary relationships.
The authors first extracted mtDNA from horse bones (mostly toes) from Eurasia and North and South America dating back to 53,000 years ago. Their genetic analysis showed that Hippidion, stilt-leggeds, and caballines all arose from a common lineage. Even though stilt-leggeds share morphological traits with the Asian hemionids, they are genetically distinct, suggesting a convergence of form rather than a common ancestry. And because none of the Old World specimens had genetic sequences similar to sequences extracted from the stilt-legged horses, it's likely that the stilt-legged horses were North American endemics. The genetic analysis also showed that stilt-legged specimens from north and south of the ice sheets that bisected North America during long stretches of the Pleistocene belong to the same taxon, suggesting a wide-ranging species.
Like this wild horse from North Carolina, free-ranging American horses descended from European stock. Horses native to the New World became extinct about 10,000 years ago
Hippidion and stilt-leggeds cluster as sister taxa—in direct contrast to paleontological models of Hippidion's ancient form. The sequences are truly from Hippidion fossils (rather than the Equus lineage proposed in a recent report), the authors argue, because all the fossils came from Late Pleistocene cave deposits in southern Patagonia known to contain only Hippidion saldiasi specimens—and all but one specimen came from these deposits. Along with telling morphological and size characteristics of the bones, these results suggest that Hippidion emerged closer to 3 Ma ago.
A final analysis of horse specimens from the Pleistocene, historic, and recent caballines—which have been grouped as separate species based on their diverse size—suggests that all North American caballines may belong to the same species. Altogether, the results suggest that just two horse lineages—caballine and stilt-legged—may have lived in North America during the Late Pleistocene. Both lineages showed regional and temporal variations in size and form. Though these variations have been taken to represent many different species, the authors propose that the two lineages are more likely just two species whose variations reflect adaptations to different environments. If true, this model could provide a tool for exploring how environmental adaptations give rise to morphological variation.
| 0 | PMC1159167 | CC BY | 2021-01-05 08:21:25 | no | PLoS Biol. 2005 Aug 28; 3(8):e275 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030275 | oa_comm |
==== Front
PLoS BiolPLoS BiolpbioplosbiolPLoS Biology1544-91731545-7885Public Library of Science San Francisco, USA 10.1371/journal.pbio.0030276SynopsisCancer BiologyCell BiologyDevelopmentGenetics/Genomics/Gene TherapyHematologyOncologyMus (Mouse)Cancer-Causing Genes Can Convert Even the Most Committed Cells Synopsis8 2005 28 8 2005 28 8 2005 3 8 e276Copyright: © 2005 Public Library of Science.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
The Ews-ERG Fusion Protein Can Initiate Neoplasia from Lineage-Committed Haematopoietic Cells
==== Body
Healthy, normal cells follow the rules: don't crowd the neighbors, stick to your own tissue, and die when it's time. When cells no longer observe these regulations, they become cancerous, dividing uncontrollably, pushing out their healthy counterparts, and eventually invading other tissues. All cells appear to have the capacity to become cancerous, but most don't. Just what turns a healthy cell into an outlaw remains uncertain, but cancer researchers Terence Rabbitts and colleagues at the MRC Laboratory of Molecular Biology in Cambridge, United Kingdom, think the genetic changes regularly observed in cancer cells can provide some clues.
Tumor cells commonly exhibit chromosomal abnormalities. One type of aberration occurs when DNA strands break and segments between two different chromosomes are swapped, a process called chromosomal translocation. Depending on where these chromosomal breaks occur, the newly fused DNA can produce novel genes called fusion genes. Fusion genes are the result of explicit chromosomal changes associated with different cell types and result in distinct types of cancers. In human connective tissue cancers (called sarcomas), genetic exchange between Chromosomes 21 and 22 produces the EWS-ERG fusion gene; this translocation is thought to initiate tumor formation in undifferentiated “progenitor” cells called mesenchymal cells.
Progenitor mesenchymal cells are long-lived and self-renewing, and can give rise to many specialized cells, including muscle cells, bone cells, and connective tissue cells. Because the EWS-ERG fusion gene in humans was observed only in sarcomas derived from progenitor lineages, researchers thought it initiated a differentiation program that transformed uncommitted non-cancer cells into committed cancer cells.
Genetic exchange between chromosomes can cause cells to become cancerous, like these cells from metastasized Ewing's sarcoma (Image: Lance Liotta Laboratory)
To determine whether the Ews-ERG fusion protein could initiate tumorigenesis in other lineages, particularly lineages of committed cells not typical of human cancers, Rabbitts and colleagues genetically engineered mice to express an Ews-ERG fusion protein exclusively in committed B and T immune cells. Mice that expressed the fusion protein developed T cell tumors, demonstrating for the first time that Ews-ERG could cause blood-borne cancers from committed cells, but B cell tumors were not observed. Since the Ews-ERG fusion protein was expressed in both B and T cells in the mutant mice, these results suggest that other factors influence the fusion protein's potential to cause cancer.
These results reveal important information about the way tumors are generated. The cancer-causing effects of the Ews-ERG fusion protein are neither specific to a given tissue type, nor do they exclusively activate tumor cell differentiation in progenitor cells, as originally thought. Rather, EWS-ERG may be able to act as a universal cancer-causing gene, inappropriately activating signaling pathways responsible for regulating the cellular lifecycle. These observations, if shown to extend to other known cancer fusion genes, may indicate that the apparent tissue specificity of Ews-ERG and other similar fusion proteins stems from a particular propensity for chromosomal translocations in a given cell type, rather than from the specificity of the resulting fusion proteins.
| 0 | PMC1159168 | CC BY | 2021-01-05 08:21:26 | no | PLoS Biol. 2005 Aug 28; 3(8):e276 | utf-8 | PLoS Biol | 2,005 | 10.1371/journal.pbio.0030276 | oa_comm |
==== Front
BMC AnesthesiolBMC Anesthesiology1471-2253BioMed Central London 1471-2253-5-51590449810.1186/1471-2253-5-5Study ProtocolSynergistic effect of intrathecal fentanyl and bupivacaine in spinal anesthesia for cesarean section Bogra Jaishri [email protected] Namita [email protected] Pratima [email protected] Deptt. of Anesthesiology, Kings" George Medical University, Lucknow, India2 Deptt. of Pharmacokinetics and Metabolism, Central Drug Research Institute, Lucknow, India2005 17 5 2005 5 5 5 20 12 2004 17 5 2005 Copyright © 2005 Bogra et al; licensee BioMed Central Ltd.2005Bogra et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Potentiating the effect of intrathecal local anesthetics by addition of intrathecal opiods for intra-abdominal surgeries is known. In this study by addition of fentanyl we tried to minimize the dose of bupivacaine, thereby reducing the side effects caused by higher doses of intrathecal bupivacaine in cesarean section.
Methods
Study was performed on 120 cesarean section parturients divided into six groups, identified as B8, B10 and B 12.5 8.10 and 12.5 mg of bupivacaine mg and FB8, FB10 and FB 12.5 received a combination of 12.5 μg intrathecal fentanyl respectively. The parameters taken into consideration were visceral pain, hemodynamic stability, intraoperative sedation, intraoperative and postoperative shivering, and postoperative pain.
Results
Onset of sensory block to T6 occurred faster with increasing bupivacaine doses in bupivacaine only groups and bupivacaine -fentanyl combination groups. Alone lower concentrations of bupivacaine could not complete removed the visceral pain. Blood pressure declined with the increasing concentration of Bupivacaine and Fentanyl. Incidence of nausea and shivering reduces significantly whereas, the postoperative pain relief and hemodynamics increased by adding fentanyl. Pruritis, maternal respiratory depression and changes in Apgar score of babies do not occur with fentanyl.
Conclusion
Spinal anesthesia among the neuraxial blocks in obstetric patients needs strict dose calculations because minimal dose changes, complications and side effects arise, providing impetus for this study. Here the synergistic, potentiating effect of fentanyl (an opiod) on bupivacaine (a local anesthetic) in spinal anesthesia for cesarian section is presented, fentanyl is able to reduce the dose of bupivacaine and therefore its harmful effects.
==== Body
Background
Spinal anesthesia is the preferred means for cesarean section, being simple to perform, economical and produces rapid onset of anesthesia and complete muscle relaxation. It carries high efficiency, involves less drug doses, minimal neonatal depression, awake mother and lesser incidences of aspiration pneumonitis. However, it also produces a fixed duration of anesthesia, postdural puncture, headache, hypotension and lesser control of block height [1].
Bupivacaine, an amide type of local anesthetic, has high potency, slow onset (5–8 minutes) and long duration of action (1.5–2 hours). For cesarean section intrathecal dose of hyperbaric bupivacaine is 12 to 15 mg [2]. Cesarean delivery requires traction of peritoneum and handling of intraperitoneal organs, resulting in intraoperative visceral pain. With higher doses of hyperbaric bupivacaine, incidence of intraoperative visceral pain associated with higher blocks [4] is reduced [3].
Opiods have been a choice in regional (intrathecal and epidural routes) anesthesia to improve the antinociceptive effect of local anesthetics. Morphine [5], and fentanyl [6], are being used intrathecally, together with local anesthetics in cesarean section.
This study aims to monitor the effect of intrathecal fentanyl + bupivacaine on reduction of higher blocks incidence simultaneously improving the quality and avoiding the complications of higher doses of local anesthetics used in spinal anesthesia in cesarean section. The study can be implicated to select the best possible combination of local anesthetics used in spinal anesthesia in cesarean section.
Methods
The study was cleared from the Ethics Committee of the University and written consent was taken from patients who participated in this study. All the patients taken for this study belonged to American Society of Anesthesiology (ASA) grade 1 or 2. None of the patients had any contradiction for spinal anesthesia. Complicated pregnancies such as multiple pregnancies, pregnancy induced hypertension and placenta previa were excluded. Also the antenatal patients with acute fetal distress were excluded, keeping the respiratory depressant effect of fentanyl in mind.
Prospective single-blind study was performed on 120 parturients. In the first group of 60 parturients we tried to find the optimal dose of intrathecal bupivacaine, which was not associated with visceral pain. In the second group of 60 parturients intrathecal fentanyl was added to varying dose of bupivacaine. The second group was made with the idea to find out the lowest dose of bupivacaine-fentanyl combination that was not associated with visceral pain. The first group of 60 parturients was further subdivided into 3 subgroups receiving 8,10 or 12.5 mg of 0.5% hyperbaric intrathetal bupivacaine respectively. The second group also of 60 parturients was again subdivided into 3 subgroups receiving 8,10 or 12.5 mg of 0.5% hyperbaric intrathetal bupivacaine mixed with 12.5 μg of intrathecal fentanyl. All parturients were given rapid fluid infusion of 1 to 1.5 litres ringer lactate via 18 gauge venous catheter. Spinal anesthesia was given in lateral position. For lumbar puncture 25 gauge needle was used. Immediately after the block each parturient was placed with 10 cm wedge under right hip. Pulse and noninvasive blood pressure were measured every 5 minutes for the first 30 minutes and thereafter every 10 minutes. If the systolic blood pressure fell below 90 mm HG, additional vasopressor support was given.
Sensory block was tested by pinprick at the left midclavicular line till the block reached T6 when the surgical incision will be allowed. Degree of motor block was assessed by using Bromage scale (BS).
Intraoperative pain was checked whenever the parturients complained of discomfort or pain during operation and expressed as visual analogue scale (VAS 0–1000 mm) by the parturient. Each time VAS exceeded 30–50 μg fentanyl was given IV.
Muscle relaxation was assessed clinically and rated as poor, fair, good or excellent and score of 1,2,3 or 4 was given for each description respectively.
Incidence of nausea, vomiting, itching, shivering pruritus and sedation during operation, the time required for sensory recovery to T10, motor recovery to B0 and the onset of postoperative pain was recorded. All these time variables were measured from the beginning of the spinal injection. Apgar score of all the babies at 1, 5 and 10 minutes and maternal depression were also recorded.
Results
The study was performed on 120 parturients undergoing caesarian section under spinal anesthesia. Six group of twenty each were made as follows according to the intrathecal medication.
Group I-B8 (8 mg 0.5% hyperbaric bupivacaine)
Group II-B10 (10 mg 0.5% hyperbaric bupivacaine)
Group III-B 12.5 (12.5 mg 0.5% hyperbaric bupivacaine)
Group IV-FB8 (8 mg 0.5% hyperbaric bupivacaine+12.5 μg fentanyl)
Group V-FB10 (10 mg 0.5% hyperbaric bupivacaine +12.5 μg fentanyl)
Group VI-FB 12.5 (12.5 mg 0.5% hyperbaric bupivacaine +12.5 μg fentanyl)
All six groups were almost of similar age (30 ± 4 years), height (158 ± 6 cm) and weight (54 ± 11 kg).
Onset of sensory block to T6 occurs faster with increasing bupivacaine doses in the alone or < combination groups with fentanyl. Muscle relaxation (degree of muscle relaxation i.e. 4) was excellent in all the patients of all the groups. About 90–100% parturients had complete motor block, with no significant change in different groups. No visceral pain was noticed in any of the combination treated group as well as Group III receiving 12.5 mg of bupivacaine. Incidence of visceral pain was significantly higher in B8 and B10.
Maximum fall in the systolic blood pressure was noticed after 25 minutes in all the groups. Depending upon the % of systolic blood pressure fall the following series can be drawn: FB 12.5>B 12.5>FB10>B10>FB8>B8. On comparing the hemodynamic stability of equipotent dose of bupivacaine and bupivacaine-fentanyl, we found that the later is more stable. We found the intraoperative hypotension increases with increasing doses of bupivacaine however, alongwith fentanyl it increased more. Bradycardia was found in 10–15% cases in each group. There exists no statistical significance in the incidence of Bradycardia in different groups.
There was no intraoperative sedation in bupivacaine whereas 75–90% of parturient of bupivacaine-fentanyl combination groups were drowsy but arousable. Incidence of vomiting was more in bupivacaine alone group than in combination. Duration of post-operative analgesia increased with increasing dose of bupivacaine. Also addition of fentanyl to bupivacaine increased the postoperative analgesic effect. Motor recovery takes longer with increasing doses of bupivacaine alone or in combination with fentanyl. Data of the above-mentioned clinical parameters (if they are significantly different from each other) are given in Fig. 1 and 2.
Figure 1 Alterations in different clinical profiles of six groups
Figure 2 Alteration in different clinical syptoms in 6 groups
Apgar score (9.85 ± 0.37 at 10 minutes) of babies was unaffected when additive of 12.5 μg of intrathecal fentanyl was used in cesarean section.
Discussion
Recent trends of obstetric anesthesia show increased popularity of regional anesthesia amongst obstetric anesthetists. General anesthesia is associated with higher mortality rate in comparison to regional anesthesia. However, regional anesthesia is not without risk. Deaths in regional anesthesia are primarily related to excessive high regional blocks and toxicity of local anesthetics. Reduction in doses and improvement in technique to avoid higher block levels and heightened awareness to the toxicity of local anesthetics have contributed to the reduction of complications related with regional anesthesia [7]. Spinal anesthesia among the neuraxial blocks in obstetric patients needs more strict dose calculations as the drugs are directly injected in intrathecal space. With minimum dose changes, the chances of complications and side effects are enhanced [1]. Above-mentioned factors provided us the impetus to perform this study.
These days 0.5% heavy bupivacaine is used commonly for spinal and epidural anesthesia. It was decided to combine it with intrathecal fentanyl to provide adequate depth of anesthesia with lesser doses of bupivacaine [8]. Fentanyl is a lipophilic opiod and is preferred for having a rapid onset and short duration of action with lesser incidence of respiratory depressions. Our results of onset time of sensory block to T6 corroborrate with that of Randalls et. al (1991) [9] which state that the onset of sensory block to T4 gets faster with increasing bupivacaine doses, whereas, it differs from the observations of Singh et. al (19 95) [10].
Complete motor block was achieved in 90–100% of patients with our study; this is in accordance with the results of Pederson et.al (1989) [3] and Choi et.al (2000) [2]. Visceral pain is a common problem in cesarean section under spinal anesthesia. In our study we found no pain in B 12.5 group however, visceral pain was not fully abolished with lower doses of bupivacaine. Our results corroborates fully with Choi et.al (2000) [2].
It is evident from the results that the depth of anesthesia in FB8 group is equivalent to B 12.5 group. This proves that by adding fentanyl adequate depth of spinal anesthesia can be achieved at much lower doses of bupivacaine. Incidence of hypotension as well as fall in the systolic BP increases with the dose of bupivacaine. However, no significant difference was noticed in the fentanyl added group when compared with their nonadded counterpart.
Bradycardia results from the blockade of sympathetic cardio accelerator fibers and decreased venous return to the heart. In our study bradycardia occurrence was overall 7% with no significant intergroup variation. This is in accordance with Singh et.al (1991) [10]. About 75–90% of the patients become drowsy but arousable with the intrathecal fentanyl addition as compared to those without fentanyl addition. However, there is report where fentanyl addition do not cause any change in the sedation [10], however, similar to the findings of Randalls et. al (1991) [10] we found significant reduction in the incidence of nausea by the addition of fentanyl to bupivacaine. Further, negligible incidences of pruritus, shivering or respiratory depression were observed. Also, the apgar score of the babies remained same in all the groups. There was longer duration of postoperative analgesia in fentanyl-bupivacaine groups; this also increases with the increasing dose of bupivacaine. However, motor recovery was not affected by the addition of fentanyl.
Conclusion
The above study can be concluded as bupivacaine -fentanyl combination leads to abolishment of the visceral pain, reduction in nausea incidence increased hemodynamic stability and increases the duration of post-operative analgesia; however no effect can be seen on bradycardia, nausea, vomiting, shivering, maternal or neonatal respiration. Thus, overall the combined effect of fentanyl-bupivacaine is superior over just bupivacaine alone as fentanyl apart from positive effects, retards the negativity as well as reduces the doses of bupivacine too.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JB is guide and with NA helped in execution of work and PS helped in experiment design and manuscript preparation. All authors read and approved the final manuscript. The authors do not have any financial or non-finanacial competing interests.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors are thankful to the Prof. B.K. Singh, Head, Deptt of Anesthesiology and the patients of the Kings' George Medical College University, Lucknow, India.
==== Refs
Caplan RA Ward RJ Posner K Cheney FW Unexpected cardiac arrest during spinal anesthesia: a closed claims analysis of predisposing factors Anesthesio 1988 68 5 11
Choi DH Ahn HJ Kim MH Bupivacaine sparing effect of fentanyl in spinal anesthesia for cesarean delivery Regional Anesthesia Pain medicine 2000 25 240 245 10.1016/S1098-7339(00)90005-1
Pedersen H Santos AC Steinberg ES Schapiro HM Harmon TW Finster M Incidence of visceral pain during cesarean section: The effect of varying doses of spinal bupivacaine Anesthesia & Analgesia 1989 69 46 49 2742167
De Simone CA Leighton BL Norris MC Spinal anesthesia for cesarean delivery. A comparison of two doses of hyperbaric bupivacaine Regional Anesthesia 1995 20 90 94 7605770
Abouleish E Rawal N Fallon K Harnandez D Combined intrathecal morphine and bupivacaine for cesarean section Anesthesia & Analgesia 1988 67 370 374 3354872
Benhamou D Thorin D Brichant JF Dailland P Milon D Schneider M Intrathecal clonidine and fentanyl with hyperbaric bupivacaine improves analgesia during cesarean section Anesthesia & Analgesia 1998 87 609 613 9728839 10.1097/00000539-199809000-00022
Courtney MA Bader AM Hartwell B Hauch M Grennan MJ Datta S Perioperative analgesia with subarachnoid sufentanil administration Regional Anesthesia 1992 17 274 8 1419940
Kang FC Tsai YC Chang PJ Chen TY Subarachnoid fentanyl with dilute small dose bupivacaine for cesarean section Acta Anaesthesiol Sin 1998 36 207 214 10399516
Randalls B Broadway JW Browne DA Morgan BM Comparison of four subarachnoid solutions in a needle-through needle technique for elective cesarean section Br J Anaesth 1991 66 314 318 2015147
Singh H Yang J Thornton K Giesecke AH Intrathecal fentanyl prolongs sensory bupivacaine spinal block Can J Anesth 1995 42 987 91 8590509
| 15904498 | PMC1159169 | CC BY | 2021-01-04 16:28:05 | no | BMC Anesthesiol. 2005 May 17; 5:5 | utf-8 | BMC Anesthesiol | 2,005 | 10.1186/1471-2253-5-5 | oa_comm |
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-451594638810.1186/1472-6963-5-45Research ArticleThe factors associated to psychosocial stress among general practitioners in Lithuania. Cross-sectional study Vanagas Giedrius [email protected] Susanna [email protected] Preventive Medicine dept., Kaunas University of Medicine, Kaunas, Eiveniu 4, Lithuania2 Nordic School of Public Health, Gothenburg, Nya Varvet 25, Sweden2005 10 6 2005 5 45 45 14 1 2005 10 6 2005 Copyright © 2005 Vanagas and Bihari-Axelsson; licensee BioMed Central Ltd.2005Vanagas and Bihari-Axelsson; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
There are number of studies showing that general practice is one of the most stressful workplace among health care workers. Since Baltic States regained independence in 1990, the reform of the health care system took place in which new role and more responsibilities were allocated to general practitioners' in Lithuania. This study aimed to explore the psychosocial stress level among Lithuanian general practitioner's and examine the relationship between psychosocial stress and work characteristics.
Methods
The cross-sectional study of 300 Lithuanian General practitioners. Psychosocial stress was investigated with a questionnaire based on the Reeder scale. Job demands were investigated with the R. Karasek scale. The analysis included descriptive statistics; interrelationship analysis between characteristics and multivariate logistic regression to estimate odds ratios for each of the independent variables in the model.
Results
Response rate 66% (N = 197). Our study highlighted highest prevalence of psychosocial stress among widowed, single and female general practitioners. Lowest prevalence of psychosocial stress was among males and older age general practitioners. Psychosocial stress occurs when job demands are high and job decision latitude is low (χ2 = 18,9; p < 0,01). The multivariate analysis shows that high job demands (OR 4,128; CI 2,102–8,104; p < 0,001), patient load more than 18 patients per day (OR 5,863; CI 1,549–22,188; p < 0,01) and young age of GP's (OR 6,874; CI 1,292–36,582; p < 0,05) can be assigned as significant predictors for psychosocial stress.
Conclusion
One half of respondents suffering from work related psychosocial stress. High psychological workload demands combined with low decision latitude has the greatest impact to stress caseness among GP's. High job demands, high patient load and young age of GP's can be assigned as significant predictors of psychosocial stress among GP's.
==== Body
Background
Lithuania is one of the three Baltic States which regained independence in 1990. Back in 1989 a Congress of Physicians of Lithuania took place in which the necessity to reform the health care system was discussed. To implement these reforms a National Health Care Conception was adopted in 1991 by the Parliament. The main goal of the reform is to optimise the health care resources and services for better health of the population. The development and reformation of Primary Health Care was underlined as a key factor of a whole Health Care Reform. The main concept argues development of the primary health care services reorienting them from disease centered episodic activities to patient needs, continuity, comprehensiveness, health promotion and disease prevention.
Primary health care services in Lithuania are delivered in primary health care centers, GP's, both school and community medical posts (paramedical centers), ambulatories and polyclinics, women's consultancies, nursing hospitals, as well as by the ambulance service (stations and divisions). A health reform goal was that all practicing physicians in primary health care (PHC) level (district internists and pediatricians) should be replaced by family doctors (GPs) by 2005.
More than 300 public and private GP clinics are in operation. At the moment, the vast majority of health care facilities are publicly owned, but there are plans to partially privatize primary care. For the most part private primary care takes the form of solo or small group physician-owned practices [1]. Currently, more than 1000 GPs practicing in primary health care. Many GP's have more than 2000 patients in the list when the reform statement it is suggested that an appropriate average should be of 1600 patients per one GP. Not all of GP graduates are practicing family medicine; there is still lack of GP's in rural areas. A financing principle of the health care system is based on compulsory health insurance and does not support to cover practice needs. PHC services are covered by capitation fees only, which average is about €20 per capita per year.
In 1996–1997 operational service standards for GP's were defined. To GP's services were defined new task to deliver pediatry, gynaecology and many other services as primary health care. Since 1998, an existing partial gate-keeping role for family doctors was switched in 2002 to complete gate keeping. This new role tasks increased workload and responsibilities for Lithuanian GP's. The primary role of the Lithuanian GP today continues to be in diagnosis and ongoing management of medical conditions, with consultations accounting for about 50% of their workload. GP's are overloaded by patient's list and paperwork, consuming other 50% of their working time.
Number of studies has been argued that general practice has become an increasingly stressful place to work [2-11] by the increasing demands and constraints [6,9,12-17]. It is showed that about half of the investigated general practitioners (GP's) were not satisfied with their work [18-22] due to high job requirements. In recent literature important sources of psychosocial stress for GP's are mentioned: excessive paperwork, health reforms, bureaucratic interference (6), job demands, decision latitude [9], workplace location [23] job pressure, patient load [6,18,24,25], lack of organizational support [26-30], dealing with difficult patients [31,32] and objective personal characteristics such as age, gender and workers marital status [33-36].
Independence and flexibility as provided by the continuing Lithuanian health care reform with regard to the primary care services as a small business are now being undermined by high workload requirements to general practitioners [9,37] due to new tasks, excessive paperwork and high patient load, which can lead to intension to quit general practice, lower intensions to attend higher professional standards, and can take turnover to quality of care and to patient's satisfaction with services [5,6,32,38-40].
Aim of this paper is to explore the psychosocial stress level among Lithuanian general practitioners and to examine the relationship between psychosocial stress and work characteristics. Main investigated research questions were: What is prevalence of psychosocial stress regarding the sociodemographic characteristics of GP's? How do work demands and decision latitude influence the presence of psychosocial stress and low quality of life among GP's? What characteristics can be predictors of psychosocial stress among GP's?
Methods
Target group
Lithuanian general practitioners.
Study design
Cross – sectional study. A mailed survey of random national samples. Computerized randomization was performed from the registry of Lithuanian physicians. The data collected through the questionnaires filled in by the GP's.
Questionnaire
The study involved the development and administration of a questionnaire for GPs. The questionnaire was designed using the Karasek and Reeder. A questionnaire was pilot tested among a group of ten GPs.
Sample size
Sample size was calculated using EpiInfo 2000 Statcalc software whish argued the sample size of 192 GP's with the 95% confidence level. From the previous studies the expected response rate was 63%. Therefore, it was decided to send questionnaires to 300 Lithuanian GP's. Our observed response rate was 66%. We collected 197 filled in questionnaires.
Assessmen of psychosocial stress
Psychosocial stress in this study was investigated by a questionnaire based on the Reeder scale [41,42]. The Reeder scale uses four statements experienced in everyday stressful situations as "usually tense or nervous", "daily activities are extremely trying and stressful". The respondents should indicate whether each of the statements describe them. Each question has four alternative responses, which were coded using Likert-like scale. For scoring we used the simple summation method [43].
Assessment of work characteristics
Work characteristics were investigated with the Karasek scale [44,45]. This model, also known as the "job strain" model. (fig. 1)
Figure 1 Prevalence of psychosocial stress among GP's by sociodemographic characteristics.
The Lithuanian version of Karasek's scale of 11 questions was adopted by prof. A. Gostautas in 1992. This scale measures job character – decision latitude and psychological workload demands.
The first scale – decisions latitude scale is composed of two subscales: skill discretion and decision making authority available to the worker (Table 1).
Table 1 Basic components of R. Karasek JSQ model
Component Demand
Decision latitude
Skills discretion Job requires learning new things
Job requires high level of skills
Job requires creativity
Job entails a variety of things to do
Decision authority Job allows making one's own decision
Job provides a lot of freedom as to how the work gets done
Job demands Job requires very hard work
Job requires very fast work
Job requires excessive work
Job involves conflicting demands
Jon involve not having enough time to get the job done
Skill discretion, measured by six items such as "keep learning new things', "can develop skills", "job requires skills", "task variety", "repetitious", and "job requires creativity", and decision authority, measured by three items such as "have freedom to make decisions", "choose how to perform work", and "have a lot of say on the job".
The second scale is psychological job demands, defined by five items such as "excessive work", "conflicting demands", "insufficient time to work", "work fast", and "work hard".
A four point Likert – like scale was used with the coding from 4 to 1 for series, so that the responses were summarized to give a score [46].
Data were also collected on supplementary aspects of stress and work characteristics, including: practice characteristics (partnership size, workplace location, patient load); and personal characteristics (gender, age, marital status).
Statistical analysis
The data were computed – coded and analyzed using Statistical Package for the Social Sciences for Windows version 11.0 (SPSS Inc). The analysis included descriptive statistics, interrelation analysis and multivariate logistic regression as useful tool to predict the presence or absence of a characteristic or outcome based on values of a set of predictor variables. Logistic regression coefficients were used to estimate the odds ratios for each of the independent variables in the model.
Nonparametric tests were used to test for significant differences at the p = 0.05 level.
Results
Descriptive statistics
Of the 197 respondents, 162 (82.2%) GP's were female, and 35 (17.8%) male. The GP ages ranged from 31 to 66 years (mean 44.2 years, 95% CI 42.9 – 45.4). This reflects to the whole GP population in Lithuania. Significant gender difference was found for mean age (males 47.1 years, 95% CI 43.5 – 50.7; females 43.5 years, 95% CI 42.2 – 44.9; p < 0.03). All descriptive measures are shown in Table 2.
Table 2 Characteristics of general practitioners who responded to questionnaire survey of stress in general practice (N = 197).
Variables Number %
Gender
Male 35 17.8
Female 162 82.2
Age (years)
less 45 90 45.7
45–54 85 43.1
more 54 22 11.2
Years worked as GP
less 8 40 20.3
8–28 115 58.4
more 28 42 21.3
Practice ownership type
Solo practice 56 28.4
Group practice 141 71.6
Workplace
City 123 62.4
Rural 74 37.6
Patient load (patient/day)
less 18 21 10.7
18–28 140 71.1
more 28 36 18.3
Prevalence of psychosocial stress among GP's by sociodemographic characteristics
Fourthy-eight percents of respondents could be classified as suffering from work related psychosocial stress by the Reeder scale. In figure 1 we see that there are considerable variations in psychosocial stress measures regarding to sociodemographic characteristics of GP's. The highest percentage of psychosocial stress by sociodemographic characteristics was found among widowed, single and female GP's.
Work demands and decision latitude influence presence of psychosocial stress among GP's
The job strain model suggests that high job demands, and low job control, are indicators of psychosocial stress. In figure 2 showed statistically significant interrelationship between job demands, decision latitude and psychosocial stress. Our results confirmed Job strain model's hypothesis that psychosocial stress occurs when job demands are high and job decision latitude is low.
Figure 2 Interrelationship between job demands, decision latitude and psychosocial stress (χ2 = 18,9; p < 0,01).
Predictors of psychosocial stress among GP's
The multivariate analysis (Table 3) shows that gender, work place location, practice ownership type, low ability to use skills and low decision latitude did not exhibit a statistically significant effect on psychosocial stress caseness and did not have a significant effect even when no other variables were controlled for. The model highlighted high job demands, patient load more than 18 patients per day and young age of GP's as significant predictors for psychosocial stress.
Table 3 Multivariate logistic regression model to predict psychosocial stress among Lithuanian GP's (n = 197)
Variables B P-value OR 95,0% CI for OR
Constant -4.782 <0,001
Female gender (male – reference) 0,465 0,337 1,593 0,616–4,117
Rural workplace (city – reference) 0,261 0,478 1,298 0,632–2,665
Solo practice (Group practice – reference) 0,342 0,373 1,408 0,664–2,987
Age less 45(reference to age group 45–54) 1,928 0,024 6,874 1,292–36,582
Age more 54(reference to age group 45–54) 0,010 0,980 1,010 0,459–2,226
Practice duration 8–28 (more 28 – reference) 1,770 0,015 5,873 1,407–24,523
Practice duration less 8 (more 28 – reference) 1,627 0,054 5,089 0,975–26,552
Patient load 18–28 p/d (less 18 p/d -reference) 1,769 0,009 5,863 1,549–22,188
Patient load > 28 p/d (less 18 p/d -reference) 1,845 0,014 6,330 1,450–27,630
Low ability to use skills 0,198 0,609 1,219 0,571–2,600
Low decision latitude 0,317 0,343 1,373 0,713–2,644
High job demands 1,418 <0,001 4,128 2,102–8,104
OR – Odds Ratio
Discussion
GPs are the professionals who are at the forefront of helping patients to manage urgent health problems, and as gatekeepers they have to make decisions on patient's health; whether to send them to hospitals. As a consequence of health care reform, GP's are required to have more competence in diagnosis and ongoing management of medical conditions. This means increased responsibilities, which may contribute to higher psychosocial stress for Lithuanian GP's. This can be regarded as stressors related to the new demands and sometimes it can interfere with personal life that can cause negative feelings about work, frustration, tension and lack of time to make appropriate decisions [47].
Our study has highlighted that fourthy-eight percents of respondents could be classified as suffering from work related psychosocial stress by the Reeder scale. Highest psychosocial stress prevalence among widowed, single and female GP's. Lowest psychosocial stress prevalence among was among males and older age GP's. The greatest risk to physical and mental health from stress occurs to GP's facing high psychological workload demands combined with low decision latitude in meeting those demands. High job demands, patient load more than 18 patients per day and young age of GP's can predict a statistically significant effect on psychosocial stress caseness. Gender, work place location, practice ownership type, low ability to use skills and low decision latitude does not exhibit for GP's in Lithuania a statistically significant effect on psychosocial stress development.
Main weaknesses of the present study can be mentioned: cross-sectional nature of the study, self reporting scales and generalisability of the questionnaire. Cross-sectional nature precludes an evaluation of temporal precedence and causality of the observed associations. Job Strain Model guided our hypotheses about causal relationships psychosocial stress and other work characteristics. The explored causal relations should be interpreted carefully and longitudinal studies should be carried out in the future research.
Second limitation – our reliance on self-reported rating scales can raise the issue of systematic positive or negative response tendencies. Several authors have argued that this phenomenon is not a major threat if interactions has been found [27,48]. Furthermore, as no scale is perfectly reliable, the associations between self-reported measures and self-reported workload appear to be weaker than they could be in reality.
Karasek's JCQ questionnaire was designed to be broadly applicable to a wide range of occupations, this means that factors that are specific to particular occupations may be overlooked. For example, job demands as it has been conceptualized and operationalised in this survey would not take into account some emotional demands that could be source of stress to GP's such as dealing with difficult patients or caring for the dying patients [30,31].
Otherwise, on the positive side, it is important to mention that generalisability of Karasek's model allows us to make comparisons among different medical and non-medical occupational groups and this has been an important factor in selecting the job strain model. Our results were obtained among a sample of people working in general practice. As strength of the investigation can be seen similar education level that respondents had. The sample size was sufficient regarding to sample size calculation and to allow exploration of tendencies. The participation rate was acceptable, and the scales used were previously validated instruments that retained their psychometric properties in our population. Findings from this research have hopefully emphasized the importance of examining changes and associations between work characteristics and psychosocial stress among GP's before health care reform in Lithuania will be definitely implemented.
Conclusion
One half of respondents suffering from work related psychosocial stress. High psychological workload demands combined with low decision latitude has the greatest impact to stress caseness among GP's. High job demands, high patient load and young age of GP's can be assigned as significant predictors of psychosocial stress among GP's.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
GV designed the study, abstracted data, made data analysis, drafted and revised the manuscript.
SBA participated in initial study design, participated in data analysis and revised the manuscript.
All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
Thanks for Collegium of General Practitioners of Lithuania for expressed kind interest in this study.
==== Refs
European observatory on health care systems. Health care systems in transition Lithuania 2000 Copenhagen, World health organization
Dowell AC Westcott T McLeod DK Hamilton S A survey of job satisfaction, sources of stress and psychological symptoms among New Zealand health professionals N Z Med J 2001 114 540 543 11833946
Firth-Cozens J Individual and organizational predictors of depression in general practitioners Br J Gen Pract 1998 48 1647 1651 10071396
Iversen L Farmer JC Hannaford PC Workload pressures in rural general practice: a qualitative investigation Scand J Prim Health Care 2002 20 139 144 12389749 10.1080/028134302760234573
Sibbald B Bojke C Gravelle H National survey of job satisfaction and retirement intentions among general practitioners in England BMJ 2003 326 22 12511457 10.1136/bmj.326.7379.22
Simoens S Scott A Sibbald B Job satisfaction, work-related stress and intentions to quit of Scottish GPS Scott Med J 2002 47 80 86 12235914
Sundquist J Johansson SE High demand, low control, and impaired general health: working conditions in a sample of Swedish general practitioners Scand J Public Health 2000 28 123 131 10954139
Vanagas G Bihari-Axelsson S Vanagiene V Do age, gender and marital status influence job strain development for general practitioner? Medicina (Kaunas ) 2004 40 1014 1018 15516827
Vanagas G Bihari-Axelsson S Interaction among general practitioners age and patient load in the prediction of job strain, decision latitude and perception of job demands. A Cross-sectional study BMC Public Health 2004 4 59 15581430 10.1186/1471-2458-4-59
Wilhelmsson S Foldevi M Akerlind I Faresjo T Unfavourable working conditions for female GPs. A comparison between Swedish general practitioners and district nurses Scand J Prim Health Care 2002 20 74 78 12184716 10.1080/02813430215560
Williams ES Konrad TR Linzer M McMurray J Pathman DE Gerrity M Schwartz MD Scheckler WE Douglas J Physician, practice, and patient characteristics related to primary care physician physical and mental health: results from the Physician Worklife Study Health Serv Res 2002 37 121 143 11949917
Norman P Fitter M Wall T General practitioners' subjective experience of surgery workload Soc Sci Med 1991 33 161 166 1887279 10.1016/0277-9536(91)90176-D
Peters B Primary care: new demands for a new era Mich Health Hosp 1995 31 18 20 10145053
Schieman S Van Gundy K Taylor J Status, role, and resource explanations for age patterns in psychological distress J Health Soc Behav 2001 42 80 96 11357720
Shanley BC Schulte KM Chant D Jasper A Wellard R Factors influencing career development of Australian general practitioners Aust Fam Physician 2002 31 49 54 11840889
Sturmberg JP Renumeration systems and GPs: can we adapt to future demands? Aust Fam Physician 1999 28 650 10484763
Wilhemsson S Faresjo T Foldevi M Akerlind I The personal doctor reform in Sweden: perceived changes in working conditions Fam Pract 1998 15 192 197 9694174 10.1093/fampra/15.3.192
Schattner PL Coman GJ The stress of metropolitan general practice Med J Aust 1998 169 133 137 9734508
Chambers R Wall D Campbell I Stresses, coping mechanisms and job satisfaction in general practitioner registrars Br J Gen Pract 1996 46 343 348 8983252
Cooper CL Rout U Faragher B Mental health, job satisfaction, and job stress among general practitioners BMJ 1989 298 366 370 2493939
Gosden T Williams J Petchey R Leese B Sibbald B Salaried contracts in UK general practice: a study of job satisfaction and stress J Health Serv Res Policy 2002 7 26 33 11822258 10.1258/1355819021927647
Jacobs S Addressing the problems associated with general practitioners' workload in nursing and residential homes: findings from a qualitative study Br J Gen Pract 2003 53 113 119 12817356
Sexton R Maintaining the wellbeing of rural GPs BMJ 2003 326 S101 12663427 10.1136/bmj.326.7391.S101
Charles-Jones H Houlker M Out-of-hours work: the effect of setting up a general practitioner cooperative on GPs and their families Br J Gen Pract 1999 49 215 216 10343426
Scott A Wordsworth S The effects of shifts in the balance of care on general practice workload Fam Pract 1999 16 12 17 10321389 10.1093/fampra/16.1.12
Branthwaite A Ross A Satisfaction and job stress in general practice Fam Pract 1988 5 83 93 3391360
de Jonge J Mulder MJ Nijhuis FJ The incorporation of different demand concepts in the job demand-control model: effects on health care professionals Soc Sci Med 1999 48 1149 1160 10220016 10.1016/S0277-9536(98)00429-8
Young G Spencer J General practitioners' views about the need for a stress support service Fam Pract 1996 13 517 521 9023527
Rout U Stress among general practitioners and their spouses: a qualitative study Br J Gen Pract 1996 46 157 160 8731621
Sutherland VJ Cooper CL Identifying distress among general practitioners: predictors of psychological ill-health and job dissatisfaction Soc Sci Med 1993 37 575 581 8211271 10.1016/0277-9536(93)90096-M
Calnan M Wainwright D Forsythe M Wall B Almond S Mental health and stress in the workplace: the case of general practice in the UK Soc Sci Med 2001 52 499 507 11206648 10.1016/S0277-9536(00)00155-6
McGlone SJ Chenoweth IG Job demands and control as predictors of occupational satisfaction in general practice Med J Aust 2001 175 88 91 11556426
Barnes-Farrell JL Rumery SM Swody CA How do concepts of age relate to work and off-the-job stresses and strains? A field study of health care workers in five nations Exp Aging Res 2002 28 87 98 11928214 10.1080/036107302753365577
Chambers R Campbell II Anxiety and depression in general practitioners: associations with type of parctice, fundholding, gender and other personal characteristics Fam Pract 1996 13 170 173
Estryn-Behar M Kaminski M Peigne E Bonnet N Vaichere E Gozlan C Azoulay S Giorgi M Stress at work and mental health status among female hospital workers Br J Ind Med 1990 47 20 28 2310704
Preston DB Marital status, gender roles, stress, and health in the elderly. Health Care Women Int 1995 16 149 65 7759346
Vanagas G Bihari-Axelsson S Cross-sectional study on quality of life,work demands and psychosocial stressof Lithuanian general practitioners Eur J Gen Pract 2004 10 179 80
Evans J Lambert T Goldacre M GP recruitment and retention: a qualitative analysis of doctors' comments about training for and working in general practice Occas Pap R Coll Gen Pract 2002 iii 33 12049026
Luce A van Zwanenberg T Firth-Cozens J Tinwell C What might encourage later retirement among general practitioners? J Manag Med 2002 16 303 310 12463646 10.1108/02689230210445112
Pitts J Vincent S General practitioners' reasons for not attending a higher professional education course Br J Gen Pract 1994 44 271 273 8037983
Metcalfe C Smith GD Wadsworth E Sterne JA Heslop P Macleod J Smith A A contemporary validation of the Reeder Stress Inventory Br J Health Psychol 2003 8 83 94 12643818 10.1348/135910703762879228
Schar M Reeder LG Dirken JM Stress and cardiovascular health: An international cooperative study- II. The male population of a factory at Zurich. Soc Sci Med 1973 7 585 603 4731469 10.1016/0037-7856(73)90027-9
Heslop P Smith GD Carroll D Macleod J Hyland F Hart C Perceived stress and coronary heart disease risk factors: the contribution of socio-economic position Br J Health Psychol 2001 6 167 178 14596732 10.1348/135910701169133
Karasek R Brisson C Kawakami N Houtman I Bongers P Amick B The Job Content Questionnaire (JCQ): an instrument for internationally comparative assessments of psychosocial job characteristics J Occup Health Psychol 1998 3 322 355 9805280 10.1037//1076-8998.3.4.322
Levi L Bartley M Marmot M Karasek R Theorell T Siegrist J Peter R Belkic K Savic C Schnall P Landsbergis P Stressors at the workplace: theoretical models Occup Med 2000 15 69 106 10620787
Bosma H Marmot MG Hemingway H Nicholson AC Brunner E Stansfeld SA Low job control and risk of coronary heart disease in whitehall ii (prospective cohort) study BMJ 1997 314 558 9055714
Grol R Mokkink H Smits A Van Eijk J Beek M Mesker P Mesker-Niesten J Work satisfaction of general practitioners and the quality of patient care Fam Pract 1985 2 128 135 4043602
Dollard MF Winefield AH A test of the demand-control/support model of work stress in correctional officers J Occup Health Psychol 1998 3 243 264 9684215 10.1037//1076-8998.3.3.243
| 15946388 | PMC1159170 | CC BY | 2021-01-04 16:31:49 | no | BMC Health Serv Res. 2005 Jun 10; 5:45 | utf-8 | BMC Health Serv Res | 2,005 | 10.1186/1472-6963-5-45 | oa_comm |
==== Front
Nutr Metab (Lond)Nutrition & Metabolism1743-7075BioMed Central London 1743-7075-2-121590721310.1186/1743-7075-2-12ResearchProlonged exercise testing in two children with a mild Multiple Acyl-CoA-Dehydrogenase deficiency Takken T [email protected] J WH [email protected] G [email protected] L [email protected] PJM [email protected] Koning TJ [email protected] Department of Pediatric Physical Therapy & Exercise Physiology, University Hospital for Children and Youth 'Het Wilhelmina Kinderziekenhuis', University Medical Centre Utrecht, Utrecht, The Netherlands2 Department of Metabolic Diseases, University Hospital for Children and Youth 'Het Wilhelmina Kinderziekenhuis', University Medical Centre Utrecht, Utrecht, The Netherlands2005 20 5 2005 2 12 12 23 2 2005 20 5 2005 Copyright © 2005 Takken et al; licensee BioMed Central Ltd.2005Takken et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Multiple Acyl-CoA-Dehydrogenase deficiency (MADD) is an inherited metabolic disorder characterized by impaired oxidation of fatty acids and some amino acids.
Methods
We were interested whether children with MADD could tolerate a prolonged low-intensity exercise test and if this test could have any additional diagnostic value. Therefore, we performed a maximal exercise test and a low-intensity prolonged exercise test in 2 patients with MADD and in 5 control subjects. During a prolonged exercise test the subjects exercised on a cycle ergometer at a constant workload of 30% of their maximum for 90 minutes and heart rate, oxygen uptake, fuel utilization and changes in relevant blood and urinary parameters were monitored.
Results
The tests were tolerated well. During the prolonged exercise test the fatty acid oxidation (FAO) was quite low compared to 5 control subjects, while characteristic metabolites of MADD appeared in plasma and urine.
Conclusion
We suggest that the prolonged exercise test could be of diagnostic importance and might replace the fasting test as a diagnostic procedure in some cases, particularly in patients with anamnestic signs of intolerance for prolonged exercise.
==== Body
Introduction
Multiple Acyl-CoA-Dehydrogenase deficiency or Glutaric Acidemia Type II (MADD; McKusick # 231680) is an inborn error due to a deficiency of electron transfer flavoprotein (ETF) or of ETF-ubiquinone oxidoreductase, and is characterized by impaired oxidation of fatty acids (FAO) and some amino acids. Severely affected patients present in the newborn period with multiple congenital abnormalities in association with hypotonia, hepatomegaly, hypoglycemia and severe metabolic acidosis [1]. Patients with milder forms of the disorder present with a variety of symptoms such as hypoglycemia, Reye-like episodes, intermittent vomiting, acidosis and myopathy [1]. In adults, MADD can manifest itself as a lipid storage myopathy [1]. In patients with metabolic diseases, such as glycogen storage diseases [2-4], and FAO disorders such as carnitine palmitoyltransferase II deficiency and very long-chain acyl-coA-dehydrogenase deficiency [5,6] prolonged exercise tests are increasingly getting more attention. However, data in patients with FAO disorders are still scarce. This is the first report investigating fuel utilization and exercise capacity in patients with a mild form of MADD using indirect calorimetry and changes relevant in blood and urinary parameters during low-intensity prolonged exercise.
Methods
We diagnosed two brothers with a milder form of MADD. The younger patient presented himself after an intercurrent illness at the age of 3.1 yr with a coma, hypoglycemia and metabolites, indicative for MADD. Assessment of his family revealed that the older brother was affected as well and in both boys the diagnosis was confirmed in cultured skin fibroblasts (ETF assay 0.30 and 0.22 nmol·min-1·mg-1protein for the older and younger brother respectively, controls 1.08 ± 0.16, Dr Vianey-Saban, Lyon). Both patients had regular carnitine suppletion (50–100 mg·kg-1·day-1) since the diagnosis.
The patients reported to the exercise laboratory after a light breakfast, and performed a graded maximal exercise test on the bicycle ergometer to determine the exercise intensity for the prolonged exercise test. The workload was increased in constant increments of 10 watts every 1 minute. This protocol continued until the patient stopped due to volitional exhaustion, despite strong verbal encouragement. During the test the subjects breathed through a facemask (Hans Rudolph Inc., Kansas City, MO) connected to a calibrated respiratory gas analysis system (Oxycon Champion, Jaeger, Bilthoven, The Netherlands) which measured and or calculated breath-by-breath minute ventilation (VE), oxygen uptake (VO2), carbon dioxide production (VCO2), and respiratory exchange ratio (RER; = VCO2·VO2-1) using conventional equations. Heart rate (HR) was measured continuously during the maximal exercise test by a bipolar electrocardiogram. Peak oxygen consumption (VO2peak) was taken as the average value over the last 30 seconds during the maximal exercise test. Efforts were considered to be at a maximum level if HRpeak was > 180 per minute and RERpeak was > 1.0 [7].
Urine samples were taken before and after (3 hours) the exercise test and analyzed for organic acid analysis. Blood samples were taken before and after exercise and analyzed for glucose, lactate, free fatty acids (FFA), ketone bodies, creatine kinase (CK), organic acids, and acyl-carnitines. In the afternoon, after complete recovery, isometric muscle strength from the quadriceps muscle was measured using a hand-held myometer (Penny and Giles, Christchurch, UK), as described by Bäckman [8], to exclude that the results obtained during the maximal exercise test were influenced by muscle weakness. The results of myometry were compared to current reference values [8].
The prolonged exercise test was performed on another visit to prevent interference of the previous test. This test consisted of 90 minutes cycling at a constant workload at 30 % of Wpeak. During this test the following variables were monitored: VO2, VCO2, HR and RER, using the same setting as reported above. For each period of gas collection, whole-body rates of carbohydrate and fat oxidation (g·min-1) were calculated from the rates of CO2 production and O2 uptake using the non-protein RER values, according to the following equations [9]:
CHO oxidation = 4.585 VCO2 (l·min-1) - 3.226 VO2 (l·min-1);
and
FAO = 1.695 VO2 (l·min-1) - 1.701 VCO2 (l·min-1);
These equations are based on the assumption that VO2 and VCO2 accurately reflect tissue O2 consumption and CO2 production. The energy provided from CHO and fat oxidation was calculated from their energy potentials (3.87 and 9.75 kcal·g-1, respectively) [9].
Blood samples were taken at regular time intervals (t = 0, 30, 60, 75 and 90 minutes after the start of exercise). Glucose, lactate, CK, ketone bodies, FFA, organic acids, and acyl-carnitines were tested and measured. Urine samples were taken 3 hours after exercise and analyzed for organic acids, carnitine and myoglobin.
Organic acid concentrations in urine and plasma were determined by gas chromatography-mass spectrometry as their trimethylsilyl derivatives (Hewlett Packard 5890 series II gas chromatograph linked to a HP 5989B MS-Engine mass spectrometer (Hewlett Packard, Avondale, PA). The coefficients of variation for the various measured organic acids varied between 10–15%. Analysis of acylcarnitines in plasma as their butyl esters was performed by electrospray tandem mass spectrometry, (ESI-MS-MS; Micromass Quattro Ultima, Micromass Ltd., UK) equipped with an Alliance HPLC system (Waters, Milford, MA, USA). Also for these analyses the coefficients of variation for the determined acylcarnitines were 10–15%.
The results of the exercise tests in these two boys with MADD were compared to the results of the same test protocol used in 5 boys with complaints of fatigue and exercise intolerance, in whom inborn errors of metabolism were excluded. Their characteristics can be appreciated from Table 1.
Table 1 Subject characteristics, exercise capacity and muscle strength
Subject 1 Subject 2 Control subjects Mean (SD)
Age (years) 10.2 8.7 12.6 (2.6)
Weight (kg) 33 (SD-score 0) 25 (SD-score -1) 42.6 (6.5) (SD-score 0.01)
Height (cm) 143 (SD-score 0) 133 (SD-score 0) 154.6 (10.6) (SD-score -0.4)
Wpeak (Watt) 115 70 145 (70)
HRpeak (b·min-1) 195 195 185 (5)
RERpeak 1.27 1.25 1.22 (0.11)
VO2peak (L·min-1) 1.4 (SD-score -1.5) 1.2 (SD-score -1) 1.5 (0.46) (SD-score -1.6)
VO2peak (ml·kg-1·min-1) 42.8 (SD-score -1.65) 50.2 (SD-score 0.06) 35.3 (5.3) (SD-score -2.17)
Peak blood lactate 3 min after the exercise test (mmol·l-1) 7.3 6.0 6.5 (3.6)
Quadriceps muscle strength (N) 207 (SD-score 0.2) 156 (SD-score 0.8) NM
Abbreviations: Wpeak: peak workload, HRpeak: peak heart rate, RERpeak: peak Respiratory Exchange Ratio, VO2peak: peak oxygen uptake, SD-score: standard deviation score, NM: not measured.
Informed consent was obtained from the parents after being fully informed about the test procedures and the possible risks involved. All procedures were approved by the Medical Ethics Committee of the University Medical Center Utrecht.
Results
The results of the maximal exercise test and myometry are shown in Table 1. Both children had normal quadriceps muscle strength. Also, both children met the criteria for a maximal exercise test, i.e. a HR > 180·min-1, RER > 1.0 and increased blood lactate concentrations. Before and after the maximal exercise test none of the MADD specific metabolites were found in organic acid analysis and plasma acylcarnitines in subject 1 (16 mmol·mol-1 creatinine in subject 1, normal value < 20), with in subject 2 only a marginal elevated urinary excretion of ethylmalonic acid (30 mmol·mol-1 creatinine) in combination with normal plasma acylcarnitines.
Results of the prolonged exercise tests are shown in Figure 1 and Tables 2A and 2B. All patients did not reach the ventilatory anaerobic threshold during the 90-minutes exercise test, and had a VO2 of approximately 50% of VO2peak. All patients were very exhausted at the end of the prolonged exercise test. As can be seen in Figure 1 the FAO was very low in the two patients with MADD compared to the control subjects, although the exercise intensity was quite low (<50% VO2peak). Especially from 40 minutes onwards the difference in FAO was significant.
Figure 1 Percentage energy derived from fatty acid oxidation during the prolonged exercise test for patient 1 (filled circle), patient 2 (open circle) and 5 control subjects (black triangles). Values are mean ± SD.
Table 2A Results prolonged exercise test subject 1
T = 0 T = 30 T = 60 T = 75 T = 90 T = 120
HR (b·min-1) 70 140 145 145 150
VO2 (ml·min-1) 300 700 700 720 720
RER (VCO2·VO2-1) 0.8 0.95 0.95 0.9 0.9
Glucose (mmol·l-1) 7.1 5.6 4.9 4.8 5.5 4.8
Lactate (mmol·l-1) 1.1 1.2 1.0 1.3 1.1 1.0
CK (u·l-1) 103 98 105 119 117
3-keto-B (mmol·l-1) 0.07 <0.05 <0.05 0.05 0.1 0.15
3-OH-B (mmol·l-1) 0.09 <0.02 <0.02 0.05 0.12 0.26
FFA (umol·l-1) 0.76 0.28 0.49 0.88 1.03 0.93
Urine
Ethylmalonic acid in Mmol·mol-1 creatine T = -60–0: 16 T = 120–300: 74
Abbreviations: HR: heart rate, RER: Respiratory Exchange Ratio, VO2: oxygen uptake, CK: creatine kinase, 3-keto-B: 3-ketobutyric acid, 3-OH-B: 3-hydroxybutyric acid, FFA: Free Fatty Acids.
Table 2B Results prolonged exercise test subject 2.
T = 0 T = 30 T = 60 T = 75 T = 90 T = 120
HR (b·min-1) 80 120 120 120 120
VO2 (ml·min-1) 260 450 620 620 600
RER (VCO2·VO2-1) 0.9 0.9 0.95 0.95 0.9
Glucose (mmol·l-1) 5.2 4.6 4.3 4.4 4.5
Lactate (mmol·l-1) 2.0 0.9 0.9 1.0 1.2
CK (u·l-1) 58 80 78 78
3-keto-B (mmol·l-1) <0.05 <0.05 0.07 0.14
3-OH-B (mmol·l-1) <0.02 <0.02 0.04 0.24
FFA (umol·l-1) 0.15 0.31 0.62 0.91 1.34
Urine
Ethylmalonic acid in mmol·mol-1 creatinine T = -60–0: 29 T = 120–300: 59
Abbreviations: HR: heart rate, RER: Respiratory Exchange Ratio, VO2: oxygen uptake, CK: creatine kinase, 3-keto-B: 3-ketobutyric acid, 3-OH-B: 3-hydroxybutyric acid, FFA: Free Fatty Acids.
Glucose, lactate and myoglobin levels did not change significantly during the tests. In both children levels of CK raised but stayed within normal limits. Under normal (resting) conditions plasma of both patients did not show any abnormality. However, ethylmalonic acid excretion rose to 74 mmol·mol-1 creatinine in subject 1 and 59 mmol·mol-1 creatinine in subject 2 and acylcarnitines showed characteristic short and medium chain acylcarnitines (details given later).
The urine and plasma of subject 1 and 2, collected after the prolonged exercise test, showed high excretions of metabolites diagnostic of MADD. In detail our analyses showed the following: in addition to ethylmalonic acid discussed earlier abnormal urinary compounds were found consisting of octanoic acid, 2-methyl-3-hydroxy-butyric acid, isobutyrylglycine, isovalerylglycine, 2-methylbutyrylglycine and hexanoylglycine (see tables 2A and 2B). After the prolonged exercise test (at T = 120 to 300), abnormal carnitine-ester profiles were seen in plasma samples of both patients. Elevations were noted of short and medium chain acylcarnitine esters (C-4-, C-6-, C-8-, C-10-, C-10:1-, C-12- and C-14:1-carnitine). Also free cis-decenoic acid was increased in the plasma samples after the prolonged exercise test (32 umol·L-1 for subject 1 and 27 umol·L-1 for subject 2, normal not detectable). Twenty-four hours after the exercise test, blood was drawn and analyzed for CK values and which showed normal results not different from baseline values.
After exercise the increased presence of ethylmalonic acid in urine with dicarboxilic acids combined with the abnormal short and medium chain acylcarnitine esters in plasma is diagnostic of dysfunctioning of multiple acyl-CoA dehydrogenases and therefore MADD. In the urine and plasma of the control subjects, no abnormal metabolites were found in organic acid analyses and acylcarnitines.
Discussion
Little data are available on the risks and benefits of exercise in children with defects in cellular metabolism. It is well known that in milder forms of MADD, myopathy can be a prominent feature and there are no clear guidelines as to whether physical exercise should be avoided or advocated. Recent reports suggest that physical exercise might be beneficial for some patients with defects in energy metabolism [10,11]. It needs to be stressed that before exercise tests are being performed, a careful clinical evaluation is needed, including a complete cardiac evaluation, as cardiomyopathy is frequently present in patients with MADD. To evaluate the risks and advantages of exercise and training, exercise protocols need to be developed for different age groups and disorders. We think that an exercise protocol such as our protocol can be used to evaluate physical fitness and exercise risks in individual patients with a metabolic disease. However, optimal time and intensity of the prolonged exercise test remains to be determined.
The exercise tests that are currently used in our clinical setting (e.g. ischemic fore arm test and maximal exercise test) are not designed and therefore not appropriate for testing endurance exercise capacity. The same holds true for maximal exercise tests. Patients with MADD, even when their myopathy is mild, will probably not experience problems during the first 8–12 minutes of exercise. After maximal exercise, biochemical test results in patient 1 were not informative. In patient 2, before the maximal exercise test ethylmalonic acid was elevated 30 mmol·mol-1 creatinine and did not rise during the test. Etylmalonic acid was the only abnormal metabolite present in the urine sample.
The recently used prolonged low-intensity exercise tests are of more value for evaluating patients with FAO disorders [5,6]. A prolonged exercise test designed for fatty acid oxidation disorders should be of low intensity < 70% of VO2peak, since below < 70% the fat oxidation is maximal [12]. We choose a duration of 90 minutes since this period was also used in previous exercise studies in patients with metabolic muscle diseases [13]. Recently others have used 60 minutes of cycling at 50% of VO2peak in patients with FAO disorders [6]. We have chosen to set the constant workload as a % of maximal workload instead of oxygen uptake, since oxygen uptake drifts during exercise [14]. From Figure 1 can be derived that the length should be at least 45 minutes to show significant differences between patients and controls. However, the greatest differences between the two patients and controls in FAO was at 60 minutes of exercise. Optimal time and intensity of the prolonged exercise test remains to be determined.
One would expect problems during prolonged endurance exercise. During low intensity exercise, the dominant substrates for the muscular activity are plasma free fatty acids (FFA) and muscle triglyceride, plasma glucose and muscle glycogen have only a minor role in the energy provision [15], this role is even lower in children [16]. As relative exercise intensity increases, there is a decrease in the proportion of the energy requirement derived from fat oxidation and an increase in that provided by carbohydrate oxidation [15]. During prolonged low-intensity exercise that can be maintained for 90 minutes or longer, there is a progressive decline in the proportion of energy derived from muscle glycogen and muscle triglyceride, while the oxidation of plasma free fatty acids will increase [15]. As to be expected, a maximal graded exercise test showed no abnormalities in our patients. However, when the prolonged exercise test was performed the characteristic metabolites of MADD appeared in urine and plasma (increased presence of ethylmalonic acid in urine combined with the abnormal mid-chain carnitine ester profile in plasma), without evident clinical or biochemical signs of muscle damage. Moreover, we found a low FAO during the low intensity prolonged exercise test as measured from indirect calorimetry in the two subjects with MADD compared to the 5 control subjects. The FAO oxidation was also lower compared to a previous study in healthy children [16], although in that study, the subjects even exercised at a higher exercise intensity (70% of VO2peak). Moreover, our data indicate that the low FAO is caused by an impaired FAO, and not caused by an impaired mobilization of plasma FFA, since plasma FFA levels were increasing during prolonged exercise.
Our results are comparable with the study of Orngreen et al [6] who tested 2 subject with very long-chain acyl-coa dehydrogenase deficiency (VLCAD) using a 60 endurance exercise test. In both patients with MADD and VLCAD subjects FAO was very low during exercise, and plasma glucose levels were decreasing during exercise, while plasma FFA concentrations were increasing during exercise.
When FAO disorders or MADD are suspected in a patient and the investigations of biological fluids does not reveal significant clues to the diagnosis, which sometimes is the case, "stress-tests" like a fasting test can be used. A fasting test can last up to 24 hours and is a less controlled tests compared to a standardized exercise test. Moreover, a fasting test is much more a burden to the patient than exercise tests. Furthermore, fasting tests do not give any guidance to clinicians in advising physical exercise possibilities or limitations in their patients. Given the fact that in the patients discussed in the manuscript the diagnosis MADD could be made after exercise suggests the diagnostic value of prolonged exercise testing.
Conclusion
We performed a prolonged low-intensity exercise test in two boys with a mild form of MADD and 5 control subjects. The test was tolerated well in the 2 patients and the 5 control subjects, although they were exhausted at the end of the test. During the test the FAO was very low in the MADD compared to the control subjects, while specific metabolites of MADD appeared in plasma and urine samples. We suggest that this test could be of diagnostic importance and might replace the fasting test as a diagnostic procedure in some cases, particularly in patients with anamnestic signs of intolerance for prolonged exercise. Furthermore, based on the moment and rate of appearing of specific metabolites, a tailored counseling regarding physical activity can be given.
Competing interests
The author(s) declare that they have no competing interests
Authors' contributions
TT and JWHC carried out the exercise test and drafted the manuscript. TK, GV, LD carried out the analysis of blood and urine. TT, JWHC, TK, GV, LD, and PJMH conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.
==== Refs
Tein I Metabolic myopathies Semin Pediatr Neurol 1996 3 59 98 8795843 10.1016/S1071-9091(96)80038-6
Vissing J Haller RG A diagnostic cycle test for McArdle's disease Ann Neurol 2003 54 539 542 14520671 10.1002/ana.10725
Ong HY O'Dochartaigh CS Lovell S Patterson VH Wasserman K Nicholls DP Riley MS Gas exchange responses to constant work-rate exercise in patients with glycogenosis type V and VII Am J Respir Crit Care Med 2004 169 1238 1244 15070817 10.1164/rccm.200307-974OC
Mundy HR Georgiadou P Davies LC Cousins A Leonard JV Lee PJ Exercise Capacity and Biochemical Profile during Exercise in Patients with Glycogen Storage Disease Type I J Clin Endocrinol Metab 2005 15671110
Orngreen MC Duno M Ejstrup R Christensen E Schwartz M Sacchetti M Vissing J Fuel utilization in subjects with carnitine palmitoyltransferase 2 gene mutations Ann Neurol 2005 57 60 66 15622536 10.1002/ana.20320
Orngreen MC Norgaard MG Sacchetti M van Engelen BG Vissing J Fuel utilization in patients with very long-chain acyl-coa dehydrogenase deficiency Ann Neurol 2004 56 279 283 15293280 10.1002/ana.20168
Binkhorst RA van 't Hof MA Saris WHM Maximale inspanning door kinderen; referentiewaarden voor 6-18 jarige meisjes en jongens [Maximal exercise in children; reference values girls and boys, 6-18 year of age] 1992 Den-Haag, Nederlandse Hartstichting 1 64
Backman E Methods for measurement of muscle function. Methodological aspects, reference values for children, and clinical applications Scand J Rehabil Med Suppl 1988 20 9 95 3074485
Peronnet F Massicotte D Table of nonprotein respiratory quotient: an update Can J Sport Sci 1991 16 23 29 1645211
Mahoney DJ Parise G Tarnopolsky MA Nutritional and exercise-based therapies in the treatment of mitochondrial disease Curr Opin Clin Nutr Metab Care 2002 5 619 629 12394637 10.1097/00075197-200211000-00004
Taivassalo T Haller RG Implications of exercise training in mtDNA defects--use it or lose it? Biochim Biophys Acta 2004 1659 221 231 15576055 10.1016/j.bbabio.2004.09.007
Achten J Gleeson M Jeukendrup AE Determination of the exercise intensity that elicits maximal fat oxidation Med Sci Sports Exerc 2002 34 92 97 11782653 10.1097/00005768-200201000-00015
de Meer K Roef MJ de Klerk JB Bakker HD Smit GP Poll-The BT Increasing fat in the diet does not improve muscle performance in patients with mitochondrial myopathy due to complex I deficiency J Inherit Metab Dis 2005 28 95 98 15702410 10.1007/s10545-005-1485-8
van Loon LJ Greenhaff PL Constantin-Teodosiu D Saris WH Wagenmakers AJ The effects of increasing exercise intensity on muscle fuel utilisation in humans J Physiol 2001 536 295 304 11579177 10.1111/j.1469-7793.2001.00295.x
Romijn JA Coyle EF Sidossis LS Gastaldelli A Horowitz JF Endert E Wolfe RR Regulation of endogenous fat and carbohydrate metabolism in relation to exercise intensity and duration Am J Physiol 1993 265 E380 91 8214047
Timmons BW Bar-Or O Riddell MC Oxidation rate of exogenous carbohydrate during exercise is higher in boys than in men J Appl Physiol 2003 94 278 284 12391100
| 15907213 | PMC1159171 | CC BY | 2021-01-04 16:37:46 | no | Nutr Metab (Lond). 2005 May 20; 2:12 | utf-8 | Nutr Metab (Lond) | 2,005 | 10.1186/1743-7075-2-12 | oa_comm |
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-471592706310.1186/1465-9921-6-47ResearchShort-term variability of biomarkers of proteinase activity in patients with emphysema associated with type Z alpha-1-antitrypsin deficiency Stolk Jan [email protected] Barbara [email protected] Laura [email protected] Chiara [email protected] Elly M [email protected] Kuppevelt Toine H [email protected] Willem [email protected] Paolo [email protected] Maurizio [email protected] Department of Pulmonology, Leiden University Medical Center, Leiden, The Netherlands2 Department of Medical Statistics, Leiden University Medical Center, Leiden, The Netherlands3 Laboratory of Capillary Electrophoresis, Department of Biochemistry, University of Pavia, Italy4 Laboratory of Biochemistry and Genetics, Department of Respiratory Disease, IRCCS San Matteo Hospital, Pavia, Italy5 Department of Biochemistry, 194, University Medical Center, NCMLS Nijmegen, The Netherlands6 TNO-Prevention and Health, Gaubius Laboratory, Leiden, The Netherlands2005 31 5 2005 6 1 47 47 3 2 2005 31 5 2005 Copyright © 2005 Stolk et al; licensee BioMed Central Ltd.2005Stolk et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The burden of proteinases from inflammatory cells in the lung of subjects with type Pi ZZ of alpha-1-antitrypsin deficiency is higher than in those without the deficiency. Cross-sectional studies have shown increased levels of biomarkers of extracellular matrix degradation in vivo. Longitudinal variability of these biomarkers is unknown but desirable for clinical studies with proteinase inhibitors.
Methods
We measured three different types of biomarkers, including desmosines, elastase-formed fibrinogen fragments and heparan sulfate epitope JM403, in plasma and urine for a period of 7 weeks in a group of 12 patients who participated in a placebo-controlled study to assess the safety of a single inhalation of hyaluronic acid.
Results
Effect of study medication on any of the biomarkers was not seen. Baseline desmosines in plasma and urine correlated with baseline CO diffusion capacity (R = 0.81, p = 0.01 and R = 0.65, p = 0.05). Mean coefficient of variation within patients (CVi) for plasma and urine desmosines was 18.7 to 13.5%, respectively. Change in urinary desmosine levels correlated significantly with change in plasma desmosine levels (R = 0.84, p < 0.01). Mean CVi for fibrinogen fragments in plasma was 20.5% and for JM403 in urine was 27.8%. No correlations were found between fibrinogen fragments or JM403 epitope and desmosines.
Conclusion
We found acceptable variability in our study parameters, indicating the feasibility of their use in an evaluation of biochemical efficacy of alpha-1-antitrypsin augmentation therapy in Pi Z subjects.
alpha-1-antitrypsinemphysemaJM403desmosinesbiomarkers
==== Body
Background
Polymorphonuclear leukocytes (PMNs) play a major role in the pathogenesis of chronic obstructive pulmonary disease (COPD), in particular in emphysema [1]. In subjects with Pi Z type of alpha-1-antitrypsin deficiency (AATD) the burden of PMN and other inflammatory cells in the lung is higher than in those without the deficiency [2,3]. The serum levels of alpha-1-antitrypsin (AAT) found in deficient AAT subjects with phenotypes ranging between Null/Null and MZ correlate with clinical severity of emphysema and suggest that AAT is the most important inhibitor of protease activity in the lung [4]. Proteinases released by inflammatory cells such as PMN and macrophages are able to degrade the extracellular matrix components in lung interstitium, including elastin, proteoglycans and collagens [5]. Although active degradation is difficult to demonstrate in vivo, immunohistochemical studies in resected human lung have shown PMN elastase and other proteases present on extracellular matrix components, suggesting that enzyme is in contact with its substrate for degradation [6]. In patients with AATD, such degradation is thought to be more active in the absence of AAT.
The assessment of inflammatory cell-mediated extracellular matrix degradation in vivo partly suffers from the lack of specific biochemical markers that reflect proteolysis and thus protease activity in vivo. For example, neutrophil elastase can be measured in plasma as antigen concentration or in complex with its inhibitor alpha-1-antitrypsin, but this is only an indication of PMN degranulation and may not be representative of functional extracellular proteolytic activity in vivo.
In the past five years, three different concepts of biomarkers of protease activity of extracellular matrix degradation around inflamed alveoli have been published. First, the heparan sulfate specific epitope JM403 was found 10-fold reduced in urine of patients with COPD compared to healthy controls [7]. The decreased urinary content of a specific epitope of heparan sulfate, together with a normal content of heparan sulfate richly present in basement membranes of alveoli suggest a structural alteration or an altered processing of the heparan sulfate molecule in the lungs of patients with emphysema. In view of the biological functions of heparan sulfate, this could lead to destabilisation of the extracellular matrix, facilitating the development of further proteolytic damage to other matrix components [7].
Second, elastin breakdown products were demonstrated in urine and plasma, as a footprint of the degradation of cross-linked elastin [8-10]. Third, large fibrin(ogen) fragments formed by neutrophil elastase-mediated degradation (PMN-FDP) were significantly elevated in plasma of AATD subjects compared to healthy controls, indicating an imbalance in the protease-antiprotease ratio, which allows elastase activity in vivo at sites of inflammation where fibrin(ogen) is deposited [11,12].
The aim of the present study was to measure the above three types of biomarkers in a short-term pharmaceutical safety study to assess biomarker variability between and within patients.
Materials and methods
Subjects and study design
Twelve patients with Pi ZZ type of AATD participated in a double blind, randomised, placebo-controlled phase I study to investigate the safety and tolerability of a single inhalation of hyaluronic acid (HA), using a Pari Boy compressor and LC nebuliser [13]. Patients were randomised for a single inhalation of a solution of HA (0.003 or 0.01% ETX-100 from CoTherix, Belmont, CA, USA) or placebo. This resulted into 3 blocks of treatment, a block of 4 patients who inhaled 0.003% ETX-100 or placebo, a block of 4 patients who inhaled 0.01% ETX-100 or placebo and another block of 4 patients who inhaled 0.003 or 0.01% ETX-100. These dosages were calculated from dosages that have been used in standard toxicity studies in mice and rats, required by the Food and Drug Administration of the USA, to demonstrate a no effect level in standardized pathological examination. The two inhalations were 15 days apart. The first single inhalation was at baseline visit with overnight stay in the clinic, followed by a control visit on day 9. The second single inhalation was at day 15, followed by another overnight stay in the clinic and by a control visit on day 23. A final visit was scheduled 44 days after the baseline visit. Citrate plasma samples were taken at baseline and days 1, 9, 15, 16, 23 and 44. Urine samples of 24 h period of collection were taken on days 1, 9, 16, 23 and 44.
Patient characteristics are shown in Table 1 (see AdditionalfileRespirResTable1). Although no chest CT's were available, chest X-rays of all patients showed signs of panlobular emphysema and none showed bronchiectasis. All patients were in stable clinical condition in the four weeks preceding the baseline visit and none have ever been treated with AAT replacement therapy.
The study was approved by the Ethical Board of LUMC and all patients gave written informed consent. The study was conducted according to Good Clinical Practice.
Desmosine assay
The determination of desmosines in urine was performed essentially as previously described [14]. For analysis of these compounds in plasma, the above method was slightly modified. Briefly, plasma samples (1 ml) were deproteinized by addition of 0.45 M trichloroacetic acid (400 μl) and centrifuged at 14,000 rpm for 10 min. Finally, a 100 μl aliquot of each sample was placed into pyrex tubes, evaporated to dryness in vacuo and hydrolyzed by refluxing with 500 μl of twice distilled constant boiling HCl (6 M) at 106°C for 24 h. The hydrolyzed samples were dried under a nitrogen stream, the residue washed four times in de-ionized water and neutralized with 0.5 M Na2CO3, pH 8.7 to give a final volume of 500 μl. After centrifugation for 15 min at 13,000 rpm, the supernatant was diluted with water (1:1) and labeled for 5 hrs at 45°C in the dark by addition of 50 μl 0.5 mM fluoresceine isothiocyanate (FITC) solution prepared fresh immediately before of use. Samples were then analyzed with capillary electrophoresis and laser-induced fluorescence detection (CE-LIF) using a Beckman P/ACE 2200 (Fullerton, CA, USA) automated system equipped with a LASER MODULE 488 (Beckman) consisting of a 3 mW and a 488 nm air-cooled argon-ion laser. Samples were injected at 3.5 kPa for 10 sec (approximately 10 nl injected) and the separation was carried out at a temperature of 25°C applying a voltage of 30 kV for 25 min. The background electrolyte was 20 mM sodium tetraborate pH 9.0 containing 60 mM sodio dodecyl sulfate and 15% (v/v) methanol. Data were collected and processed using the Beckman System Gold software. The assay is measuring the combination of isodesmosine and desmosine present in both plasma and urine samples. Throughout this manuscript it is referred to as desmosines. The calculated analytical interassay coefficient of variation (CVa) was 4.2 %. Desmosines concentration in urine is normalized to urine creatinine (μg/g creatinine), whereas desmosines concentration in plasma is expressed as μg/L.
JM403 assay
The heparan sulfate JM403 epitope, defined by monoclonal antibody JM403 was quantified using an inhibition enzyme immunoassay (EIA) as previously described [7]. The EIA was preceded by urine preparation as described [7]. The CVa of the assay was 5 %.
PMN-FDP assay
Frozen citrate plasma was warmed for 5 – 10 min in a 37°C water bath until the sample was a clear solution. Capture-antibody-coated microtiter plates from Kordia BV, Leiden, The Netherlands were filled with samples diluted in PBS/Tween and incubated for 2 h at room temperature. Bound PMN-FDP was quantified by incubation for 1 h at RT with peroxidase-labelled Mab DD13 in PBS/Tween with 0.1% (w/v) BSA, and the subsequent conversion of TMB/H2O2 as described [11]. The CVa of the assay was 10%.
Statistical analysis
After the code of treatment was broken, the change over time from baseline to day 15 and from day 15 to day 44 for each of the treatment groups was calculated for each of the three different biomarkers. To test for a treatment effect of hyaluronic acid on each of the biomarkers, the differences between two time points for a given biomarker in the treatment groups was compared with the differences in the group treated with placebo using an univariate general linear model. A t-test for the inter-subject variability for each of the biomarkers in each of the three treatment blocks showed no statistically significant difference.
Mean values and standard deviations of desmosines, JM 403 and PMN-FDP of all available plasma or urine samples for each of the days of collection were calculated in SPSS version 11. Coefficient of variation within and between patients was calculated.
The correlation between baseline FEV1, Kco and baseline value of each of the biomarkers was assessed by the Spearman's rank correlation coefficient.
Results
Analysis of change over time of any of the three biomarkers for a given treatment with hyaluronic acid did not show a statistically significant difference compared to change by placebo treatment in a univariate general linear statistical model. Therefore, all data of each biomarker are shown for the 12 patients as a single study group.
The study medication was well tolerated and only 5 mild adverse events were possibly treatment-related. There were no clinically important changes in lung function or safety variables.
Assay results
The mean values of desmosines and PMN-FDP of the 12 patients for each of the study days were all in the same range (Figure 1, 2, 3). Previously published values in healthy individuals are mentioned in the legend to the figures. Except for patient 4 and 10, JM403 values were within the range of values for patients with general COPD [7]. The JM 403 epitope in urine of patient 4 (FEV1 46% pred., Kco 40% pred.) and 10 (FEV1 30% pred., Kco 50% pred.) showed a markedly elevated value that was in the range of previously published values for healthy subjects (Figure 4) [7].
Figure 1 Plasma desmosines levels (median ± quartels presented as box together with minimal and maximal values) determined on samples collected on indicated days from the patients. Values of plasma desmosines in healthy individuals (n = 15) range between 40 and 60 μg/l, mean 43 ± 3 μg/l (mean ± SD, reference 16).
Figure 2 Urine desmosines levels (median ± quartels) determined on samples collected on indicated days from the study patients. Ten healthy individuals had mean values of 22.70 ± 1.66 μg/g creatinine (mean ± SD, reference 14).
Figure 3 Plasma fibrinogen fragments generated by PMN (PMN-FDP) determined on samples collected on indicated days from the study patients (median ± quartels). Ten healthy individuals had mean values of 35 ± 12 ng/ml (mean ± SD, reference 11).
Figure 4 Urine JM403 values corrected for urine creatinine concentration on samples collected on indicated days from the study patients. Individual values are shown to appreciate the differences between patients. Values are expressed as Units/mg creatinine. One unit is defined as the amount of JM403 epitope present on 1 microgram kidney heparan sulfate. Sample on Day 9 of patient X is missing because he forgot to collect the sample. The top line of asterixes represents data from a patient that also had stable chronic pancreatitis, a condition not known to be associated with type Z alpha-1-antitrypsin deficiency. The other patient with normal values had no other known conditions and had a stable lung function for the past 15 years as measured in our clinic.
Spearman's rank correlation coefficient between baseline lung function values of the patients and each of the baseline biomarkers is presented in Table 2 (see DOC AdditionalfileRespirResTable2). Spearman's correlation of change over time between plasma desmosines and urine desmosines in the 12 patients was 0.84 (P < 0.01) and is shown in Figure 5. No such significant correlation was present for the two other biomarkers.
Figure 5 Change in plasma desmosines level and change in urine desmosines level during the 44 days of the study assessed for each of the 12 participating patients. Spearman's rank correlation coefficient between individual change in plasma and urine desmosines is 0.84 (P < 0.01).
The coefficient of variation (CV) calculated within and between patients is presented in Table 3 (see DOC AdditionalfileRespirResTable3). Mean coefficient of variation within patients (CVi) for plasma and urine desmosines was 18.7 to 13.5%, respectively. Mean CVi for fibrinogen fragments in plasma was 20.5% and for JM403 in urine was 27.8%.
Discussion
The results of this 44-day follow up study showed for the first time the variability of biomarkers of matrix degradation within and between patients with Pi ZZ type of alpha-1-antitrypsin deficiency and emphysema. Furthermore, the present baseline results confirm values of similar patients in previously published cross-sectional studies where the outcome was compared with control subjects [7-9,11]. The patients in the present study were carefully controlled during the conduct of the study, including two overnight stays in the clinic and three visits in the out patient clinic, allowing optimal conditions for sample collection and preparation.
The assays were performed in three different laboratories, each with their own expertise built up in the past 5 years. For the JM403 epitope assay, urine samples needed specific pre-treatment [7]. Glycosaminoglycans were first isolated by an ion-exchange column and quantified. From the glycosaminoglycan eluate, a sample was used for the JM403 EIA. When values of our patients 4 and 10 with high JM403 levels are taken out, a mean of 0.7 ± 0.2 U/mg creatinine is highly comparable to the previously published value of 0.6 ± 0.1 U/mg creatinine, indicating stable assay performance of the EIA despite a complicated sample preparation. This is in contrast to the PMN-FDP assay [11]. We learned that cryo-precipitation of fibrinogen and large fibrinogen fragments in plasma samples requires careful sample handling, including snap freezing and rapid thawing to 37°C to prevent the formation of cryoprecipitates within the sample. In a previous report on this assay, our values for Pi ZZ patients were ten-fold lower, as these samples were thawed to and measured at room temperature [11]. To minimize solute losses in the system and to decrease the degree of variability of data, the determination of urinary desmosines was performed on urine samples not submitted to any pretreatment procedure other than filtering the sample. For all three assays, it is possible to measure values with an analytical inter-assay coefficient of variation (CVa) below 10%. This is well below 15% which is the value derived from the criterion that the CVa should be below half of the intra-individual coefficient of variation (CVi), with ranges between individuals reported in Table 3. Therefore, all three assays have acceptable variability and can be used for studies that aim to evaluate the effect of drugs on the level of these biomarkers [15].
The key question about all three of our assays is whether they only reflect matrix degradation present in lung interstitium or whether significant other parts in the body also contribute to elevated plasma levels of the biomarkers. Recent data showed abnormal high levels of desmosines in plasma and urine of patients affected by Pseudoxanthoma elasticum, an inherited disorder of connective tissue characterized by severe alterations of its components [16]. As COPD/emphysema is now recognised not only as a pulmonary condition, but also as a systemic disorder, including skeletal muscle wasting and skin atrophy, it is possible that other tissues in the body also contributed to the levels of our biomarkers due to the same processes. Furthermore, almost all patients with COPD/emphysema have a history of cigarette smoking that may also result in vascular atherosclerosis, a condition that is able to affect the levels of our biomarkers. To our knowledge, no lung specific protein, sensitive to proteolytic cleavage, has been identified yet to serve as a useful substrate for proteolysis to act as a footprint of inflammation-induced matrix destruction specific for the lung. Therefore, the most appropriate way to study the contribution of inflammation-induced lung matrix degradation would be the use of specific protease inhibitors in patients and to measure their effects on levels of our biomarkers. So far, urinary desmosines only has been used to monitor the efficacy of the AAT replacement therapy in short-term, open label trials in AATD subjects [17,18]. The latter study showed a small, but not significant, decrease in desmosines excretion after 24 weeks of replacement therapy, which suggests that reversal of chronic proteolysis in AATD cannot be achieved quickly. Just like in the above-mentioned studies, also our study was of short-term duration and no change in spirometry or CO diffusion capacity is to be expected. Therefore, a definitive answer to the question whether the increased urinary excretion in AATD-associated COPD is mostly due to accelerated proteolysis within the lung would require a more extended period of replacement therapy (probably more than one year) or a higher dose than currently used. Apart from above mentioned manuscripts, all the other reports dealing with matrix degradation biomarkers in COPD have been designed in a cross-sectional fashion [8,9,19].
A unique feature of the present study is that three different laboratories have united to measure three different biomarkers of matrix degradation in two different biological fluids being serially collected from a well-defined study population. Such collaboration arose from the Alpha1 International Registry (AIR), a consortium of scientists with special interest in AATD [20]. In this study, urinary and plasma desmosines have been detected simultaneously for the first time in these patients. Plasma normally contains peptides derived either from tropoelastin or from degraded crosslinked mature elastin. It has been reported that these circulating peptides have a wide variety of sizes, peaking at 70 kD, but with a significant proportion of peptides with lower molecular weight. The latter are expected to more easily filter into urine. Chromatographic separation of urine elastin peptides in humans has detected material over a wide range of MW, from < 5 kD to > 50 kD, peaking at 10–50 kD [21]. Detection of desmosines allows discrimination, among elastin peptides, of those derived from the breakdown of mature elastin from those derived from nascent elastin. The good correlation between urinary and plasma desmosines here shown is consistent with the correlation between urinary and plasma elastin peptides previously reported, suggesting that plasma and urine provide generally comparable estimates of elastin turnover, at least in a short-term [21]. The high cross-sectional correlation between baseline CO diffusion capacity and both urine and plasma desmosines and the absence of such correlation with either PMN-FDP or JM403 raises the question as to whether one is better than the other. Differential changes in any of these biomarkers during acute exacerbations or AAT supplementation therapy may answer this question. In addition, all three parameters do not seem to correlate with each other. This may in part be explained by a difference in clearance from the body. The half-life of desmosines-containing elastin fragments in mice is about 2 h [22], about 5 days for PMN-FDP in humans [23] and is unknown for JM403.
Conclusion
We have shown, compared to previously reported healthy individuals, significantly altered levels of three different surrogate markers representing footprints of matrix degradation simultaneously present in patients with type Z alpha-1-antitrypsin deficiency and clinically significant emphysema. In addition, acceptable variability in our study parameters occurred. The footprints are recommended for use in an evaluation of biochemical efficacy of alpha-1-antitrypsin augmentation therapy in Pi Z subjects before initiation of studies involving functional tests of the lung and CT scan lung density assessment [24].
Abbreviations
AAT, alpha-1-antitrypsin
AATD, alpha-1-antitrypsin deficiency
CE-LIF, capillary electrophoresis and laser-induced fluorescence detection
COPD, chronic obstructive pulmonary disease
FITC, fluoresceine isothiocyanate
HA, hyaluronic acid
EIA, inhibition enzyme immunoassay
Kco, CO diffusion capacity per volume unit of alveolar ventilation
PMN-FDP, neutrophil elastase-mediated degradation
PMNs, polymorphonuclear leukocytes
Competing interests
Financial support: The study was conducted as a phase I study in collaboration with CoTherix, Belmont, CA, USA who financially supported the study.
All authors declared no conflict of interest with the content of the manuscript.
Authors' contributions
All authors contributed to the content of the manuscript. The desmosines were measured by LA, CZ and PI. WN was responsible for the PMN-FDP assay. TK and EV performed the JM403 assay and BV performed the statistical analysis. JS and ML were responsible for the clinical work and JS wrote the manuscript. All authors have read the text and contributed corrections.
Acknowledgements
The antibody for JM403 was kindly provided by prof. J.H.M. Berden, Department of Nephrology, Nijmegen University Medical Center. Discussions with members of the Alpha1 International Registry (AIR, ) were much appreciated. Part of the work on desmosines was funded by Fondazione Cariplo (Milan, Italy) # 2003 1643/10.8485.
==== Refs
Barnes PJ Shapiro SD Pauwels RA Chronic obstructive pulmonary disease: molecular and cellular mechanisms Eur Respir J 2003 22 672 688 14582923
Morrison HM Kramps JA Afford SC Burnett D Dijkman JH Stockley RA Elastase inhibitors in sputum from bronchitic patients with and without alpha 1-proteinase inhibitor deficiency: partial characterization of a hitherto unquantified inhibitor of neutrophil elastase Clin Sci 1987 73 19 28 2440636
Parmar JS Mahadeva R Reed BJ Farahi N Cadwallader KA Keogan MT Bilton D Chilvers ER Lomas DA Polymers of alpha(1)-antitrypsin are chemotactic for human neutrophils: a new paradigm for the pathogenesis of emphysema Am J Respir Cell Mol Biol 2002 26 723 730 12034572
Cox DW Levison H Emphysema of early onset associated with a complete deficiency of alpha-1-antitrypsin (null homozygotes) Am Rev Respir Dis 1988 137 371 375 3257661
Mahadeva R Shapiro SD Chronic obstructive pulmonary disease * 3: Experimental animal models of pulmonary emphysema Thorax 2002 57 908 914 12324680 10.1136/thorax.57.10.908
Damiano VV Tsang A Kucich U Abrams WR Rosenbloom J Kimbel P Fallahnejad M Weinbaum G Immunolocalization of elastase in human emphysematous lungs J Clin Invest 1986 78 482 493 3525610
Van de Lest CH Versteeg EM Veerkamp JH Berden JH van den Born J Heunks L Lammers JW van Herwaarden CL Dekhuijzen PN van Kuppevelt TH Altered composition of urinary heparan sulfate in patients with COPD Am J Respir Crit Care Med 1996 154 952 958 8887591
Stone PJ Gottlieb DJ O'Connor GT Ciccolella DE Breuer R Bryan-Rhadfi J Shaw HA Franzblau C Snider GL Elastin and collagen degradation products in urine of smokers with and without chronic obstructive pulmonary disease Am J Respir Crit Care Med 1995 151 952 959 7697272
Viglio S Iadarola P Lupi A Trisolini R Tinelli C Balbi B Grassi V Worlitzsch D Doring G Meloni F Meyer KC Dowson L Hill SL Stockley RA Luisetti M MEKC of desmosine and isodesmosine in urine of chronic destructive lung disease patients Eur Respir J 2000 15 1039 1045 10885422 10.1034/j.1399-3003.2000.01511.x
Kucich U Christner P Lippmann M Kimbel P Williams G Rosenbloom J Weinbaum G Utilization of a peroxidase antiperoxidase complex in an enzyme-linked immunosorbent assay of elastin-derived peptides in human plasma Am Rev Respir Dis 1985 131 709 713 3890638
Bos R van Leuven CJM Stolk J Hiemstra PS Ronday HK Nieuwenhuizen W An enzyme immunoassay for polymorphonuclear leucocyte-mediated fibrinogenolysis Eur J Clin Invest 1997 27 148 156 9061309 10.1046/j.1365-2362.1997.810634.x
Pacholok SG Davies P Dorn C Finke P Hanlon WA Mumford RA Humes JL Formation of polymorphonuclear leukocyte elastase: alpha 1 proteinase inhibitor complex and A alpha (1–21) fibrinopeptide in human blood stimulated with the calcium ionophore A23187. A model to characterize inhibitors of polymorphonuclear leukocyte elastase Biochem Pharmacol 1995 49 1513 1520 7763294 10.1016/0006-2952(95)00031-T
Turino GM Cantor JO Hyaluronan in respiratory injury and repair Am J Respir Crit Care Med 2002 167 1169 1175 12714341 10.1164/rccm.200205-449PP
Annovazzi L Viglio S Perani E Luisetti M Baraniuk J Casado B Cetta G Iadarola P Capillary electrophoresis with laser induced fluorescence detection of urinary desmosines Electrophoresis 2004 25 683 691 14981697 10.1002/elps.200305607
Harris EK Statistical principles underlying analytic goal-setting in clinical chemistry Am J Clin Pathol 1979 72 374 82 474517
Annovazzi L Viglio S Gheduzzi D Pasquali-Ronchetti I Zanone C Cetta G Iadarola P High levels of desmosines in urine and plasma of patients with pseudoxanthoma elasticum Eur J Clin Invest 2004 34 156 164 14764080 10.1111/j.1365-2362.2004.01306.x
Gottlieb DJ Luisetti M Stone PJ Allegra L Cantey-Kiser JM Grassi C Snider GL Short-term supplementation therapy does not affect elastin degradation in severe alpha(1)-antitrypsin deficiency. The American-Italian AATD Study Group Am J Respir Crit Care Med 2000 162 2069 2072 11112116
Stoller JK Rouhani F Brantly M Shahin S Dweik RA Stocks JM Clausen J Campbell E Norton F Biochemical efficacy and safety of a new pooled human plasma alpha(1)-antitrypsin, Respitin Chest 2002 122 66 74 12114340 10.1378/chest.122.1.66
Ma S Lieberman S Turino GM Lin YY The detection and quantitation of free desmosine and isodesmosine in human urine and their peptide-bound forms in sputum Proc Natl Acad Sci U S A 2003 100 12941 12943 14563926 10.1073/pnas.2235344100
Luisetti M Miravitlles M Stockley RA Alpha1-antitrypsin deficiency: a report from the 2nd meeting of the Alpha One International Registry, Rapallo (Genoa, Italy), 2001 Eur Respir J 2002 20 1050 1056 12412702 10.1183/09031936.02.00302502
Schriver EE Davidson JM Sutcliffe MC Swindell BB Bernard GR Comparison of elastin peptide concentrations in body fluids from healthy volunteers, smokers, and patients with chronic obstructive pulmonary disease Am Rev Respir Dis 1992 145 762 766 1554198
Stracher B Peterson B The kinetics of elastolysis: elastin catabolism during experimentally induced fibrosis Exper Lung Res 1998 25 407 424
Stolk J Nieuwenhuizen W High dose intravenous AAT and plasma neutrophil derived fibrinogen fragments Thorax 2005 60 84 (letter to the editor) 15618590
Dirksen A Dijkman JH Madsen F Stoel B Hutchison DC Ulrik CS Skovgaard LT Kok-Jensen A Rudolphus A Seersholm N Vrooman HA Reiber JH Hansen NC Heckscher T Viskum K Stolk J A randomised clinical trial of alpha(1)-antitrypsin augmentation therapy Am J Respir Crit Care Med 1999 160 1468 1472 10556107
| 15927063 | PMC1159172 | CC BY | 2021-01-04 16:23:26 | no | Respir Res. 2005 May 31; 6(1):47 | utf-8 | Respir Res | 2,005 | 10.1186/1465-9921-6-47 | oa_comm |
==== Front
Biomed Eng OnlineBioMedical Engineering OnLine1475-925XBioMed Central London 1475-925X-4-151574828410.1186/1475-925X-4-15ReviewOn the methodological unification in electroencephalography Durka Piotr J [email protected] Institute of Experimental Physics, Warsaw University, ul. Hoża 69 00-681 Warszawa, Poland2005 4 3 2005 4 15 15 1 3 2005 4 3 2005 Copyright © 2005 Durka; licensee BioMed Central Ltd.2005Durka; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
This paper presents results of a pursuit of a repeatable and objective methodology of analysis of the electroencephalographic (EEG) time series.
Methods
Adaptive time-frequency approximations of EEG are discussed in the light of the available experimental and theoretical evidence, and applicability in various experimental and clinical setups.
Results
Four lemmas and three conjectures support the following conclusion.
Conclusion
Adaptive time-frequency approximations of signals unify most of the univariate computational approaches to EEG analysis, and offer compatibility with its traditional (visual) analysis, used in clinical applications.
==== Body
1 Background and main ideas
1.1 Historical background
"Animal electricity" has been subject to scientific research since the end of the 18th century, when Galvani and Volta performed their famous experiments [1]. Electrical activity of the brain was first mentioned in 1875, in a grant report by R. Caton [2]. In 1929 the first electroencephalogram (EEG) was recorded from the surface of human scalp by Hans Berger [3].
Year 1935 witnessed birth of the major fields of today's clinical electroencephalography. F. Gibbs and H. Davis [4] showed association of 3/sec spike-wave complexes in EEG with epileptic petit mal absences, and A. L. Loomis et al [5] methodically studied human sleep EEG patterns and the stages of sleep. Also in 1935, the first electroencephalograph (Grass Model I) started the era of contemporary EEG recording: galvanometers, used in earlier decades to record EEG traces on photographic paper, were replaced by 3-channels preamplifier, and the recording was drawn by ink writer on rolls of paper. These rolls were later replaced by folded paper, and, currently, by digital storage and display of EEG traces. Also, contemporary amplifiers provide higher sensitivity and number of channels, but all these changes are quantitative rather than qualitative.
Finally, by the end of forties, Dawson [6] recorded first evoked potentials. He constructed an advanced mechano-electrical (analog) device for averaging brains potentials triggered by a stimulus [7]. Averaging was indispensable to show the event-related activity, which is normally invisible in the on-going EEG background.
These techniques, combined with visual analysis of recorded EEG traces, constitute the canon of contemporary clinical electroencephalography.
1.2 Computational techniques of EEG analysis
First attempts were made by Hans Berger in 1932 [8]; he was assisted by the physicist Dietsch, who applied Fourier analysis to short EEG sections. Then, quoting [9]: The 1950s saw the early generation of automatic frequency analyzers approaching and eventually saw the end of these magnificent but mostly unused machines.
Digital storage of the EEG time series opened unlimited possibilities of offline analysis, and triggered significant efforts towards algorithmic solutions. Studies published in last decades cover the whole spectrum of possible signal processing methods. This is partially due to the fact, that "first" application of given signal processing method to an old experimental paradigm or dataset usually fulfills the requirement of scientific novelty, and justifies publication.
1.3 The need for a unification
As mentioned in section 1.1, nowadays technology of EEG recording is quite satisfactory: we can record simultaneously more channels than the number of electrodes which would fit on a scalp, with sampling (in both time and amplitude) exceeding the assumed information content of the signal. Unresolved issues relate to the analysis of these huge amounts of data.
One of the major problems in biomedical sciences is the inter-subject variability. Especially in neurosciences, it may dramatically exceed the effects attributed to the investigated phenomena. This naturally leads to an unrealistic postulate that, to draw significant conclusions, each new hypothesis or method must be evaluated on a sufficiently large dataset, that is involving enough (probably at least thousands) human subjects. Otherwise, a coherent progress may be achieved only via comparison of different studies, performed and analyzed in a compatible way.
The only paradigm, up to now applied to a significant amount of cases, is the classical, visual analysis of EEG, involved in clinical and research applications. Classic EEG evaluation always involved measuring frequency and/or amplitude with the help of simple rulers [10]. Some of the EEG structures are even defined using terms like "cycles per second" and "microvolts" [11]. Nevertheless, new methods proposed for EEG analysis are mostly incompatible with such description. Therefore, they cannot be directly related to this indispensable knowledge base. Also, results offered by most of the advanced methods are seldom compatible with each other, so they do not seem to form a new, coherent knowledge base.
This situation, at least in relation to the clinical applications, was summarized in [12] as presented in Figure 1.
Figure 1 70 years of progress in clinical electroencephalography: from visual analysis of EEG traces on paper (background picture) to visual analysis of EEG traces displayed on CRT (front). Figure and caption reprinted from [12], © 2001 IEEE.
1.4 Univariate EEG analysis
This section briefly recalls the univariate linear and quadratic transforms, which were used in a significant amount of EEG studies.
1.5 Linear expansions
To explore the properties of an unknown signal f – in the absence of a quantitative model of its generation – we may use a linear expansion in a set of known functions {g1(t), g2(t),..., gN(t)}:
If {gi} form an orthonormal basis, ai are given simply by
Traditionally, orthogonal {gi} sets were limited to Dirac's deltas δi and complex exponentials eiωt, corresponding to the time-domain and spectral analyses. In recent decades, new orthogonal bases were discovered, like e.g. wavelet packets or local cosines. The first and most popular orthogonal wavelet bases are built from dilations and translations of a single function . Only very special choices of ψ give an orthogonal basis, c.f. [13] – it was not before 1980. that first such orthogonal sets were discovered. Scaling a single function implies the characteristic "zooming property" of wavelets, that is increased time resolution and decreased frequency resolution at high frequencies. However, in the EEG analysis this seems to be more of a drawback than advantage.
To interpret (1), we usually relate those gi, that correspond to the largest |ai|, to some properties of f. For example, large value of (spectral peak at ω0) suggests, that f contains periodic activity with base frequency near ω0. However, presence of the same periodic activity will not be so evident if |ai| are calculated e.g. in a wavelet basis, since in this case it will be diluted across several wavelets.
1.6 Time-frequency distributions
Increasing need to study the detailed picture of event-related changes of signal energy – e.g. in the event-related EEG (de)synchronization studies [14] or research on brain-computer interfaces [15] – turns our attention to the quadratic estimates of the time-frequency energy density. The most natural (at least for a physicist) is the Wigner-Ville transform :
However, it suffers from severe interferences, that is cross-terms. Cross-terms are the area of a time-frequency energy density estimate, which may be interpreted as indicating false presence of signal's activity in given time and frequency coordinates. Their origin can be deduced e.g. from equation (11). Several reduced interference distributions (RID) were proposed to minimize this drawback. It is achieved via the smoothing kernel φ, which decreases the influence of cross-terms at a cost of lower time-frequency resolution:
The most popular time-frequency distributions are spectrogram (squared modulus of the short-time Fourier transform) and scalogram (squared modulus of continuous wavelet transform). All of them can be derived from (4) with different kernels φ, and they give better or worse results for different types of signals. Therefore, it is possible to obtain a clear time-frequency picture of a given signal f, if we find an optimal kernel φ for its particular content. However, there is no general recipe for this choice. Also, these quadratic representations do not offer the direct parameterization of signal's features such as formula (1).
1.7 Adaptive time-frequency approximation
In the ideal situation we would like to have a linear expansion like (1), but with functions gi representing all the relevant structures present in f. Mere expansion of the set {gi} beyond a basis does not solve the problem. For the representation of each signal f we must choose the set functions {}i = 1..M adapted to represent its particular content. As a result, we get an approximation:
Approximation appears instead of the exact expansion, because to choose from a set large enough to represent efficiently relevant signal's structures, we must give up the orthogonality. Adaptively chosen orthogonal representations, available in terms of wavelet packets, do not provide enough flexibility. Criterion of choice can be based upon the minimal norm of the residue. We can define an optimal M-approximation as an expansion, minimizing the following error ε:
Finding such an optimal approximation is NP-hard (intractable). It means that its computational complexity is growing with the size of the problem (in this case dimension of the signal) exponentially – in computer science this term is used as synonym of "faster than any polynomial". Such problems, even for moderately-sized input data, require times easily exceeding e.g. the age of the Universe.
We can formulate an almost equivalent decision problem, asking whether for given D, ε1 and f exists an M-approximation giving error (6) smaller than ε1. It can be proved to belong to the NP-complete class [16], that is problems possesing the two following properties [17]:
• Finding a solution requires exponential (non-polynomial time), but given solution can be checked in polynomial (P) time. The name "nondeterministic-polynomial" stems from the lack of a deterministic algorithm finding this answer in polynomial time – "guessing" the right answer is non-deterministic.
• Solution of each of the problem from the NP-class can be translated into solution of any other problem from this class in a polynomial time.
This class encompasses a variety of important problems, from the classical traveling salesman problem to e.g. factorization of large numbers, which has direct implications in cryptography. It is generally believed, that NP-complete problems do not have polynomial-time solutions, yet this basic theorem (P ≠ NP) is still not proved (it may be also undecidable from the currently applied axioms).
Therefore, we rely on a suboptimal solution, which can be found by means of an iterative procedure like the matching pursuit algorithm (MP), proposed in [18]. However, even this sub-optimal solution is computationally intensive, so the first practical applications were not possible before the last decade.
1.8 Matching pursuit algorithm
In the first step of MP, the waveform which best matches the signal f is chosen from the dictionary D. In each of the consecutive steps, the waveform is matched to the signal Rn f, which is the residual left after subtracting results of previous iterations:
Orthogonality of Rn+1 f and in each step implies energy conservation
For a complete dictionary the procedure converges to f
1.9 Time-frequency dictionaries of Gabor functions
Time-frequency dictionaries are composed – apart from the complete Dirac and Fourier bases – from the Gabor functions, since these functions provide optimal joint time-frequency localization [13]. Real valued Gabor can be expressed as
where K(γ) is such that ||gγ|| = 1. Parameters γ = {u, w, s} of the possible Gabor functions constitute a 3-dimensional (phase φ is optimized separately in practical implementations), continuous space, from which a finite dictionary must be chosen for an implementation of the procedure (7). Originally (in [18]), dictionary's parameters were chosen from predefined dyadic sequences. However, any fixed scheme of subsampling the space of possible dictionary's functions leads to a statistical bias of the resulting decompositions. A solution proposed in [19] relies on stochastic dictionaries, in which parameters {u, w, s} are drawn from uniform distributions across ranges defined by sizes of the signal and the dictionary.
1.10 Estimate of signal's energy density
If we find the signal's approximation (5), we can also solve the problem of cross-terms present in the distributions from section 1.6. Namely, calculating the Wigner distribution of expansion (5) would yield
where (, ) is a cross-Wigner transform of and given by
Double sum in (11) contains all the cross-terms. Owing to the representation (5), we can omit them explicitly and construct the time-frequency representation of signal's energy density from the first sum, containing auto-terms:
Energy conservation of this distribution is easily demonstrated (c.f. [18]).
2 Unification
In this section we collect and discuss the statements, needed for the acceptance of the thesis formulated in the Introduction:
Adaptive time-frequency approximations of signals unify most of the univariate computational approaches to EEG analysis, and offer compatibility with its traditional (visual) analysis, used in clinical applications.
2.1 Lemma: Matching pursuit sub-optimal solution to the problem of adaptive approximation is suitable for EEG analysis
Sub-optimal solution of an intractable problem (section 1.7) must have its price. Mathematical examples of failures in pattern recognition, due to the greedy strategy (7) applied by the matching pursuit, were presented e.g. in [20,21]. These cases did not relate to structures encountered in biomedical signals, but the lack of a relevant counterexample does not deny it's existence. Figure 2 gives an example of a case referring directly to the transient oscillatory activity of the kind which may be actually present in the EEG [22].
Figure 2 A failure in feature extraction: R0 – analyzed signal, g0 – function fitted in the first iteration by the MP algorithm. Horizontal – time, vertical – amplitude, both in arbitrary units (simulated signals).
Signal (R0) in Figure 2 is composed from two Gabor functions, both of them actually present in the dictionary D used for the decomposition (Section 1.9). In spite of that, we observe (in the right column) that the first function(g0) fitted to the signal is completely different from either of the two functions, from which the signal was composed! According to (7), MP algorithm has chosen function gi giving the largest product in a single step. Of course, taking into account the next steps, this decision was definitely not optimal. Choosing the two Gabors, which were exactly represented in D, would explain 100% of signal's energy in only 2 iterations, contrary to the residues produced in the left column of Figure 2 as a consequence of the first choice.
However, such an effect occurs only if both Gabors present in the signal have not only the same frequencies, but also exactly the same phase. Such a "coincidence" would be possible in a biological signal only if both the structures were produced by the same generator. And still, MP represents them jointly only if their time centers are close enough. Their larger displacement would result in separate representation even in such a synchronized case. Therefore, we may argue that this effect, mathematically classified as a failure of the sub-optimal procedure, is actually a welcome feature in the analysis of physiological signals. It was explored in [23] for detecting the synchronized spiking activity, as opposed to series of unrelated EEG spikes.
Even if we were able to calculate the optimal adaptive decomposition minimizing error (6), it would have one more disadvantage: namely, the set of M functions chosen to optimally represent signal f, may significantly differ from the optimal set of M + 1 functions chosen for the same signal f (from the same dictionary D). With the iterative MP solution (7), {}i = 1..M will be the same in both decompositions. Considering that the features related to the sub-optimality of the MP solution turn to be advantages rather than drawbacks in the analysis of biomedical signals, we conclude that matching pursuit algorithm is the correct implementation for an adaptive time-frequency approximation of EEG.
2.2 Lemma: Matching pursuit decomposition is asymptotically well defined and does not depend on arbitrary settings
Calculating representation (5) of a signal f via matching pursuit in given dictionary D is a deterministic procedure, given by (7). Therefore, for a given signal, representation (5) depends only on the dictionary D. For the Gabor dictionaries discussed in section 1.9, the only free parameter is their size. It influences MP decomposition in a predictable way: larger dictionaries improve the quality of the decomposition, that is less waveforms are needed to achieve the same error ε in (6). However, starting from a certain density of dictionary's waveforms in the space of their parameters {u, w, s}, this improvement reaches asymptotically a saturation. At this point, which we shall tentatively term a "sufficient density of a Gabor dictionary", decompositions in larger dictionaries, or in different stochastic realizations of the same dictionary, are indistinguishable (of course this relates only to the structures coherent with the dictionary, c.f. [18]).
This qualitative reasoning – although still not backed up by a quantitative proof – is well supported by numerical experiments and years of real-world applications. Problems with mathematical derivations stem from the non-linearity of the MP procedure (7) and variable content of the analyzed signals. Therefore, the notion of "sufficient density" of a Gabor dictionary is up to now defined only empirically. Nevertheless, for the following we shall use it as a requirement for (5). For reasonable dimensions of signals, such densities are achieved with the dictionary sizes easily implementable on today's computers.
2.3 Conjecture: Relevant content of the EEG signal is coherent with Gabor time-frequency dictionaries
This statement depends on the meaning of "relevant EEG content", which, unfortunately, does not have a strict definition.
EEG originates from postsynaptic potentials of firing neurons; only if a large ensemble is working synchronously, the signal can reach a level measurable from the scalp electrodes. A single electrode records a potential from an arbitrary number of such ensembles, spatially filtered by different conductivities of cortex, dura mater, skull and scalp. Recorded EEG is not only affected by possibly non-linear interactions between these ensembles, but the signal may also contain extra-cerebral – biological or not – potentials (artifacts).
Rhythmic activity has been recognized as a prominent feature of the signal since the beginnings of EEG recordings. It appears as transient – rather than permanent – oscillations. These oscillations are produced by large masses of neurons (typically 104 – 107 [24]). Synchronization and desynchronization of these neural masses cannot be instantaneous, so the oscillations exhibit some waxing and waning. Gabor functions provide a concise general model for such signals.
"Relevant content of EEG" can be also translated to "relevant structures used to date in EEG analysis". That would refer mainly to the visual analysis, discussed in lemma 2.6 and 2.7.
Finally, non-permanence of this statement – underlined by using "conjecture" rather than "lemma" – relates also to the fact that, if any other structures would appear to be important in this context, they can be added to the dictionaries without a negative impact on the convergence of the MP procedure (7).
2.4 Conjecture: All the relevant EEG structures, which can be found via linear expansions in different bases, are included in the MP parameterization
As discussed in section 1.5, exploratory value of linear expansions relates to gi with the largest weights in expansion (1) – only those are usually treated as reflecting the significant signal's structures. Accepting Conjecture 2.3 we assume, that all relevant EEG structures are well represented in the Gabor dictionary. Also, waveforms used in linear expansions (1) can be more or less exactly assimilated by smooth functions (10). E.g. some wavelets contain discontinuities, but in the context of spatially-filtered EEG (c.f. Lemma 2.3) smooth Gabor functions should be more adequate for its description. Therefore, we may expect that if a signal contains a mixture of oscillatory and transients activities, and each of them would be highlighted in a different expansion (1), then all of them will be efficiently represented in the MP approximation (5).
As an example, we may quote a study of pharmaco-EEG [25], where traditional spectral estimators of spindling activity between 12 and 14 Hz were replaced by the energy of relevant structures estimated from MP expansion. It provided full compatibility with the traditional approach and better accordance with physiological expectations, owing to the increased sensitivity of MP-based estimator.
However, the correspondence to which we refer in this Conjecture, relates only to EEG structures, which should be parameterized by the MP decomposition as efficiently as in any of the linear expansions (1). But certain statistical properties of the signal can be derived only from its full representation e.g. in wavelet or Fourier bases. On the other hand, properties of MP expansion can be used to construct completely new measures, like e.g. the Gabor atom density (GAD), proposed for prediction of epileptic seizures in [26].
2.5 Lemma: Matching pursuit provides the most general solution to the problem of non-unique estimates of time-frequency energy density, applicable for EEG analysis
All the time-frequency distributions from section 1.6 are estimates of the same magnitude, that is energy density of the signal. However, estimates derived from different approaches (e.g. Wigner, RID, CWT or spectrogram) differ significantly, mainly due to the variable contribution of cross-terms. Also, estimates obtained from the same approach may look quite different depending on their parameters, like e.g. the length of the time window in spectrogram, different wavelets in CWT or smoothing kernels φ in RID.
Estimate (13), derived from the MP decomposition, contains the auto-terms, corresponding to structures present in expansion (5), and no cross-terms (section 1.10). Therefore, for a signal coherent with the dictionary used in the MP decomposition, estimate (13) approximates the intersection of its different quadratic estimates of energy density. This common part corresponds to the generalized auto-terms, contrary to the cross-terms variable between estimates. Therefore, if we accept Conjecture 2.3, this lemma is also proved. Even if we doubt the coherence of EEG signal with the Gabor dictionary, this approach is currently the best candidate for a unifying approach, owing to the uniqueness discussed in Lemma 2.2, since all the other quadratic estimates depend on an arbitrary choice of parameters.
Recent studies of brain electrical activity, recorded from the scalp [14,27] and from the cortex [28], confirmed also that (13) provides resolution superior to the previously applied quadratic methods.
2.6 Lemma: Among the contemporary methods of signal analysis, adaptive approximations with time-frequency dictionaries provide the best correspondence to visual EEG analysis
The knowledge, verified through decades of clinical applications, consists mainly of associations between appearances of rhythmic activities and certain waveforms in EEG with physiologically relevant states or pathologies, like e.g. level of alertness, sleep depth or epileptic seizure. These rhythms and waveforms constitute the dictionary of clinical EEG analysis. Due to a limited repeatability of their visual detection, the field is often considered an art rather than science [24]. Mapping this dictionary into mathematical terms will allow for application of repeatable scientific methods.
EEG rhythms (δ, θ, α, β, γ) are approximately assigned to frequency bands. Therefore, if we assume that those frequency bands are fixed, we can detect their presence from the relative power of the corresponding frequency band in the Fourier spectrum. However, power spectrum is a property of the whole analyzed epoch, and does not provide the information on the time extent of the given oscillatory phenomenon.
Appearance and disappearance of a rhythm can be clearly detected from a time-frequency map of signal's energy density, obtained from one of the quadratic methods discussed in section 1.6. However, stating that oscillations with frequency f last from the time t1 to t2 requires an analysis (post-processing) of a redundant map. Results will depend on the combination of the applied post-processing method and the chosen estimate of the energy density.
On the contrary, unequivocally defined (Lemma 2.2) MP expansion (5) gives explicitly the time span and frequency of detected oscillations.
Transient EEG structures – sleep spindles, K-complexes, epileptic spikes and many others – are obviously undetectable from the spectral estimates. They may exhibit certain, more or less typical, time-frequency signatures; we may imagine using the complex transforms (3) or (4) for their detection, but as previously mentioned we are confronted with continuum of possible transforms and post-processing algorithms.
Adaptive approximations (5) provide today the only method of an explicit description of a significant variety of waveforms. As presented in [23], this description is efficient also for non-oscillating structures like epileptic EEG spikes.
2.7 Conjecture: The correspondence from Lemma 2.6 can bridge the art of visual EEG analysis and reproducible methods of signal processing
Adaptive time-frequency approximation is currently the only signal processing method, offering description of structures, present in a signal, explicitly in terms of their physical parameters like amplitude, frequency or time width. Similarly to the way in which an expert evaluates EEG, the algorithm concentrates on most prominent, local structures, rather than general properties of the whole analyzed epoch.
Visual analysis of EEG relies primarily on detecting occurrences of rhythms and other waveforms, called graphoelements. In some cases, their classical definitions are given already in time-frequency terms. For example, sleep spindles are defined (c.f. [11]) as structures of least 0.5 sec duration, frequency between 12 and 14 Hz and amplitude above 15–25 μV. Such a definition can be directly applied to filter the decomposition (5) in the space of the parameters of functions , fitted to the analyzed signal. Waveforms , conforming to such ranges of parameters, represent sleep spindles – as verified by comparison with their visual detection in [29].
However, not all the graphoelements have such strict, mathematical definitions. The very name "graphoelement" (rather than waveforms) suggests, that their visual detection may rely on different aspects of its shape. Nevertheless, most of these shapes can be effectively approximated by Gabor functions. In such cases we may a posteriori adapt the ranges of the parameters of , to maximize the concordance with visual detection. This resembles the classical paradigm of expert systems, however, using the MP parameterization (5) has the major advantage of operating in the space of physically meaningful and universal parameters, like time width, amplitude or frequency. An example of a successful implementation of such an approach and its broader discussion can be found in [23].
3 Discussion
Sections 2.3, 2.4 and 2.7, necessary for the strict proof of the main thesis, are currently conjectures rather than lemmas. Before we discuss this issue any further, let us gain a proper perspective by considering another theorem, which is generally believed to be true, and, although still not proved, is used by most of us on a daily basis:
As mentioned in section 1.7, we still do not have a strict mathematical proof, that the problems belonging to the NP class do not have a polynomial-time solution (P ≠ NP?). Factorization of large numbers also belongs to the NP class. Strength of the public key cryptography depends on factorization. Therefore, if for any problem from the NP class, a polynomial-time solution is found – which is, in theory, not impossible – it can be used to break the contemporary public key cryptography. Nevertheless, we are still using this system for securing transmission of sensitive data (e.g. credit card numbers) over Internet. Our belief is based upon the lack of counterexamples, experimental evidence given by years of application, and up-to-date security offered by these methods.
4 Conclusion
Similarly to the above example, the main thesis of this Dissertation is not yet backed up by a rigorous proof, because:
• Conjecture from Section 2.7, stating that proposed approach can bridge the art of visual EEG analysis and reproducible methods of signal processing, needs to be confirmed on a larger number of cases, and by independent opinions of the masters of the art of electroencephalography.
• A complete model, which would explain in details the generation of all the relevant aspects of the EEG time series, still does not exist. Without a strict definition of the word relevant in this context, we cannot construct a rigorous proof of conjectures from Sections 2.3 and 2.4, which relate to the relevant content of the EEG time series.
Up to now, all the experiments and simulations confirm the main thesis. Its acceptance and wide implementation will bring a direct application of the indispensable knowledge gathered in decades of visual analysis and a significant improvement of the efficiency of EEG-related research. Hopefully, this may contribute to the renaissance of electroencephalography in research and clinical applications.
Acknowledgements
This work was partly supported by a grant of the State Committee for Scientific Research (KBN) to the Institute of Experimental Physics of Warsaw University.
==== Refs
Martonosi AN Animal electricity, Ca2+ and muscle contraction. A brief history of muscle research Acta Biochimica Polonica 2000 47 493 516 11310955
Brazier MAB A History of the Electrical Activity of the Brain, The First Half-Century 1961 39 Parker Street, London W.C. 2: Pitman Medical Publishing
Berger H Über das Elektrenkephalogramm des Menschen Arch f Psychiat 1929 87 527 570
Gibbs FA Davis H Changes in the human electroencephalogram associated with loss of consciousness Am J Physiol 1935 113 49 50
Loomis AL Harvey EN Hobart GA Potential rhythms of the cerebral cortex during sleep Science 1935 82 198 200 17844579
Dawson GD Cerebral responses to electrical stimulation of peripheral nerve in man J Neurol Neurosurg Psychiatry 1947 10 134 140
Dawson GD A summation technique for the detection of small evoked potentials Electroencephalogr and clinical neurophysiology 1954 6 65 84 10.1016/0013-4694(54)90007-3
Dietsch G Fourier Analyse von Elektroenzephalogrammen des Menschen Pflügers Arch Ges Physiol 1932 230 106 112
Niedermayer E Historical Aspects Electroencephalography: Basic Principles, Clinical Applications and Related Fields 1999 fourth Williams & Wilkins 1 14
Lopes Da Silva F EEG Analysis: Theory and Practice Electroencephalography: Basic Principles, Clinical Applications and Related Fields 1999 fourth Williams & Wilkins 1135 1163
Rechtschaffen A Kales A (Eds) A manual of standardized terminology, techniques and scoring system for sleep stages in human subjects 1968 No. 204 in National Institutes of Health Publications, Washington DC: US Government Printing Office
Durka PJ Blinowska KJ A Unified Time-Frequency Parametrization of EEG IEEE Engineering in Medicine and Biology Magazine 2001 20 47 53 11668896 10.1109/51.956819
Mallat S A wavelet tour of signal processing 1999 second New York: Academic Press
Durka PJ Ircha D Neuper C Pfurtscheller G Time-frequency microstructure of event-related desynchronization and synchronization Med Biol Eng Comput 2001 39 315 321 11465886
Durka P Finding significant correlates of conscious activity in rhythmic EEG EURASIP Journal on Applied Signal Processing 2005 Trends in Brain-Computer Interfaces. [(accepted)].
Davis G Adaptive Nonlinear Approximations PhD thesis 1994 New York University
Harel D Algorithmic: The Spirit of Computing 1992 2 Addison-Wesley
Mallat S Zhang Z Matching Pursuit with time-frequency dictionaries IEEE Transactions on Signal Processing 1993 41 3397 3415 10.1109/78.258082
Durka PJ Ircha D Blinowska KJ Stochastic time-frequency dictionaries for matching pursuit IEEE Tran Signal Process 2001 49 507 510 10.1109/78.905866
DeVore RA Temlyakov VN Some remarks on greedy algorithms Advances in Computational Mathematics 1996 5 173 187
Chen SS Donoho DL Saunders MA Atomic decomposition by basis pursuit SIAM Review 2001 43 129 159 10.1137/S003614450037906X
Durka P From wavelets to adaptive approximations: time-frequency parametrization of EEG BioMed Eng Online 2003 2 1 12605721 10.1186/1475-925X-2-1
Durka P Adaptive time-frequency parametrization of epileptic EEG spikes Physical Review E 2004 69
Niedermayer E Lopes Da Silva F Electroencephalography: Basic Principles, Clinical Applications and Related Fields 1999 fourth Williams & Wilkins
Durka P Szelenberger W Blinowska K Androsiuk W Myszka M Adaptive time-frequency parametrization in pharmaco EEG Journal of Neuroscience Methods 2002 117 65 71 12084565 10.1016/S0165-0270(02)00075-4
Jouny CC Franaszczuk PJ Bergey GK Characterization of epileptic seizure dynamics using Gabor atom density Clin Neurophysiol 2003 114 426 437 12705423 10.1016/S1388-2457(02)00344-9
Durka P Żygierewicz J Klekowicz H Ginter J Blinowska K On the statistical significance of event-related EEG desynchronization and synchronization in the time-frequency plane IEEE Transactions on Biomedical Engineering 2004 51 1167 1175 15248533 10.1109/TBME.2004.827341
Franaszczuk PJ Bergey G Durka PJ Eisenberg H Time-frequency analysis using the Matching Pursuit algorithm applied to seizures originating from the mesial temporal lobe Electroencephalogr Clin Neurophysiol 1998 106 513 521 9741751 10.1016/S0013-4694(98)00024-8
Żygierewicz J Blinowska KJ Durka PJ Szelenberger W Niemcewicz S Androsiuk W High resolution study of sleep spindles Clin Neurophysiol 1999 110 2136 2147 10616119 10.1016/S1388-2457(99)00175-3
| 15748284 | PMC1160515 | CC BY | 2021-01-04 16:37:37 | no | Biomed Eng Online. 2005 Mar 4; 4:15 | utf-8 | Biomed Eng Online | 2,005 | 10.1186/1475-925X-4-15 | oa_comm |
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1597193410.1371/journal.pmed.0020129Correspondence and Other CommunicationsBioethicsEpidemiology/Public HealthHealth PolicyCommunication in Health CareEthicsMedicine in Developing CountriesPublic HealthSociologyMedical Decision Making: The Family–Doctor–Patient Triad CorrespondenceAslam Fawad
1
Aftab Omar
1
Janjua Naveed Z
2
1Aga Khan University Medical College, Karachi, Pakistan2Aga Khan University, Karachi, PakistanE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
6 2005 28 6 2005 2 6 e129Copyright: © 2005 Aslam et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
==== Body
The importance of a person-centred approach and the intricacies of risk communication have recently been well described in PLoS Medicine [1,2]. The applicability of the patient-centred approach to Eastern countries, however, has cultural, religious, and practical impediments that demand careful consideration. The bulk of the world population lives outside the United States and western Europe. Unlike in the West where the patient takes centre stage by both tradition and law, the family–doctor–patient triad is the norm in Eastern states, in general, and Pakistan in particular [3–8].
Pakistan is a predominantly Muslim country of 150 million people. About half the population is uneducated, and more than a third lives below the poverty line. There is one doctor for every 1,432 patients, compared to one doctor for every 390 patients in the US. The health-insurance system is virtually nonexistent, and there is no concept of assisted-care living, with the care of the elderly largely taking place at homes by their families. Strongly held religious beliefs and cultural views govern everyday life and dictate the roles of every member of the society. Families consist of well-knit, supportive, and collectively earning interdependent members who take mutual decisions on all matters pertaining to life and death [3,4]. The elder members of the family command the greatest respect and authority. The family unit is the functional unit of the society, the dynamics of which need attention and respect.
Strong family systems and the authoritative position of the doctor are the governing forces of medical decision making in these countries. Illiteracy, poverty, poor awareness of patient's rights, and a lack of accountability for physicians are factors conducive to such a practice. With this background, the role of the patient is limited. Health expenditure is borne by the family, giving it a central role in decision making. The concept of the financial survival of the family is a harsh reality [3,4]. The health-care costs of one seriously ill member may jeopardize the survival of others by draining the limited resources. Due to familial, moral, and monetary support, the patient relinquishes the responsibility of decision making and gives the primary role to the family or the doctor. Women, for example, may not give consent unless they get approval from their spouses [5]. In the case of women, who may have a lesser say in the patriarchal family system, the doctor should strive for active participation. Sometimes the family aims to protect the patient from stress by withholding information, and in terminal illnesses, the doctor and family act in concert to conceal complete information from the patient. They, for example, may not mention the word cancer to patients who have malignancies [9].
In contrast to Western practice, the role of the doctor is authoritative. Doctors are regarded as instruments of God and given the final authority in decision making [3,4,10]. In such circumstances, doctors are likely to take decisions unilaterally. When they do involve patients in decision making, physicians accept the centrality of families, with some considering patients and families as one [5]. One worry regarding communication of harm is of losing patients to other physicians with a more reassuring “nothing will go wrong” attitude [5]. It is also said that more time and patience are required to explain things to the illiterate. It is perhaps impractical, therefore, to expect overworked and underpaid physicians to practice risk communication according to the book.
Thus, the concept of individual centrality that is so elementary in the West stands challenged in the East. Research is needed to formulate appropriate strategies of risk communication. Areas needing research include the patient's concept of autonomy, the role of the family as perceived by patients and doctors, the existing practices of medical decision making, and the training of doctors in communicating risk.
An economically sound and literate population, properly trained doctors, and institutional checks and balances are essential prerequisites for establishing decision making with parity of partners. The need is to find a middle ground where not only the family unit is respected, but the patient also plays a proactive role. A dynamic balance between cultural values of caring and the possibility of a more individualistic role in health care is needed and is, indeed, attainable. Doctors, being the most influential component of the family–doctor–patient triad, can play a significant role in bringing about this change.
Citation: Aslam F, Aftab O, Janjua NZ (2005) Medical decision making: The family–doctor–patient triad. PLoS Med 2(6): e129.
==== Refs
References
Alaszewski A A person-centred approach to communicating risk PLoS Med 2005 2 e41 10.1371/journal.pmed.0020041 15736997
Herxheimer A Communicating with patients about harms and risks PLoS Med 2005 2 e42 10.1371/journal.pmed.0020042 15736998
Moazam F Families, patients, and physicians in medical decision making: A Pakistani perspective Hastings Cent Rep 2000 30 28 37
Moazam F Reconciling patients' rights and God's wisdom: Medical decision making in Pakistan Responsive Community 2001 11 43 51 12374145
Jafarey AM Farooqui A Informed consent in the Pakistani milieu: The physician's perspective J Med Ethics 2005 31 93 96 15681673
Cong Y Doctor-family-patient relationship: The Chinese paradigm of informed consent J Med Philos 2004 29 149 178 15371185
Fetters MD The family in medical decision making: Japanese perspectives J Clin Ethics 1998 9 132 146 9750985
Younge D Moreau P Ezzat A Gray A Communicating with cancer patients in Saudi Arabia Ann N Y Acad Sci 1997 809 309 316 9103582
Holland JC Geary N Marchini A Tross S An international survey of physician attitudes and practice in regard to revealing the diagnosis of cancer Cancer Invest 1987 5 151 154 3607572
Kanabar P Doctor-patient partnership J Indian Med Assoc 2002 100 718
| 15971934 | PMC1160566 | CC BY | 2021-01-05 10:39:49 | no | PLoS Med. 2005 Jun 28; 2(6):e129 | utf-8 | PLoS Med | 2,005 | 10.1371/journal.pmed.0020129 | oa_comm |
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1597193510.1371/journal.pmed.0020130PerspectivesIntensive CareRespiratory MedicineRespiratory MedicineCritical Care / Intensive CareEmergency MedicineHooray for Hypoxia? PerspectivesBellingan Geoff Geoff Bellingan is Clinical Director, Department of Critical Care, University College Hospitals London and Senior Lecturer in Intensive Care Medicine, Centre for Respiratory Research, Rayne Institute, University College London, United Kingdom. E-mail: [email protected]
Competing Interests: The author declares that he has no competing interests.
6 2005 28 6 2005 2 6 e130Copyright: © 2005 Geoff Bellingan.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.The mortality of critically ill patients rises steadily as the partial pressure of arterial oxygen falls below about 11 kPa (80 mm Hg). A new animal study in the May 2005 issue of PLoS Biology showing a potential benefit for hypoxia is thus a challenge to current thinking.
A new study comes to a counter-intuitive conclusion: hypoxia may have its benefits for patients with lung injury
==== Body
The Risks of Hypoxia and Hyperoxia
Hypoxaemia kills. Recently we have shown, in a study of over 50 000 critically ill patients, that mortality rises steadily as the partial pressure of arterial oxygen (PaO2) falls below about 11 kPa (80 mm Hg) (Figure 1) [1]. Rats exposed to hypoxic conditions develop a mild inflammatory response with increased capillary leak and pro-inflammatory cytokine release [2]. Hypoxia increases leukocyte adhesion, prolongs neutrophil survival, enhances pro-inflammatory responses (in particular, increasing leukocyte IL-8 expression), and reduces lung fluid transport [3,4]. A new study by Thiel et al. in PLoS Biology that shows a potential benefit for hypoxia is therefore a challenge to current thinking [5].
Figure 1 Mortality Rises Steadily as PaO2 Falls
To put this finding into context we must acknowledge the equally accepted fact that hyperoxia is toxic. Experiments in primates show that 100% oxygen results in progressive damage to the pulmonary endothelium and epithelium. It causes free radical release, capillary leak, and impaired surfactant function in addition to arteriolar vasoconstriction and maldistribution of microcirculatory perfusion [6,7].
What, then, is the clinical balance we should strike between these extremes? Oxygenation of patients with acutely impaired lung function saves lives, but what is the right amount of oxygen? Many clinicians use the oxygen dissociation curve, aiming for saturations of above 92%. Others highlight the fact that patients with severe respiratory failure can survive despite being mechanically ventilated for days on 100% oxygen. In other words, a high percentage of inspired oxygen itself may not be harmful. (In animal studies showing harm from giving 100% oxygen, the lungs are usually normal at the beginning of the study, so the arterial oxygen is correspondingly high—typically a PaO2 of 40 kPa (300 mm Hg); thus, it may be the high PaO2 rather than the high percentage of inspired oxygen that causes lung damage.) Some researchers question the importance of oxygen toxicity at all—for example, Meier et al., using a haemorrhagic shock model, found a significant improvement in outcome from the addition of 100% oxygen to the resuscitation [8].
The New Study in Mice: Hyperoxia May Enhance Tissue Damage
It is against this background that the study by Thiel et al. is so interesting. Two of the study's authors, Ohta and Sitkovsky, showed recently that adenosine receptors are important in down-regulating inflammation [9]. They argued that a negative feedback loop operates: inflammation causes tissue damage, leading to adenosine release, which acts on G-protein-coupled adenosine A2A receptors (A2ARs) to increase intracellular cyclic AMP, which in turn reduces the activity of NFκB and down-regulates the inflammatory response (Figure 2). Indeed, inflammation was dramatically enhanced in A2AR-deficient mice compared with wild-type controls. Importantly, no significant compensatory changes in the expression of other adenosine receptors have been demonstrated in these A2AR-deficient mice.
Figure 2 Adenosine Receptors Down-Regulate Inflammation through a Negative Feedback Loop
Sitkovsky and colleagues postulate that inflammation causes tissue damage leading to adenosine release, which acts on adenosine A2ARs to increase intracellular cyclic AMP. The rise in cyclic AMP down-regulates the inflammatory response. Tissue inflammation induces local hypoxia, which regulates the adenosine-driven protective response. The figure shows the steps in the pathway where hyperoxia might act (having an inhibitory effect).
(Illustration: Giovanni Maki)
Sitkovsky and colleagues went on to propose that tissue inflammation induced local hypoxia and that this hypoxia was important in regulating the adenosine-driven protective response. As a corollary they postulated that hyperoxia interrupts the hypoxia→adenosine→A2AR lung protection pathway and hence could enhance tissue damage (Figure 2). The current paper by Sitovsky's group in PLoS Biology examines this hypothesis [5].
At the outset, using an inflammatory challenge that significantly impaired gas exchange, the authors confirmed that when mice were additionally exposed to 100% oxygen, the toxicity of the challenge was dramatically increased compared to the inflammatory challenge in normoxic conditions (21% oxygen). (Over the time frame of the experiments high concentrations of oxygen alone were found to be only mildly inflammatory.) Inflammatory neutrophils from the challenged mice expressed the A2AR, and exposure to high concentrations of oxygen reduced this expression. Moreover, an adenosine agonist could block pro-inflammatory responses.
Mice lacking the A2AR had increased neutrophil infiltration in the lungs after endotoxin challenge and poorer gas exchange. The same was seen in wild-type mice treated with an A2AR blocker. Endotoxin-challenged mice breathing 10% oxygen (hypoxia) had a low mortality and, in the survivors, lung inflammation was less severe than in mice breathing room air. Hypoxia also significantly reduced pulmonary neutrophil accumulation and improved gas exchange compared with normoxia, suggesting that hypoxia acted to protect the lung from additional inflammatory damage. Importantly, hypoxia was also associated with elevated adenosine concentrations.
In contrast to the 20% mortality of hypoxic, endotoxin-challenged wild-type mice, hypoxic, endotoxin-challenged A2AR-deficient mice had elevated local and systemic accumulation of pro-inflammatory cytokines, and all died. As hypoxia per se did not affect the A2AR-deficient mice, this suggests that both hypoxia and A2AR functionality are required to induce the anti-inflammatory response. Finally, the authors showed that an adenosine agonist also inhibited lung injury in endotoxin-treated mice, reducing leukocyte influx and capillary leak. Importantly, this agonist could dramatically improve mortality even with a severe two-hit inflammatory challenge under hyperoxic conditions.
Implications of the Study
These are exciting and provocative results. They are, however, hard to reconcile with many studies showing pro-inflammatory effects from hypoxia. Sitkovsky and colleagues attempt to reconcile this difference by saying that their study had a long time course (over 48 hours), whereas studies showing that hypoxia is harmful usually had much shorter study periods.
The authors' explanation is probably too simplistic, and many issues and questions remain. (1) Different challenges—for example, pulmonary versus extrapulmonary—might give different results. (2) There is emerging evidence that more prolonged hypoxia can affect many processes, for example, mitochondrial function, and such complicated metabolic changes may be important and need to be examined [10]. (3) We don't know the mechanism of increased extracellular adenosine. (4) Lung injury may relate to tachypnoea and tidal volume, and these have not been measured. (5) In sampling arterial oxygen 15 minutes after removal from the hyperoxia, we do not know whether the animals were exposed to excessive PaO2 levels or whether the degree of lung injury prevented the animals from having excessive PaO2 levels. In other words, although the animals were given 100% oxygen, they may not have had an elevated PaO2 (the lung injury may have prevented oxygenation of the blood). This issue is critically important, as the use of high-percentage inspired oxygen simply to achieve a normal PaO2 is much more akin to clinical practice.
Refreshingly, though, this is a study that invites us to re-evaluate our perspectives. We know hypoxia can be bad, but is it as bad as hyperoxia? The scenario is all too familiar—the patient's lung injury has rendered the patient hypoxaemic, and the paramedics have started high-flow oxygen. In the emergency room the team are heartened to see the patient “pink up” and to hear the ominous sound of the saturation monitor abate. But, beware, is the team really in a comfort zone? Is the saturation monitor really now chirping happily with 100% oxygen saturations, or is this the harbinger of future problems as the PaO2 increases unnoticed? Should we have an additional set of alarm limits, hitherto unrecognised: those of high oxygen tensions? Should the monitors be alarming not only when arterial saturation is below 90% but also when it's above 98%? The study by Thiel et al. suggests that we urgently need to find out.
Citation: Bellingan G (2005) Hooray for hypoxia? PLoS Med 2(6): e130.
Abbreviations
A2ARadenosine A2A receptor
PaO2partial pressure of arterial oxygen
==== Refs
References
Bellingan G Wunsch H Young D Rowan K Is there a difference in mortality between patients admitted to ICU/HDU with acute hypoxaemic respiratory failure from pulmonary and extra-pulmonary causes: A study of 57706 patients [abstract]? Am J Respir Crit Care Med 2004 169 A17
Madjdpour C Jewell UR Kneller S Zeigler U Schwendener R Decreased alveolar oxygen induces lung inflammation Am J Physiol Lung Cell Mol Physiol 2003 284 L360 L367 12388372
Hirani N Antonicelli F Streiter RM Weisner MS Ratcliffe PJ The regulation of interleukin 8 by hypoxaemia in human macrophages—A potential role in the pathogenesis of the acute respiratory distress syndrome (ARDS) Mol Med 2001 7 685 697 11713368
Hannah S Mecklenburgh K Rahman I Bellingan GJ Greening A Hypoxia prolongs neutrophil survival in vitro FEBS Lett 1995 372 233 237 7556675
Thiel M Chouker A Ohta A Jackson E Caldwell C Oxygenation inhibits the physiological tissue-protecting mechanism and thereby exacerbates acute inflammatory lung injury PLoS Biol 2005 2 e174 10.1371/journal.pbio.0030174
Welty-Wolf KE Simonson SG Huang YC Kantrow SP Carraway MS Aerosolized manganese SOD decreases hyperoxic pulmonary injury in primates. II. Morphometric analysis J Appl Physiol 1997 83 559 568 9262453
Tsai AG Cabrales P Winslow RM Intaglietta M Microvascular oxygen distribution in awake hamster window chamber model during hyperoxia Am J Physiol Heart Circ Physiol 2003 285 H1537 H1545 12805029
Meier J Kemming GI Kisch-Wedel H Blum J Pape A Hyperoxic ventilation reduces six hour mortality after partial fluid resuscitation from hemorrhagic shock Shock 2004 22 240 247 15316394
Ohta A Sitkovsky M Role of G-protein-coupled adenosine receptors in down regulation of inflammation and protection from tissue damage Nature 2001 414 916 920 11780065
Frost MT Wang Q Moncada S Singer M Hypoxia accelerates nitric oxide-dependent inhibition of mitochondrial complex I in activated macrophages Am J Physiol Regul Integr Comp Physiol 2004 288 R394 R400 15486095
| 15971935 | PMC1160567 | CC BY | 2021-01-05 10:39:51 | no | PLoS Med. 2005 Jun 28; 2(6):e130 | utf-8 | PLoS Med | 2,005 | 10.1371/journal.pmed.0020130 | oa_comm |
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1597193610.1371/journal.pmed.0020153Research ArticleNeurosciencePsychologyOtolaryngologyNeurologyOtolaryngology/ENTTinnitus Perception and Distress Is Related to Abnormal Spontaneous Brain Activity as Measured by Magnetoencephalography Tinnitus and Spontaneous Brain ActivityWeisz Nathan
1
*Moratti Stephan
1
Meinzer Marcus
1
Dohrmann Katalin
1
Elbert Thomas
1
1Department of PsychologyUniversity of KonstanzGermanyHazell Jonathan Academic EditorTinnitus and Hyperacusis CentreUnited Kingdom
Competing Interests: The authors have declared that no competing interests exist.
Author Contributions: NW designed the study. NW, SM, MM, KD, and TE analyzed the data. NW, KD, and TE contributed to writing the paper.
*To whom correspondence should be addressed. E-mail: [email protected] 2005 28 6 2005 2 6 e15329 11 2004 6 4 2005 Copyright: © 2005 Weisz et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Brain Activity and Tinnitus
Background
The neurophysiological mechanisms underlying tinnitus perception are not well understood. Surprisingly, there have been no group studies comparing abnormalities in ongoing, spontaneous neuronal activity in individuals with and without tinnitus perception.
Methods and Findings
Here, we show that the spontaneous neuronal activity of a group of individuals with tinnitus (n = 17) is characterised by a marked reduction in alpha (8–12 Hz) power together with an enhancement in delta (1.5–4 Hz) as compared to a normal hearing control group (n = 16). This pattern was especially pronounced for temporal regions. Moreover, correlations with tinnitus-related distress revealed strong associations with this abnormal spontaneous activity pattern, particularly in right temporal and left frontal areas. Overall, effects were stronger for the alpha than for the delta frequency band. A data stream of 5 min, recorded with a whole-head neuromagnetometer under a resting condition, was sufficient to extract the marked differences.
Conclusions
Despite some limitations, there are arguments that the regional pattern of abnormal spontaneous activity we found could reflect a tinnitus-related cortical network. This finding, which suggests that a neurofeedback approach could reduce the adverse effects of this disturbing condition, could have important implications for the treatment of tinnitus.
Regional patterns of abnormal brain activity identified in people with tinnitus suggest the presence of a tinnitus- related cortical network that it may be possible to modify by neurofeedback.
==== Body
Introduction
Tinnitus is characterised by the perception of sounds (e.g., a tone, hissing, or roaring noise, and sometimes combinations of such perceptions) in the absence of any objective physical sound source. For the affected individual, this condition often causes a considerable amount of distress. Although it is now widely accepted that the generation of tinnitus has a central basis [1–3], the neurophysiological mechanism is still not well understood. Typically, tinnitus is accompanied by an audiometrically measurable hearing loss, and the considerable overlap between hearing loss and tinnitus spectra [4] suggests that these phenomena may be related. Thus, a widespread assumption is that neuronal response properties in the auditory system change following damage to it. These changes may persist even after recovery from the peripheral lesion. However, there is no general agreement as to which changes are responsible for the perception of this auditory phantom phenomenon. Studies in humans and animals indicate that neurons in the deprived regions of the auditory cortex change both their receptive field [5–8] and their spontaneous activity [1,9,10]. Furthermore, recent positron emission tomography studies in particular have pointed to the brain areas involved in attentional and emotional regulation [11–14]. An example of such a neuroimaging study was conducted by Lockwood et al. [14], who investigated individuals with tinnitus who were able to enhance or reduce the perceived loudness of their tinnitus via oral–facial movements. Besides changes in auditory cortical activity contralateral to the affected ear, the authors reported changes in hippocampal activity related to loudness changes. The authors interpreted this as evidence for limbic system involvement. In addition to this study implicating the limbic system (and temporal) structures, studies by Mirz et al. [11,12] indicate that frontal areas have a role in tinnitus.
Surprisingly, even though spontaneous activity has been a frequent research target in animal models of tinnitus, studies in humans have been rare. Data from a group study for abnormal activity of the cochlear nerve were reported by Martin [15]. An abnormal peak in spectral power could be observed close to 200 Hz in a majority of participants undergoing cerebellopontine angle surgery, a majority of whom had tinnitus (in the present study, however, we mean cortical activity when speaking of spontaneous activity). When extracted from spontaneous magnetoencephalography (MEG)–produced images in perisylvian regions, the values of the largest Lyapunov exponent—a measure for the predictability of the time series—were generally larger for the participants with tinnitus than for control individuals, indicating different nonlinear temporal dynamics in spontaneous activity for tinnitus and control participants [16]. A further study aiming to use information offered by spontaneous ongoing brain rhythms was published by Shulman and Goldstein, who reported temporal and frontotemporal changes (increases and decreases) of relative power and coherence irregularities in severely disabled individuals with tinnitus [17]. Otherwise, attempts to study tinnitus in humans have focussed on the use of designs that measure neurophysiologic responses following sounds [8,18–23] or experimental manipulations that enhance or reduce the perceived loudness [13,14,24,25]. In the present study, we examined the power spectrum of neuromagnetic oscillatory activity during rest.
The generation of tinnitus, in most cases, can be linked to damage to the auditory system, usually to receptors of the inner ear [1–3], probably even in cases where an impairment cannot be assessed audiometrically [26]. In this sense, one can speak of the presence of a deafferentation in the system since certain areas of the brain are now deprived from their normal input. Various sources of evidence indicate that deprivation of primary input leads to the functioning of the system in a slow-wave modus, i.e., analysis of neuroelectric signals in the frequency domain reflect an enhanced power in the delta frequency range (< 4 Hz). A dramatic example of this fact is the presence of neurological damage, e.g., following an infarct or a tumor. In these cases, enhanced slow-wave activity can be detected in perilesional areas [27–31]. Another example is slow-wave sleep [32,33], during which thalamic (and thus also cortical) centers are cut off from input from prethalamic relays. This condition leads to a hyperpolarisation of thalamocortical cells [34] that activates sodium and potassium currents, gradually depolarising the cell. The depolarisation in turn triggers a calcium-mediated low-threshold spike burst. The frequency of this hyperpolarisation–spike-burst cycle is approximately in the delta to theta frequency range. Interestingly, hyperpolarisation-dependent low-threshold spike bursts in the thalamus have previously been associated with positive symptoms in various pathologic conditions including tinnitus [35,36]. Because of the corticothalamic connections, coherent slow-wave oscillations are also seen on a cortical level [36]. In the case of tinnitus, one would expect to find them within the temporal lobe. In the electroencephalogram, much of this activity would project to frontal sites because of the orientation of the sources in the auditory cortex, where it gets blurred as activity from other regions, especially radially oriented frontal sources, also contribute to and even dominate the electroencephalogram at this site. MEG recordings can be used to differentiate between radial frontal and tangential temporal sources, as MEG is largely blind to the former but not to the latter sources. Studies from slow-wave sleep also show that an enhancement of oscillatory activity in the delta band is accompanied by a reduced activity in several other frequency bands—among others, alpha, sigma, and beta [32,37]. Next to enhanced delta activity, attenuated alpha activity has been implicated in schizophrenia [38,39]. These results indicate that a change in the delta range might not be isolated phenomenon, but might instead be accompanied by reciprocal alterations in other frequency bands.
The data recorded by MEG represent the summed activity of tangential sources on a two-dimensional (sensor space) level. By applying various inverse solution strategies [40], it is possible to obtain information about potential (usually cortical) generators underlying the observed data. Using power spectrum analysis of such source-space-transformed neuromagnetic data the primary aim of the present study was to identify cortical regions with changes in underlying spontaneous activity patterns in individuals with tinnitus. Based on previous studies—summarised above—we predicted enhanced power in the delta frequency range. A second aim was to explore whether such an enhancement would also occur with a corresponding change in a different frequency band. Relations between the neurophysiologic scores and the subjective reports of tinnitus-related distress were examined by means of correlation analyses.
Methods
Participants
Seventeen individuals with chronic tinnitus (one woman; mean age ± standard error, 52.41 ± 2.70)—16 with high-frequency hearing loss and one with deafness due to transection of the hearing nerve—and 16 normal hearing control individuals (one woman; age 45.88 ± 3.84) participated in the study. Five minutes of resting MEG were measured and analysed with regard to the spectral content (see “Data Acquisition and and Signal Analysis” in Supporting Information for details). Tinnitus was reported to be bilaterally equal in four individuals, bilateral but left-dominant in one individual, and right-dominant in two individuals. Ten individuals reported unilateral tinnitus, of which eight indicated that they heard their tinnitus on the left side. The majority of individuals for which etiology is known experienced either a sudden hearing loss or noise trauma (assessed using a structured tinnitus interview). A comprehensive overview of all tinnitus participants is given in Table 1. Prior to the experiment, participants gave written informed consent. Tinnitus-related distress was assessed with a standard German questionnaire (Tinnitus Fragebogen [41], an adaption of the commonly used Tinnitus Questionnaire [42]), which allows tinnitus distress to be ranked on the following subscales: emotional distress, cognitive distress, intrusiveness, hearing problems, sleeping problems, and somatic complaints. In addition to the total score (sum of scores on each subscale), the following subscales were tested against the neurophysiological data separately: emotional distress, cognitive distress, intrusiveness, and sleeping problems. The last subscale, together with the regional slow-wave distribution, should help to rule out the hypothesis that changes in frequency bands could be a trivial result of sleep disturbances. Based on the total score, participants were assigned to a certain distress category: slight (0–30 points), moderate (31–46), severe (47–59), and very severe (60–84) distress. The greatest contributions (approximately 47%) to the total distress score (mean score ± standard error, 30.0 ± 4.69) were from emotional (7.8 ± 1.4) and cognitive distress (6.1 ± 0.9) and intrusiveness (approximately 27%; 8.0 ± 1.0)—i.e., factors that involve the perception and psychological effects of tinnitus more directly. Thus, 75% of the total score can be attributed to these three scales. Only 15% of the total distress score was attributed to reported hearing problems.
Table 1 Participant Information
a Note that sometimes the participants' assumption is given together with an unknown etiology.
b Values in parentheses are total scores on the Tinnitus Questionnaire. Bil., bilateral, dom., dominant; HL, hearing level. DOI: 10.1371/journal.pmed.0020153.t001
Statistics
We employed repeated-measures analysis of variance (assuming compound symmetry; this assumption was validated and superior to a general correlation structure) to test for the presence of main and interaction effects (group × frequency band) in the overall average for each frequency. Since the result of this analysis was significant, we executed this analysis for specified regions (see above) separately. Associations with tinnitus-related distress were tested using product-moment correlations. In a first step, the average over all dipoles for each frequency band was correlated with the tinnitus questionnaire. As this analysis yielded significant associations that were specific and consistent in direction for the alpha and delta band, values from each region were correlated with the scales of the questionnaire in order to help interpret effects gained in the first step.
Results
Spontaneous Neuronal Activity Pattern
The explorative analysis of the frequency spectrum of the recorded magnetic fields revealed characteristic changes in participants with tinnitus from the usual pattern seen in normal hearing controls (Figure 1). The alpha peak, which can be seen in the control group, is strongly reduced in the tinnitus group. In contrast, the power in the delta frequency band is considerably enhanced. This pattern is spatially resolved by the results obtained by the minimum norm solution, which attributes both—the alpha attenuation and the delta enhancement—mostly to the temporal regions (Figure 2). It is clearly visible that the alpha reduction is considerably stronger than the delta enhancement. A repeated-measures analysis of variance reveals a significant group by frequency band interaction (F
2,62 = 4.47, p < 0.02, power 0.78). Differences between the values of each frequency range (alpha versus theta, alpha versus delta, and theta versus delta) were calculated for each individual, and the results were subjected to a between-subjects analysis of variance. This analysis shows that the control group exhibited a significantly higher alpha than theta power (F
1,31 = 5.84, p < 0.03), whereas the tinnitus group has stronger delta activity than alpha activity (F
1,31 = 5.11, p < 0.03). No differences were found between the groups when theta was compared with delta (F
1,31 < 1). These results statistically confirm the impression of a reduced alpha and enhanced delta peak in the tinnitus group. This interaction pattern was statistically significant for all regions, with the exception of right frontal and anterior parietal regions of both hemispheres. The foci of the altered spontaneous activity pattern in the tinnitus group are clearly found bilaterally in temporal areas (F
1,31 > 5, p < 0.01), which can be seen in Figures 2 and 3.
Figure 1 Power Spectra Averaged over All Sensors Show a Reduced Alpha Peak in Participants with Tinnitus and an Enhancement for Delta
The sharp peak centred at 16 2/3 Hz represents technical noise resulting from the 1-km-distant railway system.
Figure 2 Difference Maps between Participants with Tinnitus and Controls for Alpha and Delta
The results suggest that areas for which alpha reduction and delta enhancements are found partly overlap. Overall, the effect for the alpha band is considerably stronger (note that for this reason the scaling ranges are chosen differently).
Figure 3 Display of the Group × Frequency Band Interaction Effects Averaged over Temporal Sources
Effects for right (A) and left (B) temporal cortex, where the strongest enhancements of alpha and reductions of were found.
Association of Spontaneous Activity with Tinnitus-Related Distress
For both alpha and delta power, significant correlations were found with the Tinnitus Questionnaire (total score and subscales), with r ranging between 0.5 and 0.7 (Table 2). These correlations correspond to large effect sizes, as outlined by Cohen [43]. Note, however, that because of the large number of calculations and the increased risk of false positives, p-values help us to gain a more general impression of the present associations. The distribution of correlation coefficients for the total score on the Tinnitus Questionnaire versus alpha and delta power is shown in Figure 4. Overall, correlations for alpha appear slightly stronger than for delta. Yet the strongest effects are located in similar brain regions (note that correlations smaller than 0.4 are coloured white). To exclude the possibility that the high associations are a consequence of potentially different subgroups—i.e., one subgroup high in delta and another one with low alpha—a frequency index was calculated ([delta – alpha]/[delta + alpha]) for each individual. The correlation of this score with the Tinnitus Questionnaire is displayed in the bottom panel of Figure 4. Strongest associations with this measure are found in right temporal and left frontal regions. Results of the theta band did not correlate with any of the questionnaire data. Also, correlations between hearing-loss parameters (depth and slope of hearing loss) and tinnitus-related distress were not significant (p > 0.4 for all). Amount and slope of hearing loss were also not significantly correlated with neuromagnetic data (0.12 < r < 0.25).
Figure 4 Correlation Map between Alpha, Delta, and Tinnitus-Related Distress
Since previous analyses (see Figure 3) implicated corresponding areas for the effects found for alpha and delta, tinnitus-related distress was additionally correlated with a frequency index ([delta – alpha]/[delta + alpha]; bottom panel). Effects are largest for right temporal and left frontal sources.
Table 2 Correlations of Power with Tinnitus Questionnaire
Correlation coefficients between power in frequency bands and scores on the Tinnitus Questionnaire. Statistically significant correlations are in red. Note that significance level has not been corrected (see Methods for rationale). Significance assessed with t-test (df = 15). *, p < 0.05; **, p < 0.01.
a ant, anterior; occi, occipital; par, parietal; post fr., posteriofrontal; pos, posterior; pre fr., prefrontal; temp, temporal.
b Cog, Cognitive; Emo, Emotional; Intr, Intrusiveness; Sleep, Sleep Problems.
Discussion
To our knowledge, this is the first group study on tinnitus in humans (there have been reports on single cases [35,36]) to show marked relative enhancements in delta and concomitant reductions in alpha spontaneous cortical activity. Notable group differences are a bilateral relative enhancement in delta and an accompanying reduction in alpha power over temporal areas (extending into posterior regions in the right hemisphere) in individuals with tinnitus. These findings are in agreement with a study conducted by Shulman and Goldstein [17], who report abnormalities in spontaneous activity in all 21 investigated patients. The definition of abnormality, however, was rather broad, meaning enhancements, reductions, or coherence irregularities in different bands. Another difference from our study is that these authors investigated severely disabled patients, whereas only a minority (four out of 17; see Table 1) of our participants could be classified in a similar manner. Our results are in agreement with ideas put forward by Jeanmonod and Llinas and colleagues [35,36], who stated that positive symptoms are a consequence of a hyperpolarisation of thalamic neurons following deafferentation, leading to spike bursts around 4 Hz. A look at the power spectra in Figure 1 shows that a reduced alpha cannot be caused by normalisation, because in this case the peak is almost absent—a pattern we encounter frequently in tinnitus participants. Concerning delta, it is very unlikely that the enhancement relative to alpha is an effect of the reduced alpha, as enhancements should then not be restricted to slow oscillations but should also be seen in other frequency bands (e.g., beta; note, e.g., that the artefactual 16-Hz activation is identical for both groups). Apart from this, the general pattern (i.e., alpha reduction and delta enhancement) is also seen in the non-normalised data (available upon request). It is unclear whether this relative enhancement of delta activity compared to alpha is the “abnormal activity” that is perceived as tinnitus [44]. The underlying neural substrate of tinnitus perception could also be an edge effect, such as gamma activity that arises in regions between abnormal (delta) and normal awake activity [45]. The (methodological) problem in the latter case lies, however, in the exact definition of gamma activity. However, it would be interesting to investigate the effects of treatments presumably inducing tinnitus on slow waves in animals. The fact that regions with an increase in slow-wave activity are also the regions of decreased alpha activity resembles results found during slow-wave sleep [32] and supports the idea that the changes in spontaneous brain activity might be mediated by sensory deprivation, i.e., partial hearing loss in this case. Additional evidence is provided by the observation that delta enhancement and alpha reduction were strongly correlated with tinnitus-related distress variables with a focus on the right temporal and also left frontal cortex. The association of the abnormality index—which sets alpha and delta power in relationship to one another—and tinnitus-related distress demonstrates that the effects reported are not an epiphenomenon of normalisation. Furthermore, a deviant abnormality index can be seen at an almost individual level. We are thus confident that the main effect reported here is not due to methodological flaws concerning data analysis.
However, we can not exclude with certainty the possibility that the effects are not specific to tinnitus, a limitation of our study. One interesting finding is that compared to the normal hearing control group, the tinnitus group also had a high-frequency hearing loss. So, theoretically, the ideal control group would have exactly the same type of hearing loss without a tinnitus sensation. However, to find such a group of an appropriate sample size would be very difficult as both phenomena are strongly associated. For example, in an attempt to investigate the influence of high-frequency hearing loss (similar to that of our study) on cortical reorganisation, Dietrich et al. [8] noticed that all hearing-impaired participants also had tinnitus. Yet it cannot be disputed that occasionally individuals exist who have a similar kind of deficiency in clinical audiograms as the participants in our study, but without having a tinnitus sensation. Simply having a reduced threshold does not seem to be sufficient to trigger reorganisational processes leading to tinnitus. Why these individuals (although rare) do not develop this symptom is a very important question for understanding tinnitus. A possible way to tackle this question is to analyse underlying hearing damage in more detail (e.g., amount of inner and outer hair-cell damage) by comparing individuals with hearing loss with and without tinnitus. Another strategy would be to investigate possible predisposing factors, e.g., psychological or genetic factors. As parameters of hearing loss (depth and slope) appear to be uncorrelated with distress scores and neurophysiological data, we assume its distorting effects to be minimal. This hypothesis is further supported by the fact that only a small proportion of the distress score was attributed to hearing loss. Overall, it should be emphasised that the issue of hearing loss is less serious in our study than in studies in which individuals with tinnitus (either human or animals) are acoustically stimulated (especially in the hearing-loss range).
The association of our neurophysiologic data with distress suggests that the right temporal and left frontal cortex might be involved in a tinnitus-related cortical network, in which the temporal region is associated more with perceptual issues (i.e., aspects concerning the character of the sound, e.g., tonal or noise-like, loudness), and the left frontal region more with affective distress and motivational attention of tinnitus (i.e., the tinnitus becoming a signal of high importance, so that it draws attention of the individual). Without reference to lateralisation, Jastreboff describes the prefrontal cortex as a “candidate for the integration of sensory and emotional aspects of tinnitus” [44]. Our data lend support to this notion. Concerning the stronger effect for the right temporal area than for the left, one has to consider the higher frequency of left-sided tinnitus in our study. Thus, it cannot be excluded that this asymmetry would vanish if more individuals with right-sided tinnitus were included in the analysis. However, the fact that tinnitus is generally more common for the left ear [46] certainly opens up possibilities to speculate about potential asymmetries, either on a peripheral or central level, that could account for this finding. The underlying physiological reasons for this asymmetry on an epidemiological level have not been a matter of research so far. Left frontal activation has been linked with positive, and right frontal activation with negative, affect [47–49]. In the context of this framework, enhanced alpha (indicating hypoactivation) in the left frontal cortex should be indicative of depression. Our alpha–distress association, however, is negative, thus implying that the results obtained cannot be explained by an effect of higher depressive mood in individuals with tinnitus. This simple explanation would also not fit with recent data from our group [52], which demonstrated a negative association between left frontal delta (measured via dipole density) and depression.
In future studies, we hope to elucidate the function of the (left) frontal area, as it may point us to the role of top-down influences that presumably play a role in the perception and perhaps even generation of tinnitus. An important aspect of these results is that they potentially have clinical implications, which are currently being tested by our group. Reducing the abnormal spontaneous activity pattern reported in this study via neurofeedback might cause concomitant reductions in tinnitus distress and/or intensity (often reported to be unrelated; [51]). Besides experimentally demonstrating to what extent our results reflect a “genuine” signature of tinnitus sensation in ongoing brain activity, these investigations could ultimately be of benefit for persons affected by this condition.
Supporting Information
Data Acquisition and Signal Analysis
Five minutes of MEG under a resting condition was recorded (sampling rate: 678.17 Hz; 0.1–200 Hz band-pass) using a 148-channel neuromagnetometer (Magnes 2500 WH, Biomagnetic Technologies, San Diego, California, United States). The participant was requested to keep eyes open and to maintain gaze on a fixation mark at the ceiling of the recording chamber. Eye movements were monitored with four electrodes attached to the left and right outer canthi and above and below the right eye.
In the first step of data analysis, sampling points were reduced by a factor of ten. After noise reduction, eye-movement correction was undertaken using the algorithm proposed by Berg and Scherg [52]. As an explorative step, spectral power was calculated for each sensor via mean fast Fourier transformations (window length, 7.55 s; 50% overlap between cosine squared windows). This step was taken to focus on specific frequency bands of interest. As the emphasis of the study was on alterations of spontaneous activity patterns within the tinnitus group, data were scaled by dividing each value by the overall mean power (gained by averaging over all sensors). As can be seen in Figure 1, participants with tinnitus showed a markedly reduced alpha peak accompanied by an enhancement in the slow frequency (delta) range.
The second step consisted of the investigation of the underlying source activity via application of the minimum norm estimate [53,54] to the eye-movement-corrected continuous data. This linear estimation technique yields a solution for the current density of a configuration of dipoles (here: 197 evenly distributed) located on a spherical shell by multiplying the pseudo-inverse of the lead-field matrix (which describes the sensitivity of each sensor to the sources) with the obtained data. The lead-field matrix was computed for each participant, based on information about the centre of a fitted sphere to the digitised head shape, and the positions of the MEG sensors relative to the head. In order to be able to average the obtained minimum norm estimate solutions over different participants, we employed a shell with a fixed radius of 6 cm. The radius of 6 cm was chosen as a tradeoff between depth sensitivity and spatial resolution [53]. Each of the 197 dipole locations consisted of two perpendicular dipoles oriented tangentially to the shell surface. The source-space transformed continuous data were then entered into the same mean fast Fourier transformation algorithm as described above. For both orientations of each dipole, power was calculated in the following frequency bands: delta (1.5–4 Hz), theta (4–8 Hz), and alpha (8–12 Hz). In order to gain a single value for each dipole, the square root of the sum of squares of the power for the two orientations was calculated and scaled in the same manner as described above (using mean power of each dipole instead of each sensor). Oscillatory activity was then analysed (a) averaged over all 197 dipoles and (b) in specific regions by grouping clusters of dipoles on the sphere and averaging their values. These regions (always bilateral) were prefrontal, frontal, temporal, frontocentral, parietal (anterior and posterior), and occipital cortex.
Patient Summary
Background
Tinnitus—hearing sounds such as hissing or roaring that are not really present—is a common condition and can be very distressing. Little is know about why tinnitus starts. Though people who have it often have some hearing loss, the hearing loss does not cause the tinnitus. There are few good treatments for tinnitus.
What Did the Authors Do?
They measured spontaneous activity in the brains of 17 people with tinnitus and 16 without. The found that one type of brain activity—alpha—was decreased and another—delta—was increased in people with tinnitus, particularly in one part of the brain, the temporal region, where the brain processes sounds. The changes were particularly pronounced in people who were most distressed by the tinnitus.
What Do These Findings Mean?
It seems that tinnitus is associated with one type of abnormal spontaneous activity in the brain. One way of treating tinnitus therefore might be to try to alter the activity by providing feedback about the ongoing brain activity and training the individual to interfere with this activity, i.e., by attempting to enhance alpha activity and reduce delta activity.
Where Can I Get More Information?
MedlinePlus has a Web page on tinnitus: http://www.nlm.nih.gov/medlineplus/tinnitus.html
The Welcome Gateways has a selection of Web sites on tinnitus: http://omni.ac.uk/browse/mesh/D014012.html
The American Tinnitus Association also offers information on tinnitus: http://www.ata.org
This study was supported by a grant from the Deutsche Forschungsgemeinschaft (EL 101/20). The funder had no role in the study design, data analysis, decision to publish, or manuscript preparation and content. We thank Thomas Hartmann for help during data acquisition, and Horst Böttcher (ProAkustik) for providing a high-resolution audiometer.
Citation: Weisz N, Moratti S, Meinzer M, Dohrmann K, Elbert T (2005) Tinnitus perception and distress is related to abnormal spontaneous brain activity as measured by magnetoencephalography. PLoS Med 2(6): e153.
Abbreviations
MEGmagnetoencephalography
==== Refs
References
Eggermont JJ Central tinnitus Auris Nasus Larynx 2003 30 Suppl 7 12 12589843
Lockwood AH Salvi RJ Burkard RF Tinnitus N Engl J Med 2002 347 904 910 12239260
Moller AR Pathophysiology of tinnitus Otolaryngol Clin North Am 2003 36 249 266 12856295
Norena A Micheyl C Chery-Croze S Collet L Psychoacoustic characterization of the tinnitus spectrum: Implications for the underlying mechanisms of tinnitus Audiol Neurootol 2002 7 358 369 12401967
Rauschecker JP Auditory cortical plasticity: A comparison with other sensory systems Trends Neurosci 1999 22 74 80 10092047
Rajan R Irvine DR Neuronal responses across cortical field A1 in plasticity induced by peripheral auditory organ damage Audiol Neurootol 1998 3 123 144 9575381
Irvine DR Rajan R Brown M Injury- and use-related plasticity in adult auditory cortex Audiol Neurootol 2001 6 192 195 11694726
Dietrich V Nieschalk M Stoll W Rajan R Pantev C Cortical reorganization in patients with high frequency cochlear hearing loss Hear Res 2001 158 95 101 11506941
Norena AJ Eggermont JJ Changes in spontaneous neural activity immediately after an acoustic trauma: Implications for neural correlates of tinnitus Hear Res 2003 183 137 153 13679145
Elbert T Rockstroh B Kowalik ZJ Hoke M Angelieri F Butler SR Giaquinto S Majkowski J Clinical applications of EEG analyses using tools derived from non-linear system theory Analysis of the electrical activity of the brain 1997 London Wiley and Sons 45 67
Mirz F Gjedde A Sodkilde-Jrgensen H Pedersen CB Functional brain imaging of tinnitus-like perception induced by aversive auditory stimuli Neuroreport 2000 11 633 637 10718327
Mirz F Pedersen B Ishizu K Johannsen P Ovesen T Positron emission tomography of cortical centers of tinnitus Hear Res 1999 134 133 144 10452383
Lockwood AH Salvi RJ Coad ML Arnold SA Wack DS The functional anatomy of the normal human auditory system: Responses to 0.5 and 4.0 kHz tones at varied intensities Cereb Cortex 1999 9 65 76 10022496
Lockwood AH Salvi RJ Coad ML Towsley ML Wack DS The functional neuroanatomy of tinnitus: Evidence for limbic system links and neural plasticity Neurology 1998 50 114 120 9443467
Martin WH Vernon JA Møller AR Spectral analysis of brain activity in the study of tinnitus Mechanisms of tinnitus 1995 London Allyn and Bacon 163 180
Elbert T Rockstroh B Kowalik ZJ Hoke M Molnar M Chaotic brain activity Electroencephalogr Clin Neurophysiol Suppl 1995 44 441 449 7649054
Shulman A Goldstein B Quantitative electroencephalography: Preliminary report—tinnitus Int Tinnitus J 2002 8 77 86 14763215
Mühlnickel W Elbert T Taub E Flor H Reorganization of auditory cortex in tinnitus Proc Natl Acad Sci U S A 1998 95 10340 10343 9707649
Jacobson GP Calder JA Newman CW Peterson EL Wharton JA Electrophysiological indices of selective auditory attention in subjects with and without tinnitus Hear Res 1996 97 66 74 8844187
Jacobson GP Ahmad BK Moran J Newman CW Tepley N Auditory evoked cortical magnetic field (M100–M200) measurements in tinnitus and normal groups Hear Res 1991 56 44 52 1769924
Pantev C Hoke M Lutkenhoner B Lehnertz K Kumpf W Tinnitus remission objectified by neuromagnetic measurements Hear Res 1989 40 261 264 2793608
Hoke M Feldmann H Pantev C Lutkenhoner B Lehnertz K Objective evidence of tinnitus in auditory evoked magnetic fields Hear Res 1989 37 281 286 2708150
Weisz N Voss S Berg P Elbert T Abnormal auditory mismatch response in tinnitus sufferers with high-frequency hearing loss is associated with subjective distress level BMC Neurosci 2004 5 8 15113455
Giraud AL Chery-Croze S Fischer G Fischer C Vighetto A A selective imaging of tinnitus Neuroreport 1999 10 1 5 10094123
Lockwood AH Wack DS Burkard RF Coad ML Reyes SA The functional anatomy of gaze-evoked tinnitus and sustained lateral gaze Neurology 2001 56 472 480 11222790
Shiomi Y Tsuji J Naito Y Fujiki N Yamamoto N Characteristics of DPOAE audiogram in tinnitus patients Hear Res 1997 108 83 88 9213125
Kamada K Moller M Saguer M Ganslandt O Kaltenhauser M A combined study of tumor-related brain lesions using MEG and proton MR spectroscopic imaging J Neurol Sci 2001 186 13 21 11412866
Vieth JB Kober H Grummich P Sources of spontaneous slow waves associated with brain lesions, localized by using the MEG Brain Topogr 1996 8 215 221 8728406
de Jongh A de Munck JC Baayen JC Jonkman EJ Heethaar RM The localization of spontaneous brain activity: First results in patients with cerebral tumors Clin Neurophysiol 2001 112 378 385 11165544
Meinzer M Elbert T Wienbruch C Djundja D Barthel G Intensive language training enhances brain plasticity in chronic aphasia BMC Biol 2004 2 20 15331014
Elbert T Rockstroh B Bulach D Meinzer M Taub E [New developments in stroke rehabilitation based on behavioral and neuroscientific principles: Constraint-induced therapy] Nervenarzt 2003 74 334 342 12707702
Benoit O Daurat A Prado J Slow (0.7–2 Hz) and fast (2–4 Hz) delta components are differently correlated to theta, alpha and beta frequency bands during NREM sleep Clin Neurophysiol 2000 111 2103 2106 11090758
Gath I Bar-On E Classical sleep stages and the spectral content of the EEG signal Int J Neurosci 1983 22 147 155 6668131
Steriade M Central core modulation of spontaneous oscillations and sensory transmission in thalamocortical systems Curr Opin Neurobiol 1993 3 619 625 8219730
Jeanmonod D Magnin M Morel A Low-threshold calcium spike bursts in the human thalamus. Common physiopathology for sensory, motor and limbic positive symptoms Brain 1996 119 363 375 8800933
Llinas RR Ribary U Jeanmonod D Kronberg E Mitra PP Thalamocortical dysrhythmia: A neurological and neuropsychiatric syndrome characterized by magnetoencephalography Proc Natl Acad Sci U S A 1999 96 15222 15227 10611366
Uchida S Maloney T March JD Azari R Feinberg I Sigma (12–15 Hz) and delta (0.3–3 Hz) EEG oscillate reciprocally within NREM sleep Brain Res Bull 1991 27 93 96 1933440
Canive JM Lewine JD Edgar JC Davis JT Miller GA Spontaneous brain magnetic activity in schizophrenia patients treated with aripiprazole Psychopharmacol Bull 1998 34 101 105 9580382
Canive JM Lewine JD Edgar JC Davis JT Torres F Magnetoencephalographic assessment of spontaneous brain activity in schizophrenia Psychopharmacol Bull 1996 32 741 750 8993097
Elbert T Andrä W Nowak H Neuromagnetism Magnetism in medicine: A handbook 1998 Berlin Wiley-VCH 190 261
Goebel G Hiller W Tinnitus-Fragebogen (TF): Ein Instrument zur Erfassung von Belastung und Schweregrad bei Tinnitus 1998 Göttingen Hogrefe 90
Hallam RS Rachman S Hinchcliffe R Rachman S Psychological aspects of tinnitus Contributions to medical psychology 1984 Oxford Pergamon Press 31 53
Cohen J Statistical power analysis for the behavioral sciences, 2nd ed 1988 Hillsdale (New Jersey) Lawrence Erlbaum Associates 567
Jastreboff PJ Phantom auditory perception (tinnitus): Mechanisms of generation and perception Neurosci Res 1990 8 221 254 2175858
Llinas R Ribary U Contreras D Pedroarena C The neuronal basis for consciousness Philos Trans R Soc Lond B Biol Sci 1998 353 1841 1849 9854256
Lenarz T Feldmann H Epidemiologie 1992 Tinnitus. Stuttgart Thieme 71 75
Davidson RJ Affective style and affective disorders: Perspectives from affective neuroscience Cogn Emot 1998 12 307 330
Davidson RJ Irwin W The functional neuroanatomy of emotion and affective style Trends Cogn Sci 1999 3 11 21 10234222
Davidson RJ Pizzagalli D Nitschke JB Putnam K Depression: Perspectives from affective neuroscience Annu Rev Psychol 2002 53 545 574 11752496
Wienbruch C Moratti S Elbert T Vogel U Fehr T Source distribution of neuromagnetic slow wave activity in schizophrenic and depressive patients Clin Neurophysiol 2003 114 2052 2060 14580603
Henry JA Meikle MB Psychoacoustic measures of tinnitus J Am Acad Audiol 2000 11 138 155 10755810
Berg P Scherg M A multiple source approach to the correction of eye artifacts Electroencephalogr Clin Neurophysiol 1994 90 229 241 7511504
Hauk O Keil A Elbert T Muller MM Comparison of data transformation procedures to enhance topographical accuracy in time-series analysis of the human EEG J Neurosci Methods 2002 113 111 122 11772433
Hämäläinen M Hari R Ilmoniemi RJ Knuutila J Lounasmaa OV Magnetoencephalography—Theory, instrumentation, and applications to noninvasive studies of the working human brain Rev Mod Physics 1993 65 413 497
| 15971936 | PMC1160568 | CC BY | 2021-01-05 11:13:38 | no | PLoS Med. 2005 Jun 28; 2(6):e153 | utf-8 | PLoS Med | 2,005 | 10.1371/journal.pmed.0020153 | oa_comm |
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1597193710.1371/journal.pmed.0020154Learning ForumGastroenterology/HepatologyHematologyHematology (including Blood Transfusion)GastroenterologyA 61-Year-Old Man with Dyspepsia and Weight Loss Learning ForumBurton Catherine Catherine Burton is Clinical Research Fellow in the Department of Haematology, University College London, United Kingdom. E-mail: [email protected]
Competing Interests: The author declares that she has no competing interests.
6 2005 28 6 2005 2 6 e154Copyright: © 2005 Catherine Burton.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.In this case-based Learning Forum, Burton takes clinicians through the key steps in diagnosing and managing this patient.
==== Body
DESCRIPTION of CASE
A 61-year-old male carpenter presented with ten months' history of dyspepsia. More recently he had developed epigastric pain and had lost four kilograms in weight because of anorexia. He was otherwise asymptomatic and in particular there was no history of nausea, vomiting, abdominal distension, melaena, haematemesis, diarrhoea, constipation, night sweats, or fevers. There was no past medical history of note, and he had a performance status of zero (see Table 1). His only medication was lansoprazole. He drank no alcohol and smoked 40 cigarettes/day. Examination was unremarkable. All blood tests, including full blood count, biochemistry, and lactate dehydrogenase, were normal.
Table 1 World Health Organization Performance Status
What Was the Differential Diagnosis at This Point?
The differential diagnosis in a patient with dyspepsia, epigastric pain, and weight loss is gastro-oesophageal reflux, gastritis, gallstones, peptic ulcer, oesophageal/gastric carcinoma, lymphoma, and inflammatory bowel disease.
The patient underwent an endoscopy which showed a 4–5-cm ulcer with rolled edges on the greater curve of the antrum of the stomach. Appearances were suggestive of a malignant lesion.
What Was the Most Likely Diagnosis?
The likely diagnosis was therefore either gastric carcinoma or lymphoma—most likely non-Hodgkin lymphoma (NHL)—specifically, either gastric mucosa-associated lymphoid tissue (MALT) lymphoma or diffuse large B cell lymphoma (DLBCL).
Histology revealed a diagnosis of DLBCL (see Figures 1 and 2). The biopsy was positive for surface markers CD20 and CD79a and surface immunoglobulin (IgM). These markers are found on B cells, and positivity confirmed the B cell origin of the lymphoma (see Figures 3 and 4). The biopsy was negative for BCL-2 protein expression. Positive BCL-2 expression is associated with an adverse prognosis. Staining for Helicobacter pylori antibody was negative.
Figure 1 Low-Power Haematoxylin and Eosin Staining Showing DLBCL
DLBCL is composed of large transformed lymphoid cells, with oval to round nuclei with fine chromatin and membrane-bound nucleoli. Surface and/or cytoplasmic immunoglobulin (IgM > IgG > IgA) is demonstrated in 50%–75% of cases. BCL-2 is positive in approximately 30%–50% of cases and is associated with a poorer prognosis.
Figure 2 High-Power Haematoxylin and Eosin Staining Showing DLBCL
Figure 3 Strong CD20 Positivity
DLBCL usually expresses various pan-B markers, e.g. CD20 and CD79a.
Figure 4 High MIB-1 Expression
MIB-1 staining demonstrates proliferative activity of DLBCL. Proportion of cells stained is usually greater than 40%.
Which Further Investigations Were Indicated?
In view of the diagnosis of lymphoma, staging investigations were performed. Computed tomography (CT) imaging showed thickening of the gastro-oesophageal junction (see Figure 5). Multiple subcentimetre lymph nodes within the abdomen were also noted, but these were not enlarged by CT criteria. CT chest was normal. A bone marrow biopsy was normal.
Figure 5 CT at Diagnosis Showing Thickening of Gastro-Oesophageal Junction and Stomach Wall
The patient was therefore diagnosed with stage IE (see Table 2) DLBCL of the stomach. His international prognostic index (IPI) score was one (see Table 3).
Table 2 Ann Arbor Staging System
aSuffix “E” added to the stage denotes involvement of extra-nodal organ or site.
Table 3 International Prognostic Index
Source: [9].
aFive clinical features at presentation, indicative of a poor prognosis, used in calculating the IPI score are age greater than 60, Ann Arbor stage III or IV, elevated serum lactate dehydrogenase, performance status 2–4, and having more than one extra-nodal disease site. One point is given for each criterion met. Patients are assigned to one of four risk groups on the basis of their number of presenting risk factors.
What Was the Appropriate Treatment?
The patient was treated with R-CHOP chemotherapy (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone). He received four initial courses administered once every three weeks. His chemotherapy course was complicated by a hospital admission with Klebsiella pneumonia that responded well to antibiotics. After four courses of treatment he underwent a restaging CT and endoscopy. The CT scan showed improvement in the thickening of the gastro-oesophageal junction and no significant lymphadenopathy (see Figure 6). The repeat endoscopy and biopsy were normal.
Figure 6 Post-Treatment CT Showing Almost Complete Resolution of the Abnormal Thickening of Gastro-Oesophageal Junction and Stomach Wall
The patient went on to have two further courses of consolidation R-CHOP chemotherapy. Repeat CT imaging after six courses of chemotherapy showed no further change in the thickening of the gastro-oesophageal junction. In view of the fact that there had been no further response between the fourth and sixth courses of chemotherapy and a biopsy had been negative, no further consolidation chemotherapy was given. The final CT also showed collapse of the left lower lobe of the patient's lung. In view of these findings and the patient's history of being a heavy smoker, he underwent a fibre optic bronchoscopy. This revealed a mucus plug at the orifice of the left lower lobe, probably secondary to his previous pneumonia. The mucosa was normal. A repeat chest X ray showed re-inflation of the left lower lobe of his lung.
At follow-up this patient remains well and asymptomatic and is gaining weight. His last blood tests were normal. He remains under standard review every three months.
DISCUSSION
Symptoms of dyspepsia and epigastric pain are common, especially in the Western world. There is a one in ten lifetime risk for people in the Western world of developing a peptic ulcer [1]. If symptoms persist or clinical history is suggestive of significant pathology, an endoscopy with biopsies is mandatory. Although not at the top of the list of differential diagnoses, this case highlights the fact that lymphoma needs to be considered as a possible diagnosis.
Non-Hodgkin Lymphoma
NHL is the 11th commonest malignancy worldwide. It is increasing in incidence, with more than 300,000 cases worldwide and 9,000 cases in the United Kingdom diagnosed each year [2]. DLBCL is the most frequently occurring NHL, constituting about one-third of all adult NHL [3]. There is a slight male predominance, and median age of presentation is in the seventh decade [3,4], as in this patient, although the age range is broad.
Patients can present with nodal or extra-nodal disease. Up to 40% of cases are initially confined to extra-nodal sites [5]. The most common extra-nodal site is the gastrointestinal tract [5], especially the stomach, although any extra-nodal location can be a primary site. Lymphomas of the stomach can either be indolent MALT lymphoma or DLBCL, which may or may not be on the background of a MALT lymphoma.
For MALT lymphoma the median age of presentation is in the seventh decade—median age is 61 years—with a slight female predominance [4]. There is a close association between gastric MALT lymphoma and Helicobacter pylori infection; over 90% of cases are H. pylori positive [6]. For MALT lymphoma, if the lesion is limited to the gastric mucosa, eradication of the bacteria with triple antibiotic therapy is often sufficient to cause regression of the lymphoma [7,8], whereas DLBCL also requires treatment with chemotherapy.
Use of the IPI score [9], defined by clinical prognostic factors, is currently the standard approach to stratify patient risk in DLBCL (see Table 3). Difference in outcome, though, between groups of patients with similar IPI scores indicates that DLBCL encompasses a clinically and biologically heterogeneous group of tumours, which vary in response to chemotherapy. Combining the IPI risk score with cellular and molecular markers may improve patient risk stratification at presentation. For example, germinal cell phenotype, defined by BCL-6 and CD10 positivity, has been associated with a better outcome [10,11], and BCL-2 protein expression has been shown to be predictive of a worse outcome [11–15]. Both have been demonstrated immunohistochemically and confirmed by gene microarray analysis [16–18]. Combining cellular prognostic factors with the IPI is likely to become increasingly important in predicting outcome and tailoring therapy for DLBCL [19].
Treatment of DLBCL
DLBCL is an aggressive lymphoma but potentially curable. Treatment for DLBCL is combination chemotherapy plus the monoclonal antibody rituximab [20]. CHOP chemotherapy (cyclophosphamide, doxorubicin, vincristine, and prednisolone) is given every three weeks. Rituximab is a chimeric monoclonal antibody containing human IgG lambda and kappa constant regions with murine variable regions, which has a high affinity for the CD20 antigen. A dose of 375 mg/m2 given every three weeks with CHOP chemotherapy is now considered standard therapy for DLBCL. The mechanisms of action of anti-CD20 antibody are multiple and include induction of antibody-dependent cell-mediated cytotoxicity, induction of complement-mediated lysis, induction of apoptosis, and sensitisation of resistant lymphoma cells to chemotherapeutic agents.
Rituximab and CHOP chemotherapy have non-overlapping toxic effects, with emerging evidence of synergy [20,21]. The French GELA group (Groupe d'Etude des Lymphomes de l'Adulte) showed that addition of rituximab to CHOP (R-CHOP) had significant activity in elderly patients newly diagnosed with DLBCL [20]. CHOP plus rituximab was associated with significantly better overall response rate (94%), complete response rate (76%), event-free survival, and overall survival than CHOP alone. This survival advantage in elderly patients (>60 years) with DLBCL following R-CHOP represents the first major therapeutic advance for this group of patients since the introduction of CHOP nearly 30 years ago. Recent studies show similar benefits in younger patients [21,22], so rituximab in combination with CHOP should be used for all newly diagnosed patients with DLBCL.
Toxicity of Chemotherapy
General toxicity associated with chemotherapy includes mucositis, gastrointestinal disturbance, myelosuppression (increasing the risk of infections and bleeding), alopecia, parathesiae, and infertility. Specific side effects of chemotherapeutic agents used for treating NHL are shown in Table 4.
Table 4 Specific Side Effects of Chemotherapeutic Agents Used for Treating NHL
Side effects of rituximab include fevers, headache, nausea, lethargy, myalgia, arthralgia, skin erythema, pruritus, and hypotension. Most of these symptoms disappear upon temporary slowing or discontinuation of the treatment or after the administration of paracetamol and/or anti-allergic medication. Fatalities following severe cytokine release syndrome—which is characterised by severe shortness of breath and associated with features of tumour lysis syndrome—have occurred 1–2 hours after infusion of rituximab([23], Section 8.2.3; [24]). Patients with a high tumour burden as well as those with pulmonary insufficiency or infiltration are at increased risk and should be monitored very closely, and a slower rate of infusion should be considered as necessary.
Related to this case in particular, patients with gastric ulcers are at risk of perforation on initiation of treatment. Such patients should receive proton-pump inhibitor prophylaxis during chemotherapy. There is a need for close observation when chemotherapy is started, but the presence of a gastric ulcer is not an indication for dose reduction.
Monitoring Response to Chemotherapy
Response to chemotherapy needs to be monitored on a regular basis. Patients should be restaged after four courses of R-CHOP chemotherapy with CT scanning. If the bone marrow was involved at diagnosis, then a repeat biopsy is required. In cases such as our patient where disease was apparent elsewhere, a repeat biopsy of that lesion should be performed.
In the UK, standard practice for those in complete remission with no evidence of disease is to receive a further two courses of R-CHOP consolidation. Those in partial remission (≥50% improvement) after four courses are restaged after six courses, and if they are improving they go on to receive two further courses of R-CHOP. If there is no further improvement, consolidation radiotherapy is considered. Those in partial remission after four courses with no further improvement after six courses of R-CHOP receive consolidation radiotherapy or are considered for second-line treatment with an alternative regimen containing different chemotherapeutic agents.
Practice elsewhere in Europe varies, with up to eight courses of R-CHOP being given regardless of remission status. Although not universally available, FDG-PET (2-fluoro-2-deoxy-d-glucose positron emission tomography) is a functional imaging technique that is currently the most accurate modality to identify residual disease and differentiate it from scar tissue in patients with residual radiological abnormalities [25–27]. In general, re-biopsy of suspicious lesions should be considered.
Learning Points
In all patients a careful history of the nature of the dyspepsia and the associated symptoms should be taken. All patients over 45 years with dyspepsia should have an endoscopy performed. For younger patients, those with persistent symptoms or clinical history suggestive of significant pathology such as weight loss should also undergo endoscopy.
Not all gastric lesions require complete surgical resection. It is imperative that an endoscopy with biopsies is performed to allow histological diagnosis.
Patients with DLBCL require staging to ascertain the extent of disease and for comparison with post-therapy imaging. Determination of the IPI score is important to assign risk group and to predict prognosis.
Optimal treatment for DLBCL is CHOP chemotherapy in combination with rituximab. Patients should receive 6–8 courses.
Citation: Burton C (2005) A 61-year-old man with dyspepsia and weight loss. PLoS Med 2(6): e154.
Abbreviations
CTcomputed tomography
DLBCLdiffuse large B cell lymphoma
IPIinternational prognostic index
MALTmucosa-associated lymphoid tissue
NHLnon-Hodgkin lymphoma
==== Refs
References
Calam J Warrell D Cox T Firth J Benz E Peptic ulcer disease Oxford textbook of medicine, 4th ed 2003 Oxford Oxford University Press 558 568
Ferlay J Bray F Pisani P Harkin D GLOBOCAN 2002—Cancer incidence, mortality and prevalence worldwide, version 2 IARC CancerBase No. 5 [computer program] 2004 Lyon IARCPress
Armitage J Weisenburger D New approach to classifying non-Hodgkin's lymphomas: Clinical features of the major histologic subtypes. Non-Hodgkin's Lymphoma Classification Project J Clin Oncol 1998 16 2780 2795 9704731
The Non-Hodgkin's Lymphoma Classification Project A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma Blood 1997 89 3909 3918 9166827
Harris N Jaffe E Stein H Banks P Chan J A revised European-American classification of lymphoid neoplasms: A proposal from the International Lymphoma Study Group Blood 1994 84 1361 1392 8068936
Wotherspoon A Ortiz-Hidalgo C Falzon M Isaacson P
Helicobacter pylori -associated gastritis and primary B-cell gastric lymphoma Lancet 1991 338 1175 1176 1682595
Bayerdörffer E Neubauer A Rudolph B Thiede C Lehn N Regression of primary gastric lymphoma of mucosa-associated lymphoid tissue type after cure of Helicobacter pylori infection Lancet 1995 345 1591 1594 7783535
Fischbach W Goebeler-Kolve ME Dragosics B Greiner A Stolte M Long term outcome of patients with gastric marginal zone B cell lymphoma of mucosa associated lymphoid tissue (MALT) following exclusive Helicobacter pylori eradication therapy: Experience from a large prospective series Gut 2004 53 34 37 14684573
A predictive model for aggressive non-Hodgkin's lymphoma. The International Non-Hodgkin's Lymphoma Prognostic Factors Project N Engl J Med 1993 329 987 994 8141877
Alizadeh A Eisen M Davis R Ma C Lossos I Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling Nature 2000 403 503 511 10676951
Barrans S Carter I Owen R Davies F Patmore R Germinal center phenotype and bcl-2 expression combined with the International Prognostic Index improves patient risk stratification in diffuse large B-cell lymphoma Blood 2002 99 1136 1143 11830458
Hermine O Haioun C Lepage E d'Agay M Briere J Prognostic significance of bcl-2 protein expression in aggressive non-Hodgkin's lymphoma. Groupe d'Etude des Lymphomes de l'Adulte (GELA) Blood 1996 87 265 272 8547651
Kramer M Hermans J Parker J Krol A Kluin-Nelemans J Clinical significance of bcl2 and p53 protein expression in diffuse large B-cell lymphoma: A population-based study J Clin Oncol 1996 14 2131 2138 8683246
Hill M MacLennan K Cunningham D Vaughan Hudson B Burke M Prognostic significance of BCL-2 expression and bcl-2 major breakpoint region rearrangement in diffuse large cell non-Hodgkin's lymphoma: A British National Lymphoma Investigation Study Blood 1996 88 1046 1051 8704213
Gascoyne R Adomat S Krajewski S Krajewska M Horsman D Prognostic significance of Bcl-2 protein expression and Bcl-2 gene rearrangement in diffuse aggressive non-Hodgkin's lymphoma Blood 1997 90 244 251 9207459
Hans C Weisenberger D Greiner T Gascoyne R Delabie J Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray Blood 2004 103 275 282 14504078
Lossos I Czerwinski D Alizadeh A Wechser M Tibshirani R Prediction of survival in diffuse large B-cell lymphoma based on the expression of six genes N Engl J Med 2004 350 1828 1837 15115829
Rosenwald A Wright G Chan W Connors J Campo E The use of molecular profiling to predict survival after chemotherapy for diffuse large B-cell lymphoma N Engl J Med 2002 346 1937 1947 12075054
Mounier N Briere J Gisselbrecht C Emile J Lederlin P Rituximab plus CHOP (R-CHOP) overcomes bcl-2-associated resistance to chemotherapy in elderly patients with diffuse large B-cell lymphoma (DLBCL) Blood 2003 101 4279 4284 12576316
Coiffier B Lepage E Briere J Herbrecht R Tilly H CHOP chemotherapy plus rituximab compared with CHOP alone in elderly patients with diffuse large-B-cell lymphoma N Engl J Med 2002 346 235 242 11807147
Vose J Link B Grossbard M Czuczman M Grillo-Lopez A Phase II study of rituximab in combination with CHOP chemotherapy in patients with previously untreated, aggressive non-Hodgkin's lymphoma J Clin Oncol 2001 19 389 397 11208830
Pfreundschuh M Trumper L Gill D Osterborg A Pettengell R First analysis of the completed MabThera international (MInT) trial in young patients with low-risk diffuse large B-cell lymphoma (DLBCL): Addition of rituximab to a CHOP-like regimen significantly improves outcome of all patients with the identification of a very favourable subgroup with IPI = 0 and no bulky disease [abstract] Blood 2004 104 157A
Joint Formulary Committee British National Formulary, 49th ed 2005 London British Medical Association and the Royal Pharmaceutical Society of Great Britain
Yang H Rosove M Figlin R Tumor lysis syndrome occurring after the administration of rituximab in lymphoproliferative disorders: High-grade non-Hodgkin's lymphoma and chronic lymphocytic leukemia Am J Hematol 1999 62 247 250 10589082
Jerusalem G Beguin Y Fassotte M Najjar F Paulus P Whole-body positron emission tomography using 18F-fluorodeoxyglucose for posttreatment evaluation in Hodgkin's disease and non-Hodgkin's lymphoma has higher diagnostic and prognostic value than classical computed tomography scan imaging Blood 1999 94 429 433 10397709
Mikhaeel N Timothy A O'Doherty M Hain S Maisey M 18-FDG-PET as a prognostic indicator in the treatment of aggressive non-Hodgkin's lymphoma—Comparison with CT Leuk Lymphoma 2000 39 543 553 11342337
Spaepen K Stroobants S Dupont P Van Steenweghen S Thomas J Prognostic value of positron emission tomography (PET) with fluorine-18 fluorodeoxyglucose ([18F]FDG) after first-line chemotherapy in non-Hodgkin's lymphoma: Is [18F]FDG-PET a valid alternative to conventional diagnostic methods? J Clin Oncol 2001 19 414 419 11208833
| 15971937 | PMC1160569 | CC BY | 2021-01-05 11:13:39 | no | PLoS Med. 2005 Jun 28; 2(6):e154 | utf-8 | PLoS Med | 2,005 | 10.1371/journal.pmed.0020154 | oa_comm |
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1597193810.1371/journal.pmed.0020155Correspondence and Other CommunicationsBioethicsScience PolicyEpidemiology/Public HealthHealth PolicyHIV/AIDSMedical EthicsResource allocation and rationingEthicsInternational healthHIV Infection/AIDSMedicine in Developing CountriesAllocating Antiretrovirals in South Africa: Using Modeling to Determine Treatment Equity CorrespondenceWilson David P Blower Sally M University of CaliforniaLos Angeles, CaliforniaUnited States of AmericaE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
6 2005 28 6 2005 2 6 e155Copyright: © 2005 Wilson and Blower.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Designing an Equitable Strategy for Allocating Antiretroviral Treatments
Authors' Reply
==== Body
Recently PLoS Medicine published our paper entitled “Designing Equitable Antiretroviral Allocation Strategies in Resource-Constrained Countries” [1]. We were disappointed to find that the editorial perspective written by the World Health Organization (WHO) ethicists regarding our paper [2] was based upon a substantial misunderstanding of our novel quantitative analyses and our important results. Hence, they misunderstood the significance of the health-policy implications of our results. Thus, we wish to correct the record.
Firstly, Capron and Reis [2] misunderstood our quantitative analyses. They stated that “Wilson and Blower developed a mathematical model that could inform policy-makers' decisions regarding the optimal distribution of treatment sites to ensure equal access by all individuals infected with HIV.” However, our model does not determine the optimal distribution of treatment sites. As we clearly state in our paper [1] (and is also stated in the synopsis [3]), we developed a model that policy makers can use to make decisions regarding how to achieve the optimal allocation of scarce antiretrovirals among the available health-care facilities (HCFs) if the objective is to ensure treatment equity. We also calculated how the optimal allocation of antiretrovirals would vary if the number of HCFs utilized increased and/or the size of the catchment area that each HCF services increased [1]. Thus, we took the treatment sites (i.e., HCFs) as given, and we used their specific spatial location in South Africa as inputs to our model in order to determine optimal antiretroviral allocation strategies under a variety of conditions.
Secondly, Capron and Reis [2] misunderstood our important results. They stated that “applying this tool to the South-African province of KwaZulu–Natal, Wilson and Blower were able to confirm mathematically the intuitive assumption that using a maximum number of centers, at the least possible distance from most affected populations, would lead to the greatest fairness in the geographical distribution of ART [antiretroviral therapy].” We agree that if these had been our results, they would have been trivial and obvious. However, Capron and Reis [2] did not discuss our actual results: we determined how to decide how many drugs to allocate to each of the available HCFs in order to achieve an optimal allocation if the objective is to ensure treatment equity. This is a very complex problem and the antiretroviral allocation strategies that we calculated (by using our model) to be optimal are very complex (see Figure 3 in our paper, which graphically shows the proportion of drugs that should be allocated to each of the available HCFs). Furthermore, we also determined what catchment area each HCF should service; specifically, we calculated that each HCF should serve (if the objective is to achieve treatment equity) a catchment area of 40–60 km. Thus, our results demonstrate (to our knowledge for the first time) that patients infected with HIV will have to travel extremely large distances (i.e., 40-–60 km) in order to receive antiretrovirals, if the objective is to achieve treatment equity in South Africa. We stress that currently it is unknown what the actual size of the catchment area is around HCFs in South Africa. Catchment areas may in fact be very small. Thus, we suggested [1] that a primary goal should be to obtain empirical data of the distances that patients in South Africa are willing (or able) to travel in order to receive antiretrovirals. We have been the first to provide a quantitative assessment of the necessary size of the catchment area, and our results have identified that there is an urgent need to collect these critical data for quantifying the size of the catchment areas around HCFs. We have determined that the size of the catchment area will be a critical component in the ability to achieve treatment equity in South Africa. We also compared the optimal antiretroviral allocation strategies that we calculated with the current plan of the South African government for allocating antiretrovirals [4], and we determined that the current antiretroviral allocation strategies in South Africa will not achieve treatment equity. Taken together, our quantitative results are novel and controversial, providing important quantitative insights into a complex public-health problem.
We applaud the ambitious “3 by 5” WHO target for the antiretroviral rollout. However, the WHO has not yet devised a quantitative policy for determining how to allocate antiretrovirals in situations where the demand for drugs greatly exceeds the supply [5]. Health-policy officials in each country will have to make these important and difficult decisions, and they will all make different decisions based upon what objectives they wish to optimize and prioritize. There are a multitude of factors to consider (these factors are well described in the recent Institute of Medicine report [6]). We stress that the alternative to a quantitative rational approach for allocating scarce resources is an ad hoc approach, which is how the scarce supply of antiretrovirals is currently being distributed in many resource-constrained countries. Our operations research modeling approach is based upon spatial heterogeneity in the distribution of HCFs in South Africa and the spatial heterogeneity of the HIV-infected population. The most important “real world” result is that we show that what the South African government is currently doing is inequitable. We show them how to achieve equity, if they wish to do so. We hope that our novel approach for deciding how to allocate antiretrovirals will be of use to the WHO and also to the relevant authorities in the many resource-constrained countries who will soon have to make very difficult decisions as to who lives and who dies. Our analysis is to our knowledge the first analysis to show how a rational and scientific solution can be reached for deciding how to allocate a limited amount of antiretrovirals, if the goal is to achieve treatment equity. Clearly, other goals must be taken into consideration (and our model can be modified to include these other goals); however, we hope that treatment equity will be a very high priority during the antiretroviral rollout that is just beginning.
Citation: Wilson DP, Blower S (2005) Allocating antiretrovirals in South Africa: Using modeling to determine treatment equity. PLoS Med 2(6): e155.
==== Refs
References
Wilson DP Blower SM Designing equitable antiretroviral allocation strategies in resource-constrained countries PLoS Med 2005 2 e50 10.1371/journal.pmed.0020050 15737005
Capron AM Reis A Designing an equitable strategy for allocating antiretroviral treatments PLoS Med 2005 2 e69 10.1371/journal.pmed.0020069 15737013
Equitable allocation of antiretrovirals in resource-constrained countries PLoS Med 2005 2 e57 10.1371/journal.pmed.0020057
Tshabalala-Msimang M Statement of cabinet on a plan for comprehensive treatment and care for HIV and AIDS in South Africa 2003 South Africa Cabinet Available at http://www.capegateway.gov.za/eng/pubs/news/2004/feb/29782 . Accessed 12 May 2005
Blower SM Bodine EN Kahn JO McFarland W The antiretroviral rollout and drug- resistant HIV in Africa: Insights from empirical data and theoretical models AIDS 2005 19 1 14 15627028
Curran JW Scaling up treatment for the global AIDS pandemic: Challenges and opportunities 2004 Washington (DC) Institute of Medicine of the National Academies Available at http://www7.nationalacademies.org/ocga/briefings/Scaling_Up_Treatment_for_the_Global_AIDS_Pandemic.asp . Accessed 12 May 2005
| 15971938 | PMC1160570 | CC BY | 2021-01-05 10:39:51 | no | PLoS Med. 2005 Jun 28; 2(6):e155 | utf-8 | PLoS Med | 2,005 | 10.1371/journal.pmed.0020155 | oa_comm |
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1597193910.1371/journal.pmed.0020156Correspondence and Other CommunicationsInfectious DiseasesEpidemiology/Public HealthMalariaMedicine in Developing CountriesMalaria Diagnosis and Treatment: One Size Does Not Fit All CorrespondenceDrakeley Chris Gosling Roly Reyburn Hugh London School of Hygene and Tropical Medicine, London, England, and the Joint Malaria Program, Moshi, TanzaniaE-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
6 2005 28 6 2005 2 6 e156Copyright: © 2005 Drakeley et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Teaching Health Workers Malaria Diagnosis
==== Body
Icke and colleagues highlight an online and CD-ROM source available for the teaching of malaria diagnosis [1]. While this is a commendable initiative that provides open access to badly needed training materials, we feel that it is overambitious to propose this Web site (which gives equal weight to diagnosis, treatment, and chemoprophylaxis) as a resource that can be equally relevant to the issues of travel-related malaria, to which the Web site primarily applies, and the problems of malaria diagnosis and treatment in sub-Saharan Africa, which are described as the primary focus in the PLoS Medicine article [1].
The microscopic diagnosis of malaria is indeed an important resource in Africa to which the large majority of health facilities do not have access. Even where microscopy is available, there is evidence of substantial overdiagnosis of malaria and that better access to malaria microscopy may not result in better targeting of antimalarials [2,3]. What data there is on the accuracy of slide results suggest that microscope quality, preparation of blood films, and quality of reagents are at least as important in constraining the quality of results as the ability of slide readers to identify Plasmodium parasites from high-quality thin blood films [4–6]. In fact, thick blood films are the norm in Africa, thin films being almost exclusively confined to research, yet these are in the small minority on the Web site.
P. falciparum is overwhelmingly the dominant Plasmodium species seen in Africa, and training guides that try to give a “global overview” run the risk of giving a misleading impression of the relative importance of the four species in any given setting where malaria is endemic. The only actual setting where all species of malaria might be seen relatively frequently would be in travel clinics in developed countries. This point might seem obvious, but for many junior laboratory technicians in Africa hungry for scarce training opportunities it is potentially distracting and confusing.
The problems of simultaneously addressing the issues of malaria in endemic areas of Africa and in travellers from resource-rich countries are particularly applicable to malaria treatment. On the Web site, among the drugs recommended for the treatment of non-severe malaria are quinine, atovaquone/proguanil hydrochloride, mefloquine, and two tetracyclines. In Africa quinine is reserved for severe malaria and is associated with a high rate of failure due to poor adherence in outpatients; atovaquone/proguanil hydrochloride and mefloquine are not widely used because of their high cost; and tetracyclines are contraindicated in children and pregnant women, who bear the overwhelming burden of malaria.
The Web site describes criteria for severe malaria that are not appropriate for Africa. Thus, cerebral malaria, jaundice, renal failure, and lactic acidosis (impossible to detect in most African hospitals where “respiratory distress” is the equivalent criterion) are all listed ahead of severe anaemia, the most common manifestation of severe malaria in Africa. Repeated convulsions are not mentioned, presumably because they are rare among travellers with malaria, but they are common among children in Africa and are associated with a poor outcome. Renal failure as a manifestation of severe malaria is mentioned second but is very uncommon in Africa [7,8].
The apparently large numbers of African health-care workers who have accessed this site might legitimately feel confused by some of these descriptions, and there is no indication of which, if any, sections apply to Africa and which to travellers. While many will interpret the information critically, others may not, especially when English is not their first language and where there is a tendency to accept didactic sources in preference to what is obvious from personal or local experience.
The diagnostic component of the Web site has important potential as a training tool, but we feel that significant modification and more explicit indication of information that is country-specific are needed before its use is likely to result in improved standards of care in African hospitals and clinics.
It is true that Internet technologies can “revolutionise information technology…for healthcare professionals in developing countries”. The World Health Organization and others already have very extensive Web sites that are contributing to this. Searching the Internet on “malaria diagnosis” reveals a number of sites, many with excellent sections, but one is struck by the lack of clarity defining the target audience and context of the problem. For much information available on the Web this may not matter, but for malaria we feel it does.
Citation: Drakeley C, Gosling R, Reyburn H (2005) Malaria diagnosis and treatment: One size does not fit all. PLoS Med 2(6): e156.
==== Refs
References
Icke G Davis R McConnel W Teaching health workers malaria diagnosis PLoS Med 2005 2 e11 10.1371/journal.pmed.0020011 15736990
Barat L Chipipa J Kolczak M Sukwa T Does the availability of blood slide microscopy for malaria at health centers improve the management of persons with fever in Zambia? Am J Trop Med Hyg 1999 60 1024 1030 10403337
Reyburn H Mbatia R Drakeley C Carneiro I Mwakasungula E Overdiagnosis of malaria in patients with severe febrile illness in Tanzania: A prospective study BMJ 2004 329 1212 15542534
Amexo M Tolhurst R Barnish G Bates I Malaria misdiagnosis: Effects on the poor and vulnerable Lancet 2004 364 1896 1898 15555670
Bates I Bekoe V Asamoa-Adu A Improving the accuracy of malaria-related laboratory tests in Ghana Malar J 2004 3 38 15516269
Opoku-Okrah C Rumble R Bedu-Addo G Bates I Improving microscope quality is a good investment for under-resourced laboratories Trans R Soc Trop Med Hyg 2000 94 582 11132395
Schellenberg D Menendez C Kahigwa E Font F Galindo C African children with malaria in an area of intense Plasmodium falciparum transmission: Features on admission to the hospital and risk factors for death Am J Trop Med Hyg 1999 61 431 438 10497986
Marsh K Forster D Waruiru C Mwangi I Winstanley M Indicators of life-threatening malaria in African children N Engl J Med 1995 332 1399 1404 7723795
| 15971939 | PMC1160571 | CC BY | 2021-01-05 10:40:14 | no | PLoS Med. 2005 Jun 28; 2(6):e156 | utf-8 | PLoS Med | 2,005 | 10.1371/journal.pmed.0020156 | oa_comm |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.