Search is not available for this dataset
article stringlengths 4.36k 149k | summary stringlengths 32 3.35k | section_headings listlengths 1 91 | keywords listlengths 0 141 | year stringclasses 13
values | title stringlengths 20 281 |
|---|---|---|---|---|---|
Mammalian DNA replication starts at distinct chromosomal sites in a tissue-specific pattern coordinated with transcription , but previous studies have not yet identified a chromatin modification that correlates with the initiation of DNA replication at particular genomic locations . Here we report that a distinct fraction of replication initiation sites in the human genome are associated with a high frequency of dimethylation of histone H3 lysine K79 ( H3K79Me2 ) . H3K79Me2-containing chromatin exhibited the highest genome-wide enrichment for replication initiation events observed for any chromatin modification examined thus far ( 23 . 39% of H3K79Me2 peaks were detected in regions adjacent to replication initiation events ) . The association of H3K79Me2 with replication initiation sites was independent and not synergistic with other chromatin modifications . H3K79 dimethylation exhibited wider distribution on chromatin during S-phase , but only regions with H3K79 methylation in G1 and G2 were enriched in replication initiation events . H3K79 was dimethylated in a region containing a functional replicator ( a DNA sequence capable of initiating DNA replication ) , but the methylation was not evident in a mutant replicator that could not initiate replication . Depletion of DOT1L , the sole enzyme responsible for H3K79 methylation , triggered limited genomic over-replication although most cells could continue to proliferate and replicate DNA in the absence of methylated H3K79 . Thus , prevention of H3K79 methylation might affect regulatory processes that modulate the order and timing of DNA replication . These data are consistent with the hypothesis that dimethylated H3K79 associates with some replication origins and marks replicated chromatin during S-phase to prevent re-replication and preserve genomic stability .
The ability to turn gene expression on and off is fundamental to cell cycle progression and metazoan development . Selective gene expression requires chromatin adjustments , mediated by post-translational modifications of chromatin-associated proteins such as histones . In addition to these changes in chromatin condensation , a complete copy of the entire cellular genome must be replicated during each cell cycle . Thus , cells must coordinate replication with chromatin modifications to insure that all genetic and epigenetic information is accurately transferred to the daughter cells . It is unclear how replication proceeds along with chromatin condensation and remodeling while ensuring the fidelity of the replicated genome . In most somatic cells , DNA replication starts from consistent multiple initiation sites on each chromosome and advances in a precise temporal and tissue-specific order . It is postulated that this temporal and spatial consistency reflects a tight orchestration of replication initiation events that is necessary to coordinate replication with other chromatin transactions such as transcription . Several lines of evidence suggest that chromatin modifications play a role in coordinating replication and transcription . Mapping the locations of replication initiation events show that replication initiation sites are enriched with transcription factor binding motifs , CpG islands and G-quartets [1]–[4] . Replication preferentially starts in transcribed chromatin [5] , with the highest preference observed in moderately transcribed regions [3] , and associates with genomic regions exhibiting DNAse hypersensitivity and/or containing methylated CpG sequences [3] . Although many histone modifications were examined , no particular histone modification examined thus far showed a striking association with DNA replication . Further evidence for a potential role of chromatin modifications in DNA replication stems from genetic studies characterizing the determinants of replication initiation sites . Distal DNA elements , which do not start replication but are involved in chromatin remodeling , interact with replicators , which directly facilitate initiation of DNA replication ( for reviews , see [6] , [7] ) . Such interactions are required for initiation of replication at a number of loci , including a region 40 kb upstream of the human beta-globin ( HBB ) replication origin [8] , the promoter of the Chinese hamster Dhfr locus [9] , and an enhancer of the Th2 locus [10] . In addition , replicator sequences themselves can affect chromatin structure . For example , replicators prevent transcriptional silencing [11] by facilitating an interaction between a locus control region and a chromatin remodeling complex [12] . It is likely that chromatin modifications play a role in mediating the distal interactions that determine the locations of replication initiation events and facilitate the effects of replicators on gene expression [13] , yet whole-genome mapping of replication initiation sites had not pointed to any particular histone modification [3] , [4] , [14] . Histone H3 exhibits methylation on lysine 79 ( H3K79 ) catalyzed by the methyltransferase DOT1 ( Disruptor of Telomere silencing 1 ) enzyme ( DOT1-like , or DOT1L in humans ) that facilitates telomeric and Sir mediated silencing [15] . DOT1 promotes the mono- , di- and tri-methylation of H3K79 [16] and these methylations are involved in transcriptional elongation , DNA repair , and heterochromatin maintenance . In yeast , cell cycle regulated genes exhibit differential di- and tri-methylation of H3K79 [17] , and methylation of H3K79 is required for the G1 and intra S-phase checkpoints . H3K79Me2 interacts with CAF-1 and is particularly abundant during late S-phase [18] . In mammals , methylation of H3K79 is abundant in active genes , including the murine beta-globin locus [19] . Human DOT1L methylates H3K79 and associates with a complex that participates in Wnt signaling [20] , which includes beta-catenin , Skp1 , and TRRAP . DOT1L is required for development and plays essential roles in early erythropoiesis [21] and cellular reprogramming during development [22] . Proper functioning of DOT1L , in collaboration with H2B ubiquitination , promotes the DNA damage checkpoint [23] , likely by H3K79Me2 mediated targeting of 53BP1 to DNA damage lesions [24] . Methylation of H3K79 was implicated as a determinant of global genomic repair [25] and silencing of tumor suppressor in hematologic malignancies [26] . Aberrant methylation of H3K79 by DOT1L is associated with MLL rearrangements in leukemia [27]–[29] , possibly by mistargeting of DOT1L activity [30] . The human beta-globin locus ( HBB ) contains one of the most intensely studied replication initiation regions ( IRs ) [31]–[34] in mammalian cells . This particular origin is used in both erythroid and nonerythroid cells , but the timing of replication initiation differs between these two cell types . In erythroid cells , the HBB locus initiates DNA replication during the early stages of S-phase , while initiation in non-erythroid cells typically occurs during late S-phase [35] , [36] . The HBB IR contains two independent replicators ( Rep-P and Rep-I ) , and each can initiate DNA replication at both native and ectopic sites [31] , [33] . Detailed genetic analyses have revealed that both Rep-P and Rep-I contain an AT-rich sequence and an asymmetric purine:pyrimidine ( AG ) sequence , both of which are required for replication initiation [34] . The HBB IR , therefore , provides an excellent system to study replicator-binding proteins that both recruit the general replication machinery to specific chromatin sites and interact with the cell cycle machinery . Because of the availability of mutants that do not initiate replication , this system is ideally suited to investigate whether any particular protein-DNA interaction correlates with functional replicator activity . Here , we asked whether methylation of H3K79 is associated with replication initiation events genome wide and followed those general observations with functional studies at the well-characterized replication initiation site within the human beta-globin locus . We also studied the function of H3K79 methylation in replication initiation by depletion of DOT1L . Our results suggest that H3K79Me2 is associated with initiation of DNA replication genome wide , that the modification of H3K79 at the beta globin locus correlates with replicator activity and that H3K79 methylation might play a role in ensuring that at each locus , replication would initiate only once per cell cycle .
We have recently mapped the locations of replication initiation events genome-wide in several human cell lines . Although replication initiation events were enriched in DNAse hypersensitive sites and in methylated CpG rich regions , previous studies in our lab and others have not identified chromatin modifications that exhibited a high enrichment for replication initiation events [3] . Because replication initiation events tend to be depleted at transcription start sites and enriched just downstream of those sites [3] , we asked if H3K79 dimethylation , which exhibits a similar pattern , might mark replication initiation events . We performed chromatin immunoprecipitation followed by sequencing ( ChIP-Seq ) with an antibody specifically directed against H3K79Me2 in human eryhtroleukemia K562 cells . K562 cells express the gamma globin in which the beta-globin locus replicates early during S phase and were used for numerous replication-related studies including whole genome origin mapping [3] . In addition , ChIP-Seq data delineating biding sites of many histone modifications are available for K562 cells , and the cells are amenable to fractionation according to cell cycle stages using centrifugal elutriation ( see Table S1 for a complete list of cell lines used in this study , as well as their backgrounds and/or reasons being used ) . The genomic locations of chromatin enriched in H3K79Me2 were identified by massively parallel sequencing and visualized relative to the locations of replication initiation events , based on sequencing short , RNA-primed nascent DNAs isolated from K562 cells [3] . The frequency of initiation events at individual genomic regions was measured as the ratio between reads obtained from a nascent strand preparation and reads obtained from a corresponding control genomic DNA preparation . The reads were calculated as reads per kilobase ( kb ) per million mapped reads ( RPKM ) . The results were visualized using the Integrative Genome Viewer 2 . 1 ( Broad Institute ) . Figure 1A–E shows the replication initiation ratio ( nascent strands vs . genomic RPKM ) and H3K79Me2 in sample human loci . For each image , a chromosome map is shown at the top , and the region-of-interest is circled . The analyzed region is shown underneath the ideogram , with map coordinates indicated . Ref-Seq genes are aligned under the map coordinates . The replication tracks show the distribution of sequence reads ( aligned with the indicated region ) obtained from massively parallel sequencing of nascent strands from K562 erythroleukemia cells as described [3] . All data are shown as the ratio of reads obtained from a nascent strand preparation , and reads obtained from a corresponding control genomic DNA preparation . The y-axis indicates the nascent strand/genomic DNA ratio . The H3K79Me2 tracks indicate the distribution of reads from ChIP-Seq analyses with H3K79Me2 specific antibodies as described in the Methods section . The data show that chromatin regions exhibiting H3K79 dimethylation exhibited high enrichment for replication initiation sites , measured by the ratio of reads on nascent strands vs . genomic DNA controls . When we calculated the enrichment of genome-wide H3K79Me2 peaks obtained from the ChIP-Seq data for replication initiation events , overall , 23 . 39% of the H3K79Me2 peaks were detected in regions adjacent to replication initiation events whereas an association with replication initiation events was expected in only 7 . 26% of the size-matched randomized feature regions ( Table 1 , “overall” ) . Chromatin regions exhibiting H3K79 methylation markedly enriched in replication initiation events , showing an average whole-genome RPKM ratio of 1 . 8 ( Figure 1F ) . As shown previously [3] , replication initiation events were also enriched , to a lesser extent , in chromatin regions exhibiting acetylation of histone H3 on lysines 9 and 27 and methylated on lysine 4 , and in chromatin binding transcription factors c-Jun and c-Myc . As shown in Figure 1F , the extent of enrichment for replication initiation events in regions exhibiting methylation of histone H3 on lysine 79 is markedly higher than in regions exhibiting other histone modifications . Given the sequencing depth and the large sample size , an average value of 1 . 3 ( as calculated for H3K27Ac ) represents significant enrichment over the average whole-genome RPKM ratio ( p<10−6 ) despite the seemingly low numerical values . The statistical significance of enrichment in H3K79Me2 in replication initiation events is very high ( p<10−20 ) . Replication initiation events were markedly enriched in regions within 500 bp of an H3K79Me2 binding site , and this is not true for other regions such as JunB binding sites ( Figure 1G and Figure S1 , S2 ) . H3K79 methylation exhibited a similar enrichment with replication initiation events as DNAse hypersensitivity and CpG methylation , however these traits did not exhibit any synergy ( Figure S3 ) . To further explore the association between H3K79 dimethylation and replication initiation , we measured the levels of H3K79Me2 enrichment during the G1 , S , and G2 phases of the cell cycle in K562 cells fractionated into synchronous cell cycle populations using centrifugal elutriation ( Figure 2 , Table 1 , rows G1 through G2 , and Table 2 ) and Western immunoblots . The overall level of H3K79Me2 modification did not change during the cell cycle ( Figure S4A ) and the abundance and distribution of chromatin regions exhibiting H3K79Me2 was similar in the G1 and G2 phases of the cell cycle . However , the distribution of H3K79Me2 associated regions expanded in S-phase . Interestingly , chromatin regions that exhibited H3K79Me2 only during S-phase did not display a high association with replication initiation events ( Figure 2C and Table 1 , “S only” ) . By contrast , another modification , H3K4Me3 , did not exhibit a similar expansion of peaks in S-phase and displayed a consistent distribution throughout the cell cycle ( Figure 2D and Table 2 ) . These observations support the conclusion that the H3K79Me2 modification associates preferentially in a cell cycle specific manner with replication initiation regions . To determine if H3K79 was methylated at a replicator sequence essential for initiation of DNA replication , we performed chromatin immunoprecipitation ( ChIP ) analysis with the Rep-P asymmetric region of the endogenous β-globin locus in K562 cells ( for maps , see Figure 3A and Figure S4B ) . This region is essential for initiation of DNA replication from Rep-P , one of the two adjacent replicators that can initiate DNA synthesis at the β-globin locus [33] , [34] . As shown in Figure S4B , probes that colocalized with the Rep-P replicator ( bG59 . 8 , bG61 . 3 , AG ) exhibited higher enrichment in H3K79Me2 . We also tested the association between H3K79Me2 and late-replicating replication initiation site in Jurkat cells , which are human T-cells that do not express any gene from the human beta-globin locus and yet initiate replication from Rep-P late during the S-phase of the cell cycle . The asymmetric region ( probe marked AG ) exhibited a high level of H3K79 dimethylation in both K562 cells that express γ-globin and in Jurkat cells , which do not express any gene within the HBB cluster and replicate the entire locus late in S-phase . We then investigated whether DNA sequences required for initiation of DNA replication were also essential for enrichment of H3K79Me2 in the Rep-P region . To that end , we introduced transgene cassettes that included the LCR , an enhanced GFP expression cassette driven by the beta-globin promoter , and Rep-P variants into a site termed random locus 4 ( RL4 ) in murine erythroleukemia ( MEL ) cells . This site was used previously to assess the roles of replicator sequences in initiation of DNA replication and gene silencing [37] . The RL4 site was engineered to contain an inverted pair of LoxP sites so that any inserted transgene cassette in that locus could be exchanged with other cassettes inserted precisely at the same genomic location . This system can facilitate testing for effects of distinct mutations and sequences from the murine beta globin locus serve as controls for human inserted sequences . Previous studies have shown that the unmodified RL4 site exhibits a heterochromatin conformation and does not initiate replication , hence initiation activity detected at this site after insertion of ectopic sequences reflects sequence information encoded within the inserted transgens [37] . Using this feature , we tested a transgene cassette that included an intact Rep-P ( Rep-PWT – Figure 3A ) and a cassette that included a Rep-P variant with only 2 nucleotides mutated in AG region of the rep-P ( Rep-PAG1 ) . The intact Rep-P ( Rep-PWT ) exhibited enrichment of H3K79Me2 ( Figure 3B , probes bG59 . 8 , 61 . 3 – see Figure S4B and Table 3 for locations and sequences of primer pairs and probes ) and initiated replication at the ectopic site ( Figure 3C ) . By contrast , the mutant Rep ( Rep-PAG1 ) did not exhibit enrichment of H3K79Me2 and did not initiate replication ( Figure 3 , B and C , respectively ) . Although the orientation of the transgene at the RL4 site affects transcriptional silencing [37] , analyses of Rep-PAG1 yielded similar results when the transgene was inserted in both orientations ( data not shown ) . It should be noted that in the RL4 locus the highest enrichment in the intact Rep-P transgene was observed upstream of the asymmetric region ( primer pairs bG59 . 8 and 61 . 3 ) whereas the asymmetric locus exhibited a higher enrichment at the native locus in K562 cells ( Figure S4 ) . This shift was most likely due to the fact that the inserted transgene contained only one of the two replicators at the replication initiation region ( Rep-P without Rep-I ) . Importantly , since of H3K79Me2 was detected in the active but not the inactive Rep-P , these observations suggest an association between H3K79 dimethylation and replicator activity . That H3K79Me2 is strongly associated with the start of DNA replication implicates a causal role for H3K79 dimethylation in regulating the initiation of DNA replication . To address this question , we depleted DOT1L , the enzyme that methylates histone H3 on lysine K79 . These experiments were performed in human colon cancer HCT116 cells . HCT116 cells were chosen for this analysis because these cells exhibit a relatively stable karyotype and can be easily transfected with siRNA , and are therefore suitable for analyses aimed to test the effects of siRNA mediated gene silencing on cell cycle distribution . In HCT116 cells , siRNA mediated DOT1L depletion prevented H3K79 dimethylation ( Figure 4A ) . We used those cells to measure the rate of DNA synthesis in cells with and without DOT1L using dynamic molecular combing ( Figure 4 , B–D and Figure S5 ) . As shown in Figure 4 , C and D , depletion of DOT1L did not affect replication fork velocity , suggesting that H3K79 methylation does not directly affect the progression of replication forks . DOT1L depletion also did not affect the frequency of replication initiation events ( Figure S6 ) . We then asked whether cell populations in which DOT1L was depleted exhibited changes in cell cycle patterns . The overall distribution of cells in the G1 , S and G2/M phases of the cell cycle was similar in control and DOT1L depleted cells , but FACS analyses indicated that DOT1L depleted cells had fewer cells in the early S-phase and more cells with late S-phase DNA content than cells in which DOT1L was not depleted ( transfected with a control siRNA; Figure 5A ) . As shown in Figure 5B , DOT1L depletion also resulted in an increased frequency of apoptosis ( subG1 ) and in an increase in the fraction of cells exhibiting DNA content greater than 4N ( >G2/M ) or cells that did not synthesized DNA despite a DNA content between 2N and 4N ( S non-replicating ) . Similar results were observed in U2OS osteosarcoma cells ( Figure S7 and Table S2 ) . These observations suggested that although most cells can continue to proliferate and replicate DNA in the absence of methylated H3K79 , prevention of H3K79 methylation might affect regulatory processes that modulate the order and timing of DNA replication . Because H3K79 methylation was enriched in replication initiation sites , we asked whether the order of DNA replication was altered in cell populations in which DOT1L was depleted . To measure DNA replication patterns , cells were pulse-labeled with the nucleotide analog EdU for 60 min . DNA content ( DAPI intensity ) was measured , along with the pattern of EdU staining ( Figure 6 , A and B ) . The pattern of replication foci as exhibited by EdU incorporation was recorded for each nucleus ( for examples , see Figure 6C ) . In untreated cells , diffuse replication foci patterns are characteristic of early S-phase ( Figure 6C , ES ) , whereas nuclei exhibiting a few condensed replication foci are abundant in late S-phase ( Figure 6C , LS ) . We then tallied the frequency of early and late S-phase EdU staining patterns in cell populations exhibiting early and late S-phase DNA content measured by DAPI staining . As expected we observed that diffuse patterns were indeed abundant in cells exhibiting early S-phase DNA content , whereas clustered patterns are frequent in cells exhibiting late S-phase DNA content ( Figure 6A and B ) . However , cell populations in which DOT1L was depleted contained a small subset of cells with DNA content of more than 4N exhibiting diffuse replication patterns ( Figure 6B; examples are shown in Figure 6C , > = G2M ) . This cell population might represent cells with late S-phase DNA content that re-replicated DNA that had already been duplicated earlier during the same S-phase , thus exhibiting an early S-phase replication pattern . This pattern is consistent with the observation that depletion of DOT1L resulted in an increased fraction of cells with DNA content larger than 4N , representing cells that skipped mitosis and partially re-replicated their DNA . To investigate whether the EdU staining patterns we have observed indeed reflect re-replication of DNA in cells we have labeled cells with Bromodeoxyuridine ( BrdU ) for 18 hours that exceeds the length of a complete S-phase but is shorter than the time required for cells to undergo a complete cell cycle . We then determined the extent of BrdU incorporation into DNA using density gradients . BrdU substituted DNA is more dense ( heavier ) then unsubstituted DNA and the difference between unsubstituted ( LL ) , semi-substituted ( HL ) and fully substituted ( HH ) DNA can be observed by recording the abundance of DNA in fractionated CsCl density gradients . BrdU substituted DNA was detected using anti-BrdU antibodies . As shown in Figure 6D , cells containing active Dot1L exhibited BrdU substituted DNA in fractions corresponding to HL ( semi-substituted ) DNA consistent with the assumption that those cells have completed a single round of replication . By contrast , cells in which Dot1L was depleted exhibited BrdU substituted DNA in both the HL and HH fractions , consistent with over-replication of a fraction of the DNA during the labeling period . Importantly , these results imply that H3K79 methylation plays a role in preventing re-replication during normal cell cycle progression .
The observations reported here demonstrate that H3K79Me2-containing chromatin was enriched in replication initiation events . The methylation of histone H3 on histone 79 exhibited the highest enrichment of replication initiation events observed in any single chromatin modification that was studied . It is worth noting , however , that despite the high level of enrichment in replication initiation events , not all replication initiation sites associated with H3K79 dimethylation . The association of H3K79Me2 with replication initiation sites was independent and not synergistic with other chromatin modifications . H3K79 dimethylation exhibited a wider distribution on chromatin during S-phase , but regions of chromatin that only associated with H3K79 dimethylation during S-phase were not in replication initiation events . We also observed that H3K79 dimethylation was enriched in chromatin containing a functional replicator , but was not enriched in chromatin containing a mutant replicator that could not initiate replication . Hence , H3K79 dimethylation at the human beta-globin locus replicator was associated with replicator activity . Prevention of H3K79 methylation by depletion of DOT1L did not affect the rate of DNA replication or the inter-origin distance . Importantly , however , over-replication occurred at a higher frequency following depletion of Dot1L , the sole enzyme responsible for H3K79 methylation . Together , these results demonstrate for the first time that H3K79 methylation is not required for replication initiation but rather plays a role in preventing re-replication of DNA once initiated . Our previous studies showed that methylated CpG regions and genes undergoing moderate transcription were highly associated with replication initiation sites [3] . Other studies have also found that replication initiation sites exhibit enrichment in transcribed regions [5] , G-quartets [4] and methylated CpGs [2] . Studies have also shown that methylation of H4 lysine 20 is required for genome-wide DNA replication [38] , suggesting a potential mechanistic involvement of this histone modification in replication since Orc1 , which exhibits an association with replication origins [39] , interacts with H4K20Me2 through its BAH domain [40] . However , these studies have not identified a single histone modification that is associated with initiation of DNA replication at distinct genomic sites . Here , we have identified H3K79 methylation as a modification that is not only associated with initiation but plays a functional role in restricting replication to once per cell cycle . Our ChIP-Seq data suggest that H3K79 dimethylation might occur in regions proximal to replication initiation sites during G1 and then expand to adjacent regions during S-phase . H3K79me2 marks again cluster with initiation sites in G2 , suggesting that the S-phase specific marks , which do not associate with replication origins , might be erased and the origin-specific marks remain post-mitosis for the next cell cycle . These results imply the intriguing possibility for a mechanism to specifically and quickly remove H3K79 methylation . A precedent for an enzyme that might remove methylated histones in that way is Rph1/KDM4 , which can specifically demethylate H3K36 in yeast [41] . The observed restriction of H3K79me2 containing chromatin to replication origins after S-phase might be mediated by an equivalent demethylase capable of removing methyl groups from H3K79 in non-origin regions , or by active removal of H3K79Me2 containing nucleosomes from chromatin that is not associated with replicator activity . The mechanism ( s ) by which potential replication origins retain H3K79 methylation whereas regions that are not associated with replication origins lose H3K79 methylation are under investigation . Regardless of the mechanism , regions of potential replication origins that can undergo initiation of DNA replication specifically retain the H3K79 methylation mark during cell division . The most likely role played by replication origin associated DOT1L-mediated H3K79 methylation is to facilitate an interaction that marks origins that had started replication , where such a mark might prevent replication from initiating a second time . Consistent with this suggestion , H3K79Me2 associated with active but not with mutant inactive replication origins and we observed cells with late S-phase DNA content and early replication foci patterns following DOT1L depletion . In accordance , direct measurement of nucleotide incorporation also showed that Dot1 depletion resulted in partial genome over-replication , and cell cycle profiles of DOT1L depleted cells detected a larger population of cells with DNA content higher than 4N and cells with a late-S-phase DNA content . Such cells might have re-started replication of early-replicating origins without completing the replication of late-replicating origins ( reflected in late S-phase DNA content and early replication patterns ) , or they might have completed replication of their entire genomes and skipped mitosis to re-start replication at early replication origins ( reflected in DNA content greater than 4N ) . Taken together , over-replication of a part of the genome and early replication foci patterns in cells with late S-phase DNA content likely indicate re-replication of early replicating regions . Our observations are consistent with the hypothesis that methylation of H3K79 marks replicated regions and prevents re-initiation; when methylation is absent , cells undergo limited re-initiation of DNA replication in early replicating origins . DOT1L depleted cell populations also contain a marked fraction of cells with S-phase DNA content that are not actively replicating DNA , consistent with the suggestion that the limited re-replication we observed in the absence of H3LK79Me2 is curbed by regulatory checkpoint pathways during S-phase . H3K79Me2 associated with an active but not with a mutant inactive replication origin , but some H3K79Me2 associated genomic regions did not exhibit strong initiation activity . These apparently disparate observations might suggest that association with H3K79Me2 plays a role in regulating replication in a subset of genomic regions , or it can reflect variations in the use of initiation sites and the fact that not all potential initiation sites start replication each cell cycle [6] , [7] , [14] . Currently , we have yet to identify the distinct replication initiation regions that undergo re-replication because only a fraction of cells exhibit re-replication and current methods ( including NS-Seq and quantitative PCR-based measurements of nascent strand abundance ) are not sufficiently sensitive to detect small alterations in the abundance of nascent strands with the precision required for drawing statistically significant conclusions . It is possible that the H3K79Me2 might mark cryptic replication initiation sites , which are capable of initiation of DNA replication but only do so under distinct conditions such as exposure to DNA damaging agents [7] . If replication does occur on such cryptic origins marked by H3K79Me2 , H3K79Me2 might be available to facilitate in an interaction that will prevent re-replication from cryptic as well as constitutive origins . Because limiting replication once per cell cycle is critical in preventing genome instability , cells employ numerous strategies to prevent re-replication [42] , [43] . The mechanisms are diverse and species-specific , with components of the pre-replication licensing complex such as Cdt1 and ORC often serving as primary regulatory targets . Interestingly methylation of Tetrahymena histone H3K76 , which is orthologous to histone H3K79 in mammals , is required for replication initiation and overexpression of the Tetrahymena Dot1A methylase results in over-replication [44] . Although this observation suggests that methylation of H3K76 in Tetrahymena achieves the opposite effect than the methylation of H3K79 in mammals , and it is yet unclear whether Tetrahymena H3K76 methylation associates with replication origins , both enzymes seem to be involved in regulating replication re-initiation events and thus might both be involved in processes that mark genomic regions for initiation . Another precedent for a role of chromatin modifications in regulating replication initiation is found in plants , in which histone H3 lysine 27 ( H3K27 ) monomethyltransferases ( Arabidopsis Trithorax-Related 5 ( ATXR5 ) and ATXR6 interact with the two Arabidopsis proliferating cell nuclear antigen protein [45] and nuclei from atxr5 atxr6 double mutants exhibit evidence of over-replication associated with marked decondensation of constitutive heterochromatin at chromocenters [46] . In mammals , methylation of histone H4 on lysine 20 is associated with the onset of replication licensing in G1 and the dissociation of H4K20Me1 from replication origins in S-phase is associated with prevention of re-replication [38] . Methylation of histone H4 on lysine 20 was required for initiation of DNA replication from replication origins [38] , and although genome-wide analyses of replication initiation events did not exhibit a marked association with H4K20Me1 [3] it is possible that H4K20Me1 plays a role in establishing potential replication initiation patterns during the G1 phase . The temporal pattern of H3K79Me2 , as reported here , is distinct from H4K20Me1 . The increased frequency of apoptotic cells and cells with S-phase DNA content that do not replicate DNA suggest that the role played by H3K79Me2 in limiting re-replication is vital for cell survival and coordinated progression through the cell cycle . It is possible that the two modifications play distinct roles in regulating replication patterns and maintenance of genomic stability .
Human K562 cells , HCT1116 cells , U2OS cells and murine erythroleukemia cells ( RL4 ) containing Rep-P variants were grown at 37°C in a 5% CO2 atmosphere in Dulbecco's modified Eagle medium ( Invitrogen , Cat . no . 10564-011 ) , supplemented with 10% heat-inactivated fetal calf serum . For RL4 cells containing Rep-P variants , two transgenes containing sequences from the human beta-globin Locus Control Region ( HS432 ) , the human beta-globin promoter ( GloPro ) driving enhanced green fluorescent protein ( EGFP ) and two variants of the Rep-P replicator were inserted into a single location on murine chromosome 15 in murine erythroleukemia ( RL4 ) cells [11] . We introduced the AG1 mutation by site-directed mutagenesis . Details of the mutagenesis methods were published previously [12] . HCT116 cells were transfected with 25 nM of control siRNA ( Dharmacon , D-001810-10 or Dot1L siRNA ( Dharmacon , L-014900-01 or Qiagen , GS84444 ) with DharmaFECT 2 siRNA ( Dharmacon , T-2002 ) or RNAiMAX Transfection Reagent ( Invitrogen , 13778 ) according to the manufacturer's protocol . Cells were transfected 2 or 3 times with a 2 to 3 days interval and harvested 72 h after the last transfection . Whole cell lysates by1XSDS loading buffer were used for western-blot with anti-Histone H3 ( dimethylK79 ) antibody ( Abcam , ab3594 or Millipore , 04-835 ) to verify the knockdown efficiency of dot1L siRNA . We performed nascent-strand DNA analysis as described previously [33] . DNA was extracted from asynchronous cells , denatured by boiling for 10 minutes , incubated on ice for 10 minutes , and fractionated on a neutral sucrose gradient . We collected 0 . 5–1 kb DNA fractions , treated them with λ exonuclease to remove non-RNA-primed genomic DNA fragments , purified them , and measured the DNA concentration using a NanoDrop 1000 ( Thermo Scientific ) . We quantified nascent strand DNA by real-time polymerase chain reaction ( PCR ) in an ABI 7900 thermocycler ( primers and probes used for real-time PCR are listed in Table 3 ) . For whole genome analyses of nascent strand abundance , DNA from K562 cells was isolated using the above procedure and processed for massively parallel sequencing as previously described [3] . Three independent biological replicates were sequenced to the depth of 1 . 4×108 reads . All the data from mapping replication initiation events in K562 cells were deposited in GEO , as Series GSE28911 . We performed ChIP analyses with 1% formaldehyde-fixed K562 and RL4 cells using the Millipore ChIP assay kit ( Cat . no . 17-295 ) following the manufacturer's protocol . Anti-H3K79Me2 antibody was from Abcam ( ab3594 ) . We analyzed ChIP samples by real-time PCR in an ABI 7900 thermocycler using primers/probes listed in Table 3 . DNA combing analyses of replicating DNA were performed according to previously published methods [47] . Briefly , cells were pulse-labeled with 20 µM IdU ( Sigma , Cat . no . I-7125 ) for 20 minutes and then with 50 µM CldU ( MP biomedical , Cat . no . 105478 ) for 20 minutes before harvest . Then the cells were embedded in low-melting point agarose plugs and lysed in the plug with proteinase K lysis buffer at 50°C overnight . After agarose was digested with β-agarase ( NewEngland Biolabs ) , DNA was combed onto silanized surfaces ( Microsurfaces , Inc . ) and detected with anti-IdU ( BD , Cat . no . 347580 ) , anti-CldU ( Accurate Chemical , Cat . no . OBT0030 ) , and anti-single strand DNA ( Chemicon , Cat . no . MAB3034 ) antibodies . Images were captured with the Attovision software using the epifluorescence microscope Pathway ( Becton Dickinson ) and measured the signals using ImageJ ( open source from National Cancer Institute , NIH ) with custom-made macros [48] . Cells were cultured in 4-well chamber slides , pulse labeled with 10 µM EdU ( Invitrogen ) for 1 hour before harvest . EdU staining using the Click-iT EdU kit ( Invitrogen ) were performed according to manufacturer's protocol . Images were captured with the Attovision software using the epifluorescence microscope Pathway ( Becton Dickinson ) and measured with the Attovision software for DNA content by DAPI for cells with early S-phase and late-S phase EdU replication patterns . Asynchronous K562 cells or fractionated G1 , S and G2M K562 cells by elutriator ( see [12] for elutriation details ) were used for ChIP as described above . ChIP-Seq samples were sequenced using the Illumina GA II platform . The resultant sequences were aligned to the hg19 ( NCBI Build 37 ) human reference genome using Bowtie [49] . Alignments were converted to . bam and . tdf format using SAMtools and igvtools for visualization in Broad's Integrative Genomics Viewer , http://www . broadinstitute . org/igv/ [50] . Reads per kilobase per million aligned reads ( RPKM ) values were calculated for each sample using 100 base genomic bins [51] . A Gaussian smoothing algorithm was applied to the bin values . To correct for sequencing biases and copy number variation , an enrichment ratio was defined as the ratio of nascent strand RPKM to control RPKM , and calculated for each 100 base bin . The H3K79Me2 and H3K4Me3 ChIP-Seq data are deposited in GEO Series GSE GSE35294 including 11 sample files from experiments in unsynchronized and cell cycle fractionated cells . HCT116 cells transfected with control siRNA or Dot1L siRNA were pulse-labeled with 50 µM of BrdU for 18 hours before harvest . Genomic DNA were purified and sonicated to 500–4000 bp . Genomic DNA from HCT116 cells with no BrdU incorporation and BrdU incorporation for 48 hours was used as control . 100 µg of DNA was fractionated with 6 ml CsCl ( 1 g/ml in TE ) . Samples were spun at 45000 rpm in a Ti75 rotor ( Beckman ) for 66 hours . Fractions of 250 µl were collected and the refractory index was measured to confirm the formation of the gradient . Same volume samples from each fraction were loaded to a positive charged nylon membrane by a Slot Blot Filtration Manifold ( PR648 , GE Healthcare life sciences ) . 0 . 5 ng , 5 ng , 50 ng , 500 ng and 5000 ng of total genomic DNA labeled with BrdU for 48 hours were also loaded to the membrane as standard for BrdU . BrdU were detected with anti-BrdU antibody . | Before each cell division , cells must accurately duplicate their chromosomes . It is critical that cells coordinate the replication of DNA with the packaging of DNA into chromosomes to insure that all genetic and epigenetic information is accurately transmitted to the next generation . In eukaryotes , replication starts at multiple sites , called “replication origins , ” which are distributed throughout the genome and initiate replication in a strict order to maintain genomic stability and prevent cancer . Previous studies looked at the effect of chemical modifications on histone proteins , which affect chromosome packaging , on replication but no particular histone modifications distinctly associated with replication start sites . Here , we took advantage of recent advances in whole genome sequencing to map replication origins and histone modifications for the entire DNA in human cancer cells . One of the histone modifications we tested , methylation of lysine 79 on histone H3 , was remarkably enriched at a group of replication origins . Inhibiting the enzyme that catalyzes this histone modification caused some DNA to replicate more than once during a single cell cycle , suggesting that methylation of histone H3 on lysine 79 might play an important role in controlling DNA replication . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"gene",
"function",
"genome-wide",
"association",
"studies",
"genome",
"sequencing",
"functional",
"genomics",
"cell",
"division",
"rna",
"interference",
"nucleic",
"acids",
"cell",
"growth",
"genetics",
"epigenetics",
"biology",
"genomics",
"molecular",
"cell",
"biolog... | 2013 | Methylation of Histone H3 on Lysine 79 Associates with a Group of Replication Origins and Helps Limit DNA Replication Once per Cell Cycle |
Traditional faecal-based methods have poor sensitivity for the detection of S . stercoralis , therefore are inadequate for post-treatment evaluation of infected patients who should be carefully monitored to exclude the persistence of the infection . In a previous study , we demonstrated high accuracy of five serology tests for the screening and diagnosis of strongyloidiasis . Aim of this study is to evaluate the performance of the same five tests for the follow up of patients infected with S . stercoralis . Retrospective study on anonymized , cryo-preserved samples available at the Centre for Tropical Diseases ( Negrar , Verona , Italy ) . Samples were collected before and from 3 to 12 months after treatment . The samples were tested with two commercially-available ELISA tests ( IVD , Bordier ) , two techniques based on a recombinant antigen ( NIE-ELISA and NIE-LIPS ) and one in-house IFAT . The results of each test were evaluated both in relation to the results of fecal examination and to those of a composite reference standard ( classifying as positive a sample with positive stools and/or at least three positive serology tests ) . The associations between the independent variables age and time and the dependent variable value of serological test ( for all five tests ) , were analyzed by linear mixed-effects regression model . A high proportion of samples demonstrated for each test a seroreversion or a relevant decline ( optical density/relative light units halved or decrease of at least two titers for IFAT ) at follow up , results confirmed by the linear mixed effects model that showed a trend to seroreversion over time for all tests . In particular , IVD-ELISA ( almost 90% samples demonstrated relevant decline ) and IFAT ( almost 87% ) had the best performance . Considering only samples with a complete negativization , NIE-ELISA showed the best performance ( 72 . 5% seroreversion ) . Serology is useful for the follow up of patients infected with S . stercoralis and determining test of cure .
Strongyloides stercoralis infection is widely distributed in tropical , subtropical countries and even in areas of temperate climate [1] . Strongyloidiasis probably affects at least 370 million people worldwide [2] and represents a threat for immunosuppressed people , who tend to develop the fatal complications of the infection [1 , 3] . Therefore , it is mandatory to diagnose the infection during the chronic phase , which is often indolent and can be more easily treated [3] . The diagnosis of S . stercoralis infection is characterized by poor sensitivity of fecal-based methods [4] . Therefore , other diagnostic tools have been developed and demonstrated better sensitivity [4 , 5] . Polymerase chain reaction ( PCR ) is still based on in-house techniques [6–8] , performed only in reference centers , and is not necessarily more sensitive than fecal culture[9] . Serology is more sensitive , though not 100% specific [4] . Some serology kits are commercially available [10 , 11] . A high sensitivity is also necessary when evaluating the response to the treatment , as treatment failures leave the patient exposed to the risk of developing a potentially fatal , disseminated strongyloidiasis at any time in his/her life [2] . Negative fecal-based methods cannot safely exclude persistence of infection [4 , 12] , therefore the follow up of infected patients should also rely on more sensitive techniques as markers of cure . Although some authors have observed a decline of optical density ( OD ) /titers of serology tests over time , a wider comparative evaluation has not been carried out so far , and a clear definition of cure has not yet been established [13–20] . We recently published the results of a study comparing the accuracy of five serologic tests for the diagnosis of S . stercoralis infection [5]: two commercial ELISA tests ( Bordier ELISA , IVD-ELISA ) , two tests based on the recombinant antigen NIE ( ELISA and luciferase immunoprecipitation system , LIPS ) and one in-house indirect immunofluorescence antibody test ( IFAT ) . The study demonstrated a good performance of the tests , and in particular NIE-LIPS demonstrated the best accuracy for the diagnosis of S . stercoralis . The same tests were also evaluated on sera collected pre and post treatment in the present study . Thus , the aim of this study was to compare the performance of the five tests for the follow up of patients after treatment in order to identify if antibody decline could be used a surrogate marker for cure , in addition to stool negativization .
This was a retrospective study on archived , anonymized sera available at the Centre for Tropical Diseases ( CTD ) . Samples were classified according to a composite reference standard ( a procedure suggested for evaluation of diagnostic tests when there is no gold standard ) [21 , 22] as a ) positive: positive fecal tests and/or at least 3/5 positive serologic tests; b ) negative: negative fecal tests and less than 3 positive results out of the 5 serologic tests . The inclusion criteria were: samples resulting positive before treatment , according to the composite reference standard ) , and available follow up serum sample/s , from 3 to 12 months after treatment . Treatment administered was ivermectin ( stat dose of 200 μg/kg ) , with the exception of 6 cases treated with thiabendazole ( two daily doses of 25 mg/kg for two days ) in the earlier period . The exclusion criterion was travel history to endemic areas between treatment and follow up . The results of stool examination/agar culture were registered and entered in the study database . Parasitological tests used were: at least 3 stool samples examined with microscopy ( formol-ether concentration ) and Koga agar plate culture [23 , 24] . These methods were performed at the CTD . The serology tests evaluated were: the CTD in-house immunofluorescence technique ( IFAT ) [13] , two commercial ELISA tests ( Bordier ELISA [10] and IVD ELISA [11] ) and two techniques based on the recombinant antigen NIE ( NIE-ELISA [25] and NIE-LIPS [26] . IFAT and the two commercial ELISA tests were executed by senior staff of the CTD Negrar ( Verona ) , Italy , while NIE-LIPS and NIE-ELISA were up to senior staff of the National Institute of Allergy and Infectious Diseases ( NIAID ) of the National Institutes of Health ( NIH ) , Bethesda , US and of the Instituto de Investigaciones en Enfermedades Tropicales of the University of Salta/CONICET , Oran , Argentina . Lab staff were blinded to the patients’ data and to the results of the other tests . Cure was operationally defined by negative composite reference standard ( see above ) at follow up or at least by: negative stool examination/coproculture and decrease of at least half of initial eosinophil count . For the evaluation of each test , we assessed , over the denominator of patients cured according to the operational definition reported above: a ) the proportion of initially positive tests that were negative at follow up; b ) the proportion of those showing a decrease of at least half of initial OD/relative light units ( RLU ) values ( for ELISA tests and LIPS , respectively ) or decrease of at least two titers ( for IFAT ) . This was taken as an empirical measure of response to therapy . The STARD flow chart ( Fig . 1 ) describes the selection of the samples tested . Among the 130 subjects responding to our definition of positive , 8 were excluded because follow up samples were not available . Of the remaining 122 , 6 had a positive fecal result at follow up . Of the 116 testing negative at follow up , 98 met the criterion of cure as defined above , of which: 57 were negative according to the composite reference standard , and 41 showed a decrease of at least half of initial eosinophil count . Two subjects were excluded because their follow up sample was collected less than 3 months after the baseline sample . Eventually , 96 subjects were included in the analysis . Primarily , the performance of each test was calculated as the proportion of samples demonstrating seroreversion or a quantitative decrease ( as indicated above ) over all positive samples ( for the same test ) at baseline . Uncertainty was quantified using 95% confidence intervals . To reduce the limitations due to the different time intervals between treatment and observation ( from 3 to 12 months ) , we used the following methods . The associations between the independent variables age and time and the dependent variable value of serological test ( for all five tests ) , were analyzed by linear mixed-effects regression model . Linear mixed model is a generalization of traditional linear regression , which adjusts for the correlation between repeated measurements within each subject and finds the best linear fit to the data across all individuals [27 , 28] . More specifically , a unique identification number for each subject and time was treated as a random effect in the model and age was treated as fixed effect . Time was entered as random effect because measurements of the value of serological tests over time were not taken at regular time points . Interaction term between age and time was evaluated to include in the regression model by using Likelihood Ratio Test . Introduction of an interaction term is necessary where the effect of one variable ( time ) is affected by the presence or value of another variable ( age ) . Unstructured covariance matrix was selected since this is the structure that appears to fit the data the best , based upon the Akaike’s information criterion ( AIC ) . Analyses were done by using SAS ( version 9 . 1; SAS Institute , Inc , Cary , NC ) . We considered differences to be statistically significant when the p-value was <0 . 05 . Although this was a retrospective study on anonymously coded , cryo-preserved samples , the study protocol was nevertheless submitted to the Ethics Committee of the Coordinating Site ( Comitato Etico Provinciale di Verona ) for approval . The latter acknowledged the study protocol and formally authorized the study ( protocol n . 13286/09 . 11 . 01 of 24th April , 2012 ) .
The sample selection and the laboratory analyses were performed during the second semester of 2012 . The median age of the population considered was 42 years ( IQ range 22 . 5–67 ) . Table 1 shows the time ( in months ) elapsed from baseline to follow up . Every patient had a baseline evaluation both with serology and with parasitological methods . Only 9/96 ( 0 . 9% ) patients had negative stools at baseline; according to the composite reference standard , these patients were included in the analysis because they had at least 3 out of 5 positive serologic results . All but these 9 patients , had also parasitological evaluation at the time of collection of the follow up serum sample . All had negative stool microscopy and culture at follow up ( data not reported in Table 1 ) , as this was the first required criterion for the definition of cure . For each time frame , it is also showed the number of patient who had positive versus negative stool microscopy and culture at baseline . Table 2 shows , for each test , the percentage of serum samples showing response according to the pre-defined criteria . For each serologic test , we considered for this analysis only the samples that were positive at baseline . For instance , among the 96 samples resulting positive at baseline according to the composite reference standard , 91 had a positive IFAT result ( see column “Positives at baseline” ) . The column “Negativization” comprises the samples which were positive at baseline and negative at follow-up , while the column “Response” includes the latter , plus the samples that , albeit remaining positive , showed a decrease of at least half of initial OD/relative light units ( RLU ) values ( for ELISA tests and LIPS , respectively ) or two titers ( for IFAT ) . IVD-ELISA ( almost 90% samples demonstrated response ) and IFAT ( almost 87% ) had the best performance . When considering only samples with a complete negativization , NIE-ELISA showed the best performance ( 72 . 5% of seroreversion ) . Total baseline samples positive at composite reference standard: 96 . Negativization concerns for each test the samples that were positive at baseline and negative at follow-up; response also includes samples that , albeit not yet negative at follow-up , showed a decrease in OD , RLU or titer , respectively , as explained in the text . Figs . 2 to 6 show the results of the mixed effects model for all five serological tests . They represent the prediction of the trends of the values of serology , from the baseline evaluation ( 0 on the x-axis ) to the moment in which the result became negative ( 0 on the y-axis ) . Thus , significantly negative trends over time were detected for all tests . Moreover , the intersection of the interpolation line with the x-axis predicts the average time ( days ) required to obtain the negativization of the serology test . Therefore , NIE-ELISA and IVD-ELISA showed the most rapid predicted negativization ( about 1 year from baseline evaluation ) . Interaction terms between age and time were not statistically significant , meaning that effect of time was not affected by age in the outcome variable .
Based on the operational case definition of cure , we obtained the denominator of “cured” patients on which we assessed the decline in titer of the different serologic tests . In the absence of a gold standard for cure , we cannot rule out that some patients might have been misclassified , i . e . considered cured when they were not , also considering that the eosinophil count can fluctuate . It is therefore possible that in some cases the lack of serologic response to cure could be due to misclassification . Moreover , the follow up samples were available at different time intervals after treatment , because of the retrospective design of the study . A three-month time could be a period of time too short to observe a decrease in the values of serology , therefore it cannot be excluded that a longer and more homogeneous period of observation would have demonstrated better performance of the tests in terms of percentage of seroreversion ( as seen in Table 2 ) . However , the application of the mixed effects model permitted to have a prediction of the decrease over time , making it possible to demonstrate a tendency to seroreversion for all tests . Another limitation is related to the different treatment used ( ivermectin or thiabendazole ) . Although the two drugs demonstrated a comparable efficacy [30] we cannot exclude a difference in the rapidity of the response to treatment . However , the patients treated with thiabendazole were just a few ( 6 subjects ) , thus not allowing a separate analysis . Our results demonstrate that each of the serology tests considered can be used for monitoring patients who received a treatment for S . stercoralis infection . Serology , in combination with fecal-based methods , should be used as the preferred tool for the follow up . Validation of PCR techniques for the follow up might be a useful support for situations of uncertainty ( such as patients with serology values that do not seem to decrease over time ) . Further investigations are necessary to extend these considerations to endemic areas , where re-infection might be an issue . | Patients infected by S . stercoralis are at risk of fatal complications . It is therefore mandatory to demonstrate complete response to therapy . Post treatment evaluation should be done with highly sensitive diagnostic methods , which can exclude the persistence of the infection . Serology is more sensitive than fecal examination and coproculture . In this study , we compare the post-treatment performance of five serology tests , and suggest that they can be useful for the follow up of patients with S . stercoralis infection , especially in non-endemic areas , where there is no risk of reinfection . In fact , the results of the tests show a progressive decrease , towards negativization , of the values ( expressed in different units , depending on the specific test ) through time . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Accuracy of Five Serologic Tests for the Follow up of Strongyloides stercoralis Infection |
Many patterns of affiliative behaviour have been described for primates , for instance: reciprocation and exchange of grooming , grooming others of similar rank , reconciliation of fights , and preferential reconciliation with more valuable partners . For these patterns several functions and underlying cognitive processes have been suggested . It is , however , difficult to imagine how animals may combine these diverse considerations in their mind . Although the co-variation hypothesis , by limiting the social possibilities an individual has , constrains the number of cognitive considerations an individual has to take , it does not present an integrated theory of affiliative patterns either . In the present paper , after surveying patterns of affiliation in egalitarian and despotic macaques , we use an individual-based model with a high potential for self-organisation as a starting point for such an integrative approach . In our model , called GrooFiWorld , individuals group and , upon meeting each other , may perform a dominance interaction of which the outcomes of winning and losing are self-reinforcing . Besides , if individuals think they will be defeated , they consider grooming others . Here , the greater their anxiety is , the greater their “motivation” to groom others . Our model generates patterns similar to many affiliative patterns of empirical data . By merely increasing the intensity of aggression , affiliative patterns in the model change from those resembling egalitarian macaques to those resembling despotic ones . Our model produces such patterns without assuming in the mind of the individual the specific cognitive processes that are usually thought to underlie these patterns ( such as recordkeeping of the acts given and received , a tendency to exchange , memory of the former fight , selective attraction to the former opponent , and estimation of the value of a relationship ) . Our model can be used as a null model to increase our understanding of affiliative behaviour among primates , in particular macaques .
Patterns of affiliative behaviour have long puzzled primatologists . One of the most frequent behavioural acts is grooming . It has been explained as serving several functions , such as cleaning the fur [1] , reducing anxiety , tension and stress [2] , social bonding [3] , repairing relationships [4] and social reciprocation and exchange [5] . As regards the mechanisms of exchange , individuals have been supposed to direct grooming up the hierarchy in order to receive more effective support in return , and due to competition for partners of high rank they may end up grooming others of similar rank [6] . Besides , they were also supposed to groom others of similar rank , because individuals of similar rank have similar needs [7] . Grooming between two former opponents immediately after a fight has been interpreted to function as a means to repair the relationship or ‘reconcile’ , because it occurred significantly earlier after a fight than otherwise in matching control periods the next day . Besides , individuals appeared to reconcile in particular with those partners that appeared more valuable to them , the so-called ‘valuable-relationship hypothesis’ [8] . To complicate matters , the degree of exchange and reciprocation [9] appeared to differ between egalitarian and despotic species . Applying market theory [10] , [11] , this was explained by assuming that the exchange rate of services differed between the two competitive regimes [9] . Further , the co-variation hypothesis explained the lower conciliatory tendency in despotic societies by the greater danger involved in reconciliation in these species [12] . Many specific cognitive considerations have been suggested to underlie these affiliative patterns . For instance , as regards reciprocity and exchange , the individuals are supposed to keep records of the acts of grooming and tune them to frequencies of receipt of being groomed or another act , such as support [13] , and to use their knowledge of the ranks of others to obtain more effective support [6] , [14] . Besides , individuals have been supposed to be attracted to others of higher rank [6] and to others of similar rank [7] . The supposed cognition underlying reconciliation consists of the ability to remember the former opponent and of selective attraction to the former opponent or a conciliatory disposition [15] , [16] . As to their inclination particularly to reconcile fights with opponents that are of greater value to them , the so-called ‘valuable-relationship hypothesis’ [8] , [17]–[23] , three key components are supposed to influence the quality of a relationship , namely its security , its value , and the compatibility of both partners [8] , [24] . According to Silk [25] this implies that assessing the value of a relationship over the long-term requires cognitive sophistication , because it asks for a precise memory of what happened in the past and for a correct evaluation of the relationship in the long run . These theories of affiliation pose several problems . First , evidence for each of these theories is not conclusive [5] , [26]–[30] . Second , from a scientific perspective , these numerous different theories for specific patterns of affiliation ( such as exchange and reconciliation ) must be integrated in some way . Third , the use of grooming as a ‘currency of exchange’ is dangerously anthropomorphic according to us and others [25] , [31] , [32] . As a more parsimonious alternative , we suggest to follow a more distributed approach based on local interactions and rules of thumb [31] , [33]–[36] . Fourth , even though primates are obviously intelligent [37] , [38] it seems much to ask of primates to combine intentionally all these rational considerations in the distribution of their affiliative behaviour ( e . g . to consider what incidence of grooming was used in exchange for something , and what for reconciliation or maintenance and development of social bonds ) . Fifth , often simple rules suffice to cause many of the observed patterns and herewith an integrative theory [39] , [40] . Therefore fewer cognitive processes may suffice as shown for instance in a model for dominance style [32] , [41] . A similar integrative approach based on fewer cognitive processes is also suggested by the co-variation hypothesis ( or theory of social epigenesis ) . In this theory part of the behavioural acts is supposed to be forced by constraints due to the specifics of the social structure [12] . For these reasons , we use in the present paper a computer model to develop an integrative approach to patterns of social affiliation in primates . We first precede this by a survey of the precise patterns of dominance style and affiliation found in the literature . In the model , we assume very little cognitive deliberations by the individuals to groom others: Individuals merely groom others out of fear of being defeated and to reduce their own anxiety . Individuals do not intend to reconcile fights nor to exchange or reciprocate grooming . Our model is an extension of our earlier model of grouping and competition , called DomWorld [42] , [43] . We choose DomWorld , because it has reproduced many of the patterns of aggression , dominance and spatial structure that have been observed in despotic and egalitarian societies of primates , in particular of macaques . These have arisen merely as a side effect of local rules for grouping and competition through the feedback between hierarchical development and spatial-social structure with dominants in the centre and subordinates at the periphery [35] , [41] , [44]–[46] . Note that the hierarchy develops via self-reinforcing effects of victory and defeat , which have been described for many species including primates [45] , [47]–[50] . Through these self-reinforcing effects , occasional victories of low ranking individuals may lead to rank reversals . This is important , because dominance hierarchies in empirical data are not entirely stable [51]–[55] . Interactions in our new model , called GrooFiWorld ( a contraction of groom and fight ) , are extended with the option to groom . When individuals meet each other at close proximity , they will consider whether to groom , to fight or to rest . As to the order of what to do first , we are led by four observations: first , those on baboons by Kummer [56] who inform us that upon their first encounter individuals first fight and later groom; second , by the empirical finding that an individual builds up anxiety ( as indicated by the increased heart rate ) when approaching an opponent by whom it may be defeated [rhesus monkeys , 57]; third , that anxiety increases after a fight as is indicated by the increase in frequency of scratching and heart rate in both opponents [25] , [58]–[64]; fourth , that anxiety may subsequently be reduced ( in many species ) by the receipt of affiliative behaviour as indicated by the reduced heart rate and the rate of self-directed behaviour [57] , [61] , [62] , [64] and to a lesser degree by active grooming [65] . Furthermore , our model is informed by empirical studies on grooming and opiate administration which indicate that not being groomed for some length of time reduces the concentration of endorphins and increases the motivation to be groomed , and that grooming increases the level of endorphins in the brain and reduces the motivation to groom [66]–[71] . In sum upon encountering someone else , an individual in our model first deliberates whether or not to attack . This decision depends on the risks involved ( whereby risk concerns the chance of losing a fight ) , as is the case among primates [72] , and as in our earlier model: a fight is only initiated when the individual expects to win [41] , [73] . If defeat is expected , its fear of losing makes the individual consider grooming the other . Its decision whether or not to groom depends on its degree of anxiety: an individual that is more anxious is more inclined to groom ( instead of resting close by ) . After being groomed by another and ( a little less ) after actively grooming another , its anxiety and therefore its tendency to groom diminishes . Its anxiety also increases after a fight and after a period of not having been involved in grooming . Note that we do not distinguish between anxiety , tension or stress . In order to compare the patterns of affiliative behaviour in our model with those in real primates , we used the same statistical measures as applied in empirical data and we confined ourselves to macaques for two reasons . First , because their social behaviour has been studied extensively and shown to differ in interesting ways between the typical egalitarian and despotic societies [74] , [75] . Second , because in our earlier model , DomWorld patterns of dominance and aggressive interaction were remarkably similar to those of macaques [41] , [45] . Since GrooFiWorld is an extension of this model , we assume it to also be suitable for comparing to macaques . The paper is organized as follows . First , we summarise the literature on the common patterns of affiliative behaviour in females of egalitarian and despotic species of macaques ( Table 1 ) . Second , we tune the percentage of grooming time and the unexpectedly emerging percentage of reconciliation to empirical data for despotic societies . Third , by varying the intensity of aggression we show the emergence of all these common patterns of affiliation and their differences between typical egalitarian and despotic macaque species in GrooFiWorld . Fourth , in order to understand how these patterns emerge , we remove different assumptions in turn , such as the self-reinforcing effects of victory and defeat and effects of spatial proximity . Fifth , the explanation of the causation of these patterns in the model leads to new hypotheses about the interconnection between other traits which we confirm in the model . Part of these predicted patterns appear also to be found in empirical data described by scientists in other contexts . Other patterns still need to be tested empirically . Since for all patterns empirical data are insufficient , we list them together in Table 2 so that the relevance of our model to empirical data may be tested in the future .
A demo of our model can be seen in Video S1 . The model is individual-oriented and event-driven [76] . It has been written in C++ , as an extension of a previous model of grouping and competition , called Dom-World [41] , [42] , [77] , [78] which has been reimplemented in C++ by Hanno Hildenbrandt . The extension consists of the addition of grooming behaviour ( for default parameters see Table 3 ) . Therefore , we call it ‘GrooFiWorld’ . The individuals are provided with three tendencies: 1 ) to group , 2 ) to perform dominance interactions and 3 ) to display affiliative behaviour . Why individuals actually group ( whether to avoid predators or because resources are clumped ) is not specified and irrelevant to the model . The same holds for dominance interactions which may reflect competition for resources such as food and mates , but these resources are not explicitly specified in the model . Individuals groom to reduce Anxiety , as suggested for real primates [66]–[71] , [79]–[81] . GrooFiWorld consists of a ‘world’ ( without borders ) containing its interacting individuals . The space of the ‘world’ is continuous , i . e . individuals are free to move in any direction . Individuals have a certain angle of vision ( VisionAngle ) and a maximum distance of perception ( MaxView ) . At the start of each run they occupy random locations within a predefined circumference , InitRadius , which is the product of an arbitrary constant times the number of individuals . Activities of individuals are regulated by a timing regime in which each individual receives a random waiting time from a uniform distribution and the individual with the shortest waiting time is activated first . This regime is combined with a biologically plausible timing regime reflecting a kind of social facilitation [e . g . , see 82] in which the waiting time of an individual is shortened when a dominance interaction occurs close by ( within the individual's NearView ) . Whenever an individual does not see another one close by ( within its personal space , PersSpace ) , grouping rules come into effect . The individual starts looking for others at greater and greater distances ( NearView and MaxView ) . If , even then , no one else is in sight , it turns over a SearchAngle in order to scan for others . In this way individuals tend to remain in a group ( Figure 1 and Table 3 ) . If , however , an individual spots another one close by , within its personal space ( PersSpace ) , a social interaction may take place . Upon encountering someone else the individual first deliberates whether or not it will perform a dominance interaction on the basis of the risk of losing the fight [following the so–called ‘risk–sensitive attack strategy’ , 43] . Only if it expects to be defeated , it will consider grooming . In real primates , motivation to groom depends on opiate concentrations as well as on other physiological conditions such as stress levels , and we have coded these factors together as Anxiety [66]–[68] , [71] , [83] ( Figure 1 ) . Thus , in GrooFiWorld , first , the more anxious an individual is the more likely it is to groom ( instead of resting close by ) ; second , after being groomed and ( a little less ) after actively grooming another , an individual's anxiety and thus its tendency to groom declines; third , after not having been involved in grooming for some time an individual's anxiety builds up again; and fourth , an individual's anxiety grows after a fight . Thus anxiety reflects the psychological and physiological state of an individual . Dominance interactions are modelled as before [41] , [46] and they are an extension of the DoDom rules of Hogeweg [46] . First , an individual i estimates whether it will win on the basis of a ‘mental battle’ ( Equation 1 ) . It may do so once [84] or repeatedly depending on its degree of sensitivity to risks ( RiskSens Table 3 and Parameters and Experimental Setup ) . Higher values of RiskSens indicate that individuals need to win several mental fights before starting an actual interaction . Here , individuals i and j observe each other's capacity of winning , i . e . their dominance values Domi and Domj . The probability of winning for individual i is greater if it is higher in rank , and this is proportional to the Dom-value of individual i relative to that of its opponent j ( Equation 1 ) . It expects to be victorious if its relative dominance value is greater than a random value drawn from a uniform distribution between zero and one . If this is the case , a dominance interaction takes place . During the actual dominance interaction , the individual i compares its relative dominance value again with another value randomly drawn and if its relative dominance value is greater than a new random number , it wins ( wi = 1 ) , else it loses ( wi = 0 ) : ( 1 ) The stochastic effect is introduced to allow for dominance reversals . To reflect the self-reinforcing effects of victory and defeat [45] , [85] , dominance values are updated by increasing the dominance value of the winner and decreasing that of the loser by the same amount: ( 2 ) This positive feedback is ‘dampened’ because a victory of a higher ranking opponent increases its relative Dom-value only slightly , whereas an ( unexpected ) success of the lower ranking individual increases its relative dominance value by a greater change . To keep Dom-values positive , their minimum value is , arbitrarily , set at 0 . 01 . The change in Dom-values is multiplied by a scaling factor , called StepDom , which varies between 0 and 1 and represents the intensity of aggression [41] , [84] ( see Parameters and Experimental setup ) . High values imply a great change in Dom-value after a fight , and thus indicate that single interactions ( e . g . involving biting ) may strongly influence the future outcome of conflicts . Conversely , low StepDom-values represent low impact ( e . g . threats or slaps ) . Winning an interaction includes chasing the opponent over a distance of one unit and then turning randomly 45 degrees to right or left in order to reduce the chance of repeated interactions between the same opponents . The loser responds by fleeing under a small random angle over a predefined FleeingDistance . If an individual meets another in its PersSpace and when it has decided on the basis of a ‘mental’ battle that it is too dangerous to attack , the individual considers whether or not to groom its partner ( Figure 1 ) . Grooming behaviour is induced by the level of Anxiety , which ranges from very relaxed to very tense , represented by a scale from 0 to 1 . If the Anxiety value is lower than a random number , the individual will display ‘non-aggressive’ proximity; otherwise , if Anxiety is higher , it will groom its partner ( Figure 1 ) . After grooming both partners turn over a small angle ( 45° ) randomly to the right or left in order to avoid repeated interactions with the same partner . Grooming reduces Anxiety . In line with empirical evidence [61] , [62] , [64]–[71] , it does so more strongly in the groomee ( with AnxDcrGree ) than in the groomer ( with AnxDcrGrmr ) ( Table 3 ) . During periods without grooming Anxiety increases , which is consistent with opiate-dependent motivation to groom in real primates [67] , [71] . This increase is updated after every activation with AnxInc . Furthermore , inspired by the observed increase in scratching after a fight in real primates [8] , in the model , after a fight Anxiety increases with AnxIncFght for both opponents . Many parameters that have been used in earlier studies were kept at the same value , namely the NearView , MaxView , FleeingDist , SearchAngle and StepDom values . Note that StepDom values ( that reflect intensity of aggression ) differ between the sexes ( on the basis of the stronger muscular structure of males than females ) and between dominance styles reflecting the tendency of individuals in despotic societies to bite relatively more ( and slap and threaten less ) than in egalitarian ones [41] , [42][78][84][86] ( Table 3 ) . We used 12 individuals to represent the number of adults in a group of primates . Since in empirical studies the percentage of females is lower in egalitarian macaques with approximately 70% females than despotic macaques with approximately 80% females , we have set the sex ratio at low and high aggression intensity accordingly ( with 8 females , 4 males at low intensity and 10 females and 2 males at high intensity ) [87]–[89] . The initial dominance values we set at 16 for females and 32 for males , reflecting the initially higher winning chance of males due to sexual dimorphism in fighting power resulting from differences in body weight , physiology and weaponry . In order to tune the frequency of grooming to 20% of the time as indicated for despotic societies of real primates by Dunbar [90] , we had to increase PerSpace from 4 to 8 units ( reflecting a tendency to interact with others over larger distances ) , to increase the risk-sensitivity of individuals by increasing the number of mental battles ‘ego’ had to win before starting a real dominance interaction ( in order to reduce the frequency of aggression ) ( RiskSens , Table 3 ) and to tune the Anxiety-related parameters ( see Table 3 ) . To understand what caused the patterns of affiliation in the model , each of four assumptions related to grooming and fighting were switched off in turn . The simulation was run in turn 1 ) without the self-reinforcing effects of winning and losing fights , 2 ) without the grooming inducing effect of anxiety , 3 ) without the dependence of grooming on the risks to attack and 4 ) without the selection of interaction partners on the basis of spatial proximity . First , when switching off the self-reinforcing effects of winning and losing , we gave the individuals Dom values that were constant . We took these values from runs with the corresponding intensity of aggression , because hierarchical differentiation was greater at a high than at a low StepDom . We took the values from the middle ( i . e . period 230 ) of the interval between periods 200 and 260 , because in this interval the Dom values were considered to have stabilised [91] , since the average Dom values between period 200 and 230 are significantly correlated with those between 230 and 260 ( Kendall Tau , N = 10 , High intensity Tau = 0 . 88 *** , Low intensity Tau = 0 . 85** two tailed probability ) . Second , to switch off the grooming inducing effect of Anxiety implies that we made grooming independent of the value of Anxiety . In this case , the individual always groomed its partner whenever it refrained from attack because of high risks . Third , switching off fear-based grooming , implied that we made grooming independent of the risks of defeat , i . e . upon meeting another individual in its PerSpace we gave the individual a 50% chance to either consider grooming it or attacking it . After choosing between attacking and grooming , the risk-sensitive decision procedure was used to decide whether the individual actually attacked and the anxiety-based rule was used to decide whether it actually groomed . If the individual decided to refrain from interacting , it rested at its location . Fourth , to switch off proximity-based interactions , interaction partners were chosen at random independent of their proximity in space . Every run consisted of 260 periods and each period consisted of 240 activations ( the number of individuals ( i . e . 12 ) times 20 ) . Data were collected from period 200 to 260 to exclude any bias caused by transient values . Data consisted of every change in spatial position and in heading direction of each individual and , as regards social interactions , we recorded ( 1 ) the identity of the attacker and its opponent , ( 2 ) that of the winner/loser , ( 3 ) the updated Dom values of the individuals , ( 4 ) the identity of the groomer and groomee and ( 5 ) the updated Anxiety value of the individuals . For each model ( the complete model and the four controls with a missing assumption ) 10 independent simulations were run for each of the two aggression intensities ( high and low ) . The results are shown here per condition as the average statistic of these 10 runs with their combined probability using the improved Bonferroni procedure [92] . Patterns apparent in empirical studies of egalitarian and despotic macaques ( Table 1 ) were tested for by means of ( combined ) one-tailed probabilities ( Tables 4 and 5 ) , all the other patterns were tested with two-tailed probabilities ( Tables 6 and 7 ) . To test for differences in patterns between high and low intensity of aggression , we used the Mann Whitney U test whereby we compared the statistics between 10 runs at a high and 10 runs at a low intensity of aggression ( see Tables 4 , 5 , 6 ) . The percentage of time females spend in fighting ( or in grooming ) is calculated as the number of fights ( grooming ) in the group divided between the total number of activations . The percentage of interaction time spend in grooming is the frequency of grooming divided by that of grooming plus that of fighting . The hierarchical differentiation among all females was measured by the coefficient of variation of Dom values ( standard deviation divided by the mean ) . For each run the average value was calculated ( over period 200–260 ) and this was averaged over 10 runs . Higher values indicate greater rank distances among individuals [41] . Hierarchical differentiation is also reflected in the empirical behavioural measure of the degree of unidirectionality of aggression [12] , [93] , which we show also ( Table 4 ) . The diversity of partners with whom a female interacts is measured by the Berger-Parker dominance index [94] by dividing an individual's frequency of grooming with its most favourite partner by its total grooming frequency . The rank of group members we calculated as the average Dom value of each individual per run over periods 200–260 . We used an average measure , because we correlated it with an average measure of aggressive and affiliative acts , i . e . data were summed over the whole interval of period 200–260 . Apart from the average dominance value as a measure of rank we applied also a behavioural measure used in empirical studies , namely the average percentage of winning [95] . The degree to which dominant individuals ( both males and females ) occupy the centre was measured by a correlation between an individual's average Dom value and the average spatial direction of others around it . The centrality of each individual is calculated by means of circular statistics by drawing a unit circle around ego and projecting the direction of other group members as points on the circumference of this circle [96] . The connection of these points with ego's location results in vectors . The length of the mean vector represents the degree to which group members relative to ego form a cluster . Thus , longer mean vectors indicate a more peripheral , and hence , less central location of ego . Therefore , centrality of dominants is represented by a negative correlation between rank and the length of average vector ( indicating the average direction of others ) . In empirical studies reconciliation has been measured often by the PC-MC method ( i . e . Post-Conflict versus Matched-Control ) . Here , we have used its improved version [16] , [97] . In it a comparison is made between the moment in which grooming occurs during a short interval after a conflict , the so-called Post-Conflict period , and the moment it occurs in a control period of the same length ( ten minutes ) , the Matched-Control period , taken a day later at the same time . Because our model is event-driven , we use its average number of fights over the interval 200–260 ( of 2196 acts at high intensity ) and the average frequency of fights per hour of rhesus monkeys of 0 . 2 per hour per individual [98] and a day length of about 13h to estimate that the interval of ten minutes is approximated by three activations in the model ( one activation is about 3 . 8 min ) and the interval of one day is approximately represented by one period . These are of course abstractions but results appear to be robust ( see Sensitivity to parameter changes ) . Dyads were classified as ‘attracted’ , when grooming happened earlier in the Post-Conflict period than in the Matched-Control . Pairs were classified as ‘dispersed’ , when grooming happened in the reverse way , and as ‘neutral’ , when grooming happened exactly at the same time or did not happen at all . Following [97] , we calculated the conciliatory tendency , CT , of the group as:To measure the conciliatory tendency of each female with each of its group members , we calculated per pair the number of times they groomed sooner after a fight than in the matched control ( attracted events ) minus the cases where the opposite happened ( dispersed events ) divided by the total number of fights of the pair . Correlations between the distribution of grooming , proximity , aggression and reconciliation among females , and between social interactions and rank were measured by means of the Tau-Kr correlation as described by Hemelrijk [93] , [99] , which is frequently used in studies of animal behaviour [100]–[102] . The advantage of this statistic is that it is animal-centred , because it takes variation in grooming and aggression between individuals into account by measuring the correlation between the corresponding rows of two social interaction matrices and because it accounts for the dependence of data in an interaction matrix . The level of significance was calculated using 2000 permutations [93] , [99] . We tested for unidirectionality of attack and reciprocity of grooming by correlating an actor and receiver matrix with the Tau-Kr correlation . Note that unidirectionality and reciprocity are opposite correlations: a significantly negative correlation implies unidirectionality , whereas a significantly positive correlation implies reciprocity [93] . Whether grooming was directed up the hierarchy and to partners of similar rank was computed by the Tau-Kr correlation between , on the one hand , the grooming matrix and , on the other hand , respectively , the partner-rank-matrix ( with the average Dom values of grooming partners in the rows ) and the similar-rank-matrix ( filled with zeros apart from the partners closest and second closest in rank , which are indicated as 1's ) . Note that the higher-ranking individuals have higher Dom values . Thus , a significantly positive correlation with the partner-rank-matrix corresponds to grooming being directed up the hierarchy , and a significantly positive correlation with the similar-rank-matrix corresponds to a high degree of grooming among individuals of similar rank . To test the valuable-relationship hypothesis , we defined valuable relationships on the basis of the grooming frequency as is done by primatologists [e . g . 20] , [21] , [e . g . 23] . We used the frequency of grooming that occurred per dyad outside of the context of reconciliation in order for correlations with reconciliation not to be circular . We determined the degree of reconciliation with valuable partners by means of the Tau-Kr correlation between the matrices of the conciliatory tendency per dyad and that of the frequency of grooming per dyad outside the context of fighting ( by subtracting the acts of conciliatory grooming from the matrix with all grooming acts ) . A significant positive correlation reflects that reconciliation is more frequent with partners that are more valuable .
As regards our distinction of macaques in egalitarian and despotic , we updated the classification of Thierry [103] , [104] with new data on Tibetan macaques ( Macaca thibetana ) [51] and Assamese macaques ( Macaca assamensis ) [105] . Therefore , we rated as egalitarian Barbary macaques ( Macaca sylvanus ) , bonnet macaques ( Macaca radiata ) , stumptailed macaques ( Macaca arctoides ) , lion-tailed macaque ( Macaca silenus ) , Celebes crested macaque ( Macaca nigra ) and tonkean macaques ( Macaca tonkeana ) and as despotic we classified long-tailed macaques ( Macaca fascicularis ) , Japanese macaques ( Macaca fuscata ) , rhesus macaques ( Macaca mulatta ) , pigtailed macaques ( M . nemestrina ) , Tibetan macaques and Assamese macaques . Regarding the dominance style ( Table 1 ) , the frequency of unidirectional aggression , which is a measurement related to the hierarchical gradient in macaques , appears to be higher in despotic than in egalitarian species [12]; further , frequency of aggression is lower [103] and average distance among all females is greater [103] , [106] . Besides , for the despotic Japanese macaques , it has been reported that dominants are in the center of the group [107]–[109] . As to affiliative patterns , reconciliation occurs in both types of species , and is more frequent in egalitarian species [12] . Grooming is reciprocated in both egalitarian and despotic species . Further , grooming is directed up the hierarchy and to others of similar rank only in despotic species . Reconciliation is directed significantly more often to partners that are more valuable in several despotic species and according to a single study also in an egalitarian species , Macaca arctoides [19] . As described in the methods , we first tuned the percentage of time spent on grooming at a high intensity of aggression so that it resembled that of empirical data for despotic macaques [90] . Subsequently , we , unexpectedly , observed reconciliation . Since there are more precise data on the conciliatory tendency of despotic macaques than on their percentage of time spent on grooming , we subsequently tuned the conciliatory tendency to that of despotic macaques by adjusting the risk sensitivity further ( 7 in Table 4 ) . As to the two dominance styles in our model , we first confirmed that they still emerged , like they did in the earlier DomWorld model without grooming [41] , [43] . In GrooFiWorld , at a high intensity of aggression , the hierarchy appeared to be significantly steeper than at a low intensity , aggression was more unidirectional , time spent on fighting was less , average distance among all females was greater and the spatial structure with dominants in the centre and subordinates at the periphery was more conspicuous ( 1–5 in Table 4; 1 , 5–7 in Table 2C ) . We confirm the resemblance of the affiliative patterns in the model to empirical data ( Tables 1 , 4 ) : The conciliatory tendency appeared to be significantly higher at a low aggression intensity than at a high one ( 7 in Table 4 ) ; grooming appeared to be reciprocated at both intensities ( 8 in Table 4 ) ; a number of significant correlations were confined to a high aggression intensity , namely individuals direct their grooming significantly 1 ) up the hierarchy , 2 ) to others of similar rank , and 3 ) they reconcile more often with more valuable ( grooming ) partners ( 9–11 in Table 4 ) . The only difference to empirical data concerns the absence in the model of more frequent reconciliation with valuable partners at low aggression intensity ( 11 in Table 4 ) . However , in empirical data this correlation for the valuable relationship hypothesis was found only in a single empirical study of an egalitarian species [19] and it was based on a different method , i . e . the time rule method , whereas in the model we use the MC-PC method . In order to understand what caused these patterns of affiliation in the model , we took out four different assumptions in turn ( see Parameters and Experimental Set-up ) . This reduced the number of emergent patterns . Switching off the self-reinforcing effect of the outcome of a fight did not affect the patterns qualitatively , but switching off the grooming-inducing effect of Anxiety changed three patterns of the 28 ( 11% ) ( indicated in bold in Table 5 ) . Making grooming independent of fear of defeat changed seven patterns ( 29% ) and choosing partners at random independent of spatial proximity changed 20 patterns ( 75% ) . Thus patterns arose mainly from the social-spatial group structure and secondly from grooming being dependent on fear of defeat . To explain the emergence of each of the affiliative patterns in the model ( Table 4 ) , we proceed now by studying the effects of each of the four above-mentioned assumptions by taking them out ( Table 5 ) . This process leads to a number of model-based hypotheses for empirical data ( Table 2 ) . The emergence of grooming up the hierarchy depended on grooming being based on fear of being defeated ( without this assumption the pattern disappeared ) and on the intensity of aggression ( since it is absent at a low intensity of attack ) . This arises because the hierarchical differentiation is stronger at a high aggression intensity compared to a low one , and aggression is more unidirectional ( 1 , 2 in Tables 4 and 5 ) . Thus , when grooming depends on fear of defeat and the difference in rank between the partners is high , lower ranking ones will usually groom higher ranking ones and rarely attack them ( as a consequence of Eq 1 ) . Grooming reciprocation ( 8 in Table 4 , 5 ) arose from spatial structure , because it was disrupted only by taking out the socio-spatial structure , but not by taking out any of the other three assumptions . This means that , because certain individuals are often in close proximity , they will groom each other mutually , resulting in reciprocation . Furthermore , reciprocation appeared to be strongest in the experimental control condition where grooming did not depend on fear of defeat , and next strongest at a low aggression intensity . This arose because reciprocation was weakened less by differences in dominance , because these are smaller at low intensity of aggression ( 1 , 2 in Table 4 and 5 ) . Besides , at high aggression intensity , but not at a low one , individuals automatically more often groomed partners that were similar in rank . This was due to grooming being based on fear of defeat , and due to spatial structure ( 10 in Tables 4 and 5 ) . At a high intensity of aggression , not only a steep hierarchy develops , but also a spatial structure with dominants in the centre and subordinates at the periphery that is clearer than at a low intensity ( 1 , 5 in Table 4; 1 , 7 in Table 2C ) . Therefore individuals of similar rank are closer together . Thus , at high aggression intensity because individuals will groom up the hierarchy , while meeting mostly others of similar rank , this means that everyone grooms those of similar rank more often than those of lower rank , and , those of similar rank approximately as often as those of higher rank . Therefore , a correlation for grooming among individuals of similar rank results . At a low intensity of aggression , spatial centrality of dominants is absent ( 5 in Table 4 ) and due to the small rank differences grooming is directed neither up the hierarchy , nor to others of similar rank ( 9 , 10 in Table 4 ) . The occurrence of reconciliation in our model is a side-effect of spatial proximity , since it is almost absent if interaction partners are chosen at random ( 7 in Table 5 ) . Thus , reconciliation in the model is largely due to the higher probability of two opponents to be close to each other immediately after a fight ( i . e . Post-Conflict ) than otherwise ( during the Matched-Control ) . Furthermore , the conciliatory tendency is reduced at high intensity of aggression as a side-effect of the spatial structure and the dependence of grooming on the fear of defeat; without these assumptions the conciliatory tendency is independent of intensity of attack ( 7 in Table 5 ) . This happens for three reasons ( to be tested in empirical data , Table 2 ) : at a high aggression intensity the spatial structure is more static ( 10 in Table 2C ) , average inter-individual distance is larger ( 6 in Table 2C ) , and centrality of dominants is greater ( 7 in Table 2C ) . First , spatial structure is more static at high aggression intensity , which is apparent from the stronger spatial assortment by rank of individuals ( 5 in Table 4 ) , from the lower diversity of partners at high intensity of aggression than at a low one ( 16 in Table 6 , 10 in Table 2C ) , and from the fact that the correlation between proximity and conciliatory tendency is significantly stronger at a high aggression intensity than at a low one ( 22 in Table 6; 11 in Table 2C ) . Therefore , former opponents may have been more often close to each other not only immediately after the conflict ( in the post conflict period ) but also in the matched control . Consequently , it is more likely that they groom each other in the matched control . If this happened at an earlier moment than after the conflict ( in the post conflict period ) it reduced the conciliator tendency . Second , due to the greater differences in rank , individuals are further apart ( 1 , 4 in Tables 4 and 5 ) and groom less often both as calculated as the percentage of time and the percentage of interactions at a high than at a low intensity of aggression ( 13 , 14 in Table 6; 8 , 9 in Table 2C ) . Thus , they will automatically also groom less often immediately after a conflict , thus reconcile less than at a low aggression intensity ( 7 in Tables 4 , 5 ) . Third , at a high intensity of aggression grooming and conciliatory tendency are reduced because of the combination of spatial structure and the fear of defeat: If the fear of defeat is removed , the conciliatory tendency at a high intensity of aggression is higher than in the complete model ( 4 in Table 7 ) , because spatial assortment according to dominance ( i . e . spatial centrality of dominants ) is weaker than in the complete model ( 3 in Table 7 ) . Thus , dominants are relatively less often activated ( to fight ) and this increases the relative frequency of grooming because subordinates are aggressive less often ( 2 in Table 6 ) . Thus without fear of defeat the percentage of time spent and interaction time spent on grooming is higher ( 13 , 14 in Table 6 ) , so that it is higher than it is at a high intensity of aggression in the complete model ( 1 , 2 in Table 7 ) and thus the percentage of time spent on reconciliation is higher also ( 5 , 6 in Table 2A ) . Similarly , in the complete model , because at a lower intensity of aggression spatial structure is weaker than at a high intensity also , the percentage of grooming of the total number of interactions ( aggressive plus grooming ) is higher at a low than high intensity of aggression ( 14 in Table 6 ) . Thus the conciliatory tendency is lower at a high than low intensity also ( 5 , 6 in Table 2A ) . Further , at a high intensity of attack reconciliation was directed mostly to those partners that are more valuable ( in terms of grooming outside the context of reconciliation , 11 in Table 4 ) and this was stronger than at low intensity ( 23 in Table 6 ) . This is due to ( 1 ) stronger effects of spatial proximity ( 2 ) high intensity of attack , and ( 3 ) fear of defeat , because without these traits there is no reconciliation with valuable partners ( 11 in Table 5 ) or it is significantly weakened ( 23 in Table 6 ) . As regards spatial proximity , the stronger correlation for valuable relationships arises because the spatial structure at a high intensity is more rigid and therefore both reconciliation and grooming are correlated stronger with the proximity between partners than at a low intensity ( 20 and 22 in Table 6; 11 in Table 2C ) ; thus , the two patterns of grooming and reconciliation are correlated too at a high , but not at a low intensity ( 11 in Table 4 ) . As to the second and third cause , at a high intensity of aggression ( due to the strong hierarchical differentiation ) conciliatory tendency like grooming behaviour appears to be directed up the hierarchy ( 9 in Table 6; 1 in Table 2B ) , although this holds only when grooming is done out of fear of defeat ( 9 in Table 6 ) like in ‘normal’ grooming which does not occur after a conflict ( 9 in Table 4 , 5 ) . There are other patterns in the model that are of interest in it self and for study in empirical data ( Table 2 ) . For instance , in the model higher ranking individuals appear more aggressive due to the lower risk involved ( 2 in Table 6 ) , and less anxious ( but only at high intensity of aggression ) ( 5 in Table 6; 12 in Table 2C ) because they have lost fewer fights ( 4 in Table 6 ) and these effects are stronger at a high than low aggression intensity ( 10–12 in Table 6; 4 in Table 2C ) . Further , both at a high and a low intensity of aggression , there is no correlation between grooming and rank ( 6 in Table 6; 2 in Table 2A ) . This is remarkable at high intensity of aggression , because lower ranking individuals are more anxious and therefore , they may be expected to groom others more often . The absence of this correlation arises from the fact that a high grooming frequency by low ranking individuals is counteracted by the spatial social structure ( 5 in Table 5 ) ; due to their peripheral positions , low ranking individuals have fewer opportunities to interact with others than dominants do and therefore , despite their greater tendency to groom , they do not groom more often than dominants . As regards the sensitivity to changes of parameter , the affiliative patterns were insensitive to different values of parameters related to Anxiety . Values ranging from 0 . 001% to 10% for AnxInc and values from 0 . 05 to 0 . 15 for AnxIncFight , AnxDcrGree and AnxDcrGrmr ( whereby AnxDcrGree was kept at higher values than AnxDcrGrmr ) changed the level of anxiety , but did not change results qualitatively . To obtain a sufficiently high number of interactions ( both of grooming and fighting ) to detect affiliative patterns statistically , a Perspace 8 was necessary , whereas a value of 4 was too low . Furthermore , two mental fights ( Equation 1 ) before initiating a dominance interaction ( RiskSens = 2 ) were needed in order to make the frequency of grooming higher than that of fighting like in empirical data . Besides , in empirical data the percentage of time spend fighting was lower in fierce than mildly aggressive species . This was true when comparing the percentage of fighting at high versus low intensity of aggression in the model for RiskSens 1 and 2 , but not for higher values of RiskSens . Results of reconciliation were similar if we prolonged the period of Matched control from three activations to five and to ten activations ( Puga-Gonzalez et al in prep ) . Since in the empirical data average dominance cannot be accessed directly like in our model , we also tested all correlations with a measure of dominance , i . e . their average percentage of winning , which can be measured in real behaviour [95] . All results of Table 4 and 6 remain similar ( also in the strength of the significance ) , apart from two correlations in Table 6: when correlating with the average percentage of winning as a measure of dominance , at a high intensity of aggression , higher ranking individuals groom others significantly less and at a low intensity of aggression , the negative correlation between aggression received and dominance is no longer significant ( data available on request ) . It should be noted however , that to explain patterns of our simulation , the correlations with average dominance value are of greater interest than with percentage of fight won because the average dominance value is a more direct cause of behaviour in the model .
The causation of each of the affiliative patterns in the model is as follows . First , grooming up the hierarchy results when aggression intensity is high and the hierarchy is steep because individuals seldom dare to attack higher ranking ones , and therefore in order to reduce their anxiety they groom up the hierarchy instead . When aggression intensity is low , the hierarchy is weak , thus individuals experience a smaller risk to attack higher ranking ones and therefore there is no such pattern . Other patterns depend on the spatial configuration . Because interactions take place in space , individuals are more likely to be close to those they have recently interacted with than to others . Therefore , they are more likely to groom one another after an interaction than at other times . Thus , we observe patterns of both reconciliation of fights and reciprocation of grooming . In the model , aggression determines the spatial structure of the group [43] . At a high intensity of aggression a spatial structure develops through the continuous fleeing of low ranking individuals . Therefore , subordinates end up at the periphery and dominants are located in the centre , and thus individuals are closer to others of similar rank . Such a rank-assortment is virtually absent at a low intensity of aggression [41] . Therefore at a high intensity of aggression , since individuals are closer to others of similar rank , they usually groom others of similar rank . Furthermore , at a high intensity of aggression dominants interact more often than subordinates , because dominants are surrounded at all sides by others due to their spatial centrality . Consequently , because dominants are more often aggressive than subordinates are , the percentage of interaction time spent in grooming is lower at a high than at a low intensity of aggression . Because individuals groom relatively less often , this causes less reconciliation at a high than at a low aggression intensity . Furthermore , due to the fact that the spatial structure is relatively more rigid at a high aggression intensity , individuals are more often close to the same partner and this increases the chance that they are close to a former opponent at all times . Therefore , the frequency with which individuals groom with former opponents sooner after a fight than in the matched control ( MC-PC method ) declines . This reduces the rate of reported reconciliation . Besides , due to the relatively rigid spatial structure at a high aggression intensity , individuals more often reconcile with the same partners as they groom with and thus , they reconcile with valuable partners more often than at a low aggression intensity . In sum , aggression structures the spatial configuration of individuals in the group and ( together with grooming out of fear of defeat ) this structures the affiliative patterns . Although in empirical data rank is not measured by an internal Dom value ( like in our model ) , similar results were obtained in the model if rank was computed by the empirical measure , the average percentage of winning [95] . In the model , the correlations with rank and 1 ) aggression given , 2 ) aggression received and 3 ) fights lost appeared to be stronger at a high intensity of aggression than at a low one ( 10–12 in Table 6 ) . Whether this difference may serve as a new indication of the degree of despotism for real primates , needs further study ( 2–4 Table 2C ) . The relevance of the model to affiliative patterns of primates is supported by the following empirical evidence ( Table 2 ) . First , in many species grooming up the hierarchy appears to be stronger the steeper the gradient of the hierarchy when comparing between groups of a single species [115] ( conform 3 in Table 2A ) . Further , the larger inter-individual distance at high versus low aggression intensity in the model ( 6 in Table 2C ) is confirmed in empirical data at several levels of comparison , not only by a comparison between species , namely between rhesus and tonkean macaques [103] , and between rhesus and stump-tailed macaques [106] ( see Table 1 ) , but also within groups intense conflicts result in larger distances between opponents than do mild conflicts in both a group of Japanese macaques [116] and wild chimpanzees [117] . The correlation between proximity and grooming ( 1 in Table 2A ) is supported in lion-tailed macaques and tonkean macaques [52] , [118] and by the difference in distance and grooming frequency between despotic rhesus monkeys and egalitarian stump-tailed macaques [106] . The combination of spatial configuration and proximity induced grooming leads to reciprocity of grooming . This mechanism may underlie the so-called ‘symmetry-based’ reciprocity [119] where the correlation results from a common underlying variable , namely proximity . As to the extent to which closer proximity between former opponents after a fight explains the occurrence of the higher grooming tendency after a fight ( which is interpreted as reconciliation ) , a number of empirical studies confirm this . These studies concerned stump-tailed macaques , rhesus macaques [19] , [120] , Japanese macaques [116] , Moor macaques [121] and a comparison between studies of several species in captivity vs . natural conditions [122] . However , a number of studies conclude that closer distance after a fight cannot explain the conciliatory tendency exhaustively , because when controlling for distance by matching ( to some degree ) the distance in the matched control to that after the fight ( in the post conflict period ) , these studies show that a certain conciliatory tendency still remains after controlling for distance [21] , [63] , [123]–[126] despite a great reduction in the conciliatory tendency in some studies [120] . Whether this also happens in the model if we control for distance in the matched control of the MC-PC method , we will study in future . Further , as in the model ( 21 in Table 6; 1 in Table 2B ) , females of a group of baboons reconciled more often with higher ranking victims than lower ranking ones [127] . In the model , this arises at a high aggression intensity , because individuals groom others of higher rank more often , since they are afraid to attack them . Thus , they also groom high ranking ones after a fight more often and thus reconcile with others of higher rank more . Note that this finding also may be interpreted in the frame of the most valuable relationship hypothesis , because the higher the rank of the partner ( due to the effective support it can give , for instance ) the more valuable the individual is to reconcile with . Further , as in the model , a correlation between rank and grooming is lacking ( 2 in Table 2A ) in the study of baboons and vervets ( which are despotic species ) [128] , [129] , but such a correlation is found in lion tailed macaques ( in this study this species appears to be despotic ) [52] . Since the absence of this correlation in the model is due to spatial centrality of dominants , we expect spatial social structure to be stronger in baboons and vervets than in lion tailed macaques . At a high intensity of attack , but not at a low one , lower ranking females are more anxious ( 5 in Table 6; 12 in Table 2C ) , because they more often receive aggression and lose fights than higher ranking individuals in the model ( 10–12 in Table 6; 3–4 in Table 2C ) . This is confirmed by correlations between the frequency of receipt of aggression , the level of anxiety , and anxiety-induced arthrosclerosis in the fiercely aggressive despotic macaque species , rhesus and long-tailed macaques [130]–[132] . It is of interest to see whether in empirical data , like in the model ( 5 in Table 6 ) this correlation between rank and anxiety is weaker in egalitarian species ( 12 in Table 2C ) . Thinking along the lines of dominance relations , our model may also change our explanations for two other phenomena . Firstly , in female-bonded species , in primate groups that are more female-biased females appear to groom less frequently . This is explained by the assumption that in female-bonded groups not every female needs to groom every other [133] . According to our model , however , reduced grooming by females in a group that is female-biased may be a side-effect of the rule that individuals groom the others out of fear of defeat: Because in a female-biased group females meet other females more often and they fear defeat less if they meet a female than if they meet a male , they will attack more than in a group with more males . Second , the fact that female macaques groom males more often than vice versa [118] , [134]–[137] is explained by our model as a consequence of their subordinance to males . From this we may derive another prediction: since in despotic species females are dominant over a higher number of males than in egalitarian species [45] , we expect that ( for the same adult sex ratio in despotic and egalitarian groups ) females of despotic species groom males less than females of egalitarian species do ( 13 in Table 2C ) . Our model shows the four different levels of complexity of social behaviour distinguished by Hinde [138]: Individual behaviour , interactions , relationships and social structure . In agreement with Hinde'suggestion , each level can be described in terms of the level below it , and levels influence each other mutually . For instance , the nature of the behaviour of the participants influences their relationship and these relationships in turn , also influence the behaviour of the participants . Also related to this view is that observed social structure can vary dramatically with circumstances , without any changes in the underlying motivational mechanisms or strategies . For instance , here we show that patterns of reconciliation differ depending on intensity of aggression and in our former paper we showed that female dominance increases with the percentage of males in the group [45] . A criticism made against DomWorld by Bryson and co-authors [139] has been that the dominance hierarchy in the model was not as stable as that of real primates . The dominance hierarchy in GrooFiWorld is stable , however , because average dominance values between periods 200 and 230 are significantly correlated with those between 231 and 260 ( see methods ) . Further , in GrooFiWorld we have shown that even if we keep the hierarchy 100% stable ( by omitting the self-reinforcing effects of winning and losing fights ) all patterns remain similar ( Table 4 , 5 ) . Another criticism concerned the directional inconsistency of aggression [140] . The directional inconsistency of aggression at high aggression intensity in DomWorld appeared to be lower than that in empirical data . In the present paper , in GrooFiWorld , the directional inconsistency is higher than in DomWorld . 0 . 73 vs 0 . 55 respectively , because in GrooFiWorld the individuals think twice before they attack , whereas in DomWorld they think only once and thus , attack higher ranking individuals more often . How it compares exactly to empirical data is not clear , because the matrices tested by de Vries sometimes comprise of males , sometimes of females and sometimes of both sexes and the directional inconsistency probably depends on the group composition . However , despotic macaque species show an average directional inconsistency of 0 . 89 , which still is above that of GrooFiWorld . To study this in more detail is beyond the scope of this paper . Yet , there are other shortcomings in our study of the model that will be amended in future . There are a number of patterns related to reconciliation that have been found in studies of real primates that we do not yet treat in the model [25] , [110] , [127] , [141]–[143] , we used the time rule method [58] neither to test for reconciliation nor for the valuable relationship hypothesis , nor did we control for proximity in our study of reconciliation [63] . The rule of grooming out of fear of defeat may be interpreted by assuming that individuals groom others to calm these partners down and to forestall the chance of receiving aggression from them; thus , it could be viewed as an exchange of grooming for tolerance . However , in the present model grooming others does not influence whether or not the groomee will subsequently attack the groomer . The model also does not represent cases in which grooming can be rejected by the receiver , nor pre-existing differences between individuals , such as are apparent , for instance , between primates of different personality [144] , nor what individuals compete for such as sex or food . It omits kin-relations and offspring among partners as well as coalitions . Besides , we have not yet studied effects of different sex ratios , whereas primate groups of the same species may differ in sex ratios , and this has been shown to have an influence on their affiliative patterns [133] , [145]–[147] . These are natural variations and extensions that will need to be added to our model , as we intend to do in future studies . As to cognition , our model does not at all reflect the behavioural and cognitive complexity of primates . Regarding affiliation , it is confined only to the representation of an anxiety reducing effect of grooming in the context of a competitive regime . Because of the resemblance of the emergent affiliative patterns in our model to those of primates , similar processes may cause these affiliative patterns in primates also . Whether or not primates may ( sometimes ) use the more complex cognitive rules that have been suggested by primatologists before , our model cannot decide . Instead , our model may be used as a null-model that indicates what patterns we should expect in the absence of the usual cognitive rules regarding reciprocation , reconciliation etcetera . Thus , it does not deny that primates are intelligent as has been shown in many experimental studies [37] , [38] , [148] , but it questions whether these primates use all aspects of their intelligence in all contexts . It illustrates that apart from the here reproduced patterns at a group level in the model , extra evidence , is needed as proof of 1 ) intentional reciprocation , 2 ) competition for higher ranking grooming partners , and 3 ) intentional exchange and 4 ) intentional reconciliation . Further , our model points to the need for more studies of the spatial distribution of monkeys within a group . Of these studies [108] , [149] ( Girod , Thierry , Hemelrijk , in prep ) , there have been only a few so far . In sum , we have shown that without the specific cognitive assumptions for the creation of each pattern of grooming , cognitively simple local interactions and self-organization suffice to produce many of the affiliative patterns that are typical of egalitarian and despotic primate societies ( Table 1 , 4 ) and also a number of other patterns ( Table 6 ) . The main finding is that the spatial configuration associated with the competitive regime and grooming out of fear of defeat or out of anxiety structure the patterns of grooming such that we measure patterns of reciprocation , exchange and reconciliation . This leads to a number of model-based hypotheses for real primates ( Table 2 ) . Because the model generates many of the behavioural patterns found in real primates , but does so without the usually assumed cognitive processes , it can be used as a null model for studying primate affiliative behaviour . | Individual primates distribute their affiliative behaviour ( such as grooming ) in complex patterns among their group members . For instance , they reciprocate grooming , direct it more to partners the higher the partner's rank , use it to reconcile fights and do so in particular with partners that are more valuable . For several types of patterns ( such as reconciliation and exchange ) , a separate theory based on specific cognitive processes has been developed ( such as individual recordkeeping , a tendency to exchange , selective attraction to the former opponent , and estimation of the value of a relationship ) . It is difficult to imagine how these separate theories can all be integrated scientifically and how these processes can be combined in the animal's mind . To solve this problem , we first surveyed the empirical patterns and then we developed an individual-based model ( called GrooFiWorld ) in which individuals group , compete and groom . The grooming rule is based on grooming out of fear of defeat and on the anxiety reducing effects of grooming . We show that in this context this rule alone can explain many of the patterns of affiliation as well as the differences between egalitarian and despotic species . Our model can be used as a null model to increase our understanding of affiliative patterns of primates . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"neuroscience/behavioral",
"neuroscience",
"neuroscience/cognitive",
"neuroscience",
"ecology/behavioral",
"ecology",
"computational",
"biology",
"neuroscience/animal",
"cognition",
"computational",
"biology/systems",
"biology"
] | 2009 | Emergent Patterns of Social Affiliation in Primates, a Model |
The most prominent developmental regulators in oocytes are RNA-binding proteins ( RNAbps ) that assemble their targets into ribonucleoprotein granules where they are stored , transported and translationally regulated . RNA-binding protein of multiple splice forms 2 , or Rbpms2 , interacts with molecules that are essential to reproduction and egg patterning , including bucky ball , a key factor for Bb formation . Rbpms2 is localized to germ granules in primordial germ cells ( PGCs ) and to the Balbiani body ( Bb ) of oocytes , although the mechanisms regulating Rbpms2 localization to these structures are unknown . Using mutant Rbpms2 proteins , we show that Rbpms2 requires distinct protein domains to localize within germ cells and somatic cells . Accumulation and localization to subcellular compartments in the germline requires an intact RNA binding domain . Whereas in zebrafish somatic blastula cells , the conserved C-terminal domain promotes localization to the bipolar centrosomes/spindle . To investigate Rbpms2 functions , we mutated the duplicated and functionally redundant zebrafish rbpms2 genes . The gonads of rbpms2a;2b ( rbpms2 ) mutants initially contain early oocytes , however definitive oogenesis ultimately fails during sexual differentiation and , rbpms2 mutants develop as fertile males . Unlike other genes that promote oogenesis , failure to maintain oocytes in rbpms2 mutants was not suppressed by mutation of Tp53 . These findings reveal a novel and essential role for rbpms2 in oogenesis . Ultrastructural and immunohistochemical analyses revealed that rbpms2 is not required for the asymmetric accumulation of mitochondria and Buc protein in oocytes , however its absence resulted in formation of abnormal Buc aggregates and atypical electron-dense cytoplasmic inclusions . Our findings reveal novel and essential roles for rbpms2 in Buc organization and oocyte differentiation .
Two major objectives of oocyte development are to produce haploid gametes through meiosis , and to prepare the ovulated egg for successful fertilization and early embryonic development . Unlike most developmental programs that are regulated by transcription factors , the developmental programs of oocyte maturation , egg fertilization , and early embryonic development take place while the oocyte and early embryonic genomes are transcriptionally silent ( reviewed in [1 , 2] ) . During this period , RNA-binding proteins ( RNAbps ) are the predominant post-transcriptional regulators that coordinate localization and translation of the RNA molecules encoding the proteins that govern processes essential to oogenesis and early embryogenesis . The RNAbp RNA-binding protein with multiple splicing , RBPMS , family is generally represented by two paralogs in vertebrates , RBPMS and RBPMS2 [3] . The RNA recognition motif of RBPMS family members contains two ribonuclear protein domains , RNP1 and RNP2 , which contain the 6–8 residue structural elements which bind to RNA [4–6] . RBPMS proteins associate with poly-adenylated mRNAs in vitro [7] , and PAR-CLIP followed by RNA sequencing identified the 3’UTR of target RNAs as the primary region to which RBPMS proteins bind ( ~ 35% ) , followed by intronic regions ( ~ 20% ) and coding sequence ( ~10% ) [3] . Interestingly , the association with intronic regions suggests that RBPMS proteins can interact with pre-mRNA , and indeed , RBPMS/RBPMS2 can shuttle between nuclear and cytoplasmic fractions [3] . In germ cells , RNAbps associate with RNAs into supramolecular complexes called RNPs ( ribonucleoproteins ) , which further aggregate into granules that are a hallmark feature of primordial germ cells ( PGCs ) , and oocytes of various stages ( reviewed in [8 , 9] ) . In primary oocytes , a transient structure called the Balbiani body ( Bb ) is a single , large , cytoplasmic aggregate of RNPs , scaffolding proteins , and other patterning molecules which indicates the future vegetal pole of the oocyte [10] . The RNAbp RNA-binding protein with multiple splicing ( Rbpms ) , or hermes in Xenopus , localizes to the Bb of frog and zebrafish oocytes [11 , 12] , where it interacts with Bucky ball protein ( called Velo1 in Xenopus ) , the only vertebrate gene known to be required for Balbiani body formation [13–16] . Zebrafish Rbpms2b also binds to the bucky ball transcript , which contains numerous predicted Rbpms2 RNA recognition elements within its introns and 3’UTR [14] . In spite of Rbpms2 localization to the Bb of oocytes and the presence of these important biochemical interactions , the function of Rbpms2 in oocyte development or Bb formation has not been well elucidated . In this work , we characterized the localization of wild-type and mutant Rbpms2 proteins to cellular RNA granules , including germ granules of PGCs , the Bb of oocytes , and granules within somatic cells . Rbpms2 localization to germ granules and the Bb of oocytes is dependent on its RNA binding domain . In zebrafish somatic cells , this domain is sufficient for granule localization , while the C-term domain promotes association with the bipolar spindle at the expense of granules . In HEK 293 cells , RNA binding is dispensable for granule localization , indicating Rbpms2 uses different domains to achieve its subcellular localization in diverse cell types . To investigate Rbpms2 functions , we generated zebrafish mutants disrupting the duplicated rbpms2 genes , rbpms2a and rbpms2b using Crispr-Cas9 mutagenesis . These analyses revealed that Rbpms2 is essential for oocyte development as rbpms2 double mutants ( hereafter rbpms2 mutants ) develop exclusively as fertile males . Zebrafish rbpms2 mutants have normal germline development until the onset of sexual differentiation , at which time rbpms2 mutants can initiate , but not complete , oocyte differentiation . Testis differentiation is unimpaired in rbpms2 mutants . Ultrastructural analysis revealed early oocytes of rbpms2 mutants can asymmetrically accumulate mitochondria like their normal siblings , but have large cytoplasmic inclusions that were not present in wild-type . Based on the presence but unusual localization of Bucky ball protein in early oocytes of rbpms2 mutants , we conclude that Rbpms2 is dispensable for Buc translation; however , Rbpms2 seems to be required for structural integrity of Buc aggregates . In addition , Rbpms2 has a Buc-independent function in ovary maintenance . This data reveals a novel and essential role for rbpms2 in oogenesis and Bucky ball organization within early oocytes .
In Xenopus , hermes is expressed in maturing oocytes , and hermes RNA and protein are both localized to the Balbiani body of primary oocytes [12] . Similarly , in primary zebrafish oocytes , the Xenopus anti-Hermes antibody detects a Bb-localized Hermes homolog [11 , 15] . Zebrafish have three Hermes homologs encoded in their genome: Rbpms ( hereafter Rbpms1 for clarity ) , and the similar proteins Rbpms2a and Rbpms2b , the products of rbpms2 gene duplication in zebrafish [17] . Additionally , Xenopus Hermes interacts with the Bb-localized Bucky ball homolog Velo1 [16] , as do all three protein products of the zebrafish Rbpms family ( Fig . S1M in S1 Supporting Information ) [14] . Phylogenetic analyses of RBPMS proteins have previously been described [3 , 18 , 19]; however , none directly compared zebrafish RBPMS proteins with Xenopus Hermes . Thus , we compared protein similarity of Hermes and the zebrafish Rbpms family , which revealed that zebrafish Rbpms2a/b proteins clustered more tightly together with Hermes , and thus likely represent the closest Xenopus Hermes homolog ( Fig . S1A in S1 Supporting Information ) [20 , 21] . Maternal RNA transcripts for rbpms1 , rbpms2a , and rbpms2b were all abundant in early embryonic stages prior to the maternal-zygotic transition , after which their levels were severely reduced ( Fig . S1B–E” in S1 Supporting Information ) . Examination of rbpms2 transcripts during zebrafish embryogenesis by whole-mount in situ hybridization ( ISH ) revealed that rbpms2a and rbpms2b were expressed in the embryonic heart , retina and pronephros , an evolutionarily conserved expression pattern similar to that previously reported for hermes homologs in other vertebrates , including Xenopus , chicken , and mouse ( Fig . S1H–K in S1 Supporting Information ) [7] . Low levels of rbpms1 expression were ubiquitous throughout most tissues of 24-48hpf embryos ( Fig . S1F–G in S1 Supporting Information ) . Next we examined rbpms1 and rbpms2a/b expression in zebrafish oocytes . We found that while all three transcripts were expressed in primary oocytes ( Fig 1A–1D and Fig . S1L in S1 Supporting Information ) , only rbpms2a transcripts were enriched within the Bb ( Fig 1A , arrow ) . In contrast , rbpms2b transcripts appeared to be enriched at the oocyte cortex or within the granulosa cell layer ( Fig 1C , arrowheads ) . RT-PCR on FACS-sorted GFP-positive granulosa and theca cells from Tg[cyp19a1a:GFP] transgenic zebrafish ovaries , revealed no rbpms2 expression in these cell populations [22] ( Fig 1E ) . However , because Cyp19a1a:GFP is only expressed in granulosa cells of stage II follicles or later [22] , we cannot exclude the possibility that rbpms2 RNAs are transiently expressed in granulosa cells of stage I follicles , then rapidly down-regulated prior to stage II . Finally , we tested whether both Rbpms2 proteins were expressed in primary oocytes ( stage I ) and localized to the Bb . We used two commercial antibodies raised against human Rbpms2 which we predicted would recognize both Rbpms2a and Rbpms2b based on protein similarity ( Fig . S2A in S1 Supporting Information ) . Stained ovaries from zebrafish mutants for rbpms2a ( Fig 1F and Fig . S3A , E in S1 Supporting Information ) or rbpms2b ( Fig 1I , Fig . S3B , F in S1 Supporting Information ) both exhibited staining of Rbpms2 in the Bb , likely representing the ohnolog protein , or alternatively a truncated mutant protein ( mutants described in next section ) . The absence of signal in rbpms2 double mutants indicates antibody specificity for Rbpms2s ( Fig . S3C in S1 Supporting Information ) . Rbpms2 protein in single mutant ovaries is localized to the Bb along with Bucky ball ( Fig 1F–1K ) . Thus , consistent with the expression of Xenopus Hermes , we conclude that both Rbpms2a and Rbpms2b proteins are localized to the Bb of zebrafish oocytes . We anticipated that rbpms2a and rbpms2b might function redundantly based on protein similarity ( 92% identity ) and their comparable gene expression profiles ( Fig 1 and Fig . S1 in S1 Supporting Information ) . Therefore , to study rbpms2 functions we generated zebrafish mutant lines disrupting both rbpms2a and rbpms2b using CRISPR/Cas9 mutagenesis [23] . After numerous unsuccessful attempts to mutagenize upstream regions of the rbpms2a and rbpms2b genes , we ultimately succeeded in mutating analogous regions of exon 5 in rbpms2a/2b using CRISPR sites predicted by the web-based ZiFiT targeting program ( https://crispr-cas9 . com/96/zifit-targeter-crispr-cas9/ ) ( Fig 2A and Fig . S2B , C in S1 Supporting Information ) . We recovered germline-transmitted mutant alleles for each rbpms2 gene . Sequencing of genomic regions as well as mutant cDNAs revealed that the isolated mutations were as follows: the rbpms2aae27 allele had an in-frame deletion of 9 amino acids , the rbpms2aae30 allele had a 15bp deletion and 17bp insertion resulting in a truncated protein ( Fig . S2B in S1 Supporting Information ) , and the rbpms2bae32 allele had a 20bp insertion resulting in a truncated protein ( Fig . S2C in S1 Supporting Information ) . Additionally , we characterized a splice-site mutation from the Sanger Institute’s Zebrafish Mutation Project [24] , allele rbpms2bsa9329 , which was found to contain a T>A point mutation in the 5’ splice site between exon 3 and 4 that causes in-frame skipping of exon 3 and is predicted to partially disrupt RNP1 ( Fig . S2C in S1 Supporting Information ) . We found that zebrafish embryos homozygous for mutations in rbpms2a or rbpms2b had no observable morphological differences from their wild-type siblings . Adult fish mutant for a single rbpms2 gene had no apparent phenotypes , and differentiated into fertile fish of either sex ( Fig 2B ) . Moreover , no overt maternal-effect , paternal-effect or maternal-zygotic phenotypes were observed in the progeny of single mutant adults for rbpms2 genes ( n>20 adult mutants per allele examined ) . To test the possibility that the rbpms2 genes were functionally redundant , we made rbpms2a;rbpms2b double mutants . At 3 days post fertilization ( d3 ) no phenotypes were observed when a single functional copy of rbpms2 was present; however , double mutant embryos for the truncation alleles ( rbpms2aae30/ae30;rbpms2bae32/ae32 ) or for the rbpms2a truncation allele and the rbpms2b exon-skipping allele ( rbpms2aae30/ae30;rbpms2bsa9329/sa9329 ) displayed cardiac edema phenotypes ( Fig 2C and 2D ) . This cardiac phenotype is consistent with the conserved expression of rbpms2 in the embryonic heart ( Fig . S1H–K in S1 Supporting Information ) [25 , 26] , as well as the previously published cardiac phenotypes caused by hermes/rbpms2 overexpression in Xenopus [7 , 25 , 26] . Consistent with previous findings that demonstrate Rbpms proteins function as dimers [6 , 7 , 27] , we found that zebrafish Rbpms2a and Rbpms2b can both homodimerize , and Rbpm2a can heterodimerize with Rbpms2b ( Fig . S6B in S1 Supporting Information ) . Thus , there is likely no requirement for heterodimers and no difference in Rbpms2a or Rbpms2b homodimer function since complete loss of a single rbpms2 gene is still sufficient for normal Rbpms2 function . Approximately half of the embryos with cardiac edema at d3 recovered by d5 and were raised to adulthood ( 47±18% , based on quantification of 71 total edematous embryos at d3 from 9 parental in-crosses ) ( Fig 2E–2G ) . In contrast , no cardiac edema phenotypes were observed in rbpms2aae27/ae27;rbpms2bsa9329/sa9329 double mutants or rbpms2a trans-het;rbpms2b double mutants ( rbpms2aae27/ae30;rbpms2bsa9329/sa9329 ) ( Fig 2B ) . This analysis suggests that the rbpms2aae27 allele retains sufficient activity to fulfill Rbpms2 functions; whereas , rbpms2aae30 , rbpms2bae32 , and rbpms2bsa9329 are loss-of-function mutations . Therefore , we focused subsequent analyses on the double mutants rbpms2aae30/ae30;rbpms2bae32/ae32 or rbpms2aae30/ae30;rbpms2bsa9329/sa9329 , hereafter referred to as rbpms2 DM ( double mutants ) or simply rbpms2 mutants . To examine the stability of the mRNA for the different rbpms2 alleles , we performed RT-PCR on RNA extracted from maternal-zygotic ( MZ ) mutant embryos at the 4-cell stage . This embryonic stage precedes zygotic genome activation; thus , the contributions of maternal transcripts from rbpms2a or rbpms2b single mutant mothers can be examined . We found that the rbpms2aae30 and rbpms2bae32 transcripts were detectable , but appeared to be less abundant than wild-type , indicative of potential nonsense mediated decay ( NMD ) ( Fig 3A , pink asterisks ) . This is consistent with descriptions of NMD in zebrafish , which does not typically lead to complete degradation of transcripts containing early stop codons [28] . The finding of NMD further supports the notion that rbpms2aae30 and rbpms2bae32 are likely null alleles . Mammalian RBPMS has previously been studied in HEK293 cells , where it has been demonstrated to localize to stress granules along with poly-adenylated RNAs and the stress granule marker G3BP1 [3] . Therefore , we used this established localization assay to examine the stability and activity of the Rbpms2 proteins encoded by the zebrafish mutant alleles . First , we examined the localization of the wild-type zebrafish Rbpms2 proteins in HEK293 cells using GFP protein fusions . We found that GFP-Rbpms2a and GFP-Rbpms2b were each predominately localized to small punctate structures resembling the previously described stress granules ( Fig 3B and 3D ) . Stable GFP fusion proteins for GFP-Rbpms2aae30 and GFP-Rbpms2bae32 were detected in HEK293 cells , indicating that the mutant RNAs and proteins were sufficiently stable to be visualized in this context; however , in contrast to the punctate localization of the WT proteins , the GFP-Rbpms2aae30 and GFP-Rbpms2bae32 proteins were diffusely cytoplasmic ( Fig 3C and 3E ) . GFP-Rbpms2bsa9329 exhibited an intermediate localization phenotype , displaying both diffuse GFP fusion protein as well as localization to somatic cell granules ( Fig 3F ) , suggesting this allele may behave as a hypomorphic reduction-of-function allele . Taken together these data indicate that the C-terminus is required for localization to granules in this somatic cell type . Next , we examined the localization of Rbpms2 fusion proteins in somatic cells of zebrafish embryos . Wild-type GFP-Rbpms2 was detected near the nucleus ( ~20% ) and associated with the centrosomes or spindle ( 9% ) , and as in HEK293 cells , GFP-Rbpms2 was localized to granules in 65% of somatic cells of the zebrafish embryo ( Figs 3G–3H and 4 1; n = 489 cells , 19 embryos ) . To determine the identity of these Rbpms2 positive granules , we examined endogenous markers , including the classical stress granule marker Tial-1 and the p-body marker Dcp2 , using antibodies previously validated in zebrafish [29] . We detected no overlap with Tial-1 ( Fig 3G; n = 172 cells , 8 embryos ) and partial overlap with Dcp2 ( Fig 3H; n = 317 cells , 11 embryos ) , indicating that the granules to which GFP-Rbpms2 localizes in zebrafish somatic cells are not stress granules , but that a small subset are GFP-Rbpms2 and Dcp2 positive granules . Therefore , we designate these heterogeneous GFP-Rbpms2 granules as granules of somatic cells . In most GFP-Rbpms2bae32 expressing cells the protein was detected in the nucleus , throughout the cells , and in granules ( 73% ) ( Figs 3J and 4; n = 232 cells , 8 embryos ) . In GFP-Rbpms2aae30 fewer cells with GFP granules were detected ( 43% ) . GFP-Rbpms2aae30 was also detected throughout the cytoplasm ( 33% ) or in the nucleus , or in a bipolar pattern likely representing the centrosomes ( 14% and 10% respectively ) ( Figs 3K and 4; n = 177 , 8 embryos ) . Strikingly , the sa9329 protein , which disrupts the RNA binding domain was not detected in granules , but strongly localized to the centrosomes/spindle ( Figs 3I and 4; n = 311 cells , 9 embryos ) . These data indicate that the RNA binding and C-terminal domains differentially contribute to Rbpms2 subcellular localization in somatic cells . To test if Rbpms2 proteins can localize to granules in the zebrafish germline , we used a commonly employed assay in which in vitro transcribed RNA encoding GFP-rbpms2a/2b was injected into fertilized zebrafish eggs . For many germ plasm RNAs and interacting proteins , the translated product of the exogenous RNA will localize to germ granules of primordial germ cells ( PGCs ) in 24-30hpf embryos ( reviewed in [1 , 30] ) . PGCs were marked by cytoplasmic expression of RFP-nanos3’UTR . Co-injection of RFP-nanos3’UTR and either wild-type protein , GFP-Rbpms2a or GFP-Rbpms2b , demonstrated that Rbpms2 fusions localize to germ granules of PGCs . ( Fig 5A , 5B and 5F ) . Next , we checked the germ granule localization capacity of the rbpms2 mutant alleles: GFP-Rbpms2aae30 , GFP-Rbpms2bae32 and GFP-Rbpms2bsa9329 . Like wild-type Rbpms2 proteins , all mutant proteins became restricted to PGCs ( Fig 5C–5E ) ; however , the GFP-Rbpms2aae30 mutant protein and the GFP-Rbpms2bsa9329 were cytoplasmic rather than enriched in granules ( Fig 5C , 5D and 5F ) . Similarly , the GFP-Rbpms2bae32 fusion protein was diffuse throughout the cytoplasm in 35% of expressing cells; whereas , in the majority of the cells ( 65% ) GFP-Rbpms2bae32 cells was both diffusely cytoplasmic and enriched in germ granules ( Fig 5D and 5F ) . The PGC-enrichment of the mutant GFP-Rbpms2 alleles suggests that the injected RNAs , or more likely the protein products , are capable of interacting with germ cell factors that stabilize them in germ cells . Because only Rbpms2bae32 retains partial ability to localize to germ granules this suggests that an intact RRM and sequences adjacent to the RRM are required but not sufficient for germ granule localization in PGCs . To further investigate the mechanisms that mediate localization to germ and somatic cell granules , we asked if Rbpms2 localization was dependent on its interaction with RNA . To test this , we constructed a mutant version of Rbpms2b , called GFP-rbpms2bΔRNP1 , that lacks seven amino acid residues of the RNP1 ( ribonuclear protein ) domain that are essential for making direct contact with the RNA molecule ( Fig 6A ) [4 , 5] . We injected in vitro transcribed GFP-rbpms2b and GFP-rbpms2bΔRNP1 to test if the encoded proteins were able to localize to germ granules of PGCs . Wild-type GFP-Rbpms2b was present in germ granules where it localizes with the endogenous germ granule component Vasa; however , GFP-Rbpms2bΔRNP1 did not localize and did not appear to be enriched in PGCs ( Fig 6C and 6D” ) . To determine if failure to localize to germ granules was due to instability of the mutant protein , we examined the injected embryos earlier , at 3-4hpf , and found that both WT and mutant proteins had robust observable GFP signals ( Fig 6G and 6H ) . Furthermore , when these constructs were transfected into HEK293 cells , both wild-type and Rbpms2bΔRNP1 could localize to somatic cell granules ( Fig 6I and 6J ) . Therefore , we determined that GFP-Rbpms2bΔRNP1 produces a stable protein; however , the protein lacking RNP1 cannot localize to the germ granules . Germ granules in PGCs are similar to the Bb of stage I oocytes , in that both are germline aggregates of RNP particles , and many germ plasm components are localized in both cell-specific structures ( reviewed in [1] ) . To determine fusion protein localization in oocytes , we constructed a transgenic zebrafish line expressing wild-type Rbpms2b under the ovary-specific buc promoter [14] , Tg[buc:RFP-rbpms2b] , and a transgenic line that harbors the same RNA-binding deficient mutation described above , Tg[buc:RFP-rbpms2bΔRNP1] ( Fig 6B ) . When we observed dissected oocytes from adult transgenic females , we found that the wild-type transgenic protein localized to the Bb similar to endogenous Rbpms2 . However , as in PGCs , RFP-Rbpms2bΔRNP1 was not localized , and RFP protein was not detectable in oocytes , although transcripts were detectable for both transgenes by in situ hybridization ( Fig . S4 in S1 Supporting Information and Fig 6E and 6F ) . The ability of Rbpms2 fusion protein to localize to germ granules or the Bb is likely independent of interaction with Bucky ball , since myc-Rbpms2bΔRNP1 , which does not accumulate in germ granules or the Bb can still immunoprecipitate with GFP-Buc in co-IP experiments ( Fig 6K ) . Thus , Rbpms2b requires the RNA-binding RNP1 domain to accumulate in and localize to RNP aggregates of the germline . It remains to be determined if RNP1 contributes to stabilization , efficient translation the protein , or both; however , this domain is dispensable for stability and granule localization in somatic cells such as HEK293 . This further suggests that the mechanisms used to localize Rbpms to granules are distinct between germline and soma . We raised rbpms2 mutants that had recovered from the edema phenotype in order to assess possible rbpms2 roles in reproductive development . The adult progeny ( >2 . 5 months ) of in-crosses between double heterozygous rbpms2aae30/+;rbpms2bsa9329/+ fish were genotyped , and rbpms2a or rbpms2b single mutants and double heterozygotes or heterozygote-mutants displayed typical 50/50 male-female sex ratios ( Fig 7A–7D and 7G ) . However , no rbpms2 DM adults were identified among 129 fish , although approximately 8 would be expected according to Mendelian genetics ( at a prevalence of 1:16 ) . To overcome a potential survival disadvantage of rbpms2 mutants , we sorted mutants based on their transient edema phenotype at d3 to determine if rbpms2 mutant adults could be recovered when reared separately from their wild-type siblings . Using this strategy , we recovered 33 rbpms2 mutants to adulthood , all of which were fertile males ( Fig 7E–7G ) . Fertility of rbpms2 mutant males was indistinguishable from wild-type males based on mating assays ( fertilization of eggs ) . Additionally , histological evaluation of mutant testes , through H&E staining and expression of a transgene [ziwi:GFP] that marks germ cells of both sexes [31] , revealed no differences from normal male siblings ( Fig 7C–7F ) . Undifferentiated spermatogonia were noted on the basement membrane of the testicular tubules , as well as differentiated spermatozoa within tubule lumens , indicating that progression of spermatogenesis is normal in rbpms2 mutants . Zebrafish will differentiate as males if germ cell numbers are diminished or oogenesis fails . Early in embryonic development , germline differentiation to the male fate can result when there are few or no PGCs [32–35] . To test the possibility that the all-male phenotype of rbpms2 mutants was caused by insufficient PGC numbers , we stained d3 rbpms2 mutants and their siblings for the germ cell marker Vasa [36] . PGCs were examined by confocal imaging of the lateral side followed by manual counting through the Z series . We found no significant difference in PGC numbers between non-mutant siblings and rbpms2 mutant embryos , with both groups having approximately 20 PGCs per side ( Fig 8A–8C ) . Next , we examined d21 bipotential gonads in the [ziwi:GFP] transgenic line and determined that development of primitive oocytes/gonocytes and meiotic initiation were comparable between rbpms2 mutants and their non-mutant siblings ( Fig 8D and 8E ) . Next , we examined rbpms2 mutants and siblings at d35 when sexual differentiation of the gonads has been initiated . In non-mutant siblings , we found readily distinguishable ovary tissue indicative of female differentiation with numerous stage I and II oocytes ( Fig 8F ) , or testis differentiation with a few gonocytes but mostly small spermatogonia-like cells ( Fig 8G ) . In contrast , rbpms2 mutant gonads were either wholly testis-like ( resembling gonads of their non-mutant male siblings at d35 ) ( Fig 8I ) , or were of mixed character with some spermatogonia-like cells , and also significant numbers of oocyte-like germ cells based on cell diameter and nuclear morphology ( Fig 8H ) . We quantified the size of [ziwi:GFP]-labeled oocytes at d35 and found that non-mutant oocytes reached average diameters of up to 57±5 μm ( n = 26 oocytes , 3 fish ) ; whereas , the largest rbpms2 mutant oocytes reached average diameters of up to 31±4 μm ( n = 23 oocytes , 4 fish ) . These diameters are consistent with rbpms2 mutants failing in oogenesis sometime during the growth phase of stage Ib oocytes ( which typically have diameters of 20–140 μm ) [37] . Consistent with this notion , ultrastructural examination of d21 gonads by transmission electron microscopy revealed no discernable differences in nuclear or cytoplasmic morphology of gonocytes ( n = 3 rbpms2 mutants , n = 4 wild-type ) ( Fig 9A and 9B ) . Taken together , these results suggest that intersex rbpms2 mutants initiate oogenesis , but are in the process of transitioning to the male phenotype , since ultimately all recovered rbpms2 mutants are fertile males . In zebrafish the transition to male fate has been reported to involve apoptosis of germ cells [38] . Consistent with this notion , disrupting the key regulator of apoptosis , Tp53 , can suppress loss of the female germline and the eventual male-only phenotypes of zebrafish mutants disrupting zar1 , brca2 or fancl [39–41] . These studies indicate that apoptosis of ovary-like cells of the bipotential gonad facilitates the transition to testis . Therefore , we reasoned that suppressing tp53-mediated apoptosis by genetically eliminating tp53 may support sustained development of oocytes in rbpms2 mutants . To investigate whether loss of tp53 could support oocyte development , we generated triple mutants with rbpms2 and tp53M214K mutations ( Berghmans et al . , 2005 ) and examined the gonads of rbpms2aae30/ae30;rbpms2 ae32/ae32;tp53M214K/M214K fish at d35 , when gonad differentiation into testis or ovary has occurred . At this stage , we detected no differences between the gonads of rbpms2 mutants and rbpms2 mutants lacking Tp53 ( n = 5 ) ( Fig . S5 in S1 Supporting Information ) , and more advanced stages of oogenesis were not recovered . Moreover , rbpms2;tp53 triple mutants developed exclusively as fertile adult males , like rbpms2 double mutants ( n = 3 ) . Based on this analysis , we conclude that loss of tp53 is not sufficient to support further ovary differentiation in rbpms2 mutants . Moreover , differentiation of rbpms2 mutants as males occurs by a mechanism that is independent of the p53-mediated apoptotic pathway . To examine the subcellular compartment of rbpms2 mutant oocytes more closely , we compared the ultrastructure of d35 gonads from wild-type females ( n = 2 ) and rbpms2 mutants containing oocyte-like germ cells ( n = 2 ) . We found both wild-type and rbpms2 mutant oocytes had accumulated asymmetric mitochondria on one side of the nucleus , a hallmark of early oocyte polarization and Bb formation ( Fig 9C and 9D , mitochondrial accumulation outlined in white ) [15 , 42] . In general , oocytes in the wild-type gonads tended to be larger; therefore , we focused our analysis on oocytes with similar diameters in both genotypes ( ranging from 20–40 μm ) ( Fig 9E ) . We also examined these oocytes for the presence of characteristic nuage: discrete areas of cytoplasm that appear as electron-dense , fibrous accumulations , closely associated with the nuclear envelope and mitochondria [43] . We found similar numbers of cytoplasmic nuage accumulations in wild-type and rbpms2 mutant oocytes ( Fig 9C , 9D and 9F , yellow arrows ) . However , rbpms2 mutant oocytes also contained atypical cytoplasmic inclusions with electron density distinct from nuage that excluded mitochondria ( Fig 9D , 9D’ and 9G ) ; atypical cytoplasmic inclusions >1μm were rarely observed in wild-type oocytes ( p = 0 . 01 , unpaired t-test ) . In rbpms2 mutants , cytoplasmic inclusions were often significantly large , with three of the eight analyzed oocytes containing structures measuring over 3μm across . Although these large electron dense structures are a feature of rbpms2 mutant oocytes , the molecular contents of these cytoplasmic bodies remain to be determined . We have previously shown that Rbpms2 C-terminus interacts with Bucky ball protein and buc transcripts ( Fig . S6A in S1 Supporting Information ) [14]; therefore , we assessed whether loss of Rbpms2 affects Bucky ball abundance or localization by examining Bucky ball protein in Rbpms2 mutants . As previously reported , no specific Buc staining was apparent in d35 testes , including in larger gonocytes of males ( Fig 10A ) , and Buc protein was expressed in pre-Bb stage oocytes at d35 ( Fig 10B ) [14] . Accordingly , no Buc staining was observed in d35 rbpms2 mutant gonads that had already undergone testis differentiation ( Fig 10C ) . Consistent with the notion that a subset of rbpms2 mutants initiated oogenesis as judged by this marker , we detected asymmetric Buc staining in germ cells of rbpms2 mutants that retained oocyte-like cells . As previously reported , the morphology of the Buc stained structures in non-mutant primary oocytes was compact and perinuclear ( Fig 10B and 10B’ white arrows ) . Interestingly , rbpms2 mutant oocytes had more dispersed Buc distribution that formed a ring-like structure ( Fig 10D and 10D’ white arrow heads ) . Although perinuclear localization of Vasa was detected , Vasa was also more dispersed in rbpms2 mutant oocytes ( Fig . S3 in S1 Supporting Information ) . Whereas another Balbiani body localized protein , Macf1/Mgn [44] was not detected in these early stage oocytes ( Fig . S3 in S1 Supporting Information ) . Therefore , Macf1 protein may only accumulate in more mature Balbiani bodies , consistent with its role in Balbiani body dispersal [44 , 45] . Thus , we conclude that Rbpms2 is not required for initial Buc or Vasa translation , but may play a role in localizing Buc protein or regulating Bb morphology .
Stress granules are ribonucleoprotein particles formed in the cytoplasm of somatic cells during stress responses that stall the initiation of protein translation; for example , stress granules form in response to environmental stress such as heat shock , translation-initiation blocking drugs , or in response to overexpression of RNA-binding proteins that inhibit translation ( reviewed in ( Buchan and Parker , 2009; Panas et al . , 2016 ) ) . Stress granules store mRNAs that are in the process of translation initiation , and typically contain numerous components of translation initiation complexes including poly ( A ) positive mRNAs , the 40s ribosomal subunit , numerous eukaryotic initiation factor proteins ( eIF2 , 3 and 4 ) , as well as RNA helicases , and other RNAbps ( reviewed in ( Buchan and Parker , 2009; Kedersha and Anderson , 2009 ) . Previous work on human RBPMS and RBPMS2 has found that these proteins partially overlap with poly ( A ) + mRNA and the stress granule marker G3BP1 ( GTPAse SH3-Domain Binding Protein/ Stress Granule Assembly factor 1 ) , consistent with a possible role in translational repression in the stress granule ( Farazi et al . , 2014 ) . We found that zebrafish fusion proteins containing wild-type GFP-Rbpm2a or GFP-Rbpms2b localize to punctate structures in the cytoplasm of HEK 293 and zebrafish somatic cells ( Fig 10A and 10B ) . In zebrafish somatic cells , these granules partially overlapped with the p-body marker Dcp2 but not the stress granule marker Tial-1 . Granule localization was disrupted for GFP-Rbpms2aae30 and GFP-Rbpms2bae32 , which are expressed in the cytoplasm of HEK 293 and in the cytoplasm and nucleus of zebrafish blastomeres . In HEK 293 cells , localization was only partially impaired in GFP-Rbpms2bsa9329 , which was not in granules but instead showed a striking localization to the centrosome in zebrafish somatic cells ( Fig 11A and 11B ) . Interestingly , the C-terminal part of Rbpms2 that is missing from GFP-Rbpms2aae30 and GFP-Rbpms2bae32 ( which have intact residues 1–107 and 1–116 , respectively ) has previously been shown to be dispensable for RNA-binding in electrophoretic mobility shift assays ( EMSAs ) ( Farazi et al . , 2014 ) , and for dimerization ( Sagnol et al . , 2014 ) . Furthermore , GFP-Rbpms2bsa9329 , which partially lacks RNA-binding residues coded for by the ( skipped ) third exon , and GFP-Rbpms2bΔRNP1 , which lacks all the RNA-binding residues of RNP1 , can still localize to granules in HEK 293 cells and to the centrosome of zebrafish somatic cells , where the wild-type protein can also be detected ( Fig 10A and 10B ) . Thus , the abnormal localization of the truncated mutant proteins GFP-Rbpms2aae30 and GFP-Rbpms2bae32 , is likely not due to inability to bind RNA or dimerize . This suggests that perhaps another protein interaction that is mediated through the C-terminus of RBPMS2 facilitates recruitment of these proteins to somatic cell granules . The germ granules of PGCs share many components and general features of somatic cell stress granules and p-bodies , but also contain RNAs and proteins that are unique to the germline ( reviewed in ( Voronina et al . , 2011 ) ) . Injection of exogenous GFP-tagged rbpms2a/b RNAs resulted in the localization of the translated proteins to the PGC germ granules . Localization of all three mutant proteins , GFP-Rbpms2aae30 , GFP-Rbpms2bae32 , and GFP-Rbpms2bsa9329 , was absent or severely reduced in germ granules although diffuse expression in the germ cell cytoplasm was robust ( Fig 11A and 11B ) . However , unlike in somatic cell granules , the lack of the RNP1 motif in GFP-Rbpms2bΔRNP1 prevented this mutant fusion protein from becoming enriched in PGCs ( Fig 10A and 10B ) . Based on this result , we speculate that interaction with RNA is likely required for enrichment or stabilization of Rbpms2 proteins to the PGC cytoplasmic compartment; however , localization within the granules depends on a separate mechanism . A similar phenomenon may operate for the localization and stability of Rbpms2 in the oocyte Balbiani body because the fusion protein coded for by the transgenic line [buc:RFP-rbpms2b] localizes to the Bb like the endogenous protein; however , the product of [buc:RFP-rbpms2bΔRNP1] transgenics does not localize to the Bb or yield detectable stable protein , despite comparable transcript expression from both transgenes ( Fig 11A and 11B ) . Therefore , Rbpms2 interaction with germ plasm RNAs may be required for efficient translation of rbpms2 RNA or stability of the protein in germ cells including primordial germ cells and the primary oocyte . We predicted that the zebrafish rbpms2a and rbpms2b , might function redundantly in the germline based on their overlapping expression domains , and their extremely similar protein sequences . Furthermore , Rbpms2a and Rbpms2b can both interact with Bucky ball protein in co-transfection and co-IP experiments . Thus , we strategically targeted both rbpms2 genes for Crispr-Cas9 mutagenesis . Indeed , single mutants for rbpms2a or rbpms2b had no discernable embryonic or adult phenotypes , supporting our prediction of functional redundancy between these genes . However , loss of both rbpm2a and rbpms2b resulted in cardiac phenotypes in zebrafish embryos , consistent with the reported role for hermes in embryonic heart development ( Gerber et al . , 2002; Gerber et al . , 1999 ) . Furthermore , loss of both rbpms2 genes resulted in defective oocyte differentiation , and a male only phenotype , confirming that these genes are likely redundant in their reproductive functions . Of the adult rbpms2a;rbpms2b double mutants ( rbpms2 mutants ) that escaped the embryonic cardiac phenotype , we could recover only fertile males . Due to the complex manner in which zebrafish sexual differentiation is regulated , we reasoned that this all-male phenotype was likely the result of a defect in an rbpms2-dependent aspect of germline development . Nonetheless , we observed comparable gonad development between rbpms2 mutants and their non-mutant siblings during early embryonic development ( d3 ) when mutants and siblings had similar numbers of PGCs , and later at the bipotential stage ( d21 ) when mutants and siblings had comparable germ cell morphologies and composition of meiotic-stage gonocytes . At the onset of sexual differentiation ( d35 ) , approximately half of the wild-type siblings will differentiate ovaries containing numerous primary oocytes , and the other half develop primitive testes containing numerous spermatogonia that are characteristically small in size , with high nuclear-to-cytoplasmic ratios , and present in large clusters . However , gonads of rbpms2 mutants do not complete differentiation as ovaries and instead were categorized as either testis-like gonads resembling those of their male siblings , or intersex gonads with some spermatogonia , but also many germ cells that morphologically resemble oocytes . Many of the rbpms2 mutant oocytes were significantly large , reaching diameters comparable to the oocytes of their siblings . Furthermore , rbpms2 mutants express the female-specific marker Bucky ball in their oocytes ( Bontems et al . , 2009; Heim et al . , 2014 ) . Importantly , it should be noted that the role of rbpms2 in promoting oogenesis is Bucky ball-independent , since zygotic buc mutants progress normally through oogenesis , and can become adult females with mature ovaries ( although buc mutant eggs are un-patterned ) ( Bontems et al . , 2009; Dosch et al . , 2004; Heim et al . , 2014; Marlow and Mullins , 2008 ) . In contrast , rbpms2 mutants initiate oocyte differentiation and maturation; however , mutant oocytes ultimately fail to complete differentiation and are lost in a tp53-independent manner , resulting in apparently normal spermatogenesis . Rbpms2 protein interacts with the Bucky ball protein and RNA transcript ( Heim et al . , 2014 ) ; therefore , we examined Bucky ball protein in rbpms2 mutant oocytes to determine if Bucky ball abundance or localization is Rbpms2-dependent . The presence and asymmetric pattern of Buc localization in rbpms2 mutant oocytes indicates that Rbpms2 is not required for translation of the bucky ball transcript . This is in agreement with other studies suggesting that Xenopus Hermes and human RBPMS proteins are likely to be translational repressors , rather than activators ( Farazi et al . , 2014; Song et al . , 2007 ) ; however , it is also possible that other RNA-binding proteins , like the related Rbpms ( 1 ) , act redundantly with Rbpms2 to promote buc translation . In wild-type oocytes prior to the Bb stage , Buc staining typically appears as a crescent abutting the oocyte nucleus and then coalesces to form a spherical Balbiani body ( Heim et al . , 2014 ) ( Fig 11C ) . However , in rbpms2 mutant oocytes we observed more diffuse Buc staining that formed a ring-like structure not observed in wild-type oocytes . Therefore , Rbpms2 may play a role in localizing Buc protein through its C-terminus , possibly by bringing Buc molecules into proximity to promote their oligomerization , or regulating Bb morphology by interactions with other Bb localized proteins , or by preventing translation of a protein that promotes Bb disassembly ( Fig 11C and 11D ) . The only protein known to promote Bb disassembly in zebrafish is Mgn/Macf1 [44 , 45]; however , we did not detect Mgn/Macf1 in the early stage oocytes that are present in rbpms2 mutants , although Vasa and Buc were detected . Thus , it seems unlikely that premature activation of Macf1/Mgn accounts for the dispersed Buc and Vasa in rbpms2 mutants . Examination of ultrastructure with TEM in primary oocytes determined that asymmetric mitochondrial localization or accumulation of nuage is not dependent on rbpms2; however , the presence of atypical cytoplasmic inclusions is significantly enhanced in the mutant population . It seems plausible that these electron dense structures might interfere with proper Bb coalescence; thus , Rbpms2 would limit a factor that is disruptive to Bb assembly . Identification and further characterization of additional Bb-localized components is needed to determine if this peculiar localization pattern is specific to Bucky ball protein , or reflects a pattern of exclusion of Bb components such as mitochondria and germ plasm by the large electron dense structures present in rbpms2 mutants . In summary , using transient and transgenic localization assays we identified unique domains of Rbpms2 that mediate its localization to subcellular structures of germ cells and somatic cells . We determined that an intact RNA binding domain is dispensable for Rbpms2 localization to granules of HEK 293 cells , and Rbpms2 subcellular localization to the centrosome in somatic cells of the zebrafish blastula , but is required for Rbpms2 association with germ granules in primordial germ cells and the Balbiani body of oocytes . Using CRISPR-Cas9 mutagenesis we showed that the two zebrafish Rbpms2 genes function redundantly in heart development and in oogenesis . Establishment of oocyte polarity and initial translation of Buc and Vasa proteins do not require Rbpms2 protein; however , Rbpms2 is required for normal Balbiani body structure and to prevent formation of aberrant cytoplasmic bodies . In addition to its role in preserving Balbiani body architecture , we discovered an independent and novel role for Rbpms2 in maintaining ovary fate . Further analyses are required to determine how Rbpms2 fits into the current framework of factors known to promote oogenesis .
Wild-type zebrafish embryos of the AB strain were obtained from pairwise matings and reared according to standard procedures [46] . Embryos were raised in 1X Embryo Medium at 28 . 5°C and staged as described [47] . All procedures and experimental protocols were in accordance with NIH guidelines and approved by the Einstein ( protocol #20140502 ) and ISMMS ( protocol # 17–0758 INIT ) IACUCs . The zebrafish rbpms2bsa9329 allele was obtained from the Sanger Institute’s Zebrafish Mutation Project [24] . The zebrafish transgenic reporter lines ziwi:GFP and cyp19a1a:GFP were obtained from Bruce W . Draper [22 , 31] . Zebrafish rbpms2aae30 and rbpms2bae32 mutants were made using CRISPR-Cas9-mediated mutagenesis as described [23] . Guide RNAs ( gRNA ) targeting the fifth exon of rbpms2a and rbpms2b were designed using the open-access ZiFiT Targeter website ( https://crispr-cas9 . com/96/zifit-targeter-crispr-cas9/ ) , resulting in the targeting sequences of 5’-GGGTGCAGGTTGGAAGGGTT-3’ for rbpms2a and 5’-GGGTGGATATTTGTGGGATT-3’ for rbpms2b . The rbpms2 targeting oligos were ligated into the gRNA expression vector pDR274 ( Addgene Plasmid #42250 ) . The gRNAs were synthesized using MAXIscript T7 Kit ( Life Technologies , AM1312M ) . The Cas9 RNA ( Addgene Plasmid #42251 ) was synthesized using mMESSAGE MACHINE SP6 Transcription Kit ( Life Technologies , AM1340 ) and Poly ( A ) Tailing Kit ( Life Technologies , AM1350 ) . 100ng of gRNA was co-injected with 300ng of cas9 RNA into the cytoplasm of 1-celled zebrafish embryos . T7 Endonuclease I assays ( NEB , M0302 ) and sequencing were used to confirm mutagenesis at d2-3 . Injected embryos were raised to adulthood and their progeny were screened to identify founders with germline mutations . Identified alleles were outcrossed to wild-type AB fish prior to incrossing . Genomic DNA was extracted from adult fins or single embryos using standard procedures [46] . The genomic region surrounding rbpms2aae30 was amplified using the primers 5’-GGGAAGCACCGCTTACAATA-3’ and 5’- TTTGACTCACATGGGTCTCG-3’ , followed by digestion of the wild-type strand with the enzyme BsurI ( ThermoFisher , FD0154 ) . The genomic region surrounding rbpms2bae32 was amplified using the primers 5’- GCGTGTAGTTTGTGTCCACC-3’ and 5’- TGTGGGCCGGAAACTTACAT-3’ , followed by digestion of the mutant strand with the enzyme EcoRV ( ThermoFisher , FD0304 ) . Finally , the genomic region surrounding rbpms2bsa9329 was amplified using the dCAPs primers 5’- CACTTATCAAGCTAACTTCAAAGCAGC-3’ and 5’- TGAAAGGGGACAAATAAGTCA-3’ , followed by digestion of the wild-type strand with the enzyme HpyF3I ( ThermoFisher , FD1884 ) . After 40 cycles of PCR at 60°C annealing , samples were digested for one hour using specified restriction enzyme . Digested PCR products were resolved using a 1 . 5% Ultrapure agarose ( Invitrogen ) and 1 . 5% Metaphor agarose ( Lonza ) gel . Genotyping for the tp53M214K allele was performed as previously described ( Berghmans et al . , 2005 ) . For whole mount ISH , embryos treated with PTU ( to prevent pigment formation ) were collected at specified stages , fixed in 4% paraformaldehyde overnight at 4°C , washed in PBS , and dehydrated in methanol . ISH was performed according to standard protocols [48] , except hybridization was performed at 65°C , maleic acid buffer ( 100mM maleic acid , pH 8 , 150mM NaCl ) was substituted for PBS during antibody incubations with alkaline phosphatase ( AP ) - conjugated anti-DIG antibodies ( Roche , 11093274910 ) , and BM Purple was used to develop the chromogenic reaction ( Roche , 1442074 ) . For fluorescent ISH in oocytes , whole ovaries were fixed as described above and the same ISH protocol was used until the point of chromogenic detection , at which point Fast Red Tablets ( Roche , 11496549001 ) dissolved in 2 ml 0 . 1M Tris-HCl was used for the fluorescent detection of AP enzyme activity . To generate antisense probes for rbpms2 transcripts , a linear template with T7 promoter was amplified from pcs2-rbpms2a and pcs2-rbpms2b plasmids using the primers: 5’-CAAACGCTGCGTCTGGAGT-3’ ( rbpms2a F ) , 5’-TAATACGACTCACTATAGGGCATGTCTCCACCTTTCA-3’ ( rbpms2a T7 R ) , and 5’-GCTAAGGCCAACACGAAGAT-3’ ( rbpms2b F ) , 5’-TAATACGACTCACTATAGGGCACTGGGCTACACTTC-3’ ( rbpms2b T7 R ) . For rbpms1 sense and antisense probes , rbpms1 amplicon ( see RT-PCR for primers ) was TOPO-cloned into pCR2 . 1 ( ThermoFisher , K450001 ) , and plasmids were linearized with HindIII ( ThermoFisher , FD0505 ) . Probes were in vitro transcribed with digoxigenin-UTP labeling kit ( Roche , 11175025910 ) . Whole-mount ISH of embryos was imaged using an Olympus SZ61 dissecting microscope with a high- resolution digital camera ( model S97809 , Olympus America ) and Picture Frame 2 . 0 software . Constructs used for transient expression assays and co-immunoprecipitation assays were made using the Gateway Recombination LRII Cloning Enzyme Mix ( Invitrogen , 11791 ) to insert the coding sequence of wild-type or mutant rbpms2a or rbpms2b into the destination vectors pCS-GFP Dest ( Addgene Plasmid #13071 ) or pCS-MT Dest ( Addgene Plasmid #13070 ) [49 , 50] . The coding sequences of rbpms2a and rbpms2b were amplified using Easy-A High Fidelity Taq polymerase ( 600400 , Agilent ) and the PCR fragments were TOPO cloned into pCR8/GW/TOPO ( K250020 , Invitrogen ) . Wild-type rbpms2a and rbpms2b were PCR-amplified from AB strain ovary cDNA and sequenced , while the mutant alleles rbpms2aae30 , rbpms2bae32 and rbpms2bsa9329 were PCR-amplified from cDNA from 4-cell stage maternal-zygotic ( MZ ) mutant embryos and sequenced , using the primers: 5’- ATGAGTCTGAAGTCAGATTCAGAGAC-3’ ( rbpms2a ATG F ) and 5’- TTAACAGAACTGACGTGATTTCC-3’ ( rbpms2a stop R ) , or 5’- ATGAGTGTCAAGTCCGACTC-3’ ( rbpms2b ATG F ) and 5’- TTAACAGAACTGTCGGGATTTCC-3’ ( rbpms2b stop R ) . The constructs used to generate stable transgenic lines ( Tol2-cmcl2GFP-bucP-mApple-rbpms2-3’UTR and Tol2-cmcl2GFP-bucP-mApple-rbpms2bΔRNP-3’UTR ) were made using multiple-fragment cloning with the Gateway Recombination LRII+ Cloning Enzyme Mix ( Invitrogen , 12538120 ) to insert the coding sequence and 3’UTR of wild-type rbpms2b into the Tol-2 destination vector pDestTol2CG2 ( Tol2 kit v1 . 2 #395 ) behind the bucky ball promoter ( p5E-bucP ) and mApple RFP sequence ( pME-mApple , Tol2 kit v2 . 0 #763 ) [14 , 49] . The rbpms2b coding and 3’UTR sequence was TOPO cloned using amplified products from ovary cDNA into pCR8 as described above using the same F primer ( rbpms2b ATG F ) and 5’-GCATTCCAATTTTAATACTATCATACAGTAGTTTCTTT-3’ ( rbpms2b 3’UTR R ) . pCS-GFP-rbpms2bΔRNP1 and Tol2-buc-mApple-rbpms2bΔRNP-3’UTR constructs were made using the QuikChange II Site-Directed Mutagenesis Kit ( Agilent , 200523 ) and mutagenesis primers: 5’-ATCAAGCTAACTTCAAAGGACAGTCGTTCTGGCGCT-3’ ( rbpms2bΔRNP1 F ) , and 5’-AGCGCCAGAACGACTGTCCTTTGAAGTTAGCTTGAT-3’ ( rbpms2bΔRNP1 R ) . The rbpms2bΔCterm constructs were made from pcs2-MT-rbpms2b using the QuikChange II Site-Directed Mutagenesis Kit ( Agilent , 200523 ) and mutagenesis primers 5’-GGACTCGTCCCAGCCTGGATTAAGATTCGAGCCTCTAGAAC-3’ ( rbpms2bΔCterm F ) and 5’-GTTCTAGAGGCTCGAATCTTAATCCAGGCTGGGACGAGTCC-3’ . For assays of localization in somatic cells and PGCs , GFP-Rbpms2 and RFPnos3UTR [51] plasmids were linearized and then transcribed using the mMessage mMACHINE SP6 Transcription Kit ( AM1340 , Invitrogen ) . For GFP-rbpms2 and RFPnos3UTR plasmids , 200pg of RNA was injected into 1-cell embryos . For analysis of somatic cells , injected embryos were fixed at sphere stage in 4%PFA , and for PGC analysis embryos were fixed at 30hpf . Antibody staining was performed as indicated below in the Immunofluorescence section . For whole-mount IF of embryos or ovaries , tissue was fixed in 4% paraformaldehyde overnight at 4°C , dehydrated in MeOH , and placed at—20°C . Two anti-Rbpms2 antibodies were used in this study , mαRbpms2 ( Abcam , ab169394 ) a mouse polyclonal raised to full-length human protein ( amino acids 1–209 NP_919248 ) , and rαRbpms2 ( abcam , ab170777 ) a rabbit polyclonal antibody raised to a peptide within amino acids 120–149 were diluted at 1:500 , Anti-Bucky ball y1165 at 1:500 [14] , and anti-GFP antibody ( Invitrogen , A10262 ) was used at 1:500 . Chicken anti-Vasa antibody was a gift of Bruce W . Draper and used at 1:1000 dilution . Rabbit anti-ACF7/Macf1 was used at 1:1000 [52] . Rabbit anti-DCP2 ( Novus Biologicals , NBP2-16109 ) and rabbit anti- TIAL-1 ( Novus Biologicals , NBP1-79932 ) were used at1:2000 as in [29] . Alexafluor488 , Alexafluor568 , CY3 , C5 ( Molecular Probes ) secondary antibodies were diluted at 1:500 . Images were acquired using a Zeiss Axio Observer inverted microscope equipped with Apotome or ApotomeII and a CCD camera , or Zeiss Live DuoScan ( line-scanning ) Confocal . Images were processed in ImageJ/FIJI , Adobe Photoshop and Adobe Illustrator . In zebrafish , Vasa protein localizes in a perinuclear ring of staining around each PGC nucleus that can be used much like a nuclear marker to identify and count individual cells . Z-series image stacks of one lateral half of embryonic gonads were obtained using a Zeiss Live DuoScan ( line-scanning ) confocal microscope , and cells in the stack were manually counted by analyzing the slices and nuclear morphology comprising each Z-stack with ImageJ/FIJI . Co-IPs were performed as previously described ( Heim et al . , 2014 ) . Briefly , sub-confluent HEK293 cells ( 1×106 ) were transfected with 3 μg pCS3-MT-Rbpms1/2a/2b , or the specified pCS3-MT-Rbpms2b mutation/truncation , and 3 μg pCS3-GFP-Buc or pCS3-GFP ( control ) with 3 to 1 ratio of polyethylenimine:DNA , overnight . IP was performed on supernatant from cell lysates ( including mRNA ) with 1 mg of anti-GFP 3E6 antibody ( A11120 , Invitrogen ) and Protein G magnetic beads ( S1430 , NEB ) . Precipitated proteins were separated by SDS-PAGE , and transferred to ImmobilonP ( Millipore ) . Membranes were blotted using 1:2000 anti-Myc ( 9E10 , Santa Cruz ) or 1:2000 anti-GFP ( 11814460001 , Roche ) , washed in TBST , then stained with 1:25 , 000 goat anti-mouse HRP ( 12–349 , Millipore ) prior to ECL detection ( GE Healthcare ) . To visualize granules in HEK293 cells , the same transfection protocol was used with pCS3-GFP-Rbpms wild type and mutant constructs , and the cells were plated onto cover-slipped dishes and photographed live . For semithin and ultrathin sections , samples were fixed with 2 . 5% glutaraldehyde , 2% paraformaldehyde in 0 . 1 M sodium cacodylate buffer , postfixed with 1% osmium tetroxide followed by 2% uranyl acetate , dehydrated through a graded series of ethanol and embedded in LX112 resin ( LADD Research Industries , Burlington VT ) . Ultrathin sections were cut on a Reichert Ultracut UCT , stained with uranyl acetate followed by lead citrate and viewed on a JEOL 1200EX transmission electron microscope at 80kv . Haematoxylin and Eosin ( H&E ) staining was performed as previously described ( Heim et al . , 2014 ) . Briefly , tissue was fixed in 4% paraformaldehyde overnight at 4°C , dehydrated in MeOH , and placed at—20°C . After paraffin embedding and sectioning onto slides , tissue was deparaffinized , stained with H&E , coated with Permount solution ( Fisher Scientific ) , coverslipped , and imaged using an AxioSkop2 microscope and AxioCam CCD camera . Total RNA was extracted from pooled embryos ( n = 20-30/stage ) or pools of 2–3 adult organs using Trizol ( Life Technologies , 15596 ) . cDNA was prepared with SuperScript III Reverse Transcription Kit ( Life Technologies , 18080–051 ) . RT-PCR was performed using the primers 5’-ATTCACCTCTAAACAGCCGGT-3’ and 5’-AGGCTAGGCTAATCATTACACTG-3’ for rbpms1 , 5’-ATGCGTTAAATGGCATCCGC-3’ and 5’-GTCCTCAGCATCTCTACCGC-3’ for rbpms2a , 5’-CAACGCATCTGAGCATGAAG-3’ and 5’-GATCCAGTCGCACTTTAAGGA-3’ for rbpms2b . The primers for ef1α , vasa and cyp19a1a are as described in [14] . FASTA protein sequences corresponding to the longest known alternative transcript for each of the zebrafish rbpms genes and Xenopus hermes were obtained from the Ensembl website ( www . ensembl . org ) and analyzed using ClustalW for multiple alignments and JalView software to visualize . Phylogeny tree is calculated for neighbor joining using percent identity . Aligned sequences include rbpms ( herein rbpms1 ) ( ENSDART00000127288 ) , rbpms2a ( 20ENSDART00000067514 ) , rbpms2b ( ENSDART00000006619 ) and xhermes ( ENSXETT00000024635 ) . | Oocyte development relies on posttranscriptional regulation by RNA binding proteins ( RNAbps ) . RNAbps form large multi-molecular structures called RNPs ( ribonucleoproteins ) that further aggregate into regulatory granules within germ cells . In zebrafish primary oocytes , a large transient RNP aggregate called the Balbiani body ( Bb ) is essential for localizing patterning molecules and germline determinants within oocytes . RNA-binding protein of multiple splice forms 2 , or Rbpms2 , localizes to germ granules and the Bb , and interacts with bucky ball , a key factor for Bb formation . We show that Rbpms2 requires RNA binding for localization within germ cells , and that the C-term and RRM contribute to Rbpms2 subcellular localization in distinct somatic cell types . To investigate Rbpms2 functions we mutated the duplicated zebrafish rbpms2 genes . Consistent with redundant functions , rbpms2a and rbpms2b gene expression overlaps , and single mutants have no discernible phenotypes . Although rbpms2a;2b double mutants have cardiac phenotypes , those that reach adulthood are exclusively fertile males . Genetic analysis shows that rbpms2 mutant oocytes are not maintained even when Tp53 , a regulator of cell death is absent . Initial oocyte polarity is established in rbpms2 mutants based on asymmetric distribution of Buc protein and mitochondria; however , abnormal Buc structures and atypical cytoplasmic inclusions form . This work reveals independent Rbpms2 functions in promoting Bb integrity , and as a novel regulator of ovary fate . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"rna-binding",
"proteins",
"fish",
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"gonads",
"vertebrates",
"animals",
"xenopus",
"animal",
"models",
"osteichthyes",
"germ",
"cells",
"developmental",
"biology",
"oocytes",
"model",
"organisms",
"amphib... | 2018 | rbpms2 functions in Balbiani body architecture and ovary fate |
JIP3/UNC-16/dSYD is a MAPK-scaffolding protein with roles in protein trafficking . We show that it is present on the Golgi and is necessary for the polarized distribution of synaptic vesicle proteins ( SVPs ) and dendritic proteins in neurons . UNC-16 excludes Golgi enzymes from SVP transport carriers and facilitates inclusion of specific SVPs into the same transport carrier . The SVP trafficking roles of UNC-16 are mediated through LRK-1 , whose localization to the Golgi is reduced in unc-16 animals . UNC-16 , through LRK-1 , also enables Golgi-localization of the μ-subunit of the AP-1 complex . AP1 regulates the size but not the composition of SVP transport carriers . Additionally , UNC-16 and LRK-1 through the AP-3 complex regulates the composition but not the size of the SVP transport carrier . These early biogenesis steps are essential for dependence on the synaptic vesicle motor , UNC-104 for axonal transport . Our results show that UNC-16 and its downstream effectors , LRK-1 and the AP complexes function at the Golgi and/or post-Golgi compartments to control early steps of SV biogenesis . The UNC-16 dependent steps of exclusion , inclusion and motor recruitment are critical for polarized distribution of neuronal cargo .
The secretory pathway in a cell involves the synthesis and trafficking of proteins through the ER-Golgi network and their subsequent targeting to different sub-cellular compartments . Generation of a defined transport carrier , with a characteristic protein and lipid composition , along the trafficking pathways is known to involve at least three steps ( a-c , see below ) , several occurring at the trans-Golgi network ( TGN ) . ( a ) Protein sorting where the secretory cargo is segregated away from Golgi resident proteins [1–4] . For example , segregation of different Regulated Secretory Proteins ( RSP ) such as POMC occurs via receptor-mediated sorting [5 , 6] . ( b ) A post-sorting step where clustered cargo undergoes budding and separation to form a vesicular compartment from the donor membrane . In part , the adaptor protein ( AP ) complexes regulate such steps by recognizing signal sequences on proteins and ensuring that they are sorted into appropriate compartments [7 , 8] . The AP-1 complex recruits proteins like Clathrin , which causes membrane deformation followed by budding and scission from the TGN and post-Golgi compartments [9 , 10] . ( c ) A third step is the recruitment of specific motors , dependent on the characteristic membrane composition , constituting proteins and/or lipids , of the newly formed cargo . For example , AP-1 interacts with the Kinesin-3 motor KIF13A to coordinate endosomal sorting during melanosome biogenesis [11] . These events ensure the formation of a defined transport carrier that gets targeted to a specific sub-cellular compartment . Protein sorting occurs at post-Golgi compartments as well . For example , during the multi-step maturation of secretory granules , sorting of proteins also occur post-Golgi at intermediate compartments known as the immature secretory granules ( ISGs ) [12] . However , the genes that regulate such processes remain to be well understood . Synaptic vesicle proteins ( SVPs ) are essential for neurotransmission and synaptic vesicles at the synapse are known to have a defined composition [13] . Several SVPs have transmembrane domains and are trafficked out of the TGN through the regulated secretory pathway to the synapse . Each synaptic vesicle found at the pre-synaptic bouton has an assortment of proteins important for processes such as neurotransmitter filling ( VGLUT1 [14] ) , docking and neurotransmitter release ( SNB-1 , SNT-1 [15–17] , and fusion of synaptic vesicles with the plasma membrane ( UNC-13 [18] ) . It is not clear whether all of these proteins are found on a single SVP transport carrier as it exits the cell body . An earlier study indicates that different SVPs , for example Synaptophysin and SV2 are associated with distinct pools of membranous organelles [19] . This could imply that different SVPs might travel in separate compartments before they come together in a mature synaptic vesicle that is found at the synapse . However , another study suggests that most or all SVPs , including Synaptobrevin and SV2 , are transported in a single transport carrier to the presynaptic active zone [20] . Studies using PC12 cell lines and in vivo studies in C . elegans have shown that the AP-3 complex is required for synaptic vesicle biogenesis , potentially directly from the Golgi , and for the axonal targeting of SVPs respectively [21–23] . SVPs are also known to require the molecular motor KIF1A for their exit from the cell body and transport to the synapse in multiple systems [24 , 25] . Thus , although proteins such as AP-3 and KIF1A have been implicated in SVP transport carrier biogenesis and trafficking , several aspects of these early steps remain unclear . For example , it is not fully understood how the AP complexes confer specificity to the sorting of different SVPs . Mechanisms that ensure separation of SVPs and/or lipids of the cargo membrane from those integral to the Golgi membrane is also poorly understood . Further , we do not know in what order these multiple steps are carried out during SVP trafficking and transport carrier biogenesis JIP3/UNC-16/dSYD , a JNK-signaling scaffold protein present at the Golgi in Drosophila , is thought to have roles in protein trafficking [26–28] . The C . elegans unc-16 mutants also show mis-trafficking and mis-accumulation of multiple neuronal cargo such as SVPs , dendritic receptors , lysosomes and early endosomes [28 , 29] . It is unclear how UNC-16 , a molecule known to interact with and scaffold multiple kinases , including MAPK family members , carries out its trafficking roles . Among the UNC-16-interacting proteins , LRK-1 , the C . elegans homolog of LRRK2 , has been previously implicated in the polarized trafficking of SVPs and is also found at the Golgi , like UNC-16 [30 , 31] . Thus , UNC-16 and LRK-1 could play roles in the early steps of SVP protein trafficking . In this study , we show that the SVP transport carriers formed in unc-16 and lrk-1 mutants have an altered composition and size . These mutants also show defects in polarized distribution of SVPs , which mis-localize to the dendrites [29 , 30] . We show that UNC-16 regulates the composition via LRK-1 by excluding Golgi enzymes from the SVP transport carrier and by increasing the incidence of co-transport of SNB-1 and RAB-3 in the same transport carrier . We show that size of the SVP transport carriers formed is determined by UNC-16 via LRK-1-dependent localization of the μ-component ( UNC-101 ) of the AP-1 complex at the Golgi . We also show that the UNC-16 through LRK-1 and the AP-3 complex ensures that specific SVPs are included in the same transport carrier . Thus , based on genetic and biochemical evidence , we propose a novel role of UNC-16 in synaptic vesicle protein trafficking wherein it functions via LRK-1 to regulate exclusion and inclusion of proteins and consequently motor recruitment on the carrier . The AP-1 and AP-3 complexes likely function at the Golgi and/or post-Golgi intermediate compartments downstream to UNC-16 and LRK-1 , to regulate respectively the size and composition of the SVP transport carrier formed in the neuronal cell body .
UNC-16/JIP3/dSYD has been implicated in the trafficking of multiple proteins such as those associated with synaptic vesicles and lysosomes [28 , 29 , 32] . To investigate the role it plays in regulating protein trafficking , we first examined the localization of UNC-16 . Using two independent Golgi markers , Mannosidase-II ( Man-II ) and RUND-1 [33] , we show that UNC-16 localizes to the Golgi in C . elegans neuronal cell bodies ( Fig 1a ) . A similar localization of mammalian dSYD/JIP3 has been reported in CV-1 epithelial cells [26] . We isolated a new allele of unc-16 ( tb109 ) , with an early stop codon at amino acid 423 , that caused mis-trafficking of the trans-membrane VAMP Synaptobrevin-1 ( SNB-1 ) to the dendrite of the amphid sensory neurons ( Methods , S1a and S1b Fig , S1 Table ) . This phenotype is identical to those reported in other unc-16 alleles [29] . The dendritic mis-localization of SNB-1 in tb109 was rescued by the transgenic expression of wild type UNC-16 ( S1b Fig , S1 Table ) . Upon examination of a dendritic receptor ODR-10 in the AWB neuron we found that , unlike in wild type animals , ODR-10 is ectopically localized to the axonal compartment in unc-16 mutants ( S1c Fig ) . Loss of UNC-16 thus leads to the loss of polarized distribution of cargo in neurons . This mis-trafficking is not dependent on the orientation of microtubules , which is similar in both unc-16 and wild type ( S1d Fig ) . The defects in axonal and dendritic targeting along with the reported trafficking defects in dendritic , endosomal and lysosomal proteins suggests that UNC-16 acts as a general regulator of early events in the trafficking pathway [27–29 , 34] . To understand the nature of early defects in unc-16 , potentially occurring at the Golgi , we examined both—the localization of Golgi enzymes and whether cargo such as SVP transport carriers had altered composition . Unlike wild type , in unc-16 mutants , the Man-II enzyme mis-localizes to dendritic tips of the ASI neuron ( S2g Fig , S1 Table ) and both Man-II and Sialyl transferase ( ST ) are present as discrete compartments throughout the touch receptor neuron ( TRN ) process , up to the synapse ( Fig 1b , S2b Fig ) , similar to reported observations [28 , 34] . Comparable to wild type , in unc-16 mutants Golgi resident enzyme Man-II and ST continue to localize as 1–3 large puncta in the cell body ( Fig 1b , S2b and S2d Fig ) . Other Golgi markers , such as RUND-1 and RAB-6 . 2 , show a punctate distribution in the neuronal cell body of unc-16 mutants , very similar to those observed in wild type animals , with no gross changes in the number or position of Golgi puncta ( S2d Fig ) . This suggests that potentially only a subset of Golgi markers is mis-trafficked into the axons of unc-16 animals . In order to check if the mis-trafficking of Golgi proteins seen in unc-16 alters SVP transport carriers , we carried out dual colour imaging of both Man-II and RAB-3 . About 86% of RAB-3 containing compartments emerging from the cell body carry the Golgi enzyme Man-II , unlike in wild type where only ~ 5% of RAB-3 marked compartments co-transport Man-II ( Fig 1d , S1 and S2 Movies ) . Similar observations were also made with the Golgi enzyme ST ( S2a Fig ) . The mis-trafficking of Man-II along with RAB-3 into the neuronal process in unc-16 mutants is rescued by the transgenic expression of wild type UNC-16 ( Fig 1d ) . Thus , UNC-16 is necessary to restrict Golgi resident proteins to the cell body and to exclude them from SVP transport carriers . The defects seen in unc-16 are likely due to disruption of retention/retrieval mechanisms , leading to mis-trafficking of Golgi enzymes . To assess the membrane composition of the atypical SVP transport carriers formed in unc-16 , we examined the co-transport of two synaptic vesicle proteins , RAB-3 and SNB-1 , in TRNs . In wild type animals , the incidence of co-transport of RAB-3 and SNB-1 is ~ 39% of the mobile SVP transport carriers . In unc-16 however , the frequency of co-transport of RAB-3 and SNB-1 reduces by half to ~19% ( Fig 1c and 1e ) . This was corroborated by the reduction in co-localization of endogenous transmembrane Synaptotagmin ( SNT-1 ) and RAB-3 at non-synaptic regions of the sub-lateral cord from ~95% in wild type to ~60% in unc-16 animals ( S1e Fig ) . Unlike in wild type , these erroneous SVP transport carriers exit the cell body as long tubular compartments . These long compartments show a ~1 . 5-fold increase in average length compared to those found in wild type and are three times more frequent in multiple alleles of unc-16 across several neuronal cell types ( Fig 2a , S3a Fig , S2 Table , S3–S5 Movies ) . Electron micrograph analyses showed an increase in the width of vesicles in non-synaptic regions of the dorsal and ventral nerve cord in unc-16 compared to wild type animals ( Fig 2b ) . Such larger vesicular profiles have been previously seen at the synapses of unc-16 animals [34] , [35] . The longer SVP transport carriers we see in our live imaging could contribute to these wider vesicular profiles observed in our electron micrographs ( S4 Movie ) . We also verified that mutants in known interactors of UNC-16 such as jnk-1 , unc-116 and dhc-1 [29 , 32 , 36] do not show these large moving tubular profiles carrying either SNB-1 or RAB-3 , nor do they mis-localize the Golgi enzyme ST into the TRNs ( S2 Table , S2b Fig ) . Thus , UNC-16 facilitates “exclusion” of Golgi enzymes ( see above ) from SVP transport carriers , “inclusion” of specific SVPs into the same transport carrier and regulates the size of such compartments exiting the cell body . To identify other genes in the UNC-16-mediated SVP trafficking pathway , we examined mutants in LRRK2/LRK-1 , known to be present on the Golgi [30] . LRRK2/LRK-1 regulates the trafficking and distribution of synaptic vesicles in presynaptic boutons [37] and lrk-1 mutants show mis-trafficking of SNB-1 into dendrites of C . elegans chemosensory neurons , similar to phenotypes seen in unc-16 animals [30] . In lrk-1 animals , Golgi enzymes Man-II and ST occasionally exit the cell body ( Fig 3b and S2a Fig ) . However , unlike in unc-16 , Man-II does not mis-localize to the dendrite in lrk-1 animals ( S1 Table , S2c Fig ) . In lrk-1 animals the incidence of compartments carrying both Golgi enzyme and SVP is closer to wild type , with only ~18% of the mobile compartments co-transporting Man-II and RAB-3 ( Fig 3b , S2a Fig ) . The incidence of co-transport of RAB-3 and SNB-1 in the same compartment reduces in lrk-1 animals , with the severity of the phenotype similar to that observed in unc-16 ( Fig 3c ) . Additionally , there is a ~5-fold increase in the frequency of longer SVP transport compartments seen exiting the cell body in both alleles of lrk-1 examined ( Fig 3d and 3e , S3b Fig ) . As reported earlier , we also found that the dendritic marker ODR-10 shows a wild type-like polarized distribution in lrk-1 animals [32] . Taken together , our observations suggest that LRK-1 plays a critical role in regulating the composition as well as the size of the SVP transport carriers formed at the cell body , potentially at the Golgi . However , lrk-1 mutants lack the aberrant distribution of dendritic and Golgi proteins seen in unc-16 . Since lrk-1 has trafficking defects that are similar to unc-16 , we tested whether UNC-16 and LRK-1 genetically function in the same pathway . We built double mutants and firstly assessed phenotypes present in unc-16 but absent in lrk-1 . The Golgi enzyme Man-II is mis-trafficked into RAB-3-containing compartments similar to unc-16 single mutants ( Fig 3b ) . Thus , the lrk-1; unc-16 double mutants are similar to unc-16 single mutants alone . Further , comparable to unc-16 and lrk-1 single mutants , the lrk-1; unc-16 double mutants show loss of polarized distribution of SNB-1 ( Fig 3a ) , reduced co-transport of RAB-3 and SNB-1 in the TRN process ( Fig 3c ) and an increased frequency of long moving compartments carrying RAB-3 ( Fig 3e ) . We next determined whether UNC-16 acts upstream of LRK-1 in the trafficking of SVPs by overexpressing transgenic LRK-1 in unc-16 animals . Overexpression of LRK-1 greatly reduces the dendritic mis-localization of SNB-1 in unc-16 animals ( Fig 4a , S2 Table ) . Over-expression of LRK-1 was also sufficient to exclude Golgi enzyme Man-II from RAB-3 containing SVP transport carriers ( Fig 4b ) , to restore incidence of co-transport of RAB-3 and SNB-1 to wild type ( Fig 4c ) and to reduce the frequency of long transport carriers to frequencies seen in wild type ( Fig 4d and 4e , S3c Fig ) . Interestingly , the overexpression of LRK-1 in unc-16 animals does not suppress the mis-trafficking of the dendritic marker ODR-10 into the axon ( S1c Fig ) nor does it completely suppress the exit of Man-II into the neuronal process ( although these no longer travel in RAB-3 containing transport carriers ) . This suggests that transgenic LRK-1 only ameliorates the SVP-specific trafficking defects observed in unc-16 . As LRK-1 and UNC-16 are both present on the Golgi , we examined the localization of each protein and found that in unc-16 animals the punctate localization of LRK-1 on the Golgi is reduced in neuronal cell bodies ( Fig 4f ) . In unc-16 , a 10-fold increase was seen in the number of cell bodies showing a completely diffuse localization of LRK-1 , compared to wild type animals ( S4d Fig ) . In addition , LRK-1 is also mis-localized into the dendrites of sensory neurons in unc-16 animals ( S4c Fig ) . On the other hand , UNC-16 localization remains unaffected in lrk-1 animals ( S4e Fig ) . We further examined if both these proteins were part of the same complex . Immunoprecipitation experiments from C . elegans expressing LRK-1::FLAG and UNC-16::GFP show that LRK-1 and UNC-16 are present together in a complex in vivo ( S4a and S4b Fig ) . Given that overexpression of LRK-1 in unc-16 is sufficient to restore the processes of exclusion , inclusion and size regulation , our data suggest that unc-16 functions genetically upstream of lrk-1 . This , along with the observation that lrk-1 mutants have modest exclusion defects but inclusion defects similar in severity to unc-16 , suggests that exclusion may precede inclusion during SV biogenesis . The presence of UNC-16 likely facilitates the localization of LRK-1 on the Golgi and both together , possibly in a complex , regulate the trafficking of multiple SVPs . Dendritic trafficking of proteins is known to require the AP-1 complex [10 , 23 , 38] . The mis-trafficking of SVPs into dendrites in lrk-1 mutants is dependent on UNC-101 , the μ-chain of the Adaptor protein-1 ( AP-1 ) complex in C . elegans [30] . Therefore , we tested if UNC-101 is involved in the UNC-16 and LRK-1-mediated regulation of the composition and size of the SVP transport carrier . In unc-101 animals , like in wild type TRNs , Man-II and ST are restricted to the cell body and the incidence of co-transport of SVPs RAB-3 and SNB-1 is about 40% ( Fig 5b and 5c , S2a Fig ) . Thus UNC-101 , unlike UNC-16 and LRK-1 does not appear to play a role in regulating the composition of SVP transport carriers . However , nearly 40% of the RAB-3 containing vesicles in unc-101 animals were longer in the PLM neuron ( Fig 5d and 5e , S3d Fig , S6 Movie ) , similar to unc-16 ( Fig 2a , S4 and S5 Movies ) and lrk-1 ( Fig 3d and 3e ) mutants . These longer transport carriers were seen in several other neuronal types ( S7 Movie ) as well . Thus , UNC-101 is involved in regulating the size of the SVP transport carrier leaving the cell body . In order to genetically position UNC-101 relative to UNC-16 and LRK-1 , we built double mutants with both unc-16 and lrk-1 . The mis-trafficking of the Golgi enzyme Man-II in the double mutants unc-101; unc-16 and unc-101; lrk-1 animals were similar to unc-16 or lrk-1 single mutants respectively ( Fig 5b ) . Further , the rescue of size defects in unc-16 mutants by LRK-1 over-expression was found to require UNC-101 ( Fig 5e , S3d Fig ) . Thus , UNC-101 acts downstream of both UNC-16 and LRK-1 in regulating size of the transport carrier but does not appear to influence sorting of SVPs into the carrier . Previous studies have shown that the aberrant trafficking of SNB-1 into the dendrite of the sensory neuron in lrk-1 animals is dependent on UNC-101 [30] . Consistent with this , we observed that in unc-101 , lrk-1 animals SNB-1 is excluded from the dendrite ( S2 Table ) . In unc-101; unc-16 double mutants , SNB-1 continues to be mis-localized to the dendrite in ~ 20% of the animals , unlike in unc-16 animals where mis-localization is seen in 100% of the animals ( Fig 5a , S2 Table ) . The MAN-II mis-localization into the dendrites in unc-16 and lrk-1 also requires UNC-101 ( S2c Fig ) . Thus , even in absence of LRK-1 or UNC-16 , UNC-101 continues to regulate the trafficking of proteins into dendrites . Our data , thus , suggests two roles for UNC-101 – ( i ) regulation of dendritic trafficking and ( ii ) regulating the size of axonally trafficked SVP transport carrier . The AP-3 complex is known to function downstream to LRRK2 in the trafficking of lysosomal membrane proteins and to sort axonal proteins away from dendritic proteins [23 , 39] . Thus , we examined SVP trafficking in apb-3 mutants defective in the AP-3 β-subunit . Unlike unc-101 animals , apb-3 mutants do not have longer SVP transport carriers , suggesting that AP-3 is not involved in size regulation ( Fig 5e ) . Further , similar to wild type and unc-101 mutants , the Golgi protein Man-II largely stays restricted to the cell body in apb-3 mutants with little or no co-transport of Man-II with RAB-3 ( Fig 5b ) . However , the apb-3 mutants show inclusion defects wherein incidence of co-transport of RAB-3 and SNB-1 is reduced to ~14% ( Fig 5c ) , similar to unc-16 or lrk-1 single mutants . Both apb-3; unc-16 and apb-3 , lrk-1 double mutants show reduced co-transport of RAB-3 and SNB-1 , similar in severity to that seen in unc-16 or lrk-1 single mutants alone ( Fig 5c ) . Additionally , over-expression of LRK-1 in unc-16 mutants is unable to rescue the inclusion defects in absence of APB-3 ( Fig 5c ) . Thus , the AP-3 complex acts downstream of LRK-1 and may have roles in UNC-16 and LRK-1-mediated regulation of composition of the SVP transport carrier formed . As the phenotype of long moving compartments observed in unc-16 , lrk-1 and unc-101 is similar , we investigated whether there were changes in reported Golgi localization of UNC-101 in unc-16 and lrk-1 [23] . UNC-101 is not present as a defined puncta in unc-16 neuronal cell bodies while in lrk-1 animals , the puncta are reduced in both number and intensity as compared to wild type ( Fig 6a , 6b and 6c ) . The overexpression of transgenic LRK-1 in unc-16 animals restores the localization of UNC-101 on the Golgi ( Fig 6a , 6b and 6c ) . Since overexpression of LRK-1 rescues the size defects seen in unc-16 but not in an unc-101; unc-16 background , the LRK-1 mediated Golgi localization of UNC-101 is required for maintaining the size of the SVP transport carriers Figs 4e and 5e ) . Thus , the altered size of the synaptic vesicle precursors observed in unc-16 and lrk-1 depends on the presence of UNC-101 ( or the AP-1 complex ) on the Golgi . SVP transport carriers in unc-16 are thought to recruit non-canonical motors [29 , 36] , which may arise as a consequence of altered membrane composition . The molecular motor UNC-104 is known to transport synaptic vesicles protein transport carriers in C . elegans [24 , 39] . In concordance with this role , in unc-104 mutants , the SVPs RAB-3 or SNB-1 are found to be trapped in the cell body and absent from synapses ( Fig 7bii ) [15 , 40–42] . A previous study has shown that SVPs formed in unc-16 are transported independently of the UNC-104 motor in the DD and VD motor neurons [29] . This loss of motor-cargo specificity is also observed in TRN neurons such that SVP transport carriers travel up to the synapse in unc-16; unc-104 animals ( Fig 7biii and 7biv ) . In lrk-1; unc-104 and in apb-3; unc-104 double mutants RAB-3 was observed in the proximal portion of the neuronal process but not at synapses , suggesting that the SVP carriers formed in lrk-1 and apb-3 are only partially dependent on UNC-104 ( Fig 7bv , 7bvi and 7bx ) . The RAB-3 containing vesicles in unc-101; unc-104 were restricted to the cell body of the TRNs like in unc-104 alone ( Fig 7bix ) suggesting that unlike in unc-16 or lrk-1 mutants , the SVP transport carriers in unc-101 mutants were completely dependent on the UNC-104 motor . We tested if the overexpression of LRK-1 was sufficient to restore the dependence of the SVP transport carrier on UNC-104 in unc-16 mutants and found that in these animals the SVP transport carriers are partially dependent on UNC-104 and are unable to reach the synapse ( Fig 7bviii ) . Conversely , transgenic expression of UNC-16 does not significantly affect the altered dependence of the SVP transport carrier on UNC-104 in lrk-1 animals ( Fig 7bvii ) . These observations demonstrate that the overexpression of LRK-1 in unc-16 sufficiently changes the composition of the SVP transport carrier such that it is now largely dependent on UNC-104 . This further suggests that composition of the SVP transport carrier is critical to allow sufficient UNC-104 motor recruitment or to potentially exclude other motors along with excluding other proteins .
The mechanisms by which synaptic vesicle proteins ( SVPs ) are sorted into transport carriers and trafficked out of the cell body are not yet clearly understood . It is thought that like other membranous cargo such as secretory granules and dense core vesicles , SVPs are sorted at the Golgi and post-Golgi compartments and the transport carrier that is formed moves out of the cell body by recruiting specific motors , such as KIF1A/UNC-104 [1 , 4 , 5 , 43–45] . Using the model system C . elegans , we have uncovered a novel role for UNC-16/JIP3 in the trafficking and biogenesis of SVP transport carriers . Given that UNC-16/JIP3 localizes to the Golgi in C . elegans ( Fig 1a ) as well as in mammalian epithelial cells [26] , it likely has conserved functions at the Golgi . We show that the SVP transport carrier biogenesis roles of UNC-16 occur via LRK-1 . The sorting roles of UNC-16 and LRK-1 lead to the polarized distribution of SVPs as well as regulation of its composition and size . These two proteins regulate the composition of SVP transport carriers via the exclusion of Golgi resident enzymes and inclusion of relevant SVPs . These roles depend on the AP complexes where the AP-1 complex regulates the size of the carrier and the AP-3 complex regulates composition . Our study shows that UNC-16 is essential to prevent Golgi resident enzymes such as Mannosidase-II and Sialyl transferase from entering the SVP transport carrier as inferred from the observed mis-trafficking of these enzymes along the neuronal process in unc-16 mutants ( Fig 1b and 1d and S2a Fig ) . Previous studies have shown that retention of Golgi enzymes could occur through several mechanisms such as binding to scaffolding proteins on the Golgi or through the regulation of the lipid membrane composition to favour partitioning of different membrane proteins [46–48] . UNC-16 potentially acts as a scaffolding molecule to recruit effectors such as LRK-1 on the Golgi ( Fig 4f ) , through which it may function to retain certain Golgi resident enzymes . Although lrk-1 mutants themselves do not show exclusion defects ( Fig 3b ) , an excess of LRK-1 in an unc-16 background is sufficient to bypass the requirement of UNC-16 to exclude Golgi resident enzymes from the SVP transport carrier ( Fig 4b ) . This suggests that LRK-1 may have redundant roles in exclusion of Golgi enzymes from SVP transport carriers . Alternatively , LRK-1 could act in an analogous manner to LRRK2 , the mammalian homolog of LRK-1 , that has been implicated previously in regulating the retromer complex , which sorts proteins from the endosome-lysosome degradation pathway retrogradely to the Golgi complex [49 , 50] . This likely occurs through the action of certain RABs such as RAB-7 and RAB-9 and through LRK-1’s interaction with VPS35 of the retromer complex [49–51] . In unc-16 mutants Golgi-resident proteins may not be excluded from other compartments as well , such as lysosomes . Our data suggests that UNC-16 mediated LRK-1 localization facilitates the exclusion of Golgi enzymes specifically from the SVP transport carrier as Man-II continues to be ectopically present along the axon potentially in other compartments in unc-16 mutants overexpressing LRK-1 . An important step during protein sorting is the clustering or segregation of proteins to be placed into the same compartment away from the donor compartment proteins . For example , the SVP Synaptobrevin-II and Synaptophysin interact with each other resulting in their co-trafficking [52 , 53] . Our findings suggest that both UNC-16 and LRK-1 are required for ensuring certain SVPs ( such as SNB-1 , SNT-1 , and RAB-3 ) are sorted together and included more frequently in the same transport carrier ( Figs 1c , 1e and 3c and S1e Fig ) . Moreover , the inclusion defects seen in unc-16 can be rescued by overexpression of LRK-1 and this rescue depends on presence of a functional AP-3 complex ( Figs 4c and 5c ) . Along with the observation that lrk-1 genetically lies downstream to unc-16 and that exclusion defects are seen only in the unc-16 mutants , we hypothesize that exclusion of Golgi enzymes likely precedes inclusion during the early steps of SVP sorting . In unc-16 animals , the localization of LRK-1 itself at the Golgi is disrupted ( Fig 4f , S4d Fig ) suggesting that unc-16 may act as a potential hypomorph of LRK-1 . This , along with our biochemical evidence that UNC-16 and LRK-1 are present in the same complex ( S4a and S4b Fig ) , suggests that UNC-16 may scaffold LRK-1 at the Golgi in a physical complex that specifically regulates sorting of SVPs . In addition to protein sorting , a crucial step in the formation of a transport carrier involves the regulation of its size . Previous studies have shown that proteins present on the surface of the TGN are crucial for the recruitment of the machinery involved in size regulation such as the Adaptor protein complexes [8 , 9 , 54–57] . We found that UNC-101 , the μ-chain of the AP-1 complex in C . elegans , is indeed engaged in regulating the size of the SVP transport carrier formed ( Fig 5d and 5e , S3d Fig ) . Importantly , the localization of UNC-101 on the Golgi is regulated by UNC-16 and LRK-1 ( Fig 6 ) . Considering that the Golgi enzyme Man-II in all of these mutants show the presence of 2–3 large puncta juxtaposed to the nucleus in the cell body , like in wild type , the Golgi is likely intact and the appearance of Golgi resident enzymes in the neuronal process can be accounted for due to errors in sorting/retrieval of these proteins . Furthermore , the size regulation by UNC-101 acts downstream to the early sorting steps regulated by UNC-16 and LRK-1 since overexpression of LRK-1 was able to rescue the size in unc-16 animals but not in unc-16; unc-101 mutants ( Figs 4e and 5e , S3d Fig ) . Since exclusion or inclusion defects were absent from unc-101 animals , this also suggests that regulation of size and membrane composition could be independent processes and having an unusually large size does not necessarily incorporate other proteins ( Golgi enzymes ) typically excluded from these carriers . Earlier studies have also shown that the AP-1 complex is required for trafficking of proteins into dendrites [10 , 23 , 38] . Consistent with these observations , the dendritic mis-localization of SVPs in our mutants was largely dependent on UNC-101 ( Fig 5a , S2 Table ) . We report an additional role where it regulates the size of the axonal cargo viz . SVP transport carriers . A recent study showed that the AP-3 complex is necessary at the Golgi for axonal localization of proteins [23] . AP-3 has previously been implicated in the sorting of proteins from early endosomes to lysosomal compartments [58–60] . Our study suggests that , unlike AP-1 , the AP-3 complex does not regulate the size of SVP transport carriers ( Fig 5e ) . On the other hand , compared to wild type animals , the apb-3 mutants show inclusion defects wherein the co-transport of SVPs , SNB-1 and RAB-3 , is reduced ( Fig 5c ) , which is similar to but more severe than that observed in unc-16 mutants . A recent study showed that AP-3 complex acts downstream of LRK-1 in the endo-lysosomal trafficking pathway [61] . Since overexpression of LRK-1 is able to suppress the defects in unc-16 but not in an apb-3; unc-16 double mutant ( Fig 5c ) , AP-3 likely functions downstream to both UNC-16 and LRK-1 in regulating composition of some SVP transport carriers . Our data suggests that LRK-1 may be a general means to recruit different AP complexes to the appropriate membrane surface . We further hypothesize that UNC-101 at the Golgi is required for formation of an intermediate compartment whose size itself may be regulated by the AP-1 complex . At such an intermediate compartment the AP-3 complex may function to regulate the composition but not the size of vesicles arising from this precursor . Alternatively , UNC-101 may also have an additional role at the intermediate compartment to regulate size of vesicles arising from this compartment . LRK-1 , itself may exert its retromer function at such an intermediate compartment for retrieval of Golgi enzymes . Based on all our observations , the SVP transport carriers formed in unc-16 seem to be visibly aberrant in nature ( Fig 8b ) . These observed defects likely arising due to an altered surface composition might also lead to the recruitment of multiple motors as suggested by Byrd et al . , 2001 [29] . Consistent with this , we observed that the aberrant SVP transport carriers formed in unc-16 were no longer dependent exclusively on UNC-104 ( Fig 7b ) . Further , it also appears to depend at least partially on other motors such as UNC-116/Kinesin-1 ( [29]; S4f Fig ) . This suggests that the motor-cargo specificity is lost in unc-16 perhaps as a consequence of the altered surface composition that might now contain adaptors for multiple other motors . The transport carriers formed in lrk-1 and apb-3 , have an altered composition , as suggested by defects in the ability to include different SVPs , and are only partially dependent on UNC-104 ( Fig 7bvi and 7bx ) . This could again imply that due to an altered composition the carriers in lrk-1 and apb-3 mutants are not able to stably recruit sufficient numbers of the SV motor . Over expression of LRK-1 in unc-16 completely restores co-transport of RAB-3 and SNB-1 , but is still insufficient to make the SVP transport carriers completely dependent on the motor ( Fig 7bviii ) suggesting the existence of additional factors necessary for forming an UNC-104 motor dependent transport carrier . Previously , UNC-16 has been suggested to have a “clearance function” wherein it regulates the retrieval of cell soma organelles such as lysosomes and endosomes from the axon , thereby acting as an “organelle gatekeeper” [27 , 28] . Considering the multiplicity of mis-sorting defects we see in unc-16 , we postulate that the unusual accumulation of different organelle proteins in the axon of these mutants could also be contributed by early defects in protein sorting at the Golgi rather than from a retrieval defect alone . Our study supports previous ideas that UNC-16 is involved in multiple trafficking pathways . UNC-16 may achieve these functions via different downstream effector molecules that it can potentially scaffold , regulating different subsets of trafficking pathways . Our study shows that UNC-16 acts via downstream molecules LRK-1 , AP-1 and AP-3 to control biogenesis of the SVP transport carriers in the cell body ( Fig 8a ) . Previous biochemical studies in mammalian cells have indicated that LRRK2 was co-purified with both JIP3 and components of clathrin [31 , 62] , suggesting that similar relationships may be conserved in mammals . We uncover a likely hierarchical series of processes regulated by these proteins , which occurs early on at the Golgi and post-Golgi compartments , for the sorting and regulation of SVP trafficking ( Fig 8a ) . Both , UNC-16 and LRK-1 proteins have been implicated to have roles at the synapse . UNC-16 has been shown to regulate RAB-5-mediated membrane trafficking and contribute to SV maturation at the synapses [34] . Lee et . al . , 2010 and Piccoli et . al . , 2011 have proposed presynaptic roles for LRRK2 where it regulates synaptic morphology and SV recycling dynamics respectively [62 , 63] . Thus , both UNC-16 and LRK-1 could have trafficking roles at the synapse in addition to or as a consequence of altered membrane composition arising at the Golgi .
C . elegans strains were grown and maintained at 20°C on NGM plates seeded with the E . coli OP50 strain using standard methods [64] . L4 or 1-day adult animals were used for imaging in all cases . Strains used are listed in S4 Table . Some of the strains were provided by the CGC ( https://cgc . umn . edu/acknowledging-the-cgc ) . The unc-16 ( tb109 ) allele was isolated from a behavioral suppressor screen carried out in unc-104 ( e1265 ) worms . This was a non-clonal screen of approximately 60 , 000 haploid genomes . Animals were mutagenized using 50 mM EMS for 4 h . F1 and F2 progenies were screened for improved locomotion . Suppressors were further identified by improved localization of GFP::RAB-3 in PLM neuron ( Kumar et al . , 2010 ) . The tb109 allele was separated from the background mutation and mapped to chromosome III and tested for non-complementation with unc-16 ( e109 ) . The tb109 phenotypes could be rescued by the expression of Punc-16::UNC-16::GFP . Sequencing revealed that allele tb109 contains a point mutation in exon 9 leading to a stop codon at amino acid 423 ( Arginine to opal as stop codon ) . Young adults of unc-16 ( tb109 ) and N2 were fixed for electron microscopy by high-pressure freezing ( HPF ) technique [65] . Serial sections were cut and the dorsal and ventral nerve cord regions were imaged using Gatan side mount camera on Tecnai G2 12 BioTwin electron microscope ( FEI Company ) . The cross-sectional width of all the vesicles present in each section was measured using ImageJ [66] . Immunostaining was performed as described previously [16] . For double labeling of the SV proteins , sample was first incubated with mouse anti-RAB-3 ( 1:2000 ) , followed by incubation with rabbit anti-SNT-1 ( 1:500 ) antibody . Appropriate secondary antibodies ( 1:350 ) ( Alexa 488 , Alexa568; Molecular probes ) were added and incubated for two days . Images were captured using a Zeiss Axiovert inverted microscope . Images were processed using ImageJ [66] . Samples were prepared by mechanical homogenization using homogenization buffer ( 15mM HEPES-NaOH pH7 . 4 , 10mM KCl , 1 . 5mM MgCl2 , 0 . 1mM EDTA , 0 . 5mM EGTA , 0 . 05mM sucrose and protease inhibitors ( Roche ) ) followed by mild sonication at 4°C . Sample was then incubated with mouse monoclonal anti–Flag ( 1:50 ) ( Biovision ) or mouse monoclonal anti-GFP sera ( 1:10 ) ( Genei , Merck ) at 4°C for 4 h . The antigen-antibody complex was incubated with Protein-A agarose beads ( Genei , Merck ) for another 4 h . Antibodies used for Western blots to probe for UNC-16::GFP and LRK-1::Flag were rabbit anti-GFP ( 1:500–1000 , Santacruz , Abcam ) and mouse anti-flag ( 1:1000–2000 , Sigma ) respectively . In all graphs , data are presented as mean values ± SEM . Statistical analysis was performed using GraphPad Prism 6 ( GraphPad Software ) . Wherever possible , an independent t-test was used , or for multiple comparisons , one-way ANOVA followed by Dunett’s multiple comparison tests was done . For grouped datasets , two-way ANOVA , followed by Tukey’s multiple comparison tests was used . Differences were considered significant when P < 0 . 05 ( * , P < 0 . 05; ** , P < 0 . 01; *** , P < 0 . 001 ) . | Synaptic vesicles ( SVs ) have a defined composition and size at the synapse . The multiple synaptic vesicle proteins ( SVPs ) found on these vesicle membranes are synthesized at and trafficked out of the cell body in distinct transport carriers . However , we do not yet understand how different SVPs are sorted and trafficked to the synapse . We show that UNC-16/JIP3 plays a critical role , in a series of essential steps , to ensure proper membrane composition and size of the ensuing SVP carrier exiting the cell body . These processes are “exclusion” of resident Golgi enzymes followed by the “inclusion” of synaptic vesicle proteins in the same transport carrier . Regulation of composition and size seems to occur independently of each other and depends on two distinct AP complexes acting downstream to LRK-1 . Our study further indicates that the composition of the transport carrier formed is important for the recruitment of motors and consequently for the polarized localization of SVPs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"protein",
"transport",
"medicine",
"and",
"health",
"sciences",
"vesicles",
"nervous",
"system",
"cell",
"processes",
"enzymology",
"electrophysiology",
"neuroscience",
"synaptic",
"vesicles",
"golgi",
"apparatus",
"cellular",
"structures",
"and",
"organelles",
"neuronal... | 2017 | UNC-16/JIP3 regulates early events in synaptic vesicle protein trafficking via LRK-1/LRRK2 and AP complexes |
In June 2015 , a cholera outbreak was declared in Juba , South Sudan . In addition to standard outbreak control measures , oral cholera vaccine ( OCV ) was proposed . As sufficient doses to cover the at-risk population were unavailable , a campaign using half the standard dosing regimen ( one-dose ) targeted high-risk neighborhoods and groups including neighbors of suspected cases . Here we report the operational details of this first public health use of a single-dose regimen of OCV and illustrate the feasibility of conducting highly targeted vaccination campaigns in an urban area . Neighborhoods of the city were prioritized for vaccination based on cumulative attack rates , active transmission and local knowledge of known cholera risk factors . OCV was offered to all persons older than 12 months at 20 fixed sites and to select groups , including neighbors of cholera cases after the main campaign ( ‘case-triggered’ interventions ) , through mobile teams . Vaccination coverage was estimated by multi-stage surveys using spatial sampling techniques . 162 , 377 individuals received a single-dose of OCV in the targeted neighborhoods . In these neighborhoods vaccine coverage was 68 . 8% ( 95% Confidence Interval ( CI ) , 64 . 0–73 . 7 ) and was highest among children ages 5–14 years ( 90 . 0% , 95% CI 85 . 7–94 . 3 ) , with adult men being less likely to be vaccinated than adult women ( Relative Risk 0 . 81 , 95% CI: 0 . 68–0 . 96 ) . In the case-triggered interventions , each lasting 1–2 days , coverage varied ( range: 30–87% ) with an average of 51 . 0% ( 95% CI 41 . 7–60 . 3 ) . Vaccine supply constraints and the complex realities where cholera outbreaks occur may warrant the use of flexible alternative vaccination strategies , including highly-targeted vaccination campaigns and single-dose regimens . We showed that such campaigns are feasible . Additional work is needed to understand how and when to use different strategies to best protect populations against epidemic cholera .
Oral cholera vaccine ( OCV ) is an effective tool to prevent and control cholera both in endemic settings and in response to outbreaks [1 , 2] . On 23-June-2015 , the Republic of South Sudan Ministry of Health ( MoH ) declared a cholera outbreak in Juba , the nation’s capital . Initial cases were traced back to 18-May in the United Nations Protection of Civilians Camp , where approximately 28 , 000 internally displaced people ( IDP ) resided . By the time the epidemic was declared , cases had been confirmed throughout the city and public health officials believed that Juba was at risk for a large cholera outbreak , with the threat of spread to other areas of the country . The MoH convened the National Cholera Taskforce to guide a comprehensive outbreak response involving case management , water and sanitation interventions , health education and hygiene promotion . Following a situation assessment and in light of the 2014 cholera outbreak with 6 , 269 reported cases and 156 deaths in multiple areas of the country [3] , the MoH , supported by Médecins sans Frontières ( MSF ) , decided to integrate OCV into the cholera response in Juba . Only 270 , 000 doses were released from the global emergency OCV stockpile due to severe supply limitations despite the much larger at-risk population ( 500 , 000–1 , 000 , 000 people ) . The MoH decided to use an off-label , single-dose , regimen in a targeted vaccination campaign . The rationale was based on preliminary results from a large randomized clinical trial [4] demonstrating significant 1-dose protection in Bangladesh , immunogenicity studies [5] and modelling analyses [6] showing that even with a significantly less effective one-dose regimen , one-dose campaigns may save more lives than their two-dose counterparts when supply is limited . The goal was to quickly provide protection to the maximum number of people at highest risk with a , perhaps less effective , single-dose regimen , rather than covering half the number of people with the standard two-doses . The possibility of providing a second dose , to potentially increase the effectiveness and extend the duration protection , when supplies were available was not ruled out . Two targeted vaccination approaches were used with an aim to halt transmission and shorten the duration of the epidemic . First , OCV was targeted to neighborhoods with evidence of significant transmission just prior to the campaign and vulnerable groups at higher risk of cholera including IDPs , prisoners and health care workers . After the main campaign , sporadic case reports continued , mostly in unvaccinated neighborhoods . Given that the risk of cholera has been shown to be highly elevated among those living around a cholera case in the days after the case presents for care at a clinic [7 , 8] , the remaining vaccine was delivered to neighbors living around suspected cholera cases together with water sanitation and hygiene measures ( case-triggered interventions ) . Further details related to the decision-making process , timeline and vaccine effectiveness are described in detail elsewhere [9 , 10] Here , we describe the operational details and vaccine coverage of these spatially targeted OCV delivery approaches , from campaigns that represent , to our knowledge , the first field-use of a single-dose of OCV in response to an epidemic , and the first use of case-triggered cholera interventions including OCV . We explore vaccine uptake in the different areas targeted , including neighborhoods and smaller areas around the households of suspected cases , identify difficult-to-reach population groups and discuss alternative campaign strategies to improve vaccine coverage .
The campaign setting was particularly challenging amid a humanitarian crisis with significant population displacement . Accurate population estimates were not available , so we used two approaches to define the target populations when planning the campaigns . First , we extrapolated population estimates for different areas of the city using data from the latest population census in 2009 [11] . This census was conducted prior to independence and civil war and extrapolated population size estimates are believed to vastly underestimate the true population size . Next , we used estimates of the number of built structures in each area from recent digitized satellite images ( http://wiki . openstreetmap . org/wiki/WikiProject_South_Sudan ) . We initially assumed 70% of the structures with a roof footprint from 5 to 250 m2 were residential and that each had an average of 6 people [11] . In this rapidly evolving epidemic , the decisions of where to target vaccine were based on the most up to date cumulative attack rates , recent incidence and local knowledge of known cholera risk factors . Three main areas were identified , using unofficial boundaries and referred to here as Kator , Northern Juba and Gumbo , with combined population estimates ranging from 53 , 543 , from census data , to 368 , 136 from satellite imagery ( Fig 1 ) . The three areas differed greatly by socio-demographics . Kator is a densely populated area including a semi-commercial part of the city that had experienced a spike of suspected cholera cases just prior to the campaign and a large slum-like area bordering the Nile river . Northern Juba is a fairly-isolated settlement next to a large military base , predominantly inhabited by military members and their families . Gumbo is an area with moderate population density on the south-eastern side of the river Nile with predominantly poor-housing and persistent notification of cholera cases preceding the campaign . In addition to the three targeted areas of the city , IDPs living in informal camps , inmates in Juba’s prison , health care workers and residents living close to suspected cholera patients presenting after the main campaign were targeted by mobile teams . A single dose of OCV ( Shanchol® , Shantha Biotechnics Ltd , Hyderabad , India ) was offered to all persons older than 12 months presenting at vaccination sites , regardless of her/his area of residence . Twenty fixed vaccination sites operated from 08:00–17:00 from 31-July to 5-August each with a team of approximately 20 people per site ( 3–4 vaccinators , 3–4 individuals preparing the vaccine , 8–10 registrars filling out vaccine cards and tally sheets , 1 security guard , 2 health promoters and 1 team supervisor ) . As the number of individuals coming to the sites slowed ( 3-August ) , vaccination teams split into semi-mobile units and set up mini-vaccination sites to reach those not yet covered . Vaccines were stored under cold chain ( 2–8°C ) using a refrigerated truck , transported to the vaccine site in their original Styrofoam box with icepacks and then used at ambient temperature the day of vaccination . We distributed a vaccination card to each vaccinee indicating her/his name , age , vaccination location , date of vaccination and vaccine lot number . We avoided the widespread use of radio and other media to publicize the campaign due to concerns that offering limited vaccine only in selected parts of the city could spark civil unrest . Trained health promoters disseminated information regarding the campaign in the targeted communities , and members of the target populations were recruited to spread key messages using megaphones . To assess vaccine coverage in the neighborhood-targeted ( main ) campaign , a random sample of the population living in each of the target areas was selected using a stratified spatial sampling approach , with households serving as the primary sampling unit . A total of 128 households , including all household members , were required in each of the three target areas to estimate each area-specific coverage with a precision of ±5% . Detailed methods for the coverage survey , conducted 9–14 August , are provided in S1 Text . From 13–26 August , after the main campaign but before cholera case reports had stopped within the city , OCV was included as part of a case-triggered comprehensive targeted intervention ( CTI ) approach . Vaccine was offered together with soap , water purification tablets , a leaflet on cholera prevention and health promotion by mobile teams made up of staff from multiple organizations , including the MoH , South Sudanese Red Cross , Oxfam and MSF . This activity required close coordination with multiple governmental and non-governmental actors to ( 1 ) detect and test suspected cases , ( 2 ) rapidly communicate results from cholera diagnostic tests , ( 3 ) locate the case’s household and decide on the location for intervention in conjunction with local leaders and ( 4 ) deploy the intervention in a timely manner . Patients from Juba reporting to any of the cholera treatment centers or oral rehydration posts with a stool sample positive for cholera using the Crystal VC rapid diagnostic test ( RDT ) , either directly or after a 4–6 hour enrichment in alkaline peptone water [12] , were put on a list for CTI eligibility . Due to limited human resources , cases coming from areas that had not been covered in the neighborhood-targeted OCV campaign and those testing positive to the , more specific [12] , enriched RDT were prioritized . When possible , teams also conducted CTI in previously vaccinated areas where it appeared that cholera transmission may have continued . These CTIs were also targeted to the homes of individuals who died of acute watery diarrhea , either in the community or in a health-facility , even if no stool sample had been tested . The Juba County Health Department’s rapid response team and MSF staff travelled to the home of CTI-eligible suspected cases . Together with a community leader , they identified a suitable intervention site as close as possible to the home of the patient and recruited four community members: two to assist with security/crowd control and two for going door-to-door informing the neighbors about the intervention and encouraging residents to come to the sites . Volunteers were not informed of the specific rationale behind the location of intervention site ( i . e . details of the suspect cholera case triggering the CTI ) to respect the patients’ privacy , but they were given a specific geographic focal area . The site was typically set-up the following day by a 5-7-person team ( 1 site supervisor , 1–2 vaccinators , 1–2 registrars for completing vaccination cards and tally sheets and 2–3 people delivering the water/sanitation/hygiene intervention ) . All individuals coming to the site were eligible for OCV regardless of whether they had been vaccinated in the main vaccination campaign . To assess vaccine coverage in each case-centered targeted intervention cluster , we selected 30 spatially random points within 350-meters ( assumed as the catchment area of the interventions ) of suspected case households ( Fig 1 , S1 Text ) . As with the population-based survey for the main campaign , the closest household to each GPS point was included in the survey , but instead of ascertaining the vaccination status of all individuals living in the household as done in the main coverage survey , one person was selected at random from those residing in ( but not necessarily present at the time of the first survey visit ) the household . We collected data on age , sex , vaccination status ( both verbal and confirmed with card ) and reasons for non-vaccination ( when applicable ) for everyone included in the coverage surveys . We also collected household-level variables including the number of household members at the time of the campaign , the number of built structures included in each household and their spatial coordinates . We estimated mean vaccination coverage and 95% confidence intervals for individual vaccination target areas ( both neighborhoods and areas around case-triggered interventions ) and for the entire target population . In secondary analyses , we estimated the coverage by age group , sex and distance to the closest vaccination site . Individuals with missing information on vaccination status were excluded from the analysis . Relative risks and 95% confidence intervals were estimated using a generalized linear model with a log link . All confidence interval estimates for vaccine coverage and relative risks took into account the survey design ( clustering by household and vaccination area ) using the svy commands in Stata 12 . 0 ( College Station , TX , USA ) . This was a public health intervention designed to prevent the spread of cholera , informed consent for participation was not required . The activities presented in this study were conducted as standard monitoring and evaluation exercises , thus approval from ethical review committees was not obtained . Although written informed consent was not solicited for the coverage surveys , all interviewees provided verbal consent and no identifiable information was collected other than household coordinates .
Most of the vaccines ( 91 , 953 doses , 65 . 6% ) were distributed in the targeted area of Kator . In the neighborhoods targeted in Northern Juba , 21 , 039 individuals received OCV , which was greater than the population estimated by both census and using satellite imagery ( Table 1 ) . Just over half of those receiving vaccine during the neighborhood-targeted campaign were male ( 71 , 945 , 51 . 3% ) and 75 , 638 ( 53 . 9% ) were at least 15 years old ( Table 2 , S1 Table ) . A total of 371 households were included in the coverage survey of targeted neighborhoods ( Fig 2 ) . No households refused to participate in the survey . Household size varied from 1 to 20 with a median of 6 ( S2 Table ) . The mean number of built structures per household was 2 ( S2 Table ) . All but two households with available coordinates were within 1 kilometer from the closest vaccination site ( Fig 3 ) , with a median distance to the closest vaccination site of 156 meters . We ascertained the vaccination status for 96 . 8% ( 2578/2662 ) of individuals , with 94% of those who reported to have been vaccinated providing a vaccination card . We estimated the vaccine coverage ( self-report ) to range from 60–70% ( S1 Table ) with an overall population-weighted coverage across the three targeted areas of 68 . 8% ( 95% CI: 64 . 0–73 . 7 , Table 2 ) . Vaccination status between individuals in the same household was more correlated than expected , with a survey design effect of 7 . 3 . In Northern Juba , nearly 1 in 3 households ( 30% ) reported that no household members were vaccinated ( S2 Table ) . The proportion of household members vaccinated decreased with distance to the closest vaccination site ( Fig 3 ) . Coverage was highest among children 5–14 years ( 90 . 0%; 95% CI: 85 . 7–94 . 3 ) . While overall vaccine coverage was similar between women ( 68 . 9%; 95% CI: 63 . 7–74 . 0 ) and men ( 64 . 7%; 95% CI: 57 . 7–71 . 7 ) on average; adult women tended to have higher coverage than adult men ( RR 0 . 81 , 95% CI: 0 . 68–0 . 96 ) , with less than half the men 15 years or older reporting to have been vaccinated ( Fig 4 , S3 Table ) . The main reasons for non-vaccination in the main campaign were; ( 1 ) not being aware of the campaign ( 256 , 30% of unvaccinated individuals ) , ( 2 ) being absent during the time of the campaign ( 202 , 23% ) , and ( 3 ) not having time ( 129 , 15% , S4 Table ) . Of the 54 suspected cholera cases from Juba screened by direct and enriched RDT during the CTI period , 17 were positive by enriched RDT . We carried out CTI at the homes of 14 ( 82% ) of these enriched RDT positive cases . One additional direct RDT positive household ( out of 9 direct-RDT positive only ) was included in CTI as it occurred on a day with no other priority activities for the teams . The remaining two CTIs occurred around the residence of individuals who had reportedly died due to acute watery diarrhea for whom no sample was available ( 1 community- and 1 facility-death ) . Two additional deaths were reported during this period although the team was unaware of these deaths at the time of the activities . All but two of the CTIs took place in areas that had not been covered in the main campaign . The CTIs occurred 1–6 days after the suspected cholera case had presented at the health facility ( mean delay 3 . 4 days ) . Ten CTIs were single-day events and the remaining 7 took place over a two-day period . Based on tally sheets collected at the CTI sites , 11 , 491 ( 54 . 4% ) of those who received the intervention were female , 4 , 091 ( 19 . 4% ) were children 1–4 years old and 7 , 886 ( 37 . 3% ) were children 4–14 years old . Coverage surveys were carried out in 13 of 17 CTIs ( exact location of the patient’s home was unavailable for 3 and one was outside of Juba town ) , with a total of 390 individuals sampled . Vaccine coverage per CTI site ranged from 30% ( 95%CI 12 . 6–47 . 4 ) to 86 . 7% ( 95%CI 73 . 7–99 . 6 ) . Overall , the coverage was 51 . 0% ( 95%CI 41 . 7–60 . 3 ) , with no significant difference between those sites targeted inside and outside the main campaign target area . Coverage patterns were like those observed in the main neighborhood-targeted campaign . Adult men were less likely to have received the vaccine compared to adult women ( RR 0 . 73; 95%CI 0 . 60–0 . 89 ) . Overall coverage was 45 . 7% ( 95%CI 35 . 5–55 . 8 ) among men and 55 . 8% ( 95%CI 45 . 5–66 . 1 ) among women . Coverage was also highest among school-aged children ( Table 2 ) .
We provided a single-dose of OCV through spatially-targeted campaigns to over 160 , 000 individuals in Juba , South Sudan . We achieved nearly 70% vaccine coverage within the main , neighborhood-targeted , campaign and had no significant challenges in using a targeted strategy within this large urban setting . Our experience should ease concerns about targeting specific populations with OCV in urban settings , even during an outbreak . Similarly , targeted OCV campaigns have been successfully implemented in urban slums of Haiti [13] . These findings support the possibility of targeting particular neighborhoods that may be responsible for driving urban cholera epidemics , which may provide an efficient way to minimize cost and maximize public health impact [14] . We also demonstrated that it is feasible for multiple actors ( e . g . , MoH and humanitarian organizations ) to work together to rapidly provide a suite of cholera control interventions to the high-risk group living near cholera cases with moderate coverage . While this type of approach is intuitively appropriate for cholera control given the evidence of elevated risk around cases [7 , 8] , evaluations of the effectiveness of similar interventions in the future are needed . Here , the case-triggered CTI approach was used at the end of the outbreak , when cases were sporadic , with the hopes of quelling the outbreak . Consequently , numbers were low and given the approach was hastily devised during the outbreak , there was limited time for detailed planning to optimize impact and to incorporate any detailed evaluation of effectiveness . More work is needed to best define the best mix of components to include in CTI , including the possibility of prophylactic antibiotics , to halt cholera transmission . This approach is not likely to be a silver bullet for cholera control , but may prove to be an efficient strategy in periods of low transmission , perhaps seasonally as has been proposed in Haiti [15] , or to accelerate the end of an outbreak after mass campaigns . Although some OCV campaigns have achieved higher coverage , our estimates are consistent with others in urban areas [16–19] . Despite initial concerns , vaccine sites were not over-run with population from elsewhere in the city , and even in the targeted areas , coverage was less than expected . One potential reason that public interest in vaccine was lower than expected may be due to ‘cholera fatigue , ’ where after the much larger 2014 outbreak [3] , individuals and the media paid much less attention to cholera in 2015 . While vaccine coverage was lower in adult men , use of the administrative data alone masked this difference and suggested roughly equal OCV coverage by sex . This was especially apparent in Kator where a substantial proportion of those that received the vaccine during the campaign were adult men ( 53 . 6% per tally sheets ) , but little over 40% of the men who lived in the area received the vaccine ( S1 and S2 Tables ) . Being a commercial part of town , it is possible that these were male businessman working in the area during the day but who lived elsewhere in the city . A door-to-door strategy may have been more appropriate for this highly targeted campaign , however other campaigns using a mixture of fixed sites and door-to-door vaccine delivery report a similar coverage among urban populations and similar challenges reaching adult men [19] . Keeping vaccination sites open later ( security situation dependent ) , and perhaps moving them near places where people congregate in the early evening could prove a useful strategy to improve coverage among men , given that lack of time was the most common reason for non-vaccination . On the other hand , if the reason for not having time was due to work commitments , campaigns could target workplaces during the day . The high coverage among school-aged children likely reflects the success of using schools as vaccination sites , as has been observed in other settings [20] . We observed heavy clustering of vaccination status within households of the main coverage survey . Over 15% of the households sampled had no vaccinated individuals , despite most households visited being well within walking distance from a vaccination site ( e . g . , more than half being within 160 meters ) . On the other hand , a third of the households had 100% coverage among eligible members . The most common reason for non-vaccination in the neighborhood campaign was not being aware of the campaign , perhaps reflecting the limited use of radio and other measures to publicize the campaign . Ensuring at least one person in every house is aware of the OCV campaign could be an important approach to increasing overall coverage . A door-to-door strategy for social mobilization rather than for vaccine delivery could help increase household-level knowledge of the campaign . This experience in Juba highlights the key challenge of designing public health interventions in low-resource , volatile settings such as South Sudan , where accurate and up-to-date demographic information is not always available . The use of satellite imagery has been used to estimate population size in unstable settings [21] and innovative initiatives like MissingMaps ( www . missingmaps . org ) make the task more feasible , even in the world’s most vulnerable populations . Nevertheless , local information regarding the different observable characteristics of residential and non-residential built structures and the number of persons per built structure are needed to obtain accurate estimates . Developing standardized methods for gathering and sharing information to aid population estimation in low-resource , data-poor settings is a key priority for efficient public health programming . Our findings come with several limitations . We based our spatial sampling on building density from recent satellite images rather than true population density . It is possible that some areas of the city , especially the most vulnerable , overcrowded areas may have more people but less built structures and could therefore be underrepresented . Furthermore , we interviewed the senior household member for information regarding vaccination status of the other household members in the main coverage survey . This may have led to information bias and contribute to our findings of high intra-household clustering . This also may have led to less precise and accurate estimates of the reasons for non-vaccination within the household . In conclusion , we showed that targeting OCV in response to an outbreak within a large urban population both to neighborhoods and neighbors of cholera cases is feasible and well accepted by the population . Developing and testing new ways to reach traditionally hard-to-reach groups , including adult men , remains a priority . While cholera continues to strike in complex settings with mobile populations and dynamic security constraints , flexible targeted approaches and alternative dosing schedules , like the one described here , are needed to maximize the potential impact of the vaccine . | Oral cholera vaccine ( OCV ) is becoming part of the standard cholera-control toolkit , although experience in deploying OCV is limited . Adapting vaccination strategies to the global availability of vaccines and the local context ( i . e . , population movement , security constraints , etc . ) is key to maximize the impact of OCV as a cholera-control tool . Here we describe the operational details of the first field use of a single-dose of OCV , which was deployed in a targeted manner , both at high-risk neighborhoods and then to neighbors of suspected cases after the main OCV campaign when sporadic cholera case reports continued . We show that it is feasible to conduct micro- and macro-targeted vaccination campaigns in urban areas like Juba with moderate to high coverage and without social unrest due to vaccinating some groups and not others . Flexible and context-adapted OCV dosing regimens and strategies should be considered in future deployments of the vaccine . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"built",
"structures",
"engineering",
"and",
"technology",
"immunology",
"tropical",
"diseases",
"census",
"vaccines",
"preventive",
"medicine",
"bacterial",
"diseases",
"age",
"groups",
"research",
"design",
"adults",
"neglected... | 2017 | Neighborhood-targeted and case-triggered use of a single dose of oral cholera vaccine in an urban setting: Feasibility and vaccine coverage |
The diversity of a highly variable RNA plant virus was considered to determine the range of virulence substitutions , the evolutionary pathways to virulence , and whether intraspecific diversity modulates virulence pathways and propensity . In all , 114 isolates representative of the genetic and geographic diversity of Rice yellow mottle virus ( RYMV ) in Africa were inoculated to several cultivars with eIF ( iso ) 4G-mediated Rymv1-2 resistance . Altogether , 41 virulent variants generated from ten wild isolates were analyzed . Nonconservative amino acid replacements at five positions located within a stretch of 15 codons in the central region of the 79-aa-long protein VPg were associated with virulence . Virulence substitutions were fixed predominantly at codon 48 in most strains , whatever the host genetic background or the experimental conditions . There were one major and two isolate-specific mutational pathways conferring virulence at codon 48 . In the prevalent mutational pathway I , arginine ( AGA ) was successively displaced by glycine ( GGA ) and glutamic acid ( GAA ) . Substitutions in the other virulence codons were displaced when E48 was fixed . In the isolate-specific mutational pathway II , isoleucine ( ATA ) emerged and often later coexisted with valine ( GTA ) . In mutational pathway III , arginine , with the specific S2/S3 strain codon usage AGG , was displaced by tryptophane ( TGG ) . Mutational pathway I never arose in the widely spread West African S2/S3 strain because G48 was not infectious in the S2/S3 genetic context . Strain S2/S3 least frequently overcame resistance , whereas two geographically localized variants of the strain S4 had a high propensity to virulence . Codons 49 and 26 of the VPg , under diversifying selection , are candidate positions in modulating the genetic barriers to virulence . The theme and variations in the evolutionary pathways to virulence of RYMV illustrates the extent of parallel evolution within a highly variable RNA plant virus species .
Parallel evolution is the evolution of similar or identical features independently in related lineages when subjected to similar selection pressures [1 , 2] . Parallel evolution has been reported extensively both in natural isolates and in experimental populations of many microbes , most often viruses [2–4] , but also bacteria [5] , yeast [6] , and protozoa [7] . With viruses , similar amino acid replacements often occurred in immune or antiviral escape variants . This is usually interpreted as the fixation of a mutation with a beneficial effect . However , differences in founding genotypes may result in divergent evolutionary trajectories [5] . So , patterns of adaptation to selective constraints may also be dependent on intraspecific polymorphisms . This is well documented for HIV resistance to antiretroviral agents , where pathways of viral evolution towards drug resistance may proceed through distinct steps and at different rates among different HIV subtypes [8 , 9] . The objective of this article is to assess the extent of parallel evolution in a highly variable RNA plant virus species . Then , the diversity of Rice yellow mottle virus ( RYMV ) was considered to determine the range of virulence substitutions , to reconstruct the evolutionary pathways to virulence , and to test whether intraspecific diversity modulates virulence pathways and propensity . The breakdown of host plant resistance is a well-documented case of virus adaptation [10] . Translation initiation factors of the eIF4E and eIF4G families are major determinants of recessive resistance to plant viruses [11 , 12] . Several studies showed that one or a few amino acid substitutions in the genome-linked viral protein VPg are independently responsible for virulence ( see [12] for a recent review ) . In none of these studies however , was the diversity of the virus species considered . Furthermore , no longitudinal studies ( samples collected from the same host plant at sequential times ) have been conducted to unravel the accumulation and interplay of mutations associated with virulence over time . RYMV is an appropriate model to address these issues: RYMV is a highly variable virus species ( up to 11% in nucleotide [nt] difference in the full genome ) [13] , an eIF ( iso ) 4G-mediated resistance was identified in rice [14] , and the viral genome-linked protein ( VPg ) of RYMV was determined as the virulence factor [15] . RYMV of the genus Sobemovirus [16] occurs in all rice-growing African countries , where it causes heavy yield losses [17] . Its genome harbors four open reading frames ( ORFs ) [18] . ORF1 , which is located at the 5′ end of the genome , encodes the protein P1 involved in virus movement and gene silencing . ORF2 , which encodes the central polyprotein , has two overlapping ORFs . ORF2a encodes a serine protease and the VPg . ORF2b , which is translated through a −1 ribosomal frameshift mechanism as a fusion protein , encodes the RNA-dependent RNA polymerase . The coat protein is expressed by a subgenomic RNA at the 3′ end of the genome . RYMV is transmitted by contact during cultural practices and by animal vectors , mainly beetle species ( Coleoptera ) of the Chrysomelidae family . Very few rice varieties are resistant to RYMV . The highest level of resistance , characterized by an undetectable virus titer in ELISA tests and the absence of symptoms , is provided by Oryza sativa indica cv Gigante and a few Oryza glaberrima cultivars of the Tog series [19 , 20] . Recently , another highly resistant indica cultivar , named Bekarosaka , was found in Madagascar [21] . Resistance is controlled by the recessive gene Rymv1 [19] , which maps on chromosome 4 [22] and encodes the translation initiation factor eIF ( iso ) 4G [14] . Three alleles of Rymv1 were identified , one in the O . sativa cvs Gigante and Bekarosaka ( allele Rymv1-2 ) , and two in the O . glaberrima accessions Tog5681 ( Rymv1-3 ) and Tog5672 ( Rymv1-4 ) . Compared to susceptible varieties ( allele Rymv1-1 ) , the Rymv1-2 resistance allele is characterized by one amino acid substitution in the central domain of the eIF ( iso ) 4G gene [14] . Rymv1-2 was introgressed into widely grown indica cultivars that are now propagated in the fields . In this paper , we defined virulence as the genetic ability of a pathogen to overcome genetically determined resistance and to cause a compatible interaction leading to disease [23 , 24] . Avirulence is the antonym of virulence . With RYMV , virulence is characterized by pronounced and generalized symptoms , full systemicity , high virus titer , and high mechanical transmissibility to resistant plants . The high resistance of cv Gigante was effective against a range of isolates of different RYMV strains [19] . Recently , however , the resistance of Gigante was overcome by isolates of several geographic origins [25–27] . Sequencing one such isolate revealed that a single substitution in the VPg specifically differentiated the wild avirulent type from the evolved virulent variant . Directed mutagenesis indicated that this substitution conferred the virulence [15] . This work , carried out with a single isolate , is extended here to a range of isolates of the main strains . Isolates representative of the genetic and geographic diversity of RYMV in Africa were inoculated to Rymv1-2-resistant accessions made of cultivars Gigante and Bekarosaka and of four nearly isogenic lines ( NILs ) . The accumulation and the interplay over time of mutations associated with virulence were followed . Altogether , we found one theme , yet several isolate- or strain-specific variations , in the evolutionary pathways to virulence of RYMV .
A total of 22 isolates representative of the geographic and genetic diversity of RYMV was fully sequenced . The dN/dS value of the VPg was lower than that of the P1 and of the coat protein ( 0 . 05 versus 0 . 15 and 0 . 14 , respectively ) and similar to that of ORF2b coding the polymerase ( 0 . 04 ) . We sequenced the VPg of an additional set of 37 wild avirulent isolates collected in the fields . Analysis of the 59 wild isolates revealed that 17 of the 79 amino acids of the VPg ( 22% ) varied naturally in wild-type isolates ( both within and between strains ) . A total of 29 codons was under conservative selection pressure , most of them in the N-terminal half of the VPg ( Figure 1 ) . Only three codons were under positive selection with a Bayes factor for positive selection higher than 100 . Codon 49 was under high diversifying selection as apparent from both the REL and IFEL methods , with dN/dS = 20 . Codon 26 , and codon 62 adjacent to the sobemovirus WAD conserved motif at positions 64–68 , were under moderate diversifying selection with dN/dS = 2 . 8 and 1 . 8 , respectively , and only significant with the less conservative method REL . In all , 114 isolates representative of the main strains of RYMV were inoculated to Rymv1-2-resistant accessions . Ten of them ( c . 9% ) became virulent ( Table 1 ) . By large-scale inoculation , we generated several virulent variants from isolates CI4 , Mg16 , and Tz209 , and we analyzed , respectively , 12 , 19 , and three of them . From each of the seven other isolates , only one virulent variant was studied . Altogether , 41 virulent variants were analyzed . Amino acid changes associated with virulence were deduced from comparison between the sequences of the VPg of the wild avirulent and of the evolved virulent forms of each isolate . With three isolates ( Ma145 , Tz225 , and Tz230 ) , however , only the VPg of the evolved virulent form was sequenced , and putative changes associated with virulence were assessed by comparison with the range of sequences of the wild avirulent isolates . Virulence was always associated with non-synonymous changes . They were located within a stretch of 15 codons in the central domain of the 79-aa-long VPg at codons 38 , 42 , 43 , 48 , and 52 ( Figure 2 ) . The amino acids associated with virulence were never found in the 59 wild avirulent isolates . Codon 43 was under neutral evolution , and codons 42 and 48 were under conservative selection , whereas codons 38 and 52 were strictly conserved in the avirulent wild isolates ( Figure 1 ) . Substitutions associated with virulence always involved a change of biochemical class ( Figure 3 ) . Most substitutions were transitions ( Table 2 ) . The transversion/transition ratio of 0 . 15 calculated from the changes associated with virulence was similar to that of 0 . 13 ( +/- 0 . 03 ) estimated by maximum likelihood from the corpus of 59 wild isolates . Substitutions associated with virulence were fixed mostly at codon 48 ( 33/41 ) , and to a lesser extent at codons 52 ( 6/41 ) and 42 ( 2/41 ) ( Table 2 ) . Changes at codon 48 occurred in all strains , and changes at codon 52 and 42 were found in several strains ( Figure 2 ) . Variation was polymorphic at codon 48 with six types of substitutions of the wild arginine ( Figure 3 ) . Glycine and glutamic acid substitutions were the most frequent , isoleucine and tryptophane were isolate-specific , and valine and threonine were transient ( Table 2 ) . At codon 42 , the wild asparagine was displaced by tyrosine and also transiently by isoleucine . Changes at the other codons were monomorphic . Overall , this indicated that despite the large diversity of isolates assessed , virulence was associated with a restricted number of substitutions within the central region of the VPg , codon 48 being overrepresented , and amino acids G48 and E48 the most frequent . The ultimate sequencing of the VPg of 14 of the 19 virulent variants of isolate Mg16 revealed a glutamic acid ( Table 2 ) . The time needed for E48 to emerge ranged from 3 to 7 mo after inoculation ( Table S1 ) . Once fixed , E48 remained stable over time . Before E48 fixation , there was either the wild arginine , an R/G mixture , a glycine , or a G/E mixture ( Table S1 ) . Changes in codon 48 occurred alongside substitutions in other codons five times , twice with codon 42 , twice with codon 52 , and once with both ( Table S1 ) . In four instances , the changes occurred first in the “alternative” codons 42 or 52 , but ultimately E48 was fixed and the alternative substitution was displaced . Overall with isolate Mg16 , E48 was the most prevalent virulent variant , and most often emerged after G48 and displaced the other mutants . Similarly , one CI4 virulent variant had a glycine at position 48 and a glutamic acid at a later stage of infection ( unpublished data ) . Altogether , this suggested two successive transitions—from R to G , then G to E—at codon 48 to gain virulence ( Figure 4 ) . Glycine and glutamic acid were found in 29 of the 41 virulent variants generated from isolates of all strains with the notable exception of strain S2/S3 ( Table 2 ) . Glycine and glutamic acid at codon 48 were fixed in resistant cultivars Gigante and Bekarosaka , and in the four NILs containing the Rymv1-2 resistance gene . G48 and E48 were generated both in growth chambers and in greenhouse conditions in West and East Africa . Altogether , the two-step R/G/E mutational pathway at codon 48 ( named mutational pathway I ) was the most prevalent , and occurred in most strains , whatever the genetic background and the experimental conditions . The two other mutational pathways at codon 48 were isolate-specific . Earlier experiments revealed that the resistance breakdown of cv Gigante after inoculation with isolate CI4 was caused by a R48I substitution in the VPg [15] . Isoleucine emerged three additional times in the present experiments , again after inoculation of isolate CI4 to cv Gigante . The virulent variant CI4 with isoleucine was reinoculated to Gigante . In several instances , sequencing the VPg 5 mo after infection revealed a mixture of isoleucine ( ATA ) and valine ( GTA ) . The isoleucine-valine mixture was still apparent after re-inoculation of resistant plants . This indicated that valine was stable after emergence , but did not displace isoleucine . The successive R/I/V amino acid emergence suggested a second mutational pathway at codon 48 ( named mutational pathway II ) gained by two transversions . In all , 70 S2/S3 isolates were inoculated to Rymv1-2 accessions and more than 1 , 000 plants were tested . Only one virulent variant emerged from isolate Ma203 . Arginine at codon 48 , coded by AGG as in most S2/S3 isolates versus AGA in the other strains , was displaced by tryptophane ( TGG ) ( Table 2 ) . This amino acid was never observed in any wild-type isolates or in any of the 40 other virulent isolates . There was no substitution elsewhere on the genome . This one-step mutational pathway ( named mutational pathway III ) involved a single nucleotide transversion . The genetic barrier to virulence was defined , by analogy to viral drug resistance [28] , as the propensity for the virus to overcome the resistance by developing virulence mutations . Marked differences in virulence propensity were found among isolates and strains of RYMV . Only one isolate of the 70 S2/S3 isolates ( 1 . 4% ) tested gained virulence . This percentage was significantly lower than that of the other strains ( χ2 = 9 . 8 , df = 1 , p = 0 . 001 in the Yates corrected χ2 test ) , where nine of the 44 isolates ( 20 . 5% ) overcame the Rymv1-2 resistance , a percentage similar to the 17 . 5% estimated earlier using a collection of 280 S1 and Sa isolates from the Sudano-savannah zone [27] . Considering that more large-scale experiments were conducted with S2/S3 isolates ( over 1 , 000 plants were tested in total ) , the difference between S2/S3 and other strains was still underestimated . Accordingly , the genetic barrier to virulence was higher in S2/S3 than in any other strain . Failure of the S2/S3 strain to follow mutational pathway I did not reflect its specific codon usage at codon 48 , as glycine ( GGG ) and glutamic acid ( GAG ) could also be coded by two successive transitions from arginine ( AGG ) . However , strain S2/S3 was the only strain with the motif TK at codons 49 and 50 . Other strains had the motifs ER or EK , except some variants of the S1 and S1-ca strains ( Figure S1 ) . Three substitutions distinguished the two motifs , two non-synonymous substitutions at codon 49 ( nt 145 and 146 ) , and one at codon 50 ( nt 149 ) ( Figure S1 ) . ( AGG ) 48 was found when ( ACG ) 49 was fixed , which itself occurred only with ( AAG ) 50 . The Pagel correlation test indicated that these changes were coordinated ( p < 0 . 01 ) , and the highly conservative concentration changes test further suggested that the changes were concentrated ( p = 0 . 07 ) . Altogether , failure of the S2/S3 strain to follow mutational pathway I was associated with the presence of T49 . Interestingly , codon 49 was under diversifying selection . Isolate Mg16 ( strain S4 ) , which originated from the northwest of Madagascar , overcame Rymv1-2 resistance at the exceptional rate of 80% , whereas only 1%–20% of the plants gained virulence in the other isolates , a percentage consistent with that estimated earlier [27] . In one additional survey , 13 other isolates from Madagascar were inoculated to resistant plants . Only isolates Mg11 and Mg35 broke the resistance at a rate as high as that of Mg16 ( 8/20 and 9/20 infected resistant plants , respectively ) . As isolate Mg16 , they originated from the northwest of Madagascar in the region of Marovoay and had the same genetic determinism of virulence ( E48 was fixed in Mg35 , whereas there was a G/E mixture at codon 48 or a Y52 substitution in Mg11 ) . Isolates Mg11 , Mg16 , and Mg35 differed from any other isolates in their VPg by a lysine at codon 26 and a glutamic acid at codon 28 ( Figure 2 ) . In another survey , seven out of 54 isolates from 16 sites of eastern Tanzania readily overcome the resistance ( unpublished data ) . Six of them belonged to the S4 strain , and like isolate Tz225 , originated from the north of Lake Malawi and had fixed a G48 or E48 . They all shared a valine at codon 26 , an amino acid never found in any wild or resistance-breaking isolates . Altogether , high virulence propensity in two geographically localized variants of the S4 strain was associated with lysine and valine at codon 26 . Interestingly , codon 26 was under diversifying selection . The virulence substitution R48I in the VPg had been identified by comparison of the full sequences of the evolved and the wild isolate CI4 , and validated by mutation of the infectious clone CIa [15] . Similarly , a virulent variant from isolate CI4 with G48 was fully sequenced and compared to the wild CI4 isolate . The virulent mutant differed specifically from the wild CI4 by only one substitution in the VPg: R ( AGA ) 48G ( GGA ) . The four other changes , three synonymous ( C1547T , T2464C , T3223C ) and one non-synonymous ( C3395T ) substitutions , all located outside the VPg , were frequent in wild avirulent isolates . Full sequencing of this virulent variant at a later stage of infection revealed an E48 in the VPg instead of a G48 and no changes elsewhere in the genome . Subsequently , G48 and E48 were the only candidate substitutions to virulence . The substitutions A1728G and G1729A were introduced by directed mutagenesis into the VPg of the infectious CIa clone to produce a non-synonymous R48E substitution in order to validate the role of E48 in resistance breaking . The mutated clone CIa with E48 was fully infectious in the resistant cultivars Gigante and Bekarosaka . This indicated that E48 caused virulence and that the S2/S3 genetic background of the CIa clone did not interfere with infection . The substitution A1728G was also introduced into the VPg of the infectious clone to produce a non-synonymous R48G . By contrast , the mutated clone with G48 was not infectious either in the susceptible cv IR64 or in the resistant cv Gigante . Doubling the amount of transcript at inoculation failed to induce successful infection . Another attempt , where the infectious clone was mutated to produce G48 with the alternative codon GGA ( instead of GGG ) , also failed to infect either the susceptible cv IR64 or the resistant cv Gigante . At codon 52 , the causal role of substitution H52Y in resistance breaking was successfully validated by introduction of the T1740C into the CIa infectious clone .
The theme and variations found in the evolutionary pathways to virulence of RYMV illustrate the extent of parallel evolution within a widely variable RNA plant virus species . In particular , the independent and repeated occurrence of E48 to gain virulence in most strains of RYMV , whatever the Rymv1-2 genetic background and plant growth conditions , showed a high degree of parallel evolution . The high frequency of parallel mutations during intervals of virulence acquisition indicates both that RYMV is able to rapidly explore the adaptive landscape , fixing favorable mutations to virulence , and that there are a limited number of pathways across the adaptive landscape . Our results provide insights into the ways RYMV explores the adaptive landscape and into the constraints restricting the number of mutational pathways . RYMV most often gained virulence through two-step mutational pathways between adjacent states: R ( AGA ) to G ( GGA ) to E ( GAA ) ( mutational pathway I ) , and R ( AGA ) to I ( ATA ) to V ( GTA ) ( mutational pathway II ) . In mutational pathway I , glutamic acid arose and replaced glycine , whereas in mutational pathway II valine arose but co-existed with isoleucine . Competitive exclusion of emerging variants also occurred between codons , with substitution E48 displacing alternative virulence substitutions . In a few plant viruses , creation of chimeric viruses or directed mutagenesis of infectious clones showed that two mutations in two different codons were necessary to gain virulence [29] . However , host–plant adaptation , such as virulence through an ordered succession of substitutions , moreover within the same codon , had never been reported . These results suggest that virulence was most often gained through a process of stepwise competitive exclusion of emerging variants . This stepwise and ordered accumulation of substitutions between adjacent states did not support the alternative scenario of a selection of pre-existing virulent isolates within a pool of variants . Consistently , work with cloned infectious proviruses of the simian immunodeficiency virus showed that the env gene evolved along similar paths in different individual hosts and that the parallel mutations were generated de novo rather than selected from viral quasispecies [30] . The critical role of virus polymorphism in shaping evolutionary pathways was most apparent with strain S2/S3 . Virulence mutational pathway III with tryptophane at codon 48 was specific to isolate Ma203 of the S2/S3 strain . In isolate Ma203 , as in most S2/S3 isolates , arginine at position 48 was coded by AGG instead of AGA in the other strains . Then , tryptophane ( TGG ) can be coded by a single substitution from AGG codon , whereas two substitutions are necessary from AGA . This difference in codon usage likely explains why mutational pathway III was not followed except in strain S2/S3 . Moreover , strain S2/S3 did not follow the major mutational pathway I . G48 was never observed experimentally in S2/S3 isolates . The S3 clone CIa mutated with G48 was not infectious . As G48 conferred virulence to Rymv1-2-resistant lines in isolates of any other strains , G48 may be unfit in the S2/S3 genetic context . This possibly blocked the major mutational pathway to virulence as simultaneous double mutation from arginine ( AGG ) to glutamic acid ( GAG ) was most unlikely . Similarly , PVX variants with single mutations , intermediate in production of doubly-mutated resistance-breaking isolates , were counter-selected , explaining the durability of the Rx1 resistance [31] . Consistently , from a given influenza A virus hemagglutinin ( HA ) sequence , several mutations were required to yield an antigenically distinct HA , but little or no fitness advantage was conferred by any subset of these mutations [4] . The extent of parallel evolution to gain virulence reflects the specificity of the VPg–eIF ( iso ) 4G relationship and the restricted number of ways to restore compatible interactions between RYMV and the Rymv1-2 resistance gene . Although the diversity of the highly variable RYMV was considered , the substitutions conferring virulence were localized within a 15-aa-long stretch within the central domain of the VPg . This suggested that this domain ( aa 38 to 52 ) , especially position 48 , played a critical role in the interaction with the eIF ( iso ) 4G protein involved in Rymv1-2 resistance . Similarly , the central domain of the VPg of several potyviruses was involved in the interaction with host translation initiation factors [32] . Glutamic acid substitution at codon 48 was the final stage of the prevalent mutational pathway I . An E/K substitution at codon 309 of the eIF ( iso ) 4G gene in rice conferred resistance to RYMV [14] . Arginine and lysine are basic amino acids , whereas glutamic acid is acidic . Accordingly , the E48–K309 interaction between the virulent isolate and the resistant rice would restore more efficiently than any other variants the charge complementarity of the R48–E309 interaction between the avirulent isolate and the susceptible rice . This coordinated pattern of substitutions between the resistance gene and the virulence determinant suggested a direct binding between position 48 of the VPg and position 309 of the eIF ( iso ) 4G that should be tested experimentally . A high genetic barrier to virulence of strain S2/S3 was associated with threonine at position 49 . High propensity to virulence of localized geographical S4 variants in northwestern Madagascar and western Tanzania was linked , respectively , to lysine and valine at position 26 . Codons 49 and 26 of the VPg , candidate positions in modulating the genetic barrier to virulence , were both under diversifying selection . Then , they played a critical role in virus evolution , possibly through virulence . Similarly , with the human rhinovirus , a large proportion of diversifying residues was found in the vicinity of domains influencing the RNA/VPg primer binding [33] , and sites of the VPg of some plant viruses under positive selection were also involved in virulence [34 , 35] . Possibly , the adjacent T49 in the S2/S3 strain , instead of E49 in the other strains , altered the ability of G48 to bind site 309 of the eIF ( iso ) 4G gene . This functional dependency between adjacent codons , a possible case of sign epistasy [36] , was also observed with HIV-1 [37] . Control of RYMV in Africa through propagation of Rymv1-2 resistance was initiated recently . Our results suggest that this strategy is most likely to be successful in the forested parts of West Africa where isolates of the S2/S3 strain with a high genetic barrier to virulence exclusively occurred . By contrast , in several other regions , the resistance is likely to be challenged . In savannah and sahelian regions of West Africa , the proportion of isolates able to overcome the resistance of cv Gigante reached 15% [27] . The proportion of virulent isolates was higher in Central than in West Africa . Our results further introduced the idea of a contrasted geographical distribution of localized variants with a high propensity to virulence in regions with a majority of isolates with a high genetic barrier to virulence . The risk of selection and spread of these variants after propagation of Rymv1-2 resistance should be assessed . The strategy of an RNA plant virus such as RYMV to gain virulence against host resistance showed striking parallels with HIV resistance against antiviral treatments . ( i ) When antiviral therapy fails to be fully suppressive , viral variants with decreased susceptibility to protease inhibitors ( PIs ) can emerge [38] . Similarly , emergence of virulent RYMV isolates by mutation assumed residual multiplication of the wild isolate in the Rymv1-2 accessions . Accordingly , low but significant multiplication of wild isolates in resistant rice cultivars was detected in Q-RT-PCR ( N . Poulicard , A . Pinel , and E . Hébrard , unpublished results ) . Subsequently , the combination of a small amount of replication of the wild-type isolate and the strong selection pressure imposed by host plant resistance allows the virulent variant to emerge and to displace the wild type . ( ii ) The development of resistance to PIs is usually a gradual process , and the development of high levels of resistance usually requires an ordered accumulation of multiple mutations in the viral protein [39] . With RYMV , the stepwise R/G/E and R/I/V substitutions at codon 48 and the displacement of other changes in alternative codons over time also illustrate the gradual process leading to virulence . ( iii ) Upon PI treatment , differences in baseline polymorphism between HIV-1 subtypes may result in the evolution of drug resistance along distinct mutational pathways , or in differences in the incidence of these specific pathways [8 , 9 , 28 , 40] . Interestingly , synonymous genetic polymorphism between HIV-1 subtypes at key resistance mutations also influenced mutational routes to drug resistance [41] . Similarly , genetic diversity of RYMV—including synonymous polymorphism—affected the mutational pathways and the virulence propensity . ( iv ) The primary mutations against PIs did not occur in wild-type polymorphism , but developed during the course of antiviral treatment failure [8] . Consistently , RYMV virulence mutations were not found in wild isolates but emerged throughout the process of infection in resistant plants . ( v ) Additional ( novel , minor , secondary ) mutations , some of them in the close environment of key virulence mutations , modulated the genetic barrier to the development of drug resistance [38 , 40 , 42] . Similarly , amino acids at codons 49 and 26 in the VPg of RYMV were candidate positions to modulate the genetic barrier to virulence among strains and variants . The similarities in the processes of evolutionary changes between RYMV and HIV-1 to gain , respectively , virulence against a host plant resistance and resistance to antiviral treatments illustrate that common mechanisms operate in RNA virus evolution and that similar forces shape the genetic structure of their populations [43] .
The response to RYMV of the Rymv1-2 allele of resistance was tested in six genetic backgrounds made of two cultivars and four NILs . Cultivars Gigante and Bekarosaka are two indica cultivars that share the Rymv1-2 allele of resistance [21] . Cultivar Bekarosaka originated from Madagascar , whereas cv Gigante is assumed to be from Mozambique . Rymv1-2 was introgressed into three widely grown susceptible indica cultivars ( BG90–2 , Bouaké189 , and IR64 ) , and into one partially resistant japonica cultivar ( Nipponbare ) to derive NILs . Cultivars Gigante and Bekarosaka were challenged with isolates of the major strains , whereas the NILs were inoculated only with isolate CI4 of strain S1 . The plants were kept in a growth chamber under 12-h illumination at 120 μEm−2s−1 of PAR at 28 °C and 90% humidity . In all , 114 isolates representative of the main strains of RYMV and originating from various regions of Africa were collected on susceptible plants in the fields and inoculated to the Rymv1-2 accessions: West Africa ( strain S1 from savannah , S2/S3 from forest , and Sa from sahelian regions ) , East Africa ( S4 , S5/S6 ) , and Madagascar ( S4 ) ( Table 1 ) . Strains S2 and S3 , earlier considered as distinct , were gathered together ( referred to as S2/S3 strain ) , as more intensive survey revealed a continuum between the two strains . In each experiment , the isolate was inoculated to ten to 20 plants per Rymv1-2 accession . Larger scale experiments with inoculations of 50–200 plants were also conducted either to generate a virulent isolate from a recalcitrant strain ( S2/S3 ) , or to obtain several virulent variants from the same wild isolate ( CI4 , Mg16 , Tz209 ) in order to assess the intra-isolate spectrum of substitutions associated with virulence . Inoculum was prepared by grinding infected frozen leaves in 0 . 1 M phosphate buffer ( pH 7 . 2 ) ( 0 . 1g/ml ) . Extracts were mixed with 600-mesh carborundum and rubbed on leaves of 14-d-old rice seedlings . Symptoms were monitored weekly and virus content was assessed by ELISA over time . Plants were kept up to 13 mo after inoculation . Such a length of time is biologically realistic , even for annual cultivated rice with a growing season of c . 3 mo , as regrowth of infected rice stubble is frequent after harvesting . Isolates that induced high virus content and/or generalized symptoms on Rymv1-2 accessions were collected . A total of 59 isolates from the different geographical regions in Africa and all of the strains of RYMV were collected in the fields ( referred to as the wild type ) and analyzed , usually after multiplication in the susceptible indica cultivar IR64 in greenhouses in order to increase virus content ( Table S2 ) . Twenty-two of the 59 isolates were fully sequenced as described previously [18] . The VPg and its 5′ and 3′ neighboring regions of the 37 other isolates were sequenced as done earlier [15] . For phylogenetic purposes , the coat protein of these isolates was also sequenced as described elsewhere [44] . The VPg of virulent variants generated by infection of Rymv1-2-resistant accessions ( referred to as the evolved virulent variant ) from seven of these 59 isolates was sequenced . Complementarily , the VPg of three virulent isolates generated in greenhouse conditions at the experimental sites of the INERA research station of Kamboinsé near Ouagadougou ( Burkina Faso , West Africa ) and of the Dar Es Salaam University Department of Botany ( Tanzania , East Africa ) —where temperature , light , and humidity were less controlled than in growth chamber , but closer to field conditions—was studied . The same plants were tested before and after resistance breakdown . The last fully developed leaf was sampled . Altogether , 41 virulent variants were analyzed . Multiple peaks at the same position in a sequencing electrophoregram are currently interpreted as reflecting nucleotide polymorphism [45] . In our experiments , electrophoregrams with distinct double peaks at a position were regarded as indicating a mixture of two nucleotides . Displacement was further suggested if one nucleotide was identified singly at an early stage and the other at a later stage of infection . Full sequence comparison and directed mutagenesis were done as described elsewhere [15 , 18 , 46] to validate the role of the major substitutions to virulence . The phylogeny of the 22 fully sequenced isolates representative of the geographic and genetic diversity of RYMV was reconstructed by maximum likelihood method with the HKY model [47] using an heuristic search implementing a tree bisection and reconnection swapping algorithm applied in PAUP [48] . The transversion/transition ( tv/ti ) ratio and the alpha parameter of the gamma distribution of the among-site variation were estimated by maximum likelihood . The bootstrap support of the nodes was estimated by 100 replicates with the full heuristic search . Analysis of the genetic diversity of the VPg was based on 59 isolates collected in the fields representative of the geographic and genetic diversity of RYMV , including the 22 fully sequenced reference isolates . The ratio of non-synonymous ( dN ) versus synonymous ( dS ) substitutions of the VPg of the 22 reference isolates was calculated as implemented in DnaSP [49] and compared to ORF1 , ORF2a , ORF2a+b , VPg , and ORF4 . The dN/dS ratio at individual codons in the VPg was calculated on the corpus of 59 isolates using two maximum likelihood methods , Random Effect Likelihood ( REL ) and Internal Fixed Effect Likelihood ( IFEL ) [50 , 51] , to determine on each of the 79 codons whether the selection pressure was conservative ( negative ) ( w < 1 ) , neutral ( w = 1 ) , or diversifying ( positive ) ( w > 1 ) . REL is an improved variant of the Nielsen-Yang approach , which allows both dS and dN to vary independently across sites . IFEL is a new likelihood method to fit an independent dN and dS to every site in the context of codon substitution and test whether dN ≠ dS . The analyses were conducted with the VPg sequence ( nt 1587–1823; 237 nt ) , and with the VPg and the flanking regions ( nt 1526–2065; 540 nt ) in order to increase the statistical power of the tests . The tv/ti ratio in the VPg was estimated from the corpus of 59 isolates by maximum likelihood as implemented in HYPHY . Evidence for correlated evolutionary change in two characters ( i . e . , nucleotides or amino acids at different positions ) was tested by the Pagel correlation test [52] as implemented in Mesquite [53] . The concentration changes test [54] as implemented in MacClade [55] was applied to determine whether changes in one character ( the dependent character ) are concentrated on branches of a tree that have a particular state of a second character ( the independent character ) . | Parallel changes in independently evolving lineages are important , but their contribution to pathogen evolution has not been assessed at the species level . We investigated the extent of phenotypic and genotypic parallel evolution in a highly variable RNA plant virus species , Rice yellow mottle virus ( RYMV ) . Isolates representative of the genetic and geographic diversity of RYMV in Africa were inoculated to several rice cultivars with eIF ( iso ) 4G-mediated Rymv1-2 resistance . The theme and variations in the evolutionary pathways to gain virulence found in the VPg of RYMV illustrate the frequency of parallel evolution . The repeated occurrence of the R48E substitution in the VPg of most strains , whatever the Rymv1-2 background and plant growth conditions , showed the specificity of parallel evolution that operated through the same pathway , locus , and mutation . The frequency and specificity of parallel mutations indicate , respectively , that RYMV is able to rapidly explore the adaptive landscape , fixing favorable mutations to virulence , and that there are a limited number of pathways across the adaptive landscape . Our results provide insights into the ways an RNA virus species explores the adaptive landscape and into the constraints restricting the number of mutational pathways . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"viruses",
"virology"
] | 2007 | Theme and Variations in the Evolutionary Pathways to Virulence of an RNA Plant Virus Species |
Mad2 is a key component of the spindle assembly checkpoint , a safety device ensuring faithful sister chromatid separation in mitosis . The target of Mad2 is Cdc20 , an activator of the anaphase-promoting complex/cyclosome ( APC/C ) . Mad2 binding to Cdc20 is a complex reaction that entails the conformational conversion of Mad2 from an open ( O-Mad2 ) to a closed ( C-Mad2 ) conformer . Previously , it has been hypothesized that the conversion of O-Mad2 is accelerated by its conformational dimerization with C-Mad2 . This hypothesis , known as the Mad2-template hypothesis , is based on the unproven assumption that the natural conversion of O-Mad2 required to bind Cdc20 is slow . Here , we provide evidence for this fundamental assumption and demonstrate that conformational dimerization of Mad2 accelerates the rate of Mad2 binding to Cdc20 . On the basis of our measurements , we developed a set of rate equations that deliver excellent predictions of experimental binding curves under a variety of different conditions . Our results strongly suggest that the interaction of Mad2 with Cdc20 is rate limiting for activation of the spindle checkpoint . Conformational dimerization of Mad2 is essential to accelerate Cdc20 binding , but it does not modify the equilibrium of the Mad2:Cdc20 interaction , i . e . , it is purely catalytic . These results surpass previously formulated objections to the Mad2-template model and predict that the release of Mad2 from Cdc20 is an energy-driven process .
The process of mitosis is designed to deliver the faithful , equational division of the replicated genome . The generation of tightly connected replicated chromosomes ( sister chromatids ) during S-phase is prerequisite to the division process . Sister chromatid cohesion is then removed synchronously and irreversibly at the so-called metaphase-to-anaphase transition , when the individual sisters are distributed to the daughter cells in two equal masses [1] . Before sister chromatid cohesion is removed , every pair of sister chromatids must have achieved bipolar orientation at the metaphase plate of the mitotic spindle . A biochemical device named the spindle assembly checkpoint ( SAC ) exercises tight control over the timing of anaphase , and ensures that anaphase only takes place after all sister chromatid pairs are bioriented [2] . The loss of sister chromatid cohesion at the metaphase-to-anaphase transition requires the activity of the Ub-ligase anaphase-promoting complex/cyclosome ( APC/C ) and of its activator Cdc20 [3] . To synchronize anaphase with the completion of bipolar orientation of all sister chromatid pairs , the SAC targets and inactivates Cdc20 . Two SAC components , the Mad2 protein and the BubR1:Bub3 complex ( we indicate a complex of two or more proteins with the names of the proteins separated by a colon ) , are able to bind Cdc20 directly [2] . Together , these proteins form the mitotic checkpoint complex ( MCC ) , which is thought to act as a pseudosubstrate inhibitor of the APC [4 , 5] . Central questions in the checkpoint field concern the mechanism of checkpoint activation early in mitosis and the mechanism of checkpoint inactivation prior to anaphase . Unattached kinetochores are believed to play a major role in SAC activation [2] . Kinetochores are complex protein scaffolds assembled on mitotic chromosomes and are responsible for microtubule capture [6] . Kinetochores of sister chromatid pairs that have not attained bipolar attachment are responsible for checkpoint activation and maintenance [7] . Consistently , all key players of the SAC ( including , among others , Mad1 , Mad2 , Mps1 , Bub1 , BubR1 , Bub3 , and Cdc20 itself ) localize at unattached prometaphase kinetochores [2] . Eventually , once the last kinetochore has attached , Cdc20 becomes able to activate the APC towards cyclin B and securin , its critical substrates at the metaphase to anaphase transition , and the cell exits from mitosis [3] . The mechanism whereby unattached kinetochores regulate the binding of Mad2 to Cdc20 , possibly the first step in the assembly of the MCC , is still unclear . Mad2 comes in two different conformations , open ( O-Mad2 ) and closed ( C-Mad2 ) , that differ for the position of the C-terminal tail ( Figure 1A and 1B , reviewed in [8] ) . Before SAC activation , most Mad2 is found in the monomeric open form and is not bound to Cdc20 . Upon SAC activation , O-Mad2 binds to its Cdc20 target and switches to the closed form [9 , 10] . Throughout the cell cycle , a remaining 10%–25% of Mad2 is engaged in a very stable complex with Mad1 [10–12] in which Mad2 holds the C-Mad2 conformation ( Figure 1B ) . The role of the Mad1:C-Mad2 complex in the sequestration of Cdc20 by Mad2 is underlined by four key observations . First , the tight Mad1:C-Mad2 complex acts as the kinetochore receptor of cytosolic O-Mad2 , via the “conformational” dimerization of C-Mad2 and O-Mad2 , a reaction that creates the trimer Mad1:C-Mad2:O-Mad2 ( Figure 1C ) [13 , 14] . Second , Mad1 is required for Mad2 to bind Cdc20 , at least in a normal cell cycle ( reviewed in [2] ) . Third , Mad2 mutants that are able to bind Mad1 , but that are impaired in the formation of Mad2 conformational dimers , cannot complement the SAC defect when expressed in a mad2-deleted strain [13 , 15] . Fourth , Mad1:C-Mad2 has been recently shown to convert O-Mad2 into C-Mad2 , and evidence for its role in the formation of Cdc20:C-Mad2 has been provided [16] . On the basis of these data , it has been suggested that Mad1-bound C-Mad2 starts a catalytic amplification of the checkpoint by binding to O-Mad2 in the Mad1:C-Mad2:O-Mad2 trimer , and thus facilitating the conversion of O-Mad2 into Cdc20-bound C-Mad2 ( the network is described by reactions 1–3 in Figure 1D ) . This still speculative model of Mad2 activation is named the Mad2-template model [13] . A possible twist to the model comes from the observation that Cdc20-bound C-Mad2 is a structural copy of Mad1-bound C-Mad2 . Once released in the cytosol , Cdc20-bound C-Mad2 ( located on the overlap between the yellow and the grey hexagons in Figure 1D ) might therefore propagate the O-Mad2 conversion away from kinetochores through an autocatalytic loop ( Figure 1D , reactions 4 and 5 ) . As the molecular mechanisms of the SAC are investigated at increasingly deeper detail , it becomes progressively more attractive to develop mathematical models to rationalize SAC behaviour and to predict the effects of its manipulation . Recent theoretical studies proposed that the autocatalytic loop of the Mad2-template model would force the cell to be permanently arrested in a state of operational SAC , with most Cdc20 sequestered by Mad2 [17 , 18] . Other studies argued that the contribution of the autocatalytic loop to the SAC is negligible [19] . The previous studies , however , have neglected two fundamental kinetic and thermodynamic implications of the conversion of Mad2 from the open to the closed conformation . The kinetic implication is that the dramatic structural rearrangement of Mad2 can be expected to translate in a very slow , natural on-rate of binding to Cdc20 . Were this true , the acceleration predicted by the “template” might be required to accelerate the formation of Mad2:Cdc20 complexes required to halt progression into anaphase . The thermodynamic implication is that the template model , as it is cast in Figure 1D , does not imply irreversible reactions ( Figure 1E ) . Thus , the rate equations of the template model describe the influence of Mad2 dimerization on the rate at which the Mad2:Cdc20 complex forms , but do not imply a modification of the equilibrium concentrations of the Mad2:Cdc20 complex . Here , we show through a combination of experimentation and mathematical modelling that catalytic amplification of Mad2:Cdc20 complex formation is required as a first step of checkpoint activation to overcome the kinetic barrier built in the Mad2:Cdc20 interaction .
To gain insight into the basal rate of the interaction of O-Mad2 and Cdc20 ( reaction 1 in Figure 1D ) , it is important to remove the possible effects of dimerization . Previously , we have described several Mad2 point mutants that are impaired in dimerization , but that retain the ability to bind to Cdc20 in vitro [13 , 15 , 20 , 21] . In one such mutant , Phe141 of Mad2 is replaced by Ala ( Mad2F141A ) . Mad2F141A is unable to sustain the “conformational” dimerization of Mad2 and is unable to complement a mad2 deletion strain of Saccharomyces cerevisiae [15 , 21] . In a first set of experiments , we sought evidence that wild-type Mad2 ( Mad2wt ) and Mad2F141A bind Cdc20 with similar affinity . This expectation is sensible when considering that Phe141 is localized at the interface between O-Mad2 and C-Mad2 , and that the dimerization interface occupies the opposite end of Mad2 from where Cdc20 binds ( Figure 1A and 1B ) . By using recently described approaches , Mad2wt and Mad2F141A were purified to homogeneity in a monomeric state and in the O-Mad2 conformation ( Figure 2A and 2B ) . A 1 μM concentration of these species was then incubated with a 1 μM concentration of a construct encompassing the Mad2-binding region of Cdc20 fused to glutathione S-transferase ( GST; Figure 2C ) . Indeed , at equilibrium ( i . e . , after a 24-h incubation ) , we found similar amounts of Mad2wt and Mad2F141A on the GST-Cdc20 beads , indicating that Mad2F141A , like other Mad2 mutants that are impaired in conformational dimerization , has a very similar binding affinity for Cdc20 as Mad2wt ( Figures 2C and S1 ) . We therefore proceeded to analyze the rate of binding of Mad2wt and Mad2F141A to GST-Cdc20 . Mad2wt had reached maximal binding between 1 and 3 h . Conversely , it took Mad2F141A between 12 and 24 h to reach maximal binding ( Figure 2D ) . These results suggest that the abrogation of the ability of Mad2 to form conformational dimers slows down the binding to Cdc20 in this assay . Indistinguishable results were obtained with another Mad2 dimerization mutant , Mad2R133A ( unpublished data ) . In summary , the fact that Mad2F141A is impaired in Mad2 conformational dimerization , and that its overall binding affinity to Cdc20 is unchanged relative to Mad2wt , supports our argument that the rate of binding of Mad2F141A to Cdc20 represents the basal rate of binding of Mad2 to Cdc20 in the absence of Mad2 dimerization . Strong additional evidence in favour of this proposition is provided in the next sections . To quantify the association rate between Cdc20 and Mad2 in vitro , we developed a real-time assay based on the binding of Alexa Fluor 488–labelled Mad2 ( Alexa-Mad2 ) to a surface containing immobilized Cdc20 in a flow cell . The method is conceptually similar to the Biacore method , but is in principle amenable to multicolour analysis . Alexa-Mad2F141A and Alexa-Mad2wt retained their monomeric O-Mad2 conformation after covalent fluorescent labelling ( Figure 3A ) . As they bind to Cdc20 on the surface of the flow cell , they convert into C-Mad2 ( Figure 3B and 3C ) . As the reaction proceeds , the signal in solution decreases , while the signal on the surface increases . By measuring the fluorescence on the surface ( or the signal in solution; Figure S2 and Text S1 ) by confocal microscopy , we followed the binding kinetics at different concentrations of Mad2F141A ( Figure 3C ) . We first attempted to interpret the kinetics of Mad2F141A binding to Cdc20 on the basis of reaction 1 in Figure 1D . Following the increase of signal on the surface , we could fit the experimental data over a 4-fold change of [Mad2F141A] using kbind , on = 4 . 83 × 10−5 μM−1 s−1 and kbind , off = 4 . 83 × 10−6 s−1 , confirming that the rate constants for the interaction of Mad2F141A are exceptionally small ( Figure 3D ) . By comparison , this on rate is almost four orders of magnitude slower than that of Mad2 dimerization ( see below ) . The small rate constants are consistent with the semiquantitative binding experiments carried out with unlabelled Mad2F141A ( Figure 2D ) . As the dissociation constant ( KDbind ) is given by the ratio kbind , off/kbind , on , our kinetic analysis predicts a KD of 100 nM for the interaction of Mad2F141A with Cdc20 , in excellent agreement with previous analyses [9 , 13 , 16 , 22] . In summary , the kinetic analysis on Mad2F141A confirms the hypothesis that the Mad2 conformational change is very slow and likely rate limiting for checkpoint activation . Several factors might contribute to accelerate the basal rate of binding revealed by our experiments . For instance , it has been reported that several kinases , including checkpoint kinases , target Cdc20 , thus possibly contributing through phosphorylation to its susceptibility to be inhibited by the SAC [23–27] . Although there might be numerous additional factors impinging on the velocity of formation of Mad2:Cdc20 complexes , we decided to test the concept that conformational dimerization of Mad2 is important to accelerate this binding reaction , as advocated by the Mad2 template model . To measure the effect of dimerization in the binding between Mad2 and Cdc20 , we repeated the real-time binding experiments with Mad2wt , which—compared to Mad2F141A—has the ability to dimerize . In agreement with our hypothesis , the half-time of binding of Mad2wt to Cdc20 at 2 μM Mad2 was five to six times faster than that of Mad2F141A ( Figure 4A ) . The binding kinetics were sigmoidal , typical for an autocatalytic reaction with a slow initial phase dominated by the binding of O-Mad2 to Cdc20 , which is followed by a faster reaction based on activation by Mad2 dimerization . We asked whether we could account for these results in silico by decomposing the reaction into three steps: binding , dimerization , and catalysis ( i . e . , reactions 1 , 4 , and 5 in Figure 1D . ) Based on the fact that Mad2F141A and Mad2wt have the same affinity for Cdc20 ( Figures 2C and S1 ) , kbind , on and kbind , off were assigned values previously determined for Mad2F141A . Association and dissociation constants for dimerization , reaction 4 , have been chosen to be compatible with the values measured experimentally: 0 . 3 μM−1 s−1 and 0 . 45 s−1 , respectively [13 , 28] ( Table 1 ) . Instead , the rates of the hypothetical catalytic reaction ( reaction 5 ) are unknown . However , it should be noted that reactions 4 and 5 form a closed loop whose net result is reaction 1 ( Figure 1E ) . Furthermore , these two reactions give rise to an autocatalytic reaction whereby Cdc20:C-Mad2 induces its own synthesis . Due to the principle of microscopic reversibility , KDcat = kcat , off/kcat , on = KDbind/KDdim , which implies that the only unknown parameter in our fitting is kcat , on . We find good fitting to the experimentally determined curves when we set kcat , on = 3 . 0 × 10−3 μM−1 s−1 ( Figure 4B and Table 1 ) . This value is approximately two orders of magnitude larger than kbind , on , the noncatalysed reaction that leads to Cdc20 sequestration . Importantly , if we simply model the reaction with two terms , where Mad2wt can only bind Cdc20 and form dimers ( reaction 1 and 4 in Figure 1D ) , but cannot catalyse the conversion of O-Mad2 ( reaction 5 ) , it is impossible to fit the experimental data ( see , for example , the simulation for [Mad2] = 1 μM in Figure 4B ) . This result strongly implies that additional reactions besides binding and dimerization take place with Mad2wt as opposed to Mad2F141A . We note that the basal rate equation ( reaction 1 , Table S1 ) is based on kinetic parameters derived from experiments carried out with Mad2F141A . Thus , our results provide very strong evidence that the rate of Cdc20 binding by Mad2F141A represents the actual basal rate of Mad2wt binding to Cdc20 in the absence of catalysis . Our reaction scheme makes two key predictions . The first prediction is that the binding of O-Mad2 to Cdc20 should occur faster if the C-Mad2 catalyst was present at the beginning of the reaction . This reflects the fact that the natural formation of C-Mad2 is a slow process , and that C-Mad2 accelerates it . To illustrate this point , we compared the fitting curve for the total concentration of Mad2wt equal to two μM ( a fitting curve already shown in Figure 4B , which takes into account reactions 1 , 4 , and 5 ) with a simulation in which the interaction between Mad2wt and Cdc20 was helped by the presence of Mad1:C-Mad2 via dimerization and catalysis according to reactions 2 and 3 in Figure 1D . ( The two curves are shown in black and red in Figure 5A ) . Due to the structural similarities between Mad1:C-Mad2 and Cdc20:C-Mad2 , we assumed that the two species induce catalysis in the same way ( i . e . , reactions 3 and 5 share the same kinetic parameters ) , and that the same holds true for dimerization ( reactions 2 and 4 ) . The simulations show that the presence of preformed C-Mad2 overrides the lag phase in the binding of Mad2wt to Cdc20 , as it provides sufficient initial “catalyst” for the reaction ( Figure 5A ) . To test the prediction , we monitored the binding of 2 μM Alexa-Mad2wt to approximately 1 μM Cdc20 ( a concentration determined by fitting and experimentally , as described in Figure S4 and Text S1 ) on a surface with interspersed 0 . 25 μM of Mad1:C-Mad2 . Mad1:C-Mad2 forms a very tight complex that cannot undergo significant dissociation within the time frame of our binding experiment [13 , 20 , 22 , 28] . The ratio of free O-Mad2 to Mad1:C-Mad2 chosen for these experiments is similar to that believed to exist in living cells ( see [28] and discussion therein ) , although we suspect that the active concentration of Mad1:C-Mad2 , i . e . , the pool of this complex that can be recruited to kinetochores , is probably smaller ( see below ) . In quantitative agreement with the simulations , the reaction had no lag phase and produced an overall 3-fold acceleration relative to the noncatalysed rate ( Figures 5B and S3 ) . Thus , preformed C-Mad2 accelerates the binding reaction , providing a very strong indication that in the absence of Mad1:C-Mad2 , the lag phase represents the slow accumulation of C-Mad2 bound to Cdc20 . Similar effects were observed when an initial “seed” of C-Mad2 was created by letting substoichiometric amounts of unlabeled O-Mad2 bind Cdc20 on the surface prior to the addition of labelled O-Mad2 ( Figure 5C and 5D ) . The second prediction is that any interference with conformational dimerization should reduce the rate of binding of Mad2wt to Cdc20 . To test the prediction , we sought to impair the dimerization of Mad2 . To this aim , we used p31comet , a structural mimic of O-Mad2 that binds to C-Mad2 with an approximately 40-fold higher affinity relative to O-Mad2 and in a manner that is competitive with the binding of O-Mad2 [21 , 28–30] . The association and dissociation rate constants for the interaction of p31comet with Mad2 have been previously determined [28] . Once this reaction is introduced into our reaction scheme ( reaction 6 , Figure 5E ) , the simulations show that the interference with conformational dimerization reduces the rate of binding of Mad2wt to Cdc20 to the rate observed with Mad2F141A , as it “poisons” the reaction catalyst by efficiently competing with conformational dimerization ( Figure 5F ) . We tested this prediction by mixing 10 μM p31comet with 2 μM Alexa-Mad2wt , and by monitoring the binding of Alexa-Mad2wt to Cdc20 in our real-time binding assay . As shown in Figures 5G and S3 , the experiments were in quantitative agreement with the prediction of our model . RNA interference ( RNAi ) experiments on SAC proteins in HeLa cells have demonstrated that the checkpoint must be already fully active approximately 10–12 min after nuclear envelope breakdown ( NEB ) [31] . In the absence of Mad2 , for instance , HeLa cells undergo a precocious anaphase approximately 10–12 minutes after NEB . At present , we do not know precisely at which point , within the approximately 10–12-minute timeframe , the checkpoint response is mounted . We are also ignorant of the mechanisms that allow cells to remain in mitosis for approximately 10–12 min when the SAC is defective , although it has been proposed that this timing might reflect cyclin A degradation , which must be completed before cells can proceed into anaphase [32 , 33] . For the purpose of our analysis , we assume that it might take between 5 and 10 min to mount a full SAC response after NEB . We used the full model in Figure 1D , with the parameters estimated in vitro , to simulate the binding between Mad2 and Cdc20 using published cellular concentrations ( 0 . 1 μM for Cdc20 and 0 . 2 μM for Mad2 . See [28] and discussion therein ) . As for the concentration of Mad1:C-Mad2 , we assumed that each unattached kinetochore binds approximately a thousand Mad1:C-Mad2 molecules ( a similar estimate has been provided for Cdc20 in PtK1 cells [34 , 35] ) . Assuming a system with 22 kinetochores and a total volume of 6 pl in ptK1 cells , we introduced a concentration of active Mad1:C-Mad2 ( i . e . , localized at the unattached kinetochores ) of 22 kinetochores × 0 . 28 nM = 6 . 16 nM . This value is based on the assumption that at the beginning of prometaphase , all kinetochores are unattached , and all contribute to the activation of the checkpoint . First , we run a simulation with Mad2F141A , which is unable to form conformational dimers and is unable to sustain the SAC [15 , 21 , 36] . The simulation shows that at equilibrium , more than 50% of Cdc20 is sequestered , which is possibly compatible with checkpoint maintenance and thus at odds with the observed checkpoint defect . However , the result can be explained looking at the time course of the reaction: sequestration rate would have a t1/2 of 9 . 5 h ( Figure 6A , blue line ) , which is clearly incompatible with rapid checkpoint activation . The substitution of Mad2F141A with Mad2wt , i . e . , the addition of dimerization and catalysis as measured in vitro , partly alleviates the problem ( t1/2 of 5 . 3 h , Figure 6A , red line ) , but does not resolve it . Thus , the checkpoint response would be unreasonably slow if we were to use the parameters measured in vitro . Incidentally , we notice that if the binding of Mad2 and Cdc20 would reach completion in living cells , the SAC would be constitutively operational , impairing cell viability . We hypothesize that in cells , such a fate is avoided by energy-dependent reactions that induce the release of Cdc20 from Mad2 , as suggested recently [37 , 38] . We can envisage two reasons for the disagreement between our in vitro data and the observations in vivo . First , our simulations might underestimate the actual concentration of active Mad1:C-Mad2 . If the kinetochores would serve as a platform where Cdc20 and Mad1:C-Mad2 react , then the local concentrations of these molecules would greatly increase . As a rough estimate , the radius of one kinetochore is 0 . 2 μm , to give a volume of 3 . 3 × 10−5 pl . This provides a local concentration for Mad1:C-Mad2 and Cdc20 on one unattached kinetochore of 46 . 6 μM . If we introduce this modification , simulations show that free Cdc20 decreases to less than 50% with t1/2 of 17 min ( unpublished data ) . The second reason might be that our model , which is exclusively based on measurements in vitro , lacks additional factors required to accelerate the binding in vivo , and therefore underestimates the extent of the catalytic component of checkpoint activation . For instance , the phosphorylation of Cdc20 , which is neglected in our modelling , might be an additional factor of acceleration flanking conformational dimerization [23–27] . We tried to quantify the increase in catalytic efficiency of the SAC machinery at the kinetochore , additionally to that provided by Mad2 dimerization , required to obtain reasonable activation kinetics . Simulations with our model show that when the catalytic potency of the kinetochores is increased 300-fold or more ( to 0 . 9 μM−1 s−1 or faster ) then the checkpoint response can be mounted within the limit of 10 min ( Figure 6A , orange trace ) . The identification of the actual molecular mechanism underlying this necessary increase in catalytic efficiency is a subject for further investigation .
Barkai and coworkers ( Doncic et al . [17] ) have discussed a “self-propagating inhibition” model as a possible representation of the Mad2-template model ( Figure 6B ) . In the self-propagating inhibition model , a protein c required for cell cycle progression , is turned at the kinetochore into an inactive species c* that can diffuse freely in the cytosol to convert more c into c* in an autocatalytic and irreversible process . It is difficult to reconcile the molecular mechanism of Mad2 activation with the formalism of the self-propagating inhibition model , and we therefore will not discuss it here in detail . Certain aspects of the Mad2-template mechanism can be approximated by an alternative model , the “emitted inhibition” model [17] ( Figure 6B ) . The model posits that kinetochores are responsible for the activation of a species e to an active inhibitory form , e* . The e* inhibitor diffuses away from kinetochores and binds to the target c , the protein required for cell cycle progression . This creates , through an irreversible step , the inhibited form c* , which can then decay spontaneously and irreversibly into c and e . Although Barkai and collaborators [17] did not define the molecular identities of e , e* , c , and c* , we propose that e is O-Mad2 , c is free Cdc20 , e* is the active form of Mad2 that binds Cdc20 , and c* is the inhibited form of Cdc20 ( i . e . , bound to C-Mad2 ) . There are two main differences between the emitted inhibition model and the Mad2 template model . The first difference concerns the presence of an autocatalytic loop in the Mad2 template model . The autocatalytic loop is based on the ability of Cdc20-bound C-Mad2 to help the conversion of O-Mad2 , and it stems naturally from the structural similarities between the Mad1-bound and the Cdc20-bound C-Mad2 . Previously , the existence of such an autocatalytic loop was ruled out based on the notion that it would instate a steady state with an operational SAC from which cells could not escape [17] . We argue that this effect was an inevitable consequence of the ( unproven ) assumption that the formation of the inhibitory complex is irreversible . Thus , a second difference between our model and the emitted inhibition model ( Figure 6B ) is that in our model , the hypothesized catalytic reactions ( autocatalysis and catalysis , reactions 3 and 5 in Figure 1D ) do not require any energy sources , and thus in the system analyzed here , they do not create or affect the steady state of the binding reaction between Mad2 and Cdc20 . All reactions are fully reversible and indeed are expected to drive the system towards a state in which a large fraction of Cdc20 is bound to Mad2 ( Figure 6A ) , but that is primarily due to the low KDbind . The steady-state Mad2-bound Cdc20 would be basically the same even in the absence of the catalytic reactions ( i . e . , only in presence of reaction 1 in Figure 1D ) This feature of the SAC might explain the ability of cells to maintain the SAC for very long times ( ∼24 h in human cells ) , and up to several days in some cases . At the same time , it poses the problem of how the SAC can be switched off in vivo . We predict , in agreement with the emitted inhibition model , that this occurs via an energy-driven process that creates a new steady state with SAC OFF ( high concentration of free Cdc20 ) . Most likely , the new steady state is achieved by inducing the energy-dependent release of Cdc20 from the complex Cdc20:C-Mad2 . Recent data showing that Cdc20 ubiquitination ( i . e . , an ATP-dependent process ) is required to free Cdc20 from bound Mad2 support our hypothesis [37 , 38] . If we recast the Mad2 template model using the formalism adopted by Barkai and colleagues [17] , we end up with the scheme shown in Figure 6B . Here , catalysis simply accelerates the completion of the reaction , as schematized in Figure 6C , and all reactions are reversible , except for the dissociation of the inhibited complex c* , which reflects the above-mentioned energy-dependent reaction that releases Cdc20 . The maintenance of a SAC OFF state—in which the Cdc20:C-Mad2 complex cannot accumulate—might be rather cheap from the energetic standpoint . This is because the very large activation energy built into the Mad2 conversion works to maintain Mad2 in the O-Mad2 conformation , at least until O-Mad2 is allowed to dimerize with C-Mad2 ( Figure 6C ) . In this scenario , the fact that the initial , noncatalysed association and dissociation rates are very slow guarantees the possibility to prevent the accumulation of Cdc20:C-Mad2 complexes . The proposed energy-dependent mechanism of Cdc20:C-Mad2 dissociation might contribute to the removal of Cdc20:C-Mad2 complexes that might form , slowly but spontaneously , in cell cycle phases other than M phase . The need of a tight Mad1:C-Mad2 complex at the start of mitosis , when the checkpoint needs to be activated , fits nicely into this picture . Mad1:C-Mad2 allows the rapid initial rate of Mad2 binding to Cdc20 , and avoids wasting time for the slow step of the reaction , the conversion from O-Mad2 to C-Mad2 . The rapid formation of Cdc20:C-Mad2 triggered by conformational dimerization and other mechanisms exceeds the ability of the energy dependent mechanisms to continuously create free Cdc20 , leading to the accumulation of Cdc20:C-Mad2 . Previous analyses have suggested that the activation of Mad1:C-Mad2 early in prometaphase might require the inactivation of p31comet , a negative regulator of the SAC that acts by binding to C-Mad2 in the Mad1:C-Mad2 complex , thus preventing its interaction with O-Mad2 [21 , 29 , 30] . The molecular details of p31comet temporary inactivation , however , remain unclear . The role of Mad1:C-Mad2 in SAC activation is illustrated with a simulation of a hypothetical system in which Mad1 binds to kinetochores as a monomer , and only later it recruits O-Mad2 to form the Mad1:C-Mad2 complex ( Figure 6A , green trace ) . Because the noncatalysed binding of O-Mad2 to Cdc20 is very slow , the binding of O-Mad2 to Mad1 , which implicates the same conformational change of Mad2 , is expected to be equally slow . Even if Mad1:C-Mad2 were endowed with increased catalytic power ( e . g . , 300-fold additional acceleration ) , its contribution would be very significantly delayed relative to that achieved with preformed Mad1:C-Mad2 ( Figure 6A , orange trace . The t1/2 would be 175 min and 10 min , respectively ) . The formation of Cdc20:C-Mad2 , in turn , is likely to result in the formation of a larger inhibitory complex ( the MCC ) with the Bub3:BubR1/Mad3 complex . Although this paper focuses on what is possibly the first step in MCC formation , hereafter we provide a speculative overview of the series of reactions that lead from the isolated components to the formation of the MCC ( Figure 6C ) . Previous analyses have shown that the presence of Mad2 favours the formation of a Bub3:BubR1/Mad3 complex in vitro and in vivo [5 , 40–45] , even if a Cdc20:Bub3:BubR1/Mad3 complex can be formed from purified components in vitro [43 , 44] . Conversely , to our knowledge , there is little or no evidence that BubR1/Mad3 is necessary for assembling a complex of Mad2 with Cdc20 . The requirements on Mad2 for the incorporation of Bub3:BubR1/Mad3 in the MCC is reminiscent of the requirement of Mad1:C-Mad2 for the formation of the Cdc20:C-Mad2 complex . It leads us to speculate that the network of interactions that are responsible for SAC activation might consist of two consecutive kinetically limited steps . In this scheme , the initial activation of Mad2 will allow Bub3:BubR1/Mad3 to bind Cdc20 and the APC/C as shown in Figure 6C . The diagram in Figure 6C clarifies that the formation of the MCC might not require external energy sources , but it requires catalysis . In the absence of catalysts , a checkpoint defect will arise due to the rapid APC/C-mediated destruction of cyclin B and securin , which will drive cells out of mitosis . This model provides a molecular network to the hypothesis that kinetochores act as catalysts in the formation of an anaphase inhibitor [46] . The autocatalytic loop supposedly active in the cytoplasm could have an important role to guarantee a robust SAC with a small number of unattached kinetochores . In conclusion , we provided evidence that catalysis drives the formation of Cdc20:Mad2 complexes as initially suggested by the Mad2-template model . We found that catalysis not only is possible , but is indeed required for an efficient and functional checkpoint . Indeed , our studies show that although the catalytic conversion of O-Mad2 based on its dimerization with C-Mad2 might be necessary to accelerate O-Mad2 binding to Cdc20 , it is unlikely to be sufficient . Future studies will have to concentrate on the important problem of identifying by which additional mechanisms , besides Mad2 dimerization , the interaction of Mad2 with Cdc20 can be accelerated .
Full-length Mad2wt , Mad2F141A , and p31comet were purified as described [14] with the exception of the final size-exclusion chromatography separation , which was performed on a Superdex-75 column ( GE Healthcare ) equilibrated in PBS buffer . Biotinylated Mad1:C-Mad2 was expressed and purified using an AviTag system ( Avidity ) as described [28] . The 100 μM purified Mad2wt and Mad2F141A were incubated at 4 °C for 12 h with a 10-fold molar excess of Alexa 488 C5 maleimide ( Molecular Probes ) and 2 mM tris ( 2-carboxyethyl ) phosphine ( TCEP ) in PBS . Size-exclusion chromatography on a Superdex-75 column was used to separate the unbound fraction of dye from labelled proteins . Ten-microlitre chambers ( μ-slide V; Ibidi ) were coated with neutravidin . Chambers were equilibrated with 150 μl of MilliQ water and 15 μl of neutravidin 1 mg/ml ( A-2666; Molecular Probes ) in TRIS-EDTA buffer . A total of 15 μl of 10 mg/ml BSA in MilliQ water was added step by step with 30-min incubations followed by 150 μl of MilliQ washing . After equilibration with 150 μl of PBS , 15 μl of synthetic biotin-Cdc20111−138 at a concentration of 25 μM in PBS was added to the chamber and incubated for 30 min . Finally , the chambers were washed with 150 μl of PBS . The time courses were performed by injecting Alexa-488-Mad2wt or Alexa-488-Mad2F141A at concentrations in the 1–8 μM range . Fluorescence localization on the surface was monitored on a Leica Microsystems TCS SP2 confocal microscope equipped with a 63×/1 . 40 ( OIL CS HC3PL APO ) objective lens . Imaging was controlled by Leica Confocal Software ( v . 2 . 61 ) , and acquisition was carried out with the 488-nm line of an Ar/ArKr laser . ImageJ ( National Institutes of Health , http://rsb . info . nih . gov/ij/ , 1997–2004 ) and Omogen software programs were used to calculate mean pixel intensities at the surface . Omogen is a software developed by the authors and available upon request . Purified biotinylated Mad1:C-Mad2 complex was immobilized onto the surface of the μ-slide V . The surface was then coated with neutravidin as described above . After equilibration with 150 μl of PBS , 15 μl of 0 . 250 μM biotinylated Mad1:C-Mad2 complex in PBS and 15 μl of 25 μM synthetic biotin-Cdc20 peptide111−138 in PBS were added and incubated for 30 min . The surface was then washed with 150 μl of PBS . Parameter estimation was performed with PET , a free software developed by Dr . Jason Zwolak ( Virginia Polytechnic Institute and State University ) . Numerical simulations were carried out with XPP-AUT , a free software program developed by Prof . Bard Ermentrout ( Department of Mathematics , University of Pittsburgh; http://rd . plos . org/pbio . 1000010 ) Additional details on parameter estimation are to be found in Text S1 . | Mitosis , the partition of chromosomes from a mother cell to the two daughter cells , is based on the formation of attachments between chromosomes and the mitotic spindle . Cells enter mitosis with replicated chromosomes ( sister chromatids ) that are held together by a cohesive force . Upon attachment of the sister chromatids to the mitotic spindle , the cohesive force that holds them is removed , and the sisters are parted to opposite poles of the spindle . It is essential for the long-term viability of cells that chromosomes not be lost in the process . For this purpose , cells have evolved a molecular device ( the spindle assembly checkpoint or SAC ) , which prevents loss of sister chromatid cohesion until all sister chromatids are properly attached to the mitotic spindle . An outstanding question concerns the way the SAC signal is amplified away from chromosomes that are not yet attached to the spindle . Such an amplification mechanism has been predicted on the fact that as few as a single unattached kinetochore is able to prevent sister chromatid cohesion . In this paper , we show that the properties of the SAC protein Mad2 are ideally suited to provide a mechanism of amplification to the SAC . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"mathematics",
"cell",
"biology",
"computational",
"biology"
] | 2009 | The Influence of Catalysis on Mad2 Activation Dynamics |
Wolbachia are alpha-proteobacteria known to infect arthropods , which are of interest for disease control since they have been associated with improved resistance to viral infection . Although several genomes for different strains have been sequenced , there is little knowledge regarding the relationship between this bacterium and their hosts , particularly on their dependency for survival . Motivated by the potential applications on disease control , we developed genome-scale models of four Wolbachia strains known to infect arthropods: wAlbB ( Aedes albopictus ) , wVitA ( Nasonia vitripennis ) , wMel and wMelPop ( Drosophila melanogaster ) . The obtained metabolic reconstructions exhibit a metabolism relying mainly on amino acids for energy production and biomass synthesis . A gap analysis was performed to detect metabolic candidates which could explain the endosymbiotic nature of this bacterium , finding that amino acids , requirements for ubiquinone precursors and provisioning of metabolites such as riboflavin could play a crucial role in this relationship . This work provides a systems biology perspective for studying the relationship of Wolbachia with its host and the development of new approaches for control of the spread of arboviral diseases . This approach , where metabolic gaps are key objects of study instead of just additions to complete a model , could be applied to other endosymbiotic bacteria of interest .
Wolbachia are obligate intracellular alpha-proteobacteria , member of the Rickettsiales order known to infect nematodes and arthropods by developing diverse complex interactions with their hosts , such as supplementation with vitamins , cytoplasmic incompatibility and parthenogenesis [1–4] . The nature of these interactions is influenced by the strain and organism involved and they have been reviewed extensively [4] . In some organisms , Wolbachia has shown to impart a fitness advantage to arthropod host such as better reproductive traits or improved resistance to virus infection [5] . Particularly , interactions between this endosymbiont with its hosts range from metabolite supplementation , particularly biotin [6] , riboflavin [7] , to pathogenic interactions such as those of the strain wMelPop [8] . Several studies have been developed towards the understanding of Wolbachia interactions with their hosts driven by potential applications in development of novel control strategies for the spread of arbovirus-derived diseases such as Yellow fever , Zika , Chikungunya and Dengue [9] . Genomes of several Wolbachia strains have been sequenced [8 , 10–12] and comparatively analyzed [8] in order to explain host-symbiont features of interest such as cytoplasmic incompatibility . Wolbachia are known to have a reduced genome size as a result of their adaptation to depend on other organism to their survival . Thus , it is expected that they exhibit a small and rather incomplete metabolic network , as it has been previously observed for other endosymbiotic bacteria [13 , 14] . We hypothesized that a thorough analysis of Wolbachia metabolism could be achieved by analyzing a representation of their wide metabolic capabilities given by genome-scale models . Genome-scale models ( GEMs ) have emerged as a powerful tool for studying cellular metabolism since they provide a global representation of all biochemical transformations that could be carried out by a specific organism based on their genome annotation [15] . In this work , we developed genome-scale models for four Wolbachia strains found in insect arthropods , wAlbB ( from Aedes albopictus ) [11] , wVitA ( from Nasonia vitripennis ) [10] , wMel [12] and wMelPop [8] ( both from Drosophila melanogaster ) . We hypothesized that the analysis of the metabolic gaps in the curation stage could reveal potential candidates to explain the endosymbiotic nature of Wolbachia as shared metabolites between both species .
Four strains of Wolbachia pipientis were selected based on their characteristics , such as: their ability of causing cytoplasmic incompatibility , to infect mosquito species and pathogenicity . wAlbB ( Aedes albopictus ) , wVitA ( Nasonia vitripennis ) and wMel and wMelPop ( Drosophila melanogaster ) genomes were used to generate draft genome-scale models using modelSEED [13] . Gap filling was performed considering a complete media extracellular environment , which means that the model can consume any nutrient for which a transport reaction is available to the model . This gap filling was analyzed to determine potential gene gaps which could putatively explain Wolbachia pipientis' symbiotic relationship with its hosts . We proposed a draft biomass composition for Wolbachia based on phylogenetic information and predicted pathways in the annotation process [13] . The obtained composition was modified based on DNA and amino acid composition and updated fatty acid content based on reported concentration of phospholipids in Wolbachia , Rickettsia and Escherichia coli [16–19] . Computation of the stoichiometric coefficients associated with each component in the biomass function was achieved based on the protocol proposed by Thiele et al . , ( 2010 ) [15] . Integration of the four obtained Wolbachia metabolic reconstructions was achieved using COBRApy [20] . Gene protein reaction ( GPR ) associations are modified to integrate gene identifiers of all the studied strains . The obtained gaps were confirmed using BlastP to detect top hits to unannotated gene products , which were selected and incorporated into the genome-scale model when E-values were lower than 0 . 001 and coverage was above 85% . Flux Balance Analysis ( FBA ) simulations were achieved using COBRApy [20] .
We analyzed previously reported metabolic features for the wMel strain , which has been described to rely mainly on amino acids to support their energetic requirements , having limited carbohydrate synthesis capacity and being unable to transport ATP directly from its host and to synthesize Lipid A [12 , 22] . Our findings support these affirmations , with a Wolbachia metabolic reconstruction exhibiting a complete glycolysis pathway starting from fructose 6 phosphate towards phosphoenolpyruvate and a complete TCA ( tricarboxylic acid ) cycle . Additionally , Wolbachia presents a highly conserved pentose phosphate pathway ( PPP ) for sugar nucleotide synthesis , contrary to what has been reported for close organisms such as Rickettsia [14] and an amino acid metabolism , characterized by the presence of amino dipeptidases . This suggests that dipeptide transport is required by this endosymbiont for acquiring amino acids derived from its host metabolism . The riboflavin synthesis pathway , another putative symbiosis-determining pathway , is highly conserved with one missing step that is common to all studied strains , FMN hydrolase/5-amino-6- ( 5-phospho-D-ribitylamino ) uracil phosphatase . However , this reaction has been found to be associated with the gene ribD reported to be found in wMel , wRi , wHa , and wAu [7] . Therefore , this is not an incomplete metabolic pathway in any of the studied strains . We also analyzed the biomass reaction given by modelSEED which is generated from phylogenetic information and the predicted pathways based on their genome annotation [13] . Particularly , this stoichiometric composition is derived from the Gram-negative bacteria template and the presence of predicted metabolic pathways such as ubiquinone biosynthesis , fatty acid biosynthesis , and polyamine metabolism among others . However , Wolbachia is closer to the Rickettsia genus rather than Escherichia coli , the exemplary Gram-negative species . Subsequently , the biomass composition was modified to include specific phospholipid and fatty acid composition derived from studies made in Escherichia coli and particularly in Rickettsia prowazekii [16–19] , in order to find specific network spots where the symbiotic nature of the predicted gaps in this metabolic reconstruction could be found . We performed an analysis of the gap filling process to determine candidates for the metabolic relationship between Wolbachia and its hosts ( S3 File ) . In this analysis all four metabolic reconstructions shared the same 127 gap fill reactions ( 75% ) which are presumably requirements for synthesis of biomass components that no Wolbachia strain is able to perform . The added reactions correspond approximately to 20% of the total number of reactions in the analyzed metabolic reconstructions . This high proportion as a single fact putatively explains single handedly the obligate endosymbiotic nature of Wolbachia , as it has been discussed for other organisms of similar nature [13] . Most of these reactions are associated with transport of essential metabolites for cell growth instead of de novo synthesis since internalization of external metabolites requires the addition of a lower number of reactions to the model , a feature that was preferred in our reconstruction framework ( Fig 2 ) . However , it is worth mentioning that transport reactions are only about one third of the gap reactions of the model , which means that real gaps are also present in our reconstructions that are not solved by importing metabolites into the cell . An inspection of the added reactions for wAlbB and wVitA showed that potential metabolic candidates for explaining their symbiotic relationship with their hosts are mainly associated with fatty acid , lipopolysaccharide and amino acid metabolism . Arginine , glutamine and alanine are particularly important for wAlbB and asparagine , glutamine , histidine , isoleucine and leucine for wVitA . Additionally , wAlbB requires transport of TTP , hexadecanoate and dGTP for survival while wVitA requires transport of histidine and CTP . wMel and wMelPop share over 90% of their metabolic reactions , which is expected due to their phylogenetic closeness . An analysis on these reactions showed that biosynthesis of antibiotic precursors , lipopolysaccharide biosynthesis and alanine , serine and glycine metabolism are metabolite candidates to explain Wolbachia interaction with its hosts . Wolbachia pipientis is known to be highly dependent on amino acid metabolism of its hosts [12] , which is consistent with the obtained gaps in the analyzed strains . In support of this , transport of dipeptides and amino acids has been added in the gap filling process , which can be another factor to explain the symbiotic relationship of Wolbachia with their hosts . The obtained metabolic reconstruction also includes gaps associated with lipid A synthesis , which has previously been hypothesized to be absent in wMel and wBm [22] . In fact , transformations of amino acids initially found to be gaps in this model are part of the synthesis of this precursor of lipopolysaccharide ( LPS ) synthesis . Given the phylogenetic closeness of genera Wolbachia and Rickettsia , the obtained metabolic gaps for the different Wolbachia strains were compared to the ones reported for Rickettsia [14] . An analysis showing metabolic features of Wolbachia pipientis and predicted imported metabolites derived from gap analysis of this genome-scale model ( Fig 3 ) that although both are members of the Anaplasmataceae family , the reductive genome evolution has led to the loss of different functionalities in each phylogenetic branch . Rickettsia shows depleted metabolic pathways for B vitamins , several cofactors and the pentose phosphate pathway [14] , contrary to what has been obtained in this work , where these metabolic pathways are highly conserved ( Fig 3 ) . The obtained metabolic reconstruction is able to synthesize S-adenosyl methionine ( SAM ) from methionine but includes gaps associated with SAM metabolization for heme synthesis , which was initially reported to also be missing in the nematode infecting strain wOo [23] but then experimentally found to be fully present in Wolbachia [24] . On the other hand , common metabolic gaps between Wolbachia and Rickettsia have also been identified . The obtained Wolbachia model has reactions without associated genes involved in the synthesis of isopentenyl phosphate ( IPP ) and 4-hydroxybenzoate from lanosterol ( S2 and S3 Files ) while Rickettsia is known to import IPP and 4-hydroxybenzoate so we infer that Wolbachia must also acquire these metabolites from its host . Tetrahydrofolate has also been proposed as an imported metabolite in Rickettsia , which is consistent with our findings for Wolbachia in the analysis of Wolbachia's metabolism .
Wolbachia are endosymbiotic bacteria of interest due to their interactions with their arthropod hosts , such as pathogen blocking , which includes blocking of human arboviral pathogen spread , and cytoplasmic incompatibility . In this work we present metabolic reconstruction of four Wolbachia strains known to infect arthropods , wMel , wMelPop , wVitA and wAlbB . An analysis of these metabolic reconstructions is focused on their metabolic gaps as these missing reactions may be the key interactions with its hosts that explain their obligate endosymbiont nature and the difficulty found to cultivate these bacteria in the laboratory outside their host cell . Wolbachia are known to have a metabolism that is highly dependent on amino acids for cell growth [12 , 25] . The presence of amino peptidases suggests that this endosymbiont is able to feed on peptides derived from proteolytic processes inside their host . The obtained metabolic reconstruction predicts the lack of genes associated with lipid A synthesis as it has been proposed for the strain wMel and wBm [22] . A series of confirmed gaps are associated with IPP and 4-hydroxybenzoate synthesis which are metabolic requirements for ubiquinone 8 synthesis , a highly conserved pathway in Wolbachia . Since these metabolites have been previously predicted to be imported in Rickettsia , we propose that similar mechanisms exist for their import in this alpha-proteobacteria . Regarding our proposed methodology to find metabolites linked to the endosymbiotic nature of Wolbachia , several factors are known to affect the gap filling process in genome-scale models , such as media and biomass composition . Different extracellular environments result into different sets of metabolites that could be added , as transport reactions , into the intracellular compartment of the obtained model . On the other hand , the artificial inclusion of components that are not part of the modeled organism's biomass , result in the addition of metabolic capabilities observed to be absent experimentally . In this work , gap analysis was performed based on the obtained models for four different Wolbachia strains . The media composition was assumed to be rich in nutrients due to the obligate intracellular location of Wolbachia . Additionally , the biomass function was updated to represent the experimentally determined Wolbachia phospholipid and fatty acid content . The actual existence of the obtained gaps was confirmed by searches using BlastP against non-redundant protein and translated nucleotide databases . Their essentiality was assessed based on the upgraded Wolbachia biomass composition , as well as the composition of their surrounding environment to guarantee an improved quality of the predictions made by this approach . Alternative strategies to study metabolism of endosymbionts have been published recently . Driscoll et al . , ( 2017 ) [14] reconstructed Rickettsia metabolism and predicted transported compounds based solely on observed gaps in their metabolic network without information on the requirements for cell growth given by their biomass composition . Rickettsia exhibits depleted metabolic pathways for B vitamins , and the pentose phosphate pathway contrary to our findings in Wolbachia , which presents highly conserved pathways for synthesis of vitamins and nucleotides . These differences can be associated with the endosymbiotic and rather mutualistic behavior of Wolbachia versus Rickettsia which has a pathogenic nature . Accordingly , Rickettsia presents a metabolism oriented to the import of metabolites ( SAM , ATP ) whilst Wolbachia conserves metabolic pathways for metabolic provisioning to its hosts . There is no available automatic method to distinguish between true annotation gaps , which have biological meaning and correspond to gene losses that explain the dependency of Wolbachia on its host for their survival , from annotation gaps originated from experimental errors in the process of sequencing and annotation of the DNA of the analyzed strains . Particularly , of the 1 , 174 genes found originally in the annotation process for wMel [12] , only 316 were linked to metabolic functions represented in our genome-scale model , showing that although there is valuable information regarding metabolic interactions derived from Wolbachia gene annotation , there is a considerable number of genes associated with regulatory processes that are not represented by this approach . Widely studied Wolbachia genes as the cytoplasmic incompatibility genes ( cifA , cifB ) [21] and ankyrin repeat proteins [12] are not included in genome-scale models , given that they are not associated with chemical reactions that this organism can carry out . We propose that the inclusion of information of multiple strains could help minimize this problem and provide the scientific community with a robust tool for metabolic studies of this intracellular bacteria . Additionally , this approach , where metabolic gaps are key objects of study instead of just additions to complete a model , could be expanded for other endosymbiotic bacteria of interest .
In this work we present Wolbachia genome-scale model for four strains known to infect arthropods . We propose that the metabolic gaps present in this model are key to study metabolic interactions with their hosts and could potentially influence processes that lead to cytoplasmic incompatibility and pathogen blocking . Wolbachia metabolism is characterized by the import of amino acids for energy and growth , and also particularly the import of isopentenyl phosphate and 4-hydroxybenzoate for ubiquinone 8 biosynthesis . Although , Wolbachia share several characteristics with Rickettsia , the metabolism of the analyzed Wolbachia strains presents highly conserved pathways for nucleotide synthesis and riboflavin , which can be ascribed to their mutualistic rather than parasitic behavior . We have provided a systems biology analysis framework to study Wolbachia metabolism to improve the understanding on the relationship with its host and the development of new approaches towards the control of spread of arboviral diseases . | The expansion of the geographic distribution of arthropods has led to an increase in the number of infections of Zika and Dengue . This motivates the search for new alternative approaches for disease control . Wolbachia pipientis , an obligate intracellular bacteria known to provide pathogen-blocking capabilities to its host , has been used to this end . Wolbachia-infected mosquitoes have been released into the environment as a strategy for controlling the mosquito population and hence the spread of arboviral-derived diseases . However , there is little knowledge regarding the specific interactions that occur between Wolbachia and its host , particularly those associated with its obligate intracellular nature . In this work we studied Wolbachia endosymbiosis from a systems biology perspective . By analyzing the gaps in the metabolic networks of four Wolbachia strains we were able to identify key interaction points between this intracellular bacteria and its host , which play a crucial role in their metabolic relationship . This approach focuses on what is missing to allow unveiling new information regarding the interplay between interacting metabolic networks , and it is especially useful in the case of intracellular bacteria where experimental information is scarce . This additional insight on the metabolic interactions between host and parasite hold great potential for the development of arboviral disease control strategies . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion",
"Conclusion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"rickettsia",
"wolbachia",
"invertebrate",
"genomics",
"genome",
"analysis",
"bacteria",
"bacterial",
"pathogens",
"amino",
"acid",
"metabolism",
"infectiou... | 2019 | A systems biology approach for studying Wolbachia metabolism reveals points of interaction with its host in the context of arboviral infection |
Monkeypox is a zoonotic disease endemic to central and western Africa , where it is a major public health concern . Although Monkeypox virus ( MPXV ) and monkeypox disease in humans have been well characterized , little is known about its natural history , or its maintenance in animal populations of sylvatic reservoir ( s ) . In 2003 , several species of rodents imported from Ghana were involved in a monkeypox outbreak in the United States with individuals of three African rodent genera ( Cricetomys , Graphiurus , Funisciurus ) shown to be infected with MPXV . Here , we examine the course of MPXV infection in Cricetomys gambianus ( pouched Gambian rats ) and this rodent species’ competence as a host for the virus . We obtained ten Gambian rats from an introduced colony in Grassy Key , Florida and infected eight of these via scarification with a challenge dose of 4X104 plaque forming units ( pfu ) from either of the two primary clades of MPXV: Congo Basin ( C-MPXV: n = 4 ) or West African ( W-MPXV: n = 4 ) ; an additional 2 animals served as PBS controls . Viral shedding and the effect of infection on activity and physiological aspects of the animals were measured . MPXV challenged animals had significantly higher core body temperatures , reduced activity and increased weight loss than PBS controls . Viable virus was found in samples taken from animals in both experimental groups ( C-MPXV and W-MPXV ) between 3 and 27 days post infection ( p . i . ) ( up to 1X108 pfu/ml ) , with viral DNA found until day 56 p . i . The results from this work show that Cricetomys gambianus ( and by inference , probably the closely related species , Cricetomys emini ) can be infected with MPXV and shed viable virus particles; thus suggesting that these animals may be involved in the maintenance of MPXV in wildlife mammalian populations . More research is needed to elucidate the epidemiology of MPXV and the role of Gambian rats and other species .
Currently , 10 species are known in the genus Orthopoxvirus ( OPXV ) ; 6 of them ( Ectromelia , Cowpox , Volepox , Taterapox , Monkeypox and Vaccinia virus ) have been shown to circulate in rodent species [1–5] . With the eradication of Variola virus ( the causative agent of smallpox ) , Monkeypox virus ( MPXV ) is the OPXV that is most problematic with respect to global public health concerns . Through genotyping techniques , prior studies have identified 2 distinct MPXV clades , termed West African and Congo Basin due to their geographic location [6 , 7] . Congo Basin MPXV has been shown to be more virulent within both animal models as well as humans [8–12] . Currently , the reservoir host ( s ) and the ecological parameters surrounding transmission of this zoonotic disease from native African species into local human populations is/are uncertain . Historically , MPXV was thought to have a relatively narrow range of permissive animal hosts , but subsequent outbreaks in zoological gardens and captive primate colonies expanded our knowledge of animal species that suffer acute MPXV infections [13–15] . Since the early 1970’s , field researchers have conducted ecological investigations that involved the collection of vertebrate animals living in proximity to humans suffering from monkeypox disease . During these investigations , a variety of assays were used to identify several species with serological evidence ( anti-OPXV antibodies ) of past OPXV infection , including a wide variety of mammalian taxa such as ungulates ( hoofed animals ) , non-human primates , and rodents ( including squirrels of various genera , Cricetomys , Graphiurus , and shrews to name a few ) [16–20] . However , little success was achieved in determining which of these species or groups of species was responsible for maintaining MPXV in its native range . In the Democratic Republic of Congo ( former Zaïre ) the first wild caught animal actively infected with MPXV was captured in1985 , a squirrel identified in the field as a Thomas’s rope squirrel ( Funisciurus anerythrus ) [2] . This discovery led to focused research targeting rodents as the potential “reservoir host” species of MPXV in Africa and the conclusion that squirrels were the likely reservoir and that terrestrial rodents may not play a great role in the maintenance or circulation of MPXV in Africa [19 , 21] . More recently , MPXV was isolated from a sooty mangabey ( Cercocebus atys ) that was found dead during a long-term monitoring program in Taï National Park , Cote d’Ivoire , which represents the second MPXV isolate obtained from a wild animal [22] . In 2003 , several species of imported African rodents from Ghana were involved in the introduction of MPXV into the United States and its spread into captive North American prairie dogs ( Cynomys ludovicianus ) , and subsequently to humans . Specifically , viable virus was found in the tissues of individuals of three African rodent genera ( Cricetomys , Graphiurus , Funisciurus ) and additionally , Graphiurus showed evidence of maintaining a persistent infection [23] . Unfortunately , it could not be determined at what point during the outbreak these three genera became infected; therefore it is not possible to confidently determine if any of these genera served as the index case species , only that they are susceptible to MPXV infection . Subsequent field research in Ghana Africa of wild-caught rodents revealed anti-OPXV antibodies in four genera and evidence of OPXV DNA in three genera . In only two of the genera studied ( Cricetomys and Graphiurus ) was it possible to detect anti-OPXV antibodies and OPXV DNA [20] . Members of the genus Cricetomys are native to the savannahs and rain forests of tropical Africa , they dig burrows for shelter and food storage; and can reach body lengths of >67cm and weights of >730 grams [24] . These species are commonly exploited as bushmeat [25 , 26] . A recent taxonomic revision of the genus divided the previously recognized species and identified three new species; with this , the distribution of C . gambianus is limited to the savannah region south of the Sahel from the coast of Gambia and Senegal east through Guinea , Côte d’Ivoire , Ghana , Togo , Benin , Nigeria Cameroon and Central African Republic [27] . C . gambianus were introduced to the Grassy Key , Florida due to activities related to the exotic pet trade [28] . Thus far , virtually every short-term ecological study targeting animals in MPXV endemic areas and elsewhere have had a high degree of success in collecting mammals with evidence of OPXV infection , but the potential cross reactivity of OPXV assays and the difficulty in obtaining wild-caught viremic animals has confounded searches for the true reservoir ( s ) [16 , 17 , 29] . However , based on data from the US and African outbreaks of MPXV , Cricetomys has emerged as a rodent genus of interest in the search for the sylvatic source of the disease [16 , 20 , 23] . In the current study , we conducted a laboratory challenge study to examine the effects of MPXV infection in Cricetomys gambianus to assess its capability to maintain prolonged infections with MPXV and to shed infectious virus at levels that could lead to transmission within and between other rodent species or humans . Additionally , we compared the effects of the infection with West African ( W-MPX ) and Congo Basin ( C-MPX ) clades of MPXV in this native African rodent in terms of difference in disease presentation ( if any ) as is the case with human monkeypox disease . This work complements previous studies which have similarly examined African squirrels and dormice [30 , 31] as well as studies of non-African species such as prairie dog , ground squirrels and inbred mice [11 , 32–38] as potential models for human monkeypox disease . The investigations , including the one described herein , involving African rodents represent laboratory based ecological investigations which complement the early and current field efforts of several research groups meant to elucidate elements surrounding the maintenance and ecology of MPXV in its endemic ( African ) range .
Permission was obtained from the Food and Drug Administration ( FDA ) to capture , transport and use these animals in an experimental study at the Centers for Disease Control and Prevention ( CDC ) . All animal handling followed an existing CDC Institutional Animal Care and Use Committee ( IACUC ) -approved protocol ( 1376REGRATC ) . Animals were fully anesthetized in their cage using 5% inhalant isoflourane prior to any manipulation or sampling procedures . We followed standard procedures regarding sampling animals potentially infected with viral zoonoses in field collections [39] . All animal handling followed an existing Centers for Disease Control and Prevention ( CDC ) Institutional Animal Care and Use Committee ( IACUC ) -approved protocol ( 1376REGRATC ) . Cricetomys used in this study were caught in Grassy Key , Florida; where a population of this species became established after its introduction to the area via the exotic pet industry ( originally purchased from West African populations ) described in more detail elsewhere [28] . Serum was collected from the study animals to confirm the absence of anti-OPXV antibodies via ELISA ) Described below in Serology . Two weeks prior to the start of the study , animals were completely anesthetized with 5% inhalant isoflourane and implanted with vital-view mini-mitter IP biotelemetry G2 transmitters following the manufacturer’s guidelines . The telemetry systems were set up to record the activity level of each rat as the number of position changes , and core body temperature measurements beginning one week prior to inoculation and continued at 30 minute intervals throughout the study . For analysis , the temperature and activity for each animal on each sampling day was calculated by averaging all measures ( collected every 30 minutes ) for each variable so that one average value per sampling day per animal could be compared . Additionally , averages and standard deviations for these variables were calculated for the different groups ( experimental and control groups ) per sampling day . Animals were cared for in accordance with CDC IACUC guidelines under an approved protocol ( 1376REGRATC ) . Ten Cricetomys were divided into two experimental groups of four animals each ( C-MPX and W-MPX ) and one PBS control group with two animal . Animals were individually housed in cages with wire tops and aerosol barrier lids and received fresh food and water daily . The cages were placed on a metal shelving unit inside a ( negative pressure ) “holding” Bioclean unit . Animals were inoculated by a sub-dermal “scarification” route with the appropriate virus ( C-MPX or W-MPX ) by placing 10l of viral preparation on shaved skin between the scapulae , and lightly pricking the skin 10 times with a 28 gauge needle . The control animals were sham infected using 10l of sterile PBS and housed in adjacent cages on 2 shelves within a Bioclean unit inside of a BSL-3 animal suite . The West African and Congo Basin MPXV strains used in this study were isolated during the 2003 U . S . outbreak ( MPXV-2003-044; West African ) and an outbreak in the Republic of the Congo in 2003 ( MPXV-2003-358; Congo Basin ) [7 , 40] . Both viruses had undergone two passages in African green monkey kidney cells ( BSC-40; originally purchased from ATCC and currently maintained by CDC Biologics Information and Ordering System ) prior to seed pool production . Sucrose-cushion purified preparations [41] of virus were used for animal challenges . Inocula titers were immediately re-confirmed by standard plaque assay and found to be 4X104 pfu ( in a total volume of 10l ) for both MPXV clades . On sampling days , sealed cages were transferred to a clean “processing” Bioclean unit which was closed ( under negative pressure ) and animals were fully anesthetized in their cage using 5% inhalant isoflourane prior to their manipulation . Once the animal was unconscious , it was weighed and placed on a stainless steel heating surface to maintain normal body temperature and the face was placed in a “nose-cone” to maintain complete anesthesia during sampling . Blood , oral , nasal and rectal swabs were collected on days 0 , and every third day through day 21 p . i . ( 8 samples ) . After day 21 p . i . sampling intervals were increased to one sample day per week . All animals were euthanized on day 70 p . i . , except for one experimental animal that died on day 13 p . i . . Tissue samples were collected from all animals including the following: liver , lung , heart , kidney , spleen , skin , primary lesion and mesenteric lymph node . These challenge experiments were conducted in a Biological Safety Level 3+ laboratory at the CDC in Atlanta , Georgia . All individual cages were housed in a negative pressure Bioclean unit within a BSL-3 animal suite . DNA from blood samples was extracted using the EZ-1 DNA extraction robot ( Qiagen ) from 200l of blood after one hour incubation at 55°C to inactivate virus . To recover DNA from swabs ( oral , nasal , rectal , scarification site lesion ) , 400l of PBS was added to each swab and the swab extraction tube systems ( SETS; Roche ) protocol was followed to recover a homogenate . DNA from swab samples was obtained from 100l of the swab homogenate ( after the homogenate was incubated with Proteinase K and Buffer G2 at 55°C to inactivate virus ) using EZ-1 DNA extraction robot ( Qiagen ) ; the remaining swab homogenate was used for virus isolation ( see below ) . Tissue samples were placed in disposable dounce homogenizers with 1 ml of PBS and ground thoroughly to create a slurry . Genomic DNA was obtained form 100l of tissue slurry ( after the slurry was incubated with Proteinase K and Buffer G2 at 55°C to inactivate virus ) with EZ-1 DNA extraction robot ( Qiagen ) and the remaining slurries were used for virus isolation ( see below ) . All sample processing and testing was performed under BSL-2 conditions with BSL-3 work practices . DNA samples prepared from sampled tissues , blood and swabs were tested for the presence of OPXV DNA by PCR using forward and reverse primers and probe designed to be complimentary to regions of the E9L ( DNA polymerase ) gene that are able to detect all Eurasian OPXVs , except for variola [42] . MPXV DNA ( 50fg–5pg ) was used as a standard curve to allow for DNA quantification , and six wells with water were used as negative controls . Reactions were placed in either an ABI 7900 or ABI ViiA7 real-time PCR system and subjected to the following thermal cycle parameters: 95°C for 10 minutes; then 95°C for 15 seconds and 63°C for 1 minute for 45 cycles . Representative samples that tested positive in the generic E9L OPX assay were used for an additional clade specific MPXV- real-time PCR assay to add specificity to the OPXV diagnosis . Forward and reverse primers , and probe for this assay were specific to the West African MPXV clade ( GSR_WA ) or to the Congo Basin MXPV clade ( C3L_assay ) ; PCR reactions were as described within Li et al . [43] and were run on an ABI 7500 PCR platform . Specimens testing positive for OPXV DNA by PCR were evaluated for viable virus by tissue culture propagation . Each swab or tissue sample was titrated using 10 fold dilutions of swab eluent or tissue slurry on BSC-40 cell monolayers , incubated at 36o C and 6% CO2 for 72 hours , and subsequently stained with crystal violet and formalin to reveal plaques . We used a modified ELISA assay to screen animal samples for presence of anti-OPXV immunoglobulin types A and G as described in Hutson et al . [11] . One half of each Microtitreplate ( Immulon II; Dynatech ) was coated with 0 . 01 g crude Vaccinia virus ( Dryvax grown in BSC-40 cells ) in carbonate buffer; the other half was coated with an equal volume of BSC-40 cell lysate diluted in carbonate buffer . After an overnight incubation at 4°C , 10 % formalin was applied for 10 min for inactivation . Plates were blocked for 30 min at room temperature with assay diluent [PBS , 0 . 01 M , pH 7 . 4 ( Gibco ) +0 . 05% Tween-20 , 5% dried skim milk , 2% normal goat serum and 2% BSA] followed by three PBST ( 0 . 05% Tween-20 ) washes . Pouched rat sera were diluted in assay diluent ( 1:100 ) and added to both sides of the plates . After a 1hr incubation ( 37°C ) they were washed and then a 1:30000 dilution of ImmunoPure A/G conjugate ( Pierce ) was added to the plates . After incubation ( 1h at 37°C ) and wash , peroxidase substrate was added followed by addition of stop solution ( 5–15 min later ) ( Kirkegaard & Perry Laboratories ) . Absorbance was read on a spectrophotometer at 450nm . Both positive and negative human anti-vaccinia sera were used as assay controls . The mean value of all the BSC-40 cell lysate portion of each plate was used to generate a cut-off value ( COV ) by two standard deviations . Specimens with values above the COV were considered positive . Change of animal weight was calculated as the percent weight difference at each sample day using day 3 p . i . as the base value . We compared core body temperature , weight change and activity between groups ( C-MPX vs . Control , W-MPX vs . Control and C-MPX vs . W-MPX ) via the Wilcoxon rank-sum test for each sampling day . Additionally , group means of core body temperature , weight loss and activity level were calculated per sample day and a paired comparison between groups was performed using the Wilcoxon signed-rank test to assess differences in these variables throughout the study . Maximum levels of viable virus and average anti-OPXV antibody were obtained from samples collected from animals of the same group and compared using the Wilcoxon rank-sum test to evaluate differences in viable virus shedding between MPXV strains and type of sample ( oral , nasal , rectal , primary lesion and secondary lesion swabs ) . All statistical analyses were performed using the MASS package [44] in R 2 . 15 . 1 [45] . The West African and Congo Basin MPXV strains used in this study were isolated during the 2003 U . S . outbreak ( MPXV-2003-044; West African ) and an outbreak in the Republic of the Congo in 2003 ( MPXV-2003-358; Congo Basin ) , respectively; and have been fully sequenced in previous works ( Accession numbers: DQ011157 and DQ011154 ) [7 , 40] .
All animals in both experimental groups developed obvious cutaneous rash illness with more severe primary lesions at the inoculation site , and less severe disseminated secondary lesions on the dorsal and ventral surfaces of the trunk as well as on the fore and hind limbs ( Fig 1 ) . The earliest onset for primary and secondary lesions was day 3 p . i . for one rat in group C-MPX; lesions for this animal began to resolve by day 9 and were fully resolved ( only hypopigmented scarring left ) by day 27 p . i . All other animals in both experimental groups had distinct primary lesions at the site of scarification by day 6 and distinguishable secondary lesions between days 9–12 p . i . Lesions began to resolve by day 15 and were fully resolved for most animals by day 27 , with the exception of one C-MPX challenged animal in which lesions did not completely resolve until day 35 p . i . Two animals in each experimental group developed tongue lesions on day 12 ( C-MPX 2 and 3 , W-MPX 1 and 3 ) , that resolved by day 18 ( C-MPX animals ) and 21 p . i . ( W-MPX ) respectively . Although the gross number of lesions was not noticeably different between viral strains , based on our observations the severity of lesion presentation ( size/appearance ) seemed more pronounced in animals from group W-MPX than animals from group C-MPX ( although this was not a quantifiable measurement ) . Additionally , animal W-MPX 4 died on day 13 although the overt illness in this animal was not noticeably different from the other animals in its experimental group; tissues samples from this animal were collected during necropsy and tested for OPXV DNA and viable virus . Viral loads in harvested tissues ranged from 3 . 1X106 ( liver ) to 2 . 4X109 ( mesenteric lymph node ) ( Table 2 ) . There was no significant difference between core body temperatures , activity levels or weight change of experimental animals compared to control animals when compared on a day by day basis ( S1 Table ) . Three animals of group C-MPX and two in group W-MPX showed an increase in temperature following viral challenge ( Fig 2 ) . Additionally , when using Wilcoxon signed-rank tests to compare daily average temperatures per group throughout the study , the C-MPX group was significantly higher ( p<0 . 001 ) than control groups , and significantly higher than the W-MPX group ( p = 0 . 003 ) . The W-MPX average temperatures were marginally significantly higher compared to the control group ( p = 0 . 04 ) . Animals C-MPX 1 and W-MPX 1 increased their activity after being challenged with MPXV; animals C-MPX 4 and W-MPX 3 showed reduced activity immediately after viral challenge but their activity increased after days 9 and 12 p . i . , respectively; the rest of the experimental animals were less active than the control animals throughout the entire study ( Fig 2 ) . Comparisons of average activity levels per day throughout the study using Wilcoxon signed-rank test were highly significant between each experimental group and the control group with reduced averages throughout the study for both clades ( C-MPX: p = 0 . 004 and W-MPX: p = 0 . 001 ) , but no significant difference was found between C-MPX and W-MPX ( p = 0 . 593 ) . Individual weights and activity were measured throughout the study . All experimental animals lost weight after MPXV challenge and started recovering weight between days 21 and 27 p . i . ( Fig 2 ) . Weight loss was statistically higher for both experimental groups compared to the control group ( p<0 . 001 for both groups ) and the difference between groups was marginally significant ( p = 0 . 041 ) with greater weight loss in the C-MPX challenged group . For most animals , the scarification site swab as well as the nasal swab , were the first samples positive for viral DNA at day 3 p . i . Interestingly for one animal within the C-MPX group , all samples were uniformly positive for viral DNA at day 3 . For all other animals , oral , rectal and secondary lesion swabs were first positive for viral DNA ranging from day 6–18 . Blood samples were only positive for viral DNA for some animals ( C-MPX 1 , 2 and W-MPX 1 , 3 , 4 ) ; for these animals detection of virus within the blood only lasted for 1–6 days . Not all blood samples were titrated for virus due to lack of adequate sample , therefore we cannot compare the data from titration of a portion of blood samples . Detection of viral DNA was still possible in some animals from a subset of samples ( oral , scarification site , nasal swabs ) out to day 56 p . i . in both experimental groups ( Table 3 ) . The length of viral shedding was similar to viral DNA results , with some animals/samples having a delay in time between viral DNA positivity and viable virus detection . Infectious viral particles from at least one sample were first obtained from one C-MPX animals ( C-MPX 1 ) and one W-MPX animals ( W-MPX 1 ) on day 3 p . i . and from all experimental animals by day 6 p . i . ( for at least one sample; Fig 3 and Fig 4 ) . Viable virus could not be detected from one rectal swab that was positive for viral DNA ( C-MPX 2 ) . Additionally , viable virus could not be detected from secondary lesion swabs from three animals that were positive for viral DNA ( C-MPX 4 , W-MPX 2 and W-MPX 4 ) . Cessation of viral shedding from all animals/samples had occurred by day 27 p . i . The mean number of shedding days was not significantly different between viral clades from the oral , nasal , lesions and rectal swabs and time of shedding in days for all samples can be seen in Table 3 . The shedding of viable virus was most prevalent in swabs of the inoculation site on day 6 p . i . ( 108 pfu/ml ) followed closely in magnitude by oral and nasal swabs 107 and 105 pfu/ml for C-MPX and W-MPX challenged animals , respectively; with peak levels seen on days 9 and 12 p . i . then decreasing steadily until the last detection on day 24 p . i . for C-MPX and day 27 p . i . for W-MPX ( Figs 3 and 4 ) . Loads of virus from rectal and scarification swabs are also depicted in Fig 3 and Fig 4 . Fig 5 summarizes the maximum and average viable virus obtained from each sample type by individual sample ( Fig 5A ) and by experimental group ( Fig 5B ) . Statistical comparison , using the Wilcoxon rank sum test , of maximum viable virus found in blood samples ( p = 0 . 4 ) , oral ( p = 1 ) , scarification site ( p = 0 . 8 ) , secondary lesion ( p = 0 . 66 ) , rectal ( p = 0 . 8 ) and nasal ( p = 0 . 67 ) swabs between experimental groups showed no significant differences . OPXV generic Ig antibodies were detected earliest in C-MPX 1 on day 6 p . i . , ( this was also the animal that first developed generalized cutaneous lesions on day 3 p . i . ) ( 1 ) . This individual also developed the highest level of antibody response of the C-MPX group with levels increasing most rapidly until day 18 p . i . and continuing to increase through day 63 p . i . ( Fig 6 ) . All other MPXV challenged animals developed a detectable antibody response by day 12 p . i . . Antibody levels varied between individual animals but differences were not significantly different between viral strains . Ig antibody levels fluctuated but increased abruptly for most animals until day 21 p . i . ; after this initial increase there was a slower but continued increase in antibody titer or plateau at study end . The W-MPX animal that died on day 13 had Ig antibody levels similar to the other animals in its experimental group at time of death .
MPXV continues to be an important human health threat , causing sporadic outbreaks within Africa as well as the potential to spread outside its endemic range , as evidenced by the 2003 US outbreak . Identification of MPXV reservoir ( s ) is important so the public can be informed of associated risks with handling and consuming those species of animals and to improve infection control measures . Additionally the search to identify a small animal model that closely resembles human monkeypox disease progression continues in order to allow evaluation of vaccines and therapeutics against systemic Orthopoxvirus infection [38] . Up until the relatively recent development of the prairie dog and CAST/Eij MPXV models , there has been a paucity of small-animal models for the study of MPXV; specifically a small-animal model that mimics systemic disease in humans including the development of the characteristic cutaneous lesions . Due to the FDA’s animal rule , having numerous animal models that can be used to test efficacy of therapeutics and anti-virals against a MPXV challenge , especially at time of rash onset , would be beneficial . Through our experimental infection of Cricetomys , we have shown that these animals may well serve as a natural reservoir of the virus and additionally could be utilized as a relevant animal model for the study of MPXV . With the exception of the prairie dog MPXV model , no other small animal model develops the cutaneous lesions that characterize human disease . Through the current study , we have shown that Cricetomys develop these skin lesions after MPXV infection and therefore may be utilized for the study of therapeutics at time of rash onset . The time until secondary lesions developed after animal inoculation was 9–12 days with the exception of one animal which developed lesions at 3 days p . i . The 9–12 day incubation before lesion onset is similar to the prairie dog MPXV model as well as the time-course believed to occur after a human is infected . Additionally , we were able to infect the animals with a plausible amount of virus to that which likely occurs in a natural transmission setting via scarification ( to mimic a bite/scratch from an infected animal ) and infected animals shed large amounts of virus from multiple secretions . During the 2003 US outbreak , people became infected with MPXV due to bites/scratches from infected prairie dogs [40 , 46] . Although we do not have the data from human exposures in Africa , we can hypothesize that a bite or scratch from an infected animal can lead to infection in people and in other animals . Additionally , it is widely accepted that MPXV is less transmissible via an aerosol route than smallpox in people as well as in an animal model of MPXV [47 , 48]; therefore a route other than intranasal is probably the most relevant when studying MPXV transmission in potential reservoirs . Animals may be infecting other animals when fighting with each other as occurs during transmission of other viruses such as Hanta virus [49] , or perhaps when sharing the same food source and/or bedding and therefore oral excreta is shared . Follow-up studies with additional challenge routes would be worthwhile to explore differences in disease progression in MPXV infected Cricetomys due to infection route . However , the results from the current study provide evidence that Cricetomys gambianus ( and by inference , probably the closely related species , Cricetomys emini ) should be further considered as a likely MPXV reservoir species as well as potential animal model of monkeypox disease . Both MPXV strains caused an overt rash illness in Cricetomys , although the morbidity and mortality were slightly lower than that reported in non-African rodents; particularly when comparing animals challenged with C-MPX [50] . Although observational data noted more pronounced lesions in the W-MPX animals , decreased activity and weight loss in both experimental groups suggest that infection with either clade of MPXV produces a systemic infection that affects the normal behavior of the animals ( i . e . , they became more stationary and consumed less food ) , but infected Cricetomys did not become moribund during the periods when they were shedding infectious virus of either MPXV strain . Slight differences were observed in body temperatures of MPXV infected Cricetomys , with C-MPX animals having a more marked febrile period and temperature difference compared to the W-MPX groups . This could suggest a more robust immune response in the C-MPX animals; however no significant differences were seen in anti-OPXV antibody levels between groups , and the course of illness was similar . Because these were wild-caught , genetically heterogeneous animals , it is not surprising that some differences in disease presentation and mortality were observed . One animal challenged with C-MPX had an earlier disease time-course with both primary and secondary lesions evident by day 3 p . i . The antibody response was also expedited in this animal with antibody detection occurring 6 days earlier than all other animals . Viral DNA and viable virus were first detected earlier in all samples compared to the other Cricetomys; however the peak loads seen were similar . There was only one animal that succumbed to disease , an animal challenged with W-MPX . Although the disease presentation until the animal perished on day 13 was similar to other animals , necropsy results revealed extremely high loads of virus within all tissues tested ( 106–109 p . f . u . /g ) . It is possible that the animal suffered from multi-organ failure and resulting death due to these high loads of virus . It is interesting that only an animal challenged with W-MPX succumbed to disease , as it is the less virulent clade within people as well as other animal models . We have previously genotyped the Cricetomys used within this study and shown that these animals originated from West Africa [Mauldin et al . manuscript in draft] . Generally viruses are believed to become attenuated within an animal host after circulation within that hosts’ geographic range [51 , 52] , therefore we believe that this one animal’s death was most likely not reflective of the W-MPX virulence potential within Cricetomys , but specific to this individual animal . The viral loads from infected Cricetomys are similar to that seen in MPXV infected non-African rodent species such as prairie dogs and ground squirrels in other laboratory challenge investigations , as well as during the 2003 MPX outbreak in the United States [11 , 23 , 31 , 53] . This level of virus should be more than sufficient to achieve transmission to other rodents or humans that might come into contact with the infected animals ( especially through a bite ) or their excretions . For example , the maximum amount of virus shed from the lesions , nose , and mouth of infected Cricetomys ( 1 . 05X108 , 9 . 63X104 and 3 . 8X107pfu/ml , respectively ) exceeds the amount of virus used to initially challenge these animals ( 4X104 pfu/ml ) . Interestingly , detection of viral DNA for most animals was not seen within blood samples until day 12 p . i . It is possible that our sample collection ( i . e . amount of blood taken and days at which blood was collected ) was not adequate for viral detection in the blood sample . Therefore we may have missed the primary/transient viremia that is believed to occur in humans and other animal models [54] [Hutson et al . 2015 Manuscript accepted BioMed Research International] . The shedding of virus detected in rectal swabs was variable but consistently lower and for a shorter period of time with respect to the levels and duration seen in other samples . Viable virus was recovered in lower titers ( 6 . 72X101-6 . 19X104 pfu/ml ) from the rectal swabs primarily between days 9 and 15 p . i . , but out to day 24 p . i . for one W-MPX individual . It is important to note that these were not fecal samples , but were swabs of the rectum; thus , although this suggests that virus could be found in the fecal pellets , it is not possible to determine how much infectious virus would actually be shed in such a manner . The detection of viral DNA unsurprisingly parallels the levels of viable virus detected in all of the samples . However viral DNA was detectable for a longer period of time and lasted until day 56 p . i . for both viral strains , compared to the last viable virus being detected at day 27 p . i . . These later viral DNA positive samples are presumably the result of sheared non-infectious viral fragments which persist beyond the infectious ( viral shedding ) stages of illness . This finding highlights the necessity of virus culture in field and laboratory studies examining the presence of enzootic and zoonotic diseases . The results of our study are consistent with work done by collaborators who used bioluminescent imaging ( BLI ) to assess Congo Basin MPXV challenge ( intradermal and intranasal inoculation ) of Cricetomys [Falendysz et al . ; submitted for Publication PLoS Negl Trop Dis] . Although the authors found a difference in clinical disease presentation depending on the route of inoculation , all animals shed high loads of virus , similar to the findings reported herein . Additionally BLI analysis allowed the investigators to demonstrate replication of MPXV in both healthy and sick animals . Thus it was also concluded by our collaborators that Cricetomys could be a potential source of MPXV infection for humans . Members of the Genus Cricetomys including C . gambianus are found throughout Sub-Saharan Africa , including many areas that are outside of the known range of MPXV . Currently there is no known mammal species whose distribution perfectly overlaps with the distribution of MPXV , and the results presented herein along with those of other laboratory studies dealing with terrestrial rodents , suggest that the maintenance and transmission cycle of MPXV in Sub-Saharan Africa may involve multiple rodent species that can amplify and transmit the virus both within and between other mammalian species including humans . Based on past serosurveys conducted in Africa , Cricetomys have shown evidence of anti-OPXV antibodies and OPXV DNA; additionally , during the 2003 US Outbreak , Cricetomys was found to be infected with MPXV [16 , 20 , 23] . Although it is likely that there is a complex relationship between multiple rodent reservoirs in the transmission and maintenance of the virus within Africa , the data from our laboratory findings agree with the African serosurveys and suggest that Cricetomys should be considered as , at least , one potential MPXV reservoir host species that is involved in the maintenance and transmission of MPXV . | Post smallpox eradication , Monkeypox virus ( MPXV ) has emerged as the most important human health threat within the Orthopoxvirus genus . Sporadic outbreaks of monkeypox within Africa , concern over the potential of the virus to move outside of its natural range , as well as the increasing proportion of unvaccinated people now susceptible to MPXV ( due to cessation of smallpox vaccination ) , makes it important to understand how the virus is transmitted to humans within Africa . Thus far the natural reservoir ( s ) of MXPV has eluded identification; however several rodent species including African pouched rats ( Cricetomys spp . ) have been implicated as possible reservoirs . Cricetomys are often found living in close proximity to humans ( and invading homes ) and additionally serve as a food source within Africa . Therefore , it is important to utilize laboratory methods to examine the course of MPXV infection in Cricetomys and thus determine this rodent species’ competence as a host for the virus . We challenged eight animals with MPXV ( 4X104 pfu ) and assessed clinical symptoms and molecular markers of disease . Our results show Cricetomys can be infected with MPXV and shed high loads of virus via multiple routes , supporting the hypothesis that they may be involved in the maintenance and transmission of the virus within Africa . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Laboratory Investigations of African Pouched Rats (Cricetomys gambianus) as a Potential Reservoir Host Species for Monkeypox Virus |
The mitochondrial β-oxidation system is one of the central metabolic pathways of energy metabolism in mammals . Enzyme defects in this pathway cause fatty acid oxidation disorders . To elucidate the role of 2 , 4-dienoyl-CoA reductase ( DECR ) as an auxiliary enzyme in the mitochondrial β-oxidation of unsaturated fatty acids , we created a DECR–deficient mouse line . In Decr−/− mice , the mitochondrial β-oxidation of unsaturated fatty acids with double bonds is expected to halt at the level of trans-2 , cis/trans-4-dienoyl-CoA intermediates . In line with this expectation , fasted Decr−/− mice displayed increased serum acylcarnitines , especially decadienoylcarnitine , a product of the incomplete oxidation of linoleic acid ( C18:2 ) , urinary excretion of unsaturated dicarboxylic acids , and hepatic steatosis , wherein unsaturated fatty acids accumulate in liver triacylglycerols . Metabolically challenged Decr−/− mice turned on ketogenesis , but unexpectedly developed hypoglycemia . Induced expression of peroxisomal β-oxidation and microsomal ω-oxidation enzymes reflect the increased lipid load , whereas reduced mRNA levels of PGC-1α and CREB , as well as enzymes in the gluconeogenetic pathway , can contribute to stress-induced hypoglycemia . Furthermore , the thermogenic response was perturbed , as demonstrated by intolerance to acute cold exposure . This study highlights the necessity of DECR and the breakdown of unsaturated fatty acids in the transition of intermediary metabolism from the fed to the fasted state .
Fatty acids are amphipathic molecules that have indispensable roles in many cellular functions . In addition to energy storage in the form of triacylglycerols , fatty acids are involved in the synthesis of membrane lipids and in signal transduction and endocrine processes . When carbohydrates are depleted as an energy source during fasting and starvation , triacylglycerol stores are mobilized and acetyl-CoAs produced by hepatic β-oxidation of fatty acids are condensed to ketone bodies to ensure an alternative fuel source for extrahepatic tissues , such as brain , skeletal muscle , and cardiac muscle . Inherited disorders of mitochondrial β-oxidation are among the most common inborn errors of metabolism affecting infants and children . Although clinical phenotypes vary , the inability to completely utilize fatty acids during periods of increased energy requirement is common to all ß-oxidation disorders . Under normal conditions , patients are usually asymptomatic , but when challenged with short-term fasting during infectious illness , severe and even fatal phenotypes arise . Disease states can manifest as one or more of the following characteristics: liver dysfunction , hypoketotic hypoglycemia , organic aciduria , skeletal myopathy , and elevated fatty acid concentrations in the serum and tissues [1] . The presence of cis double bonds in naturally occurring ( poly− ) unsaturated fatty acids poses problems for ß-oxidation , that require a few auxiliary enzymes ( for review , see [2] ) . During degradation , double bonds in odd-numbered positions ( e . g . , oleic acid ) lead to Δ3-enoyl-CoAs , which must be isomerized by an enoyl-CoA isomerase ( ECI ) ( Figure 1 , center pathway ) . Double bonds in even-numbered positions give rise to conjugated Δ2 , Δ4-dienoyl-CoAs , which cannot be hydrated by the enoyl-CoA hydratases for thermodynamic reasons [3] . In eukaryotes , they are reduced by an NADPH-dependent 2 , 4-dienoyl-CoA reductase ( DECR ) to 3-enoyl-CoA , which is then isomerized by ECI to trans-2-enoyl-CoA , suitable for further oxidation ( Figure 1 , left pathway ) . DECR may also play a role in the degradation of fatty acids containing odd-numbered double bonds because the intermediate 2 , 5-dienoyl-CoA may be isomerized by ECI to 3 , 5-dienoyl-CoA and then converted to 2 , 4-dienoyl-CoA by a specific Δ3 , 5 , Δ2 , 4-dienoyl-CoA isomerase ( Figure 1 , right pathway ) [4] , [5] . In mammals , both mitochondria and peroxisomes contain the full set of these auxiliary enzymes for the breakdown of unsaturated fatty acids [2] . Mammalian mitochondrial isoforms of DECR have been characterized at the nucleotide [6]–[8] and protein level [9]–[13] , and the structure of the human 120 kD isoform has been recently solved [14] . Another mitochondrial isoform with a molecular mass of 60 kD has been partially purified , but it has not been characterized at the molecular level . Although the impact of ( poly ) unsaturated fatty acids on human well-being has been broadly discussed in both the media and the scientific literature , our understanding regarding the β-oxidation of unsaturated fatty enoyl-CoA esters in a physiological context is limited . Most reported cases of mitochondrial β-oxidation disorders have been related to failures in the oxidation of saturated fatty acids . However , the importance of the complete oxidation of ( poly ) unsaturated fatty acids for human health has been shown by the case of a patient with a deficiency in mitochondrial DECR activity who died at the age of four months [15] . Excluding Eci null mutant mice [16] , the available mouse models address only the breakdown of saturated fatty acids . To study the role of mitochondrial DECR in mammalian metabolism , we generated a mouse model in which Decr was disrupted by homologous recombination . Disruption of Decr leads to intolerance to fasting , as indicated by hypoglycemia , hepatic microvesicular steatosis , and an altered fatty acid pattern in the liver and serum . Contrary to many other animal models of fatty acid oxidation disorders in which hypoglycemia is associated with hypoketonemia , the absence of DECR activity did not alter the ketogenic response to fasting . A compromised response to stress was also manifested by the inability to maintain a normal body temperature during cold exposure .
A replacement vector was designed as described under Materials and Methods to replace a 0 . 5-kb region in the Decr locus by homologous recombination . This region containing the first exon of Decr was replaced with a neo selection cassette in targeted RW4 cells ( Figure 2A ) . Correct targeting was verified by Southern blotting . A 1-kb probe hybridizing to the promoter region of the Decr gene , which is not present in the replacement vector , labeled a 5 . 8-kb fragment in the wild type allele formed after BamHI digestion . Hybridization of the probe to a 4 . 7-kb fragment of the digested allele from Decr−/− mice confirmed the correct insertion of the neo cassette ( Figure 2B ) . Chimeric mice were produced by microinjecting correctly targeted RW4 cells into C57BL/6 blastocysts . Chimeric mice were backcrossed onto C57BL/6 mice to produce Decr+/− and finally Decr−/− mice . The different mouse genotypes were distinguished by PCR using genomic DNA ( Figure 2C ) . Immunoblotting of mitochondrial extracts from liver , muscle and heart with an antibody against human DECR revealed a detectable signal from wild type mice , whereas no signal could be detected for homozygous null mutant mice ( Figure 2D ) . The reductase activity ( n = 3 ) measured in liver ( muscle ) mitochondrial extract was 2 . 2±0 . 6 µmol/min per mg of protein ( 2 . 6±0 . 3 µmol/min per mg protein ) and 0 . 5±0 . 1 µmol/min per mg of protein ( trace ) for wild-type and Decr−/− mice , respectively . It is likely that the observed “residual activity” represents the activity of recently characterized mitochondrial 2-enoyl thioester reductase ( EC 1 . 3 . 1 . 38 ) [17] , which functions in mitochondrial fatty acid synthesis and can also reduce 2 , 4-hexadienoyl-CoA in vitro [18] . Under standard laboratory conditions , Decr−/− mice were indistinguishable from wild type mice . Crossbreeding of Decr +/− mice produced progeny in approximately Mendelian ratios , with no gender bias ( Table 1 ) . Both male and female mutant mice were viable and fertile . They exhibited weight gain and a life-span similar to that of wild type mice . Analysis of organ weights and histological analysis of major organs , including liver , muscle , heart , kidney , lungs , spleen and intestine , showed no differences between wild type and mutant mice . A common feature of individuals affected with inborn errors of mitochondrial fatty acid oxidation is that they are asymptomatic under normal conditions . The same phenomenon is observed in several animal models of fatty acid oxidation disorders . Clinical symptoms arise only after metabolic stress , such as prolonged physical exercise or fasting , which is often associated with infectious illness . In order to study the effect of metabolic stress on Decr−/− mice , the mice were subjected to fasting for 24 or 48 h . During and after fasting , the Decr−/− mice showed a tendency to be more passive and unresponsive compared with wild type mice . Mice were sacrificed and blood and selected organs were collected for further characterization . No differences were observed in the levels of serum alanine aminotransferase , alkaline phosphatase or glutamyl transferase between wild type and Decr−/− mice , indicating intact liver cells . Concentrations of different amino acids in the sera were also comparable ( Table S1 ) . The livers of the Decr−/− mice were markedly pale , and liver weights , when determined as a percentage of body weight , were significantly ( p<0 . 01 ) greater than that of wild type mice ( Figure 3 ) . Hematoxylin and eosin-stained histological liver sections ( Figure 4A and 4B ) obtained during the fed state showed no differences . Fasted Decr−/− mice showed normal lobular architecture when compared with wild type mice , except for the presence of hepatocytes with a foamy appearance and centralized nuclei , which are characteristics of microvesicular steatosis ( Figure 4C and 4D ) . Hepatocytes with extensive microvesicular vacuolation were mainly present in periportal and midzonal regions , whereas the majority of hepatocytes in centrilobular regions appeared normal . When Oil red O staining was performed to stain neutral lipids , the sections showed massive and homogeneously distributed micro- and macrovesicular lipid droplet formation , whereas only minor microvesicular lipid droplet accumulation was present in age-matched wild type control sections ( Figure 4E and 4F ) . These data suggested that fasting results in the accumulation of lipids in the livers of Decr−/− mice . The amounts of circulating non-esterified fatty acids ( NEFA ) were analyzed in the sera of Decr−/− and wild type mice . Under normal nutritional conditions , mean serum NEFA levels were comparable between wild type and Decr−/− mice ( 0 . 43±0 . 11 mmol/l , 0 . 52±0 . 03 mmol/l , respectively ) but after fasting , the Decr−/− mice demonstrated increased serum NEFA levels , reaching 1 . 28±0 . 12 mmol/l after 48 h compared with the wild type levels of 0 . 68±0 . 16 mmol/l ( p<0 . 001 ) ( Figure 5A ) . A common symptom associated with inherited defects of mitochondrial fatty acid oxidation is the development of hypoglycemia in response to fasting , a phenomenon also observed in several animal models of disrupted mitochondrial fatty acid oxidation [19]–[21] . This effect is considered to be caused by glycogen depletion in combination with an impaired gluconeogenic response . In order to analyze whether the defect in mitochondrial oxidation of ( poly ) unsaturated fatty acids generates a similar hypoglycemic condition , serum glucose levels were determined for wild type and Decr−/− mice after 24 h and 48 h of fasting ( Figure 5B ) . In the fed state , glucose levels were comparable ( 11 . 0±1 . 4 mmol/l for wild type and 9 . 5±0 . 6 mmol/l for Decr−/− mice ) . Twenty-four-hour fasting had no effect on the serum glucose levels of wild type mice ( 10 . 9±0 . 3 mmol/l ) , whereas a significant decrease was observed in the levels in Decr−/− mice ( 6 . 6±0 . 2 mmol/l , p<0 . 01 ) . After mice were subjected to prolonged fasting ( 48 h ) , the glucose levels in Decr−/− mice were further decreased to 2 . 3±0 . 3 mmol/l , whereas the decrease in wild type mice resulted in a glucose concentration of 5 . 9±0 . 9 mmol/l . These data revealed that Decr−/− mice have an accelerated hypoglycemic response to fasting . In order to determine whether the hypoglycemic state of the Decr−/− mice after fasting is in part due to more rapid depletion of glycogen stores , the liver and muscle glycogen concentration was measured before and after 6 h , 15 h , and 24 h of fasting ( Figure 5C and 5D ) . In the fed state , liver and muscle glycogen content was similar between wild type and Decr−/− mice . As expected , fasting resulted in a gradual depletion of glycogen stores , and no significant differences in glycogen content between wild type and Decr−/− mice were found at any observation time points . Hypoketonemia is often associated with fasting-induced hypoglycemia and defective mitochondrial fatty acid oxidation . The hypoketotic state is caused by an inability of mitochondria to offer enough acetyl-CoA moieties ( products of β-oxidation ) for ketone body production during fasting . Hypoketotic hypoglycemia is a condition that is used in the diagnosis of human genetic defects to establish a link between symptoms and a fatty acid oxidation defect . To study the ketogenic response to fasting , serum β-hydroxybutyrate levels were measured in Decr−/− and wild type mice under the fed state and after 24 h of fasting ( Figure 5E ) . In the fed state , the formation of ketone bodies was very low , as indicated by values of 0 . 13±0 . 01 mmol/l and 0 . 16±0 . 06 mmol/l for wild type and Decr−/− mice , respectively . Fasting greatly increased serum β-hydroxybutyrate values in the wild type and Decr−/− mice and comparable values of 1 . 04±0 . 14 mmol/l for wild type and 1 . 11±0 . 10 for Decr−/− mice indicated that the reduced capacity for the mitochondrial oxidation of ( poly ) unsaturated fatty acids in Decr−/− mice did not prevent a normal ketogenic response . To further analyze the role of DECR disruption in hepatic lipid accumulation , liver lipids were analyzed using positive ion mass spectrometry , as described under Materials and Methods . Under fed conditions , there were no significant differences in the amount or composition of total liver fatty acids between wild type and null mutant mice , as indicated in Figure 6A . The main fatty acid species were palmitic acid ( C16:0 ) , stearic acid ( C18:0 ) , oleic acid ( C18:1 ) , linoleic acid ( C18:2 ) , and arachidonic acid ( C20:4 ) . Figure 6C further shows that under fed conditions , the proportion of saturated fatty acids ( SAFA ) , monounsaturated fatty acids ( MUFA ) , and polyunsaturated fatty acids ( PUFA ) in total liver fatty acids was comparable between wild type and Decr−/− mice . Analysis of liver fatty acids after the mice were fasted for 24 h ( Figure 6B ) indicated that fasting had a minor effect on the lipid content of wild type liver , with an overall increase of 29% in the concentration of fatty acids . The increase in palmitic ( C16:0 ) and stearic acid ( C18:0 ) concentrations contributed to the increased SAFA concentration , whereas the increased linoleic acid ( C18:2 ) concentration contributed to the increased PUFA concentration . There were no significant differences in the amount or composition of total liver fatty acids between wild type and heterozygous mutant mice . In Decr−/− mice , however , the overall concentration of fatty acids increased by 108% after fasting . This was in agreement with the lipid accumulation observed in histological sections by Oil red O staining . The most profound changes between fasted wild type and Decr−/− mice were observed for the levels of palmitoleic acid ( C16:1 ) , oleic acid , linolenic acid ( C18:3 ) and linoleic acid , which were 2 . 5- to 3 . 8-fold higher in Decr−/− mice . In comparison to the fed state , the concentrations of MUFA and PUFA increased by 288% and 254% , respectively , in Decr−/− mice ( Figure 6B ) . The increased concentrations of MUFA were due to the increase in oleic acid and palmitoleic acid concentrations , whereas the increased PUFA concentrations were due to the increased linoleic acid and linolenic acid concentrations . The effect of fasting was most pronounced for the concentrations of linoleic and linolenic acids , which were increased 5 . 5-fold and 6 . 9-fold when compared with the fed state . The concentration of SAFA remained relatively unchanged , although an increase in the concentration of palmitic acid and a decrease in the concentration of stearic acid were observed . Acylcarnitine profiling is commonly used as a biochemical tool to diagnose various inherited metabolic disorders . For example , the initial diagnosis of long-chain fatty acid oxidation disorders is most often performed by analyzing serum or plasma acylcarnitines . Disruption of mitochondrial β-oxidation of long-chain fatty acids leads to intramitochondrial accumulation of acyl-CoA esters , which leak into the blood stream as acylcarnitines after transesterification with carnitine . This means that the serum acylcarnitine profile reflects acyl-CoA esters that accumulate intramitochondrially and pinpoints the site of metabolic block in the oxidation pathway . To study the effect of Decr gene disruption on the acylcarnitine profile and whether disruption leads to the secretion of specific acylcarnitine species , serum acylcarnitine profiles were determined for non-fasted and fasted wild type and Decr−/− mice by mass spectrometry . Under non-fasted conditions , there were no significant differences in the levels of total serum acylcarnitines between wild type and Decr−/− mice ( 263±29 nM and 240±16 nM , respectively ) . In addition , no significant differences were detected in the levels of individual acylcarnitines from C8 to C20 ( Figure 7A ) . Predominant acylcarnitines in the sera were C16 and C18:1 acylcarnitines . Fasting increased the total concentration of acylcarnitines by 2-fold in wild type mice ( 567±32 nM ) ; however , in Decr−/− mice , a markedly higher 9-fold increase was observed ( 2150±230 nM ) . Compared with wild type mice , the levels of all analyzed acylcarnitine species were highly elevated in Decr−/− mice . The increase was most distinct for the level of decadienoylcarnitine ( C10:2 ) , the concentration of which was 44-fold higher in the sera of Decr−/− mice compared with wild type controls ( Figure 7B ) . To analyze the excretion of dicarboxylic acids , another marker for mitochondrial β-oxidation dysfunction , mice were housed in metabolic cages and urine was collected for 24 h ( fed state sample ) . Subsequently , food was removed and collection was continued for another 24 h ( fasted sample ) . Dicarboxylic acid contents were monitored by means of mass spectrometry . A significant excretion of molecules with masses corresponding C7:2 , C8:2 , C10:2 , C10:3 , and C14:3 dicarboxylic acids was observed in Decr−/− mice after fasting . The excretion of these dicarboxylic acids , especially C10:2 and C14:3 , which were not detected in the urine of wild type mice , was also observed in the fed state for Decr−/− mice ( Table S2 and Figure S1 ) . C8:2 and C10:2 , the expected products of linoleic acid degradation , reached concentrations of 154 µM and 72 µM , respectively , while C14:3 , which is expected product of α linolenic acid , was found up to 16 µM . In the urine of wild type mice , their concentrations , with the exception of C8:2 , were close to or below detection limit ( approx . 5 µM ) . Adaptation to fasting is partially transmitted via altered transcription of genes encoding enzymes that function in multiple pathways . These alterations are directed by transcription factors , coactivators and corepressors that act as sensors of the nutritional status of an organism . To analyze whether disturbances observed in Decr−/− mice were accompanied by altered expression of genes encoding proteins involved in mitochondrial and extramitochondrial ( peroxisomal and microsomal ) fatty acids and carbohydrate metabolism , quantitative real-time PCR method was conducted . When the expression levels of several mitochondrial β-oxidation enzymes in the liver were compared between wild type and Decr−/− mice ( Figure 8 ) , a 2-fold increase was observed in the expression level of the rate-limiting enzyme carnitine palmitoyltransferase ( CPT-1 ) in Decr−/− mice . Expression levels of other studied mitochondrial β-oxidation enzymes , long chain acyl-CoA dehydrogenase ( LCAD ) and very long chain acyl-CoA dehydrogenase ( VLCAD ) were comparable , but were slightly increased in Decr−/− mice ( Figure 8A ) . A greater change was observed in the expression levels of genes associated with peroxisomal β-oxidation , because the expression levels of acyl-CoA oxidase ( Acox ) and peroxisomal multifunctional enzyme 1 ( MFE1 ) were 2 . 3- and 3 . 4-fold higher in Decr−/− mice , respectively ( Figure 8B ) . In addition , the expression of ECI , which is one of the auxiliary enzymes that functions in the oxidation of polyunsaturated fatty acids with double bonds in odd-numbered positions , was slightly upregulated ( 1 . 5-fold ) . A significant 2 . 1-fold increase was also observed in the expression level of cytochrome P450 IVA1 ( CYP 4A10 ) , which is a key enzyme in microsomal ω-oxidation ( Figure 8C ) . Normally , enzymes involved in fatty acid synthesis and desaturation are downregulated during fasting . The expression level of acetyl-CoA carboxylase ( Acaca ) , which catalyzes the first step in the fatty acid synthesis pathway , was lower in Decr−/− mice , although it was not significant ( Figure 8C ) . However , the messenger RNA level of the enzyme responsible for synthesis of monounsaturated fatty acids , stearoyl-CoA desaturase 1 ( SCD1 ) , was markedly lower in Decr−/− mice ( Figure 8C ) . During fasting , glucose homeostasis is maintained in part by the production and utilization of ketone bodies and in part by the production of glucose via gluconeogenesis . The hypoglycemic response to fasting prompted us to study the expression of phosphoenoylpyruvate carboxykinase ( PEPCK-C ) and glucose-6-phosphatase ( G6Pase ) , key enzymes in the gluconeogenesis and glyceroneogenesis pathway , as well as mitochondrial 3-hydroxy-3-methylglutaryl-CoA synthase ( HMGCS ) , an enzyme required for ketone body synthesis ( Figure 8D ) . We found that the levels of PEPCK and G6Pase were decreased ( 2- and 2 . 2-times , respectively ) in Decr−/− mice , whereas no differences were detected in the levels of HMGCS . Glucose homeostasis is regulated systemically by hormones such as insulin and glucagon and at the cellular level by energy status . During fasting , glucagon enhances glucose output from the liver via a PKA signal transduction pathway by activating cyclic AMP-responsive element binding protein ( CREB ) , which in turn activates the expression of PPARγ coactivator-1α ( PGC-1α ) [22] . PGC-1α has been ascribed a central role in controlling the transcription of genes involved in major metabolic pathways in the liver ( mitochondrial biogenesis , fatty acid catabolism , oxidative phosphorylation , and mitochondrial biogenesis ) through the coactivation of several nuclear receptors and other transcription factors [23] . During fasting , increased PGC-1α levels in the liver induce gluconeogenesis by activating PEPCK and G6Pase promoters through direct interaction with hepatic nuclear factor 4α ( HNF4α ) and forkhead box transcription factor , FOXO1 . Interestingly , we observed significantly decreased expression levels of CREB and PGC-1α ( 2 . 1 and 2 . 8-times , respectively ) in Decr−/− mice compared with wild type mice after fasting ( Figure 8E ) . Although peroxisome proliferator activated receptor α ( PPARα ) has a central role in the transcriptional control of genes encoding fatty acid oxidation enzymes , further transcription factors are responsible for the regulation of other metabolic pathways ( e . g . , sterol regulatory element-binding protein ( SREBP ) , which regulates genes involved in lipogenesis , cholesterogenesis , and glucose metabolism and carbohydrate responsive element binding protein ( chREBP ) , which mediates the transcriptional effects of glucose on glycolytic and lipogenic genes ) . Fasting produced no differences in the expression level of PPARα between wild type and Decr−/− mice; however , the expression levels of SREBP1 and chREBP in Decr−/− mice were significantly repressed ( 0 . 3 and 0 . 25 times the level observed in wild type mice after fasting ) ( Figure 8E ) . In order to determine the effects of cold stress , mice were fasted for 20 h and exposed to a cold environment ( +4°C ) for 4 h . Decr−/− mice exhibited severe cold intolerance during acute cold exposure and the experiment was terminated when body temperature decreased below 25°C . It has been shown that mice with temperatures below 25°C do not recover , and thus this body temperature can be considered terminal without using death as an end-point [24] . After exposure for 2 h , the average body temperature of Decr−/− mice ( n = 5 ) was 23 . 4°C compared with 33 . 0°C for wild type mice , and three of the five Decr−/− mice demonstrated temperatures that had declined below 25°C ( average 21 . 3°C ) . The temperatures of the two remaining Decr−/− mice continued to decline linearly and averaged 21 . 6°C after 3 h , at which time the experiment was terminated ( Figure 9 ) . Shivering was also initially present in Decr−/− mice but decreased during the experiment and was absent after 2 h , at which time mice became lethargic . In contrast , none of the 5 wild type mice succumbed to cold and , at the end of the 4 h exposure , shivering was clearly present and the average body temperature was 34 . 5°C . The severely cold intolerant phenotype of Decr−/− mice was observed only if cold exposure was preceded by fasting ( data not shown ) . As previously mentioned , Decr−/− mice showed decreased blood glucose and elevated NEFA concentrations after fasting in comparison to wild type mice . This response was more pronounced in both mouse genotypes when the fasting period was followed by acute cold exposure . When cold exposure was terminated , Decr−/− mice demonstrated glucose and NEFA concentrations of 3 . 0±0 . 2 mmol/l and 1 . 6±0 . 02 mmol/l , respectively , whereas the values obtained for wild type mice were 6 . 3±0 . 5 mmol/l and 0 . 77±0 . 08 mmol/l . To analyze the physical activity of mice during fasting and cold exposure , a LabMaster study was conducted . This study showed that during fasting wild type mice maintained a normal activity pattern , in which activity was highly enhanced during the dark period . When total activity during the 48 h fast was determined , Decr−/− mice showed significantly lower average activity counts than wild type mice; in particular , their activity was greatly reduced during the dark period ( Figure 10A ) . Average total activity counts were 158±30 counts/30 min and 293±19 counts/30 min for Decr−/− and wild type mice , respectively ( Figure 10B ) . To assess activity during cold exposure , mice were fasted for 20 h and then exposed to the cold for 2 h . At the beginning of the exposure , Decr−/− mice displayed significantly lower activity , reflecting the effect of fasting , as shown in Figure 10A and 10B . Upon continuation of cold exposure , the activity of both mouse groups decreased in a similar manner . However , at the end of the 2 h exposure period , the activity of Decr−/− mice had decreased to close to zero , whereas wild type mice maintained a reasonable amount of activity ( Figure 10C ) . Average total activity counts for wild type and Decr−/− mice during cold exposure were 878±183 counts/15 min and 317±65 counts/15 min , respectively ( Figure 10D ) . Average heat production during cold exposure was also measured , and Decr−/− mice showed a slight but significant decrease in heat production . The average heat production of wild type and Decr−/− mice was 27 . 1±1 . 2 kcal/h/kg and 23 . 2±0 . 8 kcal/h/kg , respectively ( Figure 10E ) .
The results presented herein indicate that mitochondrial 2 , 4-dienoyl-CoA reductase activity in mice is indispensable for the complete oxidation of ( poly ) unsaturated fatty acids and for adaptation to metabolic stress . Decr−/− mice exhibited hypoglycemia during fasting , with concomitant accumulation of metabolites of unsaturated fatty acids in the liver , sera , and urine . Furthermore , a predisposition to cold intolerance and a reduction in diurnal activity during metabolic stress were observed . Insufficient adaptation to metabolic stress in Decr−/− mice is exemplified by the development of microvesicular hepatic steatosis after as little as 24 h of fasting . However , levels of serum alanine aminotransferase , alkaline phosphatase , and glutamyl transferase were similar in both Decr−/− and wild type animals , suggesting that null mutant mice did not develop liver cell membrane injury during the observation period . In the Decr−/− mice , analysis of fatty acid composition in total liver lipids revealed a specific accumulation of mono- and polyunsaturated fatty acids , with oleic and linoleic acids being the dominant species . This can be explained in terms of their impaired β-oxidation , resulting in their channeling toward esterification and leading ultimately to hepatic steatosis . The fact that saturated fatty acids , such as palmitic and stearic acids , which , together with mono- and polyunsaturated fatty acids ( especially oleic and linoleic acid ) , are the main components of triacylglycerols in adipose tissue [25] , did not accumulate suggest that they were effectively metabolized . The proceeding β-oxidation of saturated fatty acids explains the carbon source for ketogenesis , which was found to be similar in wild type and Decr−/− mice . A prominent feature of Decr−/− mice is the hypoglycemic response to fasting , which seems not to be related to differences in glycogen or amino acid metabolism , but is associated with altered transcriptional control mechanisms in the gluconeogenesis pathway . Indeed , there were no significant differences in liver and muscle glycogen content in wild type and Decr−/− mice , either in the fed or in the fasted state . This finding suggested that the observed hypoglycemia was not due to a failure in glycogen metabolism . The amino acid profile and levels in the sera of wild type and Decr−/− mice after fasting were similar ( Table S1 ) giving no metabolic implications of specific pathological states in Decr−/− mice . This indicates that the availability of substrates for gluconeogenesis , in terms of gluconeogenic amino acids , was not the limiting factor when hypoglycemia developed in Decr−/− mice . Furthermore , the levels of glucagon and insulin in the fed state and during fasting showed no marked differences between wild type and Decr−/− mice , indicating that adaptation to fasting was not affected at the hormonal level . PEPCK and G6Pase are regarded , under in vivo conditions , as unidirectional enzymes that , among other factors , control the gluconeogenetic flux and show increased activity in the fasted state . The detected low PEPCK and G6Pase mRNA levels can explain , in part , the hypoglycemia observed in Decr−/− mice via contributing to the decreased flux through the gluconeogenetic pathway in the liver . Decreased mRNA levels of coactivator PGC-1α and its inducer CREB , which together drive the expression of PEPCK and G6Pase via HNF4α and FOXO1 , suggest that signaling pathways leading to activation of gluconeogenesis during fasting are compromised in Decr−/− mice . However , which of the component ( s ) upstream of PGC-1α or CREB , especially TORC2 ( transducer of regulated CREB activity 2 ) , energy sensing kinase AMPK , and salt-inducible kinase ( SIK ) , are affected and whether this effect is mediated by certain accumulated PUFA species will be evaluated in further studies ( Figure 11 ) . Analysis of the expression levels , activity and phosphorylation status of these factors in Decr−/− mice could elucidate the regulatory link between gluconeogenesis and the disrupted breakdown of unsaturated fatty acids . The expression of several genes , including CPT-1 , Acox , and CYP4A1 , has been shown to be upregulated by polyunsaturated fatty acids [26] , [27] . In addition , unsaturated fatty acids were previously shown to impact different transcription factors that play a pivotal role in lipid and carbohydrate metabolism . Thus , the accumulated unsaturated fatty acids and their derivatives in Decr−/− mice can act as regulators of gene expression by functioning as ligands for nuclear receptors , such as PPARα , which is the major PPAR subtype present in hepatocytes and is involved in regulating genes involved in lipid and carbohydrate metabolism [28] , [29] . In accordance , the expression of the PPARα target genes CPT-1 , Acox , and CYP4A1 was markedly higher after fasting in Decr−/− mice compared with wild type mice . Unsaturated fatty acids can also affect gene expression via SREBP1 , a factor considered a key regulator of triacylglycerol and fatty acid synthesis [30] . Although it is believed that they function mainly by inhibiting the maturation of lipogenic SREBP1 [31] , [32] , they have been shown to decrease hepatic SREBP1 mRNA [33] . In line with these observations , the mRNA level of SREBP1 was reduced 3-times in Decr−/− compared with wild type mice after fasting , and the levels of SCD1 and Acaca mRNAs , which are mediated in part by SREBP1 , were decreased as well . PUFA has recently been shown to have a suppressive effect on lipogenic and glycolytic gene expression through chREBP [34] . Namely , PUFA ablates chREBP translocation from the cytosol to nucleus and accelerates chREBP mRNA decay [35] , [36] . Of note , the Decr−/− mice displayed significantly decreased ( 4-times lower ) levels of chREBP mRNA compared with wild type animals after fasting , likely reflecting the increased chREBP mRNA decay mediated by accumulated PUFA . In our experimental setting of acute cold exposure ( 4 h ) without prior acclimatization , it is unlikely that nonshivering thermogenesis plays any major role in maintaining mouse body temperature . The decreased heat production observed in Decr−/− mice in the fasted state can be explained by reduced shivering [37] . Reduced shivering can in turn be explained by the reliance of fast-twitching muscle fibers on glucose , which is depleted in Decr−/− mice under these conditions . Wild type mice display a diurnal activity pattern with higher physical activity during the dark phase , and this pattern is preserved during fasting [38] . This increased dark phase activity , however , was greatly diminished in fasted Decr−/− mice . Of note , no differences in thermoregulation were observed between wild type and Decr−/− mice in the fed state . Because both animal groups generated ketone bodies , which can be utilized as fuel in muscle tissue , the link between simultaneous hypoglycemia and the incapability to maintain body temperature in Decr−/− mice remains an intriguing open question . By inspecting the acylcarnitine profile of the sera of Decr−/− mice , one can deduce which paths are used by unsaturated fatty acids during their degradation . The mitochondrial β-oxidation of unsaturated fatty acids with preexisting double bonds at even-numbered positions , such as petroselinic acid ( C18:1 , Δ12 ) , is predicted to halt in Decr−/− mice during the fourth turn of the β-oxidation cycle after cis-4-decenoyl-CoA is dehydrogenized to trans-2-cis-4-decadienoyl-CoA ( Figure 1 , left section ) . If the double bond is in an odd-numbered position , such as in oleic acid ( C18:1 , Δ9 ) , mitochondrial oxidation can proceed to completion via the ECI-dependent route ( Figure 1 , middle section ) or can be halted during the third β-oxidation cycle at the level of trans-2-cis-4-tetradecenoyl-CoA , which can be generated from the 2 , 5-tetradecenoyl-CoA intermediate ( Figure 1 , right section ) by the combined activity of ECI and DECI [39] . For polyunsaturated fatty acids having double bonds at odd- and even-numbered positions , such as in linoleic acid ( C18:2 , Δ9 , 12 ) , the lack of DECR activity results in blocking of the β-oxidation of these fatty acids either when intermediates of odd-numbered bonds are routed to the reductase-dependent pathway or when the even-numbered double bond reaches the Δ4 position during acyl chain shortening via β-oxidation . The distinctive accumulation of trans-2 , cis-4-decadienoylcarnitine ( C10:2 ) in the sera of Decr−/− mice can be derived from the incomplete oxidation of unsaturated fatty acids of the ω-6 series . Among them , linoleic acid is the most abundant species in the animal body , whereas others , such as petroselinic acid , are substantially less frequent . Because no intermediates of the reductase-dependent pathway for unsaturated fatty acids with double bonds at odd-numbered positions were observed , the data suggest that the major route for β-oxidation of these fatty acids proceeds via the isomerase-dependent pathway in vivo . If the reductase-dependent pathway were preferred , it would lead to accumulation of tetradecatrienoylcarnitine ( C14:3 , Δ2 , 4 , 8 ) and tetradecadienoylcarnitine ( C14:2 , Δ2 , 4 ) from the incomplete oxidation of linoleic and oleic acids , respectively . In line with this notion , the accumulation of C14:2 metabolites ( e . g . , intermediates of linoleic acid β-oxidation via the reductase-dependent route ) was not observed . Although total acylcarnitine concentrations in Decr−/− mice after fasting were approximately 4-fold higher compared with those of wild type , the profile of acylcarnitines showed no distinctively accumulating species other than trans-2 , cis-4-decadienoylcarnitine . This finding agrees with studies examining the breakdown of 2 , 5-octadienoyl-CoA in isolated rat liver mitochondria , namely , most of the test substrate ( 80% ) in this system was metabolized via the ECI-dependent pathway [40] . The accumulated unsaturated fatty acids observed in Decr−/− mice could also act as substrates in alternative oxidation pathways and processes . The microsomal fatty acid ω–hydroxylation can , together with peroxisomal β-oxidation pathway , provide an alternative route to prevent the accumulation of fatty acids or their derivatives in hepatocytes during times of increased lipolysis [41] . Both ω-oxidation and peroxisomal β-oxidation are induced in Decr−/− mice , as indicated by the enhanced expression of microsomal Cyp4A10 and peroxisomal Acox and MFE1 . Consequently , processing of accumulated unsaturated intermediates by microsomes and peroxisomes in Decr−/− mice can explain the observed excretion of medium chain unsaturated dicarboxylic acids into the urine . Many of the clinical characteristics observed in human patients suffering from fatty acid oxidation disorders can be reproduced in mice , as shown for mouse models of SCAD , MCAD , LCAD , and VLCAD deficiencies [19] , [20] , [42] , [43] . The stress-induced hypoglycemia observed in VLCAD-deficient mice was recently shown to be linked to impaired gluconeogenesis , but whether impairment is caused by inhibition of certain enzymes in the pathway or due to alterations at the level of transcription remains unknown [44] . The mouse model for MCAD does not suffer from hypoglycemia , although it is cold intolerant and displays lower blood glucose . A recent study indicated that severe metabolic stress leads to specific changes in carbohydrate management in MCAD-deficient mice [45] . Null mutant mouse models for defects in fatty acid breakdown frequently display more severe phenotypes than the corresponding deficiencies in humans , although this is not the case for Decr . To date , only a single clinical case presenting the DECR deficiency has been published [15] . Metabolic studies revealed abnormal plasma and urine acylcarnitine profiles , with the dominant species corresponding to decadienoylcarnitine , hypocarnitinemia and hyperlysinemia . Despite carnitine supplementation and a change in dietary fat to mainly medium-chain triacylglycerols , the patient died at the age of four months . DECR activity measured in postmortem liver and muscle samples was found to be decreased to 40% of the normal activity in liver and 17% of the normal activity found in muscle , as measured using trans-2-cis-4-decadienoyl-CoA as a substrate [15] . Consistent with the characteristics of the clinical case , the same dominant carnitine species was observed in Decr−/− mice , although hyperlysinemia was not observed . An open question , therefore , remains as to whether the primary cause of patient death was due to DECR deficiency or whether the patient suffered another disease that remained undiagnosed . One mouse model is known in which mitochondrial β-oxidation of unsaturated fatty acids is halted at the level of their cis- or trans-3-enoyl-CoA intermediates due to disruption of ECI [16] . Similar to Decr−/− mice , Eci−/− mice are asymptomatic under fed conditions but upon fasting , accumulate unsaturated acyl groups in ester lipids and develop hepatic steatosis . ECI deficiency also led to dicarboxylic aciduria , with an accumulation of medium chain unsaturated dicarboxylic acids . However , whether Eci−/− mice have a hypoglycemic response to fasting or show cold intolerance , as observed for Decr−/− mice , was not reported . In addition , production of ketone bodies was not reported in Eci−/− mice . Thus , a lack of information prevents a thorough comparison of Eci−/− and Decr−/− mice . In the present study , we examined the physiological consequences of disruption of the mitochondrial β-oxidation of unsaturated fatty acids at the level of 2 , 4-dienoyl-CoA reductase . This mouse model provides for the first time a hint that the breakdown of PUFA is essential for switching on gluconeogenesis during fasting . Decr−/− mice may serve as a model for studying the mechanism responsible for idiopathic hypoglycemia with unimpaired ketogenesis in humans . Analysis of the expression profile of selected transcription factors and cofactors revealed the involvement of CREB and PGC1α upstream of the reduced expression of PEPCK and G6Pase in Decr−/− mice . Among accumulated acylcarnitines in the sera , trans-2 , cis-4-decadienoylcarnitine ( C10:2 ) , which is a potential novel metabolic marker for screening patients with inborn errors in polyunsaturated fatty acids breakdown , was found .
The genomic clone BACM:109-18E ( from 129/SvJ strain ) corresponding to the mouse Decr locus was obtained from Genome Systems ( St Louis , MO , USA ) . A 3 . 3 kb EcoRI–HindIII fragment upstream of the first exon was cloned into a Bluescript vector modified with SalI and ClaI sites flanking the polylinker . A 4 . 4 kb SmaI–EcoRV fragment from the first intron was cloned into a Bluescript vector modified with AscI and PacI sites flanking the polylinker . For the replacement vector , SalI–ClaI and AscI–PacI cleaved fragments were ligated to the corresponding sites of the pPGKneo/TK-2 vector ( Figure 2A ) , where they flanked the PGKneo cassette . The neomycin resistance ( neo ) and thymidine kinase ( TK ) genes were used for positive and negative selection , respectively . Linearized replacement vector was electroporated into RW4 embryonic stem ( ES ) cells ( 129/SvJ , Tyrchp/Tyrcp ) that were subsequently grown under G418 and ganciclovir selection . Correctly targeted ES cell clones were identified by Southern analysis of genomic DNA , for which the BamHI restriction fragment length polymorphism created by homologous integration was identified by a 5′-probe upstream of the targeted locus ( Figure 2B ) . Germline chimeric mice were produced by microinjecting ES cells from positive clones into C57BL/6 blastocysts at the Biocenter Oulu Transgenic core facility . Genotyping was performed by PCR analysis of tail DNA samples using forward primers for the wild type allele ( 5′- TGC GTT CTT TGC TGG GGT GTC C-3′ ) and for the mutated allele ( 5′-CTC GAG AT C CAC TAG TTC TAG CC -3′ ) and a reverse primer for both alleles ( 5′-CAA ATG AAA GTT CCC TTG TGG AG-3′ ) ( Figure 2C ) . The size of the amplified products was 382 bp and 280 bp for the wild type and the mutated allele , respectively . Four- to seven-months-old mice were used in all experiments . DECR null mice were backcrossed 9 times onto the C57BL/6 background and C57BL/6 mice were used as wild type controls . Mice were housed in an animal room with a 12-hour lighting period ( 07:00–19:00 ) and given unrestricted access to water and standard chow . For fasting experiments , the mice were housed individually and food was withdrawn for 6 to 48 hours; water was provided ad libitum . Cold tolerance was tested by exposing individually housed fasted ( 20 h ) or non-fasted mice to +4°C for a maximum of 4 hours or until their body temperature dropped below 25°C . Temperature was measured from the shaved mid-dorsal body surface using a ThermoScan thermometer ( PRO 4000 , Braun , Kronberg , Germany ) , as described earlier [46] . Oxygen consumption , CO2 production , energy expenditure , determination of food and water intake , and activity ( total activity , ambulatory and fine movement and rearing ) were measured simultaneously and continuously in housing cages utilizing an indirect open circuit calorimetry system with a dual array of infrared photo beams ( LabMaster , TSE Systems , Bad Homburg , Germany ) . Before the experiments were carried out , mice were acclimated to their new environment in training cages similar to the actual experimental cages for seven days . During the experiment , mice were individually housed in Plexiglas home cages , fresh air was supplied at a constant flow of 0 . 33 l/min , and O2 consumption and CO2 production were measured and compared with room air values . Data were collected every 30 min for 72 h . The respiratory exchange ratio was calculated by dividing the volume of CO2 production ( VCO2 ) by the volume of oxygen consumption ( VO2 ) , and energy expenditure ( heat production ) was calculated using the software provided with the instrument . For fasting studies , mice were similarly acclimated to the experimental conditions and their baseline ( mice fed ad libitum ) was analyzed over 24 hours . Food was then removed at 8 am and fasting was continued for 48 hours; the mice were analyzed continuously , as described above . For the cold exposure study in the LabMaster system , mice were fasted overnight ( 20 hours ) prior to the experiments . Plexiglas cages were placed in a refrigerated cold cabinet ( Helkama , Finland ) with a controlled temperature ( average temperature of 9 . 6°C ) for 2 hours . Data were collected every 15 minutes and analyzed using Microsoft Excel and GraphPad Prism version 4 . 03 . All animals were handled in strict accordance with good animal practice and their use in the present study was approved by the University of Oulu committee of animal experimentation . When provided , the values represent means±S . E . To isolate mitochondria from heart and skeletal muscle tissues , 200–500 mg of tissue was cut into small pieces in 10 volumes ( w/v ) of isolation buffer ( 100 mM KCl , 50 mM HEPES , pH 7 . 4 , 5 mM MgCl2 , 1 mM EDTA ) . The solution was replaced with 10 volumes of isolation buffer containing 2 mg/ml of the bacterial protease Nagarse ( Sigma , St . Louis , MO , USA ) and the tissue samples were incubated on ice for 5 min . Samples were washed with 10 volumes of isolation buffer and subsequently homogenized in 10 volumes of isolation buffer containing 2 mM ATP with a motorized glass-teflon homogenizer . The suspension was centrifuged at 3000×g for 4 min and the resulting supernatant was further centrifuged at 17000×g for 10 min to pellet the mitochondria . The resulting pellet was suspended in 1 . 4 volumes of suspension buffer ( 10 mM Tris-Cl , pH 7 . 8 , 250 mM sucrose , 0 . 2 mM EDTA ) . To produce the mitochondrial homogenate from liver tissue , 500 mg of tissue was homogenized in 10 volumes of isolation buffer , followed by centrifugation as described above and resuspension in suspension buffer . To minimize peroxisomal contamination , mitochondrial purification was completed by isopycnic density gradient ultracentrifugation on a self-generating Percoll ( Sigma ) gradient , as previously described [17] . For immunoblotting , samples containing 20 µg of protein were transferred to a nitrocellulose membrane after SDS-PAGE and detected using polyclonal antibody against rat 2 , 4-dienoyl-CoA reductase [47] as the primary antibody and goat anti-rabbit IgG horseradish peroxidase conjugate ( Bio-Rad Laboratories , Hercules , CA , USA ) as the secondary antibody , followed by ECL Western Blotting Detection Reagents ( Amersham Biosciences , Piscataway , NJ , USA ) . The 2 , 4-Dienoyl-CoA reductase activity was assayed in mitochondrial extracts by spectrophotometric measurement of NADPH consumption at 22°C using 60 µM 2 , 4-hexadienoyl-CoA as substrate , as previously described [10] . Blood samples were collected from anesthetized mice by orbital bleeding in Multivette collection tubes ( Sarsted , Nümbrecht , Germany ) and serum was separated by centrifugation after 15 min . After being bled , mice were sacrificed by cervical dislocation and tissues were weighed and collected for further analysis . Serum cholesterol , triacylglycerols , albumin , alkaline phosphatase , alanine aminotransferase , glutamyl transferase , β-hydroxybutyrate and amino acids were analyzed by the clinical laboratory of the University Hospital of Oulu , Finland . Serum glucose ( Glucose , Wako Chemicals , Neuss , Germany ) and free fatty acids ( NEFA C , Wako Chemicals ) were determined by enzymatic colorimetric methods . Insulin levels were measured with the insulin ELISA kit ( Chrystal Chem Inc . , IL , USA ) using mouse insulin as a standard . Glucagon was determined using the glucagon RIA kit ( Linco Research Inc , MO , USA ) . Glycogen content was determined by the phenol–sulfuric acid method modified from Lo et al . [48] . Portions of frozen liver and muscle ( 50–90 mg ) were weighed and placed in test tubes containing 1 . 0 ml of 5 M KOH solution saturated with sodium sulfate . The tubes were placed in a boiling water bath for 30 min to obtain a homogenous solution . Tubes were cooled on ice for 5 min , and glycogen was precipitated by the addition of 1 ml of 95% ethanol and incubation on ice for 30 min . Samples were centrifuged at 840×g for 30 min , after which the supernatants were removed and the precipitates were dissolved in 3 ml of distilled water . Aliquots of the glycogen solutions , including standards , were made up to 1 ml in water . One milliliter of 5% phenol solution and 5 ml of a concentrated sulfuric acid solution were added in rapid succession to each tube . The tubes were allowed to stand for 10 min at room temperature , their contents thoroughly mixed , and the tubes were further incubated in the water bath ( 25°C ) for 10 min , followed by measurement of their absorbances at 490 nm . Mass spectral analyses were performed using an APEX II FTICR-MS equipped with an Apollo ESI source ( Bruker Daltonics , Bremen , Germany ) in the positive ion mode . To assess the level of hepatic fatty acids , frozen liver samples were homogenized in H2O ( 1∶20 , w/v ) with a Potter homogenizer . Ten microliters of homogenate was added to Eppendorf tubes in which 20 µl of 10 mM pentadecanoic acid had been evaporated at room temperature . Then , 180 µl of CH3CN and 20 µl of 5 M HCl was added to the tube and heated for 1 h at 95°C . After being cooled to room temperature , 190 µl of 1 M KOH was added and the tubes were again heated for 1 h at 95°C . After being cooled to room temperature , 100 µl of 5 M HCl was added to the tubes , and hydrolyzed fatty acids were extracted 2 times with 0 . 5 ml of hexane . Combined extracts were evaporated at room temperature under a stream of nitrogen . The residue was resuspended in 50 µl of 5% oxalylchloride in CH3CN ( v/v ) and heated for 5 min at 50°C . The solution was evaporated at room temperature under a stream of nitrogen . Then , 50 µl of 5% dimethylaminoethanol in CH3CN ( v/v ) was added and , after 5 min , the solvent was evaporated at room temperature under a stream of nitrogen . For MS measurements , the residue was resuspended in 500 µl of MeOH/H2O/Acetic acid [49 . 5/49 . 5/1; ( v/v/v ) ] and further diluted 1∶50 in the same solvent . To determine serum acylcarnitines , 10 µl of 10 µM dodecanoyl-carnitine in H2O and 100 µl of serum were added to the CHromabond C-8 matrix ( Macherey-Nagel , Düren , Germany ) in an Eppendorf tube and mixed . The supernatant was discarded after centrifugation and the residue was washed twice with 0 . 5 ml of H2O . Lipids bound to the C-8 material ( including the acyl carnitines ) were extracted twice with 0 . 4 ml of CHCl3/MeOH ( 7/2; v/v ) . Combined extracts were evaporated under a stream of nitrogen at 65°C . The C-8 material was further extracted with 0 . 6 ml of CHCl3/MeOH ( 7/2; v/v ) and the liquid phase was combined with previously evaporated extracts . After being mixed , extracts were centrifuged ( 16000×g for 1 min ) and 550 µl of the liquid phase was carefully transferred to a new tube . The combined extracts were evaporated under a stream of nitrogen at 65°C . The dried pellets ( containing lipids including acylcarnitines ) were resuspended in 100 µl of a solution containing 1 M acetylchloride in MeOH and heated to 65°C for 15 min . Neutral lipids and acyl-methyl esters were extracted twice with 0 . 5 ml of hexane . The upper phase ( containing the methyl esters ) was discarded and the lower phase ( containing the acylcarnitines ) was evaporated at room temperature under a stream of nitrogen . Shortly before the MS measurements , the residues were resuspended in 30 µl of MeOH/2%AcOH ( 50/50; v/v ) . Mass spectral data were recorded in positive ion mode by the accumulation of 256 scans at 256K resolution . Creatinine content was determined in the urine samples using the method of Popper et al . [49] . The samples were diluted with phosphate buffered saline to a creatinine concentration of 1 mM . Methanol solutions ( 1 mM ) of various dicarboxylic acids were used as standards . Each sample was measured twice: once with standards containing odd-numbered carbon atoms ( C5- , C7- , C9- , and C11-dicarboxylic acid ) and tetradecanedioic ( C14 ) acid as a reference and once with the standard mixture containing even-numbered carbon atoms ( C4- , C6- , C8- , and C10-dicarboxylic acid ) and tetradecanedioic acid as a reference . When tested separately the concentration of tetradecanedioic in the urine samples of mice was below the detection limit of the assay used , allowing the use of this acid as a reference ( see below ) . For quantitative determination of the dicarboxylic acids , 10 µl of standard solution was evaporated under a stream of nitrogen in eppendorf tubes , followed by the addition of 200 µl of the normalized urine sample . Non-carboxylic acid lipids were removed by extraction with three times 0 . 5 ml diethylether after increasing the pH with 350 µl 0 . 3 M NaOH . Carboxylic acids were extracted three times with 0 . 5 ml diethylether after the pH was decreased with 50 µl 6 M HCl . The acid extract was evaporated in a stream of nitrogen at room temperature . The residue was dissolved in 100 µl of 20 mM N , N′-carbonyldiimidazole in acetonitrile and incubated at room temperature for 1 h . Thirty microliters of a solution containing 21 µl of acetonitrile , 3 µl of acetic acid , and 6 µl of phenylethylamine was added , and the samples were incubated for 1 h at room temperature . Finally , the entire sample was evaporated under a stream of nitrogen at 60°C . To desalt the sample , 25 µl of RP-18 ( bed-vol . ) ( ICN , Eschwege , Germany ) was prewashed with 0 . 5 ml methanol , resuspended in 0 . 5 ml of 0 . 1% TFA in acetonitrile ( 9/1 , v/v ) , and added to the evaporated assay-mixture , followed by mixing and centrifugation . The supernatant was then discarded and the residue was again washed with 0 . 5 ml of 0 . 1% TFA in acetonitrile ( 9/1 , v/v ) , and extracted with 200 µl methanol . For the MS measurements , 50 µl of the extract was diluted in 50 µl of MeOH/H2O/AcOH ( 49 . 5/49 . 5/1 , v/v/v ) . Mass spectral data were recorded by accumulation of 256 scans at 256K resolution in positive mode . The intensities of the signals in the MS spectras were determined using the Xmass software ( Bruker , Bremen ) in order to compare the concentrations of dicarboxylic acid amides in the mouse urines . The intensity of the peak corresponding to the mass of tetradecanedioic acid amide was used as reference for the calculation of the concentrations of other metabolites . The intensity of the signals corresponding to the masses of the other standard substances was used as control for the similar behavior of dicarboxylic acid compounds with different chain length . For real-time quantitative PCR analysis of Decr and several other genes involved in fatty acid metabolism , cDNA was produced using a First Strand cDNA Synthesis Kit ( MBI Fermentas , Heidelberg , Germany ) from total RNA isolated from mouse liver samples with the RNeasy Mini Kit ( Qiagen , Hilden , Germany ) . Real-time quantitative PCR was performed with a 7500 Real Time PCR System ( Applied Biosystems , Foster City , CA , USA ) using fluorogenic probe-based TaqMan chemistry with Taqman Universal PCR Master Mix ( Applied Biosystems ) according to the manufacturer's instructions . Primers and 5′ FAM-labeled probes were designed using Primer Express software ( Applied Biosystems ) and the sequences available in Genbank and were purchased from Sigma-Genosys ( Sigma-Genosys , Haverhill , UK ) . For relative quantification of gene expression , the results were normalized with GAPDH as an endogenous control for each sample and analyzed using 7500 System Software ( Applied Biosystems ) . For light microscopy analysis , samples from various tissues were fixed in 4% paraformaldehyde in potassium phosphate buffer ( 100 mM , pH 7 . 4 ) , embedded in paraffin , sectioned and stained with hematoxylin and eosin . To stain the lipids , tissue samples were embedded in O . C . T . compound ( Tissue-Tek , Zoeterwoude , Netherlands ) and frozen in liquid nitrogen . Ten micrometer cryosections were cut from frozen samples with a Reichert-Jung 2800 Frigocut cryomicrotome and stained with Oil red O and hematoxylin using standard methods . | Fatty acids released from triacylglycerols or obtained from the diet serve as a main energy provider to the heart and skeletal muscle , and when carbohydrates are scarce , fatty acids provide energy for the whole organism . Inherited disorders of mitochondrial β-oxidation are among the most common inborn errors of metabolism affecting infants and children . Under normal conditions , patients are usually asymptomatic; but when challenged with metabolic stress , severe phenotypes arise . Here we describe the generation of a mouse model in which the total degradation of unsaturated fatty acids is prevented by disruption of an auxiliary enzyme of β-oxidation . Although degradation of saturated fatty acids proceeds normally , the phenotype presented here is in many ways similar to mouse models of the disrupted classical β-oxidation pathway , but with additional unique features . The null mutant mice are asymptomatic until exposed to fasting , during which they switch on ketogenesis , but simultaneously develop hypoglycemia . A number of human patients suffer from idiopathic hypoglycemia ( hypoglycemia of unknown cause ) . Our mouse model links this disease state to a specific defect in the breakdown of polyunsaturated fatty acids . Furthermore , it shows that degradation of unsaturated fatty acids is essential for balanced fatty acid and energy metabolism , as well as adaptation to metabolic stress . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/disease",
"models"
] | 2009 | Mitochondrial 2,4-dienoyl-CoA Reductase Deficiency in Mice Results in Severe Hypoglycemia with Stress Intolerance and Unimpaired Ketogenesis |
There is little knowledge of factors and mechanisms for coordinating bacterial chromosome replication and segregation . Previous studies have revealed that genes ( and their products ) that surround the origin of replication ( oriCII ) of Vibrio cholerae chromosome II ( chrII ) are critical for controlling the replication and segregation of this chromosome . rctB , which flanks one side of oriCII , encodes a protein that initiates chrII replication; rctA , which flanks the other side of oriCII , inhibits rctB activity . The chrII parAB2 operon , which is essential for chrII partitioning , is located immediately downstream of rctA . Here , we explored how rctA exerts negative control over chrII replication . Our observations suggest that RctB has at least two DNA binding domains—one for binding to oriCII and initiating replication and the other for binding to rctA and thereby inhibiting RctB's ability to initiate replication . Notably , the inhibitory effect of rctA could be alleviated by binding of ParB2 to a centromere-like parS site within rctA . Furthermore , by binding to rctA , ParB2 and RctB inversely regulate expression of the parAB2 genes . Together , our findings suggest that fluctuations in binding of the partitioning protein ParB2 and the chrII initiator RctB to rctA underlie a regulatory network controlling both oriCII firing and the production of the essential chrII partitioning proteins . Thus , by binding both RctB and ParB2 , rctA serves as a nexus for regulatory cross-talk coordinating chrII replication and segregation .
Efficient linkage of chromosome replication and chromosome segregation is necessary for all dividing cells . It is particularly important for maintaining balanced genetic content in organisms with more than a single chromosome , which includes a number of bacterial orders ( e . g . , Vibrionaceae , Photobacteriaceae [1] ) . However , there is relatively little knowledge of factors and mechanisms that link replication and segregation of bacterial chromosomes . For the gram-negative enteric pathogen Vibrio cholerae , whose genome is comprised of two circular chromosomes [2] , distinct mechanisms that control the replication and segregation of each chromosome have been described , but no mechanisms for linking or coordinating these processes have been identified . The two V . cholerae chromosomes have distinct initiator proteins that are specific for their target chromosomes . The initiator of chromosome I ( chrI ) replication is DnaA , a conserved AAA+ ATPase protein found in nearly all eubacteria [3]–[6] . V . cholerae DnaA binds and melts the origin of replication of chrI ( oriCI ) but not that of oriCII , the origin of replication of chromosome II ( chrII ) [7] . It is likely that regulation of DnaA-mediated initiation of V . cholerae chrI parallels DnaA-dependent control of replication initiation in Escherichia coli [6] , [7] . The initiator of chrII replication is RctB , a protein that is encoded near oriCII and conserved among , but restricted to , the Vibrionaceae/Photobacteriaceae ( Figure 1A ) [3] . RctB specifically binds and opens oriCII DNA in vitro , and its overexpression in V . cholerae leads to overinitiation of chrII but not chrI [4] , [7] . RctB can bind and hydrolyze ATP , despite a lack of known ATP binding motifs; however , unlike other ATPase initiator proteins , the ATP-bound form of RctB is inactive for oriCII replication [7] . RctB activity is also negatively regulated by rctA , a neighboring gene [8] . Although rctA is transcribed [3] and was originally annotated as an ORF [2] , it does not seem to encode a functional protein; instead at least one role of rctA appears to be as a DNA site for binding RctB , perhaps thereby titrating the initiator from oriCII [8] . Overall , the regulation of RctB activity and chrII replication initiation , which can be modulated by the factors noted above , by transcription within the oriCII region [8] , and by additional proteins such as Dam and SeqA [3] , [9] , is complex and incompletely understood . Although distinct proteins govern initiation of chrI and chrII replication , the replication of the two V . cholerae chromosomes is thought to be coordinated with the cell cycle , which should facilitate maintenance of genomic balance [10]–[12] . Genomic integrity is also promoted by chromosome-specific par systems , which have been implicated in the subcellular localization and/or partitioning of the respective oriC regions of each chromosome [13]–[16] . These systems consist of ParA ATPases , DNA-binding ParB proteins , and cis-acting ParB binding sites , parS ( [17] , [18] for review ) . The two V . cholerae ParB proteins ( ParB1 and ParB2 , encoded on chrI and chrII , respectively ) recognize distinct parS sequences ( parS1 and parS2 , respectively ) [15] . While the nucleotide sequence of parS1 is identical to the ‘universal’ parS sequence originally described in Bacillus subtilis [19] , the nucleotide sequence of parS2 is restricted to vibrio and photobacteria species [15] , [20] . All but one of V . cholerae's 10 consensus parS2 sites lie within chrII , and most of them are located proximal to oriCII . Interestingly , one of the parS2 sites , designated parS2-B , is located within rctA , suggesting the possibility that this site could provide a basis for coordination of the control of chrII replication and segregation . Individual parS2 sites are not essential for V . cholerae viability ( [15] and data not shown ) ; however , deletion of the chrII parAB2 locus results in loss of chrII and cell death [16] . Here we explore how RctB interacts with rctA and how rctA negatively regulates chrII replication . Our observations suggest that RctB has at least two DNA binding domains - one for binding to oriCII and the other for binding to rctA . RctB lacking its C-terminus fails to bind rctA in vitro and its replicative activity is not inhibited by rctA in vivo . Notably , the inhibitory effect of rctA on RctB could also be alleviated by binding of ParB2 to the parS2 site within rctA . Furthermore , ParB2 and RctB binding to rctA inversely alter the expression of the parAB2 genes . Together , our findings suggest that fluctuations in binding of the partitioning protein ParB2 and the chrII initiator RctB to rctA underlie a regulatory network controlling both oriCII firing as well the production of the essential partitioning proteins ParA2 and ParB2 . Thus , by binding both RctB and ParB2 , rctA serves as a nexus for regulatory cross–talk coordinating chrII replication and segregation .
Previous studies have established that plasmids harboring oriCII ( defined as the region between rctA and rctB ( [3]; see Figure 1 ) as the sole origin of replication can replicate in E . coli as long as RctB is present , and that such replication can be inhibited by the presence of rctA either in cis or in trans [8] , [21] . We utilized a similar approach to further dissect the molecular basis of rctA's inhibition of oriCII-based replication . The replication capacity of plasmids that included rctB , oriCII , and various additional linked sequences were assessed by their efficiency of transformation into heterologous ( E . coli ) host strains containing either a control vector or one that overexpressed RctB ( Figure 1B ) . Using this system , we obtained transformants within 24 hrs of introducing a plasmid that lacked rctA ( pYB289 ) , regardless of whether RctB was overexpressed ( Figure 1B[a] ) . In contrast , no transformants were detectable 24 hrs after introduction of a plasmid that contained rctA ( pYB292 ) unless the RctB expression construct was also present ( Figure 1B[b] ) , consistent with the suggestion that sequestration of RctB by rctA reduces its replicative activity [8] . Notably , after ∼48 hrs , rare transformants were obtained with pYB292 even in the absence of RctB overexpression . Most of these colonies could be re-streaked , and plasmid DNA was recovered and sequenced from sixty five transformants . Sixty of these plasmids carried mutations that fell into one of three groups: 1 ) deletions of rctA ( n = 6 ) , 2 ) substitutions in the rctB sequence that result in amino acid substitutions in RctB ( n = 34 ) , and 3 ) substitutions or deletions in rctB that result in truncations of the carboxyl terminus of RctB ( n = 20 ) ( Figure 2 and Table S1 ) . In general , strains carrying these plasmids grew at wild-type rates following the initial 24 hr lag in their detection , suggesting that the mutations within RctB did not impair its replicative capacity . Notably , none of the mutations mapped to the oriCII sequence per se , an observation that is consistent with the idea that an rctA transcript or protein does not act in trans on oriCII . In the remaining 5 cases , mutations were likely present in the host E . coli chromosome , since the purified oriCII plasmids ( which did not harbor mutations ) could be re-transformed into the DH5α strain they were isolated from ( after plasmid curing ) but not into a fresh isolate of DH5α . The prevalence ( 20 of 60 clones ) of RctB truncations among pYB292 derivatives whose replication did not require RctB overexpression ( Figure 2 , Table S1 ) suggested that the C-terminal part of RctB might be required for its interaction with , and inactivation by , rctA . However , the normal replication of plasmids containing such truncations ( predicted to remove at least 41 , and at most 159 , amino acids from the C-terminus of RctB ) indicates that truncated RctB retains the capacity to melt oriCII and initiate chrII replication . Together , these observations raise the possibility that RctB has multiple sites and/or modes for interacting with DNA . To explore this hypothesis , we compared the binding of His-tagged full length RctB and RctB ( Δ500–658 ) , which lacks 159 amino acids of the protein's C-terminus ( hereafter referred to as RctB[ΔC159] ) , to oriCII and rctA , using an electrophoretic mobility shift assay ( EMSA ) . Both proteins readily bound to oriCII , and they appear to have a similar affinity for this sequence , although RctB[ΔC159] appears to have a greater tendency to form multimeric complexes on the DNA ( Figure 3A ) . In contrast , while wild type RctB bound to the rctA probe , almost no binding of RctB[ΔC159] was detected ( Figure 3A ) . Together , these observations suggest that RctB has at least two DNA binding domains; one , which binds oriCII , is contained within RctB[1–499] and can mediate oriCII-based replication , while the other , which binds rctA , is at least partially contained within , or dependent upon , sequences within RctB[500–658] . We were unable to demonstrate binding of RctB[500–658] to rctA , suggesting that additional regions of RctB likely also contribute to rctA binding . Thus , although some sequence similarity has been noted between potential RctB target sites within oriCII and rctA [8] , our data raises the possibility that RctB actually recognizes two distinct sequences . Additionally , our data provides genetic and biochemical support for the hypothesis that RctB binding to rctA is the basis for rctA's negative influence on oriCII-based replication . Similar to other chromosomal parS sites , most parS2 sites are located proximal to oriCII on chrII [15] , [21] . One of these ( designated parS2-B ) is found within the originally annotated rctA sequence ( Figure 1A , ref . [15] ) . A parS2 site is present at a similar position relative to oriCII in the genomes of multiple other vibrio species [15] , despite an overall lack of conservation of the surrounding sequence . We hypothesized that ParB2 binding to this parS2 site might influence binding of RctB to rctA , and perhaps thereby regulate oriCII-based replication . This possibility was investigated by measuring the effect of ParB2 on the efficiency with which various oriCII-related replicons could be transformed into E . coli . Overexpression of ParB2 had a minimal effect on the transformation of pYB289 , consistent with the absence of rctA/parS2B within this construct ( Figure 1B[a] ) . However , overexpression of ParB2 caused a dramatic increase in the efficiency with which the rctA-containing plasmid pYB292 could be introduced into E . coli ( Figure 1B[b] ) . The effect of ParB2 expression was abolished when an alternate plasmid , pYB558 , in which the parS2-B site was mutated to parS2X [15] , was transformed instead ( Figure 1B[c] ) . Transformants were still obtained with pYB558 when it was introduced into a strain that overexpressed RctB but not when it was introduced into a strain containing an empty vector , suggesting that the mutation in parS2-B did not interfere with binding of RctB , and thus that the two proteins do not recognize identical sequences ( Figure 1B[b] ) . Data from the transformation assay was consistent with results from EMSAs , which revealed that ParB2 bound with high affinity to wild type rctA but not rctA containing parS2X , while RctB bound to both probes ( Figure 3A ) . Overall , these data indicate that ParB2 binding to parS2-B can mask the negative effect of rctA upon replication of oriCII-based replicons . The simplest explanation for increased transformation efficiency of pYB292 in the presence of overexpressed ParB2 is that binding of ParB2 to rctA interferes with binding of RctB to this site , and thereby makes more RctB available for replication initiation at oriCII . However , EMSA analyses did not provide direct support for this hypothesis . Instead , they indicate that RctB and ParB2 can bind simultaneously to rctA ( Figure 3B and Figure S1 ) . DNase I protection experiments confirmed that RctB can bind to rctA , but a specific region of binding was not observed ( Figure 4 ) . Instead , when ∼40–80 ng of RctB were added to the assay , several non-adjacent nucleotides that were distributed irregularly throughout the rctA sequence were protected from DNase I digestion ( Figure 4 , arrowheads ) . When higher amounts of RctB ( ∼160–640 ng ) were added , the protection of individual bands became less pronounced and much of the fragment exhibited a degree of protection , including the parS2-B site . In contrast , ParB2 protected a ∼20 bp continuous stretch of DNA around the parS2-B site ( Figure 4 ) . Inclusion of both RctB and ParB2 in the DNAse I protection assays resulted in additive protection , consistent with simultaneous binding of both proteins to rctA . Additionally , DNAse I-hypersensitive sites ( Figure 4 , arrows ) observed in the presence of ParB2 alone became protected upon inclusion of RctB in the reaction , suggesting that RctB can alter rctA structure even when ParB2 is bound . Collectively the EMSA and footprinting assays show that RctB and ParB2 can simultaneously bind to rctA . However , given the similar patterns of protection of the parS2-B region by the two proteins , it is difficult to ascertain whether ParB2 interferes with RctB's binding to this domain within rctA . In order to assess the roles of ParB2 , parS2 and rctA at more physiological levels in vivo , we generated an additional construct ( pYB404 ) for the transformation assay that contained all 6 kb of DNA from parB2 through rctB ( Figure 1A ) , and thereby enabled expression of ParB2 from its endogenous promoter . In contrast to pYB292 , pYB404 replicated in E . coli even without overexpression of RctB or ParB2 , despite the presence of rctA in this construct ( Figure 5A ) . Thus , ParB2 produced from its own promoter appears to be sufficient to overcome the negative effect of rctA on oriCII-mediated replication . However , overexpression of either RctB or ParB2 in the E . coli strain did increase the number of transformants obtained , perhaps because limiting amounts of these proteins are present when the plasmid is becoming established ( Figure 5A ) . We also assessed the influence on transformation efficiency of supplying additional copies of parS2 from one of three plasmids with different origins of replication and copy numbers: F ( 1–2 copies/cell ) , pSC101 ( ∼5 copies/cell ) and p15A ( 18∼22 copies/cell ) [22] . The number of pYB404 transformants obtained was not altered if the ectopic parS2 sequences were in the F-plasmid or even in pSC101 harboring two copies of parS2 separated by 1 . 4 kb ( pYB451 , yielding ∼10 ectopic copies of parS2 ( Figure 5B ) . However , there was a marked decrease in the efficiency of pYB404 electroporation when the ectopic parS2 sites were provided from the moderate-copy-number vector p15A ( Figure 5B , p15A ori ) . In contrast , the parS2X sequences , which do not bind ParB2 , did not alter pYB404 transformation efficiency when supplied from any of the vectors . ( Figure 5B , gray bars ) . Neither wild type nor mutant parS2 sequences altered the transformation efficiency of a vector ( pYB289 ) that lacks rctA . Thus , the presence of 10 parS2 sequences ( notably , their level within the V . cholerae genome ) is compatible with replication of a oriCII-based replicon containing rctA/parS2B . However , an increase to 20 copies ( e . g . , as should happen following chrII replication ) interferes with oriCII-based replication , presumably because ParB2 is titrated away from the parS2-B site at the origin and thus can't counteract rctA-dependent repression . This finding suggests that ParB2 makes an important contribution to controlling replication as well as partitioning of V . cholerae chrII . To further explore the contributions of rctA and parS2-B to regulation of chrII replication , we constructed ΔrctA V . cholerae ( YBB995 ) and parS2-B::parS2X V . cholerae ( YBB999 ) . These mutant strains did not have detectable growth defects compared to the wild type V . cholerae strain N16961 ( Figure S2 ) , indicating that rctA or parS2-B mediated control of oriCII is not critical for V . cholerae viability in rich media . However , quantitative PCR assays measuring the oriCII : oriCI ratio revealed that these mutations influence oriCII replication . The ΔrctA cells exhibited a higher oriCII : oriCI ratio compared to wild type cells ( Table 1 ) , in agreement with a previous report [23] . In contrast , the parS2-B::parS2X strain YBB999 had a modest but statistically significant ( p = 0 . 001 ) reduction in the oriCII : oriCI ratio compared to the wild type ( Table 1 ) . Both of these results are consistent with our findings using the heterologous host and support the model that ParB2 binding to parS2-B inhibits the negative effect that rctA exerts on RctB-mediated oriCII replication . The causes and consequences of chrII overinitiation have not been thoroughly analyzed . However , previous work revealed that extreme overinitiation of oriCII mediated by the RctB mutant RctB[R269S] resulted in a block in V . cholerae cell division , manifest as cell elongation along with a marked decrease in viability [7] . In contrast , modest overinitiation of oriCII resulting from overexpression of wild type RctB had virtually no effect on cell viability or morphology ( Figure 6A , 6C; see [4] , [7] , [23] ) . Similarly , the modest overinitiation of oriCII caused by the rctA deletion in YBB995 did not have a detectable affect on cell viability or morphology ( Figure 6; [23] ) . However , deletion of rctA sensitized V . cholerae to the deleterious effects of RctB overexpression . In the ΔrctA background , RctB overexpression reduced cell viability , particularly in M9 media ( Figure 6B ) , led to an increase in the oriCII : oriCI ratio ( Table 1 ) , and led to cell filamentation ( Figure 6D ) . In contrast , in the parS2-B::parS2X background , RctB overexpression had little discernible influence on cell division or viability ( Figure 6A , 6B ) , as might be expected given that the basal level of chrII replication initiation in this strain is even lower than in the wild type strain , whose viability was also unimpaired by RctB overexpression . Collectively , these data suggest that V . cholerae can adapt to some variability in RctB levels and availability , and that numerous regulatory processes are geared towards preventing the toxic effects of overinitiating replication of chromosome II . Additional forms of cross-talk between RctB and the parAB2 locus are evident from analyses of oriCII-region transcription , which revealed that binding of either ParB2 or RctB to rctA altered parAB2 promoter activity . As has been observed for several additional parAB systems [24]–[26] , ParB2 significantly decreased the expression of a PparAB2 - lacZ fusion ( more than 4-fold; Figure 7 ) . This repression was abolished when parS2-B was mutated to parS2X ( Figure 7 ) , strongly suggesting that ParB2 binding to parS2-B is required for autorepression of the parAB2 locus . In contrast , RctB modestly enhanced expression of parAB2 ( Figure 7 , p = 0 . 0003 ) , an effect that does not appear to depend on the parS2-B site in rctA . Thus , RctB binding to rctA may , despite initially limiting the amount of initiator protein available for replication initiation , ultimately promote replication , as such binding prompts expression of ParB2 , which can counter repression of replication . Additionally , these results suggest that cross-talk between pathways controlling replication and partitioning is bidirectional , which is likely to enhance the coordination of these two critical processes .
Collectively , our observations suggest that control of V . cholerae chrII replication and segregation is linked by a regulatory circuit that involves ∼6 kb of sequence ( and its products ) that flank oriCII and includes parAB2 , rctA , and rctB . The primary agent governing replication initiation is RctB; however , initiation can also be influenced by a previously characterized partitioning protein , ParB2 , which we now show counteracts rctA's inhibitory effect upon chrII replication . Analogously , the autoregulatory parAB2 locus is the primary determinant of chrII segregation; however , this process can also be influenced by RctB , which activates parAB2 expression . It appears likely that the cross-talk between these two systems both prevents extreme fluctuations in protein and chromosome abundance , and enables coordination of chromosome replication and partitioning . Binding of the chrII replication initiator RctB to the chrII origin and surrounding sequences appears to be more complex than was previously recognized . Our analyses indicate that RctB may in fact have multiple DNA binding modes/domains , which recognize distinct sequences . RctB lacking its C-terminus ( as many as 159 amino acids ) retained the capacity to bind to oriCII and to initiate replication at this site . However , both EMSA experiments and DNase I protection assays ( Figure S3 ) revealed that RctB[ΔC159] is unable to bind to rctA . Residues outside of the C-terminal 159 amino acids are also likely to contribute to rctA binding , although they remain to be identified . The presence of distinct DNA binding domains within the N- and C- terminal parts of RctB introduces the possibility that a single RctB can simultaneously bind to rctA and oriCII . Additional studies are needed to assess whether the binding of RctB to these two sites introduces a bend in the DNA between them . Studies to assess whether the point mutations within rctB that enable a bypass of rctA-mediated replication inhibition do so by altering binding to rctA or instead alter other aspects of RctB's activity or its affinity for oriCII are also warranted . To date , precise sequences targeted by RctB have not been identified; it has been speculated that this protein recognizes some short ( 11 and 12-mer ) repeated sequences within the origin and surrounding sequences [3] , [8] . Our footprinting analyses suggest that multiple RctB proteins are interacting with the DNA; however , the repeats do not seem to be the principal target of RctB's C-terminal DNA-binding domain , as many protected sites lie outside of the repeats . The distribution of sites within rctA that are protected from DNase I digestion by RctB is unusual , in that RctB appears to interact with multiple non-continuous bases throughout this sequence . One possible explanation for this result is that RctB binding alters the secondary structure of rctA DNA . Although the DNase I protection assays suggest that multiple RctB proteins interact with rctA , especially at high protein concentrations , EMSAs only revealed a single shifted band . The different assay conditions ( e . g . the presence of competitor DNA in the EMSAs ) may explain this apparent discrepancy . Additional analyses are needed to assess the binding sites for RctB in oriCII . Our studies confirmed previous reports that rctA inhibits replication of oriCII-based replicons . The inhibitory effect of rctA can be overcome by overexpression of RctB ( Figure 1B; [8] ) . Unexpectedly , our work revealed that rctA's effect can also be mitigated by overexpression of ParB2 , which recognizes a parS2 site ( parS2-B ) within rctA . At least three non-mutually exclusive models can explain how ParB2 abolishes rctA inhibition of replication . One possibility is that ParB2 competes with RctB in binding to rctA , resulting in more free RctB that can interact with oriCII . However , both EMSA and DNase I protection assays demonstrated that ParB2 does not block all RctB binding within rctA in vitro; at most , only the subset of RctB binding sites in parS2-B are blocked by the presence of ParB2 . However , these in vitro assays may not fully reflect binding dynamics in vivo in which binding to rctA maybe influenced by the adjacent oriCII site and by additional factors such as IHF . An alternative possibility is that RctB binding to rctA alters the secondary structure of the oriCII region in a manner that inhibits replication . ParB2 binding to parS2-B could counteract RctB-mediated remodeling of oriCII , thereby promoting replication . ParB2 might also alter the extent to which rctA is transcribed , which has also been shown to influence rctA's effectiveness as a replication inhibitor [8] . It is unlikely that the effect of ParB2 upon replication is mediated by a direct interaction between this protein and RctB , as no such interaction was detected using a bacterial two hybrid system ( Figure S4 ) . Regardless of the mechanism by which it acts , it is clear that ParB2 , previously described as a key agent mediating chrII segregation , also contributes to regulation of chrII replication , thereby enabling linkage of these cellular processes . We hypothesize that the organization of this regulatory scheme is adapted to accommodate the cell cycle . As ParB2 accumulates , perhaps to amounts that are sufficient to enable chrII segregation , the repressive effects of rctA are relieved , and initiation of chrII replication ensues . Subsequently , ParB2 is re-distributed among the newly synthesized parS2 sites , and its binding to parS2-B is reduced , enabling rctA to inactivate RctB , and thereby reducing the ability of RctB to initiate replication . Cross-talk between chrII replication and partitioning is also evident at the level of transcription . The parAB2 locus is autorepressed by parB2 , as has been observed for other parAB loci [24]–[26]; in addition , we demonstrate that parAB2 transcription is activated by RctB . The contrasting effects of these two regulators are likely to rebalance ParAB levels if their abundance becomes aberrantly elevated or reduced . Undoubtedly , additional factors and mechanisms intersect with these regulatory circuits . For example , previous studies have revealed that RctB can repress its own transcription [27] , [28] as well as the transcription of rctA [28] . Transcription of rctA has been reported to inhibit the negative influence of rctA on RctB [8] . Additional regulatory processes also contribute to control of replication initiation . For example , ATP binding inhibits RctB activity by decreasing its ability to bind oriCII [7] and the methylation status of oriCII also influences RctB binding to oriCII [9] . Given the centrality of chromosome replication and segregation to the perpetuation of the species , the existence of multiple and perhaps redundant mechanisms to increase the robustness of the control of these processes is expected . Consistent with this idea , we only observed significant impairment of V . cholerae growth when RctB was over-expressed in an rctA mutant , a condition that likely allows considerable overinitiation of chrII . Overinitiation also leads to growth impairment and cell filamentation in E . coli and Caulobacter crescentus [29]–[31] . Although coordinated control of chromosome replication and segregation makes sense to ensure proper chromosome inheritance to daughter cells , little mechanistic information linking these essential processes is available . Recent work by Murray and colleagues revealed that in B . subtilis the ParA ortholog Soj can inhibit or stimulate chromosome replication initiation via interactions with the initiator protein DnaA , while the ParB ortholog Spo0J inhibits initiation of chromosome replication by blocking Soj dimerization [32] , [33] . A similar regulatory scheme was recently described for V . cholerae chrI; Chattoraj and colleagues reported that ParA1 stimulates chrI replication and ParB1 inhibits ParA1 [34] . However , ParA2 appears to govern chrII replication initiation via a distinct mechanism that does not require it to interact with the replication initiator RctB . In contrast to findings for Soj and ParA1 , which interact with DnaA , we did not detect interaction between ParA2 and RctB using a bacterial two hybrid system . Our findings , along with previous reports , suggest that further exploration of the roles of Par systems in control of chromosome replication in diverse bacteria is warranted . Since chromosomal par genes are found in ∼70% of bacterial genomes [20] , Par proteins and parS sites may commonly exert control over chromosome replication . Finally , it will be interesting to explore whether mechanisms exist to link the replication and/or segregation of the two chromosomes in V . cholerae and other bacteria with multiple chromosomes .
Most of the plasmids used in this study are listed in Table 2 . The sites and mutations present in the rctA containing oriCII-based plasmids ( discussed in Figure 1B[b] ) are shown in Table S1 . The plasmids used for the bacterial two hybrid analysis are shown in Table S2 . A two-step strategy for construction of oriCII-based plasmids was followed . First , different segments of DNA proximal to oriCII were amplified and cloned into pYB199 , a derivative plasmid of pKD4 [35] which harbors the R6K origin and genes conferring resistance to ampicillin ( bla ) and kanamycin ( aph ) . Second , the resulting plasmids were digested with XbaI and the fragment containing the oriCII region and aph was gel-purified , self-ligated and then electroporated into E . coli DH5α . Spontaneous suppressor mutants in the rctA containing oriCII-based plasmids ( shown in Figure 2 and Table S1 ) were isolated by electroporating ∼100 ng of self-ligated DNA fragments into DH5α cells . Colonies that arose after an ∼48 hr incubation were re-streaked and then plasmid DNA was purified and sequenced . In addition , to confirm that mutation in the plasmid enabled establishment , each mutant plasmid was re-electroporated into DH5α . The mutations in parS2-B yielding parS2X ( Figure 1A ) were generated using the QuickChange XL Site Directed Mutagenesis Kit ( Stratagene ) . To construct pYB141 , pYB217 , pYB447 , and pYB448 , 15-bp double stranded DNA fragments containing either parS2-A or parS2X were inserted into the EcoRI site of the vectors ( pWKS30 and pXX705 ) . Plasmids pYB193 and pYB216 were constructed in a similar fashion using the NheI and HindIII sites in the pBAD33 vector . To construct pYB451 , the second copy of parS2-A was inserted at the AflII site of pYB447 , which is ∼1 . 4 kb away from EcoRI site where the first copy of parS2-A was inserted . Plasmid pCB192-YY , a derivative of the transcriptional fusion vector pCB192 [36] in which the EcoRI site in the 3′ end of lacZ was removed by introduction of a silent mutation ( GAATTC to GAATTT ) , was used to create transcriptional fusions to the parAB2 promoter . The parAB2 promoter region was amplified and cloned into the HindIII-EcoRI site of pCB192-YY . All the relevant DNA sequences of all plasmids used in this study were determined . The sequences of the oligonucleotides used in this study are listed in Dataset S1 . Mutations were introduced on to the V . cholerae chromosome ( ΔrctA , and parS2-B::parS2X ) via allele exchange using pCVD442-based plasmids as described [37] . V . cholerae strains used in this study are listed in Table 3 . DH5α cells harboring either pGZ119 ( vector ) , or Isopropyl-β-D-1-thiogalactopyranoside ( IPTG ) inducible rctB or parB2 , pYB285 or pYB273 respectively , were grown in LB broth containing 100 µM IPTG till mid-log phase to prepare electrocompetent cells . Similarly , DH5α cells harboring plasmid-borne parS2 sequences or control plasmids were grown until mid-log phase to prepare chemical competent cells . 40 ng of self-ligated DNA or 10 ng of pYB404 DNA were introduced into the competent cells . As a control , 10 ng of plasmid pYB190 was also introduced into competent cells in each experiment . The number of colonies obtained in the pYB190 transformations were used to normalize transformation efficiencies . Means and standard deviations were derived from 3–5 independent experiments for all plasmids tested . Assays were performed in triplicate with log phase cultures as described previously [38] . Two tailed , two-sample equal t-tests were used to compare the results from 3 independent experiments ( total 9 samples each ) for the statistical analysis . Wild type RctB-His6 , RctB[ΔC159]-His6 , and ParB2-His6 proteins were purified as previously described [15] . Sequences used for oriCII , rctA , and rctA ( parS2-B::parS2X ) ( see Figure 1A ) EMSA probes were initially cloned into the pCR-Blunt II-TOPO vector ( Invitrogen ) , yielding pYB405 , pYB406 , and pYB407 , respectively . pYB190 , a pCR-Blunt II TOPO derivative containing an irrelevant 10 bp , was used to construct the negative control probe . To prepare radio-labeled probes , appropriate DNA regions were amplified from the plasmids with universal M13 forward and reverse primers [22] , end labeled with [γ-32P] ATP with polynucleotide kinase ( New England Biolabs ) , purified from 6% DNA retardation gels ( Invitrogen ) , and ethanol precipitated . In the binding reactions , 5 , 000 cpm of probe DNA containing different concentrations of RctB and/or ParB2 in a reaction buffer of 20 mM Tris-Cl ( pH 7 . 5 ) , 1 mM EDTA , 150 mM NaCl , 12 . 5 µg/mL poly ( dI-dC ) , and 0 . 1 mg/mL BSA were incubated for 10 min at room temperature . The reactions were then electrophoresed in a 6% DNA retardation gel in 0 . 5× TAE and visualized by autoradiography . DNase I footprint assays were performed as previously described with minor modifications [39] . The rctA probe was made by PCR using 5′-32P-radiolabeled rctA 5′-FP ( CGTTTAAATAACCCACATATTCTTCGATAAGG ) and rctA 3′-FP ( ATACCTATTCGCTGGAGGAAAGATAGG ) primers on a plasmid encoding parAB2-rctA-oriCII-rctB ( pYB403 ) . The probe was purified from 6% DNA retardation gels , eluted , and ethanol precipitated . 1 , 200 , 000 cpm of probe was incubated with different amounts of RctB without and with 100 ng of ParB2 in 20 µL of 20 mM Tris-Cl pH 8 . 0 , 125 mM NaCl , 1 mM DTT for 10 min at room temperature . 0 . 1 U of DNase I ( Applied Biosystems ) was added to each reaction and incubated at room temperature for 30 sec . The digestions were quenched by the addition of 6 µL of 660 mM Tris-Cl pH 9 . 5 , 66 mM EDTA , 3 . 3% SDS and placed on ice . Samples were ethanol precipitated , resuspended in recrystalized formamide , and 20 , 000 cpm of each was run on an 8% polyacrylamide gel with 8 M urea ( National Diagnostics ) in 1× TBE . The gels were then dried and visualized by autoradiography . Genomic DNA was prepared from each strain by phenol-chloroform extraction followed by ethanol precipitation . The genomic DNA was then digested with PstI and 10 pg was used for each quantitative PCR ( qPCR ) reaction . Genomic DNA from an N16961 stationary culture was used to generate the standard curve . qPCR was performed with the StepOnePlus Real-Time PCR system ( Applied Biosciences ) using SYBR Green Master mix ( Applied Biosciences ) according to the manufacturer's protocol . The primer pairs used for oriCI and oriCII were described previously [4] . Each qPCR run was done in triplicate and the ratio was calculated from three independent experiments . V . cholerae cells harboring a plasmid borne copy of an IPTG-inducible rctB or a control vector were grown in either LB or M9-Glucose media containing 100 µM IPTG at starting OD600 of 0 . 003 . Subsequently , OD600 and CFU were monitored hourly . Growth curves shown in Figure S2 are representative of at least three independent experiments . A small aliquot of cells was removed at 4 hr , fixed with 3% paraformaldehyde , and then examined with 100× alpha-plan lens on a Zeiss Axioplan 2 microscope . | There is scant knowledge of factors and mechanisms that link bacterial chromosome replication and segregation . We investigated the mechanisms that coordinate the replication and segregation of Vibrio cholerae chromosome II ( chrII ) . Our findings suggest that control of V . cholerae chrII replication and segregation is linked by a regulatory circuit that involves sequences , including parAB2 , rctA , and rctB ( and their products ) , that flank this chromosome's origin of replication . The primary agent governing replication initiation is RctB; however , initiation can also be influenced by a previously characterized partitioning protein , ParB2 , which we now show counteracts rctA's inhibitory effect upon chrII replication . Analogously , the autoregulatory parAB2 locus is the primary determinant of chrII segregation; however , this process can also be influenced by RctB , which activates parAB2 expression by binding to rctA . Thus , our findings suggest that the cross-talk between these two systems both prevents extreme fluctuations in protein and chromosome abundance and also enables coordination of chromosome replication and partitioning . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"dna",
"replication",
"nucleic",
"acids",
"chromosome",
"biology",
"cell",
"biology",
"dna",
"biology",
"molecular",
"cell",
"biology",
"molecular",
"biology"
] | 2011 | Regulatory Cross-Talk Links Vibrio cholerae Chromosome II Replication and Segregation |
The conjugative plasmid R388 and a number of other plasmids carry an operon , stbABC , adjacent to the origin of conjugative transfer . We investigated the role of the stbA , stbB , and stbC genes . Deletion of stbA affected both conjugation and stability . It led to a 50-fold increase in R388 transfer frequency , as well as to high plasmid loss . In contrast , deletion of stbB abolished conjugation but provoked no change in plasmid stability . Deletion of stbC showed no effect , neither in conjugation nor in stability . Deletion of the entire stb operon had no effect on conjugation , which remained as in the wild-type plasmid , but led to a plasmid loss phenotype similar to that of the R388ΔstbA mutant . We concluded that StbA is required for plasmid stability and that StbA and StbB control conjugation . We next observed the intracellular positioning of R388 DNA molecules and showed that they localize as discrete foci evenly distributed in live Escherichia coli cells . Plasmid instability of the R388ΔΔstbA mutant correlated with aberrant localization of the plasmid DNA molecules as clusters , either at one cell pole , at both poles , or at the cell center . In contrast , plasmid molecules in the R388ΔΔstbB mutant were mostly excluded from the cell poles . Thus , results indicate that defects in both plasmid maintenance and transfer are a consequence of variations in the intracellular positioning of plasmid DNA . We propose that StbA and StbB constitute an atypical plasmid stabilization system that reconciles two modes of plasmid R388 physiology: a maintenance mode ( replication and segregation ) and a propagation mode ( conjugation ) . The consequences of this novel concept in plasmid physiology will be discussed .
Transmissible plasmids contribute greatly to the plasticity of bacterial genomes and to the acquisition of genetic traits by host cells through the collective carriage of adaptive genes , including antibiotic resistance and virulence genes , and through the ability to disseminate them by conjugation [1] . Horizontal gene transfer may thus increase the adaptability of bacteria to changing environmental conditions , which is dramatically exemplified by the emergence and spread of multiple antibiotic-resistance plasmids in and between potentially pathogenic bacteria . Conjugative plasmids are transmitted both vertically to daughter cells and horizontally to other strains or species . Vertical transmission requires timely controlled replication and faithful assortment ( segregation ) of sister plasmid copies to daughter cells . Segregation occurs by a range of different mechanisms including control of copy number , resolution of multimeric plasmid molecules and , in the case of most low copy number plasmids , active segregation ( partition ) . Partition ( Par ) systems ensure efficient distribution of plasmid molecules to each daughter cell during division ( for reviews: [2]–[5] ) . They are composed of a cis-acting centromere-like site and two proteins , a nucleotide-binding cytomotive protein and a centromere-binding adaptor protein . Stable inheritance requires that these proteins form a partition complex on the centromere . The par centromere locus and the Par proteins are encoded by sets of homologous genes in various plasmids , phages , and chromosomes . The mechanism of plasmid conjugation in gram negative bacteria has been well characterized ( for a review: [6] ) . The overall process is accomplished by two functional multiprotein complexes encoded by two gene clusters: the set of mobility genes ( MOB ) , involved in conjugative DNA processing , and the mating pair formation cluster ( MPF ) , encoding the nucleoprotein transport apparatus . These two systems are connected by the coupling protein ( T4CP ) . The MPF is a type IV protein secretion system and implies the assembly of a multiprotein complex at the bacterial membrane [7] . Conjugal DNA processing involves the formation of a nucleoprotein complex called relaxosome at the cognate origin of conjugative transfer ( oriT ) . The relaxase catalyses a DNA strand- and site- specific cleavage at the nic site of oriT , and remains covalently attached to the 5′ end of the cleaved single-stranded ( ss ) DNA [6] . Complementary-strand synthesis is initiated from the free 3′ end of the cleaved strand through rolling-circle replication . The relaxase-ssDNA complex interacts with the T4CP , which guides the transferred strand through the DNA transfer apparatus formed by the MPF proteins throughout the membrane and into the recipient cell [8] , [9] . Fundamental questions remain unanswered concerning the spatiotemporal coordination of the DNA substrate processing , its recruitment to conjugative pores , and the subsequent DNA translocation reactions . More relevant to this work , how conjugation is integrated with plasmid maintenance functions such as replication and segregation is still poorly understood . The low-copy number conjugative plasmid R388 is the prototype of the IncW family . It represents the smallest broad host range conjugative plasmid [10] , [11] . Its mechanism of segregation is presently unknown , and no par system was annotated in its DNA sequence [12] . R388 carries a cluster of two operons , which are transcribed divergently from the region containing oriT . One contains the MOB genes , and the other includes a cluster of three genes , stbA , stbB and stbC , whose functions have not been analyzed previously . Their relative position to oriT makes these three genes the first to enter the recipient cell during plasmid conjugation . Here , we investigated the role of the stbABC operon in plasmid R388 transfer and stability in E . coli . By using a fluorescent protein to tag plasmid molecules , we found that R388 plasmid foci , most of which contain a single copy of the plasmid , are evenly distributed within E . coli cells . In contrast , a derivative of R388 lacking stbA mislocalized as clusters at the cell poles or at the cell center , and correlated with plasmid instability . In addition , we show that StbB , a putative ATPase protein , is strictly required for R388 conjugative transfer and the occurrence of plasmid foci close to the membrane cell poles . Taken together , these results suggest that stbA and stbB constitute a balancing system that integrates plasmid conjugation with the functions that ensure the efficient vertical transmission of the plasmid .
To study the role of the three stb genes in plasmid R388 , we first analyzed the stb operon by protein sequence comparison . We found that the most conserved protein is StbB , whereas StbA is poorly conserved . Besides , StbC is an orphan protein , without significant homologs in any other system . We thus used R388 stbB gene as template to search for homology . StbB homologs were usually included in operons of three genes at the leading region of conjugative plasmids of MOBF11 , MOBP11 , MOBP6 and of mobilizable plasmids MOBP13/P14 , belonging to several Inc groups ( IncW , IncN , IncP-1 , IncP-9 , IncQ and IncI-2 ) . Synteny conservation of the stb and MOB regions of representative plasmid groups is shown in Figure 1B . There was a neat bias for the presence of stbB-like genes in conjugative plasmids that carried an MPFT T4SS . In addition , an Stb-system was also found in some groups of mobilizable plasmids . In all cases , the stb genes were located at the 3′side of the nicked strand and are thus the first to enter the recipient cell during plasmid conjugation . Alignment of StbB and homologs from each MOB group showed that they shared a deviant Walker A nucleotide triphosphate-binding motif ( Figure S1A ) , also found in the ParA/Soj/MinD superfamily of ATPases [13] . Members of this superfamily include ParA , required for accurate chromosomal and plasmid DNA partitioning , MinD required for correct placement of the septa during cell division , and Soj , which plays a role in chromosome compaction required for nucleoid partition ( reviewed in [3] , [14] , [15] ) . parA and soj genes are found in their respective operons adjacent to a second gene , parB and spo0J respectively , which encodes a DNA-binding protein . However , no homology of StbA to ParB-like partition proteins was detected . On the other hand , R388 StbA showed significant homology to the TraD protein of plasmid NAH7 . Iterative PSI-BLAST also returned similarly located proteins in MOBP11 , MOBP6 and MOBP13/P14 plasmids ( including StbA_R46 , TraK_RP4 , MobC_pTC-FC2 , and YciA_R721 ) after seven iterations . Alignment of R388 StbA and several similarly located proteins is shown in Figure S2 . According to the sequence of their StbB-like proteins , plasmids can be phylogenetically assorted in two large groups . The first group includes MOBP11 , MOBP6 and MOBP13/14 plasmids , that encode StbB-like proteins ( including TraL_RP4 or MobD_pTF-FC2 ) containing the classical motif KGGXXK[T/S] found in other ParA/Soj/MinD ATPases [13] . R388 plasmid belongs to the second group together with other MOBF11 and MOBP6 plasmids . These encode StbB-like proteins that contain the slightly divergent motif SGXXGK[T/S] . MOBP6 plasmids are pervasive in both groups , perhaps suggesting that they were the first in which the Stb-system was installed . Modeling the structure of StbB using the 3D structure of Soj ( PDB ID: 2BEK ) as template predicts that the dimerization role of the signature lysine K15 in Soj , [16] could be performed by other polar residue like serine S9 in R388-StbB ( Figure S1B ) . To investigate whether the stb operon was involved in R388 transfer , as suggested by its positional conservation relative to the MOB region , we first generated a derivative of R388 deleted of the entire stb operon , R388ΔstbABC . The method of Datsenko and Wanner [17] was used to replace the stb region by a DNA fragment conferring resistance to kanamycin . Mating experiments were carried out under appropriate conditions to avoid indirect effects due to plasmid instability ( see below and Materials and Methods ) . As shown in Figure 2A , deletion of stb did not result in any noticeable effect on transfer frequencies compared to the wild-type plasmid R388 ( R388 ) . These results , which are consistent with previous data [18] , could in principle suggest that the stb operon was not involved in R388 transfer . We nevertheless examined the effect of deleting each of the three genes of the stb operon independently . We constructed three R388 derivatives , R388ΔstbA , R388ΔstbB , and R388ΔstbC , which lack stbA , stbB , and stbC , respectively , and measured their transfer frequencies ( Figure 2A ) . Surprisingly , R388ΔstbA was transferred at a frequency approximately 50-fold higher than R388 , suggesting that StbA inhibits R388 conjugative transfer . In contrast , deletion of stbB resulted in a complete block of conjugation ( transfer frequency <10E-9 , Figure 2A ) , indicating that StbB is required for R388 conjugation . The transfer frequency of R388ΔstbC was comparable to that of R388 , which indicated that StbC had no significant role in R388 conjugative transfer . Since deletion of the entire operon did not modify the conjugation frequencies , we concluded that StbB is required for conjugation only in the presence of StbA , which in turn , inhibits conjugation . StbA and StbB thus appear to have antagonistic effects in conjugation . To further examine the interactions between the different functions of stb genes , we carried out a complementation analysis . We constructed plasmids carrying either the stbA or stbB genes , driven by the Plac promoter and controlled by a lacI q gene present on the same plasmid ( pStbA and pStbB , respectively , Table S1 ) . pStbA and pStbB were introduced in donor cells by transformation , and mating experiments were carried out . Results are shown in Figure 2B . Supplying StbA in trans in donor cells harboring plasmid R388ΔstbA led to a 100-fold reduction in conjugation , which corresponded to a frequency comparable to that of R388 . This result also indicated that the stbB gene was adequately expressed in plasmid R388ΔstbA ( that is , the stbA deletion did not cause a polar effect ) . Supplying StbB in trans in donor cells containing plasmid R388ΔstbB led to restoration of the transfer frequencies to wt level . Providing StbA in trans in donor cells harboring plasmid R388ΔstbABC abolished transfer ( Figure 2B ) , further demonstrating that StbB is required for conjugative transfer only when StbA is present , and that in turn StbA inhibits conjugation in the absence of StbB . Besides , supplying StbB in trans to plasmid R388ΔstbABC resulted in an increase of conjugation frequency of approximately 4-fold . This indicated that StbB stimulates conjugative transfer in the absence of StbA . We next introduced either pStbA or pStbB in donor cells containing plasmid R388 to analyse the effects of overexpressing the corresponding stb gene . Supplying StbA in trans had no effect , while supplying StbB in trans led to enhanced transfer frequencies to a level similar to that of plasmid R388ΔstbA ( Figure 2B ) . This indicated that StbB stimulates conjugation also when StbA is present . Taken together , the results shown in Figure 2 demonstrate that StbA and StbB , but not StbC , are involved in R388 conjugation in E . coli , and that their activities are functionally connected . StbB was strictly required for conjugation only when StbA was present . Besides , StbA prevented R388 conjugative transfer , whereas StbB stimulated R388 conjugative transfer , suggesting that StbA and StbB have balancing/compensatory effect to control conjugation . Previous studies showed that the stb genes of the IncN plasmid R46 are required for stable plasmid inheritance in recombination-proficient but not in recA strains [19] . To examine whether R388 stb operon plays a role in plasmid R388 inheritance , plasmid R388ΔstbABC was subjected to stability studies in LN2666 ( recA+ ) and in FC232 ( LN2666 recA ) strains . As we obtained similar results in both genetic backgrounds , only the results with LN2666 strain are presented in Figure 3 . E . coli cells carrying R388ΔstbABC were grown in serial cultures in nonselective medium , and plasmid loss rates were measured by plating out every 20 generations to determine the proportion of cells retaining the plasmid ( Materials and Methods ) . Plasmid R388 was stably maintained in progeny cells and 100% of the cells retained the plasmid after 80 generations ( Figure 3A ) . Deletion of the entire stb operon led to a significant stability defect with a rate of loss of 5% per generation . We next examined the effects of deletion of each stb gene on plasmid stability . R388ΔstbA showed a plasmid loss rate similar to that of R388ΔstbABC ( 5 . 1% loss per generation ) . In contrast , R388ΔstbB and R388ΔstbC loss rates were close to zero ( 10−2% and 8 . 10−3% per generation , respectively; Figure 3A ) . The stability defect of R388ΔstbABC was thus fundamentally due to the absence of stbA . Complementation of R388ΔstbABC and R388ΔstbA with the StbA-producing plasmid pStbA decreased their loss frequency to 0 . 07% and 0 . 08% , respectively ( Figure 3A ) . This result confirmed that StbA is required for the stable inheritance of R388 and showed that it acts in trans . We then addressed the question of whether the instability of plasmids R388ΔstbABC and R388ΔstbA was due to a decrease in their plasmid copy number per cell when compared to R388 . However , the average number of copies per chromosome of both plasmids , as determined by real-time qPCR ( Materials and Methods ) , were found to be similar to that of R388 ( R388 , 3 . 8±1 . 0 , R388ΔstbABC , 4 . 5±1 . 2 and R388ΔstbA , 4 . 1±0 . 9 ) . This result strongly suggested that the instability of R388ΔstbA and R388ΔstbABC was due to a defect in plasmid segregation . The above results indicated that StbA is involved both in the stability and in conjugative transfer of plasmid R388 , raising the question of how StbA interacts with R388 to perform such functions . The StbA homolog TraK_RP4 was shown to bind specifically to a DNA sequence containing the region between the nic site and the traK gene [20] . We thus explored the role of the upstream region of stb , which contains two sets of five direct repeats of a DNA consensus sequence 5′ TTGCATCAT ( the stbDRs , Figure 1 ) . StbA protein was purified on a Ni-agarose resin as a StbA-His6-tagged protein ( Materials and Methods ) . Incubation of a DNA fragment containing a complete set of stbDRs with increasing quantities of StbA and an excess of nonspecific competitor DNA gave rise to retarded species ( Figure S3 ) . Thus , StbA bound specifically stbDRs containing DNA in vitro . To examine the role of the stbDRs in vivo , we constructed a derivative of plasmid R388ΔstbABC carrying a larger deletion which included the stbDRs ( R388ΔstbDRs-stbABC ) . While providing StbA in trans to plasmid R388ΔstbABC abolished transfer ( Figure 2B ) , StbA had no effect on conjugation of R388Δ ( stbDRs-stbABC ) ( Figure 2B ) , showing that the role of StbA in R388 conjugation is mediated by its binding to the stbDRs: if StbA does not bind the DRs , StbB is not required for conjugation . Besides , providing StbB in trans to R388ΔstbABC , as well as to R388Δ ( stbDRs-stbABC ) results in 4-fold increase in transfer frequencies ( Figure 2B ) , suggesting that the role of StbB does not depend on the presence of the stbDRs . We next examined the role of the stbDRs in plasmid stability . R388Δ ( stbDRs-stbABC ) showed the same instability phenotype as R388ΔstbABC ( Figure 3B ) . Providing StbA in trans , led to an almost complete reduction of R388ΔstbABC plasmid instability ( rate of loss of 0 . 07% per generation ) . Furthermore , providing StbA in trans had only a partial stabilization effect on the R388Δ ( stbDRs-stbABC ) plasmid ( rate of loss of 2 . 8% per generation ) . We concluded that , as for conjugation , the role of StbA in R388 stability is fundamentally mediated by its binding to the stbDRs . Analysis of R388 genome showed that the 9-bp motif specific of the stbDRs is found in four other promoters within the establishment region of R388 ( Figure 1; [12] ) . To examine whether these potential StbA binding sites sequences could have a role in R388 stability , we generated a series of plasmids carrying several deletions of this region . We found that , provided that the stb operon was preserved , the deletion of a DNA fragment ranging from orf14 to kfrA did not affect plasmid stability at all ( R388Δ ( orf14-kfrA ) , rate of loss of 0% , Figure 1A , Table S1 ) . This indicated that this region does not contain genes required for R388 stability and demonstrated that within the establishment region , only the stb genes are required for R388 stability . As expected , a plasmid lacking the entire establishment region ( i . e . from stbA to kfrA gene , see Figure 1 , R388Δ ( stbABC-kfrA ) ) was not stably maintained ( rate of loss of 5% , Figure 3B ) . Supplying StbA in trans led to the stabilization of such a plasmid carrying the stbDRs region ( R388Δ ( stbABC-kfrA ) , Figure 3B ) . As expected , this stabilisation effect was dependent on the stbDRs since the equivalent plasmid lacking the stbDRs , R388Δ ( stbDRs-kfrA ) , remained highly unstable upon StbA production ( 3 . 7% of loss per generation ) . We concluded that the potential StbA binding sites located outside the stbDRs are not sufficient to support StbA-dependent plasmid stabilization . To get a deeper insight into R388 segregation , we undertook a cellular localization study of plasmid R388 using the parS/GFP-ParB system [21] . The parS site of bacteriophage P1 was inserted into R388 plasmid DNA as a parS-chloramphenicol ( parS-Cm ) cassette . This allowed visualization of the cellular localization of the resulting plasmid R388parS ( R388 , Table S1 ) in live E . coli cells expressing the green fluorescent marker protein GFP-Δ30ParB from a second plasmid ( pALA2705; [21]; Materials and Methods ) . We used two different R388parS plasmids ( R388parS1 and R388parS2 ) , in which the P1 parS site had been inserted in two different intergenic regions R388 and obtained comparable results ( Materials and Methods; data not shown ) . Figure 4A shows representative fluorescence images of cells containing R388parS1 plasmid . Discrete foci could be visualized without IPTG induction , at the basal level of expression of the fluorescent GFP-Δ30ParB from the lac promoter [22] . Under these conditions , neither the parS insertions into R388 , nor the expression of the GFP-Δ30ParB protein had a noticeable effect on R388parS plasmid stability and conjugative transfer ( plasmid loss rate <0 . 01%; conjugative transfer frequency 2 . 10−1/donor ) . In the conditions used , more than 98% of the cells analyzed contained GFP-foci , showing that the efficiency of focus detection was high . There were 4 to 10 foci per cell and the average number of foci per cell increased with cell size ( data not shown ) . A majority of the cells ( 64% ) had 4 to 6 foci , and about 34% had 7 to 10 foci ( Figure 5 , R388 ) . The population average was approximately 6 foci per cell , with most small cells harboring 4 foci and longest ones harboring 8 to 10 foci . The number of foci per cell thus roughly corresponded to the copy number of R388 as calculated by qPCR , suggesting that most observed foci contained a single copy of the plasmid . To determine the subcellular position of R388 plasmids , the distance from one cell pole to each focus was measured and plotted as a function of the cell length . For the sake of clarity , only distributions of foci in cells with 5 and 6 foci are presented in Figure 6A . Foci were broadly located and we observed no evidence for preferential positioning at the cell centre or at the ¼ and ¾ cell length positions . However , foci distribution was not random . Indeed , the proportions of cells containing at least one focus in each cell quarter were significantely different from those expected for a random distribution ( Observed/expected for a random distribution: 45%/9 , 4% for 4-foci cells; 86%/23 , 4% for 5-foci cells; 89%/38 , 1% for 6-foci cells; p-values <10−4 in all cases using the χ2 test ) . Thus , foci appeared to be evenly distributed along the cells , suggesting a mechanism of active distribution of R388 plasmid copies . To obtain a global view of foci assortment , we counted the number of foci located within five fractions of half-cells length ( Figure 6E ) . The majority of foci were positioned equally within the four central slices of cells ( from 0 . 1 to 0 . 5 fractional cell length ) . However , 7% of foci were located within the most polar region ( from 0 to 0 . 1 fractional cell length , Figure 6A and 6E ) , showing that the broad distribution of foci extended to the cell poles . We next observed the subcellular localization of the unstable R388ΔstbA plasmid , using the same parS/ParB-GFP system and conditions described above . Under these conditions , neither the parS insertions into R388 , nor the expression of the GFP-Δ30ParB protein had a noticeable effect on R388parS plasmid stability and conjugative transfer ( plasmid loss rate <0 . 01%; conjugative transfer frequency 2 . 10−1/donor ) . Representative images are shown in Figure 4B . About 11% of the cells were devoid of fluorescent focus ( Figure 5 ) , a value consistent with the degree of instability of the plasmid . A majority of cells showed a number of foci in the range of 1 to 3 ( 85% , Figure 5 ) . The population average was 2 foci per cell , which is approximately 3-fold lower than R388 . As mentioned above , the copy number of R388ΔstbA was found to be similar to that of R388 , implying that most R388ΔstbA foci contained 2 or 3 plasmid copies . Thus , the decrease in number of foci is most likely due to plasmid clustering . The subcellular distribution of foci in cells harboring R388ΔstbA is shown in Figure 6B and 6E . In cells with only one focus , the single focus was primarily in the polar region ( 91% of foci located from 0 to 0 . 2 fractional cell length ) . All cells having two or three foci had at least one polar focus . In cells having two foci , the other focus was localized mainly at the opposite polar region ( 48% ) or in the cell center ( 38% ) . In cells having 3 foci , a majority had one focus at each pole and the other at the cell center ( 55% ) or one focus at a pole and the two others at the cell center ( 43% ) . All in all , approximately 33% of the focus-containing cells contained all foci in one side of the cell and no focus close to the center , i . e . , they are cells that would give rise to plasmid-free cells if the cell divided at the mid-position without any change in plasmid number or position . The location of plasmid foci relative to the nucleoid was determined by visualizing cells stained with DAPI ( Figure S4 ) . Fluorescence foci were mainly localized in nucleoid-free areas that were not occupied by the chromosomal DNA , either at the cell poles or at the cell center in the cytosol space between two nucleoids in dividing cells . Therefore , subcellular distribution of the unstable R388ΔstbA plasmid is markedly different from that of the stable R388 plasmid carrying stbA , and correlated with its instability in E . coli ( see above ) . To assay whether StbB had a role in R388 localization , we observed R388ΔstbB-harbouring cells ( Figure 4C and Figure S4 ) . The distribution of foci was almost identical to that of R388 . More than 97% of the cells contained GFP-foci , and had focus numbers in the range of 4 to 10 ( Figure 5 ) , with a population average of 5 . 6 foci per cell . The mean number of R388ΔstbB molecules per cell was found to be 4 . 0±1 . 1 , showing that , similarly to what was observed with R388 , most foci contain a single molecule of the plasmid . Figure 6C shows the subcellular positioning of R388ΔstbB plasmid molecules in cells containing 5 and 6 foci . R388 deleted of stbB was evenly distributed within the cell with the exception of the most polar region ( from 0 to 0 , 1 fractional cell ) , which appeared to contain less foci than R388 . Comparison of R388ΔstbB and R388 foci distributions within the five fractions of half-cells length ( Figure 6E ) using the χ2 test revealed that they were indeed significantly different ( χ2 = 61 . 1 corresponding to a p-value <10−4 ) . This difference can be attributed to a different polar localization of R388 and R388ΔstbB since the distribution of these plasmids within the three central slices of the half-cell ( from 0 . 2 to 0 . 5 ) were not significantly different ( p-value = 0 . 18 ) . We concluded that StbB is required for the localization of a fraction of R388 copies towards the cell poles . Without StbB , R388 molecules are excluded from the poles . To further investigate the role of StbB in intracellular positioning , we compared the localization of the R388ΔstbA and R388ΔstbABC plasmids . As in the case of R388ΔstbA ( see above ) , R388ΔstbABC formed 1 to 5 fluorescent foci per cell , with a majority ( 78% ) of cells containing 1 to 3 foci ( Figure 4D and Figure 5 ) , and showing a strong bias for location at the center and cell poles ( Figure 6D ) . However , R388ΔstbABC and R388ΔstbA foci distributions were found to be significantly different ( p-value <10−4 ) . As illustrated in Figure 6E , this difference mainly relied on the proportion of foci within the most polar region ( from 0 to 0 . 1 fractional cell length; R388ΔstbABC , 12% , compared to R388ΔstbA , 24% ) and at the center ( R388ΔstbABC , 33% , compared to R388ΔstbA , 22% ) . Thus , StbB is required for the localization of R388 at the pole and midcell positions in both the presence and absence of StbA .
In this study , we show that protein StbA is strictly required for stability and intracellular positioning of plasmid R388 in E . coli . We found that fluorescent foci of R388 , most of which contain a single copy of the plasmid , are evenly distributed along the cell , with no evidence for preferential localization . This is in contrast with other low-copy number plasmids , such as mini-F , mini-P1 , R27 and RK2 plasmids , which were reported to localize as clusters at the ¼–¾ or midcell positions [21] , [23]–[26] . In these cases , duplication of the central focus is presumed to represent active partition of plasmid copies . However , it has been recently shown for mini-P1 plasmid that more than two foci are present in most conditions , and that the behavior of foci is more dynamic than previously reported [27] . This model of segregation of mini-P1 is more consistent with the even distribution of R388 copies that we have observed . We thus hypothesize that , as proposed for mini-P1 , R388 copies segregate as single units and distribute into a dynamic and evenly spaced pattern along the cell to ensure a proper distribution of the plasmid copies at cell division . Moreover , in contrast to non-conjugative mini-F and mini-P1 plasmids , which were reported to be contained within the nucleoid region [21] , [23] , [25] , [27] , a significant fraction of R388 foci are found at the extreme cell ends . This observation may reflect plasmid R388 ability to undertake conjugative transfer . In agreement with this , we have recently shown that R388 coupling protein TrwB localizes to the cell poles ( data not shown ) . Besides , it has been reported that the T4SS apparatus of Agrobacterium tumefaciens and of plasmid pCW3 from Clostridium perfringens , assemble at the cell poles [28]–[31] . In contrast to the even distribution of R388 , R388ΔstbA plasmid foci clustered at the cell poles or at the cell center , in nucleoid-free areas . Mislocalized plasmid clusters appear to be the main cause of instability , as they are not adequately distributed in cellular spaces corresponding to future daughter cells ( Figure 7 ) . Mislocalized plasmid clusters were reported in derivatives of plasmids mini-F , mini-P1 , R27 , and R1 in which their Par regions were inactivated [25] , [32]–[34] . In these cases , plasmid foci appeared distributed randomly in nucleoid-free spaces . This similarity suggests that StbA acts as a partition system . Indeed , the stb operon shares many characteristics with Par systems implicated in plasmid and chromosome partitioning . StbA is a DNA-binding protein which binds a cis-acting sequence ( it is thus a ParB-like protein ) , the stbDRs , and StbB is a putative motor protein harboring Walker-type ATPase motifs ( thus a ParA-like protein ) . However , and in contrast to ParA-like counterparts , StbB is not required for R388 stability , suggesting that the StbAB system does not constitute a typical ParAB system . StbB may either have no role in R388 maintenance or may be replaced by an equivalent cellular function when inactivated . Alternatively , R388 segregation may not need to involve an active motor . In this view , the StbA-stbDRs complex may be used to pair plasmid molecules with the host chromosome , ensuring an even distribution of R388 copies along the nucleoid length by an unknown mechanism . Besides , we demonstrate that , although deletion of the entire stb operon does not affect conjugation , StbA and StbB , but not StbC are involved in R388 conjugation . Deletion of stbA results in an enhanced frequency of conjugation , while deletion of stbB leads to a conjugation defect , indicating that StbA and StbB have opposite but connected effects . As explained above , our results further suggest that these conjugation defects are a consequence of variations in the intracellular positioning of plasmid DNA . Whatever the way StbA promotes R388 segregation , the associated localization is certainly not convenient for maximal conjugation frequency , since StbA inactivation , associated with plasmid localization at the cell poles , strongly enhances conjugation . In addition , R388ΔstbB conjugation defect correlates with the absence plasmid foci at the cell polar membrane ( Figure 7 ) . We thus assume that the role of StbB is either to counteract StbA by locating some plasmid copies at the polar conjugal transport site , thereby allowing conjugation to occur , or StbA may modulate the activity of StbB and/or delocalize the plasmid copies . The way the relaxosomal complex is transferred to conjugative pores remains unknown . It was previously suggested that A . tumefaciens VirC1 protein , which belongs to the ParA and Soj/MinD ATPases family , spatially positions the relaxosome at the cell pole to coordinate substrate-T4SS docking [35] . StbB also shares features with ParA/Soj/MinD ATPases ( Figure 2B ) . These proteins are though to employ a principle of dynamic oscillation between specific surfaces such as membrane or bacterial chromosome to explore and mark the cellular environment [4] . Several models for the action of the Walker partition ATPases have been proposed . Formation of dynamically unstable filaments in a nucleotide-dependent manner was suggested following the example of actin-type partition ATPases [4] . Such cytomotive filaments could achieve partitioning by pushing plasmids attached to growing filaments , or by pulling plasmids attached to retracting filaments and there has been some evidence for both modes of action [36] , [37] . Alternatively , a diffusion ratchet model was proposed ( Vecchiareli et al . , 2010 ) . In this case , the motive force for plasmid positionning does not rely on the ParA ATPase polymerization , but instead is directed toward regions of high ParA concentration . These models are all consistent with our current data and such dynamic modes of action constitute appealing mechanisms for how StbB might recruit R388 molecules from a cytosolic pool to the membrane for conjugative transfer . Since inactivation of either the coupling protein ( TrwB ) or the relaxase ( TrwC ) did not affect plasmid R388 cellular localization ( data not shown ) , the StbB-dependent mechanism of transport of R388 molecules to the cell membrane is neither associated with relaxosome formation , nor it requires either the coupling protein or cleavage at oriT . StbB may interact either with plasmid DNA or with StbA bound to DNA to form nucleoprotein complexes analogous to the ParA/ParB/parS partitioning complex , but linked to conjugative DNA processing . We are presently investigating the detailed molecular mechanisms by which StbB interacts with R388 molecules to recruit them to the cell membrane prior to transfer . Our comparative genomics studies showed the conservation of synteny of three genes , of which the second gene is the most conserved , at the leading region of conjugative plasmids of mobility groups MOBF11 , MOBP11 , MOBP6 , indicating that the stbABC operon is widespread among plasmids . Moreover , the stb operon is apparently linked to MPFT T4SS systems , although not exclusively , as it is also carried by mobilizable plasmids of the MOBP13/P14 group . It remains to be explored if these plasmids require a MPFT-type T4SS for conjugative transfer . Synteny conservation may reflect a requirement for stb in plasmids carrying such conjugation machinery under natural conditions . This is supported by our observation that the presence of R388 transfer region leads to instability of a pBR322-derivative plasmid containing it ( data not shown ) . In summary , we present experimental evidence that the StbAB system constitutes an atypical plasmid stabilization system intimately linked to conjugative transfer . On one hand , StbA is strictly required for plasmid R388 stability . Its inactivation results in mislocalization of R388 copies towards the cell poles , which is correlated with a significant increase in transfer frequencies ( Figure 7 ) . On the other hand , StbB is necessary for conjugative transfer only in the presence of StbA . Its inactivation leads to conjugation defect , which is associated with the absence of plasmid molecules at the extreme cell poles ( Figure 7 ) . Our results thus suggest that the StbAB system may act as a molecular scale between two possible transmission modes of plasmid R388: vertical transmission by faithful segregation to daughter cells and horizontal transmission by conjugative transfer to a different cell ( Figure 7 ) . It would seem that an active conjugation system provokes plasmid instability per se , which could only be counterbalanced by Stb or an analogous stability system . In this case , it remains to be investigated if other plasmids use functionally analogous but phylogenetically unrelated systems to balance their propagation and vegetative modes . This is , to our knowledge , the first report of a system involved in the reconciliation of these two cellular processes .
For mating experiments , donor ( LN2666; [38] ) and recipient ( BW27783; [39] ) strains were grown overnight from single colonies in LB medium at 37°C with appropriate antibiotics . After washing , 50 µl of donor cells were mixed with 800 µl of recipient cells , the mixture centrifuged for 1 min , resuspended in 10 µl LB medium and cells placed onto a GS Millipore Filter ( 0 . 22 µm pore size ) on a LB-agar plate at 37°C for 20 min , which corresponds to a period shorter than the generation time to limit indirect effects due to plasmid instability . Bacteria were then washed from the filter , diluted in 2 ml LB medium and serial dilutions plated on selective media . Conjugation frequencies were expressed as the number of transconjugants per donor cell . When providing plasmids pStbA or pStbB ( Table S1 ) , the amount of protein StbA or StbB produced was found to be sufficient to restore the wt phenotype without the need to induce their expression with IPTG . Single colonies of LN2666 or DH5α strains ( Table S1 ) containing R388 or a derivative of it were used to inoculate LB containing selective antibiotics , and the cultures were incubated overnight at 37°C . For each strain , stability experiments were performed at least four times , starting from separate colonies . 48 . 8 µl of a 10-fold dilution of overnight cultures were transferred to 5 ml LB , containing streptomycin ( 300 µg/ml ) but lacking the antibiotic selective for the plasmid , and grown for 12 h , i . e . for 10 generations . These freshly inoculated cultures constituted time point zero . From then on , 48 . 8 µl of a 10-fold dilution of the full-grown cultures was transferred every 12 h to fresh 5 ml LB and incubated at 37°C to reach a total of 80 generation times . Each 12 h , the cultures were also diluted and plated onto LB plates . Determining the fraction of plasmid free cells in the population was done by replica-picking 100 randomly chosen colonies per culture from the LB plates onto LB plates containing the appropriate selective antibiotics , and scoring the proportion of colonies with a given resistance . The percentage of plasmid loss per generation was calculated as described in Yates et al . , 1999 [40] . A derivative of strain LN2666 ( recA+ ) containing plasmid pALA2705 ( Ap; Table S1 ) , which produces the fluorescent GFP-Δ30ParB protein [21] , was transformed with DNA of plasmid R388::parS-Cm or one of its derivatives . Neither the parS insertions into R388 ( or stb derivatives listed in Table S1 ) , nor the expression of the GFP-Δ30ParB protein had a noticeable effect on the corresponding R388parS plasmid stability and conjugative transfer ( data not shown ) . Single-colony isolates were grown overnight in M9 medium supplemented with 0 . 2% casamino acids , 0 . 4% glucose , 2 . 0 µg/ml thiamine , 20 µg/ml leucine and 20 µg/ml thymine , ( suppl . with Ap , Cm ) at 30°C . Cultures were then diluted 1/100 in the same medium and grown at 30°C to OD600 of 0 . 8 . With these constructions , foci could be adequately visualized without the need to induce expression of the fluorescent GFP-Δ30ParB protein [27] . When needed , DAPI stain ( 1 µg/ml; Molecular Probes ) was added to the culture for 20 min to label DNA . Suspensions of growing cells were directly deposited on glass slides covered by a layer of 1% agarose containing the same growth medium and examined by phase-contrast and fluorescence microscopy . Images were captured with an inverted Olympus X81 microscope equipped with a 100x oil-immersion Olympus lens ( N . A . of 1 . 3 ) and a Roper Coolsnap CCD camera , using Metamorph software . Cell length and focus position was measured manually using ImageJ . Each strain was examined in at least four independent experiments with similar results . At least 200 cells were inspected for each experimental observation . | The ability of bacteria to evolve and adapt to new environments most often results from the acquisition of new genes by horizontal transfer . Plasmids have a preponderant role in gene exchanges through their ability to transfer DNA by conjugation , a process that transports DNA between bacteria . Besides , plasmids are autonomous DNA molecules that are faithfully transmitted to cell progeny during vegetative cell multiplication . In this study , we report a system composed of two proteins , StbA and StbB , which act to balance plasmid R388 physiology between two modes: a maintenance mode ( vertical transmission ) and a propagation mode ( horizontal transmission ) . We demonstrate that StbA is essential to ensure faithful assortment of plasmid copies to daughter cells . In turn , StbB is required for plasmid R388 adequate localization for conjugation . This is the first report of a system which reconciles plasmid segregation and conjugation . Furthermore , R388 belongs to the IncW family of conjugative plasmids , which are of particular interest due to their exceptionally broad host range . We show that the StbAB system is conserved among a wide variety of conjugative plasmids , mainly broad host range plasmids . Thus , the Stb system could constitute an interesting therapeutic target to prevent the spread of adaptive genes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"escherichia",
"coli",
"prokaryotic",
"models",
"model",
"organisms",
"molecular",
"cell",
"biology",
"genetics",
"biology",
"microbiology",
"genetics",
"and",
"genomics",
"gene",
"function"
] | 2011 | The stb Operon Balances the Requirements for Vegetative Stability and Conjugative Transfer of Plasmid R388 |
Gut microbiomes play crucial roles in animal health , and shifts in the gut microbial community structure can have detrimental impacts on hosts . Studies with vertebrate models and human subjects suggest that antibiotic treatments greatly perturb the native gut community , thereby facilitating proliferation of pathogens . In fact , persistent infections following antibiotic treatment are a major medical issue . In apiculture , antibiotics are frequently used to prevent bacterial infections of larval bees , but the impact of antibiotic-induced dysbiosis ( microbial imbalance ) on bee health and susceptibility to disease has not been fully elucidated . Here , we evaluated the effects of antibiotic exposure on the size and composition of honeybee gut communities . We monitored the survivorship of bees following antibiotic treatment in order to determine if dysbiosis of the gut microbiome impacts honeybee health , and we performed experiments to determine whether antibiotic exposure increases susceptibility to infection by opportunistic pathogens . Our results show that antibiotic treatment can have persistent effects on both the size and composition of the honeybee gut microbiome . Antibiotic exposure resulted in decreased survivorship , both in the hive and in laboratory experiments in which bees were exposed to opportunistic bacterial pathogens . Together , these results suggest that dysbiosis resulting from antibiotic exposure affects bee health , in part due to increased susceptibility to ubiquitous opportunistic pathogens . Not only do our results highlight the importance of the gut microbiome in honeybee health , but they also provide insights into how antibiotic treatment affects microbial communities and host health .
Gut microbial communities influence animal health in many ways , including synthesis of vitamins , digestion of food , defense against pathogens , and modulation of behavior , development , and immunity [1] . The gut microbial community can be disturbed by several factors: one of the most potent sources of disturbance for humans and domesticated animals is antibiotic treatment , which can severely alter community size and composition [2] . Treatment with antibiotics has also been associated with the appearance of resistant pathogens such as Clostridium difficile and Salmonella enterica [3–5] . Multiple studies have shown that reduction of gut microbial diversity occurs within a few days of ingestion of antibiotics [1 , 6 , 7] , and complete recovery of initial bacterial community composition is rarely achieved [7] . In fact , it has been suggested that the overuse of antibiotics has permanently changed our microbiomes , causing an increase in “modern plagues” such as obesity , asthma , diabetes , and certain forms of cancer [8] . However , the duration and extent of antibiotic-induced disturbance in the gut microbiota remains poorly characterized , particularly at the species and strain level where the diversity of the gut community is the greatest [9 , 10] . Characterizing shifts in size and composition of the microbiota is particularly difficult in mammalian hosts , because of the complexity of their gut communities . Model organisms provide opportunities to study host—microbiome interactions with a level of experimental control that is not achievable in human studies , and models can be used to understand the generality of associations between the microbiome and disease . The gut communities in social insects , such as honeybees ( Apis mellifera ) , are particularly useful as models , because they share common features with mammalian gut communities . As in mammals , honeybees acquire their gut microbiota through social contact [11] , in contrast to many invertebrates , which acquire gut bacteria from environmental sources . Similar to humans , the gut microbiota of honeybees is composed of host-specialized bacterial species that live only in the host gut [12] and that show considerable strain diversity within individual hosts [13] . However , in contrast to humans and other mammals , bees have a relatively simple gut microbiota , dominated by only eight core bacterial species , which comprise 95%–99% of bacteria in the gut [14 , 15] . Therefore , the honeybee provides a tractable system to study the function and evolution of host-associated microbial communities . Additionally , honeybees are important globally as agricultural pollinators [16] . Since 2006 , the world's honeybee colonies have undergone elevated mortality , with multiple factors linked to the declines [17 , 18] . Several results suggest that the gut microbiome contributes to bee health [19–22] . Therefore , dysbiosis ( microbial imbalance ) may impact honeybee health and susceptibility to disease . Another parallel between the microbiomes of honeybees and humans is the long history of exposure to antibiotics , a potent source of disturbance to gut communities . Antibiotic treatment of bee colonies has been widely used for over 50 y in the United States to prevent a bacterial disease of bee larvae called foulbrood ( Paenibacillus larvae ) [23–25] . The two antibiotics most commonly used by beekeepers are tetracycline ( or the related compound oxytetracycline ) and , since 2006 , tylosin . Tetracyclines are also used for treating bacterial infections in humans and are commonly incorporated into livestock feed , resulting in the acquisition of tetracycline resistance in many bacteria , including some pathogenic taxa [26] . Likewise , the use of tetracycline in US beekeeping has resulted in an accumulation of resistance genes in microbiomes of US honeybees as compared to bumblebees or honeybees in countries that do not use antibiotics in beekeeping [27] . Widespread antibiotic resistance has been reported in P . larvae , the bacterial pathogen that causes American Foulbrood ( AFB ) [28] , and a resistance gene ( tetL ) found in P . larvae is identical in sequence to one of the resistance loci harbored by honeybee gut symbionts [27] , suggesting past horizontal transfer between commensal gut bacteria and this pathogen . In this study , we evaluate the effects of tetracycline exposure on bee survivorship and on the size and composition of honeybee gut communities . We sampled treated bees at different time-points postexposure to determine if the microbiome recovers to pretreatment status . We monitored post-treatment survival within hives to determine if gut dysbiosis impacts honeybee health , and we tested whether antibiotic exposure increases susceptibility to infection by opportunistic pathogens present in hives . Our results show that treatment with tetracycline severely alters both the size and composition of the honeybee gut microbiome . Moreover , the perturbations caused by tetracycline treatment were still evident one week after bees were returned to their hives following exposure . Our results show that tetracycline-induced dysbiosis can decrease the survival rate of bees and suggest that this reflects increased susceptibility to opportunistic pathogens .
Adult worker bees were collected from a brood frame from a single hive . Bees were fed filter-sterilized sucrose syrup ( controls ) or tetracycline suspended in filter-sterilized sucrose syrup ( treatments ) for 5 d before being returned to the hive or maintained in the laboratory under sterile ( i . e . , kept only with other tetracycline-treated bees from their cohort ) or exposed ( i . e . , with normal workers collected from their hive ) recovery conditions . In order to determine how antibiotic treatment affects longevity and the size and composition of the gut microbiome , bees were censused and sampled at several time points post-treatment . The gut microbiota was assessed for total number of bacteria and for community composition using quantitative PCR and deep amplicon sequencing of a region of the bacterial 16S rRNA gene . Antibiotic treatment resulted in major changes in community size starting on the first sampling day ( Day 0 , before reintroduction to the hive ) ( Fig 1A and 1B ) . None of the core species was completely eradicated by tetracycline treatment , but the total bacterial abundance as well as the absolute abundance of several species decreased in treated bees ( Wilcoxon test , p < 0 . 05 ) ( Fig 1B and 1C ) . Of eight core bacterial species found in the honeybee gut [14] , four were significantly affected by tetracycline treatment . The Gram-positive taxa , Bifidobacterium , Lactobacillus Firm-5 , and Lactobacillus Firm-4 , were the most affected ( Wilcoxon test , p < 0 . 0001 ) ( Fig 1C ) . The Gram-negative species , Snodgrassella alvi , was also reduced on Day 0 ( Wilcoxon test , p < 0 . 05 ) ( Fig 1C ) . Although changes in absolute abundance were seen at the first sampling time-point , no significant changes in the relative abundances of the native bacterial taxa were detected ( S1 Fig ) . After reintroduction to the hive , bees were censused to determine effects of tetracycline treatment on longevity . Recovery of treatment bees ( 32% ) from the hive on Day 3 post-treatment was significantly lower than the recovery of control bees ( 64% ) ( Chi-squared test , p < 0 . 0001 ) ( Fig 2A ) . A replicate survival experiment performed in a different hive also indicated that tetracycline-treated bees have increased mortality in the hive ( S2A Fig ) . We also performed laboratory recovery experiments in order to control for age , determine the effects of tetracycline on the survival of germ-free bees , and determine if the bee gut microbiome could recover with exposure to workers from their hive or without such exposure . The complementary experiments performed on bees kept in the laboratory showed that the decrease in survival rate is not due to side effects of tetracycline ( S2B–S2D Fig ) . For bees possessing their natural microbiota , antibiotic treatment caused a decrease in survival during recovery in laboratory-kept bees , for both sterile recovery bees and bees exposed to untreated hive workers ( Chi-squared test , p < 0 . 0001 ) ( S2B and S2C Fig ) . In contrast , for germ-free bees , antibiotic treatment did not result in increased mortality compared to controls ( S2D Fig ) . Bees were sampled on Days 3 , 5 , and 7 to evaluate the long-term effects of antibiotic exposure on gut community composition . Bees treated with tetracycline and returned to the hive displayed changes in both community composition and size at all post-treatment sampling points ( Wilcoxon test , p < 0 . 05 ) ( Fig 2B–2D ) . Control bees had , on average , five times more bacterial cells in their guts than bees treated with tetracycline , and this discrepancy was evident at all post-treatment sampling time-points ( Fig 2B and 2C ) . The four core taxa that decreased in absolute abundance at Day 0 ( Wilcoxon test , p < 0 . 001 ) ( i . e . , Bifidobacterium , Firm-4 , Firm-5 , and S . alvi ) continued to be significantly decreased at all time-points ( S3A–S3D Fig ) . Another core species , Bartonella apis , which was not significantly altered at Day 0 , was decreased at all subsequent sampling time-points ( Wilcoxon test , p < 0 . 001 ) ( S3E Fig ) . Additionally , the absolute abundance of a few other species shifted at various time-points after the bees were returned to the hive ( Wilcoxon test , p < 0 . 05 ) . Two core species ( Alpha 2 . 1 and Frischella perrara ) and one environmental species ( Lactobacillus kunkeei ) were decreased at one or more time-points ( S4A–S4C Fig ) . In contrast , several non–core taxa , including the genus Serratia and unclassified bacteria in the family Halomonadaceae , showed elevated abundance at Days 3 and 5 , respectively ( Wilcoxon test , p < 0 . 05 ) ( S4D and S4E Fig ) . Complementary experiments in which bees were kept in the lab after tetracycline treatment exhibited similar effects on community size and composition ( S5 Fig ) , but did not show an increase in non—core bacteria . In addition to an increase in non-core bacterial taxa in tetracycline-treated bees , we also observed an apparent increase in fungal sequences in treated bees at Days 3–7 , based on diagnostic PCR assays ( S6A Fig ) . However , identified fungal taxa ( see Materials and methods ) were all closely related to yeast taxa isolated from flowers ( S6B Fig ) , suggesting that these are transient in guts of these bees , and are likely more abundant and detectable in treated bees because fewer bacteria are present . In some treated bees , the typically specific fungal primers amplified plant DNA ( S6B Fig ) , suggesting that fungal DNA template is rare in the samples . After bees were returned to the hive , differences in community composition were also apparent in the relative abundances of individual species ( Fig 2D and S7 and S8 Figs ) . The mean relative abundances of bacterial species remained stable in control bees over all sampling periods , whereas treatment bees displayed a major shift in gut microbial composition that was not stable over time and did not return to the baseline composition after one week ( S7 Fig ) . Concerning the core bacteria of the gut , the relative abundances of Bifidobacterium , Firm-4 , Firm-5 , and B . apis were decreased at Days 3 , 5 , and 7 ( S7 and S8 Figs ) . However , the relative abundance of Gilliamella apicola was much higher in treated bees ( S7 and S8 Figs ) . Based on relative abundance , antibiotic treatment also caused changes in microbiota diversity within individual hosts ( alpha diversity ) and in microbiota divergence between individual hosts ( beta diversity ) ( Wilcoxon test , p < 0 . 05 ) ( Fig 3 ) . Alpha diversity , measured as Shannon’s H index , was lower in treatment bees at all time points except Day 0 ( Wilcoxon test , p < 0 . 0001 ) ( Fig 3A ) . Beta diversity , measured as the average Bray-Curtis dissimilarity , was lower among control bees than between control and treatment bees at all time-points ( Wilcoxon test p < 0 . 0001 ) ( Fig 3B ) . Principal coordinate analysis ( unweighted and weighted UniFrac , [29] ) showed that gut community compositions of treatment bees are widely dispersed in contrast to the tight clustering observed for control bees ( Fig 3C and 3D ) . Furthermore , for the bees retained in laboratory cages , the gut community compositions of treatment versus control bees displayed similar clustering patterns based on unweighted and weighted UniFrac: treated bees had more dispersed communities for both sterile bees and bees exposed to other bees in social groups ( S9A–S9D Fig ) . To evaluate the effects of tetracycline treatment on fine-scale strain and species diversity , we counted the number of 99% operational taxonomic units ( OTUs ) assigned to each genus . At Day 0 post-treatment , no significant differences were seen in 99% OTU diversity , but at Days 3 , 5 , and 7 , the total number of OTUs was significantly lower in treatment bees ( Wilcoxon test , p < 0 . 05 ) ( Fig 4A ) . In particular , Bifidobacterium , Firm-4 , Firm-5 , and B . apis showed a decrease in fine-scale diversity ( Wilcoxon test , p < 0 . 05 ) ( Fig 4B–4E ) . This is consistent with the decrease in relative and absolute abundance of these species ( Fig 1 and S3 Fig ) . Furthermore , the increase in relative abundance of G . apicola also corresponded to an increase in 99% OTU diversity at Days 3 and 7 post-treatment ( Wilcoxon test , p < 0 . 001 ) ( Fig 4F ) . In order to investigate whether opportunistic pathogens contribute to the increased mortality observed for treated bees returned to the hive , we performed infection experiments using a Serratia strain that was isolated from honeybee guts ( Serratia kz11 ) . In one experiment , we exposed age-controlled bees ( emerged in the lab on the same day ) to Serratia kz11 after treatment with tetracycline ( see Materials and methods ) . In the other experiment , we exposed non-age-controlled ( bees taken from a brood frame in the hive ) . In both experiments , the bees were exposed to Serratia kz11 through their food for 2 d . ( Viable Serratia cells can be obtained from bee bread up to 2 d after inoculation ) . In both age-controlled and non-age-controlled experiments , bees treated with tetracycline and exposed to Serratia kz11 exhibited increased mortality when compared to control bees , bees exposed to tetracycline only , or bees exposed to Serratia only ( Fig 5 , S5 Data and S6 Data ) . To confirm that Serratia kz11 can be an opportunistic pathogen of honeybees , we performed a bacterial challenge experiment in which we exposed bees to different bacteria following tetracycline treatment . We exposed control and treatment bees to i ) Serratia kz11 , ii ) Escherichia coli K-12 , iii ) S . alvi wkB2 , iv ) Lactobacillus sp . wkB8 . The latter two species are part of the core bee gut microbiome [12] , and E . coli K-12 is a nonpathogenic lab strain [31] . Only bees exposed to Serratia kz11 showed an increase in mortality in control and treated bees when compared to control bees ( Fig 6 , S9 Data ) .
Since gut community members engage in mutualistic interactions , such as cross-feeding , and antagonistic interactions , such as competition and direct killing , responses of different species and genotypes to environmental changes are interdependent . These interactions can be altered or eliminated as a consequence of antibiotic treatment , potentially impacting host health . Several studies have shown that the use of antibiotics causes alterations in the microbiomes of humans and livestock ( reviewed in [32] ) . In honeybees and bumblebees , globally important pollinators , gut communities have been implicated in both nutrition and defense against pathogens [19–22] . In relation to growing evidence for the importance of the gut microbiome in animal health [32] and the largely unexplained decline of honeybee colonies [18] , the effects of antibiotic treatment on the honeybee gut microbiome are of major interest . Tetracycline , a broad-spectrum antibiotic , targets both Gram-positive and Gram-negative bacteria and is thus expected to affect multiple members of the gut community . As predicted , we observed substantial changes in the gut microbial community composition and size following treatment with tetracycline . However , none of the core bacterial species was completely eliminated . This persistence may have been enhanced by the presence of antibiotic resistance . In US honeybees , core species of the microbiota carry tetracycline resistance genes , which persist at low frequencies , even in hives with no recent history of antibiotic treatment [27] . Thus , we expect some tetracycline resistance in our hives , which had not been treated for over 2 y prior to our study . Nevertheless , most core species declined in population size and/or diversity following treatment . An exception was G . apicola , for which relative abundance as well as strain-level diversity increased after treatment . Overall , the effects of tetracycline treatment on the gut microbiome composition were more prominent several days after treatment was stopped , probably due to a delayed effect of the antibiotics . Also , dead bacterial cells may have accumulated in the gut from lack of defecation while bees were maintained in laboratory cages , causing our DNA-based profiles to fail to reveal initial declines in numbers of living cells . Even so , significant declines in the overall community size were seen for all treatment bees starting on the first day of sampling . Moreover , the same delayed effect was observed in treated bees kept in the lab throughout the recovery period . We found that honeybees treated with antibiotics and returned to the hive had decreased survivorship when compared to untreated bees . Several studies have pointed to a role for the bee gut microbiome in protection against trypanosomatid pathogens [19 , 21 , 33] . One recent study showed that the colonization order of honeybee gut symbionts affects susceptibility to infection by the pathogenic trypanosomatid Lotmaria passim [21] , providing evidence that gut dysbiosis promotes pathogen invasion . We detected elevated levels of two groups of non—core bacteria , Serratia and an unclassified Halomonadaceae , in treated bees sampled from the hive; these could represent opportunistic pathogens able to invade the gut as a result of antibiotic perturbation . Members of the family Halomonadaceae generally inhabit high-saline and pH environments , but some have been recognized as human pathogens [34 , 35] . Halomonadaceae-related taxa have been detected in microbiome studies of the honeybee gut and pollen , but their status in bees is unknown [36] . Serratia is an opportunistic pathogen in humans and many animals , including insects [37 , 38] . Along with other Enterobacteriaceae , it is widely present at low frequencies in honeybee guts where it is considered a signifier of atypical microbiome composition in bees [15 , 39] . Therefore , one or both of these bacteria could be responsible for the increased morality in treated bees . To test this , we exposed treated and control bees to a Serratia strain isolated from honeybees . We observed that Serratia exposure resulted in elevated mortality in bees that had been treated with tetracycline . Furthermore , this Serratia strain shows relatively high resistance to tetracycline ( see Materials and methods for details ) , suggesting that it would also have a selective advantage during the course of antibiotic treatment . Dysbiosis can lead to the sudden overgrowth and pathogenic behavior of opportunistic organisms ( pathobionts ) already present in the gut [40] . Perturbations of the gut community can affect gene expression , protein activity , and the overall metabolism of the gut microbiota [41] . For example , changes in microbial community structure can alter the provision of nutrients or secondary metabolites and inhibit the removal of toxic metabolites [42] . Metagenomic analysis of colonies with colony collapse disorder ( CCD ) , showed increases in relative abundances of G . apicola , F . perrara , S . alvi , and Lactobacillus and decreases in Alphaproteobacteria and Bifidobacteria when compared to healthy hives [43] . We also observed an increase in the abundance and diversity of G . apicola in treated bees within hives . These results suggest negative effects of high abundance of G . apicola or of reduced abundances of Bifidobacterium and Lactobacillus , which are thought to be protective in humans and other animals , including honeybees [44–46] . Furthermore , bees with naturally acquired microbiomes that were treated with antibiotics and kept in sterile conditions in the lab exhibited increased mortality rates , which were not observed in treated bees lacking their microbiome ( germ-free bees ) , implying that dysbiosis alone , rather than the tetracycline treatment itself , can impact bee health . Although the lack of an effect on germ-free bees suggests that the antibiotic is not directly harmful to bees in the concentrations we used , it is difficult to disentangle effects of tetracycline on the gut microbiome from effects on the host , which may in turn alter susceptibility to pathogens . Antibiotics are commonly used in apiculture in several countries [47] and are administered to hives by mixing with powdered sugar , sugar syrup , or dietary extender patties . The recommended treatment involves feeding or dusting each hive with approximately 200 mg/oz of tetracycline three times in the spring and fall at intervals of 4–5 d . The dose we administered in this study was slightly lower than that used in apiculture . In actual hive conditions , it is unclear how much antibiotic individual bees would consume , but it is likely that some bees receive doses as high or higher than those used here . Additionally , when antibiotics are administered to hives , they can persist for long periods of time . For example , tetracycline has been detected in treated hives for up to 3 mo post-treatment [47 , 48] . Therefore , the effects on treated hives could be greater than what we report here . Our results suggest that treated bees allowed to recover in the hive without further exposure revert towards their baseline microbiome composition after 1 wk ( Fig 3 ) . Potentially , negative effects of antibiotic treatment could be reduced through the development of alternative treatment methods that allow for the removal of the antibiotic from the hive after a specified period . In this study , we found that tetracycline , a commonly used antibiotic in beekeeping and in other livestock , severely alters the gut microbiome composition of honeybees and decreases survivorship within hives . These results thus suggest a beneficial role of the normal gut microbiome in honeybee health within the hive environment . A possible implication is that the use of antibiotics in beekeeping can be detrimental because of interference with these benefits . We show that a strain of Serratia isolated from bees causes increased mortality following tetracycline treatment . Antibiotic-induced dysbiosis may also lead to increases in nonbacterial pathogens , such as viruses and eukaryotes , as shown for trypanosomatids [19 , 21] . Other consequences of dysbiosis , such as nutritional impacts or heightened susceptibility to toxins , may also contribute to the survivorship decline that we observed in antibiotic-treated bees . Although our results suggest that antibiotic treatments may be detrimental for honeybees , we emphasize that many other factors contribute to pollinator declines , which are affecting many wild pollinator populations that are not exposed to antibiotics [49] . Therefore , antibiotic treatment can only be considered a single potential factor amongst a myriad of other potential culprits , including loss of foraging habitat . Furthermore , antibiotic treatments sometimes may be highly beneficial , to prevent or control infections by foulbrood agents [23] . Overall , our findings underline the usefulness of honeybees as a model system for disentangling the fine-scale dynamics of perturbed gut communities . Furthermore , the honeybee gut system can provide fundamental information on how antibiotic treatment affects the normal microbiota and host health .
Approximately 800 adult worker bees were collected from brood frames from a single hive ( Big Top ) kept at the University of Texas in Austin ( UT ) . The bees were immobilized at 4°C and marked with green or pink Testors paint . The bees were separated into two groups , control and treatment , green and pink , respectively , and were distributed to cup cages of previously described design [50] . The cup cages were maintained in growth chambers at 35°C and 90% relative humidity to simulate hive conditions . Each cup cage contained 30 bees , with 15 replicates for each condition . Control bees were fed filter-sterilized 0 . 5 M sucrose syrup and treatment bees were fed 450 ug/ml of tetracycline suspended in filter-sterilized 0 . 5 M sucrose syrup . After 5 d , 15 bees were sampled ( one from each cup cage ) , placed in 100% ethanol , and stored at 4°C . The remaining bees ( 345 control and 340 treatment ) were then returned to their original hive . Three days after reintroduction to the hive , the bees were individually captured and temporarily kept in cup cages ( 40 bees per cup ) until all marked bees had been recovered from the hive . After counts were obtained , 15 marked bees for both control and treatment groups were collected from each hive at time-points of 3 , 5 , and 7 d following antibiotic treatment ( Nov . 6–13 , 2015 ) , placed in 100% ethanol and stored at 4°C . Within 2 wk of collection , the bees were removed from cold ethanol and the entire gut from crop to rectum was homogenized and placed into a bead-beating tube in 500 uL of 100% molecular grade ethanol and stored at −20°C . Dissections were performed with flame-sterilized forceps under aseptic conditions . DNA was extracted from the gut using established techniques , Illumina-based amplicon profiling of the V4 region of 16S rRNA gene was performed ( S11 Data ) , and the community size pre- and post-treatment was quantified by qPCR using the total number 16S rRNA gene copies , adjusting for number of rRNA operons per genome ( S1 Table ) . A replicate hive survival experiment was performed using a different hive ( Chickamauga ) kept at UT . Approximately 1 , 300 adult worker bees were collected from brood frames from a single hive ( Nov . 1 , 2016 ) . The bees were immobilized at 4°C and marked with green or pink Testors paint . The bees were separated into two groups , control and treatment , pink and green , respectively , and were distributed to cup cages . The cup cages were maintained in growth chambers at 35°C and 90% relative humidity to simulate hive conditions . Each cup cage contained 30 bees , with 22 replicates for each condition . Control bees were fed filter-sterilized 0 . 5 M sucrose syrup , and treatment bees were fed 450 ug/ml of tetracycline suspended in filter-sterilized 0 . 5 M sucrose syrup . After 5 d , the remaining bees ( 622 control and 623 treatment ) were returned to their original hive . After reintroduction to the hive , post-treatment survival in the hive was assessed by counting bees on d 3 . We individually captured bees and temporarily kept them in cup cages ( 40 per cup ) until all marked bees had been recovered . A single brood frame was removed from a hive ( Big Top ) at the UT campus and was placed in a growth chamber at 35°C and 90% humidity . Pupae were allowed to emerge naturally , and newly emerged adults ( NEWs ) were collected after 24 h . Approximately 600 NEWs were marked with yellow Testors paint and returned to their original hive in order to naturally acquire their microbiota . After 7 d in the hive , approximately 450 bees were recaptured . They were then briefly immobilized at 4°C and split into two groups , control and treatment . These bees were placed in cup cages and maintained in growth chambers . Each cup cage contained approximately 22 bees ( ten replicates for each condition ) . Control bees were fed filter-sterilized 0 . 5 M sucrose syrup , and treatment bees were fed 450 ug/ml of tetracycline suspended in filter-sterilized 0 . 5 M sucrose syrup . Any dead bees were removed during a daily census . After 5 d , ten bees were sampled from each treatment group ( one from each cup cage ) , placed in 100% ethanol , and stored at 4°C , and the remaining bees were chilled and randomly redistributed into new cup cages . The bees were then allowed to “recover” under the following conditions i ) exposed , i . e . , with normal workers collected from their hive ( 15 per cup cage ) or ii ) sterile , i . e . , kept only with other tetracycline-treated bees from their cohort . Ten bees were collected from each of the four groups ( i ) control exposed , ii ) control sterile , iii ) treatment exposed , iv ) treatment sterile at time-points of 0 , 3 , 5 , and 7 d following antibiotic treatment ( Nov . 3–10 , 2015 ) placed in 100% ethanol and stored at 4°C . Within 2 wk of collection , the bees were removed from cold ethanol , and the entire gut from crop to rectum was homogenized and placed into a bead-beating tube in 500 uL of 100% molecular grade ethanol and stored at −20°C . Dissections were performed with flame-sterilized forceps under aseptic conditions . DNA was extracted from the guts ( see below ) , Illumina-based amplicon profiling of 16S rRNA gene copies was performed , and the total community size pre- and post-treatment was quantified by qPCR using the number of 16S rRNA gene copies ( adjusting for number of rRNA operons per genome ) . In order to compare the differences in survivorship effects in the hive and in the lab , the survivorship of the lab-kept bees was also measured on d 3 post-treatment . A brood frame was collected from a hive at UT ( Blueberry ) . Late-stage pupae ( eyes pigmented but pupae lacking movement ) were removed from these frames and placed on cotton pads in sterile plastic bins . These were placed in growth chambers at 35°C and high humidity ( ∼90% relative humidity ) to simulate hive conditions , and pupae were allowed to eclose naturally . After eclosure , NEWs were briefly immobilized at 4°C , and bins were combined to randomize potential age variation . The NEWs were then placed in cup cages and maintained in growth chambers . Each cup cage contained approximately 13 bees ( four replicate cups for each condition ) . Control bees were fed filter-sterilized 0 . 5 M sucrose syrup and sterile bee bread , and treatment bees were fed 450 ug/ml of tetracycline suspended in filter-sterilized 0 . 5 M sucrose syrup and sterile bee bread . Any dead bees were removed during a daily census . After 5 d , tetracycline treatment was arrested , and all bees were fed filter-sterilized 0 . 5 M sucrose syrup and sterile bee bread . Survivorship of the germ-free bees was measured on d 3 post-treatment by counting the total number of bees alive for each group ( controls and treatments ) . To extract DNA from bee tissues , we used bead-beating with cetyltrimethylammonium bromide ( CTAB ) , method as described in [11] with the following modifications: after bead-beating , the samples were allowed to sit briefly at 56°C while the foam settled . We then added 1 ul RNase A solution ( Sigma ) and vortexed the tubes briefly and left them overnight at 56°C . We then added 0 . 75 ml phenol-chloroform-isoamyl alcohol ( 25:24:1 ) ( Ambion , Austin , TX , USA ) to each tube , shook for 30 s before placing on ice for at least 2 min , and then centrifuged at full speed for 30 min at 4°C . The aqueous phase was precipitated in alcohol , washed , and air-dried , then resuspended in 50 ul nuclease-free water . We amplified total copies of the 16S rRNA gene using universal bacterial primers with an Eppendorf Mastercycler ep realplex instrument ( Eppendorf , Hauppauge , NY , USA ) . The forward primer was 27F ( 5’-AGAGTTTGATCCTGGCTCAG-3’ ) , and the reverse primer was 355R ( 5’-CTGCTGCCTCCCGTAGGAGT-3’ ) . Reactions ( 10 ul ) were carried out in triplicate with 5 ul iTaq universal SYBR Green ( Bio-Rad , Inc . ) , 1 ul ( each ) 3 uM primer , 2 ul H2O , and 1 ul of template DNA that had been diluted 100X . The PCR cycle was 95°C ( 3 min ) followed by 40 cycles of 95°C ( 3 s ) and 60°C ( 20 s ) . Using standard curves from amplification of the cloned target sequence in a pGEM-T vector ( Promega , Madison , WI , US ) , we calculated absolute copy number for the reaction template then adjusted this based on dilution to calculate the total copy number for each sample . PCR amplifications of the V4 region of the 16S rRNA gene were performed in triplicate using 515F and 806R primers , as described previously [51] . Reaction products were purified with AMPure XP Beads ( Beckman Coulter ) . The resulting amplicons were subjected to Illumina sequencing on the MiSeq platform ( 2x250 sequencing run ) at the Genome Sequencing and Analysis Facility at UT . Illumina sequence reads were processed in QIIME [52] . FASTQ files were filtered for quality with split_libraries_fastq . py allowing a minimum Phred quality score of Q20 . Forward and reverse Illumina reads were joined using join_paired_ends . py with default settings . Chimeric sequences were removed using the usearch6 . 1 detection method implemented in the identify_chimeric_seqs . py script in QIIME . OTUs were clustered at 97% and 99% using the UCLUST algorithm as implemented in pick_open_reference_otus . py . Briefly , sequence reads were initially clustered against the July 2015 release of the SILVA [53] reference data set ( http://www . arb-silva . de/download/arb-files ) . Sequences that did not match the SILVA data set were subsequently clustered into de novo OTUs with UCLUST . Unassigned , mitochondrial , and chloroplast reads were removed from the dataset . To eliminate pyrosequencing errors all OTUs present in less than 0 . 1% abundance were removed . Because the currently available curated 16S rRNA sequence databases do not contain reference sequences for the core species of the honeybee gut microbiota , taxonomic assignment was performed using a local BLAST database of reference honeybee bacteria 16S rRNA gene sequences . Downstream analyses including alpha and beta diversity estimations were conducted using the QIIME workflow core_diversity_analysis . py , with a sampling depth of 5 , 000 reads per sample and default parameters . Rarefaction depths were chosen manually to exclude samples with exceptionally low total sequences ( see S11 Data for sample details ) . The absolute abundance of each bacterial species was estimated by multiplying the total number of 16S rRNA genes ( measured by qPCR and adjusting for rRNA operons per genome ) by the percent relative abundance of each species . The number of 16S rRNA operons was determined using the reference genome for a given bee gut bacterial species when available . When complete genomes were not available , the mean 16S rRNA operon copy number for the bacterial genus or family was obtained from the rrnDB ( https://rrndb . umms . med . umich . edu/ ) and used as an estimation of the copy number ( S1 Table ) . Statistical tests were performed using the Wilcoxon rank sum test and the Chi-squared test implemented in R . The hive ( Big Top ) recovery bees ( all treatment and control bees from d 0 , 3 , 5 , and 7 ) were screened for the presence of Fungi using diagnostic PCR with the universal fungal primers ITS1-F ( 5'-CTTGGTCATTTAGAGGAAGTAA-3' ) and LR3- R ( 5'-GGTCCGTGTTTCAAGAC-'3 ) [54 , 55] . PCR assays included positive ( purified Saccharomyces cerevisiae DNA ) and negative ( ddH2O ) controls . PCR amplification of DNA products was performed using the Taq Polymerase ( TaKaRa , Japan ) . PCR amplification was carried out as follows: 94°C for 4 min and 40 cycles of 30 s at 94°C , 30 s at 55°C , and 1 min at 72°C; and 72°C for 10 min . After completion of the PCR , 5 μL of the samples was electrophoresed on a 2% agarose gel to determine whether the DNA of interest was amplified . The amplified products were then purified using AMPure XP Beads ( Beckman Coulter ) and cloned using the pGEM-T Easy Vector Systems ( Promega , Madison , WI , US ) according to the manufacturers instructions . The recombinant plasmids were transformed into the competent E . coli DH5α . The transformation was plated on LB Agar containing 100 mg/l each of ampicillin , IPTG and X-gal . White colonies were screened using PCR with plasmid insert specific primers T7 ( 5´-TAATACGACTCACTATAGGG-3´ ) and SP6 ( 5´-ATTTAGGTGACACTATAG-3´ ) . PCR amplification was carried out as follows: 95°C for 2 min . and 28 cycles of 10 sec . at 95°C , 20 sec . at 46°C , and 90 sec . at 68°C; and 68°C for 1 min . A total of 20 clones were selected randomly , and inserts were sequenced at both ends , using Sanger sequencing services at the DNA Sequencing Facility ( DSF ) at UT . After trimming , 34 high quality end sequences were retained . These end-reads were used as queries in BLASTn searches against the NCBI non-redundant nucleotide database . A honeybee package was purchased from a commercial apiary that does not treat with antibiotics or other chemicals . The package of workers was kept in the lab for approximately one month and was provided with 0 . 5 M sucrose solution . Hundreds of honeybees were dissected at different time-points , homogenized , and preserved in 20% glycerol at −80°C . Homogenized guts were then plated onto heart infusion agar plates with sheep’s blood and placed at 37°C in a 5% CO2 chamber overnight . Single bacterial colonies were isolated and stored in 20% glycerol at -80°C . Purified isolates were then amplified using PCR with universal bacterial primers 27F and 1492R [56] and Sanger-sequenced at the DNA Sequencing Facility ( DSF ) at UT . Sequences were identified by BLAST searches against the nonredundant nucleotide database of NCBI . One Serratia strain was chosen as a potential pathogen for infection experiments ( kz11 ) . Serratia kz11 was tested for tetracycline resistance based on the Minimum Inhibitory Concentration ( MIC ) technique using 0 . 016–256 μg/mL E-test strips ( bioMérieux ) . In brief , Serratia kz11 was plated onto LB plates . When the surface of each plate had dried , one E-test strip was put on each plate . The plates were incubated with the lid-side up at 37°C for approximately 24 h . MICs were read directly from the test strip according to the instructions of the manufacturer , where the elliptical zone of inhibition intersected with the MIC scale on the strip . The MIC was >32 μg/ml for Serratia kz11 . We note that nonresistant Serratia strains are known , with MIC < 1 [57] . A single brood frame was removed from a hive ( Whiskeytown ) at the UT campus and was placed in a growth chamber at 35°C and 90% humidity . Pupae were allowed to emerge naturally , and NEWs were collected after 24 h . A total of 720 NEWs emerged and were distributed into cup cages and fed with freshly prepared worker hindgut homogenate in addition to their food , which allows the NEWs to acquire their core microbiome composition [11] . Each cup cage contained approximately 45 bees ( 16 replicate cups ) . After 6 d , the bees were immobilized at 4°C , separated into cohorts ( control and treatment ) , and distributed to cup cages . The cup cages were maintained as described above in growth chambers simulating hive conditions . Each cup cage contained approximately 43 bees , with eight replicates for each condition . Control bees were fed filter-sterilized 0 . 5 M sucrose syrup and treatment bees were fed 450 ug/ml of tetracycline suspended in filter-sterilized 0 . 5 M sucrose syrup . After 5 d , the tetracycline treatment was stopped . One day later , the bees were chilled , randomly redistributed into new cup cages , and exposed to Serratia kz11 or fed only sterile sugar syrup and sterile bee bread . This resulted in a total of four groups with approximately 150 bees ( five replicate cups with approximately 30 bees per cup ) in each group: 1 ) control ( no Serratia ) , 2 ) control + Serratia kz11 , 3 ) post-treatment ( no Serratia ) , 4 ) post-treatment + Serratia kz11 . Bees were censused every day for ten d following bacterial exposure . Kaplan—Meier survival curves were generated in GraphPad Prism version 7 . 0b for Mac OS X , GraphPad Software , La Jolla , California USA , www . graphpad . com . Statistical analyses were performed in R using the Cox proportional hazard model ( coxhp ) implemented in the “survival” package [30] . Approximately 600 adult worker bees were collected from a brood frame from a hive ( Big Top ) kept at UT ( Jul . 21 , 2016 ) . The bees were immobilized at 4°C , separated into cohorts ( control and treatment ) , and distributed to cup cages of previously described design [50] . The cup cages were maintained as described above in growth chambers simulating hive conditions . Each cup cage contained 20 bees , with 15 replicates for each condition . Control bees were fed filter-sterilized 0 . 5 M sucrose syrup and treatment bees were fed 450 ug/ml of tetracycline suspended in filter-sterilized 0 . 5 M sucrose syrup . After 5 d , the tetracycline treatment was stopped . One day after tetracycline treatment was arrested the bees were chilled , randomly redistributed into new cup cages , and exposed to Serratia kz11 isolated from honeybee guts or were fed only sterile sugar syrup and sterile bee bread . This resulted in a total of four groups with 100 bees ( five replicate cups with approximately 20 bees per cup ) in each group: 1 ) control ( no Serratia ) , 2 ) control + Serratia kz11 , 3 ) post-treatment ( no Serratia ) , and 4 ) post-treatment + Serratia kz11 . Censusing and statistical analyses were the same as for the experiment on age-controlled bees , described above . Approximately 1 , 800 adult worker bees were collected from a brood frame from a hive ( Whiskeytown ) kept at the UT campus ( Nov . 14 , 2016 ) . The bees were immobilized at 4°C , separated into cohorts ( control and treatment ) , and distributed to cup cages . The cup cages were maintained as described above in growth chambers simulating hive conditions . Each cup cage contained 45 bees , with 20 replicates for each condition . Control bees were fed filter-sterilized 0 . 5 M sucrose syrup and treatment bees were fed 450 ug/ml of tetracycline suspended in filter-sterilized 0 . 5 M sucrose syrup . After 5 D , the tetracycline treatment was stopped . One day after tetracycline treatment was arrested ( Day 1 ) , the bees were chilled , randomly redistributed into new cup cages , and exposed to i ) S . alvi wkB2 , ii ) Lactobacillus sp . wkB8 , iii ) E . coli K-12 , iv ) Serratia kz11 , or v ) no bacteria ( controls ) . This resulted in a total of ten groups with approximately 170 bees ( six replicate cups with approximately 28 bees per cup ) in each group: 1 ) control ( no Serratia ) , 2 ) control + Serratia kz11 , 3 ) control + Lactobacillus sp . wkB8 , 4 ) control + S . alvi , 5 ) control + E . coli , 6 ) post-treatment ( no Serratia ) , 7 ) post-treatment + Serratia kz11 , 8 ) post-treatment + Lactobacillus sp . wkB8 , 9 ) post-treatment + S . alvi , and 10 ) post-treatment + E . coli . Censusing and statistical analyses were the same as for the experiment on age-controlled bees , described above . Serratia kz11 and E . coli K-12 were grown in liquid LB media at 37°C overnight . S . alvi wkB2 was grown in Insectagro ( Corning ) for 3 D at 37°C in 5% CO2 . Lactobacillus sp . wkB8 was grown in MRS broth for 3 D at 37°C in 5% CO2 . The 600 nm optical densities for each bacterial culture were measured , and cells were washed three times with PBS and diluted to a concentration of 0 . 5 OD in sterile sugar syrup or in PBS . The bacteria—PBS solutions were applied to sterile bee bread ( gamma-irradiated ) in feeding troughs that were placed in the cup cages and the bacteria—sugar syrup solutions were administered in feeding vials on Day 1 post—tetracycline treatment . In order to determine how long viable bacterial cells remained on the bee bread , feeding troughs ( ten per bacterial treatment ) were filled with sterile bee bread and inoculated with a concentration of 0 . 5 OD of bacteria suspended in PBS . Control bee bread was inoculated with PBS only . Each day , one trough was sampled , mixed with 500 ul of PBS and plated out in triplicate ( LB agar for Serratia kz11 and E . coli K-12 , heart infusion agar ( Difco ) supplemented with 5% sheep’s blood for S . alvi wkB2 , MRS agar for Lactobacillus sp . wkB8 , and one of each for controls ) . The plates were checked for bacterial growth after 24 h at 37°C ( Serratia and E . coli ) or 72 h at 37°C in 5% CO2 ( S . alvi and Lactobacillus ) . Viable cells were detected for i ) Serratia kz11 up to Day 2 , ii ) Lactobacillus sp . wkB8 up to Day 3 , iii ) S . alvi wkB2 up to Day 3 , and iv ) E . coli K-12 up to Day 4 . Viable cells were never isolated from control ( l ) bee bread . 16S rRNA gene reads are deposited with NCBI Sequence Read Bioproject: PRJNA338694 . | There is growing evidence for the importance of gut microbes in animal health . Unlike most other insects , honeybees possess a highly conserved gut microbial community , which is acquired through social contact , and several results have suggested that these microbes play an important role in honeybee health . Antibiotics , which can severely disrupt gut microbial communities , are commonly used in beekeeping in several countries . However , it is unknown how antibiotic treatment affects the gut microbial communities of honeybees . Here , we evaluated the effects of antibiotic treatment on the size and composition of the honeybee gut microbiome and on honeybee health . We found that exposure to antibiotics significantly alters the honeybee gut microbial community structure and leads to decreased survivorship of honeybees in the hive , likely due to increased susceptibility to infection by opportunistic pathogens . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"antimicrobials",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"gut",
"bacteria",
"microbiome",
"pathology",
"and",
"laboratory",
"medicine",
"honey",
"bees",
"pathogens",
"drugs",
"microbiology",
"animals",
"tetracyclines",
"antibiotics",
"enterobacteriaceae",... | 2017 | Antibiotic exposure perturbs the gut microbiota and elevates mortality in honeybees |
Holometabolous insects stop feeding at the final larval instar stage and then undergo metamorphosis; however , the mechanism is unclear . In the present study , using the serious lepidopteran agricultural pest Helicoverpa armigera as a model , we revealed that 20-hydroxyecdysone ( 20E ) binds to the dopamine receptor ( DopEcR ) , a G protein-coupled receptor , to stop larval feeding and promote pupation . DopEcR was expressed in various tissues and its level increased during metamorphic molting under 20E regulation . The 20E titer was low during larval feeding stages and high during wandering stages . By contrast , the dopamine ( DA ) titer was high during larval feeding stages and low during the wandering stages . Injection of 20E or blocking dopamine receptors using the inhibitor flupentixol decreased larval food consumption and body weight . Knockdown of DopEcR repressed larval feeding , growth , and pupation . 20E , via DopEcR , promoted apoptosis; and DA , via DopEcR , induced cell proliferation . 20E opposed DA function by repressing DA-induced cell proliferation and AKT phosphorylation . 20E , via DopEcR , induced gene expression and a rapid increase in intracellular calcium ions and cAMP . 20E induced the interaction of DopEcR with G proteins αs and αq . 20E , via DopEcR , induced protein phosphorylation and binding of the EcRB1-USP1 transcription complex to the ecdysone response element . DopEcR could bind 20E inside the cell membrane or after being isolated from the cell membrane . Mutation of DopEcR decreased 20E binding levels and related cellular responses . 20E competed with DA to bind to DopEcR . The results of the present study suggested that 20E , via binding to DopEcR , arrests larval feeding and promotes pupation .
The post-embryo development of holometabolous insects involves larval , pupal , and adult stages . The transformation from the final instar larva to the adult is called metamorphosis . During metamorphosis , the larvae stop eating , start wandering , and finally become quiescent before pupating . The insect molting hormone 20-hydroxyecdysone ( 20E ) promotes metamorphosis by upregulating 20E-pathway gene expression [1] and by counteraction with the juvenile hormone [2] and insulin [3] . However , the regulatory mechanism by which larvae stop feeding is unclear . 20E initiates gene expression by binding to its nuclear receptor ecdysteroid hormone receptor B1 ( EcRB1 ) , which forms a transcription complex with ultraspiracle protein ( USP1 ) and binds to the ecdysone response element ( EcRE ) [4] . However , as the mammal estrogen transmits signal via cell membrane receptor [5] , 20E also induces signaling via G protein-coupled receptors ( GPCRs ) . In Bombyx mori , 20E , via an unidentified GPCR , increases the intracellular Ca2+ level in the anterior silk gland [6] and activates the protein kinase C ( PKC ) pathway [7] . In Helicoverpa armigera , ErGPCR-1 [8] and ErGPCR-2 [9] transmit 20E signals in the cell membrane . 20E , via GPCR , Phospholipase C ( PLC ) , and calcium-signaling pathways , regulates protein phosphorylation , including that of USP1 [10] , cyclin dependent kinase 10 ( CDK10 ) [11] , and catalytic domain of protein kinase A ( PKAC1 ) [12] to form the 20E transcription complex EcRB1/USP1 and promote gene expression during insect metamorphosis [10 , 13] . GPCRs play fundamental roles in mediating various cellular responses from the extracellular environment , including ions , light , amines , odorants , lipids , peptides , amino acids , and nucleotides [14] . However , whether GPCRs are the cell membrane receptors for 20E or other steroid hormones remains controversial because a lack of direct evidence of GPCRs binding to steroid hormones . For example , human GPER ( GPR30 ) [15] binding to Alexa 633-labeled estrogen was assayed using the COS7 cells that overexpress GPR30 [5] . The Drosophila dopamine receptor ( DmDopEcR ) binding of the 20E analog tritium-labeled ponasterone A ( [3H]Pon A ) , was assayed using the cell membranes of Sf9 cells that overexpress DmDopEcR [16] . To date , there is no direct evidence to show that an isolated GPCR can bind a steroid hormone in vitro . Dopamine receptors are typical GPCRs that localize to the cell membrane [17] . In vertebrates , dopamine receptors are mainly expressed in the central nervous system ( CNS ) , and are distributed in non-CNS tissues , such as in epicardium of the heart [18] and the nephron in the kidney [19] . Dopamine receptors are implicated in many neurological processes , including motivation , pleasure , cognition , memory , learning , and fine motor control , as well as modulation of neuroendocrine signaling [20] . Dopamine binds to its receptors to regulate important reward-motivated behaviors , including eating [21] and reward-induced eating behavior in humans [22] . The loss of dopamine neurons results in Parkinson’s disease , whereas hyperactive dopaminergic signaling might be a major factor in the positive symptoms of schizophrenia [23] . However , the regulators of dopamine receptors are not fully understood . In insects , injection of dopamine initiates gregarious behavior of Locusta [24] . Dopamine receptors function to modulate phase change in the brain of Locusta , with dopamine receptor-1 induces the gregariousness and dopamine receptor-2 mediates the solitariness [25] . DmDopEcR is necessary for L-dopa-increased sugar sensitivity [26] . Reduced function of DopEcR in Drosophila mushroom bodies ( MB ) resulted in decreased-locomotor activity [27] . DmDopEcR requires ecdysone and dopamine as ligands to regulate the perception of sex pheromones [28 , 29] . DmDopEcR bound either dopamine ( DA ) or [3H]Pon A when DmDopEcR was overexpressed in Sf9 cells . 20E can compete with [3H]Pon A to bind to DmDopEcR; therefore , DmDopEcR is considered as a cell membrane receptor of 20E [16] . However , the functions of dopamine receptors in insects are poorly understood . In the present study , we showed that DopEcR plays dual functions in insect feeding and pupation . DA , via DopEcR , regulates larval feeding and growth by promoting cell proliferation . 20E competes with DA to bind to DopEcR and uses DopEcR as one of its plasma membrane receptors to transmit the 20E signal to promote insect metamorphosis . DopEcR could bind 20E in the cell membrane and in vitro after it was isolated from the cell membrane . This study presents evidence that GPCRs function as cell membrane receptors of steroid hormones , and the interaction between the endocrine system and the nervous system .
To study the function of DopEcR in H . armigera , the developmental expression profiles of DopEcR in tissues were examined . High levels of DopEcR were detected in the brain , which increased during metamorphic molting from the sixth instar ( 48 h to 120 h ) . In addition , DopEcR was detected in the epidermis , midgut , and fat body , and increased during metamorphic molting ( Fig 1A–1C ) , which suggested that DopEcR plays roles in the brain and other tissues . To examine 20E-induced regulation on DopEcR expression , sixth instar 6 h larvae were injected with 20E . The mRNA levels of DopEcR in the head were upregulated by 20E induction in a time and dose-dependent manner , as assessed using quantitative real-time reverse transcription PCR ( qRT-PCR ) analysis ( Fig 1D ) . Western blot analysis also detected significant upregulation of DopEcR expression in the larval head in response to 20E induction in a time and dose-dependent manner ( Fig 1E and 1F ) . These results showed that 20E upregulates the expression of DopEcR . We analyzed the 20E titer in whole body and the DA titer in hemolymph from 3rd instar larvae to adults to analyze the roles of 20E and DA in larval development . The results showed that the 20E titer was low in earlier instar larvae . However , the 20E titer increased markedly from 6th instar 72 h to the late pupal stage . The highest titer was 6 μM ( 2 . 9 μg/g larval body weight ) , which was detected in the pupal stage ( Fig 2A ) . By contrast , the DA titer was high in earlier instar larvae , and declined from the 6th instar at 0 h to 120 h ( Fig 2B ) , suggesting opposite functions of 20E and DA in metamorphosis and larval feeding and growth . To examine the influence of 20E on larval food consumption , 500 ng of 20E was directly injected into each H . armigera larva at the sixth instar 6 h stage and food consumption and increment weight of body were monitored . 20E injection decreased food consumption from 21% to 30% from 24 h to 60 h ( Fig 2C ) . The average body weight increased at a significantly slower rate by 20E injection compared with that in larvae treated with dimethyl sulfoxide ( DMSO ) ( Fig 2D ) . These data suggested that the 20E repressed larval food consumption , resulting in weight reduction . In addition , we injected the inhibitor of dopamine receptor , flupentixol [16] , into the larva at sixth instar 6 h to reach the final concentration of 50 μM . The average feeding quantity and the increase in body weight decreased compared with those in the PBS-treated controls ( Fig 2E and 2F ) , suggesting that DopEcR functions in larval feeding and growth . To determine the role of DopEcR in larval feeding and development , we knocked down DopEcR by feeding first-instar larvae with Escherichia coli ( HT115 ) that expressed dsGFP or dsDopEcR , respectively . DopEcR was significantly knocked down in the epidermis , midgut , fat body , and head in larvae feed with feeding dsDopEcR-expressing E . coli ( Fig 3A ) . The death rate in the dsDopEcR group and dsGFP group showed no significant difference; however , the time to shorten the larval body was delayed ( the larval body was shortened at the normal prepupal stage ) ( Fig 3B ) . Western blotting demonstrated the efficacy of RNAi ( Fig 3C ) . 75% of the larvae showed delayed pupation for approximately 44 h ( Fig 3D ) . In addition , larval feeding was repressed and the highest consumption of food was postponed by 48 h ( Fig 3E ) . The average body weight and body length were also decreased and postponed for 48 h ( Fig 3F and 3G ) . The time to reach maximum body weight and body length was delayed for about 48 h in the dsDopEcR-treated animals . There was no significant difference in the final body weight and body length between the larvae treated with dsDopEcR and those treated with dsGFP . These results suggested that DopEcR is necessary either for larval feeding or pupation . To confirm the function of DopEcR in 20E-promoted pupation , dsRNA targeting DopEcR was injected into the sixth instar 6 h larval hemocoel to knock down DopEcR and was followed by 20E induction . DopEcR was knocked down significantly in the larval epidermis , midgut , fat body , and head compared with the level of dsGFP ( Fig 4A ) . After knockdown of DopEcR , the larvae showed delayed pupation , even after 20E injection , compared with the larvae that received dsGFP plus 20E ( Fig 4B ) . Statistical analysis showed that 20E injection accelerated the initiation time of pupation by 29 h on average . However , pupation time was delayed by 43 h after injection of dsDopEcR plus 20E injection , compared with dsGFP plus 20E injection ( Fig 4C ) . After knockdown of DopEcR and injection of 20E , the survival rate was 83% , the normal pupation rate was 9% , and the delayed pupation rate was 74% , with a significant difference compared with that for dsGFP plus 20E control ( p < 0 . 01 ) ( Fig 4D ) . These data suggested that DopEcR has a role in 20E-regulated pupation . Terminal deoxynucleotidyl transferase nick-end-labeling ( TUNEL ) staining and histochemical analyses were performed to show the involvement of DopEcR in 20E-induced apoptosis of the larval midgut and fat body . Red fluorescence was observed in the larval midgut , which was separated from the imaginal midgut in the dsGFP+20E-treated larvae , whereas red fluorescence was not detected , and the imaginal midgut was not formed in the dsDopEcR+20E-treated larvae . Similarly , the fat body of the dsGFP+20E control larvae showed apoptosis signals and initial degradation , while the fat body of the dsDopEcR+20E-treated larvae was still closely arranged and did not exhibit obvious apoptosis signals ( Fig 4E and 4F ) . These data suggested that DopEcR plays a role in 20E-induced apoptosis . The finding that 20E repressed larval feeding activity and thus induced earlier pupation prompted us to investigate the relationship between 20E and dopamine ( DA ) , as well as the roles of DopEcR in the 20E and DA pathways by studying apoptosis and cell proliferation . Compared with cells treated with DMSO and DA , caspase-3 activity was detected in 31% of HaEpi cells ( H . armigera epidermal cells ) after the addition of 5 μM 20E for 72 h , suggesting that 20E induces apoptosis . However , caspase-3 activity was significantly reduced in HaEpi cells treated with dsDopEcR and 5 μM 20E , compared with that in dsGFP and 20E treated cells ( Fig 5A and 5a ) , suggesting that 20E induces apoptosis via DopEcR . In contrast , the proliferative signal 5-ethynyl-2'-deoxyuridine ( EdU ) was detected in 33% of DA ( 10 μM ) -treated cells , compared with that in the PBS-treated cells , suggesting that DA ( 10 μM ) promotes cell proliferation . However , a low EdU signal was detected in the dsDopEcR+DA ( 10 μM ) -treated cells compared with those treated with dsGFP+DA ( 10 μM ) . 20E did not induce a proliferative signal but repressed the DA-induced cell proliferation ( Fig 5B and 5b ) suggesting that DA promotes cell proliferation via DopEcR , and 20E opposes DA’s function . Phosphorylation of protein kinase B ( AKT ) and CDK10 were examined to further confirm the antagonism of 20E and DA functions . Western blotting showed DA ( 10 μM ) induced the phosphorylation of AKT , but knockdown of DopEcR blocked DA-induced AKT phosphorylation significantly ( Fig 5Ca ) , suggesting that DA induces AKT phosphorylation via DopEcR . 20E ( 5 μM ) repressed DA-induced AKT phosphorylation; however , DA did not repress 20E-induced CDK10 phosphorylation ( Fig 5Cb and 5Cc ) . These results further confirmed that 20E antagonizes DA’s function . The subcellular location of DopEcR was analyzed to confirm that it is a cell membrane protein . DopEcR was localized in the plasma membrane in the DMSO solvent control by detection using antibodies against H . armigera DopEcR ( Fig 6A ) ; however , DopEcR was not internalized into the cytoplasm within 5 min to 1 h after 20E treatment ( S1 Fig ) . These results suggested that DopEcR exerts its roles on the cell membrane . To address the mechanism by which DopEcR functions in 20E-promoted pupation , the transcript levels of various 20E-responsive genes were examined after knockdown of DopEcR by injecting of dsDopEcR in to 6th instar 6 h larval hemocoel . The transcript levels of key genes in the 20E pathway , including ecdysone nuclear receptor EcRB1 , heterodimeric partner USP1 , and transcription factors HHR3 and BrZ7 , were decreased significantly in larval epidermis after knockdown of DopEcR , compared with that in the dsGFP control , with the housekeeping gene ribosomal protein RpL27 as the internal reference ( Fig 6B ) . Similarly , knockdown of DopEcR by incubating HaEpi cells with dsDopEcR decreased 20E-induced gene expression ( Fig 6C ) . These data suggested that 20E induces gene expression via DopEcR . 20E regulates gene expression via a GPCR-mediated increase in the intracellular Ca2+ concentration to form the EcRB1/USP1 transcription complex [10]; therefore , we detected the involvement of DopEcR in 20E-induced Ca2+ levels . 20E induced rapid Ca2+ intracellular release and extracellular Ca2+ influx . However , after DopEcR knockdown , 20E could not induce rapid Ca2+ release and influx ( Fig 6D and 6E ) . These data suggested that DopEcR participates in 20E-induced cellular Ca2+ increase . A previous study showed that 20E triggers intracellular cAMP increase via ErGPCR-2 to enhance EcRB1/USP1-regulated gene transcription [12] . Therefore , the role of DopEcR in the 20E-induced increase in intracellular cAMP levels was detected in HaEpi cells . Compared with the DMSO-treated cells , the concentration of intracellular cAMP was increased by incubation with 20E ( Fig 6F ) . However , cAMP concentrations decreased after DopEcR knockdown ( Fig 6G ) . These results suggested that 20E increases the intracellular concentration of cAMP via DopEcR . Guanine nucleotide-binding protein subunit alpha S ( Gαs ) stimulates cAMP production and G protein subunit alpha Q ( Gαq ) promotes intracellular Ca2+ increase [30] . To address the mechanism by which 20E increases intracellular Ca2+ and cAMP levels via DopEcR , the protein interaction between DopEcR and Gαq or Gαs was examined by co-overexpression of DopEcR-His and Gαq-RFP-His or Gαs-RFP-His in HaEpi cells . RFP-His and His were overexpressed to exclude the possibility of protein interaction caused by the His- or RFP-His-tag . When Gαq-RFP-His and DopEcR-His were co-overexpressed in the input , DopEcR-His was precipitated together with Gαq-RFP-His under 20E induction using anti-RFP antibodies , but not in the negative control using IgG ( Fig 7Aa ) . Similarly , DopEcR-His was precipitated together with Gαs-RFP-His using anti-RFP antibodies under 20E induction ( Fig 7Ab ) . In the tag control , the His-tag was not precipitated together with RFP-His using the anti-RFP antibodies under 20E induction ( Fig 7Ac ) . These data suggested that DopEcR directly interacts with Gαq and Gαs under 20E induction . 20E induces phosphorylation of USP1 and CDK10 via the ErGPCR-1-mediated PKC pathway for EcRB1/USP1 transcription complex formation and gene transcription [10 , 11] , and the ErGPCR-2-mediated PKA pathway induces phosphorylation of PKAC1 to enhance gene transcription in the 20E pathway [12] . Therefore , we examined the phosphorylation of these proteins to further reveal the involvement and mechanism of DopEcR in the 20E pathway . Western blotting showed that 20E induced USP1 phosphorylation . Lambda protein phosphatase ( λPPase ) treatment degraded the phosphorylation of USP1 . However , DopEcR knockdown significantly suppressed 20E-induced phosphorylation of USP1 , compared with that in the dsGFP control ( Fig 7Ba ) . Similarly , DopEcR knockdown significantly suppressed 20E-induced CDK10 phosphorylation ( Fig 7Bb ) and PKAC1 phosphorylation ( Fig 7Bc ) . These data suggested that 20E induces phosphorylation of these key proteins via DopEcR . 20E regulates gene transcription via the EcRB1/USP1 transcription complex binding to EcREs [4] . To address the role of DopEcR in 20E-induced gene transcription , we examined the binding of EcRB1-RFP-His to EcRE ( GGGGTCAATGAACTG in the 5' regulatory region of Helicoverpa HR3 ) using a chromatin immunoprecipitation ( ChIP ) assay . The qRT-PCR results showed that EcRB1-RFP-His bound more EcRE under 20E treatment than in the DMSO treatment control , using the primers for EcRE located on the EcRE-containing DNA . However , EcRB1-RFP-His bound less EcRE in dsDopEcR treated HaEpi cells , compared with that in the dsGFP-treated cells . The IgG negative control and HR3 primers located in the open reading frame ( ORF ) of HR3 did not produce the same results ( Fig 7C ) . These results suggested that 20E regulates EcRB1/USP1 transcription complex binding to EcRE via DopEcR . Computational docking of 20E to DopEcR was conducted using Surflex-Dock ( SFXC ) in the SYBYL X2 . 0 software ( Certara , Princeton , NJ , USA ) to predict the possibility of DopEcR binding to 20E . ErGPCR-1 [8] and ErGPCR-2 [9] were analyzed to compare the results . The models led to highest score of 20E binding DopEcR , ErGPCR-2 , and ErGPCR-1 near the transmembrane helix ( Fig 8A–8C ) . 20E formed six hydrogen bounds with Tyr-68 , Tyr-109 , Thr-113 , and Trp-160 of DopEcR; three hydrogen bonds with Cys-13 , Ser-113 , and Gly-142 of ErGPCR-2; and one hydrogen bond with Met-228 of ErGPCR-1 ( Fig 8D–8F ) . The scores for DopEcR , ErGPCR-2 , and ErGPCR-1 binding to 20E were 2 . 6433 , 1 . 6124 , and −0 . 5007 , respectively , in which the higher the score , the stronger the combining ability . These results predicted that DopEcR and ErGPCR-2 have a higher binding ability to 20E than ErGPCR-1 . To prove that DopEcR binds 20E , DopEcR-GFP and its mutant DopEcR-M-GFP , which was mutated for possible steroid binding sites ( S1 Table ) based on the prediction of the Surflex-Dock ( SFXC ) program from the SYBYL X2 . 0 software ( Certara , Princeton , NJ , USA ) and I-TASSER online at http://zhanglab . ccmb . med . umich . edu/I-TASSER/ ( S2 Fig ) , were overexpressed in Sf9 cells . GFP was overexpressed as a tag control . ErGPCR-1 [8] and ErGPCR-2 [9] were overexpressed for comparison of the ability of different GPCRs to bind 20E . The overexpressed ErGPCR-1-GFP , ErGPCR-2-GFP , DopEcR-GFP , and their mutants , ErGPCR-2-M-GFP and DopEcR-M-GFP , were confirmed to be located in the cell membrane ( Fig 9A ) . Binding assay by 20-hydroxyecdysone enzyme immunoassay ( 20E-EIA ) showed that the amount of 20E bound by cell membrane from ErGPCR-2-GFP and DopEcR-GFP overexpressing cells increased significantly compared with that bound by the GFP overexpressing cells . However , the amount of 20E bound by cell membranes from ErGPCR-1-GFP , ErGPCR-2-M-GFP , or DopEcR-M-GFP overexpressing cells did not increase compared with that from GFP overexpressing cells ( Fig 9B ) . These results suggested that DopEcR and ErGPCR-2 could bind 20E in the cell membrane . ErGPCR-1-GFP , ErGPCR-2-GFP , DopEcR-GFP , and their mutants , ErGPCR-2-M-GFP and DopEcR-M-GFP , were partially purified to determine their binding capacity to 20E . SDS-PAGE with Coomassie brilliant blue staining showed the GPCRs were partially purified ( Fig 9C ) . A binding assay using 20E-EIA showed that the GFP tag and the isolated ErGPCR-1-GFP bound 20E at a very low level . However , DopEcR-GFP bound 20E in a dose dependent manner , with 5 μg DopEcR-GFP in 50 μL EIA buffer binding 1 . 3 ng of 20E , and 10 μg of DopEcR-GFP in 50 μL EIA buffer binding 2 . 3 ng of 20E . Compared with wild-type DopEcR-GFP , the DopEcR-M-GFP mutant bound less 20E , with 0 . 8–1 . 6 ng of 20E per being bound by 5–10 μg of DopEcR-M-GFP in 50 μL of EIA buffer . Similarly , ErGPCR-2-GFP bound 1 . 4–2 . 5 ng of 20E per 5–10 μg of ErGPCR-2-GFP in 50 μL of EIA buffer . However , the ErGPCR-2-M-GFP mutant bound less 20E , with 0 . 7–1 . 7 ng of 20E being bound per 5–10 μg protein in 50 μL of EIA buffer ( Fig 9D ) . These data suggested that both DopEcR and ErGPCR-2 could bind 20E after they were partially purified from the plasma membrane . A saturation binding curve was constructed using 20E-EIA to further examine the affinity of GPCRs to 20E by calculating the dissociation constants ( Kd ) . The saturable specific binding of DopEcR-GFP to 20E had a Bmax of 9 . 764 ± 0 . 4953 pmol/mg protein and a Kd of 17 . 98 ± 3 . 005 nM . However , the DopEcR-GFP mutant decreased the saturable specific binding , with a Bmax of 6 . 661 ± 0 . 2764 pmol/mg protein and a Kd of 20 . 5 ± 1 . 98 nM ( Fig 10A ) . In comparison , the saturable specific binding of ErGPCR-2-GFP to 20E had a Bmax of 10 . 42 ± 0 . 6629 pmol/mg protein and a Kd of 23 . 32 ± 3 . 304 nM . ErGPCR-2-GFP mutant decreased the Bmax to 7 . 5 ± 0 . 6592 pmol/mg protein and produced a Kd of 29 . 03 ± 5 . 275 nM ( Fig 10B ) . The GPCRs used for saturation binding curve analysis were proven to be highly purified ( Fig 10C ) . The 20E-EIA assay is based on competition between the unlabeled 20E ( 20E bound to GPCR ) and AChE-labelled 20E ( Tracer ) for the limited-specific rabbit anti-20E antiserum; therefore , an inhibition or competitive curve was not detected . These data confirmed that the DopEcR-GFP and ErGPCR-2-GFP could bind 20E . We further addressed the roles of DopEcR-GFP and ErGPCR-2-GFP in 20E pathway by examining cAMP and Ca2+ levels with 20E induction . The HaEpi cells that were overexpressing mutants DopEcR-M-GFP and ErGPCR-2-M-GFP had lower intracellular cAMP levels than that in DopEcR-GFP and ErGPCR-2-GFP overexpressing cells by 20E incubation ( Fig 10D ) . Similarly , 20E induced more Ca2+ intracellular release and extracellular Ca2+ influx in HaEpi cells overexpressing DopEcR-GFP and ErGPCR-2-GFP compared with that in the cells overexpressing DopEcR-M-GFP and ErGPCR-2-M-GFP ( Fig 10E ) . These data suggested that DopEcR and ErGPCR-2-GFP participate in the rapid cellular responses induced by 20E . The competitive binding of DA and 20E to DopEcR was analyzed further to address the relationship of the two ligands to the same receptor . Based the typical kinetic and saturation binding curves simulated in GraphPad Prism using the association kinetics equation , DopEcR to DA had a Kd of 447 . 5 ± 100 . 6 nM ( Fig 10F ) , which suggested that DopEcR could bind DA . The competition displacement binding curves were constructed at a ligand DA concentration equivalent to the Kd ( 450 nM ) and various concentrations of ligand 20E . 20E displaced DA and the calculated inhibition constant ( Ki ) was 42 . 5 ± 8 . 6 nM for 20E ( Fig 10G ) . 20E displacement studies indicated that 20E could displace DA from DopEcR . These data showed that receptor had a much higher affinity for 20E than for DA .
DA receptors play important roles in animal motor function and reward-motivated behaviors , including eating [21] and reward-induced eating behavior [22] in humans . The loss of dopamine neurons results in Parkinson’s disease [23] and Alzheimer’s Disease [31] . We observed that knockdown of DopEcR at the larval growth stage repressed larval feeding , suggesting that DopEcR plays role in larval feeding , and the function of the DA receptor in food consumption is conserved from insect to mammals . Dopamine has been linked to motivated behavior and rewarding reinforcement in fruit flies [32] . Dopamine-receptor signaling pathways play roles in olfactory associations in Drosophila [33] . The repression of DA function might effect the decrease of motor function or reward-motivated behaviors and therefore decrease feeding . 20E reduces food consumption in Bombyx mori [34] , and initiates wandering behavior in Manduca by directly acting on the central nervous system in the brain [35]; however , the mechanism is unclear . We observed that 20E repressed larval feeding and promoted earlier pupation . 20E can bind DopEcR to repress the DA pathway . The 20E titer in H . armigera increased from 0 . 5 μM to 4 . 7 μM from 6th instar 48 h feeding larvae to 6th instar 120 h wandering larvae . Ecdysteroids have a much higher affinity to DopEcR compared with DA , and the larger size of ecdysteroids excludes DA binding to the receptor and inhibits the DA-mediated increase in cAMP levels [16] . The relationship between 20E and DA through the DopEcR is only verified in CHO cells and Sf9 cells [16] . In addition , DopEcR mediates the non-canonical actions of 20E and rapidly modulates Drosophila adult conditioned behavior such as courtship memory through cAMP signaling [28] . In this study , we proposed that 20E blocks larval feeding by competitively binding to DopEcR for 20E signaling . Therefore , larvae stopped feeding and initiated pupation when DopEcR was increased at the metamorphic stages in 6th instar larvae . The decrease of the DA titer before pupation was possible due to the degeneration of the larval brain . The increase of the DA titer during pupal stage was possible related to the remodeling of adult brain for pupal motion , which needs further study . Drosophila DopEcR mutants ( hypomorphic alleles ) can reach adulthood normally . DopEcR functions as a nongenomic ecdysone receptor to control experience-dependent courtship suppression in mushroom body neurons [28] , showing DopEcR mutation is not lethal . In our study , the knockdown of DopEcR caused delayed pupation , repressed larval feeding , growth , body weight , and low death rate , suggesting that DopEcR plays role in larval feeding during growth stage , and functions as 20E receptor during metamorphosis . GPCRs interact with Gαs and Gαq to increase intracellular cAMP and Ca2+ to activate the PKA and PKC pathways , respectively [36] . The GPCR-cAMP-PKA and GPCR-Ca2+-PKC pathways are involved in animal steroid hormone non-genomic pathway signaling [37] . Acute 20E feeding induces a rapid increase in cAMP levels in the MBs via DopEcR , and DopEcR modulates the adult conditioned behavior through cAMP signaling , thus mediating the non-canonical actions of 20E , such as behavioral control [28] . 20E-induced cAMP increases and PKAC1 phosphorylation regulates the phosphorylation of cAMP-response element binding protein ( CREB ) , which binds to the cAMP response element ( CRE ) to enhance 20E-regulated gene expression for pupation in H . armigera [12] . We observed that DopEcR interacts with Gαs and Gαq directly under 20E induction . 20E increased cAMP and Ca2+ levels via DopEcR , induced protein phosphorylation of USP1 , CDK10 , and PKAC1 , and increased the binding of EcRB1 to EcRE . GPCRs transmit the 20E signal in the cell membrane in H . armigera [8 , 9] . Non-genomic GPCRs , Gαq , phospholipase C ( PLC ) 1 , Ca2+ , and the protein kinase C ( PKC ) signaling cascade have been identified in H . armigera , suggesting that activation of the PKC pathway is necessary for USP phosphorylation and that 20E-responsive gene transcription occurs in the 20E genomic pathway [10 , 38] . 20E triggers lysine acetylation of USP1 by activating the GPCR , PLC1 , Ca2+ , and CaMKII signaling pathways , which is essential for the formation of the EcRB1-USP1 transcription complex and gene transcription [13] . ErGPCR-2 knockdown blocks the 20E-induced calcium increase and phosphorylation of USP1 and CDK10 , which represses the formation of the 20E transcription complex EcRB1/USP1 [9] . 20E regulates heterodimeric partner ( USP1 ) phosphorylation via phospholipase C-gamma-1 ( PLCG1 ) and connects the GPCR-mediated non-genomic pathway to the nuclear receptor EcRB1-mediated genomic pathway [9 , 10] . Our data confirm the role of DopEcR in the 20E signaling pathway; therefore , knockdown of DopEcR at later larval stages represses pupation . This is the first identification of the dual functions and mechanism of DopEcR , being either involved in larval feeding or in the 20E pathway for insect pupation . It has been suggested that GPCRs are able to bind steroid hormones in Drosophila [16] and in mammals [15] using cells or cell membranes that overexpress GPCRs . The binding affinities of steroid membrane receptors are orders of magnitude lower than those of nuclear receptors [39] . In addition , the binding affinity of the analog of 20E , [3H]Pon A , to the nuclear receptor complex of EcR/USP is one to two orders of magnitude higher than that of 20E [40] . The Kd value of the in vitro translated EcR/USP heterodimer for Pon A was 0 . 9 nM in Drosophila [41] and 1 . 1 nM in Bombyx [42] , respectively . However , the membranes of the anterior silk gland of B . mori , which contain a putative membrane receptor of ecdysone , mEcR , exhibit saturable binding for [3H]Pon A , with a Kd of 17 . 3 nM and a Bmax of 0 . 82 pmol∙mg-1 protein [7] . The membranes isolated from Sf9 cells that overexpressed DmDopEcR showed saturable specific binding for [3H]Pon A , with a Kd of 10 . 4 ± 0 . 38 nM and a Bmax of 0 . 32 ± 0 . 04 pmol∙mg-1 protein [16] . In this study , we used the 20E-EIA method to detect isolated GPCRs binding 20E directly . We detected that isolated DopEcR exhibited saturable specific binding for 20E , with a Kd of 17 . 98 ± 3 . 005 nM and a Bmax of 9 . 764 ± 0 . 4953 pmol/mg protein . ErGPCR-2 showed saturable specific binding for 20E , with a Kd of 23 . 32 ± 3 . 304 nM and a Bmax of 10 . 42 ± 0 . 6629 pmol/mg protein . The Kd values of DopEcR and ErGPCR-2 have the same orders of magnitude as the Kd detected from the membranes of Sf9 cells , however , the Bmax values of DopEcR and ErGPCR-2 were higher than those detected from the membranes of Sf9 cells . A higher Bmax suggests a higher binding capacity of GPCRs than the membranes of Sf9 cells . This may be due to the different detection methods used , in which we used the 20E-EIA method and isolated GPCRs . The Kd values of isolated DopEcR for 20E ( 17 . 98 nM ) in H . armigera is higher than that of EcR/USP heterodimer for Pon A ( 0 . 9 nM ) in Drosophila [41] , suggesting lower binding affinity of GPCRs to 20E than nuclear receptor , which might be one of the reasons that multiple GPCRs , such as DopEcR and ErGPCR-2 , are involved in 20E signaling on the cell membrane . However , 20E shows higher affinity for DopEcR than DA and can inhibit the function of DA , and induces rapid signaling effects , such as increased Ca2+ and cAMP levels to trigger 20E-pathway . Previous studies demonstrated that ErGPCR-1 and ErGPCR-2 transmit 20E signals in the lepidopteran insect H . armigera , but could not detect the binding of ErGPCR-1 and ErGPCR-2 to the 20E analog [3H]Pon A using the isotope method [8 , 9] . In the present study , we confirmed that ErGPCR-2 can bind 20E using the 20E-EIA method . However , binding of ErGPCR-1 to 20E was not detected in this study . Agonists that exert their roles without measurable binding have been observed in another study [43] . A possible reason is that GPCRs loosely or dynamically binds to the ligands [44 , 45] . In humans , dopamine receptors regulate various physiological processes , including reward , voluntary movement , and hypertension [46] . Insufficient or hyperactive dopaminergic signaling results in disease [23] . Thus , dopamine receptors are common neurological drug targets [47] . Dopaminergic neurotransmitters have been used to treat mental and neurological diseases , such as schizophrenia , bipolar disorder , Parkinson’s disease , Huntington’s disease , and Tourette’s syndrome [48] . Steroid hormone estrogens can regulate neurotransmission in the human central nervous system , including mood , reward , and motivation [49] . We observed that 20E upregulated DopEcR expression and bound to DopEcR to block its activity of DA pathway . In addition to insects , various plants produce 20E , such as Cyanotis vaga . Our finding presents the possibility to explore the use of 20E or other phytosterols to treat dopamine-related diseases . H . armigera DopEcR is a class A ( Rhodopsin-like ) receptor that shares 68% identity with Drosophila DmDopEcR . ErGPCR-1 and ErGPCR-2 belong to class B ( secretin receptors ) ( S3 Fig ) . However , H . armigera DopEcR showed increased expression levels during metamorphosis , which is different from DmDopEcR , which is rarely expressed in third-stage larvae , but is strongly expressed in the first and second larval stages and in adult heads [16] , which suggested a difference between dipteran and lepidopteran insects . 20E induces a rapid increase in cAMP levels in MBs and triggers cAMP signaling via DopEcR to modulate adult conditioned behavior in Drosophila [28] . DopEcR is mainly expressed in the brain; however , ErGPCR-1 and ErGPCR-2 are expressed widely in the epidermis , fat body , and midgut [8 , 9] . DopEcR is localized in the plasma membrane and is not internalized , which is similar to ErGPCR-1 [8] , but differs from ErGPCR-2 , which can be internalized after phosphorylation by GPCR kinase 2 ( GRK2 ) to desensitize 20E signaling [9] . The different expression levels of GPCRs in tissues might explain why various GPCRs transmit 20E signals . That different GPCRs recognize unique positions of the G-protein barcode [50] might also explain why various GPCRs are involved in the same pathway . DopEcR functions as a DA receptor during the larval feeding stage to promote AKT phosphorylation , cell proliferation , larval feeding , and growth . 20E upregulates DopEcR expression during metamorphosis and binds to DopEcR to block its function in larval feeding and cell proliferation . 20E triggers the interaction of DopEcR with Gαq and Gαs , increases intracellular Ca2+ and cAMP levels , and induces protein phosphorylation to regulate 20E-pathway-induced gene expression to effect metamorphosis ( Fig 11 ) . GPCRs can bind 20E and function as 20E cell membrane receptors .
The cotton bollworms , Helicoverpa armigera , which come from the Henan Agricultural University in China , were fed on artificial diet comprised of soybean powder , wheat germ with compound vitamins and mineral salt . And the insects were raised in an insectarium under the cycle of 14 h light: 10 h dark at 26±1 °C with 60% to 70% humidity . A cDNA fragment ( nucleotide sequence 148 bp to 1 , 020 bp , encoding 290 amino acids from 50 to 339 ) of DopEcR was connected with pET30a and expressed in Escherichia coli BL21 ( DE3 ) . The recombinant DopEcR protein formed occlusion bodies and was purified by means of 12 . 5% SDS-PAGE . The recombinant DopEcR protein was cut from the gel for preparing rabbit polyclonal antiserum [8] . Dissected age-appropriate larvae , epidermis , midgut , fat body and brain and homogenized in 500 μL Tris-HCl buffer ( 40 mM , pH 7 . 5 ) on ice with 5 μL phenylmethylsulfonyl fluoride ( PMSF , 17 . 4 mg/mL in isopropyl alcohol ) , respectively . Homogenate was centrifuged for 10–15 min at 4 °C at 12 , 000 rpm . And then the supernatant was collected . The protein concentration was measured by Bradford protein assay [51] . 20 μg proteins were separated by 7 . 5% or 12 . 5% SDS-PAGE and transferred to a nitrocellulose membrane . The membrane was blocked by blocking buffer [Tris-buffered saline ( TBS , 150 mM NaCl solution , 10 mM Tris-HCl , pH 7 . 5 ) with 2–5% skim milk] for 1 h at room temperature . The membrane was incubated in blocking buffer with primary antibodies at 4 °C overnight . The polyclonal antibodies against H . armigera DopEcR were produced in our laboratory . Antibody against H . armigera β-actin , His , GFP , and RFP was purchased from ABclonal ( Cat . AC026 , AE003 , AE012 , and AE020 , Wuhan , China ) . Antibodies against DopEcR , β-actin , His , GFP , and RFP were diluted in 5% skim milk at 1:5 , 000–1:10 , 000 . After being washed three times with TBST ( 0 . 02% tween in TBS ) for 10 min each , the membrane was incubated with 1:10 , 000 diluted secondary antibody , alkaline phosphatase-conjugated AffiniPure Goat Anti-Rabbit/-Mouse IgG ( ZSGB-BIO , Beijing , China ) . After being washed three times with TBST and TBS , the target signals were visualized in 10 mL TBS , 45 μL of P-nitro-blue tetrazolium chloride ( NBT , 75 μg/μL ) and 30 μL of 5-bromo-4-chloro-3 indolyl phosphate ( BCIP , 50 μg/μL ) in the dark for 10–20 min . The membrane was washed by deionized water and made image by software of Photoshop . Bands on the membrane were calculated by software of ImageJ ( National Institutes of Health , http://imagej . nih . gov/ij/download . html ) . Acquired data were analyzed by software of GraphPad Prism 7 ( GraphPad Software , San Diego , CA , USA ) . More than three larvae , pupae , or adults at different developmental stages were collected and weighed . The whole bodies were then freeze-dried for 16 h . Then , 80% methanol was used to grind larvae , pupae , or adults from at least three insect samples at 4 °C to extract the 20E , 1 g tissues per 1 mL methanol . The sample was centrifuged at 10 , 000 × g for 10 min at 4 °C , and the supernatant was collected and evaporated until it dried completely . The pellet was dissolved in 1 mL EIA buffer and diluted 1 , 000 times . A 50 μL sample was used to detect 20E using a 20E Enzyme Immunoassay ( 20E-EIA ) kit ( Bertin Pharma , France ) according to the instruction and operation manual . The hemolymph samples collected from larvae , pupae , or adults were immediately transferred to a 1 . 5 mL polypropylene micro-test tube with 0 . 1 mM Phenylthiourea ( PTU ) , respectively . The sample was centrifuged at 20 , 000 g for 10 min at 4 °C . The supernatant was detected the concentration of DA directly using an insect hemolymphal dopamine ELISA kit ( MLBIO Biotechnology , Shanghai , China ) according to the instruction and operation manual . 20E was dissolved in dimethyl sulfoxide ( DMSO ) to prepare a 10 mg/mL stock solution . The work solution of 20E was diluted with sterile phosphate-buffered saline ( PBS , pH 7 . 4 , 10 mM sodium phosphate , 140 mM NaCl ) at 1:100 . The 6th instar 6 h larvae were selected and injected with 20E at 500 ng/5 μL per larva . The equivalent amount of DMSO was injected into the same stage larvae as control . The larvae were suffered at 1 , 3 , 6 , 12 , and 24 h , respectively after injected 20E or DMSO . The total mRNA or protein was extracted for qRT-PCR or western blot . The data of qRT-PCR were analyzed by software of GraphPad . The bands of western blot were calculated by software of ImageJ and the acquired data were analyzed by software of GraphPad . The total RNA was extracted with Trizol reagent according to the manufacturer’s instructions ( TransGen Biotech , Beijing , China ) , and the first-strand cDNA was synthesized using M-MLV reverse transcriptase according to the manufacturer’s instructions ( TIANGEN , Beijing , China ) . qRT-PCR was performed with a CFX96 real-time system ( Bio-Rad , Hercules , CA , USA ) using 2×SYBR qRT-PCR pre-mixture ( TransGen Biotech , Beijing , China ) and the used primers were listed in S2 Table . The relative expression levels of genes were obtained using H . armigera β-actin as a quantity control . The data from three independent experiments and three technique repeats were obtained by R = 2−ΔΔCT method ( ΔΔCT = ΔCTsample-ΔCTcontrol , ΔCT = CTgene-CTactin ) [52] . RNA interference ( RNAi ) was used to study the DopEcR function because RNAi was widely used in many species of moths in 10 families [53] . Long dsRNA can be processed into smaller fragments [54] to suppress transcription of the target gene in worms [55] . The total RNA was extracted using Trizol reagent on the basis of the manufacturer’s instruction ( TransGen Biotech , Beijing , China ) , and the first-strand cDNA ( 301 bp ) from 297 bp to 598 bp was synthesized with HaDopEcR-RNAiF/ HaDopEcR-RNAiR ( S2 Table ) and M-MLV reverse transcriptase on the basis of the manufacturer’s instructions ( TIANGEN , Beijing , China ) . Transcription was carried out as follows: 1 μg cDNA template was mixed with 20 μL 5 × transcription buffer , containing 80 U T7 RNA polymerase ( Fermentas , Los Angeles , USA ) , 2 . 4 μL 10 mM A/U/C/GTP each ( Fermentas , Los Angeles , USA ) , and 120 U RNasin ( TaKaRa Bio , Otsu , Shiga Prefecture , Japan ) . RNase-free water was added to a volume of 50 μL . After incubation at 37 °C for overnight , 10 μL 40 U RNase-free DNase I ( Fermentas , Los Angeles , USA ) , 10 μL DNAase I Buffer and 30 μL RNase-free water were added and the solution was incubated at 37 °C for 60 min . After extraction with phenol/chloroform and precipitation with ethanol , dsRNAs were resuspended with 50 μL RNase-free water . The purity and integrity of dsRNAs were determined using agarose gel electrophoresis . The dsRNAs were quantified by a Micro-Spectrophotometer ( GeneQuant; Amersham Biosciences ) . RNAi was performed according to our previous article [56] . The continuous ingestion of the bacteria HT115/DE3 containing dsRNAs expressed shows higher interference efficiency compared with one-time injection of H . armigera larvae with dsRNA [57] . Interfering primers ( HaDopEcR-swRNAiF/HaDopEcR-swRNAiR ) ( S2 Table ) were used for PCR amplification of the target fragment ( 301 bp cDNA , from 297 bp to 598 bp ) . The target fragment was inserted into pPD129 . 36 ( L4440 ) vector presented by Dr . Marek Jindra ( Biology Center ASCR , Czech Republic ) . The vector was transformed into DH5α , and positive clones were screened using 50 μL ampicillin ( Amp , 25 mg/mL ) . The recombinant plasmid was transformed into HT115 which provided by Dr . Marek Jindra ( Biology Center ASCR , Czech Republic ) . Positive clones were screened with 50 μL Amp ( 25 mg/mL ) and 6 . 25 μL tetracycline ( Tet , 25 mg/mL ) and inoculated into 5 mL medium ( yeast extract 0 . 5 g , tryptone 1 g , NaCl 1 g , ddH2O 100 mL , pH 7 . 4 ) on programmable incubator shaker at 37 °C overnight . Overnight strains were inoculated into 100 mL medium at a volume of 1: 100 and shaken culture 3 h to OD600 = 0 . 4 . Isopropyl β-D-1-thiogalactopyranoside ( IPTG , 0 . 5 mM ) was added in medium . Bacteria were cultured on programmable incubator shaker at 37 °C for 4 h . Bacterial RNA was extracted and dsRNA expression was detected by 2% agarose gel electrophoresis . The remaining bacteria were suspended in 400 μL PBS . Two groups of cotton bollworm were selected , 30 in each group . The bollworm food was cut into 1 cm × 1 cm × 0 . 2 cm . 10 μL fresh HT115 expressing dsDopEcR with PBS was applied to the feedstuff surface . Larvae were fed with HT115 every day up to 6th-96 h . Control groups were fed with feedstuff containing the same amount of dsGFP expressing HT115 . Total RNA was extracted from the epidermis , midgut , head , and fat body of 6th 72 h instar , respectively . Interference efficiency was detected by qRT-PCR or western blot . The dsRNAs were diluted with PBS to 100 ng/μL . When the 6th instar 6 h larva was stiff and numb on ice , 500 ng dsRNAs were injected into larval hemocoel from the third thoracic legs with a sterile microsyringe . dsRNAs were injected three times 24-hour intervals . In the third injection , 500 ng 20E and 500 ng dsRNAs were injected to larva together . The morphology and behavior of insects were observed since the first dsRNA injection . DNA fragments ( 300–500 bp ) of GPCRs and green fluorescence protein ( GFP ) were PCR-amplified as templates for dsRNA synthesis with the primers ( S2 Table ) . The cells of H . armigera epidermal cell line ( HaEpi ) were established from H . armigera epidermis and cultivated in Grace’s medium with 10% fetal bovine serum ( FBS ) at 27 °C [58] . When HaEpi cells density reached 70% to 80% , the cells were transfected with 2–4 μg dsRNAs and 5–8 μL QuickShuttel-Enhanced transfection reagent ( Biodragon Immunotechnologies , Beijing , China ) in Grace’s medium , respectively . And the cells were cultivated for 48 h at 27 °C . Finally , the cells were re-fed in a fresh Grace’s medium with 20E at a final concentration of 1 μM for 12 h . Equivalent volume of DMSO as control . The average quantity of feeding , body weight , and body length of insects from first instar larvae ( 1-F ) to pupae at 2 day ( P-2 ) were analyzed individually based 30 insects for three repeats after different treatments . The amount of diet actually eaten was estimated by the difference of the diet weight before and after testing twice every day , individually . The average weight and length of body were measured twice every day , individually . For the assay of the food consumption and increment body weight in 20E and flupentixol used the same method . HaEpi cells were washed three times with 500 μL DPBS , and fixed with 4% paraformaldehyde which diluted in PBS for 10 min in dark . The fixed cells were washed six times for 3 min each , and treated with 0 . 2% Triton-X 100 in PBS for 15 min . The cells were washed with DPBS six times for 3 min each , covered with blocking solution ( 2% bovine serum albumin ( BSA ) in PBS ) for 1 h . The above buffer was exchanged with blocking solution contained antibodies against the target protein that was 1:100 diluted in 2% BSA/PBS for overnight at 4 °C . The cells were washed six times with DPBS for 3 min each and then the cells were treated with 200 μL goat anti-rabbit IgG Alexa Fluor 488 ( Introvigen , Carlsbad , CA , USA ) diluted with 1:1 , 000 for 1 h at 37 °C in dark . The cells washed 3 times with DPBS for 5 min each . Plasma membrane was stained with 200 μL Alexa Fluor 594-conjugated wheat germ agglutinin ( WGA ) ( 1:2 , 000 in PBS ) ( Introvigen , Carlsbad , CA , USA ) in dark at room temperature for 10 min . Nuclei were stained with 4’ , 6-diamidino-2-phenylindole ( DAPI , 1 μg/mL in PBS ) ( Sigma , San Francisco , USA ) in dark at room temperature for 10 min . The rehydrated histologic sections were stained with hematoxylin ( 1 g of hematoxylin , 10 mL of ethanol , 20 g of KAl ( SO4 ) 2 , and 200 mL of H2O ) for 10 min . Next , the sections were washed with running water for 1 min and stained with Scott’s liquid for 1 min . The sections were then washed with hydrochloric acid ethanol differentiation medium ( 70% ethanol in 1% hydrochloric acid ) for 20 s and stained with Scott’s liquid for 1 min . Finally , sections were incubated with 0 . 5% water-soluble eosin dye solution for 30 s and subsequently washed with running water . The images were observed using an Olympus BX51 fluorescence microscope ( Olympus Optical Co . , Tokyo , Japan ) . Tunel assay was conducted using the TUNEL BrightRed Apoptosis Detection Kit ( Vazyme Biotech Co . , Nanjing , China ) according to the manufacturer’s instructions . The images were observed using an Olympus BX51 fluorescence microscope ( Olympus Optical Co . , Tokyo , Japan ) . The red fluorescence intensity ( tunnel signal ) was counted by ImageJ and the acquired data were analyzed by software of GraphPad . The cells were interfered by 2 μg dsRNAs or plasmids at 27 °C for 24 h after the density of cells up to 70–80% . Then the cells are treated with fresh medium included AM ester Calcium Crimson dye ( Invitrogen , Carlsbad , CA , USA ) ( final concentration 3 μM ) for 30 min at 27 °C . The cells were washed with Dulbecco’s PBS ( DPBS; 137 mM NaCl , 2 . 7 mM KCl , 1 . 5 mM KH2PO4 , and 8 mM Na2HPO4 , pH 7 . 4 ) for three times . Then the cells were included with 20E in DPBS ( final concentration 1 μM ) , and were automatic photographed once every 6 s for 2 min . After that , DPBS which including CaCl2 ( final concentration 1 mM ) and 20E ( final concentration 1 μM ) , were put into microscope slides . Laser Scan Confocal Microscope Carl Zeiss LSM 700 ( Thornwood , NY , USA ) was used to detect the fluorescence at 555 nm every 6 s for 360 s . Finally , the fluorescence intensity of every collected photo was obtained by Image Pro-Plus software ( Media Cybernetics , United States ) and statistically analyzed . The HaEpi cells were cultured in 6-well plates to 80% density . dsRNAs were knocked down as above method . HaEpi cells were transfected with DopEcR-GFP , DopEcR-M-GFP , ErGPCR-2-GFP , and ErGPCR-2-M-GFP for 36 h . After 36 h , Grace’s medium was replaced with DPBS with 0 . 5 mM 3-isobutyl-1-methylxanthine ( Sigma , St . Louis , MO USA ) . After 30 min , the cells were treated with 2 μM 20E for 0 , 5 , 10 , 15 , 30 , 60 , and 120 min; DMSO-treated cells served as a solvent control . The medium was removed and treated with 0 . 1 M HCl . The concentrations of cAMP in the cells were detected using the cAMP ELISA Kit ( Biodragon , Beijing , China ) according to the manufacturer’s instructions . The HaEpi cells were treated with 5 μM 20E or 10 μM dopamine ( DA ) for 72 h . The 5-ethynyl-2′-deoxyuridine ( EdU ) kit ( Ribobio , Guangzhou , China ) was used to detect cell proliferation according to the manufacturer’s protocol . The NucView 488 caspase-3 assay kit ( NO . 30029 Biotium , Hayward , USA ) was used to detect the activity of caspase-3 in the HaEpi cells , according to the manufacturer’s instructions . The nuclei were stained with DAPI ( 10 μg/mL ) for 10 min at room temperature in dark , and observed with Laser Scan confocal Microscope Carl Zeiss 700 ( LSM 700 ) ( Zeiss , Thornwood , ZY ) . The ORFs of Gαs ( GenBank accession no . MK134004 ) , Gαq ( GenBank accession no . AAX56092 . 1 ) , and DopEcR were inserted into the pIEx-4-RFP-His or pIEx-4-His vectors , respectively . The reconstructed plasmids were then transfected into the HaEpi cells . The cells were incubated with 2 μM 20E for 30 min . DMSO was used as the control . The mouse monoclonal antibody against RFP ( 1 μL ) and radioimmunoprecipitation assay ( RIPA ) buffer ( 0 . 1 M Tris-HCl buffer containing 150 mM NaCl , and 1% Nonidet P-40 , pH 8 . 0 ) ( 400 μL ) was incubated with protein A resin for 30 min at room temperature . The resin was washed with 500 μL RIPA buffer thrice . The protein was extracted from cells using RIPA buffer . The supernatant was harvested by centrifugation at 10 , 000 g for 10 min at 4 °C . The supernatant was added to protein A resin to eliminate nonspecific binding and harvested by centrifugation . The supernatant was added to the resin-antibody complex and incubated for 2–4 h with gentle shaking at 4 °C . The resin was harvested by centrifugation and washed with RIPA buffer thrice . The resin was treated with SDS-PAGE loading buffer and boiled for 10 min . After centrifugation at 12 , 000 g for 2 min ( 4 °C ) , the protein samples were detected by Western blotting with a primary antibodies of mouse monoclonal antibody against RFP and His ( ABclonal , Wuhan , China ) . RFP-His was overexpressed in the HaEpi cells as a negative control as described above . The ORFs of CDK10 ( GenBank accession no . KC188798 ) , USP1 ( GenBank accession no . EU526832 ) , and PKAC1 ( GenBank accession no . KT207930 ) were inserted into the pIEx-4-His vector . The reconstructed plasmids were transfected into the HaEpi cells for 36 h . The cells were then transfected with dsDopEcR or dsGFP for 24 h , followed by incubating with 2 μM 20E or DMSO for 30 min . DMSO was used as the control . The gel concentration of SDS-PAGE was 7 . 5% . 40 μL of protein ( 2 μg/μL ) was incubated with 0 . 5 μL of λPP , 5 μL of buffer , and 5 μL of MnCl2 ( 50 μL total ) at 30 °C for 30 min according to the manufacturer’s specifications ( Millipore , Temecula , CA , USA ) . The sample was boiled for 10 min after adding the SDS-PAGE loading buffer and then detected by Western blot analysis . The ORF of EcRB1 ( GenBank accession no . EU526831 ) was inserted into the pIEx-4-RFP-His vector . The pIEx-4-EcRB1-RFP-His plasmid was transfected into HaEpi cells with a 70% density in six-well plates for 48 h . Then , the cells were transfected with dsDopEcR or dsGFP for 24 h . Finally , the cells were treated with 20E or DMSO for 3 h . The cells were cross-linked with 1% formaldehyde at 37 °C for 10 min and then added glycine to final concentration of 0 . 125 M at 25 °C for 10 min to terminate the cross-linking . The cells were washed twice with PBS and then suspended with SDS lysis buffer ( 1% SDS , 10 mM EDTA , 50 mM Tris-HCl , pH 8 . 1 ) . 200–1 , 000 bp DNA fragments were obtained by sonication . After centrifugation , supernatants were added to the Protein A resin and incubated at 4 °C for 1 h to pre-treat nonspecific binding . After centrifugation , one supernatant was used as an input sample for qRT-PCR . Other supernatants were incubated with anti-RFP antibody or mouse control IgG as a negative control at 4 °C overnight . Protein A resin was added into the immunoprecipitated protein-DNA complex and incubated at 4 °C for 2 h . The complexes were washed once with low-salt wash buffer ( 0 . 1% SDS , 1 . 0% Triton X-100 , 2 mM EDTA , 200 mM Tris-HCl , pH 8 . 0 , 150 mM NaCl ) , high-salt wash buffer ( 0 . 1% SDS , 1 . 0% Triton X-100 , 2 mM EDTA , 20 mM Tris-HCl , pH 8 . 0 , 500 mM NaCl ) , LiCl wash buffer ( 10 mM Tris-HCl , pH 8 . 1 , 0 . 25 M LiCl , 1 mM EDTA , 1% NP-40 , 1% deoxycholate ) and twice with TE buffer ( 10 mM Tris-HCl , pH 8 . 1 , 1 mM EDTA ) . The complexes were then washed with elution buffer ( 1% SDS , 0 . 1 M NaHCO3 ) . The DNA-proteins were reversely cross-linked at 65 °C overnight , followed by RNase A and proteinase K treatments . The DNA was purified by phenol/chloroform extraction and analyzed by qRT-PCR using ChIP F/ChIP R primers ( target to EcRE ) ( S2 Table ) . The ligand 20E was docked into the active site of GPCRs using the Surflex-Dock ( SFXC ) function in the SYBYLx2 . 0 software ( Tripos , St . Louis , MO , United States ) . Final figures were prepared with PyMOL program [59] . The molecular models of GPCRs binding steroid were created using the I-TASSER on-line server ( https://zhanglab . ccmb . med . umich . edu/I-TASSER/ ) . Proofreading DNA polymerase ( TIANGEN , Beijing , China ) was used to amplify the cDNA of GFP , ErGPCR-1 [8] ( full length , GenBank number: JQ809653 . 1 ) , ErGPCR-2 [9] ( 7TM region and added signal peptide , GenBank number: AKA95280 . 1 ) , and H . armigera DopEcR ( Gene Bank number: MG596302 ) using PCR with the primers ( S2 Table ) . The resulting cDNAs were inserted into pIEx-4-GFP-His plasmid ( pIEx-4 plasmid , Merck , Darmstadt , Germany , fused with GFP and a His-tag at the C-terminus by our laboratory ) . After the density of HaEpi cells to 70%–80% , the plasmid ( 5 μg/mL ) was transfected into HaEpi cells using QuickShuttle-Enhanced transfection reagent ( Biodragon Immunotechnologies , Beijing , China ) in 2 mL Grace’s medium with 10% bovine serum at 27 °C for 48 h . Sf9 cells were utilized to expression GPCRs in insect cell culture medium ESF 921 ( Expression Systems , Davis , California , USA ) with 2% bovine serum because Sf9 expressed more exogenous protein than HaEpi cells . The GPCR binding sites were predicted online at http://zhanglab . ccmb . med . umich . edu/I-TASSER/ ( S1 Table ) . GFP , GPCRs ( ErGPCR-1 , ErGPCR-2 and DopEcR ) and their mutants [ErGPCR-2-M: Serine 113 to alanine ( S113A ) , Cysteine 138 to alanine ( C138A ) , Glycine 142 to alanine ( G142A ) ; DopEcR-M: Tyrosine 68 to alanine ( Y68A ) , Tyrosine 109 to alanine ( Y109A ) , Threonine 113 to alanine ( T113A ) , Tryptophan 160 to alanine ( W160A ) ] were amplified via bridged PCR method with the site mutated primers ( The primers were based on the sequence around the mutated site , which were not showed because too many of the primers were used ) . The mutation was confirmed by DNA sequence . DopEcR-GFP , ErGPCR-1-GFP , ErGPCR-2-GFP and their mutants , ErGPCR-2-M-GFP and DopEcR-M-GFP were overexpressed in HaEpi cells in a 25-cm2 cell culture bottle , respectively . The cells were washed with DPBS for 2 min twice . The cells were incubated in Grace’s medium containing 1 μM 20E for 5 min at 27 °C to allow 20E to bind to the cell membrane . Cells were collected by centrifugation at 1 , 700 × g and 4 °C for 5 min and the pellet was resuspended in 500 μL enzyme immunoassay ( EIA ) buffer ( Bertin Pharma , Paris , France ) and sonicated for 5 min . After centrifugation at 48 , 000 × g and 4°C for 1 h , the pelleted cell membrane debris was resuspended in 100 μL EIA buffer . 50 μg cell membrane proteins with fixed 20E in 50 μL EIA buffer was added with 450 μL EIA buffer and used to quantify 20E . A 20-Hydroxyecdysone Enzyme Immunoassay kit ( 20E-EIA kit ) ( Bertin Pharma , Paris , France ) was used to detect cell membrane bound-20E according to the manufacturer’s instructions . GPCRs were overexpressed in Sf9 cells in a 25-cm2 cell culture bottle using pIEx-4-GFP-His plasmid . After 48 h , total plasma membrane proteins were extracted using the kit of the cell transmembrane protein extracts ( BestBio , Shanghai , China ) and GFP were extracted using the RIPA Lysis Buffer ( 20 mM Tris , pH 7 . 5 , 150 mM NaCl , 1% Triton X-100 ) without EDTA ( Ethylenediaminetetraacetic acid ) ( Beyotime , Shanghai , China ) . 100 μL slurry of Ni2+ resin was washed three times with charge buffer ( 50 mM NiSO4 ) and once with binding buffer ( 0 . 5 M NaCl , 20 mM Tris-HCl , pH 7 . 9 , 5 mM imidazole ) for 5 min . The overexpressed GPCRs were bound on the washed Ni2+ resin ( GE Healthcare , Pittsburgh , USA ) . Each 100 μL slurry of Ni2+ resin was added to 150 μg/50 μL of the plasma membrane proteins with 1 mL binding buffer on a four-dimensional rotating mixer for 40 min at 26 °C . Then , the Ni2+ resin was washed three times with binding buffer followed by three times with wash buffer ( 1 M NaCl , 20 mM Tris-HCl , pH 7 . 9 ) , each for 5 min . After centrifugation at 500 × g for 1 min at 4 °C , the GPCR-resin were incubated with 1 μM 20E in 1 mL binding buffer for 5 min . The GPCR and bound 20E were washed three times with wash buffer , each for 5 min . The GPCR and fixed 20E were eluted using 50 μL elute buffer ( 0 . 5 M NaCl , 20 mM Tris-HCl , pH 7 . 9 , 75 mM imidazole ) and then diafiltration were carried out three times with EIA buffer using Amicon Ultra-15 ( Merck Millipore , Temecula , California , United States of America ) to reduce the concentration of imidazole for the following experiment . The concentration of the isolated GPCR was detected as 1˗1 . 5 μg/μL using the Bradford method . The 20E bound by 5–10 μg/5–7 μL eluted GPCR was added EIA buffer to 50 μL and detected using a 20E EIA kit as above method . The 20-hydroxyecdysone enzyme immunoassay ( 20E-EIA ) is based on competition between unlabeled 20E ( free 20E ) and acetylcholinesterase ( AChE ) -labelled 20E ( Tracer ) for limited-specific rabbit anti-20E antiserum . The rabbit anti-20E antiserum was bound to a mouse monoclonal anti-rabbit antibody coated-plate . The plate was then washed using the wash buffer provided with the kit ( Dilute 2 mL of concentrated Wash Buffer #A17000 with 800 mL of UltraPure water . Add 400 μL of Tween20 #A12000 ) to remove any unbound reagent . The AChE-labelled 20E and free 20E in samples were added into the wells and the plates were incubated at 4 °C overnight . After washing three times with wash buffer , Ellman’s reagent ( an enzymatic substrate for AChE and chromogen ) was added to the wells and incubated for 1 . 5 h at room temperature on a shaker . The AChE-labelled 20E acts on substrate in the Ellman’s Reagent to form a yellow compound that strongly absorbs light at 414 nm . The intensity of the color was determined at 414 nm by spectrophotometry ( Infinite M200PRO NanoQuant , Tecan , Grödig , Austria ) ; the optical density was proportional to the amount of tracer bound to the well and inversely proportional to the amount of 20E in the samples . The quantity of 20E bound to GPCRs was determined using a 20E standard curve generated using the same method . A detail protocol is provided with the kit . GPCRs and GFP were overexpressed in Sf9 cells and isolated as above description . Proteins were stored at -80°C for saturation experiments . 20E-EIA was used for binding saturation curve analysis . GPCR bound on Ni2+ resin was incubated with 1 to 60 nM 20E in 250 μL binding buffer ( 20 mM HEPES , 100 mM NaCl , 6 mM MgCl2 , 1 mM EDTA , and 1 mM EGTA ) at 26 °C for 40 min . The GPCR and 20E bound resin was washed three times with wash buffer , each for 5 min and finally eluted in 50 μL elute buffer ( 0 . 5 M NaCl , 20 mM Tris-HCl , pH 7 . 9 , 75 mM imidazole ) . The diafiltration of eluted protein was carried out with EIA buffer using Amicon Ultra-15 for three times . Protein concentration was detected by BCA method . 10 μg GPCR and its bound 20E was added EIA buffer to 50 μL and detected using a 20E EIA kit as above method . Nonspecific binding was determined in 20E-EIA assays with GFP . Dissociation constants ( Kd ) was determined by nonlinear regression from specific and nonspecific binding data of saturation experiments by using GraphPad Prism 7 on the assumption that 20E binds to a single site . ELISA ( Enzyme Linked Immunosorbent Assay , ELISA ) plates were incubated with 10 μg of DopEcR-GFP isolated from HaEpi cells that were overexpressed DopEcR-GFP in 200 μL coating buffer ( 0 . 015 M Na2CO3 , 0 . 035 M NaHCO3 , pH 9 . 6 ) at 4 °C overnight . The plates were washed 3 times with wash buffer ( 0 . 15 M PBS , pH 7 . 4 ) and then incubated with 1% BSA in 200 μL PBS at 37 °C for 1 h . The plates were washed 3 times with wash buffer and added various concentrations of DA in 200 μL PBS . Remaining experiments were performed according to the manufacturer’s instructions of insect hemolymphal dopamine ELISA kit ( MLBIO Biotechnology , Shanghai , China ) . The ELISA plates coated with DopEcR-GFP were incubated with 20 pmol of DA , and the increasing concentrations of the different ligands , 20E or DA , were incubated with the plates , respectively , to study the competition of 20E to DA . SPSS 23 . 0 ( SPSS Inc . , Chicago , IL , USA ) was used for data analysis . All data come from at least three biologically independent experiments . Two-group datasets were analyzed by Student’s t-test . One-way analysis of variance ( ANOVA ) were used to analyzed different among three or more group by Duncan’s multiple comparison test was used at p = 0 . 05 . One asterisk was used for p < 0 . 05 , two asterisks for p < 0 . 01 . The detail was showed in the related figure legends . | The steroid hormone 20-hydroxyecdysone ( 20E ) represses insect larval feeding and promotes metamorphosis; however , the mechanism is unclear . The dopamine receptor plays important roles in animal motor function and reward-motivated behavior . Using the serious lepidopteran agricultural pest Helicoverpa armigera as a model , we revealed that 20E binds to DopEcR to block the dopamine pathway and initiates the 20E pathway . Dopamine ( DA ) binds to the dopamine receptor ( DopEcR ) , a G protein-coupled receptor ( GPCR ) , to regulate cell proliferation , larval feeding , and growth . However , 20E competes with DA to bind to DopEcR , which represses larval feeding and triggers the 20E-pathway , leading to metamorphosis . The results suggested that 20E , via binding to DopEcR , stops larval feeding and promotes pupation , which presented an example of the steroid hormone regulating dopamine receptor and behavior . Our study showed that GPCRs can bind 20E and function as 20E cell membrane receptors . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"phosphorylation",
"invertebrates",
"neurochemistry",
"chemical",
"compounds",
"neuroscience",
"organic",
"compounds",
"animals",
"hormones",
"membrane",
"proteins",
"developmental",
"biology",
"amines",
"neurotransmitters",
"catecholamines",
"cellular",
"structures",
"and",
... | 2019 | The steroid hormone 20-hydroxyecdysone binds to dopamine receptor to repress lepidopteran insect feeding and promote pupation |
It has long been proposed that much of the information encoding how a protein folds is contained locally in the peptide chain . Here we present a large-scale simulation study designed to examine the extent to which conformations of peptide fragments in water predict native conformations in proteins . We perform replica exchange molecular dynamics ( REMD ) simulations of 872 8-mer , 12-mer , and 16-mer peptide fragments from 13 proteins using the AMBER 96 force field and the OBC implicit solvent model . To analyze the simulations , we compute various contact-based metrics , such as contact probability , and then apply Bayesian classifier methods to infer which metastable contacts are likely to be native vs . non-native . We find that a simple measure , the observed contact probability , is largely more predictive of a peptide's native structure in the protein than combinations of metrics or multi-body components . Our best classification model is a logistic regression model that can achieve up to 63% correct classifications for 8-mers , 71% for 12-mers , and 76% for 16-mers . We validate these results on fragments of a protein outside our training set . We conclude that local structure provides information to solve some but not all of the conformational search problem . These results help improve our understanding of folding mechanisms , and have implications for improving physics-based conformational sampling and structure prediction using all-atom molecular simulations .
It has long been proposed that much of the information encoding how a protein folds is contained locally in the peptide chain . Indeed , the success of fragment insertion methods for ab initio folding algorithms often relies on the predicted structures of small peptide pieces of the target protein [1] , [2] . To what extent do the conformations of peptide fragments in water predict native conformations in proteins ? We are interested in this question for at least two reasons . First , accurate local structure predictions from all-atom simulations of small peptide fragments of proteins in water may be useful for physics-based “divide and conquer” strategies for protein structure prediction , such as in the “zipping and assembly” method [3]–[5] . Physics-based methods for prediction offer several potential advantages over database-driven methods , such as the ability to simulate dynamics and predict folding pathways , using transferrable forcefield models which can be applied to a wide range of other problems . Second , this work informs an ongoing discussion about how much of the native structure of a protein is encoded within local sequence information alone [6] , [7] . In the “framework mechanism” [8] , for example , local information is sufficient to reduce the conformational searching enormously . On the other hand , protein folding is highly cooperative , so models such as the “nucleation-condensation model” indicate that secondary and tertiary structure may form concurrently [9] . Elucidating the role of local structure can help improve our understanding of protein folding mechanisms in general . The question we raise here is not about the success rates of secondary structure predictions . Secondary structure prediction methods such as PSIPRED use knowledge bases of known native structures and can achieve prediction success rates near 80% ( as judged by scores ) [10] . Here we ask a question of physics . If you knew the physical structure of a peptide in water , rather than in a database of native protein structures , would it predict the conformation of the same peptide in the protein's native structure ? As an approximation to the physics , we rely on all-atom force field simulations here . Much work has shown that simulations using current all-atom forcefields can sufficiently and accurately reflect the underlying physics [11]–[13] . There are previous studies using molecular dynamics simulations of peptide fragments for structure prediction . Bystroff and Garde performed 10-ns explicit-water simulations using the AMBER ff94 forcefield for 64 8-residue fragments to show that observed helicity correlates well with I-sites predictions [14] . Ho and Dill performed REMD simulations using the AMBER ff96 forcefield with the GB model of Tsui and Case for 133 8-residue fragments from six different proteins to identify regions of local native-like structure that could serve as folding nuclei [15] . Here , we perform much more extensive tests , over a larger data set and with multiple metrics , made possible by using a high-efficiency search method , called ZAM ( Zipping and Assembly Method ) , that samples the important parts of conformational space . We perform 872 independent simulations of 8-mer , 12-mer , and 16-mer fragments from 13 test proteins , for a total of 8 . 7 CPU-years of simulation time , which is , as far as we know , the largest set of fragment simulations performed to date . We use the AMBER ff96 forcefield [16] with the GB implicit solvent model mode of Onufriev , Bashford and Case [17] , which we have found to predict the structures of a set small peptides with better accuracy than other combinations of AMBER forcefields with GB solvation models [13] . This forcefield has been used with the ZAM conformational search algorithm to predict protein structures in the CASP7 competition [4] .
To what extent did our simulations of peptide fragments sample native-like structures ? From the native structures of our target sequences , we determined alpha helical and tight turn types across each target sequence using the secondary structure classification algorithm STRIDE [18] . The turn types were further divided into two groups , one for turns in beta-hairpins , and one for everything else . We filtered the dataset for fragments that were known to have at least 7 , 12 , and 16 native contacts respectively for 8- , 12- and 16-mers . This selects a subset of fragments with known native secondary structures to which we could compare our simulation data . We find that the fragment simulations sample diverse structures . Conformational clustering ( see Methods ) produces about 10 representative cluster conformations for each fragment simulation . Figure 1A shows , for each target sequence and fragment length , the C-alpha RMSD-to-native values for all representative cluster conformations along the target sequence . These fragments typically sample native-like conformations . Figure 1B plots the fraction of cluster conformations that sample within a given RMSD of the native conformation . It is not clear that native-like sampling would necessarily be expected; it depends on the relative importance of the tertiary context [7] , [19] . Nevertheless , we find that about 65% of 8-mer alpha helical conformations are within 2 . 0Å RMSD of the native state , and about 40% of 12-mers and 16-mers are within this range . For comparison , a random distribution of C-alpha RMSD calculated from native protein structures contains only about 10% of 8-mers , 5% of 12-mers , and 2% of 16-mer conformations with RMSD within 2 . 0Å RMSD ( see Methods ) . About 40% of 8-mer and 12-mer beta hairpin turns were within 2 . 0Å of the native structure , and 40% of 16-mer hairpins were within 3 . 0Å of the native structure ( only about 5% of random native 16-mer conformations are within 3 . 0Å RMSD ) . These results suggest that beta hairpins are more context-dependent , while helices are more generally defined locally . Also , we observe that beta hairpins show more structural variation in general than helices , due to the nonlocal contact topology . Does running longer simulations lead to more native-like structures ? We found this not to be the case . On seven different hairpin fragments , we performed 20 REMD simulations ( with and without various contact constraints ) for a total of 100 ns ( Text S1 ) . In these tests , we simulated both hairpins that corresponded to native structures , and “decoy” hairpins that were predicted by our simulations , but did not correspond to native structures . We conclude that longer simulation does not produce more native-like structures in our simulations . This could be for several reasons: ( 1 ) simulations longer than 100 ns would be needed , or ( 2 ) the physical model we used is not perfect [13] , [20] ) , or ( 3 ) because tertiary context is needed to drive them into their native states . While this work does not attempt to fully resolve these issues , it does establish a lower bound on the extent to which simulations of peptide fragments predict native-like structures , which we find here to be considerable . Our data provides an opportunity to draw inferences about what physical properties of intrachain contacts are predictive of whether a peptide conformation is native or not . To do this , we train probabilistic classifier models on several contact metrics , and interrogate the results . For each set of simulated fragments ( 8-mers , 12-mers , and 16-mers ) , we explored two kinds of per-contact classification models: a naive Bayes model and a logistic regression model ( see Methods ) . To find the most predictive classifier , each model was trained on all possible combinations of per-contact metrics ( defined in Methods ) calculated from the simulations . Which classification model best predicts native or non-native contacts from short fragment simulations ? In all cases , the logistic regression model gave better classifications than the corresponding naive Bayes model , thus we present only the results from the logistic regression models . Also in all cases , contacts defined by a 7Å distance cutoff performed significantly worse than an 8Å cutoff , thus we only present results from the latter case . The best logistic regression coefficients for 8-mers , 12-mers , and 16-mers are shown in Table 1 . What metrics are the best predictors of whether a simulated fragment has formed native contacts ? We examined several metrics ( see Methods ) , each calculated on a per-contact basis from the simulation data ( Figure 2 ) : ( 1 ) contact probability ( CPROB ) , the equilibrium probability of a given contact , ( 2 ) a distance profile score ( DPROF ) quantifying interresidue probabilities as a function of distance , ( 3 ) a mutual stability score ( MSTAB ) quantifying the joint probability of a contact when making pairs with other contacts , ( 4 ) a mutual cooperativity score ( MCOOP ) quantifying cooperative interactions made with other contacts , and ( 5 ) a mesoentropy score ( MESO ) , which is a measure of the backbone conformational entropy . Since the numerical values of the five contact metrics can differ by orders of magnitude , we obtain a better sense of the relative importance of the different contact metrics by computing the model relevance , which we define as , where is the logistic regression coefficient for contact metric , and is the standard deviation of the metric . The values calculated for each regression coefficient show that the most predictive metric is the contact probability ( Figure 3 ) . This is interesting because it might be expected that including multi-body terms would be more predictive than just the pairwise contact formation probability , since protein stability is likely to involve non-additivities that could only be captured in complex terms . Instead , we find that simple pairwise terms are the most predictive , with the multi-body terms producing small negative regression coefficients . The negative coefficients can be interpreted as providing a slight correction to the over-counting due to correlation between pairwise contact probability terms . Figure 4 shows the results of increasing the number of prediction coefficients . These curves make essentially three points . First , the best first approximation , i . e . , the most predictive single term , as noted above , is CPROB , the contact probability . Second , the figure shows that the predictive power of the model increases by adding up to two additional terms . However , the added value in predictive power is quite small . And , third , it shows that adding further terms to the model , beyond three , worsens the predictive power . We also tested whether we could obtain better classification models by training on local contacts ( or nonlocal contacts ) alone . We found that , overall , the classification success for the local-only or nonlocal-only data was comparable , but never as high as the classification success using the combined data ( see Text S1 ) . Now , given the parameters obtained from the logistic-regression models described above , we can compute the probability that a given simulated peptide conformation has native contacts . Figure 5 shows the contact prediction success for all protein targets in the test set . The average percentage of correctly classified contacts ( across each protein target ) using the 8-mer data is 63 . 2% ( 72 . 3% for native contacts and 60 . 7% for non-native contacts ) . The average percentage of correctly classified 12-mer contacts increases to 71 . 3% ( 57 . 3% for native contacts and 74 . 3% for non-native contacts ) , and for 16-mer contacts the average classification success is 76 . 9% ( 56 . 3% for native contacts and 80 . 9% for non-native contacts ) . In the case where the data contains many more non-native contacts than native contacts , a high classification accuracy may not reflect a significant improvement over a random null distribution , per se . To test this possibility for our selected models , we built a null distribution of contact metrics to test the random-case performance of our models ( see Methods ) . Several statistical tests , including Matthews correlation coefficient ( MCC ) values and receiver-operator characteristic curves [21] show that our best classification models perform better than random ( for a full discussion , see Text S1 ) . Figure 6 compares the predictions to the true native structures . It shows the ‘logit’ values ( see Methods ) given by the best 16-mer logistic regression model for an example target . This quantity has the flavor of an informational equivalent of a free energy difference of native minus denatured . The darker black on the figure indicates the strongest prediction of native-like structure . The 8-mer , 12-mer and 16-mer results for all targets is shown in Text S1 . Not surprisingly , to the extent that these peptide fragment simulations predict native-like structures , helices are better predicted than hairpins . Next , we tested our model on a protein outside our test set . We tested 1whz ( PDB ID: 1whz ) , a 70-residue CASP6 target with an structure taken from Thermus thermophilus ( Figure 7 ) . REMD simulations of 8-mer , 12-mer , and 16-mer fragments were performed ( 62 , 39 , and 74 independent fragment simulations , respectively ) using the ZAM procedure , and contact predictions were made using our previously-paramterized 8-mer , 12-mer , and 16-mer logistic regression classification models . Figure 8 shows contact prediction success rates for 1whz , and the logit values for each contact estimated from the 8-mer , 12-mer , and 16-mer data . As the fragment length grows , a consensus resemblance to the native contact map begins to emerge , although incorrect in some places . The logit values are very similar to the logit values given by 8-mer , 12-mer , and 16-mer regression models trained only on contact probability , showing that the contact probability observed in our simulations contains most of the predictive information . These models make per-contact predictions . But , we are interested in predictions for whole peptide conformations . To turn our contact-based scores into conformation-based scores , we compute a score , , for a given molecular conformation as follows:Here , runs over all contacts in the conformation , and runs over all fragment simulations which contain contact . We computed conformation scores for all the cluster conformations extracted from 8-mer , 12-mer , and 16-mer 1whz fragment simulations . For 8-mers and 12-mers , we observe a correlation ( albeit noisy ) between a high value of and a near-native ( low-RMSD ) structure ( see Text S1 ) . For 16-mers , the conformation score predicts four likely secondary structures consisting of helices and hairpins along the sequence of the protein ( Figure 9 ) . Two of these secondary structures correspond to correctly predicted native structures ( the N-terminal helix and C-terminal hairpin ) , while two of the secondary structures are non-native “decoys . ” Even for the decoys , near-native conformations are sampled substantially . Interestingly , the helical decoy seen in the sequence of residues from 12–39 is also predicted by the I-Sites/HMMSTR/Rosetta structure prediction server [22] , [23] when templates from multiple sequence alignments are turned off ( see Text S1 ) , indicating structural ambivalence .
We have performed computer simulations of short peptides—8-mers , 12-mers and 16-mers—using the AMBER 96 force field and the OBC implicit solvation model . Our aim was to see whether the metastable structures of these fragments bear any resemblance to the conformations those fragments adopt in the native states of the proteins in which they appear . We find that the peptide contact probabilities in a logistic regression model lead to a 76% success rate in 16-mers in correctly classifying contacts as either native or nonnative . Across the chain lengths studied , the false negative rates ( native contacts classified as non-native ) of our best logistic regression models range from about 30–45% . The false positive rates ( non-native contacts classified as native ) vary from about 20–40% . These results show these predicted peptide conformations in water are significantly more native-like than would be expected from random conformers . Previously , Bystrof and Garde also showed a 75% success rate at predicting native helicity across 64 8-mer fragments simulated using AMBER ff94 and explicit TIP3P water [14] . This compares with our 72% success rate at classifying native contact for 8-mers . While there remain issues of the accuracy of the forcefield+solvation model [20] , and our limited simulation times ( 5–15 ns ) , nevertheless , these results indicate that , by using REMD for all-atom sampling and ZAM for conformational searching , small peptide fragments in proteins adopt conformations in solution that significantly resemble the conformations they ultimately adopt in their native proteins . Past experiments have reached similar conclusions , for specific peptide fragments [24] . These results may have useful application in physics-based methods , like ZAM [3] , [4] that aim to predict protein structures from all-atom simulations in the absence of knowledge-based secondary structure prediction methods . This work also has implications for understanding how proteins can physically fold up so rapidly to reach their native structures . It suggests that proteins can fold into globally optimal conformations by starting with locally optimal conformations first . While this idea has long been a mainstay of models of protein folding kinetics , this is , as far as we know , the first extensive demonstration in a purely physical model . However , these local propensities alone are not sufficient , at least in our simulations , to predict the native states of proteins . While our fragment simulations show that some peptide fragments sample native-like states , the sampling still produces many false positives and false negatives . This is consistent with the information-theoretic studies of Crooks and Brenner [25] which examined neural net models trained on local sequence alone , and found that “one fourth of the total information needed to determine secondary structure is available from local inter-sequence correlations . ” Similarly , our results also support the idea that cooperative , long-range tertiary contacts are crucial in determining native structure . But while local structuring alone may be insufficient to fold proteins , such information can help to narrow the conformational search . Fleming et al . has shown that while restricting a protein chain to preferred secondary structures per se generates random coil-like behavior , some simple additional logic about tertiary cooperativity and hydrogen bonding can predict native-like protein topologies and structures [26] . Moreover , bioinformatics-based protein structure prediction methods have benefitted greatly from fragment assembly methods whereby locally compatible structures dramatically reduce the conformational search problem [22] , [27]–[29] . Our results suggest local structural information from physical simulations can improve our understanding of protein folding pathways , and may be useful in physics-based structure prediction .
Our dataset of peptides was 8-mer , 12-mer , and 16-mer fragments of 8 CASP7 target sequences and 5 other protein sequences with known structures taken from the PDB ( see Table 2 ) . The 8-mer , 12-mer , and 16-mer fragments cover 100% , 88 . 7% , and 76 . 7% of the entire sequence space of the 13 proteins considered , respectively ( see also Text S1 ) . We performed computer simulations for 10 ns for each peptide , totaling about 8 . 7 CPU years in simulation time . Classification models were trained on five different contact-based metrics , calculated on a per-contact basis from the simulation data: 1 ) contact probability ( CPROB ) , 2 ) a distance profile score ( DPROF ) , 3 ) a mutual stability score ( MSTAB ) , 4 ) a mutual cooperativity score ( MCOOP ) and 5 ) mesoentropy score ( MESO ) ( Figure 2 ) . These metrics are described in detail below . Given the various metrics above , of the peptide conformations observed from the simulations in solution , we now ask if there is a way to combine those metrics to make the best possible predictions of what the peptide's structure is in the native state of the protein . For each contact observed in our database of simulated fragments , we have a set of measured contact metrics , and the known native structure of the fragment , which tells us if the contact is native or non-native . Using this data , we want to train a probabilistic model to estimate the probability of a contact being native versus non-native , given only the contact metrics observed in a peptide simulation . This is a binary pattern classification problem , where we have an unknown parameter which can be either be native or non-native , and we wish to calculate . Bayes' formula can be used to restate this posterior probability as ( 1 ) Here , represents our prior knowledge of the probability of observing a native or non-native contact , given no other information about that contact . represents the conditional probability of observing a set of metrics for a contact , given that we know whether that contact is native or non-native . The ‘naive Bayesian’ approach would be to assume that , for any contact , our set of calculated metrics are all mutually independent and uncorrelated . In this case , ( 2 ) Using Equations 1 and 2 , and taking the logarithm of the ratio of , we get ( 3 ) Since , it follows that the log-ratio can be expressed as a linear sum of ‘logit’ terms of the form . The first term on the right side of Equation 3 is a ‘logit’ for our prior , and the remaining terms are conditional ‘logits’ for our metrics of interests . Both kinds of information are empirically compiled from our database of fragment simulations , from which we extract histogram counts of each metric for native and non-native contacts . Substituting , we solve Equation 3 to obtain ( 4 ) ( 5 ) A potential improvement to the ‘naive Bayes’ model is the logistic regression method [33] , which seeks to find the best linear coefficients for the following model: ( 6 ) Solving for yields ( 7 ) In practice , these coefficients ( and their error estimates ) are found with a maximum-likelihood optimization using Newton-Raphson gradient minimization . The optimization is equivalent to least-squared linear regression in the nonlinear ‘logit’ variables . This nonlinearity sometimes makes it possible to obtain better classifications than the naive Bayesian approach . Note the similarity of the logistic regression model ( Equation 6 ) to the naive Bayesian approach ( Equation 4 ) , with acting as a ‘prior , ’ and with the magnitudes of the values of indicating the significance of each contact metric . We built both naive Bayes and logistic regression models for 8-mer , 12-mer , and 16-mer fragments separately . For the naive Bayes models , this involved empirically computing histograms in . For the logistic regression models , estimates of the best coefficients were computed directly from the data , using a freely available Python package [34] . For each of kind of model , in order to determine the best combinations of metrics on which to train the model , we built separate models for all ( 25−1 ) = 31 combinations of the five contact metrics ( CPROB , DPROF , MSTAB , MCOOP and MESO ) . In addition , for each of the models , we tested three different inter-residue distance definitions ( , and residue side chain centroid ) , and two different distance cutoffs to define a contact ( 7 . 0Å and 8 . 0Å ) , giving a total of 186 combinations to test . To avoid over-fitting , the training data used to construct each model was divided randomly into five groups so that independent models could be built for each group . Additionally , 1/5 of the data in each group was set aside for testing the model , and the other 4/5 of the data was used to train the model . This means that for each model , there were 25 independent testing and training rounds: 5 independent model-building rounds , each with 5 leave-one-out trials of testing and training . To assess which model was the best , we used a statistical hypothesis testing scheme to find a model that most successfully classifies native contacts as well as non-native contacts . Consider a test where we use the statistic to decide between two hypotheses . The hypothesis is that the contact is non-native , while the hypothesis is that the contact is native . If is less than some threshold value , then we accept and reject , and if , we accept and reject . To find the best value for , we choose the value that maximizes , where is the fraction of non-native contacts incorrectly classified as native , and is the fraction of native contacts incorrectly classified as non-native . Even though there are many more non-native contacts than native contacts , this procedure equally weights native and non-native contacts , achieving a balance of specificity and statistical power . We define the model quality ( Q ) as the maximal value of , and use the value to rate the relative predictive power of different models . Errors in were estimated by examining the sample variance across the five independent trials of the complete model-building procedure . For the naive Bayes models built for each fragment length , the model that yielded the highest model quality ( Q ) when applied to testing data was chosen as the best model . For the logistic regression models , the 25 rounds of testing and training produced a series of models across which values may correlated . Thus , instead of choosing the average for the best logistic regression model , we chose the model whose coefficients were closest to the centroid of values across the 25 testing and training rounds . For each simulation , the probability of a contact being native can be estimated by Equation 7 . However , in the case where there are multiple simulations of the same contact ( in overlapping fragment simulations ) , we can use all of the simulation data to estimate this probability . Assuming that each of simulations is statistically independent , the probability of a particular contact being native is estimated by:We use these combined estimates of with the original hypothesis testing cutoffs to classify contacts as native or non-native . The percentage of contacts correctly classified this way is what we report as our contact prediction success rate . A null distribution in C-alpha RMSD values for 8-mers , 12-mers and 16-mers was calculated by taking 10000 random pairwise samples of 8-mer , 12-mer and 16-mer fragments from a set of 3465 protein structures taken from the SCOP database [35] ( 1 structure , or 2 if existing , from each unique SCOP class ) . Because there are correlations between contact metrics due to chain connectivity , considerable care was taken to construct null distributions for contact metrics that preserved these correlations . We did this by constructing the null distribution on a fragment-by-fragment basis . For each fragment , the values of the contact metrics were retained , while the assignment of native and non-native contacts was randomized according to a per-fragment bootstrapping procedure . For each fragment , a random contact map was drawn ( with replacement ) from the full data set . This reassignment procedure , across the entire set of fragments , was repeated 1000 times to construct a distribution of random-case realizations . | Proteins must fold to unique native structures in order to perform their functions . To do this , proteins must solve a complicated conformational search problem , the details of which remain difficult to study experimentally . Predicting folding pathways and the mechanisms by which proteins fold is thus central to understanding how proteins work . One longstanding question is the extent to which proteins solve the search problem locally , by folding into sub-structures that are dictated primarily by local sequence . Here , we address this question by conducting a large-scale molecular dynamics simulation study of protein fragments in water . The simulation data was then used to optimize a statistical model that predicted native and non-native contacts . The performance of the resulting model suggests that local structuring provides some but not all of the information to solve the folding problem , and that molecular dynamics simulation of fragments can be useful for protein structure prediction and design . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"computational",
"biology/protein",
"structure",
"prediction",
"computational",
"biology/molecular",
"dynamics",
"biophysics/protein",
"folding"
] | 2009 | Predicting Peptide Structures in Native Proteins from Physical Simulations of Fragments |
In order to increase the efficient allocation of soil-transmitted helminth ( STH ) disease control resources in the Philippines , we aimed to describe for the first time the spatial variation in the prevalence of A . lumbricoides , T . trichiura and hookworm across the country , quantify the association between the physical environment and spatial variation of STH infection and develop predictive risk maps for each infection . Data on STH infection from 35 , 573 individuals across the country were geolocated at the barangay level and included in the analysis . The analysis was stratified geographically in two major regions: 1 ) Luzon and the Visayas and 2 ) Mindanao . Bayesian geostatistical models of STH prevalence were developed , including age and sex of individuals and environmental variables ( rainfall , land surface temperature and distance to inland water bodies ) as predictors , and diagnostic uncertainty was incorporated . The role of environmental variables was different between regions of the Philippines . This analysis revealed that while A . lumbricoides and T . trichiura infections were widespread and highly endemic , hookworm infections were more circumscribed to smaller foci in the Visayas and Mindanao . This analysis revealed significant spatial variation in STH infection prevalence within provinces of the Philippines . This suggests that a spatially targeted approach to STH interventions , including mass drug administration , is warranted . When financially possible , additional STH surveys should be prioritized to high-risk areas identified by our study in Luzon .
Soil-transmitted helminth ( STH ) infections are believed to affect two billion people worldwide , equating to approximately one-third of the world’s population . Ascaris lumbricoides , Trichuris trichiura and the hookworms Necator americanus and Ancylostoma duodenale , are the species responsible for most of these infections [1] . The infective stages of these parasites are found in fecally contaminated environments , which means that a lack of clean water , sanitation and hygiene ( WASH ) contribute considerably to their transmission [2] . While most STH infections are asymptomatic , heavier infections can result in abdominal pain , diarrhea , malaise , and weakness . STH infections in children have also been linked to anaemia and impaired cognitive and physical development [3–5] . These STH species are widely distributed in tropical and subtropical areas [6] . In the Philippines , prevalence data from the first national baseline survey of STH infections conducted from 2005 to 2008 showed endemic levels in the three regions of the country ( Luzon , Visayas , and Mindanao ) [7 , 8] To address the problem of STH infections , the Philippines Department of Health has been implementing the Integrated Helminth Control Program since 2006 [9] . The program has chemotherapy with albendazole or mebendazole as its cornerstone and targets all children aged 12 months to 12 years , and other special population groups including pregnant women , adolescent females , farmers , and indigenous populations across the Philippines [9 , 10] . In addition , people found to be infected with any of the STH species are treated . Anthelminthic drugs are provided free of charge . The control program also includes installation of water and sanitation facilities and educational approaches to improve hygienic practices [11] . To guide STH control programs and ensure that scarce resources are allocated efficiently , it is important to know which areas have the highest prevalence of infection . Traditional ways of implementing and evaluating parasite control programs often entail significant costs that restrict their use in developing countries like the Philippines . Previous mapping efforts of STH prevalence in Southeast Asia enabled the construction of models to predict areas of high prevalence , the estimation of the number of people needing treatment and the identification of intervention targets for national helminth control programs [12] . In recent years , spatial prediction methods that employ model-based geostatistics ( MBG ) have been developed to guide parasite control programs [13] . These methods can be used to generate species-specific predictive prevalence maps , and to estimate relationships and associated uncertainty between infection outcomes and covariates , including environmental factors such as rainfall and temperature . Imperitive to these highly aggregated parasitic diseases , MBG also has the capacity to account for the inherent clustering of infection . Examples of the application of the MBG approach to understand STH epidemiology are available for sub-Saharan Africa [13–15] and more recently in South America [16] but none have been developed in the Southeast Asian context . This study aimed to quantify the association between the physical environment and the prevalence of A . lumbricoides , T . trichiura and hookworm infections in the Philippines and to generate statistically robust spatial predictions of STH infections for the Philippines .
Ethical clearance for this analytical study was provided by the University of Queensland Human Research Ethics Committee ( Project Number 2011000692 ) . This analytical study utilised STH data collected in the National Parasitological Survey and data collected in a study in Western Samar; the research ethics procedures for both these studies are detailed elsewhere [7 , 8 , 17] . All data used in the study was anonymized . We used STH data collected during the most recent ( conducted from 2005–2007 ) national schistosomiasis survey in the Philippines; study design and data collection procedures were described in detail elsewhere [7 , 8] . In brief , the national survey used stratified two-stage systematic cluster sampling whereby stratification was done by region and then by prevalence level . The endemic provinces were initially divided into high prevalence , moderate prevalence and low prevalence groups based on the surveys conducted during the implementation of the Philippines Health Development Program in the 1990’s . The province was the primary sampling unit and the barangays the secondary sampling unit . While all known endemic provinces were included in the sampling population , among the non-endemic provinces , random selection was performed . The barangays were selected proportional to population size . Households within barangays were selected in a systematic manner from a master list of all the households in the barangay . Two stool samples were collected on separate days from each study participant . However , the submission of the second stool was erratic and only the result from the first examination was taken . The Kato-Katz thick smear was used to detect eggs of A . lumbricoides , T . trichiura , and hookworms . Demographic data including age and sex were also collected for each participant . The data from the Visayas was supplemented with data from the cross sectional component of a cohort study that included 5 , 624 residents of 50 barangays from Western Samar which had not been included in the national survey [17] . The purpose of the Samar study was to determine the effect of water management systems and non-human animal hosts on S . japonicum transmission dynamics and their role in human infection parameters [18 , 19] . The villages were sampled to represent 25 mainly rain-fed and 25 mainly irrigated villages . In each selected village , 35 households with at least one rice farmer were selected at random , and at most 6 persons per household , including at least one rice farmer , were selected at random . This dataset included test results for three consecutive days but to be consistent with available data from the national survey only data from the first day was used in the analysis . The unit of analysis was the barangay , the smallest administrative unit in the Philippines . The mean length of the longest axis of barangays was 11km ( SD:10 . 3 ) . Barangay centroids were estimated using the geographical information system ( GIS ) software QuantumGIS ( QGIS ) version 1 . 7 . 3 ( QGIS Development Team , 2011 ) . This procedure was based on combined information from shapefiles of the barangays of the Philippines , which were obtained from the geographic data warehouses DIVA GIS ( www . diva-gis . org/Data ) and PhilGIS ( www . philgis . org ) for the Philippines . A total of 214 barangays in Luzon , the Visayas and Mindanao were included in the analysis ( Fig 1 ) . Remotely sensed environmental data for land surface temperature , rainfall and NDVI and distance to perennial water bodies ( DPWB ) were obtained from WorldClim ( www . worldclim . org ) . Land surface temperature ( LST ) was considered in the analysis because A . lumbricoides , T . trichiura and hookworm have thermal thresholds outside of which the survival of the infective stages in the soil declines [1] . Rainfall and DPWB were also considered because these will affect the moisture of the soil where the helminth infective stages are found , and therefore their survival . Normalized difference vegetation index ( NDVI ) which serves as a proxy measure of rainfall for a 1 km × 1 km grid cell resolution were obtained from the National Oceanographic and Atmospheric Administration’s ( NOAA ) Advanced Very High Radiometer and was also included in the analysis . Using QGIS , the values of these environmental co-variates were extracted for each barangay . For the purpose of the analysis the presence of parasite eggs in stool identified by the Kato-Katz method was used as the outcome variable and thus all individuals were categorized into infected and not infected based on the presence of at least one egg . Initial variable selection included age and sex because these two factors have been shown to be associated with STH infections probably by influencing exposure and susceptibility to infection [1] . The environmental variables ( rainfall , DPWB , LST and NDVI ) were also considered in the initial variable screening stage . Correlations between environmental covariates were investigated using Pearson’s correlation coefficients . Scatter plots were used to assess the relationship between the barangay-level STH prevalence and the value of each of the environmental variables . Multivariable logistic regression models for a Bernoulli-distributed outcome , with cluster correction by barangay using robust standard errors , were built for each STH species for each region of the Philippines , i . e . , Luzon , the Visayas , and Mindanao using the statistical software Stata version 10 . 1 ( Stata corporation , College Station , TX ) . Residuals of the final non-spatial models were examined for spatial autocorrelation by generating a semivariogram using the geoR package of R software v . 2 . 15 . One semivariogram was generated for each STH species for each region of the Philippines to determine how much of the clustering of STH infections is explained by location , and to establish the propensity and size of geographical clusters . Bayesian logistic geostatistical models were built for each STH species for the regions of Luzon and the Visayas combined and for the region of Mindanao separately using WinBUGS ( MRC Biostatistics Unit , Cambridge , and Imperial College London , UK ) . This decision was based on the results of the semivariogram analysis ( Fig 2 ) , which indicated similar spatial dependence in STH prevalence in the regions of Luzon and the Visayas . The models included an intercept , the individual level variables age ( categorized into children aged <5 years and 5–19 years , and adults aged >20 years ) and sex , the environmental variables DPWB , LST , NDVI and rainfall , and a geostatistical random effect ( S1 Text ) . In addition , the model included adjustment for diagnostic uncertainty by modeling sensitivity and specificity as random variables . The Kato-Katz technique is widely used for detecting helminth eggs in stools . The sensitivity of the test is influenced by changes in the number of eggs excreted in the feces from day to day . True prevalence was modeled as a function of the observed prevalence and test sensitivity and specificity , with the generalised linear model fit to the true prevalence parameter . Priors for the sensitivity and specificity were specified as beta distributions; we used the alpha and beta parameters reported in previous studies [20] ( Table A in S1 Text ) . The covariate effects were summarized using the mean and 95% credible intervals ( representing the range of values that contains the true value with a probability of 95% ) ; a significant result for a coefficient is indicated by where the 95% credible interval does not cross zero . The geostatistical random effect modeled spatial correlation as a function of the separating distance between pairs of barangays . Model predictions for A . lumbricoides and T . trichiura were used to generate representative STH risk maps for males aged 5–19 years ( the subgroup with the highest prevalence for these STH ) and model predictions for hookworm were used to generate representative STH risk maps for males aged >20 years ( the subgroup with the highest risk for this STH ) across the Philippines in ArcGIS version 10 . 0 . Note the overall mean predicted prevalence is specific to the age and sex group ( i . e . choice of a different age-sex group would result in a different spatial mean ) , with spatial variation around the mean being influenced by the environmental variables and the spatial correlation component of the model . This means that the relative differences between locations are consistent and so the maps presented are representative of the spatial distribution of risk for all age groups and both sexes . The priors used for the model parameters ( spatial and non-spatial ) are given in the S1 Text . To determine the discriminatory performance of the model predictions relative to observed prevalence thresholds ( 20% and 50%; corresponding to prevalence thresholds for WHO-recommended STH control strategies [21] ) , the area under the curve ( AUC ) of the receiver operating characteristic was used ( more detail in Table B in S1 Text ) . An AUC value of >0 . 7 was taken to indicate acceptable predictive performance [22] . We estimated the mean prediction error and the percentage of the overall observed mean attributable to the error estimate .
For the purpose of the spatial modeling , STH data collected during the most recent ( conducted from 2005–2007 ) national schistosomiasis survey in the Philippines were supplemented with data from from Western Samar ( in the Visayas ) which had not been included in the national survey . In total , we included data from 35 , 573 participants from 214 barangays in Luzon , Visayas , and Mindanao ( Fig 1 ) . This corresponded to 2 , 701 individuals in Luzon , 13 , 203 individuals in the Visayas and 19 , 669 individuals in Mindanao with complete information regarding STH infection status , barangay geolocation and demography ( i . e . age and sex ) ( Table 1 ) . The mean observed prevalence of A . lumbricoides was 23 . 7% for Luzon , 38 . 4% for the Visayas and 21 . 2% for Mindanao . For T . trichiura , the mean observed prevalence was 27 . 9% for Luzon , 53 . 6% for the Visayas and 16 . 8% for Mindanao . The mean observed prevalence of hookworm was 4 . 5% for Luzon , 18% for the Visayas and 11 . 3% for Mindanao . A . lumbricoides and T . trichiura prevalence proportions showed a tendency for clustering in all three regions , unlike hookworm prevalence proportions which showed spatial clustering only in Mindanao ( Fig 2 ) . The greatest tendency for clustering was exhibited by T . trichiura in Mindanao ( Fig 2H ) and the largest estimated cluster size was for T . trichiura in Luzon ( Fig 2B ) . Model results ( Table 2 for Luzon and the Visayas and Table 3 for Mindanao ) indicated that individuals aged 5–19 years had higher prevalence of infection than individuals aged <5 years for all three STH . In contrast , the prevalence of infection was higher among individuals aged >20 years as compared to those aged <5 years for T . trichuria and hookworm , but not for A . lumbricoides . In Luzon and the Visayas , males had higher prevalence of A . lumbricoides infection compared with females . While in Luzon and the Visayas females had higher hookworm infection prevalence compared to males , in Mindanao males had higher hookworm infection prevalence compared to females . Environmental factors influenced the prevalence of the three parasites differently in the three regions . In Luzon and the Visayas , only the distance to water bodies was associated with an increase in the prevalence of hookworm infections . No other environmental variable was associated with the prevalence of the three parasites . In contrast , in Mindanao , land surface temperature was associated with an increase of the prevalence odds of A . lumbricoides and T . trichuria , and was associated with a decrease in prevalence of hookworm infections . Similarly , NDVI was associated with an increase in the prevalence of hookworm infections and was associated with a decrease of A . lumbricoides and T . trichuria . Rainfall was associated with an increased prevalence of T . trichuria and distance to water bodies was associated with an increase of the prevalence of hookworm , the latter being similar to that observed in Luzon and the Visayas . The parameter Phi ( φ ) refers to the rate of decay of spatial autocorrelation and indicates the size of clusters . The radius of a cluster in kilometers corresponds to ( 3/φ ) *111 ( note: one decimal degree is equivalent to approximately 111 km at the equator ) . For A . lumbricoides infection , the radii of the clusters measured 28 km in all regions . For T . trichiura infection , the clusters’ radii measured 215 km in Luzon and Visayas and 85 km in Mindanao . For hookworm infection , the radius of the clusters was 57 km in Luzon and Visayas and 225 km in Mindanao . The tendency for spatial clustering was the strongest for hookworm in all regions ( the higher value the spatial variance parameter the higher the tendency for spatial clustering ) ( Table 2 and Table 3 ) . The predicted geographical distribution of A . lumbricoides infection indicated that it was widespread and endemic in Luzon and Visayas , with areas of high prevalence ( >50% ) predicted for many locations in these two regions ( Fig 3A ) . T . trichiura was particularly widespread and highly endemic in the Visayas ( Fig 3B ) . Hookworm was much more focal compared with A . lumbricoides and T . trichiura in that areas of predicted high prevalence ( >50% ) of hookworm were circumscribed to Zambales in the central region , Isabela in the Cagayan Valley , Apayao in the Cordillera region , north of the Bicol region , central areas of the Mindoro island , parts of Palawan island , western Samar in the island of Samar and in Antique in the island of Panay ( Fig 3C ) . In Luzon and the Visayas , many areas of high prevalence of A . lumbricoides and T . trichiura overlapped . In Mindanao , A . lumbricoides overlapped with T . trichiura in the Zamboanga Peninsula to the west of the island . In Mindanao , most areas were predicted to reach prevalences between 10–40% for A . lumbricoides , while T . trichiura was more circumscribed with areas predicted to be highly endemic in small foci in Surigao del Norte and the Compostela Valley in the Davao region . Hookworm was predicted to have a prevalence of between 20–40% in the central provinces of Mindanao , including Cotabato , Bukidnon , Agusan del Sur and Davao .
This study represents the first to use model-based geostatistical predictive methods for STHs in a Southeast Asian context and demonstrates that STH infection prevalence in the Philippines is spatially variable . This study also shows that the predicted geographical distribution of A . lumbricoides and T . trichiura infection prevalence is more widespread than hookworm infection and in many areas the two species overlap . Finally , it indicates that the measured environmental factors are more strongly associated with the prevalence of these parasites in Mindanao than in Luzon and the Visayas . The finding that A . lumbricoides and T . trichiura infections are more prevalent in individuals aged 5–19 years compared with other age groups is consistent with the known epidemiology of these infections . Previous studies show that the relationship between age and prevalence of Ascaris and Trichuris infections are characteristically convex in shape with the highest intensities observed in school-aged children [5] . This is likely related to behaviour ( playing in the soil and possibly unsatisfactory hygiene habits ) that leads to exposure to the parasites’ infective stages . Our study also shows that in the case of hookworm , individuals aged ≥20 years had significantly higher prevalence of infection compared with other age groups , likely reflecting cumulative occupational exposures in older individuals such as farming . These are consistent with previous studies in Southeast Asia and elsewhere that describe how hookworm prevalence rises with increasing age to a plateau in adulthood probably due to the cumulative exposure to fecal-contaminated environments brought about by agricultural activities [5 , 12] . The results of our study show that the effect of sex on STH infection is less obvious and differs between regions of the Philippines . In agreement with previous studies , we found that males are at significantly higher risk of infection compared with females for A . lumbricoides in Luzon and the Visayas and for hookworm in Mindanao [12 , 23 , 24] . The effect size of males is strongest for hookworm in Mindanao , presumably because a substantial portion of the land area there is utilized for agriculture and in this region males are traditionally more involved in agricultural occupations than females . Interestingly , the relationship between sex and hookworm infection are opposite in Luzon and Visayas , with females being positively associated with hookworm infection , a feature warranting further investigation . This study also shows that the effect sizes of environmental properties , such as distance to perennial water bodies , land surface temperature , NDVI ( an index of vegetation ) and rainfall in STH infection are generally greater in Mindanao compared with Luzon and the Visayas . In Luzon and the Visayas only distance to water bodies was associated with hookworm infection suggesting that , unmeasured , small scale factors such as socioeconomic status and behavior may be important predictors of the spatial variation in STH infection in this region . In both regions , the findings support the role of the distance to water bodies in hookworm infection suggesting that the presence of infection increases further from the water bodies . In addition , the effect size for the proximity to water bodies was larger in Luzon and the Visayas as compared to that seen in Mindanao suggesting that , in addition to occupational exposure , the lower socio-economic status and reluctance to seek treatment in communities of Luzon and the Visayas may also be an important predictor of hookworm infection in this region . While increasing land surface temperature in Mindanao was shown to be associated with the increasing prevalence of A . lumbricoides and T . trichiura , the relationship was reversed for hookworm . This difference may be partly explained by the known characteristics of the hookworm lifecycle involving the survival and development of mobile larval stages in the soil , which seek optimal conditions of temperature and humidity . In addition , the results for Mindanao demonstrate that increasing vegetation indexes were associated with increased prevalence of hookworm , suggesting a potential role of land use related to agricultural activities on the transmission of hookworm . Notably , this effect was reversed for A . lumbricoides and T . trichiura . The significant unexplained variation in STH prevalence related to survey location ( as assessed by the residual semivariogram ) , particularly that estimated for Mindanao , suggested that considerable STH clustering was left unaccounted for by variables included in the non-spatial multivariable model . This finding justified the need for formally modelling spatial clustering across different regions in the Philippines using model-based geostatistics . The fact that The Philippines is an archipelago presents a technical challenge to the application of MBG methods , but the decision to model risk as a spatially continuous process across the three island groups was justified on the following grounds . First , it is impractical to model risk , or the spatial correlation structure , separately by island because of the thousands of islands in the Philippines archipelago–pragmatically , we chose to model risk separately by the three main island groups . Secondly , inter-island travel is extremely frequent in the Philippines and we consider islands in close proximity to be intimately connected . A major advantage of our approach is that it adjusted for the low sensitivity of the Kato–Katz thick smear examination in the modelling framework and produces a more accurate assessment of prevalence . The low sensitivity of Kato-Katz has been particularly prominent in low infection intensities and due to day-to-day variation in egg output of the adult worms [25] . While the Kato–Katz remains the cheapest and often the only available method in the field , its lack of sensitivity means that survey results are likely to underestimate the ‘true’ prevalence , and adjustments for measurement error should be taken into account [20] . The STH predictive prevalence maps demonstrate that almost the entire area of the Philippines is endemic for at least one STH , warranting nationwide control . The predicted prevalence of A . lumbricoides and T . trichiura show that these infections are highly endemic and have a widespread distribution . The results are consistent with the known ubiquity of A . lumbricoides and T . trichiura infections in the Philippines and the similar mode of transmission of the two species , i . e . , the ingestion of embryonated eggs . The findings also show the value of updating the current database of STH for the Philippines with new data; additional data for Western Samar has allowed the identification of areas which could not have been detected had the analysis been carried out on national survey data alone . While areas with the highest predicted prevalence of A . lumbricoides are located in Luzon , areas with the highest predicted prevalence of T . trichiura are located in the Visayas , in line with previous smaller scale studies [26–28] . Our results also show extensive areas of high endemicity in the Philippines where A . lumbricoides and T . trichiura infections overlap . The maps predicting the prevalence of hookworm differed markedly from those of the other STH , showing a focal distribution of the parasite ( corroborated by the highest tendency for clustering ) , with most of the areas with a high predicted prevalence in the inland areas of Mindanao , such as the Compostela Valley and more circumscribed areas in Northern Luzon and Western Samar . Our models also predicted the presence of STH infection in areas not represented in the survey data but which historically are known to be endemic [29 , 30] . This finding indicates that these can be environmentally suitable for the presence of infection . For example , our map predicts an extensive area of moderate ( 20–30% ) to high ( >50% ) risk of infection for most of Palawan , indicating its environmental suitability for the presence of STH infection , which should be further investigated . The Philippines Integrated Helminth Control Program has been in place since 2006 and has provisions for mass targeted and selective deworming . As indicated by the maps of predicted STH infection prevalence , there remain numerous areas that still have a high prevalence of STH infections . In these places , periodic treatment using albendazole or mebendazole remains essential . The coverage of the control program must also be considered . While school-based periodic administration of anti-helminthic drugs is part of the Philippines program , this is only typically carried out in the country’s public schools , despite findings that children in private schools are also affected by STH infections [28] . The results also suggest that it may be necessary to place greater emphasis on improving the provision of water , sanitation and the promotion of behavioral change for improved hygiene for the control and prevention of STH infections . This is particularly true for hookworm , as this infection seems to be associated with occupational exposures and therefore more likely to be missed by targeting schools . The results of the study should be interpreted in light of the studies’ limitations . First , socioeconomic factors , such as access to water and sanitation and hygiene behavior ( WASH ) are important predictors of the small-scale spatial variation of STH infection . While WASH indicators are measured in the Philippines as part of nationwide Demographic Health Surveys ( DHS ) , unfortunately water and sanitation indicators were not captured in the initial survey . Future research will investigate approaches to integration of available WASH data and STH infections to determine WASH associations and changes in time in relation to STH risk . Overall the results show that STH infections in the Philippines are a widespread , major public health problem and depending on parasite species , these show remarkable spatial variation even within known endemic areas . Further surveys should be prioritized to areas in Luzon ( including Palawan ) , which are currently underrepresented in our database . | Soil-transmitted helminth ( STH ) infections with A . lumbricoides , T . trichiura and hookworms are endemic in all 80 provinces of the Philippines , but the spatial variation in the prevalence of these infections has not been previously described . This analysis revealed that while A . lumbricoides and T . trichiura infections were widespread and highly endemic , hookworm infections were more circumscribed to smaller foci in the Visayas and Mindanao . The results also suggest that it may be necessary to place greater emphasis on improving the provision of water , sanitation and the promotion of behavioral change for improved hygiene for the control and prevention of STH infections , particularly for hookworm . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Mapping the Risk of Soil-Transmitted Helminthic Infections in the Philippines |
Although dengue is endemic in Puerto Rico ( PR ) , 2007 and 2010 were recognized as epidemic years . In the continental United States ( US ) , outside of the Texas-Mexico border , there had not been a dengue outbreak since 1946 until dengue re-emerged in Key West , Florida ( FL ) , in 2009–2010 . The objective of this study was to use electronic and manual surveillance systems to identify dengue cases in Veterans Affairs ( VA ) healthcare facilities and then to clinically compare dengue cases in Veterans presenting for care in PR and in FL . Outpatient encounters from 1/2007–12/2010 and inpatient admissions ( only available from 10/2009–12/2010 ) with dengue diagnostic codes at all VA facilities were identified using VA's Electronic Surveillance System for Early Notification of Community-based Epidemics ( ESSENCE ) . Additional case sources included VA data from Centers for Disease Control and Prevention BioSense and VA infection preventionists . Case reviews were performed . Categorical data was compared using Mantel-Haenszel or Fisher Exact tests and continuous variables using t-tests . Dengue case residence was mapped . Two hundred eighty-eight and 21 PR and FL dengue cases respectively were identified . Of 21 FL cases , 12 were exposed in Key West and 9 were imported . During epidemic years , FL cases had significantly increased dengue testing and intensive care admissions , but lower hospitalization rates and headache or eye pain symptoms compared to PR cases . There were no significant differences in clinical symptoms , laboratory abnormalities or outcomes between epidemic and non-epidemic year cases in FL and PR . Confirmed/probable cases were significantly more likely to be hospitalized and have thrombocytopenia or leukopenia compared to suspected cases . Dengue re-introduction in the continental US warrants increased dengue surveillance and education in VA . Throughout VA , under-testing of suspected cases highlights the need to emphasize use of diagnostic testing to better understand the magnitude of dengue among Veterans .
Dengue virus ( DENV ) , a flavivirus with 4 serotypes , transmitted by Aedes mosquitoes can cause a spectrum of disease from a mild febrile illness with constitutional symptoms to a severe hemorrhagic illness [1] , [2] . An important risk factor for severe dengue , dengue hemorrhagic fever ( DHF ) or dengue shock syndrome ( DSS ) , is a previous infection with another DENV serotype [2] . Dengue is seen throughout the world in tropical regions and has been endemic in Puerto Rico for many years [3] , [4] , [5] . More recently , 2007 and 2010 were recognized as epidemic years in Puerto Rico with increased rates of dengue cases reported [3] , [4] , [5] . In the continental United States ( US ) , outside of the Texas-Mexico border , there had not been a dengue outbreak since 1946 until 2009–2010 when there was an outbreak of locally acquired dengue ( DENV-1 ) in Key West , Florida [6] , [7] , [8] , [9] , [10] . Why DENV circulated in Key West , Florida , starting in July 2009 through 2010 is not entirely clear . In Florida and Puerto Rico , dengue was a reportable disease to the department of health prior to the identification of endemic cases in Florida in 2010 . However , because of the rise in the number of dengue cases and the risk of transmission in the continental US , in 2010 , dengue became one of Centers for Disease Control and Prevention's ( CDC ) nationally notifiable diseases based on the Council of State and Territorial Epidemiologists ( CSTE ) dengue case definition [11] . The Department of Veterans Affairs ( VA ) medical facilities , particularly in Puerto Rico and Florida , have started to perform surveillance for and report new cases of dengue to county and state/territory health departments . The objective of this study was to evaluate dengue cases identified in VA facilities by electronic and manual methods . We subsequently compared clinical symptoms , laboratory data , illness severity , and differences between confirmed/probable and suspected cases presenting for care in Puerto Rico from 2007–2010 ( including the epidemic years of 2007 and 2010 ) and compared these cases to dengue cases identified in Florida VA facilities between 2007 and 2010 ( including the epidemic years of 2009 and 2010 ) to better characterize dengue identified in Puerto Rico and Florida VA facilities .
This project was approved by the Stanford University Institutional Review Board . The Human Subjects Research Panel at Stanford University determined that the study entitled “Healthcare-Associated Infections and Syndromic Surveillance in the Department of Veterans Affairs” met the requirements of regulation OHRP 45 CFR 46 . 116 ( d ) : Requests for waiver or alteration of the informed consent process , in research that is not subject to FDA regulation in that: ( 1 ) The research involved no more than minimal risk to the subjects; ( 2 ) the waiver or alteration would not adversely affect the rights and welfare of the subjects; ( 3 ) the research could not practicably be carried out without the waiver or alteration and ( 4 ) the subjects would be provided with additional pertinent information after participation . This study was approved because the data used for its conduct was retrospective and would be obtained through subject electronic medical records ( EMR ) from primary care doctors , thus it was not anticipated that any situation would arise in which pertinent information would need to be shared with individual subjects . All data extracted from the medical record during the public health investigation was analyzed anonymously . We publish findings from this study and established the database regarding infection control as a tool for providers , thus patients would learn and benefit from this study through the care provided by their primary care doctors . Dengue surveillance in the VA was done by a combination of electronic and manual surveillance . VA's Electronic Surveillance System for the Early Notification of Community-based Epidemics ( ESSENCE ) biosurveillance system was used to identify outpatient and emergency department visits ( January 2007–December 2010 ) and inpatient admissions ( available data from October 2009–December 2010 ) with dengue ICD-9 codes ( 061 and 065 . 4 ) at all 152 VA medical centers and over 970 outpatient clinics in the United States and its territories , including 5 facilities in Puerto Rico and 67 in Florida [12] , [13] . Cases were also identified in Florida using VA data transmitted to CDC's BioSense system for 2010 . BioSense captured possible dengue cases using dengue ICD-9 codes as well as syndromic definitions ( fever with rash; fever with unexplained bleeding; or fever with thrombocytopenia ) [14] . We also identified cases that were manually collected by infection preventionists in Florida and Puerto Rico and reported to VA's Office of Public Health . All identified records were further verified by chart review . Extensive , standardized reviews of VA's EMR were completed for VA visit encounters and admissions of suspected dengue cases , identified by the above methods in both Puerto Rico and Florida . In Puerto Rico , epidemic years were 2007 and 2010 and in Key West , Florida , epidemic years were 2009–2010 [3] , [4] , [5] , [9] . Cases were grouped into 3 categories: confirmed ( positive DENV polymerase chain reaction [PCR] , anti-DENV Immunoglobulin M [IgM] seroconversion , ≥4-fold rise in anti-DENV Immunoglobulin G [IgG] ) ; probable ( anti-DENV IgM present with a Positivity/Negativity [P/N] ratio ≥2 ) ; and suspected ( a clinically compatible case , epidemiologically linked to a confirmed or probable case or with travel to a dengue endemic country or presence at a location with an ongoing outbreak within the previous 2 weeks of dengue-like illness , with fever and 2 or more of the following symptoms: retro-orbital or ocular pain , headache , rash , myalgia , arthralgia , leukopenia , or hemorrhagic manifestations without confirmatory laboratory testing or incomplete laboratory testing ) [11] . Laboratory testing in Florida was done at hospital or commercial labs and at health departments and in Puerto Rico dengue testing was done at the CDC Division of Vector-Borne Diseases in the Dengue Branch in San Juan , Puerto Rico . In Florida , testing was primarily serologic until 2010 when Florida Department of Health acquired DENV PCR testing capabilities . Patient demographics and clinical signs and symptoms were extracted from encounter notes in the EMR . The extracted EMR data included age , gender , whether a patient was hospitalized and/or received intensive care unit ( ICU ) care , whether they received platelet or packed red blood cell transfusions , and laboratory results such as dengue laboratory testing , platelet count , hematocrit , and white blood cell count ( WBC ) . Documented symptoms such as fever , arthralgia , myalgias , headache , eye pain , skin manifestations/rash ( including petechiae ) , gastrointestinal ( GI ) symptoms ( nausea , vomiting , and diarrhea ) , upper respiratory infection ( URI ) symptoms ( cough , nasal congestion , sore throat ) , and any documentation of bleeding were also extracted from the EMR . Patients noted to have been treated at non-VA hospitals or with other diagnoses or other reasons for their symptoms were excluded . Clinical and laboratory data from EMRs were reviewed and classified based on the 2010 CSTE dengue case definition [11] . Confirmed and probable cases were combined since they were patients that had at least one positive dengue test . Using Epi Info ( CDC , Atlanta , GA ) , Mantel-Haenszel or Fisher exact tests ( if a count was less than 5 ) were used to compare categorical count data with a p-value≤0 . 05 signifying a statistical difference . Continuous variables were compared using SAS 9 . 2 ( SAS Institute , Cary , NC ) Student's t-test .
In VA facilities in the US , including territories as well as the Philippines , there were a total of 339 VA cases of dengue identified between 2007 through 2010 . Of those 339 cases , 288 and 21 dengue cases were identified in Puerto Rico and Florida , respectively . The 30 remaining cases were acquired outside the continental US while patients were traveling or in VA facilities located in other US territories in dengue endemic areas ( Table 1 ) . All 288 Puerto Rico dengue cases are believed to have been acquired in Puerto Rico . Dengue cases in Puerto Rico presented to VA facilities in San Juan , Ponce , Mayaguez and Arecibo ( Figure 1 ) . Twelve of 21 Florida cases were acquired in Key West , Florida , and 9 were acquired while traveling outside Florida in dengue endemic areas from 2007–2010 ( Figure 2 ) . In the 2009 analysis , there was only one patient included presenting to a Florida VA who was exposed in Puerto Rico before the start of the Florida epidemic in July 2009 . In Puerto Rico , there were 65 cases in 2007 , 13 in 2008 , 30 in 2009 , and 180 in 2010 . In Florida , there were 0 cases in 2007 , 2 in 2008 , 7 in 2009 , and 12 in 2010 . Dengue surveillance in Florida started in 2010 allowing a comparison between methods of capturing cases . Out of the 12 confirmed/probable cases identified in Florida in 2010 , ESSENCE , infection preventionists , and BioSense identified 12 ( 12/12 , 100% ) , 9 ( 9/12 , 75% ) , and 9 ( 9/12 , 75% ) confirmed/probable cases of dengue respectively ( Figure 3 ) . Although not statistically different , there were 5 . 5 and 9 . 5 times as many cases during epidemic years in comparison to non-epidemic years during 2007–2010 ( 245 vs . 43 cases in Puerto Rico and 19 vs . 2 cases in Florida ) . Overall , VA dengue cases in Florida and Puerto Rico presented at a mean age of 55 years old ( range 21 to 90 years ) and 94% ( 289/309 ) were male . Reported symptoms and signs were fever ( 298/305 , 98% ) , arthralgias/myalgias ( 252/276 , 91% ) , eye pain ( 103/168 , 61% ) , thrombocytopenia [platelets <150 K/mm3] ( 270/305 , 89% ) , headache ( 203/247 , 82% ) , GI symptoms ( 181/268 , 68% ) , leukopenia [WBC <4 K/mm3] ( 188/305 , 62% ) , documented skin manifestation/rash ( including petechiae ) ( 104/227 , 46% ) , URI symptoms ( 107/285 , 46% ) , and bleeding ( 16/224 , 7% ) . Five percent of patients received a platelet transfusion . Dengue diagnostic testing being performed among suspected or confirmed/probable cases was significantly different between facilities in Puerto Rico and Florida , respectively ( 61% ( 176/288 ) and 100% ( 21/21 ) , p<0 . 01 ) . In 2010 , one patient in Florida had DENV PCR testing performed and was identified as DENV-1 , which was the serotype identified by others during this time period in Key West , Florida [10] . In contrast to Florida , Puerto Rico performed more DENV PCR testing ( Florida 1/21 cases vs . Puerto Rico 118/288 cases , p<0 . 01 ) . DENV-1 was the predominant serotype for both epidemic and non-epidemic years in Puerto Rico , accounting for 52/79 ( 66% ) of all serotyped isolates . There was an increase in DENV-4 isolates identified in Puerto Rico during 2010 vs . 2009 ( 12 vs . 1 case ) . No patients died as a result of dengue or met criteria for DHF or DSS . Confirmed/probable cases were significantly more likely than suspected cases to be hospitalized ( 69% vs . 47% , p<0 . 01 ) , to have thrombocytopenia [platelets <150 K/mm3] ( 96% vs . 84% , p<0 . 01 ) , and to have leukopenia [WBC <4 K/mm3] ( 80% vs . 51% , p<0 . 01 ) . Confirmed/probable cases were significantly less likely to report URI symptoms ( 20% vs . 50% , p<0 . 01 ) compared to suspected cases ( Table 2 ) . Forty-two percent of suspected patients had incomplete dengue diagnostic testing performed ( a single serologic test performed with no convalescent sample submitted ) . During the epidemic years ( 2009–2010 ) in Florida , 12 patients were linked to travel to or residence in Key West , Florida , while 7 patients had documented travel outside of Florida to other endemic areas prior to their diagnosis of dengue ( Figure 2 ) . The 12 patients with exposure in Key West presented to VA facilities in Key West ( 8 ) , Miami ( 2 ) , Tampa ( 1 ) and Bay Pines ( 1 ) ( Figure 2 ) . These groups were compared and no statistical difference was identified for the patient characteristics listed in Table 2; however , there was a trend towards more hospitalizations and more ICU care received by patients that had traveled outside of the continental US ( data not shown ) . The 7 patients with documented travel outside of Florida were included as part of the Florida epidemic analysis since they presented to a Florida VA facility during the evaluation period . Florida and Puerto Rico epidemic cases ( suspected and confirmed/probable ) were compared to each other ( Table 2 ) . In Florida , cases were significantly more likely to have any dengue diagnostic testing completed ( 100% vs . 62% , p<0 . 01 ) , less likely to be hospitalized ( 32% vs . 56% , p = 0 . 04 ) but more likely to receive ICU care ( 33% vs . 5% , p = 0 . 05 ) . In Florida , patients were significantly less likely to report symptoms of headache ( 53% vs . 84% , p = 0 . 01 ) and eye pain ( 30% vs . 64% , p = 0 . 05 ) . The rest of the patient characteristics were not significantly different . Although the numbers are small , the Florida and Puerto Rico non-epidemic years were compared and there were no significant differences in patient characteristics between these groups ( data not shown ) . Epidemic years of dengue in Puerto Rico and Florida were combined and compared to non-epidemic years . None of the patient characteristics were significantly different between the groups . However , there was a higher proportion of hospitalizations and documented skin manifestations/rash in the non-epidemic years ( data not shown ) .
The re-introduction of DENV in the continental US has made it an important infection for VA providers to understand , especially since approximately two-thirds of confirmed/probable cases and almost half of the suspected cases were hospitalized . The majority of endemic VA dengue cases during 2007–2010 were identified in Puerto Rico and Florida . In VA medical facilities in Florida , locally acquired dengue was limited to Key West , Florida , while all other VA cases detected in Florida during the epidemic had an exposure outside of the state . While the number of VA dengue cases in Florida decreased in 2011 to 3 imported cases and is no longer showing sustained local transmission , the potential for further spread of DENV infection in Florida and other parts of the US is possible . Therefore , it is important to understand the clinical presentation , diagnostic testing patterns and epidemiology of dengue within the VA system . Overall , our dengue cases presented with expected signs and symptoms of fever , arthralgias/myalgias , headache , skin manifestations/rashes , headache , thrombocytopenia , and leukopenia . Clinically , our Florida cases appeared to be no different than those found in Puerto Rico . The similarity of clinical symptoms was not surprising since a DENV-1 strain , related to other Central American strains of dengue , was the predominant serotype identified in the Florida epidemic [8] . Based on the 2010 CSTE dengue case definition , confirmed/probable patients had similar characteristics as suspected cases . However , as expected based on the definition of a confirmed/probable case , all of these patients had dengue diagnostic testing , while only 42% had testing in the suspected group ( no lab test was needed to meet this definition and some suspected patients had initial dengue diagnostic testing performed but did not have convalescent serologic testing to confirm a diagnosis ) . Of note , less than two-thirds of suspected dengue cases in Puerto Rico received any dengue diagnostic testing , even though free diagnostic testing is available in Puerto Rico through CDC . In some cases , testing was never ordered and in others , samples were rejected because the CDC-required paperwork was either incomplete or not filled out properly . This finding highlights the need for additional education among VA providers regarding availability of testing and collection of required patient information . We also found that hospitalization , thrombocytopenia , and leukopenia rates were higher in the confirmed/probable group . Interestingly , URI symptoms were less likely in confirmed/probable cases; implying patients with URI symptoms were less likely to have dengue [15] . However , 20% of confirmed/probable cases reported URI symptom ( s ) so their presence does not preclude the diagnosis of dengue and testing patients that have URI symptoms and are suspected of having dengue remains important . The cases identified in Florida during the Key West epidemic were similar in nature and severity to those seen during epidemic years in Puerto Rico except for the significantly increased amount of testing and increase in cases receiving ICU care in Florida . Interestingly , there was a lower incidence of overall hospitalization for dengue in Florida during the epidemic years . This suggests that even though fewer patients were hospitalized in Florida , when they were hospitalized they received ICU level care . However , although the numbers are small , the number of hospitalized cases in Florida was low and when chart reviews were performed on the patients receiving ICU level care it revealed either atypical presentations or a provider's decision for closer monitoring in an ICU setting , suggesting a relative discomfort in treating dengue in Florida facilities compared to Puerto Rico . There were a lower percentage of cases presenting with headache and eye pain in Florida , however , this may be related to fewer providers asking patients and documenting these symptoms in the EMR . In addition , the cases of dengue during epidemic and non-epidemic years had similar characteristics . Limitations of our study included a small cohort number that was predominantly older males . In addition , some Veteran cases could have been missed if they were treated at outside facilities and not reported or if they were not coded as having dengue . Dengue PCR was not as commonly used in the Florida cases so we are unable to determine the most common serotype involved with the VA dengue cases in the continental US , or to have samples to do further molecular typing and sequencing . Since we combined confirmed and probable dengue cases it is possible that we slightly overestimated the number of actual dengue cases since elevated anti-DENV IgM may be due to a cross reactivity with other flaviviruses ( West Nile virus , St . Louis encephalitis virus , Japanese encephalitis virus , and yellow fever virus ) , although this is unlikely as there are few places where these viruses co-circulate or where these conditions cannot be differentiated clinically [16] , [17] . In addition , many Veterans are vaccinated against yellow fever which can have cross reactivity with dengue serologic testing [16] . However , the number of probable cases was small therefore it is unlikely to have greatly affected our analysis . Unfortunately , primary/secondary dengue infection status is not reliably documented in our EMR making it impossible to compare prior dengue exposure to symptom severity . The primary goal of our study was not to compare different surveillance system performance for DENV detection . However , accurate DENV case finding required the combination of two electronic biosurveillance systems ( ESSENCE and BioSense ) , as well as infection preventionist manual surveillance efforts at VA facilities . These electronic biosurveillance systems currently rely on outpatient diagnostic encounter codes , ICD-9 , which can be searched by syndrome or individual codes . ICD-9 coding for outpatient visits in VA may not be completely accurate , and likely underestimates the true number of cases , particularly in those cases where confirmatory laboratory testing was not obtained , or was obtained and results were not available at the time of encounter close-out . Syndromic surveillance includes additional , non-specific ICD-9 codes ( i . e . , fever and rash ) , which can further reduce the specificity of the diagnosis . In addition , DENV or syndrome ICD-9 codes could reflect prior outpatient encounters for DENV disease , and contribute to an overestimation of the number of cases . Infection preventionists can access multiple data domains in the EMR ( including history , laboratory data , and treatment ) , which help refine whether a potential DENV case is likely to be a confirmed or probable case . In addition , infection preventionists can help facilitate obtaining convalescent blood samples to further help confirm diagnoses . As demonstrated in Figure 3 no system of identifying cases was perfect , both infection preventionists and BioSense were able to identify 9 out of the 12 confirmed/probable cases . ESSENCE was able to capture all 12 cases , however , misidentified 3 cases . Because of the reduced specificity of electronic biosurveillance systems , VA is enhancing VA's ESSENCE system by including vital signs ( temperature ) , laboratory orders and results , inpatient admission data , outpatient encounter severity codes , telephone care encounter data , and pharmacy prescription data , in addition to ICD-9 encounter codes , which will improve specificity and automate much of what infection preventionists currently must review by hand . Until our enhanced system is available , utilization of an electronic surveillance system in addition to manual surveillance by infection preventionists will remain important . Although indigenous cases of DENV infection are rare in the continental US , after the epidemic of dengue in Key West , Florida , greater attention was placed on dengue surveillance , education and public health reporting . The VA Office of Public Health , CDC and Florida Department of Health collaborated on providing educational materials including a VA Dengue Health Alert that was produced in July 2010 to help educate VA providers on the presence of dengue in Florida . The alert advised providers to be vigilant for symptoms of dengue , to report suspected cases to local and state health departments , and to obtain appropriate laboratory testing for confirmation . Laboratory testing will hopefully become more widely available now that a DENV reverse transcription polymerase chain reaction ( RT-PCR ) assay developed by the CDC has been approved by the US Food and Drug Administration [18] . Additional details on laboratory testing algorithms and clinical guidance are available on the CDC website [19] . Increased efforts are necessary to improve dengue awareness in VA through patient and clinician education , and which emphasizes the need for testing , accurate coding of potential dengue cases and appropriate reporting to county and state health officials . | Dengue is an important tropical disease seen throughout the world in tropical climate zones and is spread by Aedes mosquitoes . Most cases of dengue in the continental US are imported . In July 2009 through 2010 , dengue virus was found to be circulating in Key West , Florida ( FL ) . Dengue virus has been transmitted in Puerto Rico ( PR ) for many years . This study used electronic and manual surveillance systems to identify dengue cases in VA healthcare facilities and clinically compared dengue cases in Veterans presenting for care in PR as well as in FL . We found that FL dengue cases were similar to those in PR and that Centers for Disease Control and Prevention defined confirmed/probable cases were more likely to be hospitalized within our VA system , and have either lower platelet or white blood cell counts than suspected cases . During July 2009–2010 , FL cases were more likely to be tested for dengue and have intensive care admissions , but had lower hospitalization rates and headache or eye pain symptoms compared to PR cases . No one method of capturing dengue cases was perfect . It is important to educate healthcare workers about this disease to help with direct patient care as well as surveillance . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"clinical",
"epidemiology",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"dengue",
"fever",
"neglected",
"tropical",
"diseases"
] | 2013 | Dengue Surveillance in Veterans Affairs Healthcare Facilities, 2007–2010 |
The establishment of connectivity between specific thalamic nuclei and cortical areas involves a dynamic interplay between the guidance of thalamocortical axons and the elaboration of cortical areas in response to appropriate innervation . We show here that Sema6A mutants provide a unique model to test current ideas on the interactions between subcortical and cortical guidance mechanisms and cortical regionalization . In these mutants , axons from the dorsal lateral geniculate nucleus ( dLGN ) are misrouted in the ventral telencephalon . This leads to invasion of presumptive visual cortex by somatosensory thalamic axons at embryonic stages . Remarkably , the misrouted dLGN axons are able to find their way to the visual cortex via alternate routes at postnatal stages and reestablish a normal pattern of thalamocortical connectivity . These findings emphasize the importance and specificity of cortical cues in establishing thalamocortical connectivity and the spectacular capacity of the early postnatal cortex for remapping initial sensory representations .
A dynamic interplay exists between the processes of cortical arealization and those controlling the guidance and targeting of thalamocortical projections [1–5] . Early in development , both the thalamic field and the cortical sheet appear homogeneous in cytoarchitecture , and connections between them form in a smoothly topographic fashion , with dorsolateral thalamus projecting to caudal cortex and ventromedial thalamus to rostral cortex [6–8] . The cytoarchitectonic resolution of these fields into discrete cortical areas and thalamic nuclei occurs later [8–13] with the elaboration of many aspects of the cortical areas dependent on appropriate thalamic innervation [1–5 , 14 , 15] . Several lines of evidence have led to the theory that subcortical sorting of thalamic axons within the ventral telencephalon largely determines their final targeting within the cortex [16–20] . For example , in mutants in the transcription factor Ebf1 or in the Dlx1/Dlx2 double mutants , a subset of thalamic axons is misrouted ventrally , resulting in a caudal shift of the remaining axons within the ventral telencephalon [16] . This shift is projected onto the cortex so that at birth , caudal cortical areas are contacted by axons that would normally project to more rostral areas . The ultimate effect of this derangement on thalamocortical connectivity could not be assessed in these mutants , however , as they die perinatally . On the other hand , many experiments have revealed the existence of cortex-specific cues that control thalamocortical targeting [21–27] . For example , changes in patterning of the cortical sheet in Emx2 [21 , 24] , Fgf8 [23] , or COUP-TF1 [27] mutants lead to parallel alterations in the patterns of thalamocortical connectivity . In each of these cases , manipulations solely in the cortex dramatically affect thalamocortical targeting . Indeed , ectopic expression of Fgf8 in either the subplate or cortical plate further revealed that thalamocortical axons ( TCAs ) are responsive to guidance cues present in both the subplate and cortical plate [26] . The interplay between subcortical and cortical mechanisms in determining eventual thalamocortical connectivity thus remains to be resolved . To get a better understanding of the interactions between areal patterning and thalamic axon guidance , we have used the Sema6A gene trap mouse . As a consequence of Sema6A disruption , a large fraction of thalamic projections gets derailed within the ventral telencephalon [28] . As these mice survive to adulthood they provide a unique model with normal cortical patterning but altered thalamic input during embryonic life . Our study reveals a changing pattern of thalamocortical development in the Sema6A mutants , drawing attention to the spectacular capacity of the cortex for altering and organizing its initial sensory representation . In particular , our findings suggest that thalamic axons from the dorsal lateral geniculate nucleus ( dLGN ) can target their correct area even if they arrive there days later than normal , through alternate subcortical routes . They also indicate that dLGN axons can out-compete invading axons from inappropriate thalamic nuclei , establishing a surprisingly normal adult cortical representation .
Sema6A is broadly expressed in the thalamus at embryonic day ( E ) 14 . 5 , a time when thalamic neurons are extending axons towards their cortical targets [28] , with expression highest in the dorsolateral aspect ( n = 5; Figure 1A and 1B ) . Sema6A is also strongly expressed in the amygdala and the ventral telencephalon , and weakly expressed in the neocortex at this age , localized to the most superficial compartments ( Figure 1A and 1B ) . At late embryonic stages , Sema6A is also expressed by deep cortical plate neurons , eventually layer 5 ( unpublished data ) . Staining with the axonal marker placental alkaline phosphatase ( PLAP ) , encoded on the gene trap cassette in this Sema6A allele , revealed thalamic axons extending from the thalamus through the internal capsule and towards the neocortex ( Figure 1C and 1D ) . A previous study using PLAP staining showed that many thalamic axons were misrouted in the Sema6A−/− brains at embryonic stages [28] . To further examine the guidance of TCAs in the absence of functional Sema6A protein , we performed carbocyanine dye tracing studies . Broad injections of DiI in the thalamus ( including the dorsolateral aspect ) at E15 . 5 revealed a prominent derailment of thalamic axons at the surface of the ventral telencephalon and amygdala in Sema6A−/− embryos ( n = 4/4; Figure 2D and 2G–2I ) , compared to the normal route of navigation through the internal capsule towards the neocortex observed in wild-type embryos ( n = 4/4; Figure 2A and 2C ) . The derailment of a large proportion of TCA fibers at the ventral telencephalon in Sema6A−/− brains at E16 . 5 was confirmed by neurofilament ( NF ) immunohistochemistry ( n = 3; Figure 2E and 2F ) and by PLAP staining ( Figure S1A and S1C ) . Overlaid consecutive serial sections of Sema6A−/− brains at E17 . 5 stained for NF reveal more clearly the extent of the TCA derailment ( n = 3; Figure 2H ) . To identify the precise origin of the misrouted thalamic axons within the thalamus , we placed small DiI crystals at the site of the derailed fibers near the ventral surface of the telencephalon ( Figure 3A ) in wild-type ( n = 4 ) and Sema6A−/− ( n = 4 ) postnatal day ( P ) 0 brains . Whereas no back-labeled cells were observed in any thalamic nuclei in wild-type brains ( Figure 3B and 3C ) , several axon bundles were retrogradely traced to the dorsolateral aspect of the thalamus in Sema6A−/− brains ( Figure 3D and 3E ) . Retrogradely labeled cells were specifically found in the presumptive dLGN ( Figure 3G and 3H ) . Some labeled bundles were also observed ascending laterally towards the cortex ( Figure 3E ) . Moreover , in Sema6A−/− brains , at more-caudal telencephalic levels , DiI-labeled axons were observed running through the intermediate zone of the primary visual cortical area ( Figure 3F and 3I; and unpublished data ) , suggesting that some dLGN axons that follow this abnormal route might still reach the visual cortex . Indeed , whereas a DiA crystal placed in the internal capsule zone of wild-type brains at E17 . 5 back-labeled cells throughout the dorsal thalamus ( n = 2/2; Figure S2A–S2C ) , a similarly placed DiA crystal in Sema6A−/− brains at the same age strikingly did not label cells in the dLGN ( n = 2/2; Figure S2D–S2F ) . To further study the trajectories of the dLGN projections , we placed small DiI crystals into the dLGN of wild-type and Sema6A−/− brains at P0 . At this age , wild-type and heterozygous dLGN axons extended through the internal capsule and arrived to the visual cortex in a normal fasciculation pattern ( n = 8/8; Figure 4A–4D ) . In Sema6A−/− brains , in contrast , dLGN axons projected ventrally to the most superficial regions of the ventral telencephalon . Despite this , a small number of dLGN axon terminals were observed in visual cortex ( Figure 4G; n = 9/9 ) . These results demonstrate that the vast majority of visual TCAs are severely affected by the lack of Sema6A function in vivo , but some still reach presumptive visual cortex by birth , possibly by alternate routes . To determine whether the initial derailment of dLGN axons could affect the general topographical arrangement between neocortex and thalamus , we performed multiple cortical dye placements at E16 . 5 and P0 . We placed DiI and DiA crystals into the putative visual and somatosensory cortices , respectively , in Sema6A+/− and Sema6A−/− mutant brains ( Figure 5 ) . In Sema6A+/− animals ( n = 6/6 E16 . 5; n = 5/5 P0 ) , a DiI crystal placement in the occipital neocortex labeled thalamic cell bodies and cortical axons in the dLGN ( Figure 5A and 5B ) . Placements of DiA in the parietal neocortex labeled cells and axon terminals in a more medial thalamic domain , where the ventrobasal ( VB ) complex is located ( Figure 5A and 5B ) . Interestingly , similar placements performed in Sema6A−/− embryos showed a lateral to medial shift in the thalamocortical and reciprocal corticothalamic connectivity ( n = 8/8 E16 . 5; n = 5/5 P0; Figure 5C and 5D ) . Specifically , placements into the occipital cortex labeled large numbers of cells in the lateral part of VB in Sema6A−/− brains , whereas few , if any cells at E16 . 5 were back-labeled in the dLGN ( Figure 5I–5L; 96 . 56 ± 4 . 6% of thalamic cells back-labeled from occipital cortex in E16 . 5 Sema6A−/− brains were found in the VB , compared to just 1 . 25 ± 2 . 25% of back-labeled thalamic cells in Sema6A+/− brains , p < 0 . 0001 ) . At P0 , an increase in the number of cells back-labeled in dLGN was observed in the mutants , though this was still far fewer than in wild-type animals ( Figure 5K and 5L; 16 . 24 ± 6 . 9% of thalamic cells back-labeled from occipital cortex in P0 Sema6A−/− brains were found in the dLGN , compared to 91 . 61 ± 3 . 17% of back-labeled thalamic cells in Sema6A+/− brains , p < 0 . 0001 ) . Placements of DiA in parietal cortex in Sema6A−/− brains back-labeled VB cells , though this labeling was more medial within this nucleus than in Sema6A+/− animals ( Figure 5C and 5D ) . Moreover , the areas back-labeled within VB by placements into occipital and parietal cortex were contiguous but showed minimal overlap . We did not observe any double-labeled cells . Taken together , these data show that in Sema6A−/− embryos and neonates , axons of the dLGN specifically are misrouted to the surface of the ventral telencephalon , and the topography of thalamocortical projections from VB is expanded caudally into the unoccupied occipital cortex . Because Sema6A is also expressed in retinal ganglion cells , we wanted to test whether abnormal development of retinal projections in Sema6A−/− embryos might secondarily cause the defects observed in the projection of visual TCAs . We injected DiI and DiA crystals in the eye and V1 , respectively , in wild-type and Sema6A mutant mice embryos at E16 . 5 . In wild-type embryos ( n = 3 ) , retinogeniculate projections were observed in the ventral lateral geniculate nucleus ( vLGN ) and dLGN ( Figure 5E and 5F ) . At the dLGN , this projection overlapped with back-labeled cells and cortical axons labeled from the visual cortical injections ( Figure 5F ) . However , in Sema6A−/− embryos ( n = 3/3 ) , although the interconnectivity between visual cortex and thalamus was shifted medially , retinogeniculate projections still arrived to the vLGN and dLGN in a normal fashion ( Figure 5G and 5H ) . Thus , the defects observed in the development of visual TCAs are not due to abnormal development of the projection from the retina or to abnormal differentiation of the LGN , which can still act as a specific target for retinal axons . Similarly , to test whether changes in cortical patterning could account for the topographical defect observed in TCAs connectivity in Sema6A−/− mutants , we performed in situ hybridization of specific cortical area markers on both wild-type ( n = 9 ) and Sema6A−/− ( n = 8 ) brains at P0 . In wild-type brains EphA7 is expressed both rostrally and caudally in the cortex , but is largely absent from parietal cortex , whereas EphrinA5 is expressed in a complementary pattern ( Figure S3A and S3C ) . We did not find any changes in the expression of EphA7 and EphrinA5 in Sema6A−/− brains ( Figure S3B and S3D ) . The expression of other cortical area markers ( Cad6 , Cad8 , and RZRβ ) was also unaltered in Sema6A−/− brains at P0 ( unpublished data ) . At earlier ( E14 . 5 and E16 . 5 ) and later ( P7 ) developmental stages , there were also no differences observed in the expression of cortical area markers between wild-type and Sema6A−/− brains ( unpublished data ) . These data indicate that intrinsic cortical patterning is unaltered in Sema6A mutants . We investigated thalamocortical connectivity at P4 using multiple cortical dye placements . Small crystals of DiI and DiA were placed in the primary visual ( V1 ) and primary somatosensory ( S1 ) cortices , respectively , in wild-type and Sema6A−/− brains ( Figure 6 ) . In wild-type brains ( n = 2/2 ) , a DiI placement in V1 labeled thalamic cell bodies and cortical axons in the dLGN , whereas a DiA placement in S1 labeled cell bodies in the VB ( Figure 6A and 6B ) . Interestingly , in Sema6A−/− brains ( n = 2/2 ) , a DiI placement in V1 mainly labeled thalamic cell bodies in the dLGN ( Figures 6C , 6D , and 5J; 79 . 49 ± 25 . 3% of thalamic cells back-labeled from occipital cortex in P4 Sema6A−/− brains were found in the dLGN , compared to 99 . 49 ± 1 . 51% of back-labeled thalamic cells in wild-type brains , p = 0 . 0428 ) , whereas far fewer cells were labeled in VB , in sharp contrast to results from similar tracing experiments at E16 . 5 and P0 ( Figure 5K . In Sema6A−/− brains at P4 , just 20 . 5 ± 25 . 3% of cells back-labeled from the occipital cortex were found in the VB , compared to 96 . 56 ± 4 . 68% and 83 . 65 ± 6 . 9% in Sema6A−/− brains at E16 . 5 and P0 , respectively , p < 0 . 0001 ) . A DiA placement in S1 labeled cell bodies in the VB as expected ( Figure 6C and 6D ) . This apparent rapid recovery of the normal thalamocortical connectivity was observed despite a persistent misprojection of TCAs from the dLGN to the ventral telencephalon in Sema6A−/− brains at P4 ( unpublished data ) . Although the topographical sorting of TCAs was apparently restored at this stage , fewer back-labeled cells were detected in the dLGN after visual cortical dye placements in Sema6A−/− compared to wild-type brains ( Figure 6B and 6D ) . To characterize the extent to which any shift in thalamocortical connectivity persists in the adult Sema6A−/− mouse , we performed in vivo stereotaxic tracing studies . We injected red and green cholera toxin ( CT ) dyes in V1 and S1 , respectively , in wild-type ( n = 6 ) and Sema6A−/− ( n = 8 ) mice at P60 ( Figure 6E , 6F , 6H , and 6I ) . At this adult stage , injections of red and green CT showed a normal topographical arrangement of the thalamocortical connectivity in Sema6A−/− brains ( Figure 6I ) when compared with wild-type brains ( Figure 6F ) . We also used two different colors of retrograde tracing microspheres , injected in V1 ( red ) and S1 ( green ) in wild-type or heterozygous ( n = 5 ) and Sema6A−/− ( n = 6 ) mice at P60 . The back-labeled thalamic cells of each color were plotted in the corresponding thalamic nuclei , dLGN or VB , in wild-type ( Figure 6G ) and Sema6A−/− brains ( Figure 6J ) . These data demonstrate that the early embryonic shift in TCAs connectivity , observed at E16 . 5 and P0 , is partially recovered at P4 and totally compensated in the adult Sema6A−/− mouse . To investigate whether the misrouted TCAs from the dLGN persist in the adult mouse , we performed PLAP staining studies to reveal the pathway of Sema6A-positive axons in both Sema6A+/− ( n = 4 ) and Sema6A−/− ( n = 4 ) brains . Sema6A is expressed in oligodendrocytes in adults , and this staining thus labels all myelinated fibers . Surprisingly , in Sema6A−/− brains at P60 , a misrouted bundle of axons was still observed at the ventral-most region of the telencephalon ( Figure 7F , 7L , and 7M ) in a similar location to the misrouted TCAs shown at earlier developmental stages . Similar ectopic bundles were never observed in any heterozygous adult brains ( Figure 7A ) . Interestingly , we could follow some PLAP-labeled axons up to the level of the visual cortex in Sema6A−/− brains ( Figure 7H–7M ) . Misrouted labeled axons appear to follow one of two alternate routes: ( 1 ) they either turn laterally through the amygdala and join the external capsule , or ( 2 ) they continue to project ventrally and extend along the superficial margin of the telencephalon ( Figure 7L–7N ) . In some cases , abnormal bundles could be followed either through the marginal zone or the external capsule up to the cortex ( Figure 7L ) . At more caudal regions , misrouted axons can be seen close to the pial surface of the cortex , including the visual cortex , of Sema6A−/− brains ( Figure 7K ) . To confirm the recovery pathways observed in PLAP-stained adult Sema6A−/− brains , a solution of DiI was injected into the primary visual cortex of wild-type or Sema6A−/− brains at P7 . These injections back-labeled cell bodies in the dLGN specifically of both wild-type ( n = 4; unpublished data ) and Sema6A−/− brains ( n = 3; Figure 8A ) . In wild-type animals , the axons of these back-labeled cells could be followed as they projected through the internal capsule to the cortex ( unpublished data ) . In contrast , in Sema6A−/− brains , many labeled axons were observed projecting ventrally and rostrally in the ventral telencephalon , turning laterally to reach the external capsule at more rostral levels ( Figure 8B–8D ) . Additionally , many axons were also seen to project in a more canonical fashion through the internal capsule , with some axons initially projecting ventrally before turning dorsally again to loop back and enter the internal capsule ( Figure 8D7 ) , suggesting that many initially misrouted dLGN axons may extend a branch at later stages and extend through the internal capsule to reach the cortex . To investigate whether the massive misrouting of TCAs during early embryonic stages has an impact on the development of the dLGN nucleus and visual cortex , we performed a series of histochemical studies in adult Sema6A−/− mice . Using both Nissl and cytochrome oxidase staining , we observed a significant reduction in the volume of the dLGN in Sema6A−/− brains ( n = 4 ) compared with wild-type littermates ( n = 4 ) at P30 ( Figure 9A–9E ) . These data suggest that dLGN neurons whose axons do not reach their target on time die during development . To further test this possibility , we performed an immunostaining against a cleaved caspase-3 to detect apoptotic cells in the thalamus of wild-type and Sema6A−/− brains at P4 . At this stage , we observed an almost 3-fold increase in the number of apoptotic cells in the dLGN of Sema6A−/− brains ( n = 5 ) compared to controls ( n = 6; Figure 9F–9J ) , suggesting that many of the dLGN neurons , whose axons were misrouted , do indeed die during development . We next investigated whether the early lack of dLGN afferents to the visual cortex together with the transient invasion of somatosensory input to this cortical area might affect the final relative representation of cortical areas in Sema6A−/− adult mice . We examined the cortical area occupied by S1 and V1 in tangential sections stained for serotonin ( 5HT ) immunoreactivity in wild-type ( n = 5 ) and Sema6A−/− ( n = 5 ) brains at P7 . Although we observed no changes between wild-type and Sema6A−/− brains in the relative position of these cortical areas ( Figure 9K and 9L ) , we observed a significant reduction in the size of V1 in Sema6A−/− mouse brains ( Figure 9L and 9M ) . Moreover , we observed a consistent change in the shape of the V1 cortical domain in Sema6A−/− mouse brains compared to wild-type littermates ( Figures 9K and 9L ) . No changes were observed in the position and dimensions of the barrel field in Sema6A−/− mice . Together , these results strongly suggest that the reduction in the size of dLGN in Sema6A−/− mice leads to a reduction in the size of V1 .
Our study of the Sema6A mutants revealed an initial subcortical pathfinding defect of thalamic axons specifically from the dLGN . This results in expansion of somatosensory thalamic axons into presumptive visual cortex during embryonic stages . Due to the viability of these mutants , we were able to assess the secondary consequences of early misrouting of the visual axons on postnatal cortical specification and adult thalamocortical topography . Remarkably , many dLGN axons are able to find their way to visual cortex during early postnatal stages , following alternate routes , and can establish almost normal patterns of thalamocortical connectivity in the adult . The general implications of these findings for principles of thalamic axon guidance and cortical arealization are discussed below . The failure of dLGN axons to arrive to the occipital cortex in Sema6A−/− brains at embryonic stages results in the dramatic expansion of the domain of VB axons into this region . Importantly , we observe no changes in cortical gene expression patterns at early stages , indicating that the removal of Sema6A from the cortex does not affect global cortical patterning . This early caudal shift of thalamocortical targeting has also been observed in the other mutants with misprojected dLGN axons ( i . e . , Ebf1 , Dlx1–2 double mutants [16]; reviewed in [2 , 18] ) . We observed back-labeling of VB from injections placed in occipital cortex , but did not observe a dramatic shift in connectivity from more rostral injection sites in parietal cortex ( the normal position of S1 ) , which still back-labeled VB ( though more medially ) . This is more consistent with a caudal expansion of the innervation zone of VB axons in Sema6A mutants than with an overall shift of all thalamic connections . A current model of thalamic axon pathfinding proposes an essential role for intermediate targets , in the ventral telencephalon ( vTel ) , in guiding TCAs to specific cortical areas [2 , 17 , 18 , 20 , 29–31] . Mutations in a number of genes expressed predominantly in the vTel ( Ebf1 , Dlx1/2 ) affect thalamic projections subcortically in a manner that seems to be passively carried through in their projections to the cortex , at least at birth [16] . Analysis of the Sema6A mutants at postnatal stages reveals , however , that initially misrouted axons from the dLGN can eventually make appropriate connections to visual cortex . Remarkably , many of these projections seem to occur through alternate routes , either via the external capsule or a superficial route along the outside of the telencephalon . It is also possible that some projections are made via collaterals through the internal capsule that arise at later stages . These findings demonstrate that correct subcortical axonal sorting is not required for eventual projection to a specific cortical area and , further , that the normal temporal sequence of arrival of thalamic axons to the cortex is also not essential for correct targeting . In addition , they show that subcortical sorting is not sufficient to permanently determine connectivity as the initial shift in cortical targeting of VB axons that is apparent at embryonic stages can be corrected after birth . These conclusions are consistent with a growing body of research demonstrating the existence of cortical guidance cues for thalamic axon rearrangements [11 , 13 , 26 , 32] and suggest that the actions of these signals may be effective at a distance [22] to selectively attract misrouted dLGN axons to the appropriate cortical area . They also suggest that guidance cues within the neocortex exist not just in the subplate , but also across the developing cortical layers [26] , allowing navigation even in the marginal zone , as demonstrated by the pathway follow by some of the misrouted LGN axons in the Sema6a−/− mutants . The interpretation that subcortical sorting does not determine final cortical targeting would seem to be challenged by a number of other mutants that show early , global defects in subcortical thalamic projections , accompanied by later , highly specific defects in thalamocortical connectivity . For example , in double Ephrin-A5;EphA4 mutants [30] and in mutants in either CHL1 or Npn1 [20] , rostral thalamic axons project more caudally than normally both subcortically and up to the cortex itself at embryonic stages . In both cases , a defect in thalamocortical connectivity is also apparent at postnatal stages , involving excess connectivity of one thalamic nucleus with a particular cortical area , although it is much more selective , and differs between Ephrin-A5;EphA4 and CHL1 or Npn1 mutants . In both cases , the early defect was interpreted as the cause of the later defect , but this has not been shown directly and the selective ( and different ) nature of the defects at later stages suggests that most of the early misrouting has in fact been corrected and that the postnatal connectivity defects are more likely to reflect later functions of these genes in the cortex itself . Overall , these studies and our data are thus consistent with a model in which subcortical sorting of thalamic axons is coordinated with eventual cortical targeting , possibly using the same cues at both levels . However , subcortical targeting does not appear to be either strictly necessary or sufficient to determine final connectivity patterns as additional mechanisms exist to restore thalamocortical connectivity to a specific cortical area when alterations during embryonic development occur . The recovery of the dLGN projection to visual cortex , in spite of previous occupation of this territory by VB axons suggests that dLGN axons have an advantage in the innervation of that particular cortical area . This must be in addition to selective axon guidance to this region as arrival of VB axons to this area is clearly not sufficient to enable them to make permanent connections , at least when faced with competition from later-arriving dLGN axons . A model to explain this would be that dLGN axons and presumptive visual cortex express some matching label ( s ) that confer this advantage . One candidate for such a cue is the neurotrophin NT-3 , which is specifically required for dLGN axons to invade the cortical plate in V1 [33] . NT-3 has been shown to be most strongly expressed in presumptive visual cortex ( V1 ) from around P0 [34] , while its receptor TrkC , is selectively expressed by neurons in the dLGN . If such a matched cue is essential then VB axons that at early stages project into the subplate of occipital cortex may not be able to invade the cortical plate , allowing later-arriving dLGN axons to do so . Indeed , if the function of NT-3 in this context shares similarities with trophic signaling [35] then dLGN axons might actively secrete factors that promote withdrawal of VB axons . Axon–axon interactions mediated by surface receptors and cell adhesion molecules [36] might also actively mediate segregation of visual and somatosensory axons [37] . Activity-dependent mechanisms mediating the competitive advantage of dLGN axons for presumptive visual cortex must also be considered , especially as the process takes place during the first few postnatal days , by which time thalamic axons have normally entered into the cortex and formed fully functional synapses [32 , 38 , 39] . A number of studies have examined the potential role of electrical activity in areal targeting of thalamic axons . Intracranial infusion of the sodium channel blocker tetrodotoxin ( TTX ) caused dLGN axons to inappropriately innervate the subplate of cortical areas that they would normally bypass [40] . This could be taken as an instructive role for patterned activity in establishing areal connectivity but could alternatively be explained by an earlier effect of TTX on biochemical signaling pathways downstream of guidance receptors [41] , or by feedback onto the expression levels of guidance molecules [42] . This interpretation is more consistent with the known specificity of thalamic axon targeting from the earliest stages [10 , 11 , 43] and the lack of effects in areal targeting observed in embryonic SNAP-25 mutants [39 , 44] . Finally , although our study demonstrates spectacular plasticity of thalamocortical connectivity during early postnatal life , there are some changes in the cortical architecture that persist into adulthood . The reduction in size and change in shape of V1 in Sema6A mutants , which are far more subtle than those observed in enucleation experiments [1 , 5 , 45 , 46] , suggest that they may be an interesting model to study some less well-characterized processes , including the separation of the termination zones of primary thalamic axons into discrete areas , the innervation of intervening areas by axons from secondary nuclei , the formation of distinct borders and the hierarchical dependence of secondary and higher-order areas on correct specification of primary areas ( reviewed in: [47 , 48] ) .
All animal procedures were performed to relevant national and international licensing agreements and in accordance with institutional guidelines . Sema6A mutants were identified in a gene trap screen , as described previously [49] . Insertion of the gene trap vector pGT1PFS into intron 17 results in a fusion of upstream exons of Sema6A with TM-β-galactosidase-neomycin phosphotransferase . This fusion protein is sequestered intracellularly [28] . PLAP is cotranscribed but translated independently from an internal ribosome entry site . No wild-type transcripts are produced from this allele [28] . Brains from E16 . 5 ( n = 18 ) , P0 ( n = 45 ) , P4 ( n = 8 ) , P7 ( n = 6 ) , and P30 ( n = 26 ) were used in the study . To label thalamic and corticofugal fibers , single crystals of 1 , 1′-dioctadecyl-3 , 3 , 3′ , 3′-tetramethylindocarbocyanine perchlorate ( DiI ) and 4- ( 4- ( dihexadecylamino ) styryl ) -N-methylpyridinium iodide ( DiA ) ( Molecular Probes ) were placed with a stainless steel electrode into the visual and somatosensory dorsal thalamic nuclei or the visual and somatosensory cerebral cortex of both hemispheres of each brain . After injections , brains were kept in 2% paraformaldehyde for between 3 wk and 2 mo at room temperature in the dark . Back-labeling of dLGN neurons and their axons in P7 animals was performed under hypothermia-induced anesthesia . A small incision was made in the scalp to reveal the skull , and a fine needle was used to pierce the skull above the primary visual cortex . A Hamilton syringe was used to inject 0 . 5 μl of a 10% solution of DiI in absolute ethanol , into the primary visual cortex . The scalp was bonded with tissue adhesive ( Dermabond ) and animals were allowed to survive for 24–48 h to allow for adequate retrograde labeling before being sacrificed . Dissected brains were postfixed for 24 h in 4% PFA at 4 °C . Brains were washed in PBS ( 0 . 1 M , pH 7 . 4 ) , embedded into 4% agarose ( Sigma ) , and cut at 100 μm with a Vibroslicer ( Leica , VT1000S ) . Sections were counterstained with 2 . 5 μg/ml of bis-benzimide ( Sigma ) or with 0 . 5 μg/ml DAPI ( 4′-6-diamidino-2-phenylindole ) , mounted in PBS/glycerol or AquaPolymount ( Polysciences ) onto slides , and analyzed using an epifluorescence microscope ( Leica , DMR , or Zeiss ) and a laser scanning confocal microscope ( Leica , DMRE ) . Mice were perfused with 4% paraformaldehyde or a mixture of 1% paraformaldehyde/1 . 5% glutaraldehyde ( for the cytochrome oxidase staining ) in PBS . Brains were removed , postfixed in the same fixative overnight at 4 °C and embedded in 4% agarose . Serial 50- or 100-μm sections were cut on a vibratome ( Leica; VT1000S ) and processed for PLAP staining as previously described [28] . Alternatively , following perfusion , brains were postfixed in the same fixative for 3 h and cryoprotected with 30% sucrose in PBS . Serial 40-μm sections were cut on a freezing microtome and processed for Nissl staining ( 0 . 5% cresyl violet solution; Sema6A+/−: P0 , n = 2; P30 , n = 6 , and Sema6A−/−: P0 , n = 2; P30 , n = 6 ) . For cytochrome oxidase staining , cortical hemispheres were dissected from adult mice ( Sema6A+/− , n = 9; Sema6A−/− , n = 9 ) , postfixed between glass slides and cryoprotected before sectioning and processing . For immunohistochemistry , dissected brains were postfixed in 4% paraformaldehyde for 24 h , washed in PBS , embedded in 4% agarose , and sectioned ( 40–60 μm ) on a vibratome . P 4 Sema6A+/+ ( n = 7 ) and Sema6A−/− ( n = 5 ) mouse brains were treated for immunofluorescence with rabbit antibody to cleaved caspase-3 ( 1:200; Cell Signaling Technologies ) . Similarly , immunofluorescence with mouse antibody to neurofilament ( 1:100; DHSB ) was detected on sections from E16 . 5 Sema6A+/+ ( n = 2 ) and Sema6A−/− ( n = 2 ) mouse brains . Tangential sections were cut from flattened cortical hemispheres of P7 Sema6A+/+ ( n = 5 ) and Sema6A knockout ( KO; n = 10 ) mouse brains and incubated with antibody to serotonin ( 1:50 , 000; Immunostar ) , which was then detected with biotinylated secondary antibodies using the Elite ABC kit ( Vector ) . Results were documented using a digital camera ( Leica DC500; Canon Powershot S40 ) or an epifluorescence microscope ( Zeiss ) and digital camera ( Olympus ) , and the images compiled with Adobe Photoshop 8 . 0 or Adobe Photoshop CS software . Green and red fluorescent latex microspheres ( Lumaflor ) were used to labeled axonal projection from somatosensory and visual cortical areas , respectively , to the corresponding thalamic nuclei in Sema6A+/+ ( n = 4 ) and Sema6A−/− ( n = 4 ) mice . Animals were anesthetized with 2 . 7 mg/kg Hypnovel ( Roche ) , Hypnorm ( Janssen ) , and distilled H2O mixture ( 1:1:2 volume ratio ) , which was delivered intraperitoneally , and placed in a stereotaxis frame . After the skin was disinfected and incised , a microdrill was used to perform a craniotomy . Glass micropipettes ( Clark Electromedical Instruments ) and a binocular stereo-microscope ( Zeiss ) were used to inject a single injection of 0 . 3–1 . 0 μl of CT or microspheres into S1 or V1 . Animals were allowed to survive for 24 to 48 h to permit adequate retrograde transport of the CT or microspheres to thalamic cell somata . In situ hybridization was performed on 50-μm vibratome sections of E14 . 5 Sema6A ( wild-type [WT]: n = 2 , Sema6A+/− [HT]: n = 4 and KO: n = 4 ) , P0 Sema6A ( HT: n = 9 , KO: n = 8 ) , and P7 Sema6A ( HT: n = 4 , KO: n = 4 ) mouse brains , as previously described [50] . The following digoxigenin-labeled RNA probes were used: Sema6a ( a gift from W . Snider ) ; EphA7 , EphrinA5 , Cadherin6 , and RZRβ ( kindly provided by J . Rubenstein , with permission from the original researchers ) ; Cadherin8 ( 241–1 , 481 of mouse Cad8; GenBank accession number X95600; obtained by reverse transcription [RT]-PCR ) . The number of cells in the dLGN and the VB back-labeled from the occipital cortex were manually counted in consecutive 100-μm sections of E16 . 5 Sema6A+/− ( n = 30 sections , 4 animals ) and Sema6A−/− ( n = 41 sections , 6 animals ) , P0 Sema6A+/− ( n = 11 sections , 2 animals ) and Sema6A−/− ( n = 10 sections , 2 animals ) , and P4 wild-type ( n = 9 sections , 2 animals ) and Sema6A−/− ( n = 12 sections , 2 animals ) brains . The numbers of cells back-labeled to either the dLGN or VB for animals of a given age were compared using the Wilcoxon two-sample test and found to be significant at 99 . 9% confidence limits at E16 . 5 and P0 . As the absolute number of cells back-labeled is dependent on the size of dye crystal used , we also analyzed the proportion of labeled cells in either the dLGN or VB of a given section . The proportional values were Arcsine transformed for statistical analysis by Wilcoxon two-sample tests . The area of the dLGN and vLGN thalamic nuclei was measured in 40-μm cytochrome oxidase serial sections from Sema6A+/− ( n = 5 ) and Sema6A−/− ( n = 4 ) brains using SigmaScan Pro software ( SigmaScan ) . The volume of the dLGN and vLGN was calculated using the Cavalieri method . The relative area of V1 was measured in tangential sections of P7 of Sema6A+/− ( n = 5 ) and Sema6A−/− ( n = 10 ) brains , stained for serotonin immunohistochemistry , using Cell A software ( Soft Image System ) . | During brain development , the emergence of distinct areas in the cerebral cortex involves an interplay between patterning of the cortical sheet in the early embryo and later influences of incoming connections made from other brain areas , namely the thalamus . Connectivity between the thalamus and the cortex is initially smooth and graded , and a prominent model for how thalamocortical connectivity is established proposes thalamic axons are topographically sorted as they course through subcortical regions and then passively delivered to appropriate areas of the cortical sheet . We have used mutant mice lacking the guidance molecule Semaphorin-6A to test this model . In these mutants , Semaphorin-6A axons from the visual part of the thalamus are subcortically misrouted and fail to innervate the presumptive visual cortex , which is instead invaded by somatosensory thalamic axons . Despite this major disruption in initial connectivity , many visual thalamic axons find their way specifically to visual cortex , arriving several days later than usual . These late-arriving axons often follow alternate routes , and upon arrival are able to out-compete earlier-arriving somatosensory axons to reestablish grossly normal thalamocortical connectivity . These results argue strongly against an essential role for early subcortical targeting in the establishment of thalamocortical connectivity patterns and suggest instead the existence of highly specific target-selection mechanisms that match thalamic axons with appropriate cortical areas . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience",
"developmental",
"biology"
] | 2009 | Specificity and Plasticity of Thalamocortical Connections in Sema6A Mutant Mice |
T-DNA activation-tagging technology is widely used to study rice gene functions . When T-DNA inserts into genome , the flanking gene expression may be altered using CaMV 35S enhancer , but the affected genes still need to be validated by biological experiment . We have developed the EAT-Rice platform to predict the flanking gene expression of T-DNA insertion site in rice mutants . The three kinds of DNA sequences including UPS1K , DISTANCE , and MIDDLE were retrieved to encode and build a forecast model of two-layer machine learning . In the first-layer models , the features nucleotide context ( N-gram ) , cis-regulatory elements ( Motif ) , nucleotide physicochemical properties ( NPC ) , and CG-island ( CGI ) were used to build SVM models by analysing the concealed information embedded within the three kinds of sequences . Logistic regression was used to estimate the probability of gene activation which as feature-encoding weighting within first-layer model . In the second-layer models , the NaiveBayesUpdateable algorithm was used to integrate these first layer-models , and the system performance was 88 . 33% on 5-fold cross-validation , and 79 . 17% on independent-testing finally . In the three kinds of sequences , the model constructed by Middle had the best contribution to the system for identifying the activated genes . The EAT-Rice system provided better performance and gene expression prediction at further distances when compared to the TRIM database . An online server based on EAT-rice is available at http://predictor . nchu . edu . tw/EAT-Rice .
Rice is a major staple in the diet for more than half of the world’s human population . With the rapidly increasing pressures of both human population growth and global climate change , optimizing rice yields is critical over the next several decades . Sequencing of the rice genome , the smallest genome among the major cereal crops , was completed in 2005 [1] and from this work , rice emerged as the major monocot model plant for functional genome study and breeding improvement within cereal crops . Global crop production , especially including maize , rice , wheat and soybean yields must double by 2050 to sustain the rapid growth of the World’s population [2]; therefore , rice scientists focus on intensive improvement of rice quality and yield as a primary goal , through the investigation of rice phenomics and genomics of which approximately 36500 genes have been annotated for application to functional genomics and modern breeding [3] . The International Rice Functional Genomics Project ( IRFGP ) has proposed an international coordinated project , RICE2020 , to determine the biological function of every gene in the rice genome by 2020 [4] . Multiple methods for large-scale analysis of the biological function of genes by forward or reverse genetic approaches have been rapidly established , including bacterial artificial chromosome ( BAC ) libraries , large-scale expressed sequence tags ( ESTs ) , full-length cDNA collections , a transcriptome database , transfer DNA ( T-DNA ) or transposon-tagged rice mutant populations , and genome-wide association study ( GWAS ) [5–15] . T-DNA insertional mutagenesis distributes uniformly throughout the rice genome , but preferentially in gene-rich regions , which results in knockout/loss-of-function for the inserted gene . Hence , this method may generate two questions that lead to fewer desirable plant traits: 1 ) Plant death occurs because the function of an essential gene is absent; 2 ) A disrupted gene can functionally complement via its gene family . To solve this problem , multiple tandem copies of cauliflower mosaic virus ( CaMV ) 35S enhancers [16] were introduced into a T-DNA vector for activation/gain-of-function tagging; genes within a 40–60 kb flanking region of the T-DNA-inserted locus are probably activated . Adding four 35S enhancer sequences in series to a T-DNA construction can enhance gene expression [16–21] . Development of large T-DNA mutant populations provides a powerful genetic resource for both forward and reverse genetics studies on gene function [5–8 , 13 , 14 , 22] . The Taiwan Rice Insertional Mutant ( TRIM ) database was generated from Tainung 67 ( TNG 67 ) and contains about 93 , 000 mutant lines; 85% and 65% of TRIM mutants have phenotyping and flanking sequence data , respectively [23] , which significantly accelerates the ability to elucidate rice gene function . Three hundred genes of the flanking region of TRIM mutants were examined; 58% of these genes were activated by T-DNA insertion at differential levels [24] and demonstrating the activation of multiple activated genes became a laborious and time-consuming process . Bioinformatics has developed rapidly [25 , 26] and many biological prediction tools have been built by machine learning approaches [27–31] . Therefore , we developed a machine learning based tool for predicting the flanking gene expression around the T-DNA insertion site to assist researchers in improving the screening efficiency of activated genes . We collected the validated genes by RT-PCR and clustered them into activated and non-detectable groups . DNA sequences including UPS1K ( a 1 kb upstream sequence from the start codon ) , DISTANCE ( from the start codon of a target gene to enhancer ) and MIDDLE ( a 150 bp up- and downstream sequence around the central nucleotide of the DISTANCE region ) were retrieved to encode and build a two-layer machine learning prediction model . The features , containing N-gram , Motif , nucleotide physicochemical properties ( NPC ) , and CG-island ( CGI ) , were referenced to construct the first-layer models by support vector machines ( SVM ) [32] . Meanwhile , the logistic regression scoring , that take into account of the distance from target gene to T-DNA located site was used to weight the feature-encoding . In the second layer , because biological phenomena are caused by multiple factors , we analyzed different combinations of the four features noted above . In the second-layer models , the NaiveBayesUpdateable algorithm selected from 69 classified methods of the Waikato environment for knowledge analysis ( WEKA ) to integrate the first-layer models [33] . Our prediction platform , EAT-Rice , based on the TIGR MSU v7 . 0 genome , can predict genes within a specific range on both sides of the T-DNA insertion site and can provide a prediction outcome , confidence score , and the distance between T-DNA insertion site and target gene .
For T-DNA activation-tagging , individual insertion events were confirmed by southern blot . Plasma rescue was used to find the T-DNA insertion site , then RT-PCR to detect the expression of genes around the T-DNA insertion site activated by enhancer . Two experimental datasets were collected: the first dataset included 226 T-DNA mutants containing 293 verified genes and the second dataset included 11 mutants containing 65 verified genes . Gene expression was divided into three types: activated gene ( defined Ac ) , gene with no significant effect ( defined NE ) , and non-detectable gene ( defined ND ) ( Table 1 ) . The first dataset of gene annotations were based on The Institute for Genomic Research Rice Genome Annotation project ( TIGR ) [34] , and the second dataset was based on Rice Genome Automated Annotation System ( RiceGAAS ) [35] . Both of them in genome sequence were referenced from Oryza sativa japonica cv . Nipponbare . Each data in the dataset represents the target gene which was validated within its T-DNA mutant line; in other words , the same target gene in different mutant line was defined as the different data . Moreover , each data contained name of the mutant line , T-DNA insertion site , accession number , and the states of gene expression . Data for 30 non-detectable genes were collected but in order to ensure the quality and stability of our prediction system , these genes were removed . The no significant effect gene was defined as a non-activated gene ( named NAc ) . The first dataset contained 280 genes , defined as the training set; the second dataset contained 48 genes , defined as the independent-testing set ( Table 1 ) . Two datasets come from different research units , which means that this data was made by different experimental process . We expect that the predictive model should have compatibility and practicality for the data from different research units; therefore , we applied TDNA-DS1 as training data and TDNA-DS2 as testing data rather than mixed the two datasets together . Thus , the method could also be used to validate the model whether it works in the study or not . The ratio of positive data ( indicated Ac ) and negative data ( indicated NAc ) in training data may influence the efficiency of machine learning . First layer models of the training dataset with different proportions of positive and negative data were established . After evaluation , the optimal ratio of positive to negative data ( P/N ratio ) in 1:1 was obtained ( S1 Fig ) . To divide the positive data into two section , we used the sequence similarity grouping . One sequence was selected within the population of 190 positive data compared with others using Pair-BLAST; the average of 189 scores was defined as the sequence similarity score . The flowchart for each positive data was duplicated to ensure all data were assigned a similarity score . Scores were sorted and divided into two groups ( S2 Fig ) . To avoid losing data and optimal P/N ratio , 180 positive data was divided into two groups and merged 90 positive data in each group with the same negative data into training set of 180 data points named as training subset 1 and training subset 2 , respectively . Taiwan Rice Insertional Mutant Database ( TRIM , http://rice . sinica . edu . tw/fgb2/gbrowse/ TRIM_gb ) which were built by Taiwan Academia Sinica can accelerate the rice functional research . The projects of TRIM are establishment of the mutant population , generation of genome-wide gene knockout by T-DNA , flanking sequence analysis , seed collection and phenotype characterization , seed conservation and PCR screening , inserted site in rice genome as well as the inserted orientation on the template are included . All above are to establish a database of the insertional mutant population . Biologists can survey whether the T-DNA mutants were inserted around the target gene which they are interested in because it might be suitable for gene functional study . In this study , the T-DNA mutant lines are acquired from TRIM database , the expression levels of flanking genes were further identified . Our purpose is to effectively predict the effect of T-DNA insertions on flanking genes by the EAT-Rice , which will accelerate the research of Rice gene function by TRIM mutants . To analyze the difference in DNA sequences between activated ( indicated Ac ) and inactivated ( NAc ) genes , the three-part nucleotide sequence of the gene was retrieved , including UPS1K , DISTANCE , and MIDDLE . The three kinds of sequences retrieving followed the three hypotheses , which were supported in previous studies [36 , 37] . 1 ) Based on promoter-enhancer interaction , first part of DNA fragment was one kb of upstream sequence from the start codon , also core promoter region , named as UPS1K; 2 ) In addition , based on scanning model , second part of DNA fragment was from the start codon of target gene to enhancer named as DISTANCE; 3 ) At last , based on lopping model , third part of DNA fragment was from 150 bp of up- and downstream sequence around the central nucleotide of DISTANCE region , and total length is 301 bp named as MIDDLE ( Fig 1 ) . T-DNA insertion site at upstream of target gene is an example shown as Fig 1 . In fact , T-DNA may be inserted downstream of the target gene or intragenic . Therefore , the sequence length of the DISTANCE and MIDDLE will be changed depending on the T-DNA insertion site . To determine the association of gene activation by analyzing whether CpG-island is present in promoter [43 , 44] . The EMBOSS Newcpgreport tool from The European Bioinformatics Institute ( EMBL-EBI ) was used to predict CpG islands , and encoded by number , length , distance , CG ratio , and OE value ( http://www . ebi . ac . uk/Tools/seqstats/emboss_newcpgreport/ ) . CGI number was the number using Newcpgreport to predict CpG islands on the promoter of target gene ( Eq 9 ) . CGI length was value of the length of CG-island divided by the length of promoter ( Eq 10 ) . CGI distance was distance from CG-island to TLS of gene ( Eq 11 ) . The CG ratio of CGI was ratio of CpG dinucleotides in CG-island ( Eq 12 ) . The observed/expected ( OE ) value of CGI was ratio of number of CpG dinucleotides observed in CG-island to the expected number of CpG dinucleotides . Its formula was number of CpG dinucleotides on the promoter divided by number of cytosine nucleotide multiply number of guanine nucleotide on CpG-island ( Eq 13 ) . To reduce model complexity and shorten calculation time , we analyze the frequency of pattern occurrences of 5440 nucleotide groups of N-gram and 2087 regulatory cis-elements of Motif in the sequence , including UPS1K , DISTANCE , and MIDDLE , between Ac and NAc genes . DNA fragments with a P-value of < 0 . 05 by T-test ( implemented by R ) were selected to identify the patterns with different frequencies in the Ac and NAc sequences . For the N-gram , the UPS1K , DISTANCE , and MIDDLE , 359 , 4085 , and 349 patterns were filtered out with P-value < 0 . 05 . In the Motif , 106 patterns were identified . The selected patterns above were encoded further depending on what the N-gram or motif it derived ( An example was shown in S3 Fig ) . In the research , a formula was designed to evaluate the prediction performance of the two second-layer models from training subset 1 and subset 2 . We considered AUC , Sn , and Sp as our evaluating indicator in model , and the formula includes the value of cross-validation multiplied by the value of exchange-testing , divided by the value of self-consistency . Note , the formula indicates the lower the evaluating scores , the higher the extent of model overfitting , and vice versa . In this study , we built the prediction system about the flanking gene expression of T-DNA insertion site in rice mutants by two layers model of machine learning . A 280 training set was selected to train a model of logistic regression based on the relationship between distance from the 35S enhancer to the target gene and gene expression . LIBSVM was used to build the first layer model that adopted three kinds of DNA sequences and four kinds of features for encoding . For UPS1K , four features , i . e . , N-gram , Motif , NPC , and CGI , for encoding , while for DISTANCE and MIDDLE only N-gram and NPC were used to encode , and eight prediction models were generated ( Table A in S1 Supplement ) . The optimal P/N ratio was calculated from the average results of eight models . For the second layer , we used a different combination to integrate first layer models encoded by four features , picked out the preferred model of predictive performance , and used WEKA v3 . 6 to analyze 69 kinds of classification algorithms . NaiveBayesUpdateable was adopted to build models ( Table B in S1 Supplement ) . The accuracy of the two-layer model was evaluated with 48 independently testing data ( Fig 2 ) . A 5-fold cross-validation method and 48 verified genes were chosen as testing data to evaluate the predictive performance of the model; evaluation indictor were Accuracy ( Acc ) , Sensitivity ( Sn ) , Specificity ( Sp ) , F-score ( F1 ) , and AUC ( Area under the receiver operating characteristic curve ) . Acc can evaluate the prediction accuracy of positive and negative data; the closer to 100% , the more accurate the overall predictive performance of the model ( Eq 15 ) . Sn and Sp evaluate the accuracy of the prediction of positive and negative data , respectively ( Eqs 16–17 ) . F1 is the weighted average of Recall ( also called Sn ) and Precision ( the ratio of true positive data with true positive data plus false positive data ) of models ( Eq 18 ) . When the numbers of positive and negative data were different , Acc was not a good evaluation indicator , so we also considered AUC using an ROCR library of R language additionally . The Sn , Sp and AUC value are from 0 to 1 . The closer to 1 , the better learning of model .
When we assigned the UPS1K sequence of the gene in the T-DNA activation-tagged mutant , we discovered 55 repeat sequences of different expression states , which are the result of a single target gene affected by multiple independent T-DNA insertion events . The data of these repeats differ significantly in the distance from the T-DNA insertion site to the target gene . We grouped data based on the distance from the 35S enhancer to the TLS of the gene and calculated the ratio of gene activation in detached groups that separated by distance . Statistical analysis showed that the distance between enhancer and TLS of the gene negatively correlated with gene activation ( Fig 3A; Table C in S1 Supplement ) , implying that distance has the ability to influence the interaction between the enhancer and target gene . Previous studies have suggested that the enhancer-gene interaction was not affected by orientation , location ( i . e . , the enhancer is located on the upstream , downstream or intragenic locus ) and distance [17 , 45] . However , our analysis demonstrated that there is a statistically significant difference in distance ( P = 6 . 39e-07 ) ( Fig 3A; Table D in S1 Supplement ) . Gene orientation , T-DNA insertion orientation , and location were analyzed to assess the promoter-enhancer interaction and if the probability of gene activation was influenced by these three factors . No significant effect for the three factors on the enhancer-to-gene activation was observed ( Fig 3B–3D; Tables C and D in S1 Supplement ) . The repeat sequences of different expression states may cause contradictions in model building by machine learning . Therefore , we used logistic regression to establish a model based on the distance factor to predict the probability of gene activation . The value of the regression prediction were used as a feature-encoding weighting when the first layer modules were built to distinguish repeat sequences , and the logistic regression formula was as shown in Eq 19: π ( x ) =exp ( 1 . 448−7 . 099e−05x ) 1+exp ( 1 . 448−7 . 099e−05x ) , ( 19 ) where linear regression formula is 1 . 448–7 . 099e-05x; intercept ( fixed constant of linear regression ) is 1 . 448; independent variable parameter is -7 . 099e-05; and x indicates distance variable . π ( x ) indicates the logically transformed function of the linear regression and represents the possibility of gene activation . The evaluation results on the first layer feature model of training subset 1 showed that the models constructed by UPS1K and MIDDLE in the N-gram encoding and UPS1K in Motif encoding achieved the most desirable results ( Table 2 ) . In models of UPS1K and MIDDLE using N-gram encoding , the cross-validation was 90 . 00% and 95 . 00% on Acc , while the independent-testing result for the same models was 64 . 58% and 72 . 92% , respectively . In the Motif model using N-gram encoding , training was 82 . 22% on Acc , but for the independent-testing of the Motif model using N-gram encoding , Acc was only 50 . 00% , indicating that this model may suffer from overfitting . Model performance of training subset 2 was similar to subset 1 . The range of expected model performance with 5-fold cross-validation was 79 . 44% - 89 . 44% and independent-testing was 64 . 58% - 70 . 83% on Acc . Compared with subset 1 , the Motif model using N-gram encoding of subset 2 was >14 . 58% on Acc and was > 0 . 17 on AUC . In the N-gram encoding of subset 1 and subset 2 , the DISTANCE model was 53 . 89% and 61 . 67% for cross-validation on Acc , respectively . However , for subset 1 and subset 2 , the UPS1K model was approximately 36 . 11% and 19 . 44% greater than the DISTANCE model , and the MIDDLE model was also greater than 41 . 11% and 27 . 77% , respectively . In the NPC encoding , cross-validations of the models of UPS1K , DISTANCE , and MIDDLE in subset 1 and subset 2 averaged 55 . 18% and 54 . 63% , respectively; the average of independent-testing was 61 . 11% and 60 . 42% . In the CGI encoding , cross-validation and independent-testing were close to 50% on Acc , suggesting that CGI might not be a good classification feature . Taken together , N-gram and Motif classification performance was more valuable than NPC and CGI , indicating that some classification features have meaningful biological significance in this study . To consider the complexity of the biological mechanism , the second layer models combined four features by integration of machine learning , with an eye to improving system accuracy . In the cross-validation of subset 1 , we found that all evaluated parameters except the AUC demonstrated N-gram encoding provided a dominant contribution to classification ( Table 3 ) . The AUC value coincides with model performance; higher AUC value provide superior stability of the Ac and NAc gene classification in model performance . From these results , we selected the CGI+Motif+N-gram combination based on the highest AUC . The independent-testing results were Acc of 72 . 92% , AUC of 0 . 76 , F1 of 0 . 772 , Sn of 0 . 846 , and Sp of 0 . 591 . In the cross-validation of subset 2 , the performance of a single N-gram was similar to that of subset 1 , indicating that the contribution of N-gram in the second layer combination was more favorable . After considering the balance performance between AUC , Sn , and Sp , we selected two combinations of N-gram+NPC and CGI+N-gram+NPC . The results illustrated that both model performances were equivalent , implying that incorporation of the CGI feature did not improve accuracy . From this assessment , we selected the N-gram+NPC combination in subset 2 . The independent-testing results were Acc of 79 . 17% , AUC of 0 . 806 , F1 of 0 . 828 , Sn of 0 . 923 , and Sp of 0 . 636 . We determined the optimal models from training subset 1 and subset 2 by second layer model combination and then chose the final model by comparing the accuracy of the cross-validation on both models . However , we found for the subset 1 model that the cross-validation value for Acc was 94 . 45% and the independent-testing value was 72 . 92% on Acc . For the subset 1 model , differences of performance between cross-validation and independent-testing on Acc and AUC were 21 . 53% and 0 . 229 , respectively; for the subset 2 model , cross-validation of the subset 2 model for Acc was 88 . 33% lower than that of the subset 1 model , and differences of performance were 9 . 16% on Acc and 0 . 166 on AUC . From the above described , the subset 1 model has higher performance in learning , however , it worked not well in testing . In addition , the subset 1 model might have an overfitting phenomenon in the first layer because Motif encoding that could affect the performance of the second layer; overfitting of the subset 1 model ( or any model ) would engender poor prediction performance for data other than its own training data . To verify this issue , we used another training data from subset 2 as the testing data to evaluate the subset 1 model , and vice versa . In addition , we also used the training data from the building model as the testing data to evaluate the training quality of the model . Evaluation results indicated that self-consistency compared to cross-validation increased 0 . 55% for the Acc indicator in subset 1 . However , Acc increased by 1 . 67% in subset 2 , indicating that the training quality of the model credible . Subset 1 was 6 . 12% higher than subset 2 in cross-validation and 5 . 00% higher in self-consistency . In contrast , subset 2 was 2 . 23% higher than subset 1 in exchange-testing , indicating that subset 2 was not only fault tolerant but also accurate with respect to prediction ( Table E in S1 Supplement ) . So , we designed a formula ( see Eq 14 ) , it can calculate which training model who has the greater quality . Additionally , applying the formula , subset 2 was identified as the best-fit model for our system , because the score of the subset 2 model was higher than the score for the subset 1 model . We found a correlation between gene activation by the 35S enhancer and the distance from the 35S enhancer to the TLS of the gene , indicating that the distance factor has an important significance ( Fig 1A ) . We further analyzed the predictive performance of the EAT-Rice using different distance ranges and compared the predictive accuracy in training and independent-testing data . In addition , we compared the difference in predictive accuracy of EAT-Rice and TRIM platforms using different distance intervals . First , we grouped training data of subset 2 and 48 independent-testing data based on different distance ranges and analyzed the predictive performance of EAT-Rice ( Fig 4A; Table F in S1 Supplement ) . Among genes at >20 kb distances , Acc of training data showed an increasing trend , but independent-testing on Acc showed a decreasing trend . With the increase in the length of the DISTANCE sequence , the features generated by N-gram+NPC , the final model used for the EAT-Rice , were more consistent with sequence-specificity related to DISTANCE sequence , resulting in the observation of the overfitting phenomenon in EAT-Rice for gene over a 20 kb distance . On the other hand , the T-DNA mutant lines were obtained from TRIM , however , we found that the states of flanking gene activation we identified were different from the database . To compare the performance of EAT-Rice and TRIM , we collected and analyzed 100 activated genes not used in subset 2 from 190 positive data in the training dataset . Performances of TRIM and EAT-Rice were 39 . 00% and 94 . 00% on Acc , respectively , applying this analysis ( Data not shown ) . TRIM had reliable predictive accuracy when the gene distance was less than 10 kb , but less reliable predictive ability over 10 kb . For EAT-Rice , the performance gradually decreased , but predictive accuracy was eliminated for ranges beyond 30 kb ( Fig 4B; Table G in S1 Supplement ) , indicating that the reliable predictive range of TRIM was approximately 10 kb up- and downstream of the T-DNA insertion site , however , EAT-Rice could predict more accurately than TRIM at greater gene distances . Overall , EAT-Rice out-performed TRIM with respect to the predictive accuracy of gene activation but due to overfitting , the predictive ability of EAT-Rice was reduced at distances of more than 20 kb .
In previous studies , we thought the enhancer has no bearing on activated genes when the orientation , location , or distance is different [17 , 37 , 45 , 46] . However , our statistical results showed the distance factor may influence the probability of gene activation by the enhancer ( Fig 1A ) . We speculated three reasons might causing the difference . 1 ) Previous investigations discussed the activation of this target genes by an endogenous enhancer , but the exotic 35S enhancer could cause nonspecific gene constitutive expression in the research . 2 ) The activation of a single enhancer was the focus of prior work; in contrast , our research objective focused on different insertion sites of enhancer from many mutant lines . Comparing the intention of the past research with ours are very different in this issue . 3 ) Finally , mammalian systems have been the target in previous work whereas ours is plants; the mechanisms of enhancers would be expected to be distinctive . Previous work showed that the distance was a key factor to target gene influenced by the 35S enhancer on T-DNA activation-tagging [36] . The enhancer works only at a suitable distance and if the distance between the target gene and the enhancer is too far or too close , the enhancer-promoter interaction will be diminished [47 , 48] . A similar mechanism exists in transgenic plants , where the interaction strength depends on the intensity of the enhancer and the sensitivity of the target gene promoter , and thus determines whether the distance barrier can be overcome [49 , 50] . Although the sequence distance of suitable interaction for the 35S enhancer is unknown , prior work showed the impact of the 35S enhancer could be observed at a 78 kb distance [51] . In the first layer model , we captured three sequence fragments based on probable mechanisms of the enhancer and discovered that the rank of performance was MIDDLE > UPS1K > DISTANCE in the model built from N-gram ( Table 2 ) . We speculated the reason for DISTANCE sequence having the lowest accuracy was that it depended on the distance from the T-DNA insertion site to TLS of gene . The difference in the distance from insertion site to gene led to a varied sequence length ( 100 bp-30000 bp ) . The sequences may contain , for example , a gene coding region , promoter , or intergenic region . These sequences would produce excessive noise which would hinder classification , and led us to choose the important sequence by T-test . It is noteworthy that the accuracy of MIDDLE is better than that of UPS1K . In general , we reasoned the performance of transcription was improved mainly by the enhancer interaction with promoter . We expected that the UPS1K sequence offered a critical message to augment the efficiency of classification and thought the sequence of DISTANCE and MIDDLE would provide less value . However , the result was not as expected . To check whether the MIDDLE could offer a useful message , we obtained randomly 180 fragments that were 301 bps from the rice chromosome to replace the original MIDDLE sequence . At the same time , to avoid taking repeat sequences , like retrotransposon elements or centromere region , we used BLAST method to compare the sequences of Oryza Repeat Database v3 . 3 offered by TIGR to randomly obtain sequences [52] . Then , we built the model by N-gram encoding with these sequences . Using independent-testing , the performance of model decreased by 18 . 75% of Acc , and then AUC decreased by 0 . 246 ( Table H in S1 Supplement ) . The results illustrated that MIDDLE had a quite pronounced effect with regard to gene activation of 35S enhancer , suggesting the nearby relationship between the MIDDLE region sequence and enhancer . The result of the first layer revealed the accuracy rank of the four features was N-gram > Motif > NPC > CGI . However , the result of the second layer showed the accuracy of the N-gram+Motif combination was less than the accuracy of N-gram alone . Although the principles of N-gram and Motif are similar , both are searching for specific fragments on sequences to encode , there are several differences between N-gram and Motif . N-gram used a 3–6 bp fragment from the random combination of nucleotides to encode , so its fragment may have no known biological significance . Motif collected cis-regulatory sequence fragments that have known biological significance in the plant kingdom . We anticipated N-gram and Motif to complement each other to enhance classification performance , however , the results demonstrated N-gram was a marked improvement over other combination . Perhaps , N-gram considered all nucleotide combinations , while Motif only considered data that was already confirmed by experiment . In the plant kingdom , the regulatory elements already confirmed are finite , and N-gram may substitute for the Motif function . Since the ND gene cannot confirm whether it is affected by 35S enhancer , it is deleted ND gene when data processing in this work . Furthermore , we also participated the deleted 30 ND gene in training dataset , and following system structure to build the same model ( Table I in S1 Supplement ) . The result indicated that the gene sequences of ND phenotype might include certain biological features of activated gene and produced an incorrect classification in the model .
DNA sequence analysis and machine learning were used to build a two-layer model system . The system predicts the flanking gene expression activated by the 35S enhancer in rice mutant lines of T-DNA insertion activation-tagging . To avoid deviation caused by single machine learning , the two-layer model was implemented with LIBSVM algorithm in the first layer and NaiveBayesUpdateable algorithm in the second layer . The distance factor from the 35S enhancer to the translation start site of target gene is consider , so the possibility of target gene activation is estimated by logistic regression . Then , the feature weighting of the first layer model is based on the value of logistic regression . We retrieved three region sequences , including UPS1K , DISTANCE , and MIDDLE , and use these features including N-gram and NPC to encode . The accuracy of cross-validation is 88 . 33% , and the accuracy of independent-testing is 79 . 17% . When EAT-Rice is compared to TRIM , the accuracy of EAT-Rice is 55 . 00% greater than TRIM , and the confidence interval is in the range of 2–5 and 10–20 kb . We found a negative correlation between the distance on the genomic sequence and gene activation by the enhancer , for example , if the gene is closer to the enhancer , gene activation is more likely . For UPS1K , DISTANCE , and MIDDLE , the models constructed from MIDDLE and UPS1K contribute more to classified prediction , but the information offered from MIDDLE provided a greater contribution than UPS1K to the system for identifying activated gene in the model , suggesting the sequence context of MIDDLE may cause proteins to bind to the region and influence the interaction between the 35S enhancer and target gene . Finally , we have developed a system that predicts flanking gene expression activated by the CaMV 35S enhancer in T-DNA insertion activation-tagged rice mutants . We expect our system ( EAT-Rice ) can assist rice gene scientists in enhancing the efficiency of selecting candidate genes . | Among all the food crops , the rice is one of the staple foods in the human population , especially in Asia . However , the human population increases rapidly and the cultivated areas decrease in these decades . To solve the food crisis in the future , the rice researchers devote themselves to study on the gene function to increase the rice yield and stress tolerant ability . There are around 39000 annotated genes in rice , so scientists are hard to survey the gene functional because of the complexity and interactivity among the genes . Therefore , scientists put into a lot of manpower and funds into the field . The T-DNA ( Transfer DNA ) activation-tagging biotechnology has been wildly used on studies of rice gene function , however , it might influence the flanking genes expression when T-DNA inserted into the rice genome randomly . Thus , it will take lot of time for the researchers to validate the activation of genes by T-DNA enhancer . In these decades , as the increase of the biological data accumulation , the extraction of hidden information from this data is getting more and more important . To assist rice biologists in rapidly focusing the target gene affected by T-DNA . The application of machine learning methods in artificial intelligence ( AI ) and the establishment of prediction tool with biological data construction to correctly identify and classify target genes are of great significance in both theory and practice . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion",
"Conclusion"
] | [
"cereal",
"crops",
"rice",
"artificial",
"intelligence",
"experimental",
"organism",
"systems",
"genome",
"analysis",
"crops",
"sequence",
"motif",
"analysis",
"plants",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"computer",
"and",
"information",
... | 2019 | EAT-Rice: A predictive model for flanking gene expression of T-DNA insertion activation-tagged rice mutants by machine learning approaches |
In a cross sectional study , 19 French and 23 Colombian cases of confirmed active ocular toxoplasmosis ( OT ) were evaluated . The objective was to compare clinical , parasitological and immunological responses and relate them to the infecting strains . A complete ocular examination was performed in each patient . The infecting strain was characterized by genotyping when intraocular Toxoplasma DNA was detectable , as well as by peptide-specific serotyping for each patient . To characterize the immune response , we assessed Toxoplasma protein recognition patterns by intraocular antibodies and the intraocular profile of cytokines , chemokines and growth factors . Significant differences were found for size of active lesions , unilateral macular involvement , unilateral visual impairment , vitreous inflammation , synechiae , and vasculitis , with higher values observed throughout for Colombian patients . Multilocus PCR-DNA sequence genotyping was only successful in three Colombian patients revealing one type I and two atypical strains . The Colombian OT patients possessed heterogeneous atypical serotypes whereas the French were uniformly reactive to type II strain peptides . The protein patterns recognized by intraocular antibodies and the cytokine patterns were strikingly different between the two populations . Intraocular IFN-γ and IL-17 expression was lower , while higher levels of IL-13 and IL-6 were detected in aqueous humor of Colombian patients . Our results are consistent with the hypothesis that South American strains may cause more severe OT due to an inhibition of the protective effect of IFN-γ .
Infection with the protozoan parasite Toxoplasma gondii is a leading cause of visual impairment in numerous countries , being responsible for 30 to 50% of uveitis cases in immunocompetent individuals [1] . Ocular toxoplasmosis ( OT ) is a potential complication of both acquired and congenital toxoplasmosis [2] . The incidence of ocular toxoplasmosis has been estimated in Colombia ( Quindio region ) to be of three new episodes by 100 000 inhabitants by year [3] , while in British-born patients it has been estimated to be 0 . 4 cases per 100 , 000 population per year and the lifetime risk of disease to be 18 cases per 100 , 000 population [4] . In a Colombian study , 5 . 5% of the population in the province of Quindío exhibited retinochoroidal scars resulting from a postnatally acquired infection , with 20% of this group presenting reduced visual capacity . [3] , [5] . In a retrospective study on uveitis conducted in 693 Colombian patients , 417 of whom had a definitive diagnosis , toxoplasmosis was the most frequent cause with 276 cases ( 39 . 8% ) followed by idiopathic uveitis and toxocariasis [6] . Some differences between South American and European clinical case series were observed in terms of congenital transmission rates , probability of symptoms in congenital OT [7] , [8] , severity of ocular inflammation [9] and intraocular specific antibody levels [10] . However , no comparative clinical and biological studies have been performed yet in patients from both continents with laboratory-confirmed OT . The population structure of T . gondii in North America and Europe includes three highly prevalent clonal lineages , Types I ( haplogroup 1 , Clade A ) , II ( Haplogroup 2 , Clade D ) , and III ( haplogroup 3 , Clade , C ) which differ greatly in virulence in the mouse model . The vast majority of human and animal infections are caused by the relatively avirulent Type II strains . In contrast , heterogeneous atypical genotypes of T . gondii are associated with severe infections in humans in South America . They belong to various haplogroups: 4 , 5 , 8 10 and 15 , Clade F [11] , [12][13] . The high genetic diversity of Toxoplasma strains in the tropical zone of the Americas may partly explain why congenital toxoplasmosis is more symptomatic in South America than Europe , as was demonstrated in cohorts of congenitally infected children from different continents [8] , [14] , [15] . A comparative prospective cohort study of congenitally infected children in Brazil and Europe found that Brazilian children displayed eye lesions that were larger , more numerous , and more likely to affect the central part of the retina responsible for acute vision [7] . Anecdotal clinical cases were also reported in the literature , such as a severe atypical bilateral retinochoroiditis in a Brazilian patient , caused by a highly divergent , non-archetypal T . gondii strain [16] . Given the markedly different population structure of T . gondii in Europe and South America , it is relevant to study the implications of this diversity on human pathogenesis [17] . Therefore , we conducted a multicenter case series study in order to compare the different clinical and immunological characteristics between Colombian and French patients , collecting the same data and performing the same laboratory assays in patients with biologically confirmed OT . The findings were related to Toxoplasma strain genotyping and peptide-based strain serotyping in our patients .
We collected data from consecutive patients who consulted at the Departments of Ophthalmology at Strasbourg University Hospital ( France ) and Quindío University Health Center ( Armenia , Colombia ) between August 2008 and August 2010 . Both departments were tertiary-level centers able to perform anterior chamber paracentesis . For both patient populations , a complete ocular examination was conducted , including best-corrected Snellen visual acuity , slit-lamp biomicroscopy , tonometry , and indirect ophthalmoscopy . The clinical diagnosis of OT was based on criteria previously described by G . Holland [6] , [18] . Screened patients with clinically suspected OT and seropositive for anti-Toxoplasma immunoglobulin G ( IgG ) antibodies were subsequently submitted to biological investigations to assess the local presence of Toxoplasma DNA and/or the intraocular antibody synthesis [19] to confirm OT . Ethics Committee/Institutional Review Board ( IRB ) approval were obtained from Hôpitaux Universitaires de Strasbourg ( PHRC 2007/3964 ) and Quindio University ( ACT 14 , 2008/23-06 ) . Written informed consent was obtained from all subjects . We analyzed the clinical characteristics of 19 French and 23 Colombian patients with active uveitis and biologically confirmed OT . Patients who were immunocompromised , suffered from other ocular infections , or received local or systemic anti-Toxoplasma treatment for active uveitis , were excluded . An assessment of the inflammation level and anatomic classification of uveitis was carried out according to the criteria proposed by the International Uveitis Study Group ( IUSG ) [20] . The size of the retinochoroidal lesions was measured in disc-diameters ( dd ) . Paired samples of aqueous humor and serum were obtained from each subject at the time of clinical diagnosis for laboratory analysis . The Colombian samples were stored locally at −80°C and then shipped together on dry ice to Strasbourg for laboratory analysis . Aqueous humor samples ( 100–150 µL ) were collected through anterior chamber paracentesis and stored , along with serum samples , at −80°C until analysis . The diagnosis of OT was first confirmed by real-time PCR detection of Toxoplasma DNA [21] . Positive PCR results were quantified using a standard curve with serial 10-fold dilutions from a calibrated suspension of T . gondii RH-Strain DNA . For PCR negative patients , immunoblot ( IB ) was performed in order to detect intraocular synthesis of Toxoplasma-specific antibodies ( LDBIO Diagnosis , Lyon , France ) . If both PCR and IB were unconclusive , a modified Goldmann-Witmer test was used to prove intraocular specificantibody synthesis [22] . The Bio-Plex Human 27-Plex Cytokine Panel assay ( Bio-Rad , Marne-la-Coquette , France ) was used according to the manufacurer's recommendations to measure cytokine and chemokine levels in aqueous humor . The assay plate layout consisted in a standard series in duplicate ( 1 to 32 000 pg/mL ) , four blank wells and 20 µL duplicates of AqH samples , diluted to 50 µL with BioPlex Human serum diluent [23] . A set of Toxoplasma seropositive cataract patients were used as control , 9 Colombian and 10 French . Data were analyzed with Bio-Plex Manager TM software V1 . 1 . DNA extraction for genotyping analysis was performed directly on ocular fluid samples and indirectly on infected cell cultures for six reference strains . GT1 , PTG , and CTG strains were selected as reference Types I , II , and III strains , respectively . TgCtCo02 , TgCtCo05 , and TgCtCo07 strains were selected as reference Colombian strains [24] , [25] . T . gondii DNA samples were subjected to genotyping analysis with 15 microsatellite markers in a multiplex PCR assay , as described elsewhere [26] . Serotyping of Toxoplasma infections was performed using 5 polymorphic synthetic peptides derived from the T . gondii dense granule proteins ( GRA ) , GRA6 and GRA7 . This test detects the presence of strain specific antibodies raised against Type II or non-Type II GRA6/7 alleles in patients infected with Type II or non Type II ( NE-II ) parasites respectively , as previously described [14] , [27] . Briefly , the ELISA results presented are an optical density ( OD ) index obtained by dividing the OD value at 405 nm for each of the 5 serotyping peptides by the mean of the OD readings for the 2 control peptides . Threshold values are determined by averaging the normalized OD ratio from 100 seronegative French samples and adding 2 standard deviations , above which normalized values are considered positive . Obtained results are divided in four populations depending on their reactivity to the 5 peptides: I/III , ATYP , no reactivity ( NR ) , and II [28] . I/III , ATYP and NR are considered as NE-II [14] . Sera from pregnant women , tested Toxoplasma seropositive in our laboratories , were used to assess the Toxoplasma serotype in a larger population from each country , 45 serum samples from Colombia and 100 from France . Mann-Whitney test followed by Dunn's Multiple Comparison test was applied for comparison of clinical and laboratory characteristics for French and Colombian patients with confirmed active ocular toxoplasmosis ( P values<0 . 05 were considered statistically significant; Stata software , College Station ( Tx ) USA ) . Fisher's exact test was used to compare diagnostic performances of IB and PCR as well as the serotype prevalence . Wilcoxon matched-pairs signed rank test was performed to compare IB patterns . Mann-Whitney test was used to compare intraocular parasite loads ( P values<0 . 05 were considered statistically significant . Kruskal-Wallis test followed by Dunn's Multiple Comparison test were applied for comparison of cytokine and chemokine levels in aqueous humor between control and OT populations in both countries ( P values<0 . 05 were considered statistically significant ) ( GraphPad Prism , La Jolla , CA , USA ) .
The clinical findings for OT patients are summarized in Tables 1 and S1 . Statistically significant differences between groups were found for eight parameters , being higher in Colombian patients in all cases: i ) time between consultation and anterior chamber paracentesis ( p = 0 . 02 ) ; ii ) size of active lesions ( p = 0 . 04 ) ; iii ) unilateral macular involvement ( p = 0 . 001 ) ; iv ) unilateral visual impairment ( p = 0 . 04 ) ; v ) vitreous inflammation ( p = 0 . 00001 ) ; vi ) percentage of patients with synechiae ( p = 0 . 04 ) ; vii ) vasculitis ( p = 0 . 04 ) and viii ) bilateral involvement ( p = 0 . 04 ) . In addition , there was a trend towards higher values for the Colombian patients regarding the number of lesions , number of recurrences , and intraocular pressure ( IOP ) , although these differences were not statistically significant . We conducted a stratified analysis in order to exclude the influence of time before anterior chamber paracentesis as a possible cause of the differences in clinical findings . We compared early ( <20 days after symptom onset ) and late consultations ( >20 days after symptom onset ) . As shown in Table 2 and supplementary figure 1 , most significant clinical differences between the populations were also visible when comparing only the early-consultant groups . In Colombians , aqueous humor samples revealed the presence of T . gondii DNA in 11 out of 23 samples ( 47 . 8% ) . In French patients , T . gondii DNA could be detected in aqueous humor samples of 7 out of 19 patients ( 36 . 8% ) . This difference was not statistically significant . In contrast , parasite loads in aqueous humor were significantly higher in Colombian patients , 4 . 53 parasites ± 2 per 100 µL versus 0 . 35±0 . 13 parasites per 100 µL ( p = 0 . 0006 ) ( Figure 1 ) . Aqueous humor samples from all French patients and 14 Colombian patients had an insufficient amount of T . gondii DNA for genotyping analysis . Only 9 Colombian ocular fluid samples were submitted for multilocus PCR-DNA sequence genotyping analysis . Six had unsuccessful PCR amplification for all 15 tested markers due to low T . gondii DNA concentration . The genotype of one clinical sample ( case COL-#6 ) was closely related to a Type I strain , but harboring unique alleles at three MS loci , M102 , N83 and AA , using 15 amplified markers ( Table 3 ) . Of note , the genotype of a reference Colombian isolate ( TgCtCo07 ) collected from a cat in 2005 was also Type I-like , suggesting that Type I-like strains may not be uncommon in animals and humans in Colombia . The genotypes of the other two clinical samples ( cases COL-#26 and COL-#38 ) could not be fully determined , with only four and five successfully amplified markers , respectively . However , the results of the amplified markers showed that both genotypes were different from the Type II or III strains , which are common in North America and Europe . They present a majority of Type I alleles ( case COL-#26 ) , like TgCtCo07 but distinct at the N61 marker , and a combination of Type I , III , and atypical alleles ( case COL-#38 ) , like TgCtCo02 and TgCtCo05 , but again distinct at the N60 and N82 genetic markers . IB detected local antibody production in 19/23 Colombian ( 82 . 6% ) and 13/19 French ( 68 . 4% ) patients ( not significant ) . However , a significant difference was observed in number of bands and their recognition pattern of Toxoplasma proteins ( p<0 . 0001 ) ( Figure 2 ) . Specific proteins were recognized in 3 . 3% to 63 . 3% of Colombian patients and 3 . 8% to 53 . 8% of French patients . Colombian patients recognized most frequently a 62 kDa protein , observed in 63 . 3% of patients . In French patients , the most frequently detected protein was at 34 . 2 kDa , found in 53 . 8% of patients . As the amount of aqueous humor was insufficient for Toxoplasma strain typing using an ELISA peptide-based assay , we decided to serotype these patients using their sera . Ten OT patients from each center were assessed , all from the early consultation group . Among the Colombian patients , no Type II serotype was detected . We found 4 I/III , one atypical and 5 non reactive ( NR ) serotypes ( Table 4 ) . In contrast , all tested French OT patients showed Type II serotypes except one patient with an atypical serotype . These patterns were significantly different between the two groups ( p<0 . 0001 ) . The two cases COL#26 and COL#38 , found as suspected Type I and Type I/III by genotyping , were serotyped as NR and type I/III , respectively ( Table 4 ) . To test if certain T . gondii strains are associated with OT , we determined the overall distribution of serotypes in infected non-OT control populations from both countries . Among the 45 Colombian control patients , only 6 subjects ( 13 . 3% ) had a type II whereas 39 ( 86 . 6% ) had NE-II serotypes , which were subdivided in 6 NR , 29 type I/III and 4 atypical serotypes . Of 100 French control patients , we found 64 ( 64% ) type II , and 36 ( 36% ) with NE-II; 10 NR , 2 type I/III and 24 atypical serotypes . No statistically significant differences were observed between the control and OT groups in Colombian patients , however we found a significant difference ( P = 0 . 02 ) between the French control and OT populations , with respect to the proportion of the two types , II and NE-II . Cytokines patterns in aqueous humor of OT patients were compared to cataract controls ( Figure 3 and Table S2 in Text S1 ) . Several immune mediators were augmented in French , as well as in Colombian patients . In French patients , the Th1 type cytokines IFN-γ , IL-2 and IL-15 were expressed in all patients . This Th1 immune response was associated to a Th17 response with increased IL-17 production . Additionally , we observed a large proinflammatory response with increased levels of IL-6 , IL-1β , IL-8 , MIP-1β , MCP-1 and G-CSF . These patients also possessed a corresponding anti-inflammatory response was based on the presence of IL-4 , IL-10 , and IL-1RA . In contrast , Colombian patients had lower expression of major proinflammatory immune modulators , including IFN-γ , IL-15 , IL-17 , IL-2 , IL-10 , MIP-1β , GM-CSF and G-CSF , with the exception of elevated TNF-α and IL-6 levels . These patients also had elevated levels of the counterregulating Th2-type cytokine IL-13 .
Previously published studies found differences between South American and European clinical case series on adult patients in terms of frequency of serological markers in OT [8] , probability of symptoms in congenital infection [7] , as well as inflammation levels and IOP [9] . However , these were mostly retrospective evaluations of multiple studies . Their main limitation is their inclusion of patients with “suspected” OT , rather than biologically confirmed cases . While the ocular signs of toxoplasmic retinochoroiditis are highly suggestive of this disease , they may be mimicked by other infections [22] , while in some cases , the symptoms may be atypical [19] , [29] . Therefore , we strengthened our evaluation by inclusion of biologically confirmed OT cases only , as well as by comparing the same bio-clinical data from two different populations of OT patients , located in South America and Europe in a cross sectional study . Among the 17 criteria analyzed in the two populations , the following were significantly higher in Colombian patients: macular involvement , vitreous inflammation , strabismus , bilateral involvement and synechiae . Our findings confirm and expand the data from the retrospective study of Dodds et al . from patients with biologically unconfirmed OT which found elevated IOP , increased presence of synechiae , AC cells , flare , and vitreous humor haze [9] . In our study , one key difference between the two patient populations was the date of consultation , as Colombian patients consulted later than the French . However , when our analysis was stratified regarding this aspect , the observed clinical differences remained significant . The main hypothesis for these clinical differences is based on the idea that severe disease in humans may result from poor host adaptation to neotropical zoonotic strains of T . gondii [11] . Our study accumulated some clues supporting this hypothesis . Central strain-specific parasite virulence factors in human infections were revealed in the last years [30] . Their role in the presence of more virulent parasite genotypes in South America [11] , [12] is not yet thoroughly studied . Theses strains are rarely found in Europe [31] where Type II genotypes predominate , including in OT patients [32] . In the three Colombian OT patients where we could detect Toxoplasma DNA , we found one Type I and two atypical strains . The fact that no patient of the French group had a sufficient ocular parasite load for genotyping clearly shows the difference in ocular virulence . Additionally , we noticed that intraocular antibodies responses showed major differences in Toxoplasma antigen recognition by an immunoblotting assay . Even if this could be partly due to better detection of Toxoplasma Type I antigens used in this assay by Colombian patients , other , host immune specific factors are certainly crucial such as local antibodies , whose exact role and function should be explored . Our serotyping assay confirmed that Colombian and French patients recognize different strain-specific epitopes . Colombian OT patients recognized a heterogeneous pattern of strain specific peptides , but none of them were from type II strains . The French OT patients recognized only Type II strain specific peptides , confirming the reliability of this test in a geographic region with predominant type II strains infections [33] . The corresponding control populations presented the same serological pattern for Colombia , but a slightly different pattern for France , where some sera were non reactive to Type II antigens . The difference may due to the unequal sample sizes , so this point needs further investigation using more samples and equilibrated OT and control population . However , these data indicate that type II and non-type II strains are able to cause ocular pathology , but with a markedly different clinical picture . Concerning the Colombian strains , current serotyping techniques might be not sensitive enough to distinguish the highly variable strains . When we looked at the patients' local immunological reaction , we observed clearly different cytokine signatures . In French patients , the host-parasite relationship seems to be equilibrated between protection and inflammation . The protective effect of IFN-γ is balanced by anti-inflammatory cytokines such as IL-2 and IL-10 . The role of IL-17 is controversial . We have previously observed an early pathologic and parasite promoting role for IL-17 in French patients and in an animal model infected by a Type II Toxoplasma strain [34] . In the intraocular ocular environment , IL-17 would attract neutrophils [35] and , accompanied by IL-15 and MIP-1β/CCL4 , activates and attracts NK cells [36] and monocytes [37] . All these innate immune cells might cause retinal inflammation , but then permit to control Toxoplasma proliferation [38] , [39] . As our recent findings implicate IL-27 and the Treg subset in counterbalancing deleterious inflammatory Th17 type responses [34] , the corresponding mediators deserve to be examined more closely in future studies . In contrast , in the clinically more severe Colombian cases , IFN-γ and other major immunomodulators such as IL-17 were barely detectable , while IL-6 and IL-13 were enhanced . Virulent strains encode virulence factors able to modulate multiple immune host cell signaling pathways through polymorphic effectors secreted into the host cells such as ROP16 and GRA15 [38] , [40] . The presence of Toxoplasma effector proteins from virulent strains could explain the down-regulation of ocular IFN-γ , leading to higher ocular parasite loads in Colombian patients . The IL-17 down-regulation remains to be explained , but decreased levels of IL-17 and other immune modulators , including proangiogenic factors , could lead to a defect in the migration of leukocytes to the eyes and be another explanation for impaired control of parasites in the context of virulent South American infections . IL-6 could also antagonize the anti-microbial properties of IFN-γ by sustained activation of STAT3 , a potent inhibitor of IL-12 and IFN-γ [41] . Down-regulation of IFN-γ and its anti-Toxoplasma activity was also observed for IL-13 in human fibroblasts [42] . It is important to note here that Type I strains express a ROP16 allele associated with prolonged activation of STAT3 and STAT6 signaling , which may in part contribute to the increased IL-13 levels , whereas Type II strains activate this pathway only transiently , allowing the establishment of an inflammatory reaction [43] . This may constitute the fundamental basis for the differential cytokine response observed in our study . The theory of local T cell exhaustion may be also of interest in the settings of Colombian patients . Immune exhaustion is characterized by the modification of the CD8+ functions by reducing their polyfunctionality and their efficacy [44] . Indeed , high Toxoplasma loads associated with a decreased IFN-γ and IL-15 production and enhancement of TNF-α could be one aspect of this loss of CD8+ T cell polyfunctionality . In contrast , in French patients , elevated IL-15 is critical for homeostasis of memory CD8 T cells , and may lead to a better control of parasite proliferation and subsequent parasite latency in the retina . Taken together , our results indicate that virulent strains observed in South America may suppress host-protective pathways , opening the way to multiplication and cytolytic activity of the parasite in retinal tissues including blood vessels . The presence of TNF-α in most of these patients could also contribute by enhancing an ongoing immunopathological retinal process [45] . In contrast , in French patients , the cytokinic environment may lead to the encystation of the parasite in the retinal tissues , leading to subsequent recurrences . Of course , for ethical reasons , we were only able to take one time-point . Our results represent thus a snapshot of a developing immune response . Additionally , a multifactorial origin of the observed clinical and biological differences could not be excluded . In our study , the source of contamination may have been drinking water collected from surface water sources ( i . e . , rivers , lakes ) [46] , [47] , [48] , [49] . The more common macular involvement in Colombian patients is often associated with congenital toxoplasmosis [6] , [15] , [50] , [51] . Even if we studied adult populations , we cannot exclude a congenital origin of infection in some Colombian patients . Moreover , acute toxoplasmosis was only diagnosed in 2 Colombian and 1 French case . The remaining population was considered to exhibit chronic toxoplasmosis . Finally , individual susceptibility was previously related to variations in various genes encoding immune response players , such as IFN-γ , IL-1α , IL-10 , TLR-9 or ABCA4 , COL2A1 , and P2X7-R [52] , [53] , [54] , [55] . These genetically susceptible patients are possibly less able to cope with a more virulent strain . Further investigations with larger cohorts including an evaluation of their immunological response and their individual susceptibility to Toxoplasma are needed to address these topics . | Ocular toxoplasmosis ( OT ) , due to protozoan parasite Toxoplasma gondii , is a potential complication of both acquired and congenital infection , leading to visual impairment in numerous countries and being responsible for 30 to 50% of uveitis cases in immunocompetent individuals . In this study we confirmed the presence of more severe ocular toxoplasmosis in a tropical setting of Colombia , when compared to France . The main hypothesis for these clinical differences is based on the idea that severe disease in humans may result from poor host adaptation to neotropical zoonotic strains of T . gondii Indeed , our results are consistent with the hypothesis that South American strains may cause more severe OT due to an inhibition of the intraocular protective immune response . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | Severe South American Ocular Toxoplasmosis Is Associated with Decreased Ifn-γ/Il-17a and Increased Il-6/Il-13 Intraocular Levels |
Cohesin subunit SMC1β is specific and essential for meiosis . Previous studies showed functions of SMC1β in determining the axis-loop structure of synaptonemal complexes ( SCs ) , in providing sister chromatid cohesion ( SCC ) in metaphase I and thereafter , in protecting telomere structure , and in synapsis . However , several central questions remained unanswered and concern roles of SMC1β in SCC and synapsis and processes related to these two processes . Here we show that SMC1β substantially supports prophase I SCC at centromeres but not along chromosome arms . Arm cohesion and some of centromeric cohesion in prophase I are provided by non-phosphorylated SMC1α . Besides supporting synapsis of autosomes , SMC1β is also required for synapsis and silencing of sex chromosomes . In absence of SMC1β , the silencing factor γH2AX remains associated with asynapsed autosomes and fails to localize to sex chromosomes . Microarray expression studies revealed up-regulated sex chromosome genes and many down-regulated autosomal genes . SMC1β is further required for non-homologous chromosome associations observed in absence of SPO11 and thus of programmed double-strand breaks . These breaks are properly generated in Smc1β−/− spermatocytes , but their repair is delayed on asynapsed chromosomes . SMC1α alone cannot support non-homologous associations . Together with previous knowledge , three main functions of SMC1β have emerged , which have multiple consequences for spermatocyte biology: generation of the loop-axis architecture of SCs , homologous and non-homologous synapsis , and SCC starting in early prophase I .
Meiosis requires unique chromosome structures and dynamics , which are most prominent in the first of the two meiotic divisions ( for reviews on aspects of meiosis relevant to this study see [1] , [2] , [3] , [4] , [5] , [6] , [7] , [8] , [9] . During premeiotic replication , two pairs of sister chromatids are formed from the two homologous chromosomes . Within each pair , the two sister chromatids are linked through sister chromatid cohesion ( SCC ) . In early meiotic prophase I the two pairs of sister chromatids form axial elements ( AEs ) through association with proteins like SYCP2 and SYCP3 . The AEs start to pair and synapse , and full synapsis is reached in pachynema . The synaptonemal complex ( SC ) is generated , which in addition to AEs includes transverse elements made of SYCP1 , SYCE1 and other proteins . Consequently , SYCP1 serves as a marker for synapsis . Other proteins such as HORMAD1 are displaced from chromosomes upon synapsis and thus their association with chromosomes indicates asynapsed chromosomes or chromosomal regions [10] , [11] . We use the terms “asynapsed” for never synapsed , and the term “desynapsed” for lost synapsis . Homologous pairing requires programmed DNA double-strand breaks ( DSBs ) , generated by the type II topoisomerase-like enzyme SPO11 . SPO11-deficient mice of both sexes are infertile and Spo11−/− spermatocytes die in late zygonema/early pachynema . In Spo11−/− meiocytes , these DSBs are not introduced into DNA , and homologous synapsis of the AEs , which still form , is defective . Non-homologous associations between multiple pairs of sister chromatids are observed [12] , [13] . DSBs may give rise to cross-overs and chiasmata . The processing of DSBs requires recombinases RAD51 and the meiosis-specific DMC1 . Characteristic RAD51 and DMC1 foci are formed during leptonema and early zygonema at sites of DSBs and disappear progressively during zygonema and early pachynema as repair proceeds . In spermatocytes , the X and Y sex chromosomes behave uniquely , for they are mostly non-homologous and only synapse at a short , about 700 kbp homologous region called the pseudoautosomal region ( PAR ) [14] , [15] . Within the PAR , meiotic recombination and chiasma formation take place [16] , although repair foci appear later on PAR than on autosomes [17] . Compared to synapsed autosomes , a higher density of repair foci and chiasmata is observed at the PAR , probably to ensure chiasma formation in this small region [17] . Asynapsis along most of the sex chromosomes in males has important consequences ( reviewed in [9] , [18] , [19] ) . The sex chromosomes are packaged into a heterochromatic subnuclear structure called the “sex body” [20] , which is enriched for proteins such as HORMAD1 and the ATM/ATR-phosphorylated H2AX , called γH2AX . Gene expression of the asynapsed sex chromosomes is silenced , a phenomenon called meiotic sex chromosome inactivation ( MSCI ) , and H2AX is one of the silencing factors [21] . Gene expression on asynapsed autosomes in leptonema or early zygonema is low [22] , and in mutants that fail to completely synapse their autosomes in pachynema , these asynapsed chromosomes or chromosomal regions are silenced as well . This process , named meiotic silencing of asynapsed ( or “unsynapsed” ) chromosomes ( MSUC ) [9] , [23] , [24] , [25] , depends on proteins like γH2AX . When autosomes are unsynapsed , they accumulate silencing factors at the expense of the sex chromosomes , leading to derepression of X- and Y-linked genes such as Zfy1/2 , which causes spermatocyte death [26] . In wild-type , upon completion of synapsis in late zygonema/early pachynema , autosomal gene expression is greatly up-regulated [22] . Cohesin is essential for SCC in mitosis and meiosis ( for recent reviews see [27] , [28] , [29] , [30] , [31] , [32] . The cohesin complex consists of a V-shaped heterodimer of two Structural Maintenance of Chromosome proteins , SMC1 and SMC3 , whose open ends can be closed by a kleisin protein . A single kleisin , RAD21 ( SCC1 , MCD1 ) , is expressed in somatic cells , and two meiosis-specific kleisins , REC8 and RAD21L , are additionally expressed in meiocytes , increasing the possible number of cohesin complex variants ( reviewed in [33] , [34] ) . Other proteins associate with cohesin , most notably one of three variants of stromal antigen proteins ( SA1 to SA3 ) , of which SA3 ( STAG3 ) is meiosis-specific . Earlier , we reported the identification of a meiosis-specific SMC1 cohesin named SMC1® [35] . SMC1® is detected in spermatocytes starting in leptonema , accumulates along the AEs and SCs in prophase I , and remains at spermatocyte centromeres until the metaphase II-anaphase II transition . SMC1β-deficient meiocytes feature profoundly shortened AEs with extended chromatin loops , telomere aberrations , and partial asynapsis [36] . Smc1β−/− spermatocytes die in early/mid-pachynema ( stage IV of the seminiferous epithelial cycle ) . In young mice many Smc1β−/− oocytes survive until metaphase II and show partial loss of SCC or chiasmata in metaphase I . This loss increases with age [37] . At metaphase II no SCC exists anymore in Smc1β−/− oocytes , which then die . Overall , SMC1β is more abundant in meiocytes than the canonical SMC1α . SMC1α is present on meiotic chromosomes in prophase I , but disappears afterwards . It is unknown whether SMC1α- and SMC1β-type cohesin complexes , or only one of them , provide SCC in prophase I . Cohesin executes additional functions besides in SCC . In somatic cells cohesin is required for efficient DSB repair through homologous recombination . Cohesin is recruited to sites of DNA damage , and its SMC proteins become phosphorylated in response to DNA damage ( reviewed in [38] , [39] , [40] , [41] . Cohesin also regulates gene expression ( reviewed in [42] , [43] , [44] , [45] and is thought to promote chromosome looping to support promoter-enhancer interactions . Cohesin may assist the CTCF-mediated insulator function , and often binds to sites that overlap with CTCF binding sites , or to binding sites of other factors . Together , it remained elusive whether SMC1β contributes to SCC in early meiosis , whether SMC1β plays a particular role in sex body formation and MSCI , whether it acts in induction and processing of DNA DSB repair foci , and whether it affects meiotic gene expression . Our analyses of these features elucidate the relationship between SMC1β and SMC1α , and reveal a consistent concept of a major function of SMC1β in two fundamental processes , SCC and synapsis .
Smc1β−/− and Spo11−/− mice have been previously described [13] , [36] , [46] and are in the C57BL/6 background . Ethics statement: Animals were bred and maintained under pathogen-free conditions at the Experimental Center of the Medizinisch-Theoretisches Zentrum of the Medical Faculty at the Dresden University of Technology according to approved animal welfare guidelines , permission number 24-9168 . 24-1/2010-25 granted by the State of Saxony . Surface-spread chromosomes were prepared by detergent spreading adapted from [47] . 10 µl of a testicular cell suspension were added to 80 µl 1% Lipsol solution and spread on a glass slide . After swelling of cells for 10 min , 90 µl Peter's fixative ( 1 . 25% glutaraldehyde , 1% paraformaldehyde in 0 . 1 M cacodylate buffer , pH 7 . 2 , containing 0 . 15% Triton-×100 ) were added and cell nuclei were allowed to dry for 90 min in a wet chamber . Alternatively , in some experiments , 1 . 5 µl of single cell suspension were dropped into 7 µl of 0 . 25% of NP40 on a glass slide . Cells were allowed to lyse for 2 mins and then fixed by adding 24 µl of fixative ( 1% paraformaldehyde , 10 mM sodium borate buffer pH 9 . 2 ) . Samples were incubated for 1 h at room temperature in a wet chamber . Object slides were washed twice for 1 min with 0 . 4% Agepon ( AgfaPhoto ) and twice for 1 min with water before freezing at −80°C or further usage . For SUMO-1 immunolabeling , spread nuclei of testes were prepared according to Peters et al . , 1997 [48] . For immunostaining , cells were permeabilized in 0 . 5% Triton X-100/PBS for 30 min , quenched in 0 . 5% glycine/PBS for 5 min , both at room temperature , blocked in PBTG ( 0 . 1% BSA , 0 . 5% fish gelatine , 0 . 05% Tween-20 ) for 5 min before primary antibody incubation in PBTG for 90 min at 37°C or overnight at 4°C . Antibodies were used as follows: mouse anti-SYCP3 ( hybridoma supernatant ) , guinea-pig anti-HORMAD1 ( 1∶100 , a kind gift from Dr . A . Tóth , TUD , Dresden , Germany ) , anti RAD21L ( a kind gift from Dr . T . Hirano , Riken , Japan; [49] ) , γH2AX ( 1∶100 , Upstate , 05-636 ) , SUMO-1 ( 1∶200 Cell Signaling Technology , #4930 ) , rabbit anti-SMC3 ( 1∶100 ) , rabbit anti-STAG3 ( 1∶100 ) , rabbit anti SMC1α ( 1∶20 ) raised against the C-terminus , rabbit and rabbit or mouse monclonal anti SMC1β as described [36] , [50] . FISH was performed according to the manufacturer's protocol ( Metasystems GmbH ) . Ten µl of probe mixture was put on slides and covered with a coverslip . Both sample and probe were denatured simultaneously by heating on a hotplate at 75°C for 2 min , followed by incubation of slides in a humidified chamber at 37°C overnight for hybridization . Post hybridization washes were performed with 0 . 4× SSC at 72°C for 2 min and 2× SSC , 0 . 05% Tween-20 at room temperature for 30 sec . Slides were incubated with DAPI and signals were analyzed . Cells were identified based on their DAPI staining and staged according to their number of pericentric heterochromatin domains . Wt and Smc1β−/− pachytene cells were characterized by 5–7 pericentric heterochromatin domains [51] , [52] . For measurement of SPO11–oligonucleotide complexes , both testes from each mouse were used per experiment , that is , littermate comparisons were made on a per-testis basis . Testis extract preparation , immunoprecipitation and western blot analysis were performed essentially as described [53] . Testes were decapsulated , then lysed in 800 µl lysis buffer ( 1% Triton X-100 , 400 mM NaCl , 25 mM HEPES-NaOH at pH 7 . 4 , 5 mM EDTA ) . Lysates were centrifuged at 100 , 000 rpm ( 355 , 040× g ) for 25 min in a TLA100 . 2 rotor . Supernatants were incubated with anti-mouse SPO11 antibody 180 ( 5 µg per pair of testes ) at 4°C for 1 h , followed by addition of 40 µl protein-A–agarose beads ( Roche ) and incubation for another 3 h . Beads were washed three times with IP buffer ( 1% Triton X-100 , 150 mM NaCl , 15 mM Tris-HCl at pH 8 . 0 ) . Immunoprecipitates were eluted with Laemmli sample buffer and diluted six-fold in IP buffer . Eluates were incubated with an additional 5 µg anti-mouse SPO11 antibody 180 at 4°C for 1 h , followed by addition of 40 µl protein-A–agarose beads and incubation overnight . Beads were washed three times with IP buffer and twice with buffer NEB4 ( New England BioLabs ) . SPO11-oligonucleotide were radio-labelled at 37°C for 1 h using terminal deoxynucleotidyl transferase ( Fermentas ) and [α-32P] dCTP . Beads were washed three times with IP buffer , boiled in Laemmli sample buffer , and fractionated on 8% SDS–PAGE . Immunoprecipitates were transferred to a PVDF membrane by semi-dry transfer ( Bio-Rad ) . Radiolabelled species were detected and quantified with Fuji phosphor screens and ImageGage software . For western blot analysis , membranes were probed with anti-mouse SPO11 antibody 180 ( 1∶2 , 000 in PBS containing 0 . 1% Tween 20 and 5% non-fat dry milk ) , then horseradish-peroxidase-conjugated protein A ( Abcam; 1∶10 , 000 in PBS containing 0 . 1% Tween 20 and 5% nonfat dry milk ) , and detected using the ECL+ reagent ( GE Healthcare ) . Total testis RNA was extracted from four pairs of 12-day-old and 16-day-old Smc1β−/− and wild-type control littermate mice with TRIzol reagent ( Invitrogen Inc . ) according to the manufacturer's protocol . Briefly , testes were surgically removed and the Tunica albuginea was detached . Testes were dounce-homogenized in TRIzol reagent prior to phenol-chloroform extraction of RNA . The integrity of the RNA solubilized in water was confirmed by use of the BioAnalyzer ( Bio-Rad ) . Microarray experiments were performed in quadruplicates with littermate mouse pairs . 800 ng of RNA were applied to microarray analysis on the One-Color Microarray-Based Gene Expression Analysis System ( Agilent Technologies Inc . ) according to the manufacturer's protocols . Briefly , cDNA was synthesized using M-MLV Reverse Transcriptase followed by labeling with the fluorescent dye Cy3 . Amplified cRNA was purified with RNeasy mini spin columns ( Qiagen Inc . ) followed by hybridization to 4×44K mouse whole genome microarrays ( Agilent Technologies ) . Microarray experiments were performed in quadruplicates with littermate mouse pairs . Scanning of the chips was performed using the Agilent Microarray Scanner . Normalization of the data was performed using the software GeneSpring GX 7 . 3 ( Agilent Technologies ) . The normalization algorithm was scaling to the median of all samples . No baseline transformation was performed . Total testis RNA of Smc1β−/− and wt mice was DNase-treated ( 1 . 25 U DNase/µg RNA ) for 30 min at 37°C and reversely transcribed using Superscript II Reverse Transcriptase ( Invitrogen ) . For PCR amplification , full-length mRNA sequence primer pairs were designed using the web-based tool Primer3 version 3 . 0 [54] to yield 120 to 180 bp long intron-spanning PCR products . For each reaction , 1 µl of diluted cDNA generated from 40 ng of RNA was amplified in a 10 µl reaction volume using the QuantiTect SYBR Green PCR kit ( QIAGEN ) in a Rotorgene 3000 thermal cycler ( Corbett Research Inc . ) . Reactions were performed in duplicates and two or three mouse pairs were analyzed for each gene . Average mRNA levels of beta-actin ( Actb ) and glyceraldehyde-3-phosphate dehydrogenase ( Gapdh ) were used for normalization . PCR primer sequences are shown in Supplementary Table S1 . MiRNA expression analysis was performed using the miRXplore microarrays ( Miltenyi Biotech GmbH ) according to the manufacturer's protocol . Microarrays carried DNA oligonucleotides with a reverse-complementary sequence of mature mouse miRNAs . The 50th percentile of background intensity values was applied for data normalization and dye bias was corrected by Lowess normalization . The normalized mean ratios of Smc1β−/− versus wt were calculated . The microarray mutant versus wild-type log fold changes were normalized according to the more stable RT-qPCR measurements of 20 selected genes . The respective regression model , which was applied to all the genes , is the following . Y corresponds to the RT-qPCR measurement and x to the microarray fold change . The normalized data have been subjected to statistical clustering testing by applying a two-sided Wilcoxon rank sum test [55] as implemented in R ( function wilcox . test . R ) . A 10 gene sliding window approach was used , testing the null hypothesis that the fold changes of the genes in each window cluster better than in the rest of the genome . Application of the same approach to permuted log-ratio profiles ( 10 whole genome permutations ) identified that the detected clusters have been observed significantly more frequently in the orignal data than in the permuted ones . The respective false discovery rates ( FDR ) are plotted in dots across chromosomes 1-X . Single-cell suspensions of testes were resuspended in FACS buffer ( HBSS supplemented with 20 mM HEPES ( pH 7 . 2 ) , 1 . 2 mM MgSO4 , 1 . 3 mM CaCl2 , 6 . 6 mM sodium pyruvate , 0 . 05% lactate , glutamine , and 1% fetal calf serum ) at a density of 2 million cells/ml . Bis-benzamide Hoechst33342 ( 5 µg/ml , Hoechst ) was added before incubation of the cells for 1 h at 32°C . For exclusion of dead cells , propidium iodide ( PI , 2 µg/ml ) was added before FACS analysis . Cells were analyzed using a LSRII flow cytometer ( BD Biosciences ) using the FACSDiva software version ( BD Biosciences , version 6 . 1 . 3 ) . Hoechst was excited with the 355 nm UV laser and emission filters Emerald ( 585/42 nm , LP ) and Alexa 350 ( 505 nm ) were applied . PI was excited with the 488 nm blue laser and emission was filtered using PE ( 685 nm ) .
The current study aimed at elucidating in detail the role of SMC1β in SCC and synapsis , and therefore a number of meiotic processes and features were analyzed , all of which relate to these two processes . Previously , we showed that loss of SMC1β causes partial asynapsis of many autosomes and total asynapsis of a few [36] , [57] . In wild-type ( wt ) pachytene cells only the sex body with its large unpaired regions stains intensely for γH2AX . In Smc1β−/− spermatocytes that we classified as early/mid pachytene small clouds of γH2AX staining were observed on many chromosomes [36] . The most advanced Smc1β−/− spermatocytes show widespread SYCP1 staining , little HORMAD1 staining , and a very low percentage of cells express H1t , a mid-pachytene marker . We classified the early/mid pachytene Smc1β−/− spermatocytes based on the highest degree of axes compaction that can be observed . The sex body was mostly absent Smc1β−/− spermatocytes , and if a prominent γH2AX cloud was seen , it was one among several γH2AX clouds , and thus the occurrence of a sex body remained uncertain . To initiate the present study , a thorough analysis of spermatocyte asynapsis was performed using various combinations of staining for SYCP1 , SYCP3 , γH2AX , HORMAD1 , and SUMO-1 ( Fig . 1 , Suppl . Fig . S1 ) . SYCP1 localizes to synapsed regions of prophase I chromosomes , SYCP3 stains asynapsed and synapsed axial elements ( AEs ) , and γH2AX localizes to asynapsed regions , which are also transcriptionally silenced regions , and to DNA DSBs . HORMAD1 specifically stains asynapsed chromosomes or asynapsed regions of chromosomes and disappears from chromosomes upon formation of the SC [10] , [11] , and SUMO-1 localizes preferentially to asynapsed regions including the sex body [58] , [59] , [60] . The level of asynapsis varies considerably between individual Smc1β−/− cells as illustrated in Fig . 1A , but there were virtually no early to mid pachytene Smc1β−/− spermatocytes without at least partial asynapsis . The average number of partially asynapsed chromosomes was 1 . 8+/−0 . 9 per cell , the average number of entirely asynapsed chromosomes including the sex chromosomes was 6 . 8+/−1 . 7 per cell . Except the unsynapsed regions of the X and Y chromosomes , there are no asynapsed chromosomes in wt pachytene cells . Synapsis of X and Y chromosomes is restricted to the short PAR . In the vast majority of Smc1β−/− spermatocytes neither the PAR-synapsed X and Y chromosomes nor the typical sex body was detected by either HORMAD1 , γH2AX , or SUMO-1 staining combined with labeling the SC ( SYCP1 ) or the AEs ( SYCP3 ) ( Fig . 1B , 1C ) . HORMAD1 , γH2AX or SUMO-1 localized to several asynapsed chromosomes . Quantification of γH2AX staining revealed that a typical sex body could not be observed . Rather , two ( app . 25% of the cells ) or more ( app . 72% of the cells ) γH2AX clouds or multiple patches , of which occasionally ( app . 3% of cells ) one was more prominent , were found . In addition , we also evaluated X-Y pairing using fluorescence in situ hybridization ( FISH ) ( Fig . 1D ) . Pachytene cells were identified by light microscopy based on their characteristic DAPI-stained pattern of heterochromatin [51] , [52] ( Suppl . Fig . S1D ) . In 98% of the Smc1β−/− cells ( 59 of 60 ) , the two FISH signals were clearly apart , indicating that the X and Y were not synapsed within the PAR , consistent with the absence of a sex body in which both chromosomes ought to be present . In one of the cells the signals were very near each other , but FISH did not allow us to determine if PAR synapsis had occurred . In wt , 90% of the cells ( 36 of 40 ) showed partially overlapping signals , and in the remaining four cells , the signals did not visibly overlap but were very close to each other . This indicates that X/Y synapsis and sex body formation fails in the absence of SMC1β , and that the remaining SMC1α-type cohesin complexes alone are not sufficient for either process . This also implies that Smc1β−/− spermatocytes contain at least 21 microscopically visible chromosomes , i . e . at least 19 fully or partially formed SCs and the two separate sex chromosome AEs . Thus every cell has at least one pair of fully asynapsed chromosomes – the X and Y chromosomes – , and often many more , in addition to some partial asynapsis . The previous section showed significant levels of autosomal asynapsis and a failure to synapse the X and Y chromosomes in absence of SMC1β . High levels of autosomal asynapsis correlate with decreased expression of autosomal genes , i . e . MSUC , and with decreased MSCI . In male mice with mutations in genes essential for establishment of synapsis ( Spo11 , Dnmt3l , Msh5 , Dmc1 ) , regions of asynapsis are transcriptionally silenced through sequestration of silencing factors such as γH2AX that are no longer available to establish MSCI [24] . The deficiency to form sex bodies in Smc1β−/− spermatocytes , and thus the failure to see γH2AX accumulation on the sex chromosomes , suggests failing MSCI . In addition , the presence of autosomal asynapsis in Smc1β−/− spermatocytes renders an MSUC response likely . However , a direct contribution of cohesin to regulation of gene expression in meiotic cells appears also possible considering this role of cohesin in somatic cells . To test this , we performed microarray analysis of gene expression of wt and Smc1β−/− spermatocytes . We chose to compare mRNA expression in testes from 16 day-old mice when the first , synchronized wave of meiosis proceeds . At this stage the cells in our wt and Smc1β−/− littermate mice have just reached pachynema . There is no apoptosis yet in Smc1β−/− mice , where apoptosis commences at day 17 at the earliest [36] . Minor differences were observed in the cellularity of wt and Smc1β−/− testes at that age ( Suppl . Fig . S2 ) . Gene expression patterns of testes of four pairs of wt and Smc1β−/− littermates were analyzed . We performed eight independent gene expression measurements using the One-Color 4×44K Whole Mouse Genome Array system ( Agilent ) . The data yielded a highly significant pattern of changes in gene expression ( Fig . 2A ) . Of all 347 genes whose expression was altered in Smc1β−/− spermatocytes by at least two-fold with a stringent p-value threshold of 5×10−5 , 85% were down-regulated and 15% up-regulated . Of the aberrantly regulated autosomal genes , 91 . 2% were down-regulated , between 5 and 25 genes per chromosome ( Fig . 2 ) . In striking contrast , a large number of up-regulated genes are found on the X chromosome . Nineteen X chromosomal genes were up-regulated , while only 2 X-linked genes were down-regulated ( Fig . 2B ) . On the Y chromosome , 2 genes were up-regulated , none were down-regulated . Thus , early pachytene spermatocyte gene expression on autosomes and sex chromosomes reacts inversely to loss of SMC1β . For the validation of microarray data , reverse transcription quantitative PCR ( RT-qPCR ) was performed for top hits of several categories of mis-regulated genes . The genes were grouped into autosomal down-regulated genes , autosomal up-regulated genes and mis-regulated genes on the X- and Y-chromosome . For 21 of the genes of these groups , expression ratios from microarrays and RT-qPCRs from two or three mouse pairs are depicted in Figure 2D . Respective ratios of data from the microarray and RT-qPCR analyses correlated with a high correlation coefficient ( R2 = 0 . 88 ) ( Fig . 2E ) . Interestingly , the Y-linked Zfy1/2 genes , which were found up-regulated in Smc1β−/− spermatocytes , are known to strongly contribute to the pachynema , stage IV , quality surveillance mechanisms [26] . Expression of the transcription factor ZFY1/2 at that stage causes cell death . Up-regulation of Zfy1/2 gene expression was also confirmed by RT-qPCR ( Fig . 2D ) and suggests that this contributes to apoptosis of stage IV Smc1β−/− spermatocytes . To test if certain genomic regions in Smc1β−/− pachytene cells show specific clustering of up- or down-regulated genes we performed two types of analysis , using ( 1 ) a two-sided Wilcoxon rank sum test using a sliding window of 10 genes , and ( 2 ) a three-state Hidden Markov model ( Fig . 2C; Suppl . Fig . S3 , S4 ) . Both methods identified focal clusters of predominantly down-regulated genes on the autosomes , whereas clusters consisting of primarily up-regulated genes were identified on the X-chromosome ( the Y-chromosome was excluded from this analysis ) . However , after correcting for multiple hypothesis testing , the remaining number of significant clusters on the autosomes was relatively small because of the global down-regulation on the autosomes . Thus , due to the large number of down-regulated genes on the autosomes such clusters could theoretically emerge by chance . The patterns of changes in gene expression between leptonema and mid-pachynema are complex and not yet understood at high resolution for individual , defined stages or cell populations . Generally , it is thought that with completion of synapsis in pachynema , autosomal gene expression increases as the silencer proteins such as γH2AX retract from synapsed autosomes and assembles on the sex chromosomes . In early/mid pachytene Smc1β−/− spermatocytes a significant number of autosomal genes fails to be expressed . To determine whether the down-regulation of autosomal gene expression seen in Smc1β−/− spermatocytes correlates with synapsis , i . e . likely originates from an MSUC response rather than a direct effect of SMC1β on transcription , we prepared mRNA from testes of 12 day-old mice ( Fig . 2F ) . At that stage , spermatocytes of our mouse strains are in early zygonema and thus there is still little synapsis . If SMC1β has a direct , synapsis-independent effect on transcription it should be measurable at this stage , similar to the pachytene effect , albeit limited to those relatively few genes that are normally expressed in early zygonema [22] . Assaying the samples by the same microarray technology revealed , however , not a single down- or up-regulated gene at a p-value below 0 . 0001 and a minimum 2-fold change ( Fig . 2F , 2G ) . At a less stringent p-value of 0 . 001 eight genes were identified of which half were down-regulated , half up-regulated . In contrast , at this p-value 1080 genes , almost all down-regulated , were identified for the pachytene samples . This strongly suggests a direct correlation of autosomal down-regulation with the failure to completely synapse . Chromosomal distribution analysis of 160 mis-regulated genes ( day 12 ) identified at low stringency ( p<0 . 05; fc>1 . 5 ) shows them on all chromosomes ( Fig . 2F ) . Many of the mis-regulated genes localize to chromosome 15 with an enrichment in the proximity to the Smc1β locus , indicative of an artifact generated through alterations of the locus environment by gene targeting . We therefore disregarded these genes . No significant changes were detected in miRNA microarrays ( Suppl . Fig . S5 ) . Thus the vast majority of down-regulation in Smc1β−/− pachytene cells is very likely not caused by a direct effect of SMC1β on transcription , but rather by an MSUC response . Recently it was shown that synapsis affects localization of phosphorylated SMC3 [61] . In mitotic cells , SMC1 and SMC3 become phosphorylated in an ATM- or ATR-dependent manner within the DNA damage response [62] , . SMC1 and SMC3 localize to DSBs in somatic cells ( reviewed in [40] , [64] ) , and in this respect behave similar to γH2AX . ATR localizes to asynapsed regions in meiocytes [9] , and recently it was reported that pSer1083-SMC3 also localizes preferentially to asynapsed chromosomes , including the sex chromosomes in pachynema and diplonema [61] . We extended this observation , and asked whether pSer966-SMC1 also localizes to asynapsed regions including the sex body , and how phosphorylated SMC1α and SMC3 behave in mutant backgrounds . In leptonema and early zygonema , pSer1083-SMC3 localizes along chromosomes in a punctuated pattern , in pachynema and diplonema pSer1083-SMC3 marks the sex body ( Fig . 3A ) . Unlike HORMAD1 , which localizes along the sex chromosome axes ( Fig . 3B ) , pSer1083-SMC3 appears as a cloud similar to , but not as expanded as the γH2AX cloud ( Fig . 3A , B and Fig . 1B ) . In wt pSer1083-SMC3 is found exclusively at the sex body from late zygonema to diplonema ( Fig . 3A ) . Controls showed that the phospho-SMC1 peptide did not block anti pSer1083-SMC3 reactions , but the phosphorylated SMC3 epitope peptide eliminated the signal ( Suppl . Fig . S6A ) . No X/Y axes staining appeared even at low concentrations of the pSMC3 peptide , i . e . the pSer1083-SMC3 always appeared as a cloud . In metaphase I , pSMC3 was observed at centromeres , a prominent structure stained by SYCP3 at this stage ( Suppl . Fig . S6B ) . In Smc1β−/− spermatocytes ( Fig . 3C , D ) pSer1083-SMC3 localizes to unsynapsed chromosomes in late zygotene . In early/mid pachytene the pSer1083-SMC3 is not detected at sex bodies , which as described above do not form . Occasionally , deposits of pSer1083-SMC3 are seen that associate with part of a chromosome . On Smc1β−/− spermatocyte spreads the pSer1083-SMC3 is observed at the HORMAD1-stained asynapsed chromosomes , regardless of their number ( Fig . 3D ) . Since SMC1α is present [36] and SMC1β-deficient cells were analyzed , this indicates that an SMC1α-pSer1083SMC3 complex marks asynapsed regions , at least in the absence of SMC1β . The anti pSer966-SMC1 antibody recognizes only phosphorylated SMC1α since the Ser966 is not conserved between SMC1α and SMC1β . In wt leptonema spermatocytes , no pSer966-SMC1 was detected . In late zygonema , pSer966-SMC1 is preferentially seen at the asynapsed axes . In pachynema and early diplonema pSer966-SMC1 localizes only to the sex body , and in late diplonema pSer966-SMC1 disappears ( Fig . 4A ) . pSer966-SMC1 localizes almost exclusively to the chromosome axes and does not form clouds like pSer1083-SMC3 . This suggests that non-phosphorylated SMC1α/pSMC3 complexes localize to the sex chromosome loops and show as clouds , and that pSMC1α/SMC3 complexes associate with the X/Y axes . It is rather unlikely that an alternative SMC1β/pSMC3 complex localizes to the sex body cloud , since SMC1β was only seen at the sex chromosome axis [35] . In Smc1β−/− spermatocytes , the pSer966-SMC1 localizes along the asynapsed chromosomes ( Fig . 4B ) . Cohesin can be loaded onto mitotic cell chromosomes upon introduction of DNA DSBs [65] . The use of Spo11−/− mice lacking programmed DSBs [12] , [13] allowed us to analyze the recruitment of cohesins , including phosphorylated SMC1 and SMC3 , to meiotic chromosomes . Earlier , SMC1α and SMC3 were found to localize to chromosomes of Spo11−/− spermatocytes [66] . We observed all of the cohesin proteins tested here to localize to chromosomes of Spo11−/− cells , including SMC1α , SMC1β , SMC3 , STAG3 , RAD21 and RAD21L ( Suppl . Fig . S7A–G ) . The patterns of association differed between individual cohesins . Costaining of SMC1α with HORMAD1 revealed that in Spo11−/− spermatocytes SMC1α localizes along AEs preferentially where no HORMAD1 associates , i . e . more SMC1α localizes to regions of non-homologous associations . There is little SMC1α on non-associated chromosomes ( Suppl . Fig . S7A; quantification in Suppl . Fig . S8 ) . As published before [50] , in wt cells SMC1α localizes along every chromosome , including asynapsed chromosomes in zygonema , and synapsed chromosomes and the unpaired region of the X/Y in pachynema , and stains in comparable intensities whether synapsed or not . In Spo11−/− spermatocytes SMC3 , the only cohesin subunit present in all cohesin complexes , localizes to asynapsed regions decorated with HORMAD1 and to non-homologously synapsed regions ( Suppl . Fig . S7B and quantification in Suppl . Fig . S8 ) . In wt , SMC3 also localizes to asynapsed and synapsed regions with equal intensity [50] . SMC1β localizes to the axes of all chromosomes in wt and Spo11−/− spermatocytes . Together , this suggests that SMC1β and not SMC1α complexes localize preferentially to non-associated regions ( Suppl . Fig . S7G ) . The data also suggest that localization of SMC1α to asynapsed regions depends on SPO11 . STAG3 decorates all axes in a punctuate pattern in wt and Spo11−/− cells ( Suppl . Fig . S7C ) . RAD21L localizes to the asynapsed axes in Smc1β−/− spermatocytes ( Suppl . Fig . S7D ) . RAD21L also localizes in a somewhat punctated fashion all along the Spo11−/− axes , and in wt pachytene cells preferentially accumulates at asynapsed regions , particularly the sex chromosomes ( Suppl . Fig . S7E ) . In wt , RAD21 co-localizes in many dots with SYCP3 along autosomes , but there is little overlap with SYCP3 on the sex chromosomes ( Suppl . Fig . S7F ) , where the RAD21 signals accumulate next to the axes . In Spo11−/− chromosome spreads , very little of RAD21 , which appears as many dots throughout the chromatin , co-localizes with SYCP3 on the chromosome axes . In spermatocytes lacking SPO11 we detected no pSer1083-SMC3 on chromosomes axes , whether unsynapsed or non-homologously associated , consistent with the report by Fukuda et al . ( 2012 ) ( Fig . 3E ) . However , we found pSer1083-SMC3 to form some clusters within the nuclei reminiscent of pseudo sex bodies , and to localize in dispersed spots throughout the nuclei in Spo11−/− spermatocytes ( Fig . 3E ) . In many cells there was overlap of pSer1083-SMC3 with staining for γH2AX at the pseudo sex bodies . In Smc1β−/−Spo11−/− spermatocytes , the localization of pSer1083-SMC3 in many cells is similar to that seen in Spo11−/− spermatocytes , but in about 40% of the cells , no staining for pSer1083-SMC3 was observed ( Fig . 3F ) . Together this indicates that SMC1α forms complexes with pSer1083-SMC3 in absence of SPO11 and SMC1β . We also noted moderately longer chromosome axes in Smc1β−/−Spo11−/− spermatocyte spreads as compared to Smc1β−/−spreads ( Fig . 4B , 4D; quantification in Suppl . Fig . S9 ) , although this varied and in some cells this difference was small . In Smc1β−/− spermatocytes the pSer966-SMC1 signal is seen on asynapsed autosomes ( Fig . 4B ) . This supports the above hypothesis of pSMC1α/SMC3 complexes to localize to asynaptic regions . No pSer966-SMC1 was detected in Spo11−/− or Smc1β−/−Spo11−/− spermatocytes ( Fig . 4C , D ) . These data indicate that SPO11 , and presumably SPO11-generated DSBs , are required for SMC1α Ser966 phosphorylation , consistent with its kinetics of appearance in early zygonema . Asynapsis may have consequences for the processing of meiotic double-strand breaks ( DSBs ) as suggested by the correlation of synapsis and disappearance of repair foci , which is delayed on the largely unpaired X/Y chromosomes . To address formation and processing of DSBs in Smc1β−/− spermatocytes , we first determined whether programmed DSBs are properly generated in absence of SMC1β . Thus , SPO11-dependent DSBs were measured by visualizing the SPO11-oligonucleotide intermediate product of the reaction [67] . Fig . 5A demonstrates that SPO11 protein is produced in adult wt and Smc1β−/− cells , although in the latter the SPO11α splice variant is absent as it is expressed later and may not contribute much to leptotene spermatocyte DSB formation [68] . The alternative splice isoform SPO11β is present although in much smaller quantities as in wt , probably because late pachytene cells , which contribute significantly to total SPO11 in spermatocytes , are absent . The early SPO11 enzyme is sufficient to produce substantial levels of SPO11-oligonucleotide complexes , which is also seen in the control Dmc1−/− samples that contain even less SPO11 enzyme , and whose spermatocytes arrest in late zygonema . Smc1β−/− spermatocytes develop further into pachynema and probably therefore synthesize more SPO11 . We assessed steady-state levels of SPO11-oligonucleotide complexes in whole-testis extracts from wt and mutant mice . We found a very mild reduction in adult Smc1β−/− testes ( 0 . 84±0 . 20-fold relative to wt , mean and s . d . , n = 2 littermate pairs ) . In contrast , adult Dmc1−/− testes , which are similarly reduced in size as Smc1β−/− testes , displayed a more substantial reduction in SPO11-oligonucleotide complex levels ( 0 . 40±0 . 07-fold , n = 2 ) , as previously reported [53] . To account for different sizes and cellularities in wt and mutant testes , we also examined SPO11-oligonucleotide levels in juvenile mice at day 14 . 5 post partum , when the majority of spermatocytes have not yet progressed far enough into meiotic prophase to be affected by the mid-pachytene checkpoint . At this age , SPO11β is only weakly expressed , but SPO11-oligonucleotide complexes are readily detectable . Complex levels were moderately reduced in Smc1β−/− testes ( 0 . 68±0 . 14-fold , n = 3 ) and in Dmc1−/− testes ( 0 . 77±0 . 03-fold , n = 2 ) . This does not argue for an impairment in SPO11 DSB formation , since minor variations such as slightly delayed initial spermatogenesis in the Smc1β−/− mutant may cause this phenotype . Efficient production of DSBs in the adult was confirmed by quantifying the number of initial repair foci . Repair proteins assemble at DSBs and form foci whose appearance is indicative of DSBs , and whose processing , i . e . disappearance , is indicative of proper repair . IF analysis of RAD51 and DMC1 foci revealed that the same number of foci are generated in wt and Smc1β−/− spermatocytes at the early zygotene stage ( Fig . 5B , Suppl . Fig . S10 ) . In late zygotene and early pachytene , these foci are processed and disappear in wt cells , but fail to do so efficiently in Smc1β−/− spermatocytes . Detailed analysis of the persistent foci by co-staining with HORMAD1 revealed that they localize predominantly to asynapsed regions and to sex chromosomes ( Fig . 5B , C ) . As expected , no foci were detected in Spo11−/− ( not shown ) and in Smc1β−/−Spo11−/− spermatocytes ( Fig . 5B ) . Staining of Smc1β−/−Spo11−/− spermatocyte spreads ( Fig . 4D ) unexpectedly showed 40 clearly separate , SYCP3-stained chromosome cores . Therefore , we further analyzed whether non-homologous associations in Spo11−/− spermatocytes depend on SMC1β . Staining of Smc1β−/−Spo11−/− spermatocyte spreads for HORMAD1 and SYCP3 showed 40 separate , short chromosome cores , i . e . univalents without any non-homologous association ( Fig . 6A ) . Axes shortening is a known , prominent phenotype of Smc1β−/− spermatocytes [36] . Since Smc1β−/−Spo11−/− spermatocytes die slightly earlier , in late zygotene , than Smc1β−/− spermatocytes , the axes are not yet as fully compacted and are slightly longer than in the latest stages of Smc1β−/− spermatocytes ( Suppl . Fig . S9; Fig . 4 ) . Quantification showed that 58% ( 22 of 38 ) of Smc1β−/−Spo11−/− spermatocytes carry 40 chromosome cores without non-homologous associations ( Fig . 6A ) . The remaining 42% of cells showed more than 40 axes ( Fig . 6B ) . None of the cells showed fewer axes . In addition to a failure to support non-homologous associations , this suggests loss of sister chromatid cohesion of some chromosomes in almost half of the cells ( see below ) . Thus , SMC1β is required for non-homologous associations , which SMC1α alone cannot support . Other cohesins like SMC3 , RAD21L and STAG3 still associate with the shortened axes indicating that the respective SMC1α-type complexes are present ( Fig . 6C ) . The number of DSBs depends on SPO11 dosage [69] . Would the introduction of fewer DSBs still allow homologous synapsis in the synapsis-weakened Smc1β−/− spermatocytes ? Smc1β−/−Spo11+/− spermatocytes appeared phenotypically not different from Smc1β−/− cells , i . e . showed about the same level of asynapsis when stained for SYCP3 and SYCP1 ( Fig . 6D ) or pSer966-SMC1α ( Fig . 6E ) . This indicates that wild-type levels of DSBs are not required for the level of synapsis seen in the absence of SMC1β . As expected , the number of DMC1 foci was slightly decreased in Smc1β−/−Spo11+/− spermatocytes compared to wt and persisted longer than in wt as described for the Smc1β−/− cells ( not shown ) . Further , we observed in some Smc1β−/−Spo11−/− spermatocytes but not in wt or single “knockout” cells what appears to be end-fusions of chromosomes , i . e . SYCP3-stained axes with a centromere signal at both ends ( Fig . 6B shows 2 examples ) . Mostly two , but also up to 6 , of the 40 or more chromosomes of about half of the cells ( n = 30 ) were seemingly connected or at least associated at one of their ends . It appeared as if mostly the centromere-distal ends of the two chromosomes were fused . Recently , we reported on gene dosage effects in Smc1β+/− and Rec8+/− mice [70] . A gene dosage effect of SMC1β was also observed for Smc1β+/−Spo11−/− spermatocytes . In a small subset of these cells ( 2% ) , no synapsis – whether homologous or non-homologous – was observed . In the remaining 98% of the cells , partial non-homologous synapsis typical for Spo11−/− spermatocytes was seen ( n = 30 ; not shown ) . Previously , we reported a requirement for SMC1β in centromeric SCC in an okadaic acid induced metaphase I-like stage in spermatocytes and in later stages of oocyte meiosis [36] , [37] . However , it remained unclear whether SMC1β is required for SCC in prophase I and whether it acts in arm and/or centromeric cohesion in early meiosis . In almost half of the Smc1β−/−Spo11−/− spermatocytes , such as the one shown in Figure 6B , the increased number of axes indicates loss of cohesion in some pairs of sister chromatids . However , 58% of the cells carry 40 axes , which show no signs of split sister chromatids , i . e . feature cohesion all along the chromosome arms . Thus , the remaining SMC1α complexes are sufficient for this fraction of arm cohesion at this stage . Since there is no pSer996-SMC1α in Spo11−/− cells , the non-phosphorylated form of SMC1α provides this sister chromatid cohesion . Non-phosphorylated SMC1likely acts together with non-phosphorylated SMC3 , since the pSMC3 is dispersed , not associated with the chromosome axes . Thus , both SMC1α and SMC1β can contribute to arm cohesion in early prophase I . Centromeric cohesion was analyzed in Smc1β−/−Spo11−/− and in Smc1β−/− spermatocytes ( Fig . 7 ) . The axes of Smc1β−/−Spo11−/− cells were investigated for split centromere signals indicative of loss of centromeric cohesion . Clearly split centromeres were observed in 35% of the axes ( n = 60 cells ) ( Fig . 7A ) . Chromosome surface spreading does not always resolve separate structures , and thus centromeres that were of large , unusual shape were distinctly quantified and found in 45% of the axes , excluding those with clearly separate centromeres . Perfectly overlapping centromeres , suggesting presumably conserved centromeric SCC , were present in the remaining 20% . This suggests a very significant contribution of SMC1β to centromeric cohesion . To confirm this , Smc1β−/− spermatocytes were similarly analyzed , although here the analysis is more complex , since the two pairs of sister chromatids with their four sister chromatids can be partially asynapsed , and may show two centromere signals because either of asynapsis or of loss of cohesion , which cannot be distinguished . We can conclude with certainty that centromeric SCC was lost only if there are three or four centromere signals . Some of the SCs showing two centromere signals may also have lost centromeric SCC . We observed 3 or 4 centromeres , i . e . loss of centromeric cohesion , in at least 3% of SCs . Another 5% of SCs showed 2 centromere signals and an unknown fraction of those may be derived from loss of SCC instead of asynapsis . Some SCs showed clearly asynapsis at the centromeric end with ( 30% ) or without ( 70% ) loss of centromeric cohesion . Centromeric asynapsis was as frequent as asynapsis within the chromosome arms or at the centromere-distal end . No split centromeres were observed in wt or Spo11−/− spermatocyte chromosome spreads .
This report addresses two major roles of cohesin in meiosis: sister chromatid cohesion and synapsis . All data and phenotypes reported here relate to these functions , either directly or indirectly . Previous reports implicated the meiosis-specific SMC1β cohesin in synapsis , post-prophase I SCC , chiasma maintenance and telomere protection [36] , [37] , [46] . However , important questions , which concern fundamental processes of male meiosis , remained to be resolved: ( 1 ) is SMC1β involved in SCC during prophase I of meiosis; ( 2 ) is SMC1β required for synapsis of sex chromosomes and for non-homologous synaptic associations; ( 3 ) does SMC1β affect regulation of gene expression during meiotic prophase I , and ( 4 ) is SMC1β necessary for generation and processing of meiotic , programmed DSBs ? Other , more complex questions are step-wise to be resolved as new mouse models become available , such as for the distinct roles of SMC1α and SMC1β complexes in meiosis , though this study provides several insights into the functions of these two complexes . Conclusions about the role of SMC1α complexes remain indirect , since these conclusions are based on functions of SMC1α that is present in SMC1β deficient spermatocytes . Whether SMC1α fulfills the same roles in wt cells cannot be determined with certainty . Sister chromatid cohesion as the prominent function of cohesin may be provided early in meiosis either by SMC1α- or SMC1β-based complexes . No evidence had previously been presented for a contribution of SMC1β to prophase I SCC , which did not appear to be grossly dysfunctional in Smc1β−/− prophase I cells . In addition , expression of the Smc1β gene starts only right after cells have entered meiosis , i . e . in early leptotene [35] , [71] . Thus the previous assumption was that SMC1α complexes provide most if not all SCC before metaphase I . A contribution of SMC1β to centromeric cohesion was suggested by published data , based on experimentally driving Smc1®−/− spermatocytes into a metaphase I-like state by okadaic acid treatment [36] . In this rather artificial set-up , 80 separate centromeres were observed in many Smc1®−/− cells indicative of complete loss of centromeric SCC of the 4×20 mouse chromatids [36] . However , some of the cells only partially lost SCC ( average of 40 centromere signals ) , and some showed app . 20 signals for intact centromeric SCC . We speculated that those cells suffering complete loss of SCC were of pachytene origin , while those that showed only partial loss of SCC may have originated from earlier stages . Together , it remained unclear whether SMC1β contributes to SCC earlier than metaphase I , i . e . in prophase I . Our analyses of wt versus Smc1β−/− or Smc1β−/−Spo11−/− spermatocytes demonstrate an important contribution of SMC1β to zygotene/pachytene SCC . At least one-third of centromeric SCC in Smc1β−/−Spo11−/− spermatocytes is impaired . At least 8% of SCs in Smc1β−/− spermatocytes lose centromeric cohesion , but this number , which includes chromosomes with three centromere signals only , is probably a significant underestimation . The many centromeres that show two signals cannot be assigned to either asynaptic homologs with intact centromeric cohesion or to synapsed homologs with defective centromeric cohesion . SCC of chromosome arms appeared intact in zygotene and pachytene Smc1β−/− spermatocytes , but much higher microscopic resolution may be required to determine this . The Smc1β−/−Spo11−/− strain with its complete absence of homolog associations revealed in almost half of the cells more than 40 axes , mostly between 42 and 48 axes , and thus showed individual sister chromatids . Thus , about 5 to 10% of the homologs in about half of the cells lost SCC . Thus , SMC1β also contributes to arm cohesion in early prophase I . However , arm cohesion is primarily supported by SMC1α cohesin under these conditions . We did not observe partial loss of arm cohesion , which would show as bubbles , i . e . separate strands , within homologs . This “all or none” response may indicate cooperative behavior of cohesin . It shall be explored in the future why SMC1α does not provide full centromeric cohesion , and why SMC1β does relatively little for arm cohesion , i . e . how and why this functional specialization occurs . SMC1β remains associated with centromeres until metaphase II , consistent with the hypothesis that SMC1β provides the essential centromeric cohesion beyond prophase I . Formally , the synergistic effect of deficiencies in SMC1β and SPO11 renders it possible that SPO11 itself also contributes to SCC . Given the prior knowledge on SPO11 we deem this unlikely , and rather think that the much more likely interpretation is that the deficiency in SPO11 and thus absence of homologous synapsis allowed us to reveal the SPO11-independent role of SMC1β in SCC . The analysis of Smc1β−/− , Spo11−/− , and Smc1β−/−Spo11−/− spermatocytes revealed several other important aspects of SMC1β function . It confirmed that SPO11 and thus SPO11-generated DSBs are not required for loading of cohesin proteins . The localization of RAD21 in Spo11−/− spermatocyte chromosome spreads differed from wt since much less RAD21 co-localized with the SYCP3-stained axes . More importantly , the non-homologous synaptic associations observed in Spo11−/− spermatocytes [12] , [13] entirely depend on SMC1β . SMC1α alone cannot sufficiently support these associations . Thus , a second instance of functional specialization between the two SMC1 variants has been identified . Non-homologous associations might reflect an early , transient stage in the search of the homologs to find each other even before DSBs are generated . We suggest that cohesin is involved in this early stage of partner search and may provide a platform or code that allows transient , weak “trial-and-error” interactions to enhance the chance of finding the homolog . In many Smc1β−/−Spo11−/− spermatocytes , but not in either single mutant or wt , several seemingly fused chromosomes were observed , where often the centromere-distal ends were fused . Earlier , we described telomere defects in Smc1β−/− meiocytes such as loss of telomeric sequences , SCs without telomeres , telomere stretches and extended bridges between telomeres consisting of stretched telomere DNA and telomeric proteins [46] . About 20 to 30% of the telomeres showed such aberrations , which we interpreted as loss of telomere protection in absence of SMC1β . The absence of protection of chromosome ends and the failure of the chromosomes to homologously synapse may contribute to end-fusions in the Smc1β−/−Spo11−/− spermatocytes . Homologously synapsed chromosomes are highly compacted and may have acquired a chromosome configuration prohibitive of end-fusions . This configuration does not exist in Smc1β−/−Spo11−/− spermatocytes . Telomere aberrations also cause failures in synapsis and may contribute to the asynapsis seen in SMC1β deficient cells . Average chromosome length was moderately increased in Smc1β−/−Spo11−/− spermatocytes compared to wt cells , but variation was high . This may either reflect a slightly different developmental stage of the double-deficient cells or reduced compaction in absence of SPO11-dependent synapsis . Why does SMC1α not bind efficiently to asynapsed , non-associated regions in Spo11−/− spermatocytes but binds to the largely unpaired sex chromosomes in wild-type cells ? One may speculate that there is a specific chromatin mark or other quality in wt cells that supports SMC1α binding to the sex chromosomes . This study further demonstrates that sex body formation depends on SMC1β and that in the absence of SMC1β , SMC1α alone cannot support PAR synapsis . The involvement of SMC1β may be direct in supporting synapsis for example through interaction of cohesin rings between the two pairs of sister chromatids . It can also be indirect if a certain axis length of the PAR is required for sufficient synapsis , and that length may not be available with the shortened Smc1β−/− axes . Our data also provides a likely explanation for the apoptosis of Smc1β−/− spermatocytes observed in mid-pachynema . While asynapsed autosomes are silenced , genes on the sex chromosome are transcriptionally activated . Thus , MSCI fails . This correlation between an MSUC reaction of autosomes and a failure of MSCI has been described and is attributed to the relocalisation of silencing factors such as γH2AX from sex chromosomes to asynaptic autosomes [9] . Made possible through the MSCI failure , the Y chromosome “spermatocyte killer gene” Zfy1/2 is expressed and probably substantially contributes to eliminating the mutant spermatocytes within the pachytene quality surveillance mechanism [26] ( reviewed in [72] . Our analysis of phosphorylated forms of SMC1α and SMC3 revealed a distinct pattern of localization , particularly at the sex chromosomes . While it is difficult to entirely exclude antibody staining artifacts , the measures taken such as peptide blocking experiments , single antibody staining , secondary antibody controls , and the use of well-established anti phospho SMC antibodies , as well as the specific staining patterns like staining of the axis only of unsynapsed chromosomes ( pSMC1 ) or staining only of the sex body chromatin ( pSMC3 ) render artifacts unlikely . Phosphorylation of SMC1 or SMC3 may serve as a mechanism to attribute functional properties to these cohesins at different localizations . The pSer1083-SMC3 may also serve as a marker for sex bodies , and the pSer966-SMC1α as a marker for the sex chromosome axes or asynapsed axes in mutants like Smc1β−/− . Since in about 40% of Smc1β−/−Spo11−/− spermatocytes no pSer1083-SMC3 was detected , but all Spo11−/− spermatocytes showed pSer1083-SMC3 , the phospho-form may either be unstable or these 40% of cells may have been at slightly distinct stages of spermatogenesis as their counterparts in the same mouse . The absence of pSer966-SMC1α from Spo11−/− spermatocytes indicates distinct requirements for phosphorylation of SMC1α and SMC3 . The pSer966-SMC1α is observed at asynapsed chromosomes in Smc1β−/− spermatocytes but not in Spo11−/− spermatocytes that carry extensively asynapsed chromosomes . Thus , it appears as if SPO11-generated DSBs are required for SMC1α phosphorylation . SMC1α phosphorylation in somatic cells is catalyzed by ATM , whose activation in meiocytes depends on DSBs , which could explain the failure to see pSer966-SMC1α in Spo11−/− spermatocytes . Spo11 heterozygosity rescues the prophase I arrest of Atm−/− spermatocytes , and revealed that sex body formation and H2AX phosphorylation are supported by ATR [73] , [74] . In one explanation , the reduced number of DSBs and their slower generation allows ATR to complement the ATM deficiency , particularly at zygonema and pachynema , when ATR is strongly expressed . Recent work showed , however , that ATM restricts the number of SPO11-generated DSBs and that in the absence of ATM , expression levels of SPO11 become critical [53] . Too many DSBs in an ATM-deficient , SPO11-proficient situation are deleterious and the reduced levels seen in Atm−/− Spo11+/− spermatocytes are more permissive for meiotic progression . Despite this ATM-dependent DSB-triggered feedback mechanisms , the number of DSBs depends on SPO11 dosage , since Spo11+/− spermatocytes show reduced levels of DSBs [69] . Our analysis shows that wt levels of SPO11-generated DSBs are not required for SMC1β-independent non-homologous synapsis in Smc1β−/−Spo11+/− cells . Several studies investigated gene expression patterns during male gametogenesis [75] , [76] , [77] , [78] . Most of these studies compared two or more cell populations , e . g . spermatogonia with spermatocytes , spermatocytes with spermatids , or germ cells with Sertoli cells , but did not venture into analyzing meiotic sub-populations , which , for example , represent leptotene , zygotene or pachytene cells . These are very difficult to separate and to obtain in high purity and new mouse models expressing specific markers may be needed to achieve this . Therefore , juvenile male mice were used in some studies to exploit the synchrony of the first wave of meiosis during mouse puberty . However , this first wave of meiosis may not faithfully represents adult meiosis in all of its aspects . Nevertheless , altered patterns of gene expression were observed in pachytene spermatocytes compared to spermatogonia or spermatids . Some studies used mutant mouse strains such as Spo11−/− mice [79] to reveal gene expression patterns that depend on this gene and may be indicative of genes required for meiotic progression , but a clear assignment to this function requires additional analysis . In our study , we compared wt and Smc1β−/− spermatocytes at day 12 or day 16 post partum , at which spermatocytes have reached zygotene or early pachytene , respectively . FACS analysis of testis cells at day 16 revealed no major difference between wt and Smc1β−/− spermatocytes , but in Smc1β−/− samples some cells accumulated at a different position in the analysis window , indicating altered morphology . We cannot exclude a minor effect of slightly changed cellularity of the testis on the gene expression patterns . However , we assume that such an effect would be small given the absence of increased apoptosis at day 16 dpp in wt and Smc1β−/− testis , the use of littermates , and the similarity of testis tubule structure at this age . At the p-value of <0 . 001 more than a thousand genes differed by at least 2-fold between day 16 spermatocytes from wt and Smc1β−/− spermatocytes . A large majority ( 92% ) of the differently regulated autosomal genes , were down-regulated . In zygonema , however , only 8 genes differed at that p-value and they were down- or up-regulated . At the least stringent p-value of <0 . 05 , only 64 genes changed at least 2-fold , and down- and up-regulation was equally distributed among them . Therefore , a strong correlation was revealed between down-regulation of many genes in Smc1β−/− pachytene spermatocytes , and their significant asynapsis . As asynapsis is known to cause the MSUC response , we suggest that most if not all down-regulation of gene expression seen in Smc1β−/− spermatocytes is caused by asynapsis-dependent MSUC . One may also describe the down-regulation as a failure to up-regulate the genes upon synapsis in early pachynema . Provided that this view is correct , one may speculate that there may be regions on autosomes that are particularly prone to asynapsis . The presence of specific genes that are very significantly down-regulated may indicate “asynapsishigh” regions , and thus may suggests that cohesin-dependent synapsis happens with different efficiency and perhaps timing on distinct regions of meiotic chromosomes . Our search for chromosomal clusters of aberrantly expressed genes by two different methods revealed the presence of such clusters , although only few remained significant after correcting for multiple hypothesis testing . Hence , we observed a global reduction of expression on autosomes , but we could not identify a substantial number of specific genomic regions that are particularly affected by asynapsis in Smc1β−/− pachytene cells . Genes that are not expressed before and after synapsis , very low absolute expression levels , and possible feedback mechanisms may limit the analysis and thereby may prevent the identification of putative “asynapsishigh” regions . Considering the few genes expressed in zygonema that are mis-regulated in Smc1β−/− spermatocytes and are either up- or down-regulated , and considering the few pachytene genes that are up-regulated in Smc1β−/− spermatocytes , we cannot exclude , however , a small direct effect of SMC1β cohesin on gene expression , possibly through interaction with transcription factors . Alternatively , these changes in gene expression may be caused by the altered loop-axis structure seen in absence of SMC1β . For example , certain chromosome regions may be aberrantly localized in chromatin loops or packed into the axis , which may affect gene expression . Yet another phenotype of Smc1β−/− spermatocytes correlates with asynapsis: the failure to timely process DNA DSBs . While the DSBs are formed properly as indicated by analyzing SPO11-oligonucleotide complexes and by staining for RAD51 and DMC1 foci in early zygonema , these foci persist longer in Smc1β−/− spermatocytes . The repair foci are processed more slowly on asynapsed chromosomes such as sex chromosomes , and in mutants that show asynapsis . Our analysis showed that persistent repair foci localize predominantly to asynaptic chromosomes , lending support to the hypothesis that the delay in processing the repair foci is caused by asynapsis , where recombinational repair using the homolog cannot take place . The alternative pathway to repair such foci , that is the repair through recombination between sister chromatids of one homolog , is apparently not enhanced in Smc1β−/− spermatocytes , at least not to an extent that would rescue the delay in foci processing . Alternatively or additionally , more DSBs may be introduced where a chromosome or chromosomal region fails to synapse in order to further promote synapsis . The axes in the absence of SMC1β are only about half as long as those in wild-type [36]; yet the number of DSBs is only little reduced in Smc1β−/− spermatocytes . Therefore , axes length per se is unlikely to be a prime determinant of DSB numbers . Together , this study revealed important insights into the functions of SMC1β cohesin in prophase I SCC , in autosome and sex chromosome synapsis , and it provides evidence for specific , distinct functions of SMC1β and SMC1α cohesin complexes . Taking previous reports into account , three main functions of SMC1β have emerged that have multiple biological consequences for meiocytes: generation of the loop-axis architecture of SCs , homologous and non-homologous synapsis , as well as sister chromatid cohesion starting in early prophase I . | The generation of mammalian gametes through meiosis comprises two subsequent cell divisions . The first division , meiosis I , features highly specific chromosome structures , and behavior , and requires distinct sets of chromosome-associated proteins . Cohesin proteins , of which some are meiosis-specific , are essential for meiosis , but their particular roles in meiosis are incompletely understood . We show here that SMC1β , a meiosis-specific cohesin , serves key functions already in prophase of meiosis I: SMC1β contributes to keeping sister chromatids in cohesion at their centromeres and supports synapsis of the four sister chromatids present in these cells . SMC1β is required for the synapsis of the X and Y sex chromosomes . The failure of autosomes to properly synapse in absence of SMC1β causes extensive alterations in gene expression . This leads to expression of sex chromosome-linked genes , which are lethal at this stage , explaining the death of spermatocytes in mid-prophase I . Together with the analyses of other cohesin proteins and of phosphorylated forms of SMC3 and SMC1α , this paper describes hitherto undescribed properties and functions of meiotic cohesin in sister chromatid cohesion and synapsis . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | Meiotic Cohesin SMC1β Provides Prophase I Centromeric Cohesion and Is Required for Multiple Synapsis-Associated Functions |
Rabies is a viral zoonosis that has been described in limited numbers of studies in Ethiopia at large and among pastoralists in particular . This study assessed dog demography , bite wound prevalence and management , potential risk factors of disease transmission and knowledge attitude practice towards rabies among urban dwellers , pastoralists and health workers in Awash , Eastern Ethiopia . Information was collected by means of structured questionnaires and interviews and through medical and official records from the Agricultural and Health bureaus . Respondents totaled 539 ( 471 urban , 49 pastoralists , 19 medical ) . Dog ( s ) were owned in 33% urban and 75 . 5% pastoralist households respectively . Mean dog number per dog owning household was 1 . 50 ( 95%CI: 1 . 40–1 . 60 ) in urban and 2 . 05 ( 95%CI: 1 . 51–2 . 60 ) in pastoralists sites . Human Dog Ratio in Metahara was 4 . 7:1 . No bite wounds records were kept in medical facilities , where staff recalled around 100 bites per year , 2/3 being in adults . Over 90% of the respondents claimed knowing rabies but up to 79 . 2% pastoralist did not know how dogs acquire the disease; 37 . 3% urban and 23% pastoralist did not know the symptoms of rabies in dogs; 36% urban and 44% pastoralists did not know rabies symptoms in people . Eighty percent of pastoralists did not know that the disease was fatal in people if untreated . Over half ( 58 . 7% ) of pastoralist respondents go to traditional healers if bitten , despite a health extension worker program in place in the study area . Knowledge gaps were also shown amidst medical staff . The study highlighted overall poor disease knowledge , severe under-reporting of human rabies cases , lack of record keeping and poor collaboration between the public and animal health sectors and communities in rabies control .
Rabies is a viral zoonotic neglected disease caused by a negative stranded RNA virus from the Genus Lyssavirus [1] . Although a wide range of animals can become infected and transmit the disease , only mammals from the Carnivora and Chiroptera ( bats ) Order act as reservoir for the disease [2] . Domestic dogs are considered to be the main source ( >90% ) for human rabies in Africa [3 , 4] . Once the symptoms have appeared , the disease ends almost always fatally . Transmission to people occurs predominantly via infected animal bite or scratch as well as via their saliva through mucosa and broken skin [3] . Therapy has to be initiated immediately and relies on Post-Exposure prophylaxis ( PEP ) , which consists of rapid and thorough washing of the wound , completion of post-exposure vaccination schedules plus inoculation with rabies immunoglobulin ( RIG ) for severely exposed bite-victims . The disease claims 24 , 000 human deaths annually in Africa alone [4 , 5] . Rabies burden in Tanzania was 4 . 9 human death/100 , 000 based on active surveillance data on bite incidence and 0 . 62 human deaths/100 , 000 when based on national bite statistics [4] . Ethiopia is thought to be a high-burden country for rabies [6] . However , hard data on dog demography and ecology as well as true rabies incidence in dog and people are lacking . Information is based on estimations and extrapolations , small scale studies and limited record reviews [6–8] . The Ethiopian Public Health Institute ( EPHI ) is the only laboratory facility in the country to diagnose rabies and produce PEP . Ferni vaccines ( adult sheep brain nervous tissue vaccines ) were used at the time of this study , despite the WHO recommendation in 2006 to completely replace nerve tissue vaccine with cell-cultured based anti-rabies vaccines [9] . A retrospective review from EPHI records between 2001 and 2009 showed that 1026–1580 patients per year in and around Addis Ababa were taking PEP and that the total fatality human cases was 35–58 per year [6] . A one year follow-up in Gondar based solely on clinical diagnosis revealed an incidence rate per year of 2 . 3/100 , 000 [8] . Rabies incidence however , is likely to be much higher considering the lack of accurate data and underreporting of cases [4 , 5 , 10] . Published data of rabies from rural areas of Ethiopia , including pastoralists are however entirely lacking . The aim of this study was to try to gain a picture of the rabies epidemiology ( prevalence , risk factors ) at the animal-human interface in the Awash Basin . Dog demography , bite history and knowledge-attitude-practice ( KAP ) regarding the disease was assessed amongst pastoralists , urban dwellers and health workers by means of questionnaires and/or interviews . Constraints to rabies control and prevention in the study area are discussed .
The study was carried out between February and June 2012 in Metahara , the administrative center of Fentale woreda ( Oromia region ) with a population of 25 , 670 , its neighboring town Addis Ketema and Merti ( Metahara Sugar Cane Plantation ) as well as in the neighboring pastoral villages of the Oromia and Afar region ( Fig 1 ) . Urban dwellers were from various national ethnic backgrounds . Pastoralists were Ittu-Oromo , Kereyu-Oromo or Afar . Climate is semi-arid with bi-annual rainfalls . The area has an elevation ranging from 800 to 960 meter above sea level ( National Meteorology Agency; Metahara Agricultural Bureau ) . This cross-sectional study collected solely in depth interview/questionnaire data and no clinical samples . Questionnaires , with closed and open questions were translated into the local languages Amharic , Oromifa and Afarinia and back translated into English for consistency checks . The questionnaires were pre-tested in the study site . A trained interviewer administered all interviews in the local languages . Two sets of questionnaires were prepared , one for urban/ pastoralist dwellers and one for health workers . The first questionnaire attempted to capture information on dog population structure and husbandry as well as Knowledge Attitude Practice ( KAP ) of the interviewees regarding rabies . The question categories included: general information on the interviewee , questions related to dog husbandry and demography , contact of dogs with livestock and wildlife , questions related to bites ( person bitten , bite location , bite wound treatment ) , questions related to disease awareness/attitude/practice ( e . g . transmission to animal and people , symptoms and outcome in animals and people , source of disease knowledge , rabies therapy , bite wound therapy , rabies in livestock ) , and willingness to buy vaccine for dogs if available . The second questionnaire aimed at assessing the knowledge of rabies amongst health workers . All types of medical facilities were included: health posts ( N = 3 ) , Health Centers ( N = 1 ) , private clinics ( N = 3 ) and Merti hospital . Questions included: information on the interviewee and his work ( age , sex , religion , work place , professional background , years spend in the profession ) , the treatment of bite wounds ( kind of treatment , information on the bitten patient , bite location , number of bitten patient seen ) , whether or not he/she had rabies patients in the last 12 months and information on these patients , rabies knowledge ( transmission and symptoms in animals and people ) , rabies therapy , and suggested intervention for rabies prevention in people . The respondent was never offered answer options but had to talk freely on a question . The question was first asked in broader term to assess general knowledge and subsequently narrowed down to more detailed questions . This was to ensure no unwilling biased guiding from the interviewer . In addition , verbal ( oral evidence ) and recorded data and general information on rabies were collected from both , the Health Bureau and the Agricultural Office in Metahara as well as from all medical facilities . We performed a non-probability sampling . A list of all pastoralist settlements was obtained from the Woreda Agricultural Bureaus of both Woredas; inclusion criteria were logistic feasibility ( accessibility , security ) , proximity to the National Park , and willingness of pastoralists to participate in the interviews . All families of the chosen pastoral settlements and all medical staff present in the medical facility at the day of the interview were included in the study . For the urban study a house to house visit was made through the city and all families willing to answer the interview were included . All data was entered into Microsoft Access tables and analyzed descriptively using Stata software ( version 10 . 1 , StataCorp , Texas , USA ) . A scoring between 1 and 3 was given to KAP parameters . For instance if a respondent could describe all main symptoms of rabies in dogs he would get a score of 1 , if the respondent could not describe them but rather minor symptoms or remaining generally vague , he would get a score of 2 . A score of 3 was attributed to respondents who claimed not knowing any symptoms . Pearson’s chi square statistics test was used to compare group differences for categorical variables in the KAP analysis as well as dog ownership by areas in Metahara . Associations and assessment of determinants for KAP were considered statistically significant if p<0 . 05 . The study received institutional ethical clearance from the AHRI/ALERT Ethical Review Committee ( AAERC ) , number PO04/12 . Heads of the Woreda Agricultural and Health Bureau in Metahara and Awash town were informed and permitted the project . All interviewees gave verbal informed consent .
Dogs totaled 236 in the 471 urban household and 76 among the 49 pastoralist households . Considering a population of 25 , 670 and 11 , 000 households ( City Administration Metahara , annual census report 2011 ) , the crude extrapolation for human dog ratio in Metahara was 4 . 7:1 . Dogs were evenly represented throughout the city quarters ( p: 0 . 138 ) . Table 1 shows the demography and ownership of the dog population in Metahara and the pastoralist areas . Dog ownership was not influenced by religious background ( p: 0 . 12 ) ; 112/285 ( 39 . 3% ) and 77/209 ( 36 . 8% ) owned dog ( s ) among orthodox Christian and Muslim respondents respectively . Dog ownership differ statistically between urban and pastoral residents ( p<0 . 001 ) ( Table 1 ) . All dogs were kept as guard dogs and/or for livestock protection . None had received preventive rabies vaccination , was neutered/castrated , received veterinary care or had collars on . Puppies were never intentionally killed but kept or given away . They were all fed with left-overs from human consumption and/or left to roam free for food . The majority of the dogs were free-roaming during the day ( stated by N = 114/157 ( 72 . 6% ) urban and N = 33/37 ( 89 . 2% ) pastoralist households ) as well as during the night ( N = 90 ( 57 . 3% ) urban and N = 31/37 ( 83 . 8% ) pastoralist households ) . Dogs had regular contact with livestock in 94 . 3% and 94 . 6% of the urban and pastoralist households . Two urban and 14 ( 37 . 8% ) pastoralist households had their dogs going regularly into the nearby National Park . Direct regular contact between dogs and wildlife was observed by N = 88/157 ( 56% ) and N = 28/37 ( 75 . 7% ) urban and pastoralist respondents respectively . Main wildlife species reported were hyenas , jackals and less often rabbits , monkeys and leopards . None of the 471 urban and 9 ( 18 . 4% ) pastoralist respondents experienced bite wounds in his/her household in the last 5 years . Among the pastoralists 1 was a child under 5 years , 5 were children between 5 and 15 years and 3 were adults . Bites were recalled to be located in the foot and leg ( N = 7 ) and arm and hand ( N = 2 ) . Four out of the 9 bitten patients flushed the wound with water and soap while the others went to a medical facility in town . When bitten by an animal , 58 . 7% of the pastoralists , as opposed to 1% of the urban responded that they would go to traditional healers . The rest of the interviewed pastoralists either do wound cleaning themselves or go to medical facilities . Ninety-nine percent of the urban interviewees would go directly to a medical facility . Bite wound information ( e . g . patient age , sex , bite location , severity of bite ) was not recorded in any of the medical facilities assessed . Overall , interviewed staff recalled 7 bites that they personally treated in the last 12 months , 2 from hyenas and 5 from dogs . The bite locations from dogs , as recalled , were most often in the legs followed by feet , arms and rarely the face . Hyenas on the other hand bit in the face . Total number of bite patients recalled from all health facilities was estimated to be around 100 per year , of which approximately 75% were adults . None of the health posts did treat patients with bite wounds and referred them directly to Health Centers or private clinics . Reasons given were the unavailability of water , soap and other disinfectant and medical supply . In the other visited facilities , all treating staff used disinfectants such as providone iodine to treat bite wounds as well as antibiotics . Initial flushing with water and soap/detergent has been done by 2 out of 16 respondents . This study was done during a rabies outbreak , which started in the rural areas South-West of the Awash National Park in May 2012 . A strychnine poisoning campaign , in a bid to stop the disease spread in the urban areas , killed in a couple of days 275 dogs ( personal communication , Merti hospital ) . The pastoralist communities stated that many people had been bitten and had died of rabies during the outbreak . Exact numbers , however , were unavailable . The Agricultural Bureau stated ( recalls only as no written records ) that the outbreak lasted 2 months and that 11 livestock at least had died of rabies and 4 people were on PEP therapy . The Health Bureau officially recorded 2 patients undergoing PEP treatment and stated the outbreak did last 1 week only . In May only , Merti hospital on the other hand , treated 32 patients for bite wounds and 15 patients with PEP . Dogs that are biting people are not further diagnosed as whether they have rabies or not . The decision to start PEP treatment on a bitten patient relies solely on the decision of the health worker .
Control of zoonosis and prevention of human cases is usually most cost-efficient by controlling the disease in the animal reservoir [11] . Knowledge of the dog population structure , dynamics and ecology is , however , an essential pre-requisite to achieve proper required preventive vaccination coverage in the dog population ( critical vaccination coverage varies with animal density ) , not to waste scarce financial and logistic resources and avoid large-scale campaign failure [12 , 13] . In Ethiopia , published data on dog population number , structure and dynamics is lacking . In our study , 33% of the urban ( grossly 1 dog per 5 people ) and 75 . 5% of the pastoralist households kept dogs , regardless of religious background . Despite it being a small geographical study area , it could be observed that dog ownership and demography differed between the urban and the rural households but also between the pastoralist groups ( Afar versus Oromo ) . The observed sex-and age- based dog population imbalance is in line with previous findings [14 , 15] . Respondents in our study , with the exception of the Afar , generally preferred male dogs , a trend reflected in many developing countries [15] . Dogs were not intentionally killed , therefore a steady population increase would be assumed . However , in general , free roaming dogs have short life expectancy and 2/3 die in their first year [15] . In Kenya , a study showed that life expectancy for males was 3 . 5 years and for females 2 . 4 years [12] . In our study , veterinary care was inexistent and dog husbandry poor , factors likely to lead to high dog mortality . A common perception of all respondents was the large number of stray dogs , hence un-owned dogs . This study did not investigate the number of stray dogs but we need to keep in mind they also contribute to human rabies . Pastoralists stated that they were coming from town . However , the majority of owned dogs were free-roaming ( day and night time ) , and in constant quest for food since they receive only little left-overs at home . In addition , they were not wearing collars , thus easily mistaken for stray dogs . Studies have shown that in reality most dogs are owned [16 , 17] . This is an opportunity for health intervention campaigns since the fact that dogs have owners would facilitate regular vaccination campaigns in Metehara . To support this , the majority of the urban and pastoralist respondents stated that they would be willing to regularly pay for their dog’s vaccination if the vaccine was indeed available . Large scale lethal poisoning with strychnine is a common preventive measure undertaken in dogs in Ethiopia . Four out of 16 health workers reported killing of stray dog as a means of eliminating dogs . Mass dog elimination , however , besides being unethical and hazardous to the environment , has been shown to be counter-productive as it will not affect the dog population size and will not stop the spread of rabies as dogs will engage in compensatory breeding and migrate into newly vacated territories , thus facilitating disease transmission [18 , 19] . Morters et al ( 2013 ) recently showed as well that rabies transmission is not density-dependent [20] . Rabies is one of 20 reportable diseases in Ethiopia . Our study showed however , that rabies notification was poor compared to other diseases . The authors observed a severe discrepancy between the orally recalled rabies cases , the number of used PEP bottles logged into medical facility pharmacies , and the number of cases actually officially recorded and reported . Mainly adults were recalled to have been bitten . Generally , children are known to be more at risk for being bitten [9 , 21 , 22] . Our results raises serious concern as whether children were indeed less at risk or whether they were not brought to health facilities when bitten , thus showing that rabies in children is likely to be underreported and their treatments severely neglected in these communities . Bite locations were most often in the legs followed by feet , arms and rarely the face . This picture differed from a study in Tanzania where dogs bit mostly hands and face [4] . The location of the bite and the wound severity ( scratch versus deep skin penetration ) is likely to affect the outcome into clinical rabies [4] . Our study unfortunately could not collect enough details on bite wounds . Bite records ( location , severity , patient identification ) were not kept in any of the assessed health facilities . All information on bite wounds was collected only through verbal recollection of health workers . The authors assume however that the wounds must have been severe for patients to come to health facilities considering the cost and time involved for patients . Reliable record keeping of bite wounds has been shown to be a useful epidemiological tool as a proxy for human exposure and rabies incidence in animals [2 , 10 , 23 , 24] . In our case , calculation of rabies cases—and exposure- incidences in people and animals was also impossible due to the lack of accurate record keeping , severe underreporting , lack of population census amongst pastoralists and the dog population in particular . Pastoralists of the study area are likely not to go to a medical facility when bitten; reasons include distance to health facility/logistics , mistrust in the medical system , and poor knowledge of the disease fatal outcome . Pastoral respondent showed to be a determinant for poorer knowledge of rabies particularly for rabies in people . Despite the presence of an extension health worker system , over half ( 58 . 7% ) of the respondents said they would go to traditional healers if bitten . It is estimated that the majority of human rabies deaths occurs in rural rather than in urban areas [10 , 25] . The main constraint to human rabies prevention in the study area was the quantitative lack of PEP , and its immediate inaccessibility ( the vaccine is available in Adama , 134 km away from Metahara ) thus delaying severely the start of the prophylaxis . From the moment a person was bitten to the start of PEP injections , delays of several days ( up to 5 days ) were not unusual , particularly if patients came from rural areas . Wound cleaning was rarely performed as first aid , neither by the patients themselves nor the medical staff , which is a behavior described in other developing countries in Africa and Asia [22 , 26 , 27] . However , immediate flushing of a bite wound for 15 minutes with water and soap can be lifesaving , as the virus is mechanically removed from the site or is rendered unable to invade tissue [28] . Neglecting immediate wound flushing was shown to increase the risk of developing rabies by fivefold [21] . Health posts , that are often the first health facility patients , particularly pastoralists , would visit , were not offering this simple , cheap and important service . Hence , health posts and health extension workers visiting remote rural villages are in a unique position to initiate this procedure before transferring a patient to a larger facility that could start PEP , to educate people at large and pastoralists in particular about the importance of immediate wound cleaning , as well as the outcome of an untreated patient with rabies . However , this study also highlighted knowledge gaps about rabies among the health staff . Patients , particularly pastoralists who are not knowledgeable about the disease will rely on the health practitioner’s advice for adequate treatment , as whether PEP should be administered or not , as also seen in a study in Tanzania [22] . In our study , 25% of health staff did not know that livestock can transmit the disease to people . Hence , patients bitten by rabid livestock are likely not going to be treated for rabies . Also only 2 out of 16 health workers knew that bats can transmit rabies . The role of bats in carrying and transmitting rabies in Ethiopia is not known . However , its role should be mentioned in any future awareness programs . The fact that 25% of health workers thought that rabies is a bacterial disease raises the question as whether they thought PEP is an antibiotic and/or if they would treat the patient with antibiotics . Unfortunately we did not look further into this point . These results overall show an urgent need of improved training of health workers in rabies epidemiology and treatment . On the other hand , the lack of diagnosis in dogs implies that likely a high number of patients are unnecessary treated , putting a strain on the already difficult logistics , economics and availability of PEP . Deressa et al . ( 2010 ) showed that only 10% of the dogs that had bitten people and were brought to EPHI for quarantine were actually rabid [6] . There are currently , new rapid and simple rabies diagnostic tests on the market that can be used directly under field condition , such as the Anigen rabies test , an immunochromatographic test ( ICT ) giving results within minutes , not requiring expertise nor special facilities/equipment and thus helping in the decision of PET use [29] . This would require , however , close collaboration and communication between the Agriculture Office , the Health Bureau , the veterinarians , the clinicians , and the communities . In our study , we could observe that collaboration and communication between the different stakeholders was poor to non-existent . This study highlighted overall poor disease awareness amongst non-medical respondents although over 90% stated knowing the disease . A high percentage of respondents , particularly pastoralists ( 79 . 2% ) did not know how dogs acquire rabies . Water shortage , wind and consumption of rotten food were often given as reason . Over a third of respondents did not know the symptoms of rabies in dogs and in people . The fate of a rabid person was not known by the majority of the pastoralist respondents ( up to 80% ) . This highlights that the seriousness of this fatal disease in people was poorly known . In a country that lacks hard data on epidemiology , vaccines for dogs and PEP for humans , and that is tight by logistic and financial constraints to efficient medical rabies control , disease awareness takes an important place in disease prevention . Our study showed that family nucleus and the community particularly amongst pastoralists played a central role in passing down knowledge about the disease whereas school and media played a minor role . Urban respondents may have better access to media , but pastoralists gather a lot of information from the pastoral community during transhumance with their animals . The latter were also aware of the economic burden of the disease since livestock can be affected . The use of media can be instrumental to increase rabies awareness in a population , similarly as was- and is being done with other diseases such as tuberculosis in Ethiopia; it targets large audience and can also help promoting responsible dog ownership . The overall rabies knowledge amongst health practitioners was variable and showed some knowledge and diagnosis/intervention gaps , particularly among staff from private clinics . Rabies is still sometimes misdiagnosed for malaria . This study was the first in its kind to look at a rabies situation in a defined study site from a holistic approach including as well the animal as human side , urban and pastoralist respondents , medical and non- medical people . The in-depth interviews as well as the analysis of official records gave a valuable insight in the many existing gaps on rabies knowledge , prevention and health delivery . These information are valuable before embarking in a control and or awareness program in the study area . The limitations of the study lay in the non-probability sampling method used and the small sample size as well as the limited information for KAP determinant analysis . Also all data on bites and rabies patients ( animal and human ) were solely based on recalls and not on official records calling for inevitable biases . Rabies is a 100% preventable disease but numerous challenges and constraints in Africa render its control and elimination difficult , hence relegating it to the neglected diseases [2 , 30 , 31] . In this study , it was observed that preventive dog vaccination was non-existent due to lack of vaccine availability at the time . In such a situation , disease awareness takes even more so , an important place in disease prevention as well in urban as in pastoral communities . However , this study highlighted overall poor disease knowledge and gaps in recognizing and treating rabies in people , the likelihood of severe under-reporting of cases , poor medical facility registries regarding bite wounds and rabies and a lack of collaboration between the animal and public health sectors . These factors are likely to hamper any future efficient rabies control campaign in the study area . | Rabies is a fatal viral disease of animals and people . People usually get infected via bites from an infected animal ( e . g . dog ) . Post exposure prophylaxis ( PEP ) has to be initiated immediately after bite wounds of suspected rabid animals in order to avoid fatalities . The situation of rabies is poorly known in Ethiopia , particularly in the pastoral context . We conducted questionnaire surveys in urban , pastoral and medical health worker communities around Awash National Park ( Ethiopia ) in order to capture information on dog demography , bite wound prevalence and management , potential risk factors of disease transmission and knowledge- attitude- practice ( KAP ) towards rabies among these communities . Disease knowledge was generally poor . Dog demography varied depending on the community which would affect control strategies . Health facilities did not keep bite records and there was poor recording and reporting of rabies cases . Delivery of PEP was inadequate . Communication and collaboration between the public and animal health sector was poor to inexistent regarding reporting and control of rabies cases . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"livestock",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"demography",
"pathogens",
"tropical",
"diseases",
"microbiology",
"vertebrates",
"geographical",
"locations",
"dogs",
"animals",
"mammals",
"health",
"care",
"viruses",
... | 2016 | Dog Demography, Animal Bite Management and Rabies Knowledge-Attitude and Practices in the Awash Basin, Eastern Ethiopia |
Science revolves around the best way of conducting an experiment to obtain insightful results . Experiments with maximal information content can be found by computational experimental design ( ED ) strategies that identify optimal conditions under which to perform the experiment . Several criteria have been proposed to measure the information content , each emphasizing different aspects of the design goal , i . e . , reduction of uncertainty . Where experiments are complex or expensive , second sight is at the budget governing the achievable amount of information . In this context , the design objectives cost and information gain are often incommensurable , though dependent . By casting the ED task into a multiple-criteria optimization problem , a set of trade-off designs is derived that approximates the Pareto-frontier which is instrumental for exploring preferable designs . In this work , we present a computational methodology for multiple-criteria ED of information-rich experiments that accounts for virtually any set of design criteria . The methodology is implemented for the case of 13C metabolic flux analysis ( MFA ) , which is arguably the most expensive type among the ‘omics’ technologies , featuring dozens of design parameters ( tracer composition , analytical platform , measurement selection etc . ) . Supported by an innovative visualization scheme , we demonstrate with two realistic showcases that the use of multiple criteria reveals deep insights into the conflicting interplay between information carriers and cost factors that are not amendable to single-objective ED . For instance , tandem mass spectrometry turns out as best-in-class with respect to information gain , while it delivers this information quality cheaper than the other , routinely applied analytical technologies . Therewith , our Pareto approach to ED offers the investigator great flexibilities in the conception phase of a study to balance costs and benefits .
That said , finding an informative , yet cost-efficient experimental setting out of the space of alternate designs is a nontrivial task: First , the space of options may be extensively large and secondly , several related , but potentially conflicting design objectives need to be optimized simultaneously . Here , a common solution concept is to optimize a weighted sum of the single criteria [6 , 15] . However , in real-world scenarios the objectives are hardly expressible in the same “currency” and appropriate weights to translate between them are not known before the experiment . Consequently , in scenarios where the ability to explore the whole space of design alternatives should be maintained , a fixed-weight solution cannot be utilized [16] . To overcome the limitations of weighted-sum single-objective approaches , the ED task can be casted into a multi-objective optimization ( MOO ) formulation [17] . Multi-objective ( MO ) ED comes with an important conceptual difference , compared to single-objective ED: When objectives are conflicting , instead of one specific solution , a whole set of—in terms of the objectives—equally good , compromise EDs is obtained where none of the designs is better than the others in terms of all criteria . These compromise EDs , denoted Pareto-optimal EDs , determine the Pareto front in the objective space ( Fig 1 ) . When the objectives are not in competition , a characteristic that cannot be known for real-world problems a priori , the MO-ED task degenerates to an ordinary ED problem . The trade-off decision on the experiment is then made after examining the Pareto front and inspecting the related Pareto-optimal designs where ( expert or newly available ) information or preferences can be considered in addition . However , to keep track of more than a few relations is not only intrinsically challenging , it also calls for domain-specific solutions to interrogate the high-dimensional Pareto-optimal results and to support exploration and interpretation processes . We present a universal computational methodology for the design of informative , yet cost-effective experiments . Our approach simultaneously optimizes many , potentially contradicting information and cost metrics rather than a single one , therewith generalizing traditional ED frameworks basing on the optimization of a single information criterion . To provide a visual means for result exploration of Pareto-optimal EDs in potentially high-dimensional design and objective spaces , we suggest a flexible solution using chord diagrams . To exemplify our information-economic Pareto approach , the MO-ED framework is implemented for 13C metabolic flux analysis ( 13C MFA ) , which provides a computationally challenging test bed owing to its enormous design space and diverse cost factors . Equipped with the computational tools , the questions raised above were addressed by a comprehensive investigation featuring the fungus Penicillium chrysogenum . In particular , two different scenarios were studied . First , all analytical platforms commonly applied for 13C MFA were profiled with respect to their information-economic characteristics , using a single information criterion . The study revealed that the specific measurement information delivered by tandem mass spectrometry ( MS/MS ) cannot only increase flux information , but also enabled cost savings by the choice of cheaper tracers , emphasizing the potential of our approach . In the second scenario we investigated whether including more than one information criterion could provide a benefit for the decision process . Indeed , for the P . chrysogenum showcase a variety of additional Pareto-optimal designs were offered , unlocking informed decision making . In particular for , but not limited to , the domain of 13C MFA our findings show that the use of several criteria balances shortcomings of conventional ED strategies and offers additional flexibilities for the experimenter , thus providing a methodology of direct practical relevance .
Planning cost-efficient , informative experiments requires finding the “best” experimental-analytical trade-offs that , on the one hand , maximize the information gain , possibly in view of different information facets , while , on the other hand , keep the associated costs to a minimum . Consequently , two formal ingredients are needed: Employing these criteria in the selection procedure of the ED formally amounts to a multi-objective optimization ( MOO ) problem: maxα∈ΩΦ ( α , θ ) s . t . g ( α , θ ) ≥0h ( α , θ ) =0l≤α≤u ( 1 ) where the objective vector Φ is composed of a set of information and ( negated ) cost criteria . The objective vector is a function of the design variables α , selected from the space Ω of feasible designs . Remaining design parameters , which are constant , are collected in the vector θ . Furthermore , the bounded design variables may be subject to inequality and equality constraints . Solving the MOO problem ( 1 ) means to find the set of all trade-off design solutions α* that minimize the objectives in Φ without being dominated by another solution [18] . Here , a specific design α1 dominates another one α2 , if ( and only if ) α1 is at least as good as α2 in all objectives and better with respect to at least one , formally expressed by Φi ( α1 ) ≤ Φi ( α2 ) , ∀i and ∃j: Φj ( α1 ) < Φj ( α2 ) ( gray shaded area in Fig 1 ) . The set of all non-dominated solutions is referred to as Pareto-optimal design set , and the corresponding achievable objective values are called Pareto front . Clearly , the concrete formulation of the MOO problem depends on the particular application case , namely the underlying system model and the peculiarities of the experimental settings . In this work , we selected a use-case from the domain of 13C metabolic flux analysis ( MFA ) , which is arguably the most expensive type of ‘omics’ technology , featuring dozens of design variables . Before introducing the information and intricate cost models as well as the analysis of Pareto-designs in high dimensions , the essential background to the application field is provided , in particular the formulation of the system model . Intracellular reaction rates ( fluxes ) describe the trafficking of metabolites which emerges as the final outcome of all catalytic and regulatory processes acting within living cells [19] . Here , the reactions within a biochemical network are characterized by a pair of flux values , net and exchange fluxes [20] , to express the respective proportions of material transported between the reaction’s educts and products . At steady-state , the in- and outflows of each intermediate metabolite are assumed to be constant and mass balanced , yielding the stoichiometric equation system for the flux vector v: S⋅v=b , Cieq⋅v≤cieq ( 2 ) with the stoichiometric matrix S and the vector b containing the extracellular rates ( substrate uptake , product formation or effluxes leading to biomass accumulation ) , accessible through extracellular concentration profiles and biomass quantification . In addition , the fluxes may be constrained in their allowable value range owing to physiological knowledge . Since metabolic networks contain parallel paths and cycles , fluxes are not uniquely determined by Eq ( 2 ) , at least not without additional assumptions . The indeterminacy implies that the flux vector v can be parametrized through a certain ( non-unique ) sub-set of fluxes , the so called free fluxes vfree [20] . The dimensionality of the vector vfree , i . e . , dim ( v ) − rank ( S ) , is referred to as degrees of freedom ( DoF ) . To resolve the DoFs , carbon labeling experiments ( CLEs ) are conducted . In a CLE , isotopically labeled carbon sources , like [1-13C] glucose enriched with a 13C isotope at the first position of the carbon backbone , are fed to the cells . The labeled substrate is taken up by the cells and distributed through the metabolic pathways to all intracellular metabolites , where it gives rise to characteristic labeling enrichment patterns . Thus , the labeling patterns are the convoluted result of the routes , the 13C labeled substrate takes , as well as the underlying metabolic fluxes . In isotopic steady-state 13C MFA , as used in this work , intracellular free fluxes are inferred from the equilibrated labeling patterns and external rate measurements by means of a computational flux fitting procedure that minimizes the least-squares error between observed measurements and those that are simulated by a computational network model [21] . For the model , carbon atom transitions have to be specified for each reaction step describing the fate of each carbon atom from the reactions’ educt to its corresponding product . Mass balancing of the intracellular isotopic forms then yields a high-dimensional nonlinear algebraic equation system that relates the steady-state labeling state x , the administered labeled tracer mixture xinp , and the free fluxes vfree [20] . Given vfree and xinp , the vector of steady-state labeling states x ( represented as isotopomers , cumomers , EMUs , or similar [20 , 22 , 23] ) is uniquely determined by [24]: x=x ( vfree , xinp ) ( 3 ) Note that CLEs that only differ in the tracer mixture are covered by the same formalism through duplication of the network model and equating the free fluxes . The full system-wide labeling state x is not accessible by any current measurement technology . What can be observed are linear combinations of ( relative ) abundances for some of the intracellular metabolites , such as mass isotopomer distributions or positional enrichments . Fig 2 shows characteristic sets of observations , henceforth denoted measurement groups , for the analytical platforms employed in the field of 13C MFA . All measurement groups available for an analytical device are organized in the measurement matrix Mmeasdev that , owing to Eq ( 3 ) , allows to simulate the measurement vector η: η=Mmeasdev⋅x ( vfree , xinp ) ( 4 ) which mimics the real measurements up to normalization to percentage scale [25] . Examples for measurement matrices are given in S1 Text . Real measurements are unavoidably affected by noise . In the context of 13C MFA , measurement noise is assumed to be independent , unbiased , additive , and normally distributed with expectation 0 and standard deviation σmeasdev , as represented by the measurement covariance matrix Σmeasdev [26]: Σmeasdev=diag ( σmeasdev ) ( 5 ) Since in the CLE’s planning phase real measurements are absent , from which measurement standard deviations σmeasdev can be derived , measurement error models need to be formulated , relating the measurements with their associated errors . For labeling measurements empirical rule-of-thumb approximations of the measurement precision have been derived for specific analytical setups . For instance , Crown et al . propose a precision of 0 . 4 mol% for their GC-MS setup targeting proteinogenic amino acids [27] . In general , labeling errors depend on the measurement technique , the instrument , the analytic protocols , they can vary between organisms , analytes and the degree of label incorporation [28] . To arrive at realistic error approximations that allow for a fair comparison of the analytical platforms , measurements and their standard deviations were collected from published studies featuring different organisms , platforms and various labeling contents . In total , more than 900 data points for six analytical platforms , namely GC-MS , LC-MS , LC-MS/MS , 13C-NMR , 1H-NMR , and GC-C-IRMS were extracted ( S1 Text ) . For all analytical platforms , similar to the approach by Dauner et al . for 13C-NMR [29] , a regression line was fitted to the respective data set , yielding device-specific linear measurement error models . These analytics-related error models provide empirical standard deviations σmeasdev for any given measured vector η: σmeasdev ( nrep , measdev ) =a ( nrep , measdev ) ⋅ ( b1dev⋅η+b2dev ) ( 6 ) where b1dev , b2dev are the device-specific regression coefficients ( S1 Text ) . Generally , by increasing the number of repetitions nrep , measdev ( i . e . , technical replicates ) , the error estimates are believed to become more reliable . This is accounted for in the error models ( 6 ) by a scaling factor ( a ) which tends to 1 for the case of many repetitions ( see S2 Appendix for details ) . Several statistical approaches have been developed to predict the approximate amount of information to be derived from the planned CLE or CLE series . When some pre-knowledge on the expected flux map v^free is available ( which we assume in this work ) , a widely adopted local information measure is the Fisher information matrix ( FIM ) [9 , 13 , 26]: FIM= ( ∂η∂vfree|v^free ) T⋅Σmeasdev⋅∂η∂vfree|v^free ( 7 ) whose inversion yields the flux covariance matrix: Cov ( v^free , α ) =FIM−1 ( 8 ) which depends on the design point ( v^free ) and the design parameters ( α ) . As a precondition for stable numeric calculation of the flux covariance matrix , the FIM needs to fulfill two conditions [30]: First , its minimal singular value λmin ( FIM ) needs to be larger than a threshold: λmin ( FIM ) >τ1>0 ( 9 ) and secondly , its condition number has to be bounded: cond ( FIM ) <τ2<∞ ( 10 ) The fulfillment of the conditions ( 9 ) and ( 10 ) implies that the standard deviations of the free fluxes—as represented by the main diagonal of the covariance matrix—remain bounded and , thus , the flux vector is said to be statistically identifiable . First , it should be remarked , that this is a slightly stronger variant of practical identifiability as defined by Raue et al . in [31] and secondly , statistically identifiable fluxes are per se structurally identifiable [32] . If either one of the conditions ( 9 ) and ( 10 ) is violated , fluxes causing the violation have to be excluded from the FIM . Eventually , this leads to models that vary in terms of their DoFs , a fact which needs careful treatment when comparing different experimental setups with respect to their information content . For quantifying the information content of a CLE several information quality criteria have been proposed that aggregate the covariance matrix to a single number [9 , 12 , 13] . The most prominent ones are the determinant ( D ) , the average-variance ( A ) , and eigenvalue ( E ) criteria . Ultimately , all these criteria provide a means for the shape of the confidence ellipsoid in the vicinity of a given design point ( v^free in our case ) , each emphasizing particular geometrical aspects [12] ( Fig 3 ) . For example , the D-criterion strives to minimize the volume of the confidence ellipsoid ( or the geometric mean of the flux confidence intervals ) : ΦD , p=det ( Cov ) 2⋅p ( 11 ) with p the dimension of Cov ( with arguments omitted for brevity ) while the A-criterion aims to minimize the diagonal of the smallest bounding box that contains the confidence ellipsoid ( or the arithmetic mean of the flux confidence intervals ) : ΦA , p=trace ( Cov ) /p ( 12 ) Hence , the A-criterion is expected to provide designs that are more robust against flux correlations than those based on the D-criterion . Notice that the explicit consideration of the dimension p of the covariance matrix in the formulation of criteria ( 11 ) and ( 12 ) intends to make the criterion values comparable for models differing in the number of free fluxes . In contrast , the E-criterion: ΦE=λmax ( Cov ) /λmin ( Cov ) ( 13 ) constitutes a dimension independent measure that strives to improve worst case designs by preventing the Fisher matrix from becoming singular . Besides these quantitative information measures , an obvious quality criterion is the number of free fluxes that can be statistically identified by the ED setting , expressed by: ΦDoF=dim ( Cov ) ( 14 ) With these information measures at hand , the information gain of a 13C MFA study can be influenced by the targeted selection of the input mixture compositions ( xinp ) , the measured groups observable by the analytical device ( Mmeasdev ) , as well as the corresponding measurement errors ( σmeasdev ( nrep , measdev ) ) , i . e . , the interval in which the true measurements are believed to lie in to a certain probability , triggered by the number of repeats . The choice of isotopically labeled substrate species , either in pure form or in a mixture , dictates the emerging labeling states of the observable metabolites and therefore significantly impacts flux information [25 , 33] . Several recent field studies yielded information-optimal tracers in a variety of biological systems and give evidence for a high diversity of flux standard deviations depending on the substrate or substrate mixture . For instance , Walther et al . showed that [1 , 2-13C]-labeled glucose and mixtures of [3-13C]- and [3 , 4-13C]-glucose increase statistical identifiability when used with fully labeled glutamate for lung cell carcinoma [34] . Crown et al . identified [3 , 4-13C]- and [2 , 3 , 4 , 5 , 6-13C]-labeled glucose to be favorable for elucidating reaction rates in the oxidative pentose phosphate pathway ( PPP ) and pyruvate carboxylase flux , respectively , based on a small scale network with two free fluxes [35] . Later on , the same group determined [1 , 2-13C]- , [5 , 6-13C]- , and [1 , 6-13C]-labeled glucose as best single tracers for Escherichia coli wild type [36] . A study of Metallo et al . suggested [1 , 2-13C]-labeled glucose to be the optimal commercial tracer for most fluxes in the PPP and glycolysis in lung carcinoma cell lines while uniformly labeled glutamine provided optimal results for tricarboxylic acid cycle ( TCA ) fluxes [37] . In theoretical studies , [3 , 4 , 5 , 6-13C]-glucose and [2 , 3 , 4 , 5 , 6-13C]-glucose resulted to have to best information yield in plants and mammalian cells , respectively [38 , 39] . Araúzo-Bravo et al . calculated mixtures of 70% unlabeled , 10% U-13C- and 20% [1 , 2-13C]-labeled glucose to be optimal for flux determination in the cyanobacterium Synechocystis sp . PCC6802 [40] . Schellenberger et al . applied a Monte Carlo sampling technique for experimental tracer design to a large-scale Escherichia coli network and found positional [1-13C] or [6-13C] labeled glucoses to be superior over a commonly used mixture of 20% uniform and 80% unlabeled glucose [41] . Here , unusual multi-positional labeling , in particular [5 , 6-13C]- , [1 , 2 , 5-13C]- , [1 , 2-13C]- , [1 , 2 , 3-13C]- , and [2 , 3-13C]-glucose , resulted in a higher identifiability than single positional labeling . Nonetheless , no single tracer has been found to outperform all others , an observation which was experimentally confirmed by Crown et al . comparing the outcome of 14 CLEs in Escherichia coli [27] . Importantly , the studies also disclosed a high redundancy in the measurement data , meaning that not all observations effectively contribute to the information gain , although they come at a certain cost . One option to raise flux identifiability that recently has become compelling through advances in lab standardization and miniaturization [42] , is the conduction of multiple independent , so called parallel CLEs under identical conditions , each with a different tracer [43] ( and references therein ) . Concurrent fitting of all labeling patterns with a single model obviously increases the measurement-to-flux ratio but , at the same time , also the measurement redundancies . Still , in these and other theoretical and practical studies a part of the fluxes remained non-identifiable [27 , 44] . Interestingly , a study of Bouvin et al . [45] exemplified , also using a MO-ED approach , that it is indeed possible to find CLEs with comparable information content , but considerably different tracer costs . In contrast to the work on tracer design , measurement setups have not yet been the target of ED in the field of 13C MFA . The primary analytical methods that are employed are NMR and MS . For both , analytical devices differ not only in the principally observable metabolite/isotopomer spectrum , achievable fragmentation patterns ( Fig 2 ) and the measurement accuracy and sensitivity , but also in terms of analysis speed/throughput , and purchase/maintenance costs ( S1 and S2 Text ) . Since comparative investigations on the inter-platform information content of CLEs for 13C MFA are scarce , in essence , it is still an open question which analytical platform delivers maximal flux information and what the information benefit of multiple-device applications is compared to single-device usage . For considering economic aspects , the cost contribution of the isotopically labeled substrates , the experimental setup and the analytical technologies are to be specified . Additionally , not only the measurement time on the device , but also spectra evaluation and proofreading processes , possibly with the need for manual post-correction , contribute to the workload . Consequently , such direct and hidden factors play a part in the overall CLE costs . Till now , if at all , only 13C labeled tracers have been considered in CLE costs examinations while further experimental-analytical efforts were neglected so far ( see e . g . [45] ) , meaning that a fine-grained cost function which relates all cost factors to the design parameters has to be set up . The overall cost function of a 13C MFA study is composed of three parts , the experimental , the analytical , and the modeling part . However , the modeling costs such as setting up an adequate model , working through the 13C MFA workflow , calculating and interpreting results etc . , heavily depend on the use case and are therefore not considered in the following . Together , the information and cost criteria Eqs ( 11 ) – ( 14 ) , ( 18 ) make up the set of goal functions out of which the objective vector Φ of the MO-ED problem Eq ( 1 ) is composed . The design vector α is subject to inequality and equality constraints such as the invertibility conditions on the Fisher matrix ( 9 ) and ( 10 ) , constraints for weights , as well as constraints imposed by reasonable practical resource considerations , e . g . , a maximum number of replicates . Since exact handling of integer-valued replicate numbers would result in NP-complete mixed integer nonlinear optimization problems [46] , the optimization problem is relaxed by allowing the replicates to take non-integer values . The solution for the relaxed problem is then “rounded” to integers . The full formulation of the MO-ED problem is given in S2 Text . Solving Eq ( 1 ) means to numerically approximate the ( potentially infinite ) design set α* by an ensemble of Pareto-optimal results [47 , 48] , optimally uniformly distributed covering the whole Pareto front . Particularly successful among these algorithms with respect to convergence and extensity of Pareto front approximation are those based on Particle Swarm Optimization ( PSO ) with update mechanisms to ensure that the solution ensemble is well-dispersed over the front [49] . For this work , the jMetal ( Metaheuristic Algorithms in Java ) library , a suite of state-of-the-art MO algorithms is utilized [50] . jMetal is linked to the high-performance 13C MFA simulator 13CFLUX2 [51] via a Java Native Interface ( JNI ) that enables jMetal to call 13CFLUX2 methods . While 13CFLUX2 is used to evaluate the objectives and takes care of the feasibility of the design parameters , the solution of the MO problem is steered by jMetal routines ( Fig 4 ) . Initially , all experimental , analytical and simulation settings as well as the network model ( incl . free flux set and flux values ) , measurement error models and input species with their respective costs are specified . Depending on the measurement selection proposed by jMetal , the measurement error model is evaluated for the suggested mixture composition in 13CFLUX2 while also taking the number of replicates into account . In turn , statistical flux identifiability is tested and , if one of the invertibility criteria fails , the free flux set is adapted in an iterative procedure: Non-identifiable fluxes are eliminated one-at-a-time by constraining them to their nominal values beginning with the worst determined one , eventually providing the effective number of statistically identifiable fluxes , i . e , p . From the resulting covariance matrix the local information measures ΦD , p etc . are calculated . Furthermore , the expected CLE costs ΦCostsdev are evaluated according to the cost model , given the experimental specification . The objective values are then passed to jMetal , calling the SMPSO algorithm ( the rationale for the choice of SMPSO and its parameters is given in S2 Text ) . Starting with an initial population created randomly , the swarm is evolved driven by polynomial mutation rules that trigger the choice of the design parameters . In this way , new swarm candidates are proposed out of which Pareto-optimal solutions are selected . The best Pareto solutions are stored in an archive where for each iteration the crowding distance is used to decide which swarm individuals remain in the archive to achieve maximal coverage of the designs . For the newly generated swarm members , measurement values are predicted in silico according to the 13C MFA model using 13CFLUX2 and the corresponding standard deviations are derived from the associated error models . This process cycle is restarted with the next generation of particles until the stopping criterion ( i . e . , maximum number of generations ) is reached . Finally , the archive containing the ( best known ) Pareto-optimal ensemble is returned and subjected to visual analysis . Having the Pareto front approximation at hand , the final step of ED involves decision making on the next experiment . In the context of 13C MFA , decision making means to find the most suited experimental-analytical setup out of the range of analytical platforms , input mixture compositions , sets of observable metabolites and replicate numbers . These quantities have different contextual meanings , scales and importance , in the sense of affecting the objective values . Hence , the visual interpretation of MO-ED results faces two challenges: To tackle these challenges a tailor-made visual interpretation workflow was created ( Fig 5 ) . The workflow is composed of three modules , applying different information visualization techniques that ( a ) allow for visual assessment of the Pareto front , ( b ) relate the objective with the most important elements of the design space , and ( c ) compress presentation of the less important design elements .
P . chrysogenum is the primary microbial cell factory for the production of penicillin G and V . Although metabolic engineering strategies have led to strongly improved production efficiencies , the yields of P . chrysogenum are still far below the theoretical maximum [55] . In this situation , 13C MFA is a powerful technique to detect pathway bottlenecks and to guide metabolic engineering efforts . Therefore , this case study explores the Pareto-optimal experimental design spaces in an industrially relevant setting . With the 13C MFA P . chrysogenum model of at hand , two scenarios differing in the composition of the design objectives were studied . The goal of this first scenario was to profile the analytical platforms according to their information-cost trade-offs and to explore the underlying Pareto-optimal designs . To this end , three objectives were considered , two information criteria and the cost criterion: Thus , the objective vector is represented by: Φ= ( ΦDoFΦD , p−ΦCostsdev ) T ( 19 ) Due to the number of objectives involved , the MO-ED problem ( 1 ) with ( 19 ) is hitherto denoted 3D-MO-ED task . Pareto-optimal solutions were calculated and objective values were recorded along with the identifiers of the statistically ( non- ) identifiable fluxes as well as the number of replicates for each single measurement group . Solutions obtained with models of maximal dimension , p = 21 , are discussed in the following ( the complete sets of Pareto sets and fronts are provided in S1 Data ) . The previous study revealed detailed insights into trade-off CLE designs for P . chrysogenum that relied on the commonly used D-criterion as quantitative information measure . With our second scenario we aimed to study the impact of including additional information criteria on the MO-EDs . To this end , the objective vector is extended by A- and E-information criteria: Φ= ( ΦDoFΦD , pΦA , pΦE−ΦCostsdev ) T ( 20 ) The MO-ED scenario ( 1 ) with ( 20 ) , henceforth referred to as 5D-MO-ED , was performed along the same lines as the 3D-MO-ED study . In the following , selected results are presented and related to the outcomes of the previous ED results . Detailed results are given in the S5 Text .
MO-ED enables the determination of design ensembles that seek to balance mutually exclusive information- and cost-objectives . Understanding the characteristics of the Pareto sets and the relationships between the different objectives is invaluable to guide the decision process on how to perform the next experiment . However , the sheer size of the design space and the many and various properties of the design parameters pose new challenges for the exploration procedure . First , searching for Pareto-optimal sets exhaustively over the whole , high-dimensional design space is compute-intensive and requires the efficient evaluation of the system model . Here , this challenge was tackled by connecting the high-performance simulator 13CFLUX2 with the optimization library jMetal . Second , a tailored visual analysis workflow was invented that tracks down Pareto-optimal designs thereby relating tracers , measurement groups , replicate numbers , costs , and information measures by means of graphical representations , starting from most relevant ( input species ) to less informative features ( replicates ) . This workflow aids the scientist to weigh the insights against the costs and , thus , guides decision making . ED studies were performed for a reference flux distribution representing prior information about the expected fluxes . It is , however , likely that the actual flux distribution under which a CLE is conducted differs from the assumed one . Because the information criteria used in this work rely on local statistical measures , actual Pareto-optimal designs may be widely different from the suggested ones . To investigate the robustness of the 3D-MO-ED Pareto designs in terms of information gain with respect to deviations from the reference flux values , for each platform 10 , 000 flux distributions were randomly sampled in the bounding box of the corresponding confidence ellipsoids . For the in each case most informative 3D-MO design setting , the D-information criteria values were calculated ( for instance in case of LC-MS/MS for pure [1 , 2-13C]-glucose ) . In all cases , the average information value of Pareto-optimal results remained in the upper third suggesting that the determined MO-ED designs are reasonably robust ( S4 Text and S3 File ) . The set of Pareto-optimal labeled tracers for CLEs was found to be remarkably similar across all investigated platforms and platform combinations , e . g . , [3-13C]- , [4-13C]- , and [5-13C]-glucose rarely contribute to the designs . However , the quantitative composition of the Pareto-optimal tracers varies widely . Often used , inexpensive substrate mixtures consisting of [1-13C]- , [U-13C]- , and [12C]-glucoses provide moderate statistical identifiability for LC-MS/MS . Several former single-objective ED studies found [1 , 2-13C]-glucose to be particularly informative ( cf . Sec ED approaches in 13C MFA revisited ) . Although this tracer is more expensive than standard mixtures , our results show that [1 , 2-13C]-glucose is beneficial to achieve a higher degree of flux confidence across all studied platforms . Our study also reveals that the use of other , more expensive substrate species such as [1 , 6-13C]-glucose , which seldom have been suggested by conventional ED studies before , is mandatory when a high degree of flux confidence is needed ( as measured by the D-criterion ) , especially for GC-MS , LC-MS , 13C-NMR , GC-MS/LC-MS , 1H-NMR/13C-NMR . These findings yield a generalized view on existing work that focuses on single objective ED aspects . The study delivers detailed experimental and analytical cost reports for all analytical platforms . 3D-MO-ED results demonstrate , not surprisingly , that the substrate species of choice are the main contributor of the costs . On the other hand , often almost the complete available measurement spectrum , including the maximal number of replicates , contributed to the Pareto-optimal designs arguably because , compared to the substrates , additional measurements come almost for free while they always increase the statistical information gain of the CLE . Only for inexpensive substrate mixtures some measurement groups did not contribute to the designs , most likely due to their redundancy . Hence , savings in analytical costs are possible but only achievable to a lesser extent . Interestingly , robust A-optimal designs emerge to be the most expensive ones across all investigated analytical platforms . Traditional 1D 13C MFA experimental planning techniques as first proposed by Möllney et al . [25] capture the “value” of a CLE in a single scalar measure of information content which is of limited value . By generalizing the 1D formulation to nD , strikingly , our study demonstrated that the latter gives a much more comprehensive view on Pareto-optimal designs , therewith opening up new possibilities for the experimenter in the planning phase of an experiment . The competition between single criteria is reflected in diverse , partly orthogonal designs . For instance , A-/D- and E-/D-optimal designs but no A/E-optimal designs co-exist for LC-MS/MS . The ability to account for a range of information criteria allows to pro-actively countering undesired side effects caused by ( a priori unknown ) flux correlations and , thus , could increase the design’s reliability . Importantly , these results were found to be specific to the analytical platform under consideration . Clearly , this wealth of additional insights comes at a computational cost . Here , the generalized ED framework has taken advantage of recent algorithmic advances in 13C MFA [23 , 51] , which paved the way for complex field studies such as reported in this work . Statistical flux identifiability with a comprehensive metabolic network of P . chrysogenum varies strongly among the measurement techniques . Even acknowledging long analysis times and high equipment costs , LC-MS/MS provides EDs with 50% less costs than other devices due to the use of cheaper input substrates . Simultaneously , LC-MS/MS yields up to ~300% higher information values as compared to the other techniques . Remarkably , the first scenario showed that the spread of Pareto-optimal designs has the highest coverage for LC-MS/MS , thus offering more options to the investigator than GC-MS , LC-MS , and 13C-NMR . Eventually , the goal of 13C MFA is to measure metabolic fluxes with the highest possible precision . Hence , the question arises whether the extra effort of MO-ED pays off in practice . A use case scenario may be as follows: An ED is desired with overall equally well determined fluxes and as little flux correlations as possible . Analyzing the 5D-MO-ED results with a high E-criterion value , corresponding designs may yield large flux confidence regions . In contrast , A-optimal designs indeed deliver superior designs in the sense of overall flux precision . However , by inspecting the costs associated with A-optimal designs , it becomes apparent that CLEs with A-criterion values overrun the budget . In this situation , alternate A-optimal designs satisfying certain cost constraints can be readily identified and even further ranked by their E- and/or D-criteria values . Having localized the desired Pareto-set ( s ) , the associated designs can be further explored in depth providing detailed specifications of substrate composition and the measurement setup . Operated in that way , we believe MO-ED to become a useful new tool for prospective and rational planning of experiments under full cost control . Besides deploying the framework to further application fields , there are several options to follow up this work . Technically , dependencies of the MO-EDs on the local design points should be diminished , e . g . , by incorporation of global sensitivity analysis [60] or other more advanced design techniques [61] into the framework , to handle scenarios when pre-knowledge on the model parameters is absent . Practically , introducing interactive features to the visual analysis such as browsing , querying , filtering , or sorting could boost the quick understanding relationships within and in-between Pareto sets . | Designing experiments is obligatory in the biosciences to valorize their scientific outcome . When the experiments are expensive , unfortunately , in practice often the costs emerge to be showstoppers . In this situation the question arises: How to get the most out of the experiment for your invest in terms of time and money ? We approach this question by formulating the design task as a multiple-criteria optimization problem . Its solution produces a set of Pareto-optimal design proposals that feature the trade-off between information gain , as measured by different metrics , and the costs . Then , exploration of the design proposals allows us to make the best decision on information-economic experiments under given circumstances . Implemented in the field of isotope-based metabolic flux analysis , practical application of the Pareto approach provides detailed insight into the tight interplay of plenty of information carriers and cost factors . Supported by an innovative tailored visual representation scheme , the investigator is enabled to explore the options before conducting the experiment . With a practical showcase at hand , our computational study highlights the benefits of incorporating multiple information criteria apart from the costs , balancing the shortcomings of conventional single-objective experimental design strategies . | [
"Abstract",
"Introduction",
"Methods",
"and",
"models",
"Results",
"Discussion"
] | [
"chemical",
"compounds",
"liquid",
"chromatography",
"experimental",
"design",
"social",
"sciences",
"random",
"variables",
"carbohydrates",
"organic",
"compounds",
"glucose",
"covariance",
"research",
"design",
"mathematics",
"metabolites",
"network",
"analysis",
"research... | 2018 | A Pareto approach to resolve the conflict between information gain and experimental costs: Multiple-criteria design of carbon labeling experiments |
The formation of proteins into stable protein complexes plays a fundamental role in the operation of the cell . The study of the degree of evolutionary conservation of protein complexes between species and the evolution of protein-protein interactions has been hampered by lack of comprehensive coverage of the high-throughput ( HTP ) technologies that measure the interactome . We show that new high-throughput datasets on protein co-purification in yeast have a substantially lower false negative rate than previous datasets when compared to known complexes . These datasets are therefore more suitable to estimate the conservation of protein complex membership than hitherto possible . We perform comparative genomics between curated protein complexes from human and the HTP data in Saccharomyces cerevisiae to study the evolution of co-complex memberships . This analysis revealed that out of the 5 , 960 protein pairs that are part of the same complex in human , 2 , 216 are absent because both proteins lack an ortholog in S . cerevisiae , while for 1 , 828 the co-complex membership is disrupted because one of the two proteins lacks an ortholog . For the remaining 1 , 916 protein pairs , only 10% were never co-purified in the large-scale experiments . This implies a conservation level of co-complex membership of 90% when the genes coding for the protein pairs that participate in the same protein complex are also conserved . We conclude that the evolutionary dynamics of protein complexes are , by and large , not the result of network rewiring ( i . e . acquisition or loss of co-complex memberships ) , but mainly due to genomic acquisition or loss of genes coding for subunits . We thus reveal evidence for the tight interrelation of genomic and network evolution .
Many proteins perform their functions together with other proteins to form distinct complexes which are responsible for specific processes in a cell . Understanding how , why and when proteins associate into stable protein complexes is a pivotal part of understanding cellular life . The evolution of protein complexes is intrinsically of interest , as protein complexes are important functional units . In addition , evolutionary information can help us to clean noisy high-throughput data on protein complexes and interactions [1] , [2] . In general , measuring the evolutionary dynamics of protein complexes should improve the framework for function prediction and comparative analysis of interactome networks . For example , knowledge on interactome evolution can help us to establish how reliably we can transfer measured interactions of a protein in S . cerevisiae to its ortholog in Human for function prediction . Various aspects of the evolution of protein complexes and interactomes have been studied [3] . Work on interaction networks so far has revealed that highly connected proteins tend to be more conserved than less connected proteins when looking for the presence or absence in other species [4] . Also , higher connected proteins tend to evolve slower than less connected proteins [5] . Moreover it has been shown that the subunits of protein complexes seem to evolve uncohesively: the genomes of many species contain only a subset of the genes that make up a protein complex of a particular species [6] , [7] . However , all these studies did not analyze the evolution of interactions or co-complex membership , but only the evolution of the genes . The actual conservation of protein interactions themselves is still debated , in part because information and direct measurements of interactions in multiple species is sparse . Suthram and co-workers [8] for instance , have found remarkably low overlap in interaction networks between P . falciparum and other eukaryotic interaction networks , like those of yeast and human . They also concluded that even between closer and well studied eukaryotes like S . cerevisiae and D . melanogaster , many interactions and complexes have been lost . This study , and others like it , has been careful to equate small overlap with a low degree of conservation and has pointed out that the analysis of complex evolution has been hampered by the quality of the available high throughput data . In contrast anecdotal evidence based on specific cases studied from the literature suggest high conservation of co-complex membership such as observed in the ribosome [9] . Therefore it remains unresolved to what extent protein interactions and protein complexes are conserved . When analyzing interaction conservation we need to acknowledge that proteins can keep , lose or gain interactions . To properly measure interaction conservation we need data which not only contains protein-protein interactions but also contains data on proteins which do not seem to interact [6] . The measurements as done in interaction experiments initially provided data on the former . Yet when the coverage of the data is such that it approximates ‘complete’ , the probability that a protein pair without measurable interaction does indeed not interact should increase rapidly . With the publication of two new datasets of high throughput tandem affinity purification-mass spectrometry ( TAP-MS ) experiments in S . cerevisiae [10] , [11] , data has become available which is seemingly of high enough quality [11] , [12] to warrant a new look at interaction conservation . We revisit therefore the question of how complexes evolve and how well protein-protein interactions are conserved . Measuring evolution of protein complexes obviously depends on a reasonable definition of what constitutes a complex: proteins can associate strongly to other proteins and form a stable protein complex ( e . g . proteasomes ) or proteins can associate transiently to often many other proteins ( e . g . a kinase and its substrate ) and not be truly part of one stable complex . We chose to study the evolution of the first ( stable ) type . In addition new insights propose a world view where complexes are not static entities but fluctuate in time and space [10] . Unlike the manner in which it is by necessity stored in reference databases such as MIPS or SGD , the composition of protein complexes is condition and sub cellular localization dependent . This also makes it difficult to study the evolution of protein complexes; i . e . if only a subset of the subunits is involved in a complex in another species , is the complex then conserved ? We here adapt to the latter problem by choosing as the unit of which we want to measure conservation “a pair of proteins that are part of the same protein complex” . For brevity we will refer to this as “co-complex membership” or sometimes the even shorter and arguably inappropriate term “interaction” . In this study we extend interaction data by defining non-interactions in order to examine co-complex membership conservation between S . cerevisiae and Human . Estimating the absence of interactions allows us to look at the conservation and not just the overlap between two interaction networks . The analysis reveals that the main processes of evolution for complexes are the acquisition of new or the loss of old subunits as the co-complex interaction network is highly conserved between orthologous proteins in S . cerevisiae and Human .
The new TAP-MS datasets seem to be very complete and accurate [10]–[12] . We explicitly test the completeness of the datasets by specifically analyzing to what extent different HTP datasets are able to predict all interactions and absence of interactions , i . e . the false negative rate ( type 2 error ) . A false negative will result in the observation that an association is absent while in reality the interaction is present but the experiment failed to detect it . We use the false negative rate because it is a measure of how complete the actual connectivity of a given protein is represented in the datasets . Such false negative pairs are crucial for the study of evolution , because these false negatives will erroneously lower the degree of conservation . A reference set of known complexes is needed to assess which co-complex memberships are erroneously reported as absent in the various HTP datasets ( false negatives ) . In the light of the ongoing discussion on what constitutes a complex [10] , [11] , we used different independent sources such as MIPS and SGD and their intersection ( see Table 1 ) . We use the latter as the main reference , because it provides a reference set in which both MIPS and SGD agree and therefore more is reliable in terms of co-complex memberships and complex definition . Naturally , there is a trade-off between the false negative rate and false positive rate when choosing an appropriate cut-off value for the TAP-MS datasets . The optimal cut-off value for the socio-affinity scores was determined by plotting a Receiver-Operator Curve ( see Text S1 ) . We found that a relatively low cut-off value of 0 provides an optimal balance between specificity and sensitivity for measuring complex interactions . We observe that the new datasets achieve very low false negative rates . The Gavin dataset has a false negative rate of 0 . 23 whereas the Krogan dataset has a false negative rate of 0 . 32 ( Table 2 ) . Combining the TAP-MS datasets ( both union and intersection ) does not only increase the number of true positives but also reduces the number of false negatives and consequently the false negative rate ( Table 2 ) , e . g . the intersection of the Gavin and Krogan datasets has a false negative rate of 0 . 11 ( see Figure 1 and Materials and Methods on dataset construction ) . These low false negative rates reveal that when the TAP-MS datasets report an absence of interaction only a small percentage is a “failure” of the experimental assay . The new datasets are therefore a substantial improvement for the study of co-complex membership conservation relative to what was available previously . In addition to the TAP-MS datasets we also analyzed other high-throughput Yeast-2-Hybrid datasets ( Y2H ) by Uetz et al . [13] and Ito et al . [14] in order to compare them to the new datasets ( for an overview on all datasets see Table 3 ) . We see that the false negative rate in these Y2H assays is much higher , when we define absence of an interaction from Y2H conventionally: that is to say an absence is a prey and bait pair that failed to report an interaction . The higher false negative rate of the Y2H datasets is of course to be expected because Y2H measures direct protein-protein interactions rather than co-complex memberships . Mass-spec co-purifications are expected to retrieve co-memberships more easily [15] . At the same time it might also be that Y2H does have a slightly higher natural level of false negatives as implied previously [2] . To test this , we redefined our Y2H negatives for the Uetz dataset as follows: both the bait-prey as the prey-bait has been tested and both failed to report an interaction . We see a very dramatic decrease in the false negative rate for the Uetz ‘strict’ dataset ( Table 2 ) . In fact Uetz strict has a false negative rate comparable to the intersection of the two mass-spec datasets ( 0 . 10 for Uetz strict as compared to 0 . 11 for the Intersection of the Gavin and Krogan datasets , see Table 2 ) . This shows it is possible to obtain reliable indications of the absence of an interaction from apparently less complete datasets . However , this requires specific attention to the method by which an absence of interaction is inferred from the primary data . Due to coverage of this Uetz strict dataset we cannot use it as the main source for the study of the conservation of interaction , but we can use it to test how general our findings from the mass-spec source are , and whether or not they depend on the precise experimental method for detecting interactions . The Gavin and Krogan datasets and in particular the combination of these datasets ( union and intersection ) show a very low false negative rate: i . e . only a small fraction of the true co-complex memberships are not reported by these datasets . Given that these datasets are available with substantially improved false negative rates we have an excellent starting point for comparative genomics to see to what extent co-complex membership is conserved between species . Reactome for Human [16] was used as a highly reliable reference set for calculating interaction conservation . Reactome is a high quality manually curated database based on expert opinion . Recently a Co-IP interaction dataset has been published for the Human interactome by Ewing et al . [17] . We use this dataset as complementary source to confirm our qualitative trends , rather than our main reference set , because this dataset is only slightly larger than Reactome ( 6 , 463 interactions vs . 5 , 960 ) , but has less protein pairs with orthologs in yeast ( 650 vs . 1 , 916 ) and contains experimental noise ( see Text S1 for analysis performed with the Ewing dataset ) . We extracted protein pairs that were part of the same core protein complex according to Reactome . Orthology data was extracted from Ensembl ( see Materials and Methods ) in order to transfer the yeast interaction data onto Reactome ( Figure 1 ) . This analysis revealed that out of the 5960 human co-complex memberships 4044 are absent in yeast due to the absence of either one ( 1 , 828 ) or both ( 22 , 6 ) of the interaction partners , leaving 1916 pairs with orthologs in yeast . In terms of complexes we found that 66% of human complexes contain less than 50% subunits with orthologs in yeast with an average of 35% over all complexes , which is similar to the percentage of protein pairs . These results are confirmed by orthology calculated with inparanoid [18] ( see Text S1 ) . Thus a large number of co-complex membership pairs are not conserved because either one or both of the genes was lost in fungi or acquired in animals . This is consistent with previous findings on the evolutionary cohesiveness of protein complexes [6] . Therefore a tremendous amount of flexibility in the evolution of protein complexes is not due to the evolution of the co-complex membership ( the interactions ) itself , but rather due to the acquisition and loss of subunits from the genome . We subsequently asked how many of the 1 , 916 gene pairs are also part of the same protein complex in yeast and , more importantly , we also counted how many pairs are not interacting according to our inferred non-interacting pairs . In case of inparalogs conservation of interaction was inferred when one of the inparalogs returned a positive interaction from the datasets ( see Materials and Methods ) . We observe a high rate of co-complex membership conservation: 82 . 5% to 85 . 2% for the Gavin and Krogan datasets respectively and 91 . 1% to 94 . 9% for the Inclusive and Intersection datasets respectively ( Table 4 ) . Although this seems in contrast to the Y2H datasets ( Uetz dataset reaches 24 . 1% , Ito dataset 8 . 6% ) , the Uetz strict dataset returns 84% conservation . The Y2H thus in fact confirms the observation on conservation from the TAP-MS datasets . The rate of conservation that we obtain from the protein purification experiment datasets are not based on a small subset of protein pairs but on a very large proportion of all associated protein pairs . The TAP-MS datasets have coverage of up to 90% when combined as the union of both datasets . The coverage of Reactome by the Krogan and Gavin datasets is substantial ( 81% and 68% resp . ) , whereas the Y2H datasets cover at most 2% ( Ito dataset ) of the 1916 orthologous protein pairs in Reactome . Moreover the conservation rates are based for e . g . the intersection on 133 distinct complexes ( Table 4 ) . From the high conservation rates as well as the percentage of coverage as determined from our analysis based on the TAP-MS datasets , we conclude that the evolution of protein complexes is mainly due to the acquisition or loss of subunits and not due to network rewiring . Analogous to the yeast datasets and the yeast complex definitions , we analyzed the overlap of the human Co-IP [17] dataset and Y2H datasets [19] , [20] with Reactome . To prevent bias we only took those Reactome gene pairs that have orthologs in yeast . The overlap between the human datasets and Reactome is surprisingly so small , that they perform worse than the Y2H datasets from yeast . The small coverage of the human datasets is perhaps caused by the fact that the human HTP interaction studies targeted proteins that are presumably of more interest to mammalian systems . From the high conservation rates as determined from our analysis we conclude that the evolution of protein complexes is mainly due to the acquisition or loss of subunits and not due to network rewiring . The non-conserved interactions are those associations between protein pairs that are present in yeast and human as orthologs but whose interaction seems to have been either lost in yeast or acquired in human . These associations are potentially interesting because they tell us about the evolution of new interactions . Out of the 1884 associations covered by the inclusive dataset only 167 seem to be not conserved ( see Table 4 ) . We scanned this list manually searching for possible errors in annotation , false negatives and true negatives ( actual non-conserved protein-protein interactions ) . Of the 167 protein pairs 139 pairs are present in the same complex in yeast according to GO and/or MIPS or based on literature . In other words , a large portion of these pairs seem to be a member of the same protein complex in yeast and human according to the literature , but were never co-purified in either Krogan or Gavin . I . e . these 139 are possible false negatives of the experimental assays rather than non-conserved interaction pairs . The remaining 28 non-conserved interactions ( see Text S1 ) consist of errors in orthology of one gene ( 5 interactions ) , incorrect assignment of two proteins to a complex in Reactome ( 10 interactions ) and possible neo-functionalisation after duplication in human ( 3 proteins , 13 interactions ) . Based on the analysis of the proteins pairs which did not have an interaction according to the HTP datasets , it seems that the actual conservation of co-complex membership might be higher than follows from our analysis , because we mostly ran into potential errors in orthology assignment , conceptual issues in the curated database of Reactome , or false negatives in the HTP assay . Interestingly , in this analysis the three proteins which represent potentially new complex memberships , are all proteins which have retained the same or similar function as their orthologs in yeast but have acquired additional functions and interactions in human .
We have shown that with the publication of the TAP-MS datasets by Gavin et al . [10]and Krogan et al . [11] we now have datasets which are sufficiently large to reliably estimate the level of co-complex membership conservation . Specifically , we have shown that the false negative rate of these datasets can be reduced to 7% . This means that we are now able to do comparative network studies with substantially less coverage problems for the yeast interactome than previous studies . This is important as estimates of the level of co-complex membership conservation do not only depend on reliable measures for the presence of a link but also on reliable measures for the absence of a link . Unfortunately similar interaction data is not available for other species . We have therefore chosen to use a curated interaction database called Reactome and extracted complex definitions . Combining the human Reactome complex definition and the interaction data for yeast reveals that the complex protein pairs which have been conserved in both species do not lose their interaction in contrast to what has been previously suggested [8] , [21] . We conclude therefore that evolution of protein complexes does not involve extensive network rewiring , but is mostly due to loss of subunits and the acquisition of novel proteins . This type of behavior is clearly illustrated by the eIF3 protein complex from human and its comparison to the complex in yeast ( see Figure 2 ) . The eIF3 complex in yeast ( yellow ) and human ( green ) are depicted in a network with similar topology relevant to the orthologs ( connected by red dotted lines ) . Although the eIF3 complex in human has expanded compared to yeast , all yeast proteins are also part of the same complex in human ( light green ) . Modifications of the complex during evolution have been through the acquisition of new proteins ( green ) . The high degree of co-complex membership conservation could potentially arise from some degree of circularity: the protein complexes in human have been originally identified in yeast . However , our knowledge of human complexes is not limited by what we know about complexes in yeast , as can be deduced by many human subunits which do not have orthologs in yeast such as EF3C or IF36 in the example of the eIF3 protein complex ( Figure 2 ) . In general many human interactions are disrupted in yeast due to the absence of either one ( 1 , 828 ) or both ( 2 , 216 ) of the interaction partners . All these subunits are part of a complex in human but are absent in yeast . The knowledge about these subunits is the result of direct intensive biochemical analysis in human or other animal systems . Therefore , we have a substantial degree of trust in our estimate of interaction conservation , because the knowledge on the protein complexes deposited in Reactome is the result of direct extensive experimentation in animal systems and is not only based on experimentation in yeast . An important aspect of protein-protein interaction evolution is that the physical interaction surface is often provided by distinct protein domains . In evolution of protein-protein interactions they play an important role as acquisition or loss of a particular domain can result in the combination of new interactions with new functions . Itzhaki et al . [22] report that 9% of protein-protein interactions in yeast and 20% in human can be ascribed to domain-domain interactions . It therefore bears to mind that a small part of co-complex membership conservation might not be due to the conservation of whole proteins but due to specific domains which have maintained the interaction . This would leave a conserved interaction network the freedom to add or change function without having to compromise interaction integrity . Another possible theoretical framework for our observations is given by Kirschner and Gerhart [23] , who argue that conserved mechanisms or processes are conserved because they “deconstrain” phenotypic variations in other processes . Our observations neatly fit their theory: the conserved proteins and their conserved interactions represent a “backbone” to which variable subunits are observed to be added or removed . The possible new interactions that we have found , XAB2 , PCBP1 and PABP2 , still have the same or similar function as their yeast orthologs , but have acquired new functions and new interactions in human . Additions to the functionality were made only through minor instead of radical adjustments leaving the interaction network intact and added upon . In the light of co-complex membership this might imply that it is easier to add function and interactions than it is to remove the interaction while retaining the gene . The high conservation of co-complex memberships is also support for bioinformatic function prediction by transfer of information on complex-membership between orthologs: this aspect of gene function can be reliably transferred between evolutionary divergent species such as yeast and human when the partner gene is also present . We have shown that the gain of interactions by existing proteins in complexes seems quantitatively not important in evolution . Rather the evolution of protein complexes is dominated by co-complex memberships that are acquired or lost concomitantly with acquiring or losing the gene . However , the precise order of events in the latter case is difficult to determine . If we for example suppose that the absence of an ortholog in yeast of a human protein complex member is the result of a gene loss ( deletion ) in the fungal lineage ( rather than being acquired in animals ) , then there are two scenarios than can explain this loss . On the one hand the loss of membership to a protein complex could have preceded the evolutionary loss of the gene . On the other hand a co-complex membership is by definition disrupted by the deletion of the gene coding for the subunit from the genome . For both examples divergence in transcriptional regulation could mediate less dramatic scenarios of interaction loss . Transcriptional regulation diverges significantly between relatively close species [24] and is therefore a faster process than for example gene loss or acquisition . Loss of membership could have preceded a fast transcriptional down regulation to avoid expression of potentially rogue proteins before the actual loss of the gene . If a subunit is no longer needed deletion of this subunit could have been preceded by down regulation , which could have given the organism some time to adapt ( stabilize the complex ) to the missing of the subunit before its deletion from the genome . Although gene loss preceded by interaction loss seems somewhat more likely , the high level of co-complex membership conservation that we observe in those cases were the protein pairs are present in both species , suggest a low frequency of such evolutionary intermediate stages . Because we find such low frequency of intermediate stages and a high conservation rate of interactions between conserved proteins we reveal evidence of the tight interrelation of genomic and network evolution .
All data was handled by Perl scripts ( Perl 5 . 8 . 8 ) on a 64 bit Linux machine . | Protein complexes are a pivotal part of the functioning of cells in health and disease . Studying the evolution of these essential cellular features is of great intrinsic as well as practical interest . However , the study of the evolution of protein complexes by comparative analysis is fraught with difficulties . Hence current reports that reveal low overlap in the interactome between species are often reluctant to equate this low level of overlap to a low level of conservation . Here we exploit new public data sets , which display unparalleled coverage , to study the amount of co-complex membership conservation , and we present a novel measure for the absence of interactions . We thereby observe a hitherto unreported high level of conservation of 90% of the interactions when the presence of the genes coding for the protein pairs that participate in the same protein complex is also conserved . This allows for new insights into the evolution of protein complexes: the evolutionary dynamics of protein complexes are , by and large , not the result of network rewiring ( i . e . acquisition or loss of co-complex memberships ) , but mainly due to genomic acquisition or loss of genes coding for subunits . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/bioinformatics",
"computational",
"biology/genomics",
"evolutionary",
"biology/genomics"
] | 2008 | Protein Complex Evolution Does Not Involve Extensive Network Rewiring |
Cytokinesis requires the spatio-temporal coordination of membrane deposition and primary septum ( PS ) formation at the division site to drive acto-myosin ring ( AMR ) constriction . It has been demonstrated that AMR constriction invariably occurs only after the mitotic spindle disassembly . It has also been established that Chitin Synthase II ( Chs2p ) neck localization precedes mitotic spindle disassembly during mitotic exit . As AMR constriction depends upon PS formation , the question arises as to how chitin deposition is regulated so as to prevent premature AMR constriction and mitotic spindle breakage . In this study , we propose that cells regulate the coordination between spindle disassembly and AMR constriction via timely endocytosis of cytokinetic enzymes , Chs2p , Chs3p , and Fks1p . Inhibition of endocytosis leads to over accumulation of cytokinetic enzymes during mitotic exit , which accelerates the constriction of the AMR , and causes spindle breakage that eventually could contribute to monopolar spindle formation in the subsequent round of cell division . Intriguingly , the mitotic spindle breakage observed in endocytosis mutants can be rescued either by deleting or inhibiting the activities of , CHS2 , CHS3 and FKS1 , which are involved in septum formation . The findings from our study highlight the importance of timely endocytosis of cytokinetic enzymes at the division site in safeguarding mitotic spindle integrity during mitotic exit .
During mitosis in budding yeast , many cellular processes such as sister chromatid separation and spindle elongation are controlled by the mitotic cyclin-dependent kinase ( CDK1 ) whose activity serves to activate or inactivate its substrates through phosphorylation ( reviewed in [1] ) . As the cell progresses through mitosis , mitotic CDK1 activity is eventually abolished due to the combinatory effect of mitotic cyclins proteolysis and expression of CDK1 inhibitors . The decline of mitotic CDK1 activity , also known as mitotic exit , is a tightly-regulated process involving components that are highly conserved across species . In eukaryotic cells , destruction of mitotic cyclins depends upon the conserved E3 ubiquitin ligase known as the anaphase promoting complex / cyclosome ( APC/C ) for ubiquitin-mediated proteolysis by the 26S proteasome [2] . APC/C is activated by two highly conserved proteins , Cdc20p and Cdh1p . The binding of Cdh1p to APC/C is under the control of a Hippo-like signal transduction cascade known as the Mitotic Exit Network ( MEN ) comprising of Tem1p ( a GTPase ) , Lte1p ( a GTP/GDP exchange factor ) , Cdc15p ( Hippo-like kinase ) , Cdc5p ( Polo-like kinase ) , Dbf2p/Dbf20p ( Ser/Thr kinase ) , Mob1p ( a kinase ) , and its ultimate effector Cdc14p ( Ser/Thr phosphatase ) [3] . The lowering of mitotic CDK1 activity initiates late mitotic events such as septum formation and cytokinesis . Cytokinesis is the process during which a cell physically cleaves to form two genetically identical progeny cells subsequent to nuclear division . In budding yeast , cytokinesis is accomplished by spatio-temporal coordination of the centripetal deposition of the primary septum ( PS ) by Chitin Synthase II ( Chs2p ) and acto-myosin ring ( AMR ) constriction [4–7] . During mitotic exit , the rough endoplasmic reticulum ( RER ) export of Chs2p is permitted only in the presence of low mitotic CDK1 activity , which eventually triggers the constriction of the AMR , leading to cytokinesis [8–10] . After completion of PS formation , Fks1p ( catalytic subunit of β-1 , 3-glucan synthase ) together with Chs3p ( chitin synthase III ) synthesizes the glucan-mannan rich secondary septum next to the ingressing PS [6 , 11 , 12] . These observations are consistent with the idea that Chs2p in budding yeast or β-glucan synthases in fission yeast promote AMR constriction when present at the neck [6 , 13] . Interestingly , it has been shown that during normal cell division , Chs2p and Chs3p neck localization precedes mitotic spindle disassembly at late mitosis [7]; Fks1p also localizes to the mother-daughter neck during mitotic exit prior to AMR constriction [14 , 15] . Crucially , the decreased mitotic CDK1 activity in late mitosis also promotes mitotic spindle disassembly . Mitotic exit contributes to the dismantling of the mitotic spindles in part by inactivation of mitotic effectors such as those required for spindle elongation [16–18] and in part by targeting the microtubule cross-linking proteins that are involved in mitotic spindle stabilization , such as Cin8p , Ase1p , and Fin1p , for proteaosomal degradation [18–20] . Given that mitotic exit promotes both the neck localisation of cytokinetic enzymes and disassembly of mitotic spindles , the question arises as to how cells ensure that spindles are not broken by premature AMR constriction in a normal cell division due to the activities of cytokinetic enzymes at the bud neck [6 , 13] . This is an important issue as cells in which spindle disassembly is delayed have mitotic spindles that are severed as a result of AMR constriction [21] . Indeed , in the absence of Kip3p , a kinesin-8 motor protein that has microtubule depolymerase activity needed to promote microtubule depolymerization during spindle disassembly [21–23] , mitotic spindles failed to disassemble in time and were sheared by AMR constriction [21] . This indicates that normally , mechanisms exist to ensure a tight coordination of spindle disassembly and AMR constriction to prevent untimely breakage of the mitotic spindle during mitotic exit . One relatively-unexplored aspect of the cytokinetic enzymes is how the levels of these enzymes at the neck are regulated during mitotic exit . The timely delivery of cytokinesis enzymes to the neck late in mitosis has been shown to rely upon the secretory pathway trafficking [7 , 8 , 24 , 25] . For instance , Chs2p synthesized at early mitosis is targeted to the neck during mitotic exit when the mitotic CDK1 activity is low [8 , 9 , 26] . Chs3p and Fks1p are constitutively targeted to the plasma membrane throughout all phases of the cell division cycle [11] . However , the cytokinesis enzymes also accumulate at the neck towards the end of mitosis [14 , 15 , 27] , presumably due to mitotic exit . At the end of mitosis , clathrin-mediated endocytosis ( CME ) has been implicated in the removal of Chs2p from the neck [9 , 24] . CME is the major route for protein cargo internalization from the plasma membrane in the budding yeast and occurs constitutively [28] . CME is divided into 3 main phases: the early immobile phase , intermediate/late immobile phase , and WASP/ myosin/ actin/ slow mobile invagination phase [reviewed in [29]] . The early immobile phase depends upon a range of proteins including clathrins Chc1p and Clc1p , Eps15 homology ( EH ) domain protein Ede1p , while the intermediate/late immobile phase relies upon other proteins such as Sla2p and End3p . In the WASP/ myosin/ actin/ slow mobile invagination phase , nucleation promoting factors such as Abp1p are needed to trigger the invagination of the clathrin-coated pit for the internalization of cargoes . Finally , the amphiphysins ( Rvs161p/Rvs167p ) and dynamin drive the scission of the clathrin vesicle from the plasma membrane by narrowing the neck of the invagination tip . Interestingly , while the forward trafficking of the cytokinetic enzymes to the neck has been fairly-well characterized , the removal of these enzymes by CME especially during late mitosis , has been less so . As the interplay among septation , AMR constriction , and spindle disassembly is presently poorly understood , we set out to study the mechanisms underlying timely spindle disassembly and cytokinesis . In relation to this , the regulation of the levels of cytokinetic enzymes at the neck during late mitosis was particularly interesting given the relative dearth of information on this aspect . In our present report , we provide evidence that the coordination between mitotic spindle disassembly and AMR constriction is regulated in part via timely endocytosis of cytokinetic enzymes at the division site during mitotic exit . We show using time-lapse fluorescence imaging that during a normal cell division , the cytokinesis enzymes Chs2p , Chs3p , and Fks1p are localized to the mother-daughter neck throughout mitotic exit when the spindles are still intact . Failure to endocytose cytokinetic enzymes at late mitosis results in a thickened cell wall , aberrant septation , and mitotic spindle breakage . Strikingly , when endocytosis of the cytokinetic enzymes is defective , excessive accumulation of cytokinetic enzymes results in premature AMR constriction prior to spindle disassembly . As a consequence of the mitotic spindle breakage , the spindle fails to reassemble in a proportion of cells in the subsequent round of cell division . These findings highlight the vital role of constitutive endocytosis in safeguarding mitotic spindle integrity during cytokinesis .
To understand how septation , AMR constriction , and mitotic spindle disassembly are coordinated during late mitosis , we used time-lapse microscopy to first characterize the dynamics of cytokinetic enzymes during mitotic exit . We examined the neck localization of cytokinetic enzyme Chs2p-mCherry , and constriction of the Myo1p-GFP ring , relative to mitotic spindle disassembly ( visualized by α-tubulin , GFP-Tub1p ) . The Chs2p-GFP and Myop1-GFP were functional , as W303 cells were inviable without functional Chs2p ( S1 Fig ) or Myo1p [30] . Consistent with results from a previous study [7] , in GFP-TUB1 MYO1-GFP CHS2-mCHERRY cells released from a Noc arrest , Chs2p-mCherry arrived at the neck 2 . 06 ± 0 . 80min before spindle disassembly ( 2-4min; Fig 1A ) . The constriction of Myo1p-GFP was initiated at 0 . 75 ± 0 . 84min after the disassembly of the mitotic spindle and was completed by 4 . 44 ± 0 . 88min ( n = 32 ) ( Fig 1A and 1B ) . This places a time separation between Chs2p neck localization and AMR constriction , with spindle disassembly occurring during the intervening time . Other than Chs2p that contributes to septation and AMR constriction , other cytokinetic enzymes , Chs3p and Fks1p also play a role in cytokinesis [6 , 11 , 14] . Though immunofluorescence data from a study by the Pellman group demonstrated that Chs3p-GFP neck signals were observed in cells with intact mitotic spindles labelled by GFP-Tub1p [14] , the dynamics of mitotic spindle disassembly relative to the neck localization of other cytokinetic enzymes has not been directly examined . Similarly , Fks1p-GFP has been shown to localize at the bud neck before the onset of AMR constriction [15] , though not much is clear about its localisation relative to spindle disassembly . To determine the dynamics of mitotic spindle disassembly relative to the neck arrival of these other cytokinetic enzymes , we examined the dynamics of Chs3p and Fks1p using Chs2p neck localization as the marker . Both Chs3p-3mGFP ( monomeric GFP ) and Fks1p-GFP were functional fusion proteins in our strain background ( S1 Fig ) . Chs2p-mCherry neck signals colocalized with Chs3p-3mGFP ( 0 . 94 ± 0 . 85 , n = 34 ) ( Fig 1C ) and Fks1p-GFP ( 1 . 48 ± 1 . 28 , n = 27 ) ( Fig 1D ) at the division site . These results suggest that cytokinetic enzymes Chs3p and Fks1p , together with Chs2p , localize to the neck prior to spindle disassembly , while AMR constriction occurs subsequent to spindle disassembly . The observations together indicate that mechanisms exist to coordinate septation and AMR constriction with spindle disassembly so as to prevent premature breakage of the spindles . Presumably , mitotic exit , a key event that triggers these processes , plays a role in controlling the timing of these events . However , mitotic exit by itself might not be sufficient to prevent untimely septation and constriction of the AMR , given that the cytokinetic enzymes are present at the neck even while the spindles are still intact . A possible mechanism by which cells restrain septation and AMR constriction before spindles disassemble could be through controlling the neck localisation of the cytokinetic enzymes Chs2p , Chs3p and Fks1p [7] . For instance , the levels of these cytokinetic enzymes at the neck could be regulated by altering the rate at which they are removed from the neck via processes such as CME . Indeed , Sla2p , a component of the CME machinery , was reported to play a role in the internalization of Chs2p , albeit at the end of AMR constriction [9 , 24] . None the less , given that CME has been previously noted to retrieve cargoes as soon as they arrive at the plasma membrane [31] , we sought to characterise the timings of endocytic components appearing at the neck relative to spindle disassembly and the arrival of Chs2p . We examined the dynamics of key CME proteins that are crucial in facilitating endocytosis during mitotic exit by using Abp1p ( actin binding protein that localizes to actin patches ) as an endocytosis marker . In APB1-mCHERRY GFP-TUB1 cells , we noted that Abp1p-mCherry localized to the division site prior to spindle breakage by 1 . 90 ± 0 . 99min ( n = 70 ) ( Fig 1E ) . This highlights a likely role of CME in the internalization of Chsp2 at the neck before the end of AMR constriction and raises the possibility of CME being involved in regulating the timing of septation and AMR constriction . We next investigated the relationship between Chs2p and key CME proteins . In line with the action of Sla2p in Chs2p internalization , Chs2p-mCherry neck localization precedes Abp1p-GFP neck accumulation by 1 . 9 ± 1 . 25min , n = 61 ( S2E Fig ) . Similarly , Chs2p-mCherry neck localization also preceded the neck accumulation of all key CME components examined [ ( Ede1p-GFP , 1 . 87 ± 0 . 87min , n = 63 ) , ( Sla2p-GFP , 1 . 52 ± 1 . 04min , n = 43 ) , ( Las17p-GFP , 1 . 80 ± 0 . 73min , n = 44 ) , ( Sla1p-GFP , 1 . 33 ± 0 . 92min , n = 30 ) , and ( Rvs167p-GFP , 2 . 21 ± 0 . 89min , n = 52 ) ] ( S2 Fig ) . The mass accumulation of endocytic components implies that the rate of endocytosis at the neck increases when cells exit from mitosis . Accordingly , we determined the effect on Chs2p neck internalization during mitotic exit in mutant cells where endocytosis were defective . To this end , we performed time-lapse fluorescence microscopy of CHS2-GFP ABP1-mCHERRY in key endocytosis deletion mutants such as ede1Δ , sla2Δ , end3Δ , and rvs161Δ rvs167Δ mutants that are defective in different stages of CME [29 , 32] . In wild-type cells , Chs2p-GFP localized to and was efficiently internalized from the neck as evident from the disappearance of Chs2p-GFP neck signals ( Fig 2A ) . However , in several key endocytosis deletion mutants , Chs2p-GFP internalization was compromised , as evident from the persistent Chs2p-GFP neck signals ( 3 min onwards , Fig 2A and 2B ) . The Chs2p-GFP was initially retained at the division site but slowly diffused to the plasma membrane surrounding the cell ( Fig 2A ) . To quantify the Chs2p retention at the division site , we measured the fluorescence intensity of Chs2p-GFP . In wild-type cells , the mean intensity of Chs2p-GFP neck signals gradually increased with time but started to decrease 4 min after its arrival at the division site . However , the mean neck fluorescence intensity of Chs2p-GFP in all endocytosis deletion mutants was significantly elevated at all time-points examined , and the neck signals were retained for a longer time as compared to wild-type cells ( Fig 2B ) . Collectively , these data suggest that Chs2p concentration at the neck is likely to be regulated in part by CME during mitotic exit and not merely at the end of AMR constriction . To determine whether endocytosis of the cytokinesis enzymes depends upon mitotic exit , we tested if Chs2p were internalized when mitotic exit were blocked , such as in metaphase- and telophase-arrested cells . As Chs2p export from the RER to the plasma membrane or neck is sensitive to the mitotic kinase activity [8 , 9 , 26] , we made use of the galactose-inducible Chs2p ( 6S-6A ) mutant where the serine residues phosphorylated by the mitotic CDK were mutated to alanines so that the Chs2p ( 6S-6A ) could exit the RER constitutively even in the absence of mitotic exit [9 , 26] . Spc42p-eqFP that is a known central plaque component of the spindle pole bodies ( SPBs ) [33] was used as a maker for metaphase and telophase arrest . As can be seen , in GAL-CHS2 ( 6S-6A ) -GFP SPC42-eqFP cells arrested in metaphase in the presence of Nocodazole ( Noc ) , 98 . 8 ± 0 . 5% ( n>100 ) of cells could be observed with neck or plasma membrane Chs2p ( 6S-6A ) -GFP signals due to depolarized transport from the RER to the plasma membrane ( including to the neck ) in metaphase ( Fig 3A and 3B ) . These Chs2p ( 6S-6A ) -GFP signals converted to vacuolar signals [9] over time ( Fig 3A ) due to CME-dependent internalisation , indicating that endocytosis occurred even during a metaphase-arrest . Treatment of the cells with a low concentration Latrunculin B ( Lat B ) that disrupted actin filaments and affected only endocytosis [9] resulted in the accumulation of the Chs2p ( 6S-6A ) -GFP neck or plasma membrane signals ( Fig 3A and 3B ) . To examine endocytosis of Chs2p ( 6S-6A ) -GFP during a late mitotic exit block , we used the cdc15-2 allele that prevents the complete destruction of the mitotic CDK activity when cultured at the restrictive temperature of 37°C [34] . cdc15-2 cells typically arrest at 37°C in telophase with high mitotic CDK activity [35 , 36] . In GAL-CHS2 ( 6S-6A ) -GFP SPC42-eqFP cdc15-2 cells arrested at 37°C , we observed that Chs2p ( 6S-6A ) -GFP was internalized ( Fig 3C and 3D ) . Again , the endocytosis of Chs2p ( 6S-6A ) -GFP was abolished in the presence of sub-optimal concentration of Lat B ( 89 . 6% ± 0 . 5% , n>100; Fig 3D ) . Our data showing that Chs2p is endocytosed during a block in mitotic exit imply that endocytosis of cargoes at the plasma membrane can occur independently of mitotic exit and is consistent with previous reports showing the recruitment of endocytosis components by the presence of cargoes [31] . More importantly , the data support the notion that cytokinetic enzymes can be endocytosed upon arrival at the neck as the endocytosis machinery appears to function constitutively . In wild-type cells shortly after the primary septum is laid down , Chs3p and Fks1p deposit the secondary septum on either side of the primary [4] . Also , in chs2Δ mutant cells where the PS is absent , a remedial septum is laid by Chs3p [6] . We therefore asked whether both Chs3p and Fks1p are also regulated in a manner similar to Chs2p at the end of mitosis . Unlike Chs2p that is specifically expressed during mitosis , Fks1p is constitutively expressed throughout the cell division cycle . Upon synthesis in the RER , Fks1p is delivered to the plasma membrane via the secretory pathway in a polarized fashion [25] . Similar to Chs2p , Fks1p is subsequently transported to the vacuole for degradation [37] . To study the role of endocytosis in regulating Fks1p localisation at the division site during mitotic exit , we examined the neck localization of Fks1p-GFP in large-budded cells isolated from cycling culture at 32°C . In wild-type cells , Fks1p-GFP neck localization was observed in 34 . 8 ± 1 . 0% of large-budded cells . Strikingly , we found that the incidence of Fks1p-GFP neck signals was significantly higher in all endocytosis mutants analyzed as compared to wild-type cells [ede1Δ = 49 . 9 ± 3 . 7% , sla2Δ = 74 . 2 ± 5 . 5% , end3Δ = 38 . 5 ± 0 . 6% , rvs161Δ rvs167Δ = 51 . 3 ± 3 . 4% , ( n>600 ) ] ( Fig 4A ) . These results suggest that the levels of Fks1p at the division site might be regulated through CME as defects in endocytosis contribute to abnormal Fks1p neck accumulation . In contrast to Chs2p and Fks1p that are subjected to vacuolar degradation upon arrival at the neck , Chs3p is not targeted to vacuoles for degradation . Rather , it is recycled between the plasma membrane and the chitosome [24 , 37 , 38] . Temperature-sensitive mutations that block endocytosis , end3-1 and end4-1 , resulted in a reduction of Chs3p levels in chitosomes , suggesting that Chs3p levels at the plasma membrane are regulated via endocytosis [24] . We speculated that defective endocytosis might also lead to premature accumulation of Chs3p at the mother-daughter neck . We therefore compared bud-neck localization of Chs3-3mGFP during mitotic exit between large-budded cells in asynchronously growing wild-type and endocytosis deletion mutant cells in the manner as described for Fks1p-GFP ( Fig 4A ) . In wild-type cells , about 53 . 1% of large-budded cells showed Chs3p-3mGFP neck signals . As anticipated , we found that the percentage of Chs3p-3mGFP neck signals in endocytosis deletion mutant cells was significantly higher in comparison with wild-type cells [ede1Δ = 79 . 7 ± 3 . 1% , sla2Δ = 95 . 7 ± 3 . 0% , end3Δ = 80 . 1 ± 1 . 7% , rvs161Δ rvs167Δ = 76 . 3 ± 0 . 5% , ( n>500 ) ] ( Fig 4B ) . This result implies that endocytosis also plays an important role in the retrieval of Chs3p from the neck during mitosis . Collectively , our results suggest that the levels of the cytokinetic enzymes , Chs2p , Fks1p , and Chs3p are in part regulated by CME at the end of mitosis . Based on the importance of CME at the end of mitosis , we next determined the effect of compromised endocytosis on the cell wall , given that ultrastructure of the septum in key endocytosis deletion mutants has not been documented except for the temperature sensitive mutant sla2-41 [39] . In agreement with our fluorescence microscopy data showing accumulation of Chs2p , Chs3p , and Fks1p , the cell wall of the end3Δ and sla2Δ mutants formed extra layers of chitin-rich primary septum and abnormally thick secondary septa when examined using transmission electron microscopy ( S3 Fig ) . These results suggest that the thickened septum observed in endocytosis mutants is perhaps due to a failure in the retrieval of chitin synthases and glucan synthase at the division site . Taken together , the data imply that endocytosis at the division site is important for the removal of excessive cytokinetic enzymes to prevent aberrant cell wall formation and abnormal septation during mitosis . Given that defective endocytosis of cytokinesis enzymes leads to irregular chitin deposition , we wondered if there were any defects in AMR constriction or spindle dynamics as these enzymes normally arrive at the neck prior to disassembly of the mitotic spindles . We first examined if there were defects in the progression of mitotic exit in the endocytosis mutants . From western blot analyses , we noted that mitotic exit in the endocytosis defective cells was comparable to that in wild-type cells ( S4 Fig ) . We then examined AMR constriction and spindle dynamics in endocytosis mutants harboring GFP-TUB1 MYO1-GFP . Consistent with results from a previous study [21] and our data above ( Fig 1A ) , the mitotic spindle disassembled prior to constriction of Myo1p-GFP in wild-type cells ( 95 . 2% ± 4 . 2% , n = 137 ) ( Fig 5A , 5B and 5C ) . In contrast , all of the endocytosis deletion mutants examined displayed varying degrees of what appeared to be mitotic spindle breakage as evident from the execution of AMR constriction prior to mitotic spindle disassembly . Indeed , in 38 . 8 ± 7 . 7% of ede1Δ ( n = 109 ) , 35 . 6 ± 13 . 5% of sla2Δ ( n = 85 ) , 72 . 2 ± 3 . 3% of end3Δ ( n = 119 ) and 55 . 1 ± 9 . 4% of rvs161Δ rvs167Δ ( n = 110 ) cells , AMR constriction occurred before mitotic spindle disassembly ( Fig 5A , 5B and 5C ) , resulting in possible spindle breakage . To investigate the possibility of spindle breakage in the endocytosis mutants at a greater detail , we observed the initial location of spindle-halves separation during spindle breakdown in wild-type and endocytosis mutant cells as previously described [21] . When the various strains harbouring GFP-TUB1 MYO1-GFP were released from Noc into fresh medium containing DMSO , the endocytosis mutants mostly exhibited an uneven distribution of the initial location of spindle-halves separation , with a population located nearer the bud neck . This was unlike the wild-type cells where the initial location of spindle-halves separation was widely distributed ( Fig 6 ) . This implies that the spindles in the endocytosis were indeed sheared by AMR constriction instead of undergoing normal spindle disassembly . In support of this idea , when the strains were released from Noc into medium containing Lat B that inhibited AMR constriction , the distributions of the initial location of spindle-halves separation in the endocytosis mutants became more widely distributed ( Fig 6 ) . As a further confirmation of the spindle breakage in the endocytosis mutants , we made use of a midzone marker Ase1p-GFP to analyse the ends of spindle-halves as cells exited from mitosis . ASE1-GFP mRUBY2-TUB1 MYO1-tdTOMATO and ASE1-GFP mRUBY2-TUB1 MYO1-tdTOMATO end3Δ cells were examined following release from Noc arrest . end3Δ was chosen as a representative endocytosis mutant as it is a key CME component . As can be seen , in wild-type cells , Ase1p-GFP decorates the midzone during the onset of anaphase , and dissociates from the spindle prior to constriction of AMR . ( Fig 7A ) . In end3Δ cells , we noticed the spindle was sheared by AMR in the region outside the midzone and Ase1p-GFP remained on the broken spindle halves ( 12min; Fig 7B ) . 61 . 6 ± 19 . 7% of end3Δ cells ( n = 43 ) exhibited broken spindle-halves , which was significantly higher than in wild-type cells ( n = 44 ) ( Fig 7C ) . In addition , when we measured the time from anaphase B ( spindle length >6μm ) to the completion of Myo1p-GFP constriction , we found a shorter interval in end3Δ cells as compared to wild-type cells ( Fig 7D ) . Collectively , the data hint at a loss of coordination between spindle disassembly and AMR constriction in the endocytosis mutants , leading to the shearing of the mitotic spindles . To test the idea that excessive cytokinetic enzymes in endocytosis mutants might trigger premature AMR constriction leading to the shearing of the mitotic spindles , we next examined the dynamics of AMR constriction by measuring the time taken for complete constriction of Myo1p-GFP relative to Chs2p-mCherry neck localization in wild-type and key endocytosis deletion mutants ( Fig 8Ai and 8Aii ) . In wild-type cells , the constriction of AMR was completed 6 . 52 ± 0 . 13min ( n = 50 ) after Chs2p arrival . As anticipated , all key endocytosis deletion mutants showed an accelerated AMR constriction in comparison to wild-type cells . The times taken for Myo1p-GFP constriction upon Chs2p-mCherry arrival in ede1Δ ( 4 . 16 ± 0 . 19min , n = 50 ) , sla2Δ ( 5 . 98 ± 0 . 21min , n = 50 ) , end3Δ ( 5 . 0 ± 0 . 23min , n = 50 ) , and rvs161Δ rvs167Δ ( 4 . 84 ± 0 . 18min , n = 50 ) cells were significantly shorter as compared to wild-type cells ( p-value <0 . 0001 ) ( Fig 8Aii ) . The observations support the notion that failure in CME internalization of cytokinetic enzymes might trigger premature septum deposition and consequently , premature AMR constriction . Given that endocytosis defects can cause premature AMR constriction , we tested whether increasing the efficiency of endocytosis would lead to the opposite effect . To this end , we released GFP-TUB1 MYO1-GFP and lsb1Δ lsb2Δ GFP-TUB1 MYO1-GFP cells from Noc arrest and performed time-lapse imaging to examine the dynamics of AMR constriction relative to anaphase B . LSB1 and LSB2 encode for the inhibitors of Las17p [40 , 41] and were deleted to increase the internalization of cytokinetic enzymes . In comparison to wild-type cells , lsb1Δ lsb2Δ cells exhibited a longer duration of AMR constriction ( Fig 8Bi and 8Bii ) , supporting the idea that endocytosis of cytokinetic enzymes can influence AMR constriction . This is in line with the hypothesis that endocytosis plays a role in the neck levels of cytokinesis enzymes , which then influences the dynamics of AMR constriction . As premature AMR constriction can lead to spindle breakage , we next tested the idea that perturbing the coordinated occurence of AMR constriction relative to spindle disassembly alters the incidences of spindle breakage . On the one hand , we made use of the kip3Δ mutant that failed to dismantle spindles in a timely manner [21 , 23] in combination with the end3Δ . As can be seen , the percentage of spindle breakage in the end3Δ mutant cells was increased with deletion of KIP3 ( Fig 5A , 5B and 5C ) to 93 . 7 ± 1 . 9% ( n = 79 ) . On the other hand , when we destabilized spindles in the end3Δ mutant by deleting SLK19 ( microtubule stabilizing protein ) [42] , the double mutant had greatly reduced the spindle breakage of 11 . 4 ± 4 . 4% ( n = 88 ) ( Fig 5B and 5C ) . This data supports the notion that a loss of coordination in the relative timing of AMR constriction and spindle disassembly results in spindle breakage in end3Δ cells . The results further imply that the timely turnover of cytokinetic enzymes at the neck in a continuous CME dependent manner plays a role in restraining AMR constriction prior to spindle disassembly . We therefore asked if the deletion of the genes encoding the cytokinesis enzymes would rescue the spindle breakage phenotype in end3Δ cells . As a control , we used kip3Δ cells that were shown previously to exhibit spindle breakage phenotype [21] . The phenotype was attributed to Kip3p being a kinesin-8 motor protein that promotes microtubule depolymerization during spindle disassembly [21 , 23] , and that failure to disassembly spindle in a timely manner results in spindle breakage during AMR constriction in kip3Δ cells . In our setup , 68 . 8 ± 0 . 2% of kip3Δ cells exhibited spindle breakage , in agreement with a previous report [21] . The spindle breakage in the end3Δ mutant could be rescued by deleting CHS3 as the spindle breakage percentage in the chs3Δ end3Δ double mutant dropped to 42 . 2 ± 5 . 9% ( n = 38 ) from 72 . 2 ± 3 . 3 ( n = 119 ) in the end3Δ mutant ( Fig 5B and 5C ) . Surprisingly , deleting FKS1 also rescued spindle breakage in the fks1Δ end3Δ double mutant ( 46 . 1 ± 9 . 4% , n = 91 ) ( Fig 5B and 5C ) , suggesting that Fks1p plays a far bigger role in secondary septum formation than previously described . The role of Chs2p in rescuing the end3Δ spindle breakage phenotype could not be assessed using chs2Δ cells due to compromised viability of the chs2 null mutant in the W303 background [43] . To overcome this issue , we utilized an improved version of auxin-inducible-degron system for rapid depletion of Chs2p during mitotic exit . We constructed the auxin-degradable CHS2 strain by fusing a degron containing single copy of mini-AID [minimum region of IAA17 ( 65–132 amino acids ) required for degradation] to the C-terminal end of CHS2 . In addition , the yeast-codon-optimized Oryzae sativa E3 ubiquitin ligase under the ADH1 promoter ( pADH1-yeOSTIR1 ) was integrated at the LEU2 locus [44 , 45] . To determine the role of Chs2p in shearing of mitotic spindle during mitotic exit [44 , 45] , END3 or end3Δ cells harbouring GFP-TUB1 MYO1-GFP CHS2-1xMini-AID were examined in cells release from Noc arrest . Indole-3-acetic acid ( IAA ) was added to induce Chs2p degradation before cells were released into YPD containing IAA . In our experiments , a sub-optimal concentration of IAA , 0 . 25mM was used , as complete depletion of Chs2p will cause the breakage of AMR , leading to cytokinesis defects ( S5 Fig ) [7] . In END3 cells , mitotic spindle breakage was not observed either with ( n = 240 ) or without IAA ( n = 217 ) ( Fig 9A and 9C ) . Conversely , end3Δ cells that were not treated with IAA displayed a mitotic spindle breakage phenotype with a percentage breakage of 42 . 4 ± 5 . 3% ( n = 139 ) ( Fig 9A and 9C ) . The shearing of mitotic spindle was greatly reduced to 7 . 4 ± 2 . 5% ( n = 121 ) when Chs2p concentration was depleted using IAA ( p-value< 0 . 001 ) ( Fig 9B and 9C ) . Although sub-optimal concentration of IAA was used to deplete Chs2p level , we also observed some cells exhibited breakage of AMR; the cells were excluded from our analysis . The results suggest that levels of Chs2p at the neck is a key determining factor that contributes to mitotic spindle breakage in endocytosis mutants . To confirm that the cytokinetic enzymes were indeed responsible for triggering premature AMR constriction , Myo1p-GFP and GFP-Tub1p were observed in endocytosis mutants while the activities of Chs3p or Fks1p were inhibited using nikkomycin-Z ( chitin synthase III specific inhibitor ) [46] and caspofungin [β ( 1–3 ) -D-glucan synthase inhibitor] [47] , respectively ( Fig 10A ) . The incidences of spindle breakage of the caspofungin and nikkomycin-Z treated end3Δ mutant cells was decreased to 26 . 5 ± 5 . 3% ( n = 110 ) and 26 . 4 ± 1 . 2% ( n = 113 ) respectively ( Fig 10B and 10C ) . The combinatorial use of nikkomycin and caspofungin was not possible as cells were not viable even when treated acutely . The reduction in spindle breakage incidences in end3Δ mutant cells treated separately with nikkomycin-Z or caspofungin supports the notion that accumulation of cytokinesis enzyme activities led to premature AMR constriction dynamics and untimely spindle breakage . AMR constriction time relative to Chs2p-GFP neck arrival was next examined in cells expressing CHS2-GFP MYO1-REDSTAR and treated with caspofungin or nikkomycin-Z ( Fig 11A ) . In end3Δ mutant cells , the time taken for complete Myo1p-Redstar constriction relative to Chs2p-GFP neck arrival was 5 . 07 ± 0 . 14 min ( n = 92 ) ( Fig 11B and 11C ) . The time taken for complete AMR constriction in end3Δ cells treated with caspofungin [5 . 66 ± 0 . 19min ( n = 61 ) , p-value < 0 . 05] and nikkomycin-Z [6 . 07 ± 0 . 19min ( n = 55 ) , p-value<0 . 001] was significantly longer as compared to untreated end3Δ cells ( Fig 11B and 11C ) . Taken together , the data suggest that spindle breakage in the end3Δ mutant cells might be due to excessive accumulation of chitin synthase II , chitin synthase III , and glucan synthase activities at the division site during cytokinesis leading to a premature initiation and faster AMR constriction dynamics . This consequently resulted in cytokinesis in the presence of intact mitotic spindles . Given that endocytosis mutants are viable in the presence of the broken spindle phenotype , we further examined the consequences of broken spindles during mitotic exit . We assessed the spindle morphology of wild-type and endocytosis mutants harbouring GFP-TUB1 MYO1-GFP SPC29-RFP . As a marker for SPBs , we used Spc29p , an inner plague component of the SPB [48] . Cells were cycled at 32°C for 2 hours to induce spindle breakage in endocytosis mutants , after which they were arrested using hydroxyurea ( HU ) . Typically wild-type cells would arrest in S-phase with a short bi-polar spindle [49] . Indeed , we noted that 98 . 1 ± 0 . 05% ( n = 207 ) of wild type cells displayed a short bipolar spindle phenotype ( Fig 12A and 12B ) . In contrast , the endocytosis mutants , end3Δ and rvs161Δ rvs167Δ showed a significantly higher incidence of monopolar spindle formation as compared to wild-type cells ( p-value<0 . 05 , Fig 12A and 12C ) . The end3Δ mutant showed the highest percentage of monopolar spindle ( 25 . 3 ± 9 . 3% , n = 221 ) , followed by rvs161Δ rvs167Δ ( 7 . 2 ± 2 . 1% , n = 152 ) ( Fig 12D ) . There was no significant difference in monopolar spindle formation between wild-type and ede1Δ mutant cells , due to the low incidence of spindle breakage in ede1Δ mutant cells ( Fig 12 ) . Next , we determined if the monopolar spindle formation in end3Δ was due to a SPB duplication failure or a defect in spindle elongation . Wild-type and end3Δ strains expressing GFP-TUB1 MYO1-GFP SPC42-eqFP were released from Noc and time-lapse imaging performed as described in Fig 7 , but with an extended duration of imaging until the progeny cells underwent a subsequent round of mitosis . In wild-type cells , all progeny cells completed spindle elongation successfully in the new round of mitosis ( n = 37 ) ( Fig 13A ) . However , 25 out of 38 of the end3Δ cells had broken spindles . Of these cells with broken spindles , all of their progeny cells were able to form short bipolar spindles ( Fig 13B ) . However , 18% of the 50 progeny cells failed to elongate their defective spindles ( 105min-189min; Fig 13B ) . Nonetheless , the data indicate that the monopolar spindles arising from spindle breakage in the endocytosis mutants are not likely due to SPB duplication failure but rather a problem with spindle elongation . We next tested if the monopolar spindles in the endocytosis mutants can be rescued by deletion of chs2Δ or fks1Δ . However , since chs2Δ is inviable in the W303 genetic background yeast strain , CHS2 was controlled under inducible galactose promoter to maintain the viability of chs2Δ cells . GAL-CHS2-13MYC chs2Δ GFP-TUB1 MYO1-GFP SPC29-RFP cells were grown overnight in YP/Raff medium containing 0 . 1% galactose . Consistent with the results from the spindle breakage rescue , there was a significant decrease in the percentage of monopolar spindle formation in chs2Δ end3Δ ( 4 . 3 ± 4 . 4% , n = 162 , p-value <0 . 05 ) and fks1Δ end3Δ cells ( 4 . 6 ± 3 . 1% , n = 153 , p-value <0 . 05 ) ( Fig 12A and 12D ) . Monopolar spindle formation in sla2Δ , chs3Δ and chs3Δ end3Δ cells was not examined due to the compromised viability of these mutants in HU ( S6 Fig ) . Furthermore , previous studies demonstrated that chs3Δ is synthetic lethal with endocytic components , Rvs161p , Rvs167p , and Vrp1p [50 , 51] . The compromised viability of chs3Δ end3Δ double mutant cells was likely due to prolonged arrest in HU at 32°C . The evidence indicates that spindle breakage in the endocytosis mutants leads to failure in re-establishment of spindles in G1-S phase transition . The breakage of the mitotic spindle by the AMR has been previously shown to contribute to monopolar spindle formation , presumably due to an insufficient pool of tubulin for the re-establishment of spindle in progeny cells [52] . We explored this possibility by over-expressing TUB1 and TUB2 in the GFP-TUB1 MYO1-GFP SPC29-RFP end3Δ cells . From the data , there was a low albeit statistically significant reduction in the occurrence of monopolar spindles upon galactose-induction of TUB1 and TUB2 ( Fig 13D ) . This observation points to the possibility that untimely breakage of mitotic spindles could lead to a defect in generating assembly-competent tubulin in a sub-population of the progeny cells .
Lessons from fission yeast ‘cut’ mutants ( reviewed in [53] ) indicate that after cells have committed to septation , any delay or halting of the process is unlikely even if late mitotic events are compromised ( point of no return ) . Consistent with this notion , budding yeast that exhibit defects in spindle disassembly such as kip3Δ mutant cells continue to undergo cytokinesis despite the presence of an intact spindle or a dicentric chromosome [54] , suggesting that mitotic arrest is not an option upon initiation of septation due to the absence of a post-anaphase surveillance system in budding yeast . This raises the question of how cells regulate the septation process to ensure that cytokinesis invariably occurs after spindle disassembly . Such a mechanism is especially pertinent for the coordination distinct processes to generate viable progeny , as cytokinetic enzymes involved in septation such as Chs2p , Chs3p , and Fks1p localize to the neck prior to spindle disassembly ( Fig 1A , 1C and 1D ) . Particularly , given that Chs2p is the main driver of AMR constriction [6 , 7] , the accumulation of Chs2p at the neck could in principle lead to premature execution of cytokinesis that causes mitotic spindle breakage . Moreover , it has been shown that the regulators of Chs2p activity including Inn1p and Cyk3p are also localized to the neck during mitotic exit [10 , 55–57] , implies that the Chs2p is active at the neck in the presence of low mitotic CDK activity . However , typically in wild-type cells , 95% of cells that have exited from mitosis were able to disassemble spindles successfully prior to septation and AMR constriction ( [21] and our data ) . Some hints of how cells might coordinate cytokinesis while maintaining spindle integrity during mitotic exit could be seen from the correlation between higher levels of Chs2p , premature localization of Chs3p and Fks1p at the neck during mitotic exit , and a drastic increase in mitotic spindle breakage in endocytosis mutants ( Figs 2 and 4 ) . Moreover , the spindle breakage phenotype was rescued when either single deletions of the cytokinesis enzymes ( Fig 5 ) or inhibitors of the enzymes were added to end3Δ cells ( Fig 10 ) . These data support the notion that a defect in endocytosis led to increased septation that consequently promoted AMR constriction , prematurely shearing the mitotic spindle . Interestingly , despite the presence of the skewed distribution of the spindle-halves ( Fig 6 ) and asymmetrical localisation of spindle midzone as marked by Ase1p-GFP on the broken spindles ( Fig 7 ) in the endocytosis mutants , septation and AMR constriction were not inhibited . This might appear inconsistent with the reports showing that midzone mutants triggered the NoCut checkpoint that functions to delay AMR contraction and abscission [58 , 59] . However , it should be noted that in the midzone mutants examined previously [58 , 59] , the loss-of-function of the midzone components led to unstable inter-polar microtubules that caused instability and collapse of the mitotic spindles . With defective spindles , abnormal chromosomes segregation ensued and lagging chromosomes occurred as a consequence . It was proposed that the trigger for the NoCut checkpoint that delayed AMR constriction and abscission was in fact the presence of lagging chromosomes at the midzone [58 , 59] . In the endocytosis mutants , however , presumably the mitotic spindles were functional and chromosomes segregated normally . Therefore , in the endocytosis mutants , the NoCut checkpoint was not activated and AMR constriction occurred due to the accumulated activities of the cytokinetic enzymes despite the presence of intact mitotic spindles . This supports the notion that a post-anaphase surveillance system in budding yeast might not exist once chromosomes have cleared the midzone . The findings that AMR constriction can occur in endocytosis mutants in the presence of intact mitotic spindle is significant , as cells with severed spindles could possibly enter the subsequent round of cell division to form monopolar spindles even though the SPB duplication appeared normal ( Fig 13 ) . Moreover , our data showing a marginally significant rescue of the monopolar phenotype by the overexpression of TUB1 and TUB2 implicates spindle breakage in causing a deficiency in the intracellular pool of assembly-competent tubulin available for progeny cells during subsequent mitoses as previously suggested [52] . However , while a significant number of end3Δ cells exhibited broken spindles ( 72 . 2 ± 3 . 3% , n = 119 ) , only ( 25 . 3 ± 9 . 3% , n = 221 ) had monopolar spindles ( Figs 5 and 12 ) . The figures are similar to the kip3Δ mutant and would suggest that the cells are fairly robust in terms of recovery from spindle breakage . Indeed , the synergistic effects on spindle breakage in the kip3Δ end3Δ double mutant point to the possibility that spindle assembly is a complex process that perhaps relies on several redundant pathways to ensure the formation of functional mitotic spindles in progeny cells embarking on subsequent rounds of cell division . The mechanisms underlying tight coordination of the final events of mitosis has been the subject of various studies , with a substantial focus on the mitotic kinase activity as the main determinant . Work from other labs have demonstrated that a decline in mitotic CDK1 activity and the release of the late mitotic phosphatase Cdc14p could lead to the reversal of the action of the mitotic CDK1 activity on several substrates related to these processes . Mitotic CDK1 activity therefore appears to play a key part in the coordination of spindle disassembly , septation and AMR constriction . For instance , spindle disassembly depends upon the APCCdh1 complex that is activated upon reduction in mitotic kinase activity . APCCdh1is needed for the destruction of midzone proteins Cin8p and Ase1p that function early in mitosis for spindle elongation [20 , 60] . Septation is also contingent upon the export of Chs2p from the RER when mitotic CDK1 activity is sufficiently reduced and Cdc14p released [8 , 9 , 26] . Furthermore , AMR constriction requires low mitotic kinase , in part due to the presence of Chs2p and in part due to the localization of MEN components to the neck [61] , though the functions of the MEN components at the neck remain unknown . However , our data showing that premature AMR constriction could occur during the earlier phase of mitosis by the mere presence of elevated levels of cytokinesis enzymes highlights the fact that the level of mitotic CDK1 activity is not directly/solely responsible for coordinating spindle disassembly and cytokinesis . Moreover , the fact that the cytokinetic enzymes localise to the neck during mitotic exit prior to spindle disassembly further argues against the case for the drop in mitotic kinase activity as being a key factor coordinating these mitotic events . Rather , other processes such as endocytosis that might appear not to be directly regulated by mitotic activity could also contribute to the timeliness of cell cycle events . This is important as the trigger for endocytosis appears to be due to the presence of cargo . For instance , as shown from a previous study , the presence of endocytic cargo at the membrane could trigger the endocytic process leading to the internalisation from the plasma membrane [31] . In the case of the cytokinetic enzymes , this places a negative feedback loop that regulates the levels of the enzymes that could contribute to septation and AMR constriction without compromising spindle integrity . Our model places the retrieval of endocytosis proteins within the normal regulation of septation , AMR constriction and spindle assembly ( Fig 14 ) . As such , while it has been previously shown that endocytosis of Chs2p and Chs3p occurred during late mitosis to remove them from the neck subsequent to septation [6 , 9 , 24] , our data suggest that the retrieval process is in fact active all throughout mitotic exit ( Fig 3 ) . This suggestion is borne out by our data showing that endocytosis components are found at the neck very soon after Chsp2-mCherry appears at the neck ( S2 Fig ) . In CHS2-mCHERRY ABP1-GFP cells , the Chs2p-mCherry preceded Abp1p-GFP in the neck localisation by 1 . 9 ± 1 . 25min ( S2 Fig ) , though Chs2p-mCherry and Abp1p-mCherry can be observed 2 . 06 ± 0 . 80min and 1 . 90 ± 0 . 99min respectively before spindle disassembly when compared to GFP-Tub1p . There is some variablity in the timings when examining Chs2p-mCherry and Abp1p-mCherry across different strains but it is important to note that the differences do not account for the heightened levels of Chs2p-GFP accumulation in the endocytosis mutants ( Fig 2 ) . It is clear that both Chs2p and Abp1p arrived at the neck prior to spindle disassembly ( Fig 1A and 1E ) . More crucially , at the bud necks of wild-type cells , cytokinetic enzymes such as Chs2p-GFP neck levels do not accumulate to levels that are found in endocytosis mutants at any point in time ( Fig 2 ) . We propose that during mitotic exit when cytokinetic enzymes such as Chs2p , Chs3p and Fks1p are trafficked to the plasma membrane at the neck and are internalised soon after their arrival . Presumably , the rate of endocytosis of these enzymes is slower than the forward trafficking process that eventually leads to the accumulation of these enzymes at the neck . Presently , it is unclear how the differential rates of forward trafficking and endocytosis are maintained , though the distinct rates might be inherent properties of the two pathways and the net levels of the enzymes at the neck are due to stochastic accumulation . Alternatively , it is also possible that the orientation of the actin cytoskeleton during mitotic exit towards the neck favours forward trafficking relative to endocytosis . This results in a net accumulation of cytokinetic enzymes at the neck that is sufficient to trigger AMR constriction , though the levels required for AMR constriction are reached only after spindle disassembly due to the constitutive endocytosis of the cytokinetic enzymes . At one level , our data implicates endocytosis in the maintenance of spindle integrity , a novel role not previously identified . At another level , our data demonstrates that the interplay between cell cycle-dependent events , such as directional trafficking of cytokinetic enzymes and spindle disassembly , and cell cycle-independent events , such as endocytosis of cytokinetic enzymes , could contribute towards the timely execution of late mitotic events occurring during cell division . Taken together , our data further highlight the complexity underlying the coordination of septation , AMR constriction , and spindle disassembly during late mitosis .
Yeast strains used in this study are listed in S1 and S2 tables . A combination of standard molecular biology and molecular genetics techniques such as tetrad dissection , PCR-based tagging and deletion of endogenous genes [62] were used to construct plasmids and strains . The plasmids for the GFP and mCherry cassettes were obtained from European Saccharomyces cerevisiae Archive for Functional analysis ( Euroscarf ) and Yeast Resource Centre ( YRC ) . Tdtomato ( Addgene plasmid 35193 ) [63] , 3mGFP ( Addgene plasmid # 25449 ) [64] , and pHIS3p:mRuby2-Tub1+3'UTR::LEU2 ( Addgene plasmid # 50645 ) [65] constructs were gifts from Robert Singer , Benjamin Glick and Wei-Lih Lee respectively . Constructs for auxin inducible degron ( AID ) system such as Mini-AID and ADH1-yeOSTIR1 were provided by the National BioResource Project ( NBRP ) of the Ministry of Education , Culture , Sports , Science and Technology ( MEXT ) , Japan . Further information regarding the strains and plasmids construction will be provided upon request . Yeast strains were routinely grown in yeast extract peptone ( YP ) or selective medium supplemented with 2% dextrose ( Glu ) at 24°C . For galactose promoter induction , cells were grown in YP supplemented with 2% raffinose ( Raff ) , followed by addition of galactose ( Gal ) to a final concentration of 0 . 1% . For cells synchronisation , exponential-phase cells were diluted to 107 cells/ml in growth medium at 24°C . For a typical Noc arrest , cells were arrested with 7 . 5 μg/ml Noc ( US Biological ) for 2h , followed by addition of 7 . 5 μg/ml for another 2h at 24°C . For endocytosis deletion mutants , cells were shifted to 32°C for another 0 . 5h after 4 hours Noc arrest . The drug was washed off by centrifugation of the cells . For S-phase arrest , cells were grown at 32°C for 2 h , followed by addition of hydroxyurea ( US biological ) to a final concentration of 0 . 2M . Cells were then incubated for another 5 . 5 h at 32°C and harvested for fluorescent microscopy analysis . Samples were collected from the respective time points stated . Protein lysates were prepared using TCA precipitation method as described previously [8] . Anti-Cdc28 ( 1:1000 dilution ) , anti-Clb2 ( 1:10000 dilution ) ( Santa Cruz , CA ) , anti-Chs2 ( 1:500 ) ( GeneScript ) , anti-AID ( 1:5000 dilution ) ( MBL , Japan ) and anti-Pgk1 ( 1:100000 dilution ) ( Invitrogen ) were used to probe the respective proteins . The Clarity enhanced chemiluminescence kit ( Bio-rad ) was used according to the manufacturer’s recommendations . Cells for TEM analysis were grown in YP supplemented with 2% glucose at 32°C to mid-log phase , fixed with 3% glutaraldehyde , then processed for observation as described in [54] . Samples for fluorescence microscopy were taken at time points indicated in the relevant sections . Cells harbouring fluorescence protein fusions harvested by centrifugation and washed once with dH2O . Samples were observed directly without fixation using an IX81 wide-field fluorescence microscope ( Olympus ) with 60x NA 1 . 4 oil lens , and 1 . 5x optivar . Filter sets for fluorescence microscopy were purchased from Omega and Semrock , and images were capture using a CCD camera ( CoolSnap HQ , Photometrics ) . Images acquisition was controlled by Metamorph software ( Molecular Devices ) . Typically , the exposure time for the acquisition of the images was 0 . 25s for GFP and 0 . 30s for RFP per plane . Nine-optical Z-sections at 0 . 5μm intervals were obtained for each time point . Images shown were either maximal projection of the Z-stacks or images taken at a single plane , as indicated in the relevant sections . For spindle collapsed analysis , 3D reconstruction of Z-stack captured was performed using Metamorph to ensure the monopolar spindles observed were not bi-polar spindles that positioned perpendicular to the slide . At least 50 cells were counted for each time-point from 3 independent experiments . For time-lapsed microscopy , cells released from Noc arrest were resuspended in complete synthetic medium containing glucose and mounted onto 5% agarose pad on slides . Time-lapsed images were captured as described in pervious section . Zero-drift compensation , ZDC ( Olympus ) was used for focal drift correction throughout the time-lapsed microscopy . For time-lapsed microscopy at 32°C , stage top incubator was used to control the temperature . ImageJ ( National Institutes of Health , Bethesda , MA ) and Photoshop ( Adobe , San Jose , CA ) were used for the production of the montages and figures . Cells were treated as described in the wide-field time-lapsed microscopy . Spinning disk images were captured using Olympus IX81-ZDC microscope , with a 60x NA 1 . 4 oil lens . Sapphire LP 488nm and 561nm solid-state lasers ( Coherent ) were used for the samples excitations . Filter sets for fluorescence microscopy were purchased from Omega and Semrock , and images were capture using the Photometrics 512EM-CCD attached behind the Yokogawa CSU22 connected to the microscope . GFP and RFP images were captured simultaneously using via a Dual-View image splitter ( Optical Insights ) . Images acquisition was controlled by Metamorph software ( Molecular Devices ) . Typically , the exposure time for the acquisition of the images was 0 . 2–0 . 35s per plane . 9-optical Z-sections at 0 . 5μm were obtained for each time point . Images shown were either maximal projection of the Z-stacks . For time-lapsed microscopy at 32°C , stage top incubator was used to control the temperature . ImageJ ( National Institutes of Health , Bethesda , MA ) and Photoshop ( Adobe , San Jose , CA ) were used for the production of the montages and figures . | The cytokinesis machinery that is required for physical separation of mother-daughter cells during mitosis is highly conserved from yeast to humans . In budding yeast , cytokinesis is achieved via timely delivery of cytokinetic enzymes to the division site that eventually triggers the constriction of AMR . It has been previously demonstrated that cytokinesis invariably occurs after the disassembly of the mitotic spindle . Intriguingly , Chs2p that is responsible for laying down the primary septum has been shown to localize to the division site before mitotic spindle disassembly . In this study , we show that mitotic spindle integrity upon sister chromatid separation is dependent on the continuous endocytosis of cytokinetic enzymes . Failure in the internalization of cytokinetic proteins during mitotic exit causes premature AMR constriction that eventually contributes to the shearing of mitotic spindle . Consequently , cells fail to re-establish a bipolar spindle in the subsequent round of cell division cycle . Our findings provide insights into how the levels of secreted proteins at the division site impacts cytokinesis . We believe this regulation mechanism might be conserved in higher eukaryotic cells as a secreted protein , hemicentin , has been shown recently to be involved in regulating cytokinesis in both Caenorhabditis elegans and mouse embryos . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"metaphase",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"enzymology",
"light",
"microscopy",
"mitosis",
"microscopy",
"cytokinesis",
"cellular",
"structures",
"and",
"organelles",
"enzyme",
"chemistry",
"research",
"and",
"analysis",
"methods",
"... | 2016 | Timely Endocytosis of Cytokinetic Enzymes Prevents Premature Spindle Breakage during Mitotic Exit |
The auditory pathway consists of multiple stages , from the cochlear nucleus to the auditory cortex . Neurons acting at different stages have different functions and exhibit different response properties . It is unclear whether these stages share a common encoding mechanism . We trained an unsupervised deep learning model consisting of alternating sparse coding and max pooling layers on cochleogram-filtered human speech . Evaluation of the response properties revealed that computing units in lower layers exhibited spectro-temporal receptive fields ( STRFs ) similar to those of inferior colliculus neurons measured in physiological experiments , including properties such as sound onset and termination , checkerboard pattern , and spectral motion . Units in upper layers tended to be tuned to phonetic features such as plosivity and nasality , resembling the results of field recording in human auditory cortex . Variation of the sparseness level of the units in each higher layer revealed a positive correlation between the sparseness level and the strength of phonetic feature encoding . The activities of the units in the top layer , but not other layers , correlated with the dynamics of the first two formants ( F1 , F2 ) of all phonemes , indicating the encoding of phoneme dynamics in these units . These results suggest that the principles of sparse coding and max pooling may be universal in the human auditory pathway .
Hearing is supported by a series of interconnected brain areas , collectively called the central auditory system or auditory pathway [1] . This pathway is thought to function as a series of hierarchical processing stages that encode features ranging from simple acoustic features and elementary time–frequency representations in the cochlea and inferior colliculus to complex phonetic features , phonemes , syllables , words , and grammatical features in the auditory cortex [2–8] . However , it remains unclear how neurons encode these distinctive features , especially at higher stages . The encoding mechanisms of neurons in the auditory pathway have been addressed in many studies , and various encoding mechanisms have been proposed . These include spatial coding in the cochlea [9–11] , spatial coding and temporal coding in the inferior colliculus [12–16] , and spatial coding and periodicity coding in the auditory cortex [17–19] . However , these studies only describe the experimental data , and do not explain why the experiments yield specific outcomes . A notable exception is the sparse coding model , which assumes that neurons encode external stimuli using sparse codes . The model was originally proposed to explain the properties of simple cells in the primary visual cortex [20 , 21] , but has been extended to explain the emergence of the response properties of auditory nerve fibers [22] and inferior colliculus neurons [23] . Another exception is a deep learning model that has been used to explain the emergence of response properties of neurons to speech in the auditory cortex [24] . Because the model was trained to recognize 40 English phonemes in a supervised fashion , high discrimination ability is critical for its success . These two models effectively explained certain neural response properties at different stages along the auditory pathway; however , their underlying computational principles are different . A fundamental question is whether the auditory system uses the same or different principles at different stages . We explored the former possibility in this study . Our initial assumption was that sparse coding plays an important role in shaping neural response properties along the auditory pathway . Support for this is provided by the ubiquitous sparse firing of neurons in the auditory pathway [25–27] , and sparse coding computational models have effectively interpreted experimental data recorded at certain stages of the auditory system [22 , 23] . We extended this model to multiple layers by introducing spatial pooling after each layer , resulting in an unsupervised deep learning model , which we named sparse HMAX ( SHMAX ) [28] . We studied the response properties of the computing units , called artificial neurons , in the model . After training the model on the cochleogram of speech , the spectro-temporal receptive fields ( STRFs ) of artificial neurons in lower layers in the model exhibited patterns similar to those of neurons in the inferior colliculus , whereas the responses of artificial neurons in the upper layers resembled the results of field recordings in human auditory cortex . The agreement between the model and neural data suggest that , although the features encoded at different levels of the auditory pathway differ , the encoding mechanisms are similar .
We first confirmed that the computing units in lower layers of the model could capture the firing properties of the inferior colliculus neurons in the auditory midbrain . This capability has been proven in a single-layer sparse coding model [23]; however , the time window of the bases in that model was too large ( 216 ms ) , and it was unclear whether a much smaller time window such as 20 ms , which is more compatible with physiological data [29 , 30] , would yield similar results . In addition , an overly large time window in the first layer causes difficulty in constructing deep models so that the time windows of higher-layer units agree with those of cortical neurons . Our model started with a time window of 10 ms in layer S1 and ended with a time window of 194 ms in layer C6 . In the following , we will show that , with these settings , several lower layers of the model , rather than only the lowest layer , can generate results qualitatively similar to those obtained in the previous study [23] . The response properties of an inferior colliculus neuron are usually delineated by STRFs , which are obtained by averaging the spectro-temporal structure of acoustic stimuli before a spike is fired [31] . The STRFs of the S1 units can be approximated by the bases used to reconstruct the stimuli [23] . To visualize the STRFs of higher-level units , we linearly combined the bases of units in previous layers ( see Materials and Methods; Fig 2A; S1 Fig ) . Several example STRFs of units in the first three S layers are shown in Fig 2B–2D , and the full results are shown in S2 Fig . This visualization method is simple , fast , and applicable to lower-layer units . In fact , the results are similar to the STRFs obtained by the normalized reverse-correlation method [32 , 33] ( see Materials and Methods; S3 Fig ) . However , in deeper layers , the STRFs become larger and more complex , and exhibit stronger nonlinearity , and are therefore difficult to capture with a linear method . By visual inspection , all of the first three S layers exhibited a certain degree of agreement with physiological data collected from the inferior colliculus , with layer S2 providing the best match in terms of STRF size and pattern ( a quantitative comparison will be provided later ) . Representative bases of layer S2 units are visualized in Fig 3 . Most layer S2 bases had both excitatory regions and inhibitory regions . Specifically , some units favored excitation first , followed by inhibition at the same frequency ( Fig 3A ) . This pattern has been observed in STRFs of inferior colliculus neurons in cats [34] . By contrast , some layer S2 units favored inhibition first , followed by excitation at the same frequency ( Fig 3B ) . This pattern has been observed in STRFs of inferior colliculus neurons in gerbils [23 , 35] . Some layer S2 bases had localized checkerboard patterns in their STRFs ( Fig 3C ) , also as in inferior colliculus neurons in gerbils [23 , 35] . Some layer S2 units were selective to spectral motion ( Fig 3D ) , as in certain inferior colliculus neurons in Mexican free-tailed bat that are tuned to motion cues present in conspecific vocalizations [29] . In those studies , the favored frequencies of layer S2 bases differed from those of inferior colliculus neurons because the experiments were carried out on gerbils and cats , whereas our deep learning model was trained on human speech data with a lower range of frequency content . We calculated the distributions of four parameters that characterized a STRF over all layer S1 , S2 , and S3 units , separately: best temporal modulation frequency ( Best T ) , response duration ( Duration ) , center frequency ( Center F ) , and spectral bandwidth ( Bandwidth ) . Fig 4A–4D shows the results of layer S2 units . The shapes of the distributions of layer S2 units were most similar to those of inferior colliculus neurons in cats [30] ( Fig 4F–4I ) . The scatterplot of Best T and spectral modulation ( Fig 4E ) indicated a tradeoff between the temporal modulation and spectral modulation among layer S2 units; i . e . , units with both high temporal modulation ( Best T ) and spectral modulation were scarce . This result agrees with observations made in inferior colliculus neurons of cats [34] . We hypothesized that the higher layers in the network would correspond to the auditory cortex . A previous study using cortical surface recordings in humans reported selectivity for distinct English phonetic features at single electrodes [6] . Hence , we investigated whether similar results could be obtained in higher layers of our network . Following that study [6] , we separately calculated the phoneme selectivity index ( PSI ) vectors of the units in layers S1 to C6 ( see Materials and Methods ) . Each element in the PSI vector of a unit indicates the selectivity of the unit to a phoneme; the larger the element , the more selective to the corresponding phoneme . Units in lower layers , such as S1 to S4 , did not exhibit distinctive phoneme selectivity ( S4 Fig ) , whereas those in higher layers did ( S5 Fig; Fig 5 ) . As an example , Fig 5A shows the PSI vectors of 173 active units in layer C6 whose responses to randomly selected time frames were statistically larger than their response to silence ( p<0 . 001 ) . Each of these units exhibits strong selectivity for a subset of phonemes , and the unit–phoneme map exhibits a strong clustering effect . We used hierarchical agglomerative clustering analyses [36] with Euclidean distance to determine selectivity patterns across phonemes ( Fig 5B ) and active units ( Fig 5C ) . Phonemes were clustered into six groups according to the place and manner of articulation: plosive , fricative , low back , low front , high front , and nasal . The active units were also clustered into six groups , each selective for one of the six types of phonemes . The PSI vectors of active units sharing a particular phonetic feature were averaged to quantify the feature selectivity of these units . Six groups of units exhibited distinctive roles in characterizing these features ( Fig 5D ) . Similar clustering results were obtained in layers S5 , C5 , and S6 ( S5 Fig ) . However , the dark area in the unit–phoneme plane becomes increasingly prominent from layer S5 to layer C6 ( S5 Fig; Fig 5 ) , suggesting increasing selectivity of these layers for phonetic features . As in previous studies [24 , 37] , we defined an index ( F-ratio ) that measures the overall selectivity of each hidden layer to phonetic features ( Materials and Methods ) . We found that the deeper the layer , the higher the F-ratio ( Table 1 ) . Specifically , the active units in layer C6 exhibited the highest overall selectivity . Phonetic feature categories are discrete acoustic parameters . We next investigated the encoding of continuous acoustic features that specify phonemes , including the fundamental frequency ( F0 ) , formant frequencies ( F1 , F2 ) , voice-onset time ( VOT ) , and spectral peak . We used linear regression to decode these features from the response amplitudes of model units ( Materials and Methods ) . Because F0 , F1 , and F2 vary significantly across vowels , whereas VOT and spectral peak vary significantly across consonants , we separately decoded F0 , F1 , and F2 of vowels ( Fig 6A ) and the VOT and spectral peak of consonants ( Fig 6B ) from the responses of all active units in layer C6 . A 20-fold validation scheme was used to predict each parameter . The prediction accuracies on the test sets ( 1-fold ) were defined based on the regression errors on the corresponding training sets ( 19-fold ) . The prediction accuracies for each parameter were significantly higher than those of a random decoder ( p<10−5; Materials and Methods ) . These observations suggest that the variability of these acoustic features is well represented in the responses of the units . Similar prediction accuracies were obtained based on the neural population responses in human superior temporal gyrus [6] . The same linear regression method was applied to decode the acoustic parameters F0 , F1 , F2 , VOT , and spectral peak from the cochleogram . We cut a length of 170 ms cochleogram for each phoneme instance and used it as the feature of this instance . All decoding accuracies were about 10% , much lower than those obtained from the responses of layer C6 units ( Fig 6A and 6B ) . In higher layers of the model , the units were always clustered into six phoneme groups according to PSI ( Fig 5; S5 Fig ) . Therefore , we could calculate the decoding accuracies for each group of units and compare the results in different layers . In layer C6 , we obtained significantly higher accuracies than did a random decoder ( Fig 6C ) ( p<10−5 ) . Similar results were obtained in layer S6 ( Fig 6D ) . However , the decoding accuracies were very low in layers S5 and C5 , and even lower in the earlier layers S4 and C4 . These poor accuracies were partly due to the small STRFs of units in these layers , which contain less information about the acoustic features of a phoneme . Fig 6D shows a small improvement from layer S ( l ) to layer C ( l ) , but a large improvement from layer C ( l ) to layer S ( l+1 ) . This is because STRF sizes were similar between units in layer S ( l ) and layer C ( l ) , but very different between units in layer C ( l ) and layer S ( l+1 ) ( Table 2 ) . Thus far , we have described the encoding of static acoustic features of phonemes in higher layers of the model . However , phonemes are not static . In fact , the first two formants ( F1 and F2 ) of the phonemes , especially the consonants , exhibited large variation over time ( Fig 7 ) . We investigated the encoding of the dynamic formants using the responses of the units . We defined the temporal variation index ( TVI ) of individual phonemes as the projections of their F1 or F2 contours ( time course of formant frequencies averaged over all instances of a phoneme ) onto their respective principal components over all phonemes ( Materials and Methods ) . Therefore , TVI measures the matching degree of an individual formant contour to the principal component of all formant contours . For each unit , we calculated the correlations between its responses to phonemes and the TVIs of the phonemes . A high correlation indicated that the unit was sensitive to the TVI of the phonemes—it “liked” phonemes whose F1 or F2 contour were congruent to the principal component but “disliked” phonemes whose F1 or F2 contour were incongruent to the principal component . Some units in layer C6 were sensitive to either F1 or F2 TVI , whereas others were sensitive to both F1 and F2 TVI ( Fig 7C ) . However , such units were scarce in layers S5 , C5 , and S6; in fact , for most of the units in these layers , the correlations between their responses and both F1 and F2 TVI were smaller than 0 . 5 . The hallmark of the sparse coding model [20 , 38] is the sparse activity of the hidden units , which has been proven to be essential in reproducing the tuning properties of auditory nerve fibers [22] and neurons in the inferior colliculus [23] . However , it remains unclear whether sparse activity in higher layers of SHMAX also plays a significant role in producing the phoneme encodings that we observed . Hence , we investigated how the sparseness level influenced the results in layer S5 to layer C6 . First , we adjusted the parameter λ in the sparse coding model , which controls the sparseness of the responses of the units of a particular S layer , while keeping λ in other layers at the default value of 1 . We found that sparseness and F-ratio were positively correlated in each layer ( Fig 8A ) , i . e . , the sparser the activity , the more selective the units . However , increasing the sparseness level in lower layers such as S5 and C5 did not lead to phoneme tuning as strong as that in C6 . This suggests that sparseness was not the only factor that led to the strong phoneme-encoding property in the top layer , and that the hierarchical structure also played a significant role . Although we used the lifetime sparseness measure [39] ( Materials and Methods ) here , the conclusion did not change when the population sparseness measure [40] was applied . Second , we modulated the neural encoding of acoustic parameters with different sparseness levels . The results revealed that sparseness also played a key role in producing the neural population coding of acoustic parameters ( Fig 8B ) . The other critical element in the model is max pooling . Without this element , the model would be almost linear because the only nonlinear operation is the down-sampling between layer S1 and layer S2 , which is implemented by a convolution with stride 2; one would not expect such a model to produce striking results . To examine the influence of pooling , first , we removed the max pooling layers and trained the model as before . Under these conditions , the STRFs of layer S2 and layer S3 units exhibited much simpler patterns ( Fig 9A and 9B ) , and the complex patterns obtained with max pooling ( Fig 3 ) were scarce . More importantly , the higher layers did not exhibit similar results to those of field recordings in human auditory cortex ( e . g . , compare Fig 9C with S5C Fig ) . Second , we replaced all max pooling in the model with average pooling , another widely used operation in the deep learning field . The only difference is that , in average pooling , we take the average of values , instead of the maximum value , in a region . After training the model , we obtained poor results ( e . g . , compare Fig 9D and Fig 5A ) . These results suggest that max pooling played an important role in modeling the function of the auditory pathway . In addition to sparse response and max pooling , two other factors are also important for the emergence of phoneme selectivity: the number of layers and STRF size of the units . This is because enough nonlinearity must be accumulated along the hierarchy through repeated max pooling and down-sampling , and the STRF must be large enough to cover the length of the phonemes . To better understand the roles of these two factors , we explored models with different architectures . First , we tried a different layer S5 with larger kernel size ( 20×20 ) , whose STRF size was the same as in the original layer S6 . The selectivity of the new layer S5 was weaker than that of the original layer S6 ( S6A Fig; Fig 10 ) . Because the nonlinearity of the model grows with ascending layers , this result indicates that the emergence of phoneme selectivity needs a certain degree of nonlinearity . Second , we replaced the original layer S6 ( kernel size 10×10 ) with two new layers , S6 and S7 , with kernel size 5×5 . In comparison with the original layer S6 , the phoneme selectivity of the new layer S6 was weaker , and that of the new layer S7 was similar ( S6B and S6C Fig; Fig 10 ) . The results indicate that STRF size is also important for the emergence of phoneme selectivity . Notice that there is an abrupt increase in the decoding accuracy of acoustic parameters from original layer S5 to layer S6 , as observed in Fig 6D . This is mainly because the STRFs of layer S5 units were not large enough to capture the selectivity information , while those of layer S6 units happened to be large enough . Fig 10 shows that , if the STRF size in original layer S5 is increased ( green ) , or the STRF size in original layer S6 is decreased ( red ) , the increase in decoding accuracy from the original layer S5 to layer S6 would not be as large as observed in Fig 6D .
Merely reconstructing the stimuli without sparseness regularization would not lead to biologically reasonable results . Indeed , we observed that , upon replacing the L1 regularization term in Eq ( 1 ) , which encourages sparse activities in the units , with the L2 regularization term λ∑k‖rk‖22 , the STRFs of lower-layer units had fewer semantic features , and the phonetic feature selectivity was poor even at layer C6 . This result agrees with a recent study [37] in which two unsupervised deep neural networks trained on natural speech without this regularization exhibited no clear selectivity for phonetic features . A previous study [23] using a single-layer sparse coding model reported qualitatively similar results to those obtained from the lower layers of SHMAX . The work described here extends that study in two ways: first , by setting a more biologically reasonable temporal window for learning the response properties of inferior colliculus neurons in auditory midbrain; and second , by introducing hierarchical structure for learning the results of field recording in the auditory cortex . By using control experiments , we have found that max pooling in the hierarchical structure also plays an important role in producing the results . Because the same principles , sparse coding and max pooling , are employed in different layers of the model , the results suggest that these principles might underlie computation at multiple stages of the auditory system . This is in agreement with the assumption that the auditory system evolved to optimize the representation of natural sounds [42] , of which human speech is an example . Natural sounds contain many forms of higher-order and nonlinear statistical regularities [22 , 42 , 43] , and sparse coding and max pooling are capable of extracting such regularities from natural images [28 , 44 , 45] . In this work , we demonstrate that these approaches can also extract them from human speech . Psychological studies have reported that infants are capable of distinguishing phonemes in their native language , even though they are not explicitly trained to accomplish this task [46 , 47] . Inspired by this finding , many computational models have been proposed to investigate how phoneme categories can be formed from continuous speech through unsupervised learning ( e . g . , [48–50] ) . However , these single-layer models cannot reveal how phoneme information is encoded gradually along the auditory pathway . Aided by some side information , such as lexical information and the identity of the speaker , deep neural networks can learn phonetic features from speech [51 , 52] . A critical step in these systems is time alignment of frames , which complicates the learning process and makes it hard to interpret the learning process of infants . Recently , two hierarchical models , the deep belief network and auto-encoder network , were trained on unlabeled speech corpus , but failed to obtain increasing phoneme selectivity in ascending layers [37] . By contrast , after supervised training with phoneme labels , the multilayer perceptron exhibited increasing selectivity to phoneme classes in higher layers [37] , and its top layer exhibited a feature organization pattern similar to that of human auditory cortex [24] . Nevertheless , this supervised learning model cannot explain how phoneme representation in the human brain is developed during infancy . In this study , we showed that phonetic feature representation can emerge in an unsupervised learning model trained on continuous speech data without relying on any side information . Our results oppose the view that phoneme learning occurs after [51 , 52] or concurrent with lexical learning and that the two processes cannot be addressed in isolation [53–55] . The model makes two predictions that could be explored in future experiments . The first is that the emergence of selectivity for phoneme features along the auditory pathway is not abrupt; instead , it should be a continuous process . In our model , the selectivity emerged in several layers other than the top layer , although it was weaker in lower layers such as S5 . Therefore , one may find neurons selective for phoneme features in subcortical areas , but their selectivity should not be as strong as in the auditory cortex . The second prediction is that there exist neurons in the auditory cortex that encode the formant dynamics of phonemes . As shown in Fig 7 , some layer C6 units in the model were sensitive to variation in the first formant of the phonemes , the second formant , or both . This is parallel to the findings in the higher-level cortical areas in the visual pathway , in which neurons encode abstract features such as contours [56 , 57] . The fact that such units were scarce in layers S5 to C5 suggests that neurons whose responses are correlated to formant dynamics are not common in subcortical areas . This study has some limitations . First , because SHMAX is not a biologically detailed model , it is unclear how it could be implemented in a biological system . Some neural circuits have been proposed to individually realize the two essential components of the model , i . e . , sparse coding [58] and max pooling [59 , 60] , but an approach for integrating them as a whole is still lacking . Second , this model differs from biological systems in many aspects . For example , real neurons can be excitatory or inhibitory , and they obey Dale’s law [61] , but these features are not considered in SHMAX . In addition , the auditory system contains abundant feedback and recurrent connections , whereas SHMAX is a feedforward model . Due to these differences , the results in this paper should be interpreted cautiously . Although such discrepancies may not affect the abstract computational principles of the auditory system revealed by the model , a more biologically plausible model would make the results more convincing .
Speech stimuli from the Texas Instruments/Massachusetts Institute of Technology ( TIMIT ) database [62] , which includes 6 , 300 sentences spoken by 630 speakers ( 10 sentences per speaker ) from eight major dialect regions of the United States , were used . The sample rate is 16 kHz . The sentences were first converted into cochleogram [9 , 63] , which is similar to a spectrogram but better reflects the effects of the cochlea . A total of 194 frequency filters were generated from a cochlear model [9] whose center frequencies were between 73 and 7 , 630 Hz . The sample step was 1 ms . The cochleogram was used as the input for the SHMAX model . SHMAX [28] is an unsupervised deep learning model that integrates sparse coding into a well-known cortex-inspired visual recognition model , HMAX [64] . The structure used in this study ( Fig 1 ) consists of six S layers and six C layers , which respectively perform sparse coding and max pooling , in alternation . We implemented the feedforward calculation of the model in the same manner as in a convolutional neural network [65 , 66] . The major difference is that the convolutional kernels were not learned by supervised learning but by unsupervised learning , specifically sparse coding ( see below ) . Another difference is that we did not use nonlinear activation functions as in standard convolutional neural networks . The details are as follows . Sparse coding is an unsupervised learning technique inspired by sparse firing of V1 simple cells [20 , 38] . Given a set of input signals xk ∈ Rn , where k indexes the input , the objective of sparse coding is to find a set of bases bj ∈ Rn such that xk=Σj=1mrjkbj+σk , where rjk is the weighting factor of bj and σk ∈ Rn is noise . The factor rjk is called the “response” of neuron j to the k-th input , whose receptive field is delineated by bj . The critical requirement of this technique is that rjk values are sparse ( i . e . , only a few of them are nonzero ) . A standard formulation of sparse coding is minimizeB , rk∑k‖xk−Brk‖22+λ‖rk‖1 ( 1 ) subjectto‖bj‖22≤1 , j=1 , 2 , … , m , where B ∈ Rn×m is the collection of bj , rk ∈ Rm is the collection of rjk , and λ is a constant controlling the tradeoff between the reconstruction error ( the first term ) and the sparseness ( the second term ) . ‖⋅‖2 and ‖⋅‖1 stand for the L2-norm and L1-norm of vectors , respectively . All results reported in this paper , except in Fig 8 , were obtained with λ = 1 . Without loss of generality , it was assumed that the STRFs of all units in the model were square . The bases of sparse coding in each S layer were learned from a number of 10 ×10 × u patches extracted randomly from the input of that layer ( therefore , the size of each basis was also 10 ×10 × u ) , where u denotes the number of the input channels . In layer S1 , u was equal to 1 , and in other S layers it was equal to the number of bases in the preceding C layer . The online dictionary learning algorithm was used to learn B [67] . The response of each S layer could be obtained by solving Eq ( 1 ) with learned B by inputting a sliding 10 ×10 × u patch in the previous layer as input [28] . This will result in a total of m feature maps consisting of responses of m bases bj at every location in the previous layer . In this study , we used another approach: each basis bj was convolved with the input to obtain the corresponding feature map . In doing so , bj was reshaped to 10 ×10 × u and convolved with the previous layer whose size was h × t × u , where h and t denote the height and width of the feature maps , then the j-th feature map ( a 2D matrix ) in the current layer was obtained as follows: rS ( h^ , t^ , j ) =∑p=09∑q=09∑u^=0u−1x ( sconv⋅h^+p , sconv⋅t^+q , u^ ) ⋅bj ( p , q , u^ ) ( 2 ) where sconv is the convolution stride . This approach yielded similar results but was much faster . Note that the difference between the outputs of the two approaches is unimportant because the responses obtained by the convolution approach are also sparse ( S7 Fig ) , and sparseness is the focus of this study . In addition , in the convolution approach it is natural to define bj as the receptive field of an artificial neuron because bj is used to explain the output , whereas in the optimization approach the correspondence between bj and receptive field is indirect because bj is used to explain the input . The convolution stride sconv was 2 in the first two S layers ( equivalent to vanilla convolution followed by down-sampling with ratio 2 ) and 1 in the other four S layers ( vanilla convolution ) . The value u in each S layer is indicated at the top of Fig 1C . The C layers take the responses of S layers as input and perform the max pooling operation . We slide on the feature maps in an S layer and at each location take the maximum value in a region of size spool × spool centered at that location ( for simplicity , square shapes of the regions are assumed ) . All maximum values taken in a feature map in the S layer then constitute a feature map in the subsequent C layer . Therefore , the number of feature maps in a C layer is equal to that in the preceding S layer . Clearly , the output ( or response ) of a C unit is the maximum response of some S units in a local region . The pooling results were input to the next S layer . In this study , we used overlapping pooling ( stride of 1 ) , and the feature maps in a C layer had one fewer column and one fewer row than those in the preceding S layer . Specifically , the calculation is as follows: rC ( h^ , t^ , j ) =max{rC ( h^:h^+spool−1 , t^:t^+spool−1 , j} ( 3 ) where spool = 2 . The STRFs of the units in layer S1 were approximated by the corresponding bases . The STRFs of the units in higher layers were obtained by linearly combining the bases of the units in previous layers [28] . The logic is that a unit in the current layer is locally connected to the units in the previous layer , and the favorite input pattern ( spectro-temporal receptive field , STRF ) of this unit depends on the favorite input patterns of the units in the previous layer . The bottom-up computation process is described as follows . First , it should be noted that the units in a given feature map share the same STRF . For a unit in the j-th feature map in layer S ( l ) , where l>1 , its STRF is defined as the weighted sum of the STRFs of the units in layer S ( l − 1 ) with their centers aligned according to the locations of the weights in the basis bj . If there is no down-sampling between layers S ( l ) and S ( l − 1 ) , this can be implemented by convolving bj with all of the ul−1 STRFs in layer S ( l − 1 ) . Note that the third dimension of bj is ul−1 , the number of feature maps in layer S ( l − 1 ) , and the result is a 2D matrix , which is the STRF of the units in the j-th feature map in layer S ( l ) . If there is a down-sampling operation with ratio d between layers S ( l ) and S ( l − 1 ) , one needs to first expand bj to 10d × 10d × ul−1 , and then perform the convolution as described above . The nearest-neighbor interpolation was used for matrix expansion . See Fig 2A for an illustration of visualizing an S2 basis ( STRF of an S2 unit ) . For a better illustration , in this example the first two dimensions of the S1 and S2 bases are assumed to be 3 and 2 , respectively , instead of 10 . Note that the weighted summation step in the figure is equivalent to 2D convolution ( “full” mode ) of the two input matrices . After obtaining STRFs of all S2 units , one can calculate STRFs of S3 units in the same way , and so forth . One can also calculate STRFs of units in any S layer directly , without precomputing the STRFs of units in lower S layers ( S1 Fig ) . This is equivalent to the bottom-up method ( Fig 2A ) , except for some differences in the boundaries of STRFs . The latter method was used to visualize the STRFs of S units in this study . Since a C unit takes the maximum response of four neighboring units in the preceding S layer ( the max pooling ratio was 2 ) whose STRFs are the same , its STRF is very similar to the STRFs of the preceding S units except that it is a bit larger . Its size is jointly determined by the size of the STRFs of the four preceding S units and the shifts between them . The STRF sizes of the units in each layer are shown in Table 2 . Because all STRFs are assumed to be square , only the side lengths are shown in the table . To test the validity of the visualization method described above , the STRFPak toolbox with the normalized reverse-correlation method [32 , 33] was also used to calculate the STRFs of the units in layers S1 , S2 , and S3 ( S3 Fig ) ; this took much longer time than the linear combination method described above . The time lag used for calculating the stimulus auto-correlation was 200 ms , the tolerance value was 0 . 01 , and the sparse parameter was 0 . The overall mean firing rate was removed from the neuronal response , and the space-time non-separable algorithm was used . To compare the model and experimental results [30 , 34] , singular value decomposition ( SVD ) was performed on the obtained STRFs , and the two unitary vectors corresponding to the first singular value were used to quantify the spectral and temporal response characteristics , namely the spectral and temporal profiles ( Fig 3 ) [30] . The peaks of the spectral and temporal profiles determine the center frequency ( Center F ) and best temporal modulation frequency ( Best T ) respectively . The widths of the spectral and temporal profile that account for 90% of the total energy determine the bandwidth and the duration respectively . To calculate the responses of a computing unit to a particular phoneme in speech , the TIMIT phonetic transcriptions were used to align responses to the onsets of all instances of the phoneme . Phoneme length was not normalized . The maximum absolute value of the response of a unit along the phoneme duration was defined as that unit’s response amplitude . The PSI vectors [6] were employed to characterize the selectivity of the units to phonemes . The method is briefly outlined as follows , and more details can be found in Ref . [6] . For every unit , the distribution of its response amplitudes across all samples of each phoneme was estimated first . To calculate a unit’s PSI for a particular phoneme , the non-parametric Wilcox rank-sum test was used to determine whether the response amplitude distribution of the phoneme had a larger median than those of other phonemes ( p<0 . 01 ) . The number of phonemes whose median response amplitudes were statistically smaller than the median response amplitude of a particular phoneme was defined as the unit’s PSI for that phoneme . Because 33 phonemes were selected from the dataset , PSI ranges between 0 and 32 , where 0 means no selectivity and 32 means extreme selectivity . The PSIs for all 33 phonemes form a 33-dimensional PSI vector for the unit . Active units whose responses to randomly selected time frames were statistically larger than their response to silence ( p<0 . 001 ) were selected , and the F-ratio [24 , 37] was used to quantify the overall phoneme selectivity of all active units in a layer ( Table 1 ) . The units were grouped based on the clustering of PSI vectors ( see Fig 5C for an example ) . Suppose there are m active units in total , which form n groups in a certain layer . Let Ωj denote the set of indices of active units in group j , and |Ωj| = mj . The F-ratio for that layer is defined as the ratio of between-group variability to within-group variability: F=∑j=1nmj||p−j−p−||22/ ( n−1 ) ∑j=1n∑i∈Ωj||pi−p−j||22/ ( m−n ) ( 4 ) where pi denotes the PSI vector for unit i , p− denotes the average of PSI vectors over all units , and p−j denotes the average of PSI vectors over group j . The larger the F-ratio , the better the clustering effect . Six distinctive features were used to describe the acoustic properties of each phoneme [68] . Each phoneme has only one of the six features . The PSI vectors of all phonemes that shared a particular feature were averaged to describe the feature selectivity of the units ( Fig 5D ) . A series of static acoustic parameters were estimated for phonemes that play a perceptually important role in speech perception , including F0 , F1 , F2 , VOT , and spectral peak . The values of the first three parameters were calculated as the median value of transcribed boundaries over the duration of the phoneme [6] . The VOT was extracted as the phoneme transcription boundary . The spectral peak was defined as the maximum energy along the frequency axis in the cochleogram . Acoustic parameters differ among individual instances of a phoneme; therefore , each acoustic parameter for each phoneme in the dataset is expressed as a distribution . Similar to [6] , a linear model y = wx + b was used to regress the static acoustic parameters y such as the fundamental frequency ( F0 ) , formant frequencies ( F1 , F2 ) , voice-onset time ( VOT ) , and spectral peak , based on the response amplitudes of the computing units x , where w and b are parameters to be learned . The least-square error between the prediction y and the ground truth y* was minimized on a training set , and the trained model predicted the parameter values on a test set . The root-mean-squared error was calculated on the training set . A prediction for a test sample was regarded as “correct” if the prediction error was smaller than the root-mean-squared error . The percent of correct prediction on the test set was defined as the testing or decoding accuracy . The acoustic parameters regressed were F0 , F1 , and F2 of vowels and VOT and spectral peak of consonants . Fig 6A and 6B presents the decoding results using the responses of all active units in layer C6 . Fig 6C and 6D presents the decoding results using the responses of active units , in each of the six phoneme groups separately , in different layers . For each task , a 20-fold cross-validation scheme was adopted , and therefore 20 decoding accuracies were obtained . To conduct a significance test , a random decoder was constructed . Given a test sample , the decoder output a value between the minimum and maximum values of the ground truth y* on the training samples , with a uniform probability . This random prediction was evaluated as correct or not based on the same criterion used for linear prediction . The 20-fold cross-validation scheme resulted in 20 chance-level accuracies . Student’s t-test was performed to compare the two sets of accuracies . The time course of the first formant frequency of a phoneme instance was called the F1 contour of that instance ( Fig 7A ) . The averaged F1 contours over all instances of a phoneme was defined as the F1 contour of the phoneme . Since 33 phonemes were selected in the dataset , we obtained 33 F1 contours . Principal component analysis ( PCA ) was performed on these contours . The first principal component , which had the same length as the F1 contour , was calculated ( Fig 7B ) . The projection of the F1 contour of each phoneme onto the first principal component ( a scalar ) was defined as the F1 temporal variation index ( TVI ) of that phoneme . A unit’s average response to all instances of a phoneme ( a scalar ) was defined as its response to that phoneme . The encoding of F1 dynamics in a unit was measured as the correlation between the unit’s responses to all of the 33 phonemes and the F1 TVIs of those phonemes . The same procedure was applied to measure the encoding of F2 dynamics in a unit . The definition of lifetime sparseness of a unit is as follows [39]: S=1− ( E[r] ) 2E[r2] ( 5 ) where r denotes the response of the unit and the expectation is taken across all test data . | When speech enters the ear , it is subjected to a series of processing stages prior to arriving at the auditory cortex . Neurons acting at different processing stages have different response properties . For example , at the auditory midbrain , a neuron may specifically detect the onsets of a frequency component in the speech , whereas in the auditory cortex , a neuron may specifically detect phonetic features . The encoding mechanisms underlying these neuronal functions remain unclear . To address this issue , we designed a hierarchical sparse coding model , inspired by the sparse activity of neurons in the sensory system , to learn features in speech signals . We found that the computing units in different layers exhibited hierarchical extraction of speech sound features , similar to those of neurons in the auditory midbrain and auditory cortex , although the computational principles in these layers were the same . The results suggest that sparse coding and max pooling represent universal computational principles throughout the auditory pathway . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"auditory",
"cortex",
"linguistics",
"medicine",
"and",
"health",
"sciences",
"brain",
"social",
"sciences",
"neuroscience",
"auditory",
"pathway",
"computational",
"neuroscience",
"coding",
"mechanisms",
"inferior",
"colliculus",
"sensory",
"physiology",
"animal",
"cells... | 2019 | A hierarchical sparse coding model predicts acoustic feature encoding in both auditory midbrain and cortex |
The budding yeast Srs2 is the archetype of helicases that regulate several aspects of homologous recombination ( HR ) to maintain genomic stability . Srs2 inhibits HR at replication forks and prevents high frequencies of crossing-over . Additionally , sensitivity to DNA damage and synthetic lethality with replication and recombination mutants are phenotypes that can only be attributed to another role of Srs2: the elimination of lethal intermediates formed by recombination proteins . To shed light on these intermediates , we searched for mutations that bypass the requirement of Srs2 in DNA repair without affecting HR . Remarkably , we isolated rad52-L264P , a novel allele of RAD52 , a gene that encodes one of the most central recombination proteins in yeast . This mutation suppresses a broad spectrum of srs2Δ phenotypes in haploid cells , such as UV and γ-ray sensitivities as well as synthetic lethality with replication and recombination mutants , while it does not significantly affect Rad52 functions in HR and DNA repair . Extensive analysis of the genetic interactions between rad52-L264P and srs2Δ shows that rad52-L264P bypasses the requirement for Srs2 specifically for the prevention of toxic Rad51 filaments . Conversely , this Rad52 mutant cannot restore viability of srs2Δ cells that accumulate intertwined recombination intermediates which are normally processed by Srs2 post-synaptic functions . The avoidance of toxic Rad51 filaments by Rad52-L264P can be explained by a modification of its Rad51 filament mediator activity , as indicated by Chromatin immunoprecipitation and biochemical analysis . Remarkably , sensitivity to DNA damage of srs2Δ cells can also be overcome by stimulating Rad52 sumoylation through overexpression of the sumo-ligase SIZ2 , or by replacing Rad52 by a Rad52-SUMO fusion protein . We propose that , like the rad52-L264P mutation , sumoylation modifies Rad52 activity thereby changing the properties of Rad51 filaments . This conclusion is strengthened by the finding that Rad52 is often associated with complete Rad51 filaments in vitro .
Homologous recombination ( HR ) is fundamental for the repair of DNA double-strand breaks ( DSBs ) . It is also involved in the error-free fill-in of single-strand gaps generated by replication fork stalling or incomplete DNA repair . Defects in HR are associated with many cancers , both hereditary and sporadic [1] , which underlines the essential nature of this process . The mechanisms and proteins involved in HR have been well conserved throughout evolution and much of our knowledge on HR comes from studies conducted with the yeast Saccharomyces cerevisiae ( S . cerevisiae ) ( reviewed in [2] , [3] ) . HR involves the interaction of a 3′-single stranded DNA ( ssDNA ) end with a homologous double-strand DNA ( dsDNA ) molecule , which is used as a template for DNA synthesis . In eukaryotes , the recombinase Rad51 forms a nucleoprotein filament on the ssDNA which undergoes synapsis and strand invasion of the homologous duplex DNA to form a stable D-loop ( reviewed in [4] ) . However , the presence of replication protein A ( RPA ) previously bound to ssDNA prevents Rad51-mediated strand exchange in vitro . This inhibition is overcome by the addition of Rad52 or the Rad55-Rad57 heterodimer ( the Rad51 paralogs of S . cerevisiae ) , defining these proteins as Rad51 filament mediators ( reviewed in [4] ) . Rad52 exhibits the greatest Rad51 filament mediator activity in S . cerevisiae . Via its interaction with both RPA and Rad51 , it stimulates the removal of RPA and recruits Rad51 to DNA ( reviewed in [4] ) . This mediator function is highlighted by the severe phenotypes caused by null mutations of the RAD52 gene: γ-ray sensitivity and highly reduced levels of both mitotic and meiotic HR ( reviewed in [3] ) . The Rad52 protein also has the capacity to anneal homologous ssDNA in vitro [5] . This activity is involved in Rad51-independent single-strand annealing ( SSA ) [6] and possibly in the capture of the second end of a DSB to generate double Holliday junction ( dHJ ) intermediates [7]–[9] . The S . cerevisiae Rad52 protein is subject to post-translational modifications but it is unclear how these modulate Rad52 activities . Rad52 is constitutively phosphorylated throughout the cell cycle on some serine and/or threonine residues and additional phosphorylations are induced specifically in S phase [10] . The phosphorylated residues have not yet been identified . Rad52 also undergoes sumoylation at lysines 10 , 11 and 220 after exposure to DNA-damaging agents that induce DSBs . This modification depends on the SUMO-conjugating enzyme Ubc9 and on the SUMO-ligase Siz2 [11] . ssDNA accumulation is necessary for sumoylation of Rad52 [11] , [12] . It has been reported that loss of Rad52 sumoylation decreases protein stability without significantly affecting HR levels or recruitment of Rad52 to DNA damage [11] , [13] . Sumoylation appears to facilitate the exclusion of Rad52 recombination foci from the nucleolus to maintain a low level of recombinational repair at the ribosomal gene locus [14] . Recently , it was shown that Rad52 , RPA and Rad59 are modified by a sumoylation wave leading to simultaneous multisite modification . Catalyzed by a DNA-bound SUMO ligase , this wave stabilizes physical interactions between the proteins [15] . However , Rad52 sumoylation might also restrict Rad51 filament formation through the SUMO-targeted Cdc48 segregase that can curb Rad52-Rad51 physical interaction and displace these proteins from DNA [16] . How exactly phosphorylation and sumoylation of Rad52 affect Rad51 filament formation remains to be determined . Contrasting with its primordial role in DNA repair , HR may lead to potentially lethal intermediates . This was first revealed by the study of the Srs2 helicase , a major actor in the avoidance of such intermediates ( reviewed in [17] ) . UV sensitivity of srs2Δ cells is suppressed by the ablation of the Rad51 protein [18] , which suggests that Srs2 is required for the elimination of toxic UV-induced recombination structures . Furthermore , it was shown that leaky alleles of RAD51 or RAD52 , which form abortive recombination intermediates , trigger Srs2 activity [19] , [20] . Finally , negative interactions between srs2Δ and genes involved in DNA replication or recombination , such as sgs1Δ , rad54Δ or mrc1Δ , are rescued by mutations preventing or altering HR ( rad51Δ , rad52Δ , rad55Δ and rad57Δ ) 21–23 . Thus , it was concluded that ablation of these genes can also induce the formation of lethal recombination intermediates normally eliminated by Srs2 . All these studies indicate that a key role of Srs2 is to avoid the formation of lethal structures induced by HR . Several studies point out that Rad51 filaments on ssDNA could be the lethal structures eliminated by Srs2 . First , in vitro experiments show that Srs2 can disrupt Rad51 filaments thanks to its translocase activity [24] , [25] . Second , srs2Δ strains show a three- to four-fold increase in the number of budded cells that contain a Rad51 or Rad54 focus compared with wild-type ( WT ) cells [26] . Finally , it has also been reported that some of the srs2Δ phenotypes , like the co-lethality with rad54Δ or the high sensitivity to a persistent DSB , depend at least partially on the DNA damage checkpoint [27] , [28] . Unproductive Rad51 filaments could induce this persistent checkpoint response . Srs2 is also necessary to complete DSB repair by HR in haploid cells , when the homologous sequence is located on another chromosome . In this context , the low viability of srs2Δ cells is associated with a strong increase in the level of crossing-over ( CO ) associated with gene conversion [29] , [30] . This suggests that Srs2 avoids the formation of CO by promoting synthesis-dependent strand annealing [31] . This is supported by recent in vivo work showing that Srs2 promotes the formation of non-crossing over products mostly through its helicase activity [32] , possibly by dismantling nicked HJs . Additionally , the increased sensitivity to UV and γ-ray irradiation of srs2Δ homozygous diploid cells compared with haploids was proposed to be related to the resolution of specific interactions between homologous chromosomes , probably related to HR [33] . Altogether , these data suggest that lethal recombination structures eliminated by Srs2 could also be intertwined recombination intermediates . To gain insight into the nature of lethal recombination intermediates eliminated by Srs2 , we searched for mutants that suppress the sensitivity of srs2Δ cells to the radiomimetic DNA alkylating agent methyl methanesulfonate ( MMS ) . This screen was designed to select against mutations in genes that are essential for HR because they generally confer high sensitivity to this drug . Yet we found an allele of RAD52 ( rad52-L264P ) that can completely suppress sensitivity to MMS in srs2Δ cells . Our extensive analysis indicates that rad52-L264P suppresses defects in srs2Δ cells associated with presynaptic Rad51 filaments rather than with the resolution of recombination intermediates . This suppression is related to a modulation of Rad52 mediator activity . We also show that , like rad52-L264P , Rad52 sumoylation leads to the suppression of srs2Δ cells defects , suggesting that sumoylation modulates Rad52 mediator activity .
To characterize factors involved in the formation of toxic recombination intermediates eliminated by Srs2 , we selected mutants that suppress the MMS sensitivity of srs2Δ haploid cells by plating them on rich medium containing 0 . 015% MMS . Several well-grown colonies were isolated and backcross analyses showed that the suppressor mutations responsible for MMS resistance were all monogenic . Genetic analyses showed that they affected different genes . We genetically mapped one of these mutations at 25 cM from the PIF1 gene . Around this position , we found RAD52 to be a good candidate since it was involved in DNA repair . The sequence of the RAD52 gene showed a single T to C transversion at position 790 of the open reading frame ( ORF ) , leading to a change of leucine 264 to a proline . To confirm that this mutation is solely responsible for the phenotype , directed mutagenesis was used to create an integrative plasmid which was introduced by gene replacement in a new srs2Δ strain using the pop-in pop-out technique . MMS sensitivity suppression was equivalent to that observed in the original suppressed strain ( Figure 1A ) . Rad52 is composed of a highly conserved N-terminus that forms ring structures [34] , [35] . Outside the N-terminus , the protein is less conserved . Sequence alignments show that L264 belongs to a stretch of well-conserved amino acids ( positions 261–283 ) in Hemiascomycetes , located outside the highly conserved N-terminus ( Figure 1B ) . L264 is located ten residues upstream of the QDDD motif essential for RPA binding [36] , ( Figure 1B ) . Note that while rad52-L264P is only slightly sensitive to MMS ( Figure 1A ) , rad52-Q275A/D276A/D277A/D278A is a null allele [36] . We wondered if the change to a proline and not the loss of the leucine itself was significant for the suppression phenotype . Therefore , we replaced the leucine with an alanine by directed mutagenesis . We found that rad52-L264A suppresses the MMS sensitivity of srs2Δ cells , but not as well as rad52-L264P ( Figure 1A ) . Therefore , even if changing the leucine to a proline has a more radical effect , probably because it affects the domain more extensively , it is the leucine ablation that confers the suppression phenotype . To characterize the effect of the mutation on Rad52 activities , we constructed a rad52-L264P single mutant strain . rad52-L264P is not sensitive to incubation at 16°C or 37°C ( Figure S1A ) and , unlike the deletion of RAD52 , it does not affect the rate of spontaneous mutagenesis in the CAN1 gene ( Figure S1B ) . The growth rate is also unchanged compared with WT ( 90 minutes ) , while the deletion of RAD52 leads to a significant increase of the doubling time ( 155 minutes ) . Surprisingly , cells carrying this mutation are strongly resistant to MMS , while rad52Δ makes cells extremely sensitive ( Figures 1A and 9D ) . rad52-L264P cell resistance to γ-ray and to UV is also comparable to that of WT cells ( Figure 2A and B ) . UV-induced recombination between his7-1 and his7-2 heteroalleles in rad52-L264P homozygous diploid cells is indistinguishable from that of WT cells ( Figure 2D ) , suggesting that gene conversion is not affected by this mutation . Altogether , these data show that rad52-L264P does not substantially affect the activity of the protein . rad52-L264P completely suppresses srs2Δ haploid cell growth defect on MMS plates ( Figure 1A ) . It also suppresses srs2Δ γ-ray and UV sensitivities ( Figure 2A and B ) . UV induces mostly single strand gaps , while γ-ray and MMS produce also DSBs . Therefore , the deadly recombination intermediates induced by both kinds of lesions in srs2Δ cells are not toxic or are not formed in the rad52-L264P srs2Δ background . rad52-L264P also completely suppresses the previously described UV-induced srs2Δ hyper-recombination phenotype [33] , [37] . Frequencies of UV-induced recombination in diploid cells measured between his7-1 and his7-2 heteroalleles are the same in rad52-L264P srs2Δ and in WT cells ( Figure 2D ) . Interestingly , the UV sensitivity and the hyper-recombination phenotype of srs2Δ/srs2Δ diploid cells were only partially suppressed when rad52-L264P and RAD52 alleles were co-expressed in comparison with homozygous rad52-L264P diploids ( Figure 2E ) . Therefore , the WT and rad52-L264P alleles of RAD52 are co-dominant , which implies that rad52-L264P cannot be a simple hypomorphic allele . The Rad52-L264P protein mediates the formation of enough functional Rad51 filaments for HR and DNA repair to occur without leading to accumulation of deadly recombination intermediates in srs2Δ cells . Deleting numerous genes involved in DNA replication or recombination can also induce the formation of lethal recombination intermediates in the absence of Srs2 . The deletion of those genes reveals negative interactions with srs2Δ in a RAD51-dependent manner . Some of these genes , like RAD50 and RAD54 , are involved in the normal maturation process of recombination intermediates [22] , [27] , [38] . Another set of genes is involved in DNA replication . Among them are RRM3 [39] , MRC1 and CTF18 [23] . The srs2Δ mutation is also synthetically lethal with sgs1Δ , but Sgs1 is involved in recombination and potentially in replication . Therefore , it is not clear which function of Sgs1 is required to avoid srs2Δ death [40] . We wondered if some of these negative interactions would be suppressed by rad52-L264P . A rad52-L264P srs2Δ strain was crossed with strains containing deletions of the genes interacting negatively with srs2Δ ( Figure 3A ) . Interestingly , tetrad analysis showed that all the negative interactions we tested were suppressed by rad52-L264P . Triple mutant strains' doubling times ranged from 92 to 147 min ( Table S1 ) , indicating that even if barriers to replication might persist in some background , the suppression is rather strong . We conclude from these experiments that there is a feature common to toxic recombination intermediates eliminated by Srs2 in the different recombination and replication mutants as well as in cells exposed to DNA-damaging agents . Therefore , the toxicity of these intermediates disappears in rad52-L264P srs2Δ cells or , alternatively , the intermediates themselves are not formed . The defects associated with srs2Δ in haploid cells can be suppressed by rad52-L264P . However , resistance to γ-ray and UV irradiation is only partially restored in srs2Δ rad52-L264P homozygous diploids ( Figure 2C and D ) . It has been proposed that the higher sensitivity of srs2Δ homozygous diploids compared with haploids is related to the resolution of specific interactions between homologous chromosomes , probably related to HR [33] . Thus , rad52-L264P would not suppress srs2Δ deficiency in the resolution of recombination intermediates involving homologous chromosomes . This would also mean that the important role of Srs2 in DNA damage resistance in haploid cells would not be related to the resolution of recombination intermediates . We propose rather that the defect in Srs2 observed in haploid cells is related to the formation of Rad51 filaments that are toxic because they do not achieve strand invasion ( because they are deficient or because a homologous sequence is not available to perform strand invasion ) and cannot be removed from ssDNA . These filaments can be defined as toxic presynaptic Rad51 filaments . Conversely , the fraction of srs2Δ sensitivity that cannot be suppressed by rad52-L264P in diploid cells may be related to a “postsynaptic” role of Srs2 . To test this hypothesis , we took advantage of two recombination systems allowing the repair of a single DSB created by a galactose-inducible HO endonuclease . Both systems require Srs2 to survive the formation of the DSB . The first one involves SSA between direct repeats located 25 kb apart ( Figure 4A ) and the other one uses ectopic gene conversion to repair the DSB ( Figure 4D ) [28] , [29] . Srs2 is required in SSA to remove Rad51 accumulating on ssDNA generated from DSB processing [41] , while during ectopic gene conversion srs2Δ cells fail to properly resolve recombination intermediates [29] . According to our hypothesis , rad52-L264P should only suppress the poor viability of srs2Δ cells after HO induction in the SSA system . This is exactly what we observed: rad52-L264P suppressed srs2Δ low cell viability strongly in the SSA system ( Figure 4B ) , but only marginally in the gene conversion ectopic system ( Figure 4E ) . Monitoring of DSB repair in both systems by Southern blot analysis showed that rad52-L264P restored a normal level of SSA product formation in srs2Δ cells ( Figure 4A ) . However , the gene conversion products in the ectopic system did not accumulate as much as in WT cells ( Figure 4D ) . Note that the survival rates after DSB induction in the SSA and the ectopic systems were largely unchanged in rad52-L264P cells in comparison to WT cells ( Figure 4B and E ) . Moreover , DSB repair analysis by Southern blotting showed that the kinetics of repair in both systems were unaffected by rad52-L264P ( Figure 4A and D ) . We conclude that the rad52-L264P mutation only bypasses srs2Δ defects associated with presynaptic Rad51 nucleoprotein filament formation . This finding implies that rad52-L264P cannot suppress the high level of CO observed in srs2Δ cells either [29] , [30] . Southern blot quantification ( Figure 4D ) confirmed the increased amount of CO in srs2Δ cells ( 18% ) compared with WT cells ( 5% ) . As expected , the amount of CO in rad52-L264P srs2Δ cells was still very high ( 13% ) . To confirm this result , we used the arg4 ectopic recombination system described by Robert et al . [30] . We found CO in 64% of ARG+ recombinants in rad52-L264P srs2Δ cells in this background , a value comparable to the 60% measured in srs2Δ cells ( Escartin F , De Cian A , Coïc E , Gilquin B , Le Cam E , Veaute X , unpublished data ) . Altogether , these results show that the genetic interactions between srs2Δ and rad52-L264P are related to Rad51 filament formation and not to postsynaptic resolution of intertwined recombination intermediates . Elimination of the checkpoint response in the mec1Δ sml1Δ derivative of the srs2Δ strain suppresses , like rad52-L264P , the poor viability related to the formation of the DSB in the SSA system [28] . As a consequence , we wondered if the Rad52-L264P mutant protein could somehow lower the persistence of the checkpoint related to the maintenance of extensive ssDNA formed around the DSB before repair could take place . The analysis of Rad53 by western blot ( Figure 4C ) shows that it is similarly phosphorylated two hours after HO induction in both rad52-L264P and WT strains . We confirmed that this modification disappears after 12 hours in WT and rad52-L264P cells and persists in srs2Δ cells even 24 hours after DSB formation . It was previously shown that the persistence of the checkpoint activation in srs2Δ cells is dependent on Rad51 [28] because the nucleofilament protects ssDNA from degradation [41] . rad52-L264P suppresses the checkpoint activation in srs2Δ cells as shown by the similarity of the Rad53 phosphorylation kinetics in rad52-L264P srs2Δ compared with WT . This implies that the Rad51 filaments formed by the mutated Rad52 protein allow the checkpoint to be turned off , even in the absence of Srs2 . This result strengthens our conclusion concerning the involvement of rad52-L264P in the suppression of srs2Δ defects in presynaptic Rad51 filament management and suggests that Rad52-L264P-mediated Rad51 filaments are different from Rad52-mediated filaments . During ectopic gene conversion , the srs2Δ mutant defect is associated with a slower disappearance of Rad53 phosphorylation . The retarded bands decrease in intensity after 8 hours in WT cells compared with 12 hours in srs2Δ cells ( Figure 4F ) . This slower recovery is not suppressed by rad52-L264P , confirming the absence of suppression of srs2Δ defects in this system . We carried out chromatin immunoprecipitation ( ChIP ) experiments to study the recruitment of RPA , Rad52 and Rad51 to ssDNA in order to monitor Rad51 filament formation in cells that express Rad52-FLAG or Rad52-L264P-FLAG . We used the SSA assay described above because DSB repair requires the formation of long ssDNA tails [28] , [29] . This allowed us to follow Rad51 filament formation on ssDNA for a long period of time . Quantitative PCR was carried out using primer sets that amplify DNA at 0 . 6 kb and 7 . 6 kb upstream of the DSB site during a time-course experiment to follow DSB induction . We found an increase in the relative enrichment of RPA , Rad52-FLAG and Rad51 at the site of DSB formation compared to the uncut control ARG5 , 6 locus , indicative of the formation of ssDNA and subsequently of Rad51 filaments ( Figure 5 ) . The increase of RPA binding was higher in Rad52-L264P-FLAG than in Rad52-FLAG expressing cells at positions 0 . 6 kb and 7 . 6 kb upstream of the DSB site . The highest RPA increase in Rad52-L264P-FLAG expressing cells ( 2-fold in comparison to what is detected in Rad52-FLAG expressing cells ) was observed four hours after HO induction and was associated with a decrease in Rad52-L264P-FLAG and Rad51 binding at the same time point ( 2- and 3-fold , respectively , in comparison to Rad52-expressing cells , Figure 5 ) . However , Rad51 binding was still 30-fold higher than the enrichment value observed at the ARG5 , 6 locus without DSB . The lower Rad51 recruitment in Rad52-L264P-FLAG expressing cells might be caused by a reduced mediator activity of this Rad52 mutant . Alternatively , it might be the result of a modification of the Rad51 filament properties . For example , their average length could be shorter in Rad52-L264P-FLAG than in Rad52-FLAG expressing cells . We also observed that the recruitment of RPA , Rad52 and Rad51 was lower at position 7 . 6 kb than at position 0 . 6 kb in Rad52-FLAG expressing cells ( Figure 5 ) . This could be linked to a limited amount of proteins to cover long ssDNA tails . However , western blot analysis of the cell extracts used for ChIP showed that the protein level of RPA and Rad52-FLAG remained unchanged after DSB formation ( Figure S3 ) , whereas the pool of Rad51 increased . Consequently , the lower recruitment of Rad51 at 7 . 6 kb could be related to a specific activity of Srs2 at this position . Indeed , we found that Rad51 recruitment is of the same order of magnitude at the 7 . 6 kb and 0 . 6 kb positions in srs2Δ cells . The finding that Rad51 binding at 0 . 6 kb was not significantly affected in srs2Δ cells in comparison to WT cells indicates that Srs2 does not influence significantly Rad51 recruitment at position 0 . 6 kb , while it is very active in eliminating Rad51 filaments at position 7 . 6 kb . This observation suggests that Rad51 filaments forming far away from the DSB site might be responsible for the death of srs2Δ cells upon DSB formation . Finally , reduced Rad51 binding at 7 . 6 kb was observed also in rad52-L264P cells and was confirmed in rad52-L264P srs2Δ cells as well . This lower recruitment ( or the modification of Rad51 filaments properties ) can explain the resistance of such cells to DNA damage . We then asked whether the lower recruitment of Rad52-L264P and Rad51 in rad52-L264P cells observed by ChIP could be related to a modification of the interaction of RPA or Rad51 with Rad52-L264P . Indeed , L264 is located just upstream of the QDDD motif essential for RPA binding [36] ( Figure 1B ) . Therefore , it was possible that the rad52-L264P mutation affects RPA binding [42] , [43] . We evaluated Rad52-RPA interaction by co-immunoprecipitation experiments . Rad52 was immunoprecipitated with a polyclonal antibody and the precipitated fraction was analyzed on western blots with a polyclonal antibody directed against the RPA complex ( Figure 6A ) . We first observed that the rad52-L264P mutation did not substantially affect the binding to RPA , which was expected for a mutant protein still able to manage DNA repair by HR . However , to make sure that there was no difference between the WT and the mutant protein in binding to RPA , we added increasing salt concentrations to the cell extracts to test the robustness of this interaction . As shown in Figure 6A , the amount of RPA co-immunoprecipitated was equivalent in extracts expressing the WT or the mutant protein . In both cases , the interaction was abolished at 500 mM NaCl . In the same experiment , we looked at the interaction between the mutated Rad52-L264P protein and Rad59 [44] and found no major differences between the WT and the mutant protein at 150 mM NaCl ( Figure S2 ) . Since Rad59 interaction is not crucial for Rad52/Rad51-dependent recombination [45] , we did not investigate further the effect of rad52-L264P on this interaction . We also measured the effect of rad52-L264P on Rad51 binding [44] . For that purpose , we used Rad52-FLAG and Rad52-L264P-FLAG tagged proteins and found that the mutation did not affect this interaction ( Figure 6B ) . Similarly to what is observed for RPA , the destabilization of the interaction with Rad51 by increasing NaCl concentrations was comparable when using WT Rad52 or Rad52-L264P . To characterize the biochemical properties of Rad52-L264P we purified recombinant Rad52-L264P and Rad52 from Escherichia coli ( Figure 7A ) . Using electrophoretic mobility shift assays ( Figure 7B ) , we found that Rad52 binding to ssDNA was not significantly affected by the mutation . Next , we investigated whether the L264P mutation affected Rad52 annealing activity by incubating Rad52-L264P and WT Rad52 with complementary ssDNA strands and monitoring the formation of dsDNA . Again , no significant difference was observed ( Figure 7C ) . This was further confirmed in a reaction where ssDNA was coated first with RPA to reflect the in vivo conditions ( Figure 7D ) . We also observed that Rad52-L264P annealing activity was sensitive to RPA , as previously reported for WT Rad52 [7] . Finally , Rad52-L264P and Rad52 annealing activity were similarly affected by Rad51 filaments and free Rad51 proteins [7] ( Figure S4A–C ) . To determine whether the L264P mutation has an impact on Rad52 recombination mediator activity , we used a well-established DNA strand exchange system that involves plasmid DNA substrates ( Figure 7E ) . In this system , RPA that is pre-bound to ssDNA partially inhibits strand exchange by Rad51 . The addition of Rad52 together with Rad51 increases the formation of products through Rad52 interaction with RPA ( reviewed in [4] ) . We found that Rad52-L264P was approximately 8-fold more efficient than WT Rad52 in catalyzing DNA strand exchange . As this finding indicates that the mediator activity of Rad52-L264P is modified , we studied more precisely the effect of the L264P mutation on the formation of Rad51 filaments and their stability by using electrophoretic analysis of glutaraldehyde-fixed Cy5-labeled DNA-protein complexes . First , we optimized RPA , Rad51 and Rad52 stoichiometry at 60 mM NaCl to obtain the best Rad52 antagonism of the inhibitory effect of pre-bound RPA on Rad51 nucleoprotein filament formation ( Figure S5 ) . In these conditions , Rad52-L264P showed a slightly increased mediator activity in comparison to WT Rad52 ( Figure 7F ) . Challenging Rad51 filament formation with increasing salt concentrations showed that the mediator activity of both WT and mutant Rad52 was very sensitive to salt . Indeed , the addition of only 60 mM NaCl to the reaction ( 120 mM in total ) was sufficient to reduce Rad51 filament formation from 30 to 10% of the protein/DNA complexes . Quantification of Rad51 filament formation at increasing NaCl concentrations confirmed that Rad52-L264P was slightly more efficient than WT Rad52 ( the difference was significant at 100 and 120 mM NaCl ) . However , the salt titration midpoint was the same for both WT and mutant protein ( around 90 mM ) , again indicating that the difference in the mediator activity between Rad52-L264P and Rad52 is minimal . Western blot analysis using an anti-Rad51 polyclonal antibody confirmed the slightly higher mediator activity of Rad52-L264P . Electron microscopy ( EM ) analysis of the resulting protein-ssDNA complexes confirmed the sensitivity of the reaction to NaCl concentration and the higher mediator activity of Rad52-L264P ( Figure 7G ) . We also measured the stability of the nucleoprotein filaments formed by Rad52 or Rad52-L264P by adding increasing concentrations of NaCl after Rad51 filament formation ( Figure S5B ) and found no difference . The salt-titration mid-point was around 400 mM . In addition , western blot analysis using an anti-Rad52 polyclonal antibody revealed that Rad52 remained associated with complete Rad51 filaments ( Figures 7F and S5 ) . This association was quite unstable because the addition of 60 mM NaCl after complete formation of filaments resulted in Rad52 dissociation from the complex ( Figure S5C ) . EM analysis of the nucleoprotein complexes formed with Rad52 ( at 60 mM NaCl ) showed that 55% of complete Rad51 filaments remained associated with the mediator protein ( Figures 7F and S5 ) . Rad52 was localized mostly at the end of filaments ( 75% ) , but in many cases multiple Rad52 spots were distributed all along the filament . In absence of RPA pre-bound to ssDNA , Rad52 was rarely associated with Rad51 filaments ( Figure S5A ) . Thus , Rad52 spot formation might be dependent on the previous binding of RPA to ssDNA . Rad52 might remain associated with residual RPA bound to DNA between Rad51 molecules . Alternatively , as only a few complete Rad51 filaments are formed in the absence of RPA , Rad52 association might be restricted to complete Rad51 filaments . Our results also show that Rad52-L264P association with Rad51 filaments was increased ( 74% of complete Rad51 filaments compared with 55% for WT Rad52 , Figure 7G and S5 ) , while the proportion of Rad52-L264P spots at the end of filaments was comparable ( 78% ) . This increased association of Rad52-L264P together with its higher mediator activity might modify qualitatively Rad51 filament properties and cause the suppression of srs2Δ phenotypes in vivo . In parallel to the search for point mutations suppressing the MMS sensitivity of srs2Δ mutants , we also screened a multi-copy genomic library for high-dosage suppressors . After a second round of selection , plasmids bearing a suppressor were sequenced . Strikingly , we isolated a plasmid carrying the SIZ2 gene ( Figure 8A ) . Siz2 is the only SUMO ligase involved in Rad52 sumoylation after the formation of chemically induced DNA damage ( like MMS ) [11] . Consequently , we speculated that the stimulation of Rad52 sumoylation was responsible for the suppression conferred by SIZ2 overexpression since our study of rad52-L264P shows that subtle changes in Rad52 activity can bypass the requirement for Srs2 in haploid cells . Indeed , we found that the suppression of the MMS sensitivity of srs2Δ conferred by SIZ2 overexpression is no longer observed in cells bearing a sumoylation-deficient rad52-3KR allele , where the three SUMO acceptor sites , lysines 10 , 11 and 220 are replaced by arginines [11] ( Figure 8A ) . As a control , we checked that the rad52-3KR allele was not sensitive to MMS ( Figure 9B and D ) , as previously described [11] . Therefore , the suppression of the MMS sensitivity of srs2Δ cells by the overexpression of SIZ2 occurs through the sumoylation of Rad52 . To check that SIZ2 overexpression truly stimulates Rad52 sumoylation , extracts from cells overexpressing SIZ2 and His-tagged SUMO were subjected to Ni-NTA pull-down experiments under denaturing conditions ( Figure 8B ) . The amount of Rad52 sumoylation was compared to that of cells expressing SIZ2 under normal physiological conditions . Because the FLAG tag we used to detect Rad52 contains a poly-histidine chain , the protein was also retained on the Ni-NTA beads , allowing us to quantify the amount of sumoylated protein relative to the total amount of Rad52 . Overexpression of SIZ2 increases the amount of sumoylated Rad52 3-fold compared with its basal expression . This result strengthens our genetic experiments establishing that an increase in the pool of sumoylated Rad52 allows Srs2 activity to be bypassed . Since Rad52 sumoylation leads to the suppression of MMS sensitivity of srs2Δ cells , we wondered whether the suppression by rad52-L264P was dependent on an increase in the pool of sumoylated Rad52 induced by the mutation . As shown in Figure 8B , rad52-L264P does not increase Rad52 sumoylation . We also found that the suppression conferred by rad52-L264P was not dependent on SIZ2 since the deletion of this gene in rad52-L264P srs2Δ cells did not affect MMS resistance ( Figure 9A ) . Additionally , a rad52-L264P allele that cannot be sumoylated ( rad52-3KR-L264P ) still suppresses the MMS sensitivity of srs2Δ mutants ( Figure 9B ) . Altogether , these results clearly show that Siz2-mediated Rad52-L264P sumoylation is not required to bypass Srs2 . We also checked if the Rad52-L264P protein was sumoylated by another SUMO ligase . We found that Rad52-L264P MMS-induced sumoylation was still dependent on Siz2 ( Figure S6 ) . To reconcile our findings that srs2Δ MMS sensitivity can be suppressed on the one hand by Rad52 sumoylation and on the other hand by rad52-L264P , even if sumoylation cannot occur , we propose that Rad52-L264P behaves as if it is constitutively sumoylated , even when it is not physically modified . This would mean that the mutation has the same effects on Rad52 activities , as does sumoylation . Consequently , cells that constitutively express only sumoylated Rad52 should suppress the MMS sensitivity of srs2Δ cells as strongly as rad52-L264P . To test this hypothesis , we fused the SMT3 gene , coding for the SUMO modifier , to the 3′ end of the endogenous RAD52 gene , to generate cells expressing only a Rad52 protein bearing a C-terminal SUMO fusion . The part of SMT3 coding for the last three amino acids and the two glycines required for conjugation [46] was removed to avoid subsequent conjugation of SUMO with other proteins . The resulting strain displays only mild MMS sensitivity ( Figure 9C ) , showing that the addition of SUMO does not substantially affect Rad52 activity in DNA repair . srs2Δ cells bearing the RAD52-SMT3 fusion allele were as MMS-resistant as those carrying rad52-L264P . Therefore , in cells that produce only Rad52-SUMO fusion proteins , Srs2 becomes incidental to MMS resistance . We also monitored the effect of the fusion protein on survival of srs2Δ cells in both SSA and MAT ectopic recombination systems ( Figure 4B and E ) . As for Rad52-L264P , the expression of the Rad52-SUMO protein instead of Rad52 can alleviate the strong lethality associated with srs2Δ in the SSA system but not in the ectopic system . Additionally , ChIP analysis showed that Rad52-SUMO behaves as Rad52-L264P . Using FLAG-tagged proteins , we also observed a 2-fold decrease in Rad52-SUMO recruitment to ssDNA compared with Rad52 ( Figure 5 ) . This is associated with a 2-fold decrease in Rad51 binding , as we observed for Rad52-L264P . These results reinforce the idea that Rad52 sumoylation affects Rad52 activities in the same way as rad52-L264P . Since Rad52 sumoylation can bypass the requirement for Srs2 in the avoidance of toxic Rad51 filaments , it would be expected that the expression of the Rad52-3KR protein , which cannot be sumoylated , would lead to a negative interaction with srs2Δ . This is exactly what we observed ( Figure 9D ) . This negative effect is also observed in a siz2Δ srs2Δ double mutant ( Figure 9E ) . It is interesting to note that the negative effect of rad52-3KR in a srs2Δ background is dominant to the RAD52 allele . Indeed , the introduction of the RAD52 allele in a rad52-3KR srs2Δ strain does not increase MMS resistance ( Figure 9D ) . We suggest that Rad52-3KR mediates toxic Rad51 filaments that can be disassembled only by Srs2 . It has been reported that a plasmid carrying rad52-3KR is able to suppress ( albeit weakly ) rrm3Δ srs2Δ synthetic lethality and sgs1Δ srs2Δ growth defect [11] . However , we found that rad52-L264P suppresses these negative interactions ( Figure 3A ) , while this allele codes for a protein behaving as if it were sumoylated . We confirmed our conclusion by tetrad analysis showing that a single integrated copy of rad52-3KR , or siz2Δ , cannot rescue rrm3Δ srs2Δ and sgs1Δ srs2Δ negative interactions ( Figure 3B ) . Finally , it has been reported that the Rad52-3KR protein is more prone to proteasome degradation than the WT protein , mostly in cells lacking Srs2 or Rrm3 helicases [11] . Since rad52-L264P makes the protein behave as if it were constitutively sumoylated , it was possible that it could suppress the increased instability of Rad52-3KR . Expression shut-off experiments were performed with different mutation of Rad52 in srs2Δ mutants . We found that the Rad52-3KR-L264P protein was not more stable than Rad52-3KR ( Figure S7 ) . It is possible that the decreased stability of Rad52-3KR is related to the substitution of the three targeted lysines and not to the inability of the protein to be sumoylated .
We found a novel allele , rad52-L264P , which suppresses the requirement for Srs2 activity in DNA damage tolerance . A profusion of RAD52 mutants displaying a separation of function has been studied , revealing the multifunctional nature of the protein . Some of them were affected in DSB repair activity but not in spontaneous recombination [34] , [47] . Others were differentially affected in the mediator and the ssDNA annealing activity of Rad52 [48]–[50] . Some of these mutants , which were shown to affect Rad52 mediator activity , display a partial resistance to DNA damage [49] , [50] , but their viability is still reduced by several orders of magnitude . Lastly , some have been shown to alter the choice of the donor template during spontaneous HR [51] . rad52-L264P contrasts with all these mutations because it does not confer any of the rad52Δ null mutation-associated phenotypes ( defect in vegetative growth , increased spontaneous mutagenesis , very high MMS and γ-ray sensitivities , large decrease in DSB repair by SSA and gene conversion ) . In addition , it is not a hypomorphic allele since it is co-dominant when it is combined with the WT allele . Several alleles of RAD52 that partially suppressed the sensitivity of srs2Δ cells to DNA damage have been previously described [20] . They were all highly defective in DNA repair and HR , but not as much as the null allele . The viability of cells carrying these alleles is reduced 10- to 20-fold at low doses of MMS , and HR is reduced 10-fold . They were all dominant-negative and , unlike what we observed for rad52-L264P , they have to be combined with the WT allele or overexpressed to suppress only partially the MMS or UV sensitivities of srs2Δ mutants . In addition , none of these RAD52 mutants suppressed both phenotypes , while this is the case for rad52-L264P . It was proposed that these rad52 alleles are able to rescue srs2Δ mutants partially by reducing HR efficiency . Some of these mutations code for C-terminal truncations unable to bind Rad51 . Therefore , they could bind ssDNA without forming Rad51 filaments . Others were N-terminal truncations , probably impaired in DNA binding , which suggests that their overexpression resulted in depletion of the protomer pool of Rad51 . It has been proposed that both kinds of mutations result in a large decrease in Rad51 filament formation , resulting in a limited suppression of srs2Δ MMS or UV sensitivities . However , rad52-L264P suppresses many defects of srs2Δ haploid cells without displaying any rad52Δ phenotype . Thus , in contrast with previously described alleles , rad52-L264P allows an extensive genetic and molecular investigation of the toxic recombination intermediates formed by Rad52 and eliminated by Srs2 in WT cells . We found that rad52-L264P cannot overcome the function of Srs2 in the resolution of postsynaptic recombination intermediates , but only those associated with the removal of presynaptic Rad51 filaments . It was reported previously that the Srs2 anti-recombination function in removing toxic Rad51 filaments is genetically separable from its role in promoting the resolution of postsynaptic recombination intermediates , which depends exclusively on Srs2 Cdk1-dependent phosphorylation [52] . Our results are in agreement with this finding . rad52-L264P does not suppress the increased sensitivity to UV and γ-ray irradiation of srs2Δ/srs2Δ diploid cells . It is also unable to suppress the strong lethality associated with ectopic HR in srs2Δ cells . Additionally , it cannot suppress the high rate of CO found in these cells . We propose that these defects are related to the resolution of HR intermediates between homologous or ectopic chromosomes that absolutely requires the helicase activity of Srs2 ( Figure 10A ) . Additionally , our results support the recent finding that the post-synaptic role of Srs2 is to dismantle HJs through its helicase activity [32] rather than by displacing Rad51 . However , rad52-L264P perfectly rescues the lethality associated with DSB repair by SSA between distant repeats . In this assay , a Rad51-dependent strand invasion step is not involved in repair [28] . Therefore , Srs2 is required to remove Rad51 filaments forming on ssDNA around the initial DSB to allow proper repair . By extension , our findings suggest that Rad51 filaments are also responsible for srs2Δ defects after γ-ray and UV irradiation and in cells bearing mutations in genes that impair replication or recombination ( Figure 10A ) . In all the situations where the requirement for Srs2 can be bypassed by rad52-L264P , we propose that the formation of extensive ssDNA may potentially lead to the establishment of Rad51 filaments that cannot recombine ( Figure 10A ) . These “unproductive” filaments would be a signal to trigger Srs2 translocase activity in WT cells . In the SSA assay , ChIP experiments showed that Srs2 is more active in removing Rad51 at 7 . 6 kb than at 0 . 6 kb from the DSB site , thus indicating how Srs2 might interfere with such unproductive filaments in WT cells . Indeed , Srs2 job might be to remove extended rather than short Rad51 filaments located near the 3′-end of the ssDNA tail . The length of 5′-ssDNA resection is in part related to strand invasion . Indeed , it is well documented that the formation of a DSB in a unique sequence results in extended resection [53] . Therefore , the formation of extended Rad51 filaments , which is a potential mark of strand invasion failure , would trigger Srs2 activity . The formation of filaments near the ssDNA end might be strictly controlled , while the lower recruitment of Rad52 at more distant sites might affect the quality of the filaments , triggering Srs2 activity . This would allow DSB repair by SSA to occur when HR is inefficient . In srs2Δ cells , the persistence of extended , toxic Rad51 filaments would lead to DSB repair towards a dead-end . Likewise , synthetic lethality of srs2Δ with replication mutants , like mrc1Δ , would be related to unproductive Rad51 filament formation . In mrc1Δ cells , uncoupling between DNA synthesis and DNA unwinding by the MCM helicases [54] leads to the formation of complementary ssDNAs on each newly replicated sister chromatid . In the absence of Srs2 , Rad51 filaments formed on these ssDNAs would not be productive because of the lack of a dsDNA template on each sister chromatid . Such filaments could impede replication fork restart . Considering that Srs2 is more active on Rad51 filaments not associated with a 3′-end , we suggest that Srs2 might specifically recognize and eliminate Rad51 filaments formed on parental ssDNA at the replication fork because they are not associated with such an end . The precise nature of toxic recombination intermediates in rad50Δ , rad54Δ or sgs1Δ has yet to be clarified . Rad50 has an essential role in sister-chromatid recombination ( reviewed in [55] ) and Rad54 is necessary to initiate DNA synthesis after the formation of the D-loop [56] , to extend this structure [57] and to remove the dsDNA template nucleosomes [58] . Sgs1 is probably involved in both replication and recombination mechanisms . Sgs1 could help replication progression and could prevent the formation of ssDNA on which Rad51 filaments could nucleate in srs2Δ cells [40] . Furthermore , several lines of evidence suggest that Sgs1 is also involved in the dissolution of dHJ [59] , [60] and it was proposed that in sgs1Δ cells Srs2 could prevent the formation of these structures [40] . Therefore , in all these backgrounds , toxic recombination intermediates accumulating in the absence of Srs2 could be presynaptic Rad51 filaments or unprocessed postsynaptic recombination intermediates . However , since rad52-L264P can suppress rad50Δ , rad54Δ and sgs1Δ synthetic growth defect or lethality with srs2Δ , it seems more relevant that these negative interactions are the consequence of the accumulation of ineffective Rad51 filaments . Finally , γ-ray and UV irradiation would sensitize srs2Δ cells because of the formation of unproductive Rad51 filaments on extensive ssDNA . These would accumulate because of DNA damage-induced fork stalling or unfulfilled DNA repair . In contrast , rad52-L264P cannot suppress the increased sensitivity to UV and γ-ray irradiation of srs2Δ/srs2Δ diploid cells compared with srs2Δ haploids . Nor can it suppress the inability to repair a DSB by ectopic recombination nor the high level of CO associated with gene conversion of srs2Δ haploid cells . We propose that these defects are related to the resolution of HR intermediates between homologous or ectopic chromosomes which absolutely requires the helicase activity of Srs2 ( Figure 10A ) . In vitro experiments showed that Rad52-L264P mediator and strand exchange activities are stronger than those of WT Rad52 . This intriguing gain of function of Rad52-L264P is not correlated with any modification of its binding to RPA or Rad51 in vivo . Thus , further biochemical analyses of Rad52-L264P are required to provide valuable insights on the mechanism by which Rad52 mediates Rad51 filament formation . In vivo , other Rad51 filament mediators , which might act with Rad52 as a complex , could be affected by the rad52-L264P mutation . This would explain the reduction in Rad51 binding associated with increased RPA recruitment , measured by ChIP , in rad52-L264P cells , while , in vitro , Rad52-L264P is a better mediator than the WT protein . The evidence that Rad52-L264P restricts the formation of ( or shortens ) Rad51 filaments without affecting HR efficiency could explain how this mutation can bypass Srs2 , not only in SSA , but also after irradiation or in mutants that are synthetically lethal with srs2Δ . Alternatively , Rad52-L264P could change substantially the properties of Rad51 filaments . Western blot analysis and EM images of Rad51 filaments nucleated in vitro show that Rad52 remains associated with complete Rad51 filaments . This association is weak because Rad52 is released by the addition of only 60 mM NaCl , but it could be stabilized in vivo by chaperone proteins . The presence of Rad52 within the Rad51 filament might have consequences on homology search and strand invasion . Indeed , Rad52-L264P association with Rad51 filaments is increased in comparison to WT Rad52 , a result that can be correlated with the more efficient strand exchange activity observed with the mutant protein . Rad52-L264P might also affect Rad51 filament properties in a way that would suppress their potential toxicity in srs2Δ cells . For example , such filaments would not prevent the restart of stalled replication forks , thereby bypassing the need for Srs2 . We also found that Rad52-SUMO fusion protein and Rad52-L264P show a similar ability to avoid or restrict the formation of toxic Rad51 filaments . It was previously reported that Rad52 is sumoylated simultaneously with RPA and Rad59 following treatment with a high dose of MMS [15] . This sumoylation wave might stabilize complexes engaged on their substrates rather than to promote specificity . However , recent work showed that Cdc48 interacts with sumoylated Rad52 and consequently dissociates it from DNA [16] . This interaction might be part of another Rad51 filament formation regulation process , acting in parallel with Srs2 activity . Here , we demonstrate that increasing the pool of sumoylated Rad52 suppresses srs2Δ deficiencies in haploid cells . We thus propose that Rad52 sumoylation might modulate its mediator activity and/or change the properties of the formed Rad51 filaments , maybe through Cdc48 activity , in the same way as Rad52-L264P ( Figure 10B ) . It seems likely that sumoylation of Rad52 is a conserved process because mono- and disumoylation of human Rad52 were also observed in HEK293T cells [11] . It would be interesting to know if this modification induces changes in Rad51 filament properties in mammals as it does in yeast . In this case , it might be important to explore the genetic interactions between Rad52 sumoylation and the newly characterized anti-recombinase PARI [61] , which could be the mammalian Srs2 ortholog . Lastly , it is tempting to compare the potential differences between Rad52 un-sumoylated and sumoylated proteins in yeast and those of Rad52 and BRCA2 in metazoans . Recently , it was shown in human cells that these two proteins co-exist as mediators of the Rad51 filaments [62] . Even if the roles of Rad52 in recombination in yeast and metazoans are certainly different , it is possible that in both cases different Rad51 filaments are nucleated .
For integration into the yeast genome , the rad52-L264P allele was cloned into the Yiplac211 integrative plasmid . rad52-L264P was PCR amplified from genomic DNA with primers containing restriction sites suitable for cloning . The digested PCR product was ligated into the EcoRI and HindIII sites of Yiplac211 to give YipLac211-rad52-L264P . YipLac211-rad52L264A was made by directed mutagenesis from YipLac211-rad52-L264P ( Phusion Site-Directed Mutagenesis kit , Finnzymes ) with primers changing the CCC codon coding for P264 to a GCC codon coding for A264 . The rad52-L264P mutation was introduced in the same way into pYI211::Kan-rad52-K10 , 11 , 220R ( D2535 , provided by S . Jentsch ) with primers changing the CTC codon coding for L264 to a CCC codon coding for P264 . Yep181-CUP-His7-Smt3 [63] was used to overexpress His7-Smt3 in cells in order to immunoprecipitate SUMO-conjugated proteins . To create a fusion between RAD52 and SMT3 , the SMT3 ORF was fused to the 3′ intergenic sequences of RAD52 in a vector bearing the NATMX cassette coding for the resistance to clo-NAT ( pAG25 , [64] ) . First , the SMT3 ORF was PCR amplified from FF18733 genomic DNA with primers suitable for cloning . The last three amino acids and the two glycines required for conjugation [46] were removed to avoid subsequent interaction of SUMO with other proteins . The natural SMT3 stop codon was conserved . The PCR product was ligated into the HindIII and BamHI sites of pAG25 resulting in pAG25-SMT3 . This construct created a NheI restriction site just 5 bp before the BamHI sites . The RAD52 3′ intergenic sequence was then ligated into the NheI and BamHI sites of this plasmid , generating pEC54 . pSIZ2 is a subclone of a plasmid suppressing the MMS sensitivity of srs2Δ cells isolated from an overexpression library built for this study . Strains used in this study are listed in Table S2 . Experiments were mostly conducted in the FF18733 background . Diploid cells used in survival and recombination assays were the result of crosses between two different cell backgrounds: FF18733 and FF18985 in order to monitor HR between the his7-1 and his7-2 alleles . All the deletion mutants were constructed by the one-step gene disruption method [65] . Multiple mutant strains were derived from meiotic segregants from FF18733 or FF18985 isogenic diploids . Mutations rad52-L264P and rad52-L264A were introduced into yeast cells with the pop-in pop-out technique using the integrative plasmids Yiplac211-rad52-L264P and Yiplac211-rad52-L264A . The non-sumoylable rad52-3KR and rad52-3KR-L264P alleles were targeted at the URA3 locus by transformation of the integrative plasmids pYI211::Kan-rad52-K10 , 11 , 220R ( D2535 [11] ) and pYI211::Kan-rad52-K10 , 11 , 220R-L264P . The SMT3 gene , coding for the SUMO radical , was fused in vivo to the 3′ end of the RAD52 gene . The insert of pEC54 , containing the SMT3 ORF and a NATMX resistance cassette , was PCR amplified with primers designed to introduce SMT3 in phase with RAD52 , without affecting the intergenic sequences surrounding RAD52 . Transformants were selected on clo-NAT containing medium and checked by colony PCR . The production of the fusion protein was checked on western blot ( Figure S8 ) . The strain bearing the Rad59-9xMYC fusion protein was made using the method described in [66] in the FF18733 background . The Rad52-FLAG strains were constructed as previously described in the same background [67] . Homologous sequences of S . cerevisiae Rad52 were retrieved using PSI-Blast searches against the nr database [68] , [69] . A multiple sequence alignment of the full-length sequences of these homologs was obtained using Muscle software [70] . However , within the C-terminal disordered tail the algorithm did not satisfactorily align the small linear motifs surrounding L264 and the alignment had to be manually refined . Propensities to adopt secondary structures were estimated using PsiPred restricting the alignments to subsets of species such as those from the Hemiascomycetes group [71] . The final alignment was represented using Jalview [72] . UV irradiation was performed using a 264 nm source delivering 1 J/m2/s . γ-ray irradiation was performed using a 137Cs source at a dose of 50 Gy/min . Cells growing exponentially were plated at appropriate dilutions on rich medium ( YPD ) and synthetic plates . Survival was determined as the number of cell-forming colonies on YPD at a given dose divided by the number of non-irradiated colonies . We determined HR frequencies by dividing the number of recombinant colonies growing on selective medium by the number of unselected colonies subjected to the same dose of irradiation . The values obtained after irradiation were corrected by subtracting the number of spontaneous recombinants present on the non-irradiated plates . Spontaneous formation of canavanine-resistant colonies was quantified by a fluctuation test based on a minimum of 27 independent cultures of each strain , initiated from approximately 200 cells and grown to saturation [73] . Cells were grown overnight in liquid culture containing lactate before plating . Survival following HO-induced DSB was measured as the number of cells growing on galactose-containing medium divided by the number of colonies growing on YPD . The results shown are the average of at least 3 independent experiments . Cells were grown in YPD until late exponential phase . Cells were then used to inoculate 400 ml of YPLactate . Cultures were grown to a concentration of 5 to 10×106 cells/ml . A 50 ml sample was removed for the 0 hour time-point and then galactose was added to a final concentration of 2% . Incubation was continued and 50 ml samples were removed at given times . Cells were harvested by centrifugation and washed with water . Cell pellets were then frozen at −20°C . DNA was extracted from the thawed cell pellets and digested accordingly . DNA fragments were separated by electrophoresis on 0 . 8% agarose gels , transferred to nylon membranes and hybridized with a suitable radioactive probe ( Ready-prime II , GE Health Care ) . Blots were analyzed by using a Typhoon 9600 phosphorimager ( GE Health Care ) and quantified with ImageQuant Software . The amount of product in the ectopic gene conversion system was measured at 10 hours to avoid the over-estimation of srs2Δ cells that had completed repair . The checkpoint is turned off after the completion of repair in this system and cells resume growth . This repeatedly distorts the quantification at 24 hours . However , in the SSA system , we quantified the amount of repair at 24 hours because the reaction was far from complete at 10 hours ( the first products appear at 6 hours ) and the persistence of the checkpoint impedes srs2Δ cell growth . We used the system described in [30] . However , we used a colony-PCR assay to detect the CO among ARG4 recombinant colonies ( Escartin F , De Cian A , Coïc E , Gilquin B , Le Cam E , Veaute X , unpublished data ) . Briefly , we used primers allowing the discrimination of the parental configurations of the arg4 locus on chromosome VIII from the reciprocal translocation produced by the CO associated with gene conversion . Cells were harvested during the time-course experiments previously described for the physical analysis of HO-induced SSA and ectopic gene conversion . Protein extracts were prepared by trichloroacetic acid precipitation . Proteins were separated on 10% SDS-PAGE with an acrylamide/bisacrylamide ratio of 30∶0 . 4 , for 2 hours at 150 V and transferred to nitrocellulose membrane . Membrane was incubated overnight with a goat polyclonal antibody raised against the C terminus sequence of Rad53 ( Santa Cruz Biotechnology , yC-19 ) at a 1/1300 dilution in PBS , 0 . 1% Tween , 5% milk ( w/v ) ; then incubated for 1 hour with secondary horseradish peroxidase-conjugated anti-goat antibody ( Santa Cruz Biotechnology , Sc-2020 ) at a 1/5000 dilution in the same buffer . The blot was revealed by chemiluminescence ( ECL Plus , GE Healthcare ) . Samples were collected during the same time-course experiment performed to monitor the physical analysis of HO-induced SSA ( see below ) . ChIP was carried out as previously described with minor modifications [74] . Samples were incubated with 2 µg of rabbit anti-RPA polyclonal antibody ( a gift from V . Géli ) , of mouse anti-FLAG monoclonal antibody ( Sigma ) or of rabbit anti-Rad51 polyclonal antibody ( Santa Cruz Biotechnology ) . 50 µl of Magnetic Dynabeads Protein A ( Invitrogen ) was added to each sample when treated with rabbit antibodies and 50 µl of Magnetic Dynabeads Pan mouse otherwise . After washes , elution of the proteins and reversal of crosslink , samples were treated with proteinase K followed by purification of the DNA with QIAquick PCR purification kit ( Qiagen ) . Quantitative PCR reactions of 180 bp fragments at 0 . 6 kb or 7 . 6 kb proximal to the DSB site and at the ARG5 , 6 locus were performed using Platinum SYBR Green qPCR SuperMix-UDG ( Invitrogen ) on an Eppendorf Realplex system . Yeast cells were grown in YPD medium to a concentration of 2 . 5×106 cells/ml . Cells were harvested and washed twice with PBS . Extracts were prepared as previously described [75] without DNAse treatment . The whole cell extract ( 1 mg ) was incubated for 1 hour at 4°C , either with a rabbit anti-Rad52 polyclonal antibody ( a gift from S . Jentsch's lab ) , or with 1 µg of a rabbit anti-Rad51 polyclonal antibody ( Santa Cruz Biotechnology ) . Then , 50 µl of Dynabeads coupled to Protein A ( Invitrogen ) was added , and the incubation was continued for another hour . The immunoprecipitates were washed twice with 1 ml of lysis buffer and resuspended in 30 µl of Laemmli buffer . The eluted proteins were analyzed by western blot . Proteins were separated on 10% SDS-PAGE and transferred to Hybond-C super membrane ( Amersham Biosciences ) . Proteins were detected with rabbit anti-Rad52 polyclonal antibody ( 1/2000 ) , rabbit anti-RPA polyclonal antibody ( a gift from V . Géli , 1/2500 ) , mouse anti-MYC monoclonal antibody ( Sigma , 1/1000 ) , mouse anti-FLAG monoclonal antibody ( Sigma , 1/10000 ) and rabbit anti-Rad51 polyclonal antibody ( 1/2000 ) . Blots were then incubated with a secondary antibody: horseradish peroxidase-conjugated goat anti-mouse antibody or horseradish-peroxidase-conjugated goat anti-rabbit antibody ( GE Healthcare 1/10000 ) . Protein-antibody complexes were visualized by enhanced chemiluminescence using the GE Healthcare ECL Plus system . RPA was purified from the protease-deficient yeast stain BJ5496 ( ura3-52 , trp1 , leu2Δ1 , his3Δ200 , pep4::HIS3 , prbΔ1 . 6R , can1 ) . Cells were transformed with three plasmids containing the RFA1 , RFA2 , or RFA3 ORF under the control of a GAL promoter ( a gift from R . Kolodner ) . The RPA heterotrimer was purified as described [76] . Rad51 was overexpressed in E . coli BL21 ( DE3 ) pLysS cells transformed with the pEZ5139 plasmid ( provided by S . Kowalczykowski ) and then purified as described previously [77] . Rad52 and Rad52-L264P were purified from BRL ( DE3 ) pLysS cells transformed with the pET15b-Rad52 or pET15b-Rad52-L264P plasmid . Cells were grown in 8-liters of LB broth containing 100 µg/ml ampicillin at 37°C until A600 = 0 . 8 . Protein expression was induced by addition of 0 . 1 mM IPTG followed by incubation at 30°C for 3 h . Cell lysis was carried out in 50 mM MES ( pH 6 . 5 ) , 450 mM NaCl , 1 mM DTT , 1 mM EDTA , 10% glycerol , 1 mM AEBSF , 10 mM Benzamidine and 2 mM Pepstatin by sonication . Proteins were purified as described previously [78] until the hydroxyapatite column step . Fractions containing Rad52 or Rad52-L264P were pooled and precipitated with 0 . 45 g/ml ammonium sulfate . Pellets were suspended in 20 mM Tris HCl pH 7 . 5 , 1 M NaCl , 1 mM DTT , 1 mM EDTA and 10% glycerol and then loaded onto Superdex 200 columns ( 24 ml ) . Peak fractions were diluted 20 times to obtain a final concentration of 50 mM NaCl and then loaded onto Resource S columns ( 1 ml ) . Fractions containing purified Rad52 or Rad52-L264P were pooled , diluted to a final concentration of 200 mM NaCl and finally concentrated using Amicon Ultra 3000 ultrafiltration devices ( Millipore ) . Rad52 and Rad52-L264P concentrations were determined using an extinction coefficient of 2 . 43×104 at 280 nm . Increasing amounts of Rad52 or Rad52-L264P ( Figure 7B ) were incubated with 0 . 27 µM 5′ end-Cy5-labeled XV2 oligonucleotide ( 5′-TGG GTG AAC CTG CAG GTG GGC AAA GAT GTC CTA GCA ATG TAA TCG TCA AGC TTT ATG CCG TT-3′ ) in buffer E ( 10 mM Tris-HCl pH 8 , 5 mM MgCl2 , 100 mM NaCl ) at 37°C for 10 min . Complexes were separated on 8% native polyacrylamide gels . Primary DNA annealing reactions ( Figure 7C ) were carried out using the same Cy5-labeled XV2 oligonucleotide as for the electrophoretic mobility shift assay and a reverse-complement oligonucleotide ( XV98 ) . Each primer ( 340 nM nucleotides ) was resuspended in buffer E and then mixed ( time 0 ) . The annealing reaction was started by adding different concentrations of Rad52 or Rad52-L264P ( final volume: 50 µl ) and incubation at 25°C . An aliquot of 9 µl was collected every two minutes , transferred into 6 µl of stop buffer ( 20 µM unlabeled XV2 , 0 . 5% SDS , 0 . 5 mg/ml Proteinase K ) and incubated at 25°C for another 5 min . The effect of RPA and Rad51 on the reaction ( Figure 7D and S4 ) was investigated at 30°C with primers 25 and 26 and the buffers previously described in [7] . RPA or Rad51 ( or only storage buffer for control reactions ) were incubated with the individual primers for 5 min before mixing the nucleoprotein complexes to start the reaction . All samples were separated on 8% native TBE polyacrylamide gels . Fluorescent signals were revealed with a Typhoon 9400 scanner and quantified with ImageQuant ( Molecular Dynamics ) . 33 µM ( nucleotides ) viral ( + ) strand of φX174 DNA were coated first with 1 . 1 µM RPA by incubation in SEB buffer ( 42 mM MOPS pH 7 . 2 , 3 mM Mg acetate , 1 mM DTT , 20 mM NaCl , 25 µg/ml BSA and 2 . 5 mM ATP ) in a final volume of 12 . 5 µl at 37°C for 5 min . Rad51 filament formation was initiated by adding 5 . 5 µM Rad51 and different amounts of Rad52 or Rad52-L264P ( Figure 7D ) , or storage buffers as controls . Reactions were incubated at 37°C for 15 min . The addition of 33 µM ( nucleotides ) of PstI-linearized φX174 dsDNA and 4 mM spermidine initiated the strand exchange reaction . After incubation at 37°C for 90 min , samples were deproteinized by addition of 2 µl of 10 mg/ml Proteinase K , 5% SDS solution at 37°C for 10 min and analyzed by electrophoresis ( 0 . 8% agarose gels in 1× TAE buffer ) . Gels were stained with ethidium bromide and protein bands quantified with ImageQuant ( Molecular Dynamics ) . A 5′-biotinylated 400 bp dsDNA fragment was prepared by PCR using the pBR322 plasmid as template . PCR products were loaded onto HiTrap Streptavidin HP columns ( GE Healthcare ) . The non-biotinylated Cy5-labeled strand was purified by elution with 60 mM NaOH . Salt titration of protein-DNA complex formation was performed by first incubating 82 . 5 nM RPA ( 1/30 nt ) with 2 . 5 µM ssDNA ( 5′ end-Cy5-labeled 400 nt-long fragment ) in SEB buffer in a final volume reaction of 10 µl at 37°C for 5 min . Increasing concentrations of NaCl were added ( Figure 7E ) , followed by addition of 0 . 83 µM Rad51 ( 1/3 nt ) and 90 . 5 nM Rad52 or Rad52-L264P ( 1/27 nt ) . After 15 min incubation at 37°C , reactions were stopped with 0 . 25% glutaraldehyde . 4 µl of 40% sucrose was added to facilitate loading on agarose gel . Nucleoprotein electrophoresis was carried out using 0 . 5% agarose gels in 1X TAE buffer for 1 . 5 hours at 150 mA . Fluorescent signals were revealed with a Typhoon 9400 scanners and quantified with ImageQuant ( Molecular Dynamics ) . Western blot analysis was performed after washing the gels twice with transfer buffer ( 25 mM Tris-HCl , 0 . 2 M glycine , 0 . 015% SDS ) for 20 min . Proteins were then transferred to PVDF membranes with a semi-dried blotter ( Biorad ) at 0 . 8 mA/cm2 for 1 . 25 hours . Membranes were saturated with PBS , 0 . 1% Tween , 5% milk for 1 hour . Hybridizations with anti-Rad51 , anti-Rad52 or anti-RPA antibodies were performed as described for co-immunoprecipitation experiments . Salt titrations of the protein-DNA complex stability were performed as above , but by first incubating proteins and ssDNA in the presence of 60 mM NaCl at 37°C for 15 min to allow the formation of protein-DNA complexes . Additional NaCl was then added to the indicated final concentrations ( Figure S5 ) and the protein-DNA complexes were incubated at 37°C for another 30 min . Reactions were fixed with 0 . 25% glutaraldehyde . Nucleoprotein gel electrophoresis and western blot analysis were performed as before . For transmission electron microscopy studies , a fraction of the complex formation reactions was handled as previously described [31] . Positive staining images were taken in order to monitor filament dynamics and formation , whereas negative staining images were taken to obtain structural information on the position of Rad52 ( along or at the end of the filaments ) . Rad52 sumoylation was induced in exponential phase ( 5×106 cells/ml ) by the addition of 0 . 3% of MMS for 3 hours at 30°C . Rad52 proteins were detected by western blotting as described [11] using a rabbit anti-Rad52 polyclonal antibody at 1/2000 dilution ( from S . Jentsch's lab ) . Rad52-FLAG cells over-expressing His7-SMT3 were collected in exponential phase ( 2 . 5×106 cells/ml ) . Lysates and Ni-NTA pull-Down of sumoylated proteins were carried out according to [63] . Rad52-FLAG was detected with a mouse anti-FLAG monoclonal antibody at 1/10000 dilution ( Sigma ) on western blots . Strains were grown to 2 . 5×106 cells/ml . Expression was shut-off by addition of cycloheximide to a final concentration of 50 µg/ml . For each time-point , 2 ml samples of yeast cells were harvested and protein extracts were prepared [79] . Rad52-FLAG was detected on western blots with a mouse anti-FLAG monoclonal antibody at 1/10000 dilution ( Sigma ) . | Homologous recombination ( HR ) is essential for double-strand break repair and participates in post-replication restart of stalled and collapsed replication forks . However , HR can lead to genome rearrangements and has to be strictly controlled . The budding yeast Srs2 is involved in the prevention of high crossing-over frequencies and in the inhibition of HR at replication forks . Nevertheless , important phenotypes of srs2Δ mutants , like sensitivity to DNA damage and synthetic lethality with replication and recombination mutants , can only be attributed to another role of Srs2: the elimination of lethal intermediates formed by recombination proteins . The nature of these intermediates remains to be defined . In a screen designed to uncover mutations able to suppress srs2Δ phenotypes , we isolated a novel allele of Rad52 ( rad52-L264P ) , the gene that codes for the major Rad51 nucleoprotein filament mediator . Interestingly , we observed that rad52-L264P bypasses the requirement for Srs2 without affecting DNA repair by HR . We also found that Rad52-L264P specifically prevents the formation of unproductive Rad51 filaments before strand invasion , allowing us to define Srs2 substrates . Further analysis showed that Rad52-L264P mimics the properties of the Rad52-SUMO conjugate , revealing that Rad52 assembles Rad51 filaments differently according to its sumoylation status . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Rad52 Sumoylation Prevents the Toxicity of Unproductive Rad51 Filaments Independently of the Anti-Recombinase Srs2 |
Trachoma is a major cause of blindness in Southern Sudan . Its distribution has only been partially established and many communities in need of intervention have therefore not been identified or targeted . The present study aimed to develop a tool to improve targeting of survey and control activities . A national trachoma risk map was developed using Bayesian geostatistics models , incorporating trachoma prevalence data from 112 geo-referenced communities surveyed between 2001 and 2009 . Logistic regression models were developed using active trachoma ( trachomatous inflammation follicular and/or trachomatous inflammation intense ) in 6345 children aged 1–9 years as the outcome , and incorporating fixed effects for age , long-term average rainfall ( interpolated from weather station data ) and land cover ( i . e . vegetation type , derived from satellite remote sensing ) , as well as geostatistical random effects describing spatial clustering of trachoma . The model predicted the west of the country to be at no or low trachoma risk . Trachoma clusters in the central , northern and eastern areas had a radius of 8 km after accounting for the fixed effects . In Southern Sudan , large-scale spatial variation in the risk of active trachoma infection is associated with aridity . Spatial prediction has identified likely high-risk areas to be prioritized for more data collection , potentially to be followed by intervention .
Trachoma , caused by the bacterium Chlamydia trachomatis , is the most common infectious cause of blindness and the leading cause of preventable blindness worldwide [1] , [2] . The disease is easily transmitted through transfer of ocular secretions infected with C . trachomatis to the eyes of an uninfected individual by flies , hands , towels or sharing of other personal items . Repeated infection with C . trachomatis leads to scarring of the conjunctiva and eventually entropion , causing the lashes of the inwardly-turned eyelid to abrade the corneal surface , a condition referred to as trichiasis [3] , [4] . Unless eyelid deformation is managed surgically , trichiasis causes irreversible scarring of the cornea leading to corneal opacity and , eventually , blindness . Trachomatous trichiasis ( TT ) in children is an indication of high-intensity transmission . Like all other neglected tropical diseases ( NTDs ) trachoma is associated with poverty [5] , [6] , as well as poor hygiene [7] , [8] . Prevention is partly based on improving personal hygiene by promoting facial cleanliness and providing clean water for face washing , and promoting the safe disposal of human faeces , thereby reducing fly abundance [9] . Facial cleanliness and Environmental improvement form two of the four components of the World Health Organization ( WHO ) recommended “SAFE” strategy for trachoma control , which also includes Surgical correction of trichiasis and mass drug administration ( MDA ) of Antibiotics in endemic communities [10] . Studies have shown trachoma risk to be associated with attributes of the physical and social environment [8] . Risk factors include environmental aridity , nomadic pastoral livelihoods ( i . e . predominantly livestock-rearing ) , increasing distance from water sources and household crowding [7] , [11]–[13] . Given that environmental factors are important drivers of trachoma risk , it is plausible to predict the spatial distribution of trachoma using statistical associations between disease prevalence and environmental variables . Linkage of trachoma survey data to environmental variables can be performed in a geographical information system ( GIS ) . Statistical models can then be used to estimate the relationship between trachoma risk and environmental variables , and to predict trachoma risk in non-sampled locations based on their environmental attributes . Schemann and colleagues used such ( non-spatial ) multivariate logistic regression model with trachoma data from Mali , finding that prevalence of active trachoma was negatively correlated with rainfall , in turn resulting in a north-south gradient of trachoma risk [14] . A major recent advance in risk mapping has been the development of model-based geostatistics , providing a statistically robust platform for prediction of disease risk based simultaneously on environmental covariates and functions of spatial autocorrelation [15] . The model outputs are distributions , rather than point estimates , which fully represent prediction uncertainties and enable flexible statistical inference , such as determining the probability that risk in a location is above a specific threshold [16] . Risk maps derived from model-based geostatistical predictions have been used to increase the efficiency of some NTD control programmes , such as for schistosomiasis and soil-transmitted helminths in sub-Saharan Africa , by allowing targeting of resources to areas where they were likely to have the greatest impact [16]–[23] . However , to date these epidemiological advances have not been applied to the management of trachoma control programmes . Cataract and trachoma are the two most important causes of blindness in Southern Sudan [24] . Recent surveys have found both extremely high prevalence of active trachoma ( trachomatous inflammation-follicular ( TF ) and/or trachomatous inflammation-intense ( TI ) ) and evidence of TT in children in some of the areas surveyed [25] , [26] . These findings indicate that trachoma constitutes a major problem to public health in Southern Sudan [27] . However , not all of Southern Sudan is equally at risk , as indicated by recent surveys that identified areas where trachoma is not endemic [28] . Generating a better understanding of the geographical distribution of trachoma is therefore important so that the limited available resources can be better targeted . To provide the National Trachoma Control Programme with a tool to prioritise areas for SAFE intervention we develop a model that takes account of spatial correlation in the data , aiming to identify important environmental predictors of trachoma risk in Southern Sudan and to use these to develop a trachoma risk map .
The risk mapping analysis received ethical approval from the Directorate of Research , Planning and Health System Development , Ministry of Health , Government of Southern Sudan ( MoH-GoSS ) . The study consisted entirely of secondary analysis of data from population-based prevalence surveys ( PBPS ) for which separate ethical approval had been obtained from the same institutional review board . Field survey data were obtained from PBPS conducted and previously reported by The Carter Center in Unity [26] , [29] , Jonglei [25] , Eastern Equatoria , Central Equatoria and Upper Nile States [30] , and by Malaria Consortium and the MoH-GoSS , in Western Equatoria State [28] . All PBPS used a two-stage cluster design with randomised selection of communities and individuals within communities . Details on the survey design and ethical approval are provided elsewhere [25] , [28] , [31] . Diagnosis of trachoma was based on physical examination of the conjunctiva of the survey participants by trained personnel and the stage of trachoma was graded using the simplified WHO scheme [32] . In the current study , only data on active trachoma from children aged 1–9 years were included because trachoma in this age group most likely reflected local transmission . Presence of trachomatous inflammation ( either TF or TI ) of the conjunctivae of one or both eyes was considered a positive diagnosis of active trachoma . The age and sex of the participants were recorded during each of the surveys . The field survey locations were geo-referenced using a global positioning system , or by matching community names with those in an existing geo-referenced community database compiled by the Southern Sudan Guinea Worm Eradication Program . The final dataset contained data collected between 2001 and 2009 from 6345 children aged 1–9 years in 112 communities that we were able to geo-locate . The dataset included 3181 boys and 3164 girls . The trachoma field survey data were plotted in the GIS software ArcView ( Version 9 . 2 , ESRI , Redlands , California , USA ) ( Figure 1 ) . Digital information on environmental variables was obtained from different sources . Elevation above mean sea level and interpolated long-term average monthly minimum and maximum land surface temperature and rainfall were obtained from the WorldClim project ( www . worldclim . org ) . Minimum , maximum and mean normalised difference vegetation index ( NDVI ) and land surface temperature ( LST ) for 1982–1998 were obtained from the National Oceanographic and Atmospheric Administration's ( NOAA ) Advanced Very High Radiometer ( AVHRR ) . Classified land cover variables were obtained from the International Geosphere-Biosphere Programme ( IGBP ) ( http://www . igbp . net , derived from AVHRR data ) , grouped into wooded savannah , savannah , cropland/shrubland/grassland and forest/wetland , and from the United States Geological Survey global land cover database ( http://edc2 . usgs . gov/glcc/glcc . php ) . The location of perennial inland water bodies was provided by the Food and Agriculture Organization of the United Nations and used to calculate the distance of survey locations from permanent water sources . These variables were linked in ArcView to the trachoma field data according to location . Co-linearity in the continuous environmental variables was assessed using Pearson's correlation coefficients and for all pairs of variables with correlation >0 . 7 , the variable with the highest p-value in bivariate logistic regression models ( with trachoma prevalence as the outcome ) was excluded . Variance inflation factors ( VIF ) were also examined and variables with a VIF >10 were removed . Environmental variables were selected using backwards stepwise logistic regression in Stata ( Version 10 , Statacorp , College Station , Texas , USA ) using an exit criterion of Wald's p>0 . 1 and an entry criterion of Wald's p≤0 . 05 . Selected environmental variables included long-term average annual rainfall ( continuous in mm ) and IGBP land cover ( categorical ) . Age ( in years ) and sex of survey participants were retained in the models as individual-level covariates . Logistic regression models were developed in the freely available Bayesian statistical software WinBUGS version 1 . 4 ( Medical Research Council Biostatistics Unit , Cambridge , UK/Imperial College London , London , UK ) . These models had the disease status ( positive or negative ) for active trachoma ( TF and/or TI ) in each child aged 1–9 years as the Bernoulli-distributed outcome ( where positive = 1 and negative = 0 ) . Two models were developed with the following parameters: model 1 had fixed effects for age , sex , long-term average annual rainfall and land cover and model 2 , constructed using the principle of model-based geostatistics[15] , had fixed effects for age , sex , long-term average annual rainfall and land cover plus geostatistical location-level random effects with a correlation structure defined by an isotropic exponentially decaying autocorrelation function . In this model , the environmental fixed effects are useful for explaining large-scale spatial variation ( i . e . trend ) ; and for spatial prediction , which is based both on the environmental attributes of the prediction locations and observed prevalence at nearby survey locations ( captured by the geostatistical random effect ) . The individual fixed effects are useful for adjusting the model estimates for any age or sex differences between the survey locations . These models were constructed separately to determine whether the inclusion of the geostatistical component improved the predictive ability of the model . All model parameters were given non-informative prior distributions . Model selection was based on the deviance information criterion ( DIC , a Bayesian analogue of Akiake's information criterion , for which a lower value of the DIC indicates a more favourable compromise between model fit and parsimony ) . Spatial prediction based on model 2 was done in WinBUGS by combining kriging of the random effects ( i . e . estimating their values at non-sampled locations using this geostatistical smoothing method [33] ) with application of the coefficients of the community-level environmental covariates to the values of these covariates at all non-sampled locations . Predictions were thus based on the environmental covariates and the geostatistical random effects . Spatial predictions were validated by randomly partitioning the survey locations into four approximately equal-sized subsets of survey locations . The model was built using three subsets and was used to predict prevalence of active trachoma for individuals at the locations of the fourth subset . This procedure was repeated four times , each time predicting prevalence of trachoma at the locations of a different subset . Thus , predicted prevalence values were obtained for all 112 locations . Discriminatory performance was assessed at the individual level and at the location level . For the former , the individual's predicted risk of trachoma was compared to their observed trachoma status . For the latter , predicted prevalence was compared to observed prevalence dichotomised using the following thresholds: >0% , 10% , 40% and 70% . For each comparison , sensitivity of the predicted value was plotted against one minus the specificity ( the receiver operating characteristic; ROC ) and the area under the ROC was calculated . This was calculated separately for each subset , and for the pooled values from all four subsets . Values of area under the ROC ≥0 . 9 indicate excellent model discrimination , ≥0 . 7–0 . 9 indicate moderate model discrimination and <0 . 7 indicate poor model discrimination . Mean prediction error and mean absolute prediction error were also calculated to determine model calibration . Isotropic semivariograms ( i . e . semivariograms that did not vary by direction ) were developed using the geoR library of the R statistical software package ( Version 2 . 9 . 0 , The R Foundation for Statistical Computing ) to test spatial autocorrelation in the raw prevalence data and in the Pearson's residuals of models 1 and 2 .
Prevalence of active trachoma in children aged 1–9 years was 48 . 2% , but this varied markedly between states of Southern Sudan , ranging from 2 . 2% to 77 . 6% ( Table 1 ) . No statistically significant difference was found in active trachoma prevalence between boys ( 47 . 3% ) and girls ( 49 . 1% ) , but there was a significant negative correlation between active trachoma prevalence and age ( Table 2 ) . In the study communities , the average long-term average rainfall was 979 mm ( range , 509–1470 mm ) . In both models there was a significant negative correlation between rainfall and the prevalence of active trachoma ( e . g . , model 2: OR 0 . 21 , 95% CI 0 . 08–0 . 46 , indicative of a 79% decrease in prevalence for a 100 mm increase in rainfall ) . Land cover was a significant explanatory variable in model 1 , but not model 2 . The unbounded semivariogram for the raw trachoma prevalence ( Figure 2A ) suggests a spatial trend . By contrast , the semivariogram of the Pearson's residuals of model 1 ( Figure 2B ) demonstrated second-order spatial autocorrelation ( i . e . local clustering ) . The semivariograms of the Pearson's residuals of model 2 ( Figure 2C ) did not show spatial autocorrelation . In this model , the range of spatial autocorrelation can be calculated by 3/φ and is thus 0 . 07 decimal degrees ( approximately 8 km ) . This value is indicative of the radius of trachoma clusters , as it represents the separating distance between two points at which spatial autocorrelation is <5% . Here we present spatial predictions based on model 2 , which had the lowest DIC . The map of the posterior median predicted prevalence of active trachoma ( Figure 3 ) shows high predicted prevalence throughout central , northern and south-eastern Southern Sudan . Low predicted prevalence was apparent in the south-west , which were generally areas with higher long-term average rainfall . Examination of the upper and lower quartiles of the posterior distributions of predicted prevalence ( Figures 4 and 5 ) suggest that large parts of Upper Nile , Unity , Jonglei and Eastern Equatoria States have a high probability of being endemic for trachoma , while large areas in the west of the country ( particularly Western Equatoria State , the south-western part of Central Equatoria State and the southern part of Western Bar-el-Ghazal State ) are unlikely to be at risk of trachoma . We can be particularly confident of the low predicted prevalence values in these latter states because of the low prediction standard errors . A map of the geostatistical random effects ( Figure 6 ) suggests areas of high residual risk of active trachoma ( after accounting for the fixed effects , rainfall , land cover , age and sex ) in Upper Nile , Jonglei , Unity and Central and Eastern Equatoria States , and areas of low residual risk in Western and Eastern Equatoria and Northern Bahr-el-Ghazal States . From the posterior distributions of predicted prevalence , we also determined the probability that predicted prevalence of active trachoma was >10% ( Figure 7 ) , an indication as to whether antibiotic MDA is required – actual MDA decisions are based on prevalence of only TF , not TF plus TI , in children age 1–9 years as determined through PBPS [9] . Nevertheless , our probability map indicates that prevalence of active trachoma in much of south-western Southern Sudan is likely to be below the MDA intervention threshold . Validation analysis of model 2 ( Table 3 ) at the individual level found , pooled across all subsets , an area under the ROC of 0 . 80 ( 95% CI 0 . 79 , 0 . 81 ) , indicating good discriminatory performance of the model for an individual's probability of having active trachoma . At the location level , model 2 had excellent predictive ability to discriminate prevalence of active trachoma relative to thresholds of 0% , 10% , 40% and 70% , with areas under the ROC of 0 . 96 ( 95% CI 0 . 93 , 0 . 99 ) , 0 . 96 ( 95% CI 0 . 93 , 1 . 00 ) , 0 . 92 ( 95% CI 0 . 87 , 0 . 98 ) and 0 . 80 ( 95% CI 0 . 72 , 0 . 88 ) respectively ( pooled across all subsets ) . Mean prediction error was −0 . 012 and mean absolute prediction error was 0 . 170 , indicating that , on average , the model under-predicted prevalence by 1 . 2% and model predictions were different from the observed prevalence by 17 . 0% .
The present study set out to identify areas of Southern Sudan that are of low priority with regards to trachoma control , so that the limited resources available to the National Trachoma Control Program and its implementing partners can be targeted to areas most in need of intervention . Using a Bayesian geostatistical model we determined that prevalence of active trachoma is associated with long-term average rainfall , and that the model containing this variable reliably predicted areas at risk of trachoma transmission . The resulting risk maps show that trachoma control activities need to focus on the centre , north and east of the country , and that large areas in the south-west can , for now , receive a low priority . The predictions were also consistent with prior knowledge of the distribution of trachoma in Southern Sudan . Western Equatoria State , predicted to be of low transmission risk , borders with the Democratic Republic of Congo , which is anecdotally believed to be relatively free from trachoma . Jonglei , Eastern Equatoria and Upper Nile States , in contrast , were predicted to be at risk of high transmission and border parts of Ethiopia , which is known to be highly trachoma endemic [34] . Our findings that older children have a lower prevalence of trachoma than younger children and that an individual's sex is not an important risk factor are consistent with the published literature [7] , [8] . Similarly , the finding that rainfall is an important predictor of trachoma transmission in Southern Sudan confirms earlier results of studies from Sudan and Mali , demonstrating that active trachoma was more prevalent in more arid areas [14] , [35] . Possible explanations for this observation are that dry conditions: i ) might promote trachoma by desiccating the conjunctiva , making it more susceptible to infection , and/or ii ) increase the amount of dust particles in the air , hence increasing irritation of the conjunctiva and providing a vehicle for C . trachomatis to come into contact with the eye [14] . Access to water may also be limited in dry areas , in turn affecting bodily hygiene measures , such as hand and face washing , hence increasing trachoma transmission by hand-to-eye contact . Lack of water is a known risk factor for trachoma [36] , [37] . Additionally , semi-arid areas often tend to be inhabited by seasonally nomadic pastoralists [8] who generally have very low access to sanitation facilities and often defecate in animal pens close to the living areas , hence providing an ideal habitat for the trachoma-transmitting fly Musca sorbens in or near their home compounds [38] , [39] . It is likely that livelihoods , particularly livestock raising , in addition to other ethnicity-related factors ( e . g . house construction methods , isolation of ethnic areas from health centres , and socioeconomic status ) are major risk factors for trachoma in Southern Sudan [8] . A limitation of the analysis is the static nature of our model . Seasonal variation in trachoma has been demonstrated [40] , but our models did not consider the season in which the data were collected . Substantially more data would be required to predict the spatiotemporal distribution of trachoma in Southern Sudan . A second , clear limitation is the geographical spread of the data , which in some states were obtained from clusters of neighbouring communities , resulting in uneven geographical coverage . This is not surprising given that spatial analysis was not a primary objective of the surveys at the time of their implementation . Uneven geographical coverage of Southern Sudan means that the spatial predictions are more precise , and likely to be more accurate , in areas that are in close proximity to the survey locations , and relatively imprecise and less accurate in areas where there are few data points . While we are less confident of our predictions in some areas compared to others , our analytical approach has the considerable advantage that we can quantify and harness these uncertainties to prioritise future data collection in areas of the country where our predictions are less precise . The maps developed here can be used , in the first instance , to prioritise surveys aimed at confirming suspected high-risk areas and at generating baseline data to monitor and evaluate subsequent interventions in currently non-targeted areas [28] . The risk maps thus provide a useful complementary tool to trachoma rapid assessments ( TRA ) and PBPS [41] in that they help to identify areas where collection of additional data would be most useful . Over time , the model presented here can be refined by incorporating new data collected in the identified high risk areas , in turn reducing the uncertainties of the spatial predictions . The findings presented here are in fact the result of multiple iterations , whereby additional data , generated by georeferencing additional sites from previous PBPS were used to revise the spatial models and risk maps . A similar approach could be taken in other countries where some trachoma prevalence data are already available , although this would probably require building of in-country capacity for spatial analysis and/or partnering with international experts . As in Southern Sudan , these data could form the basis for an initial model determining where additional surveys would be most informative . In countries with no or very little trachoma prevalence data it may be advisable to randomly survey individuals ( as outlined in the PBPS methodology [31] ) in a limited number of locations over a large geographical area , followed by development of a risk map . Suspected high-risk areas can then be targeted with TRAs , followed by PBPS in confirmed endemic areas . The risk maps also provide a useful tool to target SAFE interventions . The National Trachoma Control Programme in Southern Sudan now has information that allows it to categorize the south-western part of the country as low priority for further surveys , with resources being conserved for central , northern and eastern areas where trachoma is more likely to be endemic . Being able to present these findings in the form of a comprehensive risk map may also make it easier for the MoH-GoSS to engage the broad range of stakeholders that needs to be mobilized to deliver a comprehensive SAFE strategy . Once more data on other NTDs are available , such as schistosomiasis , soil-transmitted helminthiasis and lymphatic filariasis [31] , [42] , the approach used here can be applied to develop a co-endemicity map that identifies where integrated control of these diseases is warranted [21] . We have demonstrated that trachoma risk mapping , based on integration of field survey and environmental data in statistically robust , spatial statistical models , was achievable and useful in Southern Sudan . Risk mapping is therefore likely to also be applicable to other trachoma endemic settings . | Trachoma , caused by the bacterium Chlamydia trachomatis , is the leading cause of preventable blindness worldwide and a major cause of blindness in Southern Sudan . However , the trachoma distribution in Southern Sudan has only been partially established and many communities in need of intervention have not been identified or targeted . Incomplete mapping and intervention coverage is largely attributable to trachoma resources being scarce and not always deployed most efficiently . The present study aimed at improving programme efficiency by developing maps to help target the available resources for trachoma surveys and interventions to areas where these are most needed . Data on active trachoma prevalence , collected during baseline surveys between 2001 and 2009 , were incorporated into Bayesian geostatistical models to develop a national trachoma risk map . The model predicted the west of the country to be largely at no or very low trachoma risk , while most of the high-risk areas are located in the centre , north , and south-east . Risk mapping has allowed Southern Sudan's trachoma control programme to identify areas where collection of additional data would be most useful . As a direct result , baseline data were collected in March 2010 for the whole of Unity State , with antibiotic mass drug administration being scaled up from June 2010 onwards . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"public",
"health",
"and",
"epidemiology/epidemiology",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"infectious",
"diseases/bacterial",
"infections",
"mathematics/statistics"
] | 2010 | Targeting Trachoma Control through Risk Mapping: The Example of Southern Sudan |
The threat of viral pandemics demands a comprehensive understanding of evolution at the host–pathogen interface . Here , we show that the accessibility of adaptive mutations in influenza nucleoprotein at fever-like temperatures is mediated by host chaperones . Particularly noteworthy , we observe that the Pro283 nucleoprotein variant , which ( 1 ) is conserved across human influenza strains , ( 2 ) confers resistance to the Myxovirus resistance protein A ( MxA ) restriction factor , and ( 3 ) critically contributed to adaptation to humans in the 1918 pandemic influenza strain , is rendered unfit by heat shock factor 1 inhibition–mediated host chaperone depletion at febrile temperatures . This fitness loss is due to biophysical defects that chaperones are unavailable to address when heat shock factor 1 is inhibited . Thus , influenza subverts host chaperones to uncouple the biophysically deleterious consequences of viral protein variants from the benefits of immune escape . In summary , host proteostasis plays a central role in shaping influenza adaptation , with implications for the evolution of other viruses , for viral host switching , and for antiviral drug development .
RNA viruses are exceptionally efficient pathogens that leverage host machineries to replicate their genetic material , synthesize their proteins , and assemble new virions . Perhaps their most remarkable feature , however , is the capacity to rapidly evolve in the face of environmental and immune system pressures . Such rapid evolution is largely mediated by a high mutation rate [1] . Despite its adaptive benefits , rapid genetic change comes at a significant cost for evolving proteins . The majority of amino acid substitutions ( especially functionally relevant substitutions that alter or create protein activities ) are biophysically deleterious , negatively affecting either protein folding or stability [2–6] . A striking illustration of this phenomenon in RNA viruses is influenza nucleoprotein ( NP ) . NP is a globular protein that oligomerizes to encapsulate influenza genomic material and mediate its import into the host nucleus , a process that is required for transcription and replication of the viral genome [7] . NP is strongly conserved relative to the highly variable influenza surface proteins targeted by antibodies [8] . However , NP does experience significant selection pressure from the host immune system , including from innate immune restriction factors [9 , 10] . In particular , the human restriction factor Myxovirus resistance protein A ( MxA ) can prevent influenza ribonucleoprotein import [11–13] , cutting short the viral replication cycle . Adaptive mutations in NP that allow escape from human MxA are critical for the efficient replication of new influenza strains in humans following zoonotic transmission . In just the last century , nonhuman influenza NP was introduced into circulating human influenza strains in 2009 and probably in 1918 [14 , 15] , leading to the acquisition of MxA resistance and ultimately to global pandemics [16 , 17] . While NP evolution is driven by immune escape , it is nonetheless clear that a delicate balance exists between immune system resistance and NP stability and folding . Several NP substitutions known to engender immune escape are destabilizing [9 , 10] , impairing viral growth in the absence of immune pressure [16] . Apparently , the evolving virus must balance the costs of a NP folding defect with the benefits of escaping host immunity . In theory , any mechanism that allows influenza ( or other viruses ) to uncouple protein folding versus immune escape selection pressures would have tremendous benefits for the pathogen . Recent work by us and others suggests that the host’s heat shock protein 90 ( Hsp90 ) chaperone can modulate the evolutionary paths traversed by viruses [18 , 19] . Neither the mechanism of Hsp90’s effect on viral evolution nor its relevance to actual viral strains is clear . A provocative possibility , not yet experimentally explored , is that hijacked host chaperones , whether Hsp90 or any of the other dozens of chaperones beyond Hsp90 that interact extensively with influenza [20 , 21] , potentiate viral evolution directly by assisting the folding of biophysically defective NP variants that would otherwise be insufficiently fit to persist in the population . If this hypothesis is correct , it would suggest that subversion of host protein folding chaperones by viruses can make otherwise inaccessible mutational trajectories leading to immune system escape possible , specifically by promoting the folding of escape variants . Such a mechanism for host adaptation would have broad implications for the evolution of not only influenza NP but also other influenza proteins and other viruses . Rigorously testing this hypothesis requires a method to systematically and quantitatively evaluate whether and how host proteostasis modulates viral protein mutational tolerance . Here , we achieve that objective using deep mutational scanning [22] of influenza NP . We apply deep mutational scanning in the context of chemical genetic inhibition of the host’s heat shock factor 1 ( HSF1 [23] ) to create biophysically challenging , chaperone-depleted cellular protein folding environments . This high-throughput approach revealed a number of amino acid positions in influenza NP whose mutational tolerance is strongly reduced in chaperone-depleted host cells . We confirmed the strong effects of host chaperones on NP mutational tolerance at a number of these sites using head-to-head competition experiments . Most strikingly , the strongly conserved proline ( Pro ) residue at site 283 in NP is rendered highly unfit by HSF1 inhibition at febrile temperatures . Pro283 in NP is known to facilitate escape from the human innate immune system restriction factor MxA [9 , 16] , a feature that critically enhanced the pathogenicity and fitness of the 1918 pandemic influenza strain . We further show that Pro283 disrupts a key structural element in NP , rendering it unstable and aggregation-prone . Host chaperones resolve this folding problem , allowing Pro283 to persist in the viral population and thereby promoting MxA escape . Collectively , our data demonstrate that viral hijacking of host chaperones addresses critical biophysical defects that would otherwise sensitize the virus to host restriction factors . This phenomenon thus has tremendous impact on the capacity of viruses to adapt to their environments , emphasizing the central importance of a hitherto underappreciated element of the host–pathogen interaction and potentially providing new targets for antiviral intervention .
We used a deep mutational scanning strategy to systematically and experimentally quantify the fitness of nearly all viable single amino acid substitutions in influenza NP in both basal and biophysically challenging host cell environments . To this end , we employed previously reported duplicate NP mutant libraries based on the human-adapted A/Aichi/2/1968 ( H3N2 ) influenza strain [24] and competed each viral library in Madin Darby canine kidney ( MDCK ) cells . The MxA orthologs in MDCK cells are inactive against all influenza strains tested to date [25] and permit robust influenza growth . To create a chaperone-depleted host cell environment for viral propagation , we employed a highly specific , chemically inducible dominant negative form of HSF1 [23] , which is the master regulator of cytosolic and nuclear chaperone levels [26] . At both permissive ( 37°C ) and biophysically restrictive , fever-like ( 39°C ) temperatures , chemical induction of the HSF1 inhibitor reduces cytosolic chaperone transcript and protein levels in both the absence ( S1 and S2A Figs; S1 and S2 Data ) and presence ( S2A and S2B Fig; S2 Data ) of influenza . Notably , the condition in which HSF1 was inhibited ( HSF1i ) did not significantly alter the replication of wild-type influenza or host cell metabolic fitness over the course of our experiment ( S2C and S2D Fig; S2 Data ) . We anticipated that the fever-like temperature would prove moderately more challenging for NP folding , potentially restricting the accessible mutational landscape , while the depletion of cytosolic chaperones at each temperature would assess the potential function of host chaperones in regulating NP variant fitness . To maintain library diversity ( approximately 10 , 000 single amino acid substitutions ) whilst minimizing coinfection , we infected 107 cells in each host environment ( Fig 1A ) with 106 infectious virions from our biological duplicate viral NP libraries . In addition to performing deep mutational scanning in biological duplicate , we also performed technical duplicates with one of the replicate libraries ( S3A Fig ) . Following a 48-hour infection , we utilized a previously reported barcoded subamplicon sequencing strategy [27] to accurately quantify the abundance of NP variants after replication at both 37°C and 39°C with and without host chaperone depletion ( S3B–S3D Fig ) . The change in variant frequency upon selection was then normalized to that of the wild-type residue , such that the resulting differential selection [28] value provides a quantitative measure of the relative fitness of each NP variant in the conditions tested . The results of this analysis can be visualized on sequence logo plots ( see S4–S7 Figs for complete NP logo plots and S3 and S4 Data for complete differential selection data ) . As expected , we found that the fever-like temperature is restrictive , generally reducing the fitness of NP variants relative to the wild-type sequence ( see net site differential selection plots in Fig 1B and also the logo plot in S6 Fig—Basal 39°C versus Basal 37°C ) . Host chaperone depletion caused by inhibition of HSF1 modestly reduces NP variant fitness on average ( Fig 1B and S5 and S7 Figs—HSF1i 39°C versus Basal 39°C and HSF1i 37°C versus Basal 37°C ) , whereas chaperone depletion at the restrictive , fever-like temperature substantially reduces the fitness of NP variants ( Fig 1B and S4 Fig—HSF1i 39°C versus Basal 37°C; statistical analyses in Materials and methods ) . Most strikingly , chaperone depletion at an elevated temperature results in very high levels of differential selection at several specific sites in NP , including sites 283 , 334 , 353 , and 377 ( Fig 1C and S4 Fig—HSF1i 39°C versus Basal 37°C ) . At each of these sites , multiple amino acids confer strongly enhanced fitness relative to the wild-type residue . This phenotype is specifically revealed upon host chaperone depletion at 39°C , as we do not observe significant positive differential selection at these sites upon depleting chaperones at 37°C ( Fig 1C and S7 Fig—HSF1i 37°C versus Basal 37°C ) and observe only modest positive differential selection upon increasing the temperature in an environment with basal levels of chaperones ( Fig 1C and S6 Fig—Basal 39°C versus Basal 37°C ) . Competitions between thousands of variants are often inherently noisy , largely owing to differences in variant composition between replicate libraries [27] . However , we observed strong correlation for selection on NP sites between biological ( R2 = 0 . 71 ) and technical ( R2 = 0 . 79 ) replicates of our deep mutational scanning experiments ( S3E Fig and S3 and S4 Data ) . Thus , the selection strength imparted by host chaperone depletion at a restrictive , fever-like temperature substantially exceeds the experimental noise . Moreover , these deep mutational scanning results were recapitulated in all pairwise competitions we performed between individual highly selected variants and virus carrying the wild-type NP ( Fig 2 and S8 Fig ) , and no significant fitness change was observed for a synonymous variant used as a control ( Fig 2B , false discovery rate [FDR] = 0 . 4 ) . Cumulatively , these findings confirm the validity of the deep mutational scanning data and indicate that depleting chaperones at a moderately elevated temperature creates a stringent host environment that strongly selects against wild-type residues in NP at certain sequence positions . The strong differential selection we observed in chaperone-depleted host cells at a restrictive temperature suggested to us that specific wild-type NP residues in the Aichi influenza strain , the strain our libraries are based on , may entail a biophysical cost that is nonetheless tolerated under permissive conditions , perhaps owing to competing selection forces . This hypothesis is particularly compelling for the previously characterized stabilizing NP variants [10] N334H ( Fig 1C and S4 Fig ) and M136I ( S4 Fig ) , which were positively selected in biophysically challenging conditions . These variants exhibited modestly enhanced fitness upon increased temperature and significantly enhanced fitness upon chaperone depletion at increased temperature . The unmasking of the deleterious fitness effects of the wild-type sequence at these positions upon host chaperone depletion supports the hypothesis that chaperones can indeed rescue biophysically deleterious NP variants . We observed a similar and even more striking trend at site 283 , where numerous amino acid substitutions—including serine ( Ser ) , leucine ( Leu ) , threonine , and glycine ( Gly ) —were validated in our pairwise competition experiments to be significantly more fit than the wild-type Pro residue when host chaperones were depleted at a restrictive temperature ( Fig 1C and Fig 2B—HSF1i 39°C versus Basal 37°C ) . The identity of the amino acid at NP site 283 is known to critically modulate influenza sensitivity to the human antiviral restriction factor MxA , with Pro at that position contributing greatly to MxA escape [9 , 16] . Although a structure of the MxA:NP complex is not currently available , site 283 is likely located at the MxA–NP interface [29] . Moreover , Pro283 is nearly universally conserved in human influenza NP but rarely occurs in avian influenza NP . This characteristic difference between human and avian influenza strains is attributed to the necessity of Pro283 to escape human MxA , whereas the avian MxA ortholog lacks known antiviral activity [30] . Indeed , the Leu283Pro substitution enabled the 1918 pandemic influenza strain to escape MxA following zoonotic transmission [16] . Pro283 in NP has thus greatly impacted the fitness of modern human influenza strains . The observation that an adaptive amino acid substitution as important as Pro283 can be rendered unfit by biophysically challenging , host chaperone–depleted conditions motivated us to elucidate the underlying molecular basis of this phenomenon . Our hypothesis was that the depletion of host chaperones exacerbates a biophysical defect in NP folding that is caused by installation of a Pro at position 283 . Prior work has shown that NP is engaged by numerous cytosolic chaperones , including chaperones like heat shock protein 40 ( Hsp40 ) and heat shock protein 70 ( Hsp70 ) that are depleted in our HSF1-inhibited host cell environment [31–35] ( S1 and S2 Figs ) . For example , Hsp40 ( DNAJB1 ) directly interacts with NP and facilitates its nuclear import [34] , while Hsp70 is implicated in modulating NP nuclear export [35] . The heat shock proteins can also regulate antiviral responses indirectly through their interactions with NP [31–33] . Thus , the disruption of critical NP–host chaperone interactions by HSF1 inhibition-mediated chaperone depletion may indeed be the source of significant differential selection at site 283 , especially at a biophysically restrictive temperature . This possibility raises the question of whether the Pro283 NP variant is , in fact , biophysically defective relative to other variants at site 283 . Although there is currently no high-resolution structure of a NP variant containing Pro283 , crystal structures of avian influenza NP variants with Ser or Leu at position 283 are available [36 , 37] . In these structures , site 283 is in the middle of an α-helix ( Fig 3A ) . Pro is classically regarded as a “helix-breaker , ” owing in part to its inability to form an i + 4 hydrogen bond important for α-helix stability [38] . Therefore , it seemed reasonable to anticipate that the replacement of Ser ( or Leu ) with a Pro at position 283 , as is observed in human influenza strains , would indeed be biophysically problematic . To assess this hypothesis , we first performed molecular dynamics ( MD ) simulations in explicit water to evaluate whether Pro283 affects either the overall structure of NP or the structure of the α-helix centered at position 283 . Although the overall NP structure was not grossly perturbed by Pro283 , as would be expected given that this variant is still capable of supporting influenza replication , our simulations revealed that a Pro283 NP variant is significantly less α-helical at residues 282–284 than is a Ser283-containing variant ( Fig 3B , S1 and S2 Videos and S6 Data ) . The significant structural consequences of Pro283 observed in these simulations prompted us to experimentally investigate whether Pro283 affects NP stability . Following a previously reported protocol [10] , we recombinantly expressed and purified a monomeric form of NP with either Pro , Ser , Leu , or alanine ( Ala ) at position 283 ( S9A–S9C Fig ) . Circular dichroism spectra indicated that all four variants had grossly similar secondary structures ( S9B Fig ) , consistent with our simulations . Thermal denaturation of NP is irreversible , leading to rapid aggregation and precipitation of the protein . Nonetheless , fitting these irreversible thermal melts to a two-state model revealed that Pro283 NP does indeed precipitate at a significantly lower temperature than all three of the non-Pro variants studied ( Fig 3C , S9C Fig and S7 Data ) and is therefore less stable and more aggregation prone . The observation that Pro283 NP has a higher propensity to aggregate than other variants is consistent with either a kinetic or thermodynamic defect caused by substitution with Pro283 . Given that Gly is , like Pro , known as an α-helix breaker and that Gly283 is positively selected relative to Pro283 ( see Fig 1C—HSF1i 39°C versus Basal 37°C ) upon chaperone depletion at febrile temperatures , we favor a substantial contribution from a kinetic defect that may be associated with the propensity of Pro to form both cis- and trans-amide bonds [38] . The substantive biophysical defect endowed by Pro283 on NP likely explains the enhanced dependence of this variant on host chaperones and explains why other variants are significantly more fit in the absence of those key chaperones . Moreover , these results may help to explain why Pro283 is not observed in avian influenza strains [16] . Birds typically have body temperatures ranging from 39–43°C [39] , the upper end of which may be too extreme to permit chaperone-mediated rescue of the biophysically defective Pro283 NP variant . Cumulatively , these results suggest that NP variants critical for innate immune system escape , most especially Pro283 , can be folding-defective and display compromised fitness in biophysically challenging host environments ( Fig 4A , left ) . This finding motivated us to evaluate whether Pro283 fitness remains compromised under biophysically challenging conditions even upon the addition of an MxA selection pressure that normally selects strongly in favor of the Pro283 variant ( Fig 4A , right ) . To this end , we performed pairwise viral competitions between the biophysically stable Ser283 variant and the MxA-resistant Pro283 variant in each of our host environments in the presence of either active or inactive MxA ( Fig 4B and S2E Fig ) . In permissive folding environments ( Basal and HSF1i at 37°C ) , Pro283 was enriched compared to Ser283 , thereby enabling MxA escape ( Fig 4C ) . In contrast , in biophysically challenging environments ( Basal and HSF1i at 39°C ) , the stability defects of Pro283 were exacerbated , and Ser283 was enriched compared to Pro283 . Ser283 was enriched even in the presence of MxA selection pressure when chaperones are depleted at 39°C , thereby hindering immune escape . Altogether , these data reveal that HSF1-regulated chaperones can define the fitness of biophysically destabilized immune escape variants in influenza . This observation suggests a model in which enhanced fitness conferred by immune escape is often counterbalanced over the course of influenza evolution by biophysical defects that have a substantive fitness cost ( Fig 5 , left ) . Remarkably , at least in the case of influenza NP , our data show that the virus is able to hijack host chaperones to resolve these biophysical defects ( Fig 5 , right ) . By this mechanism , the virus manages to uncouple protein folding fitness costs from the advantageous consequences of immune escape , expanding the accessible mutational landscape to access essential viral protein variants capable of both folding and immune escape .
This work provides , to the best of our knowledge , the first direct experimental evidence that host chaperones mediate the accessibility of biophysically deleterious , adaptive viral protein variants . This feature of the host–pathogen interaction is apparent in multiple sites across the NP gene . Particularly noteworthy , we find that the destabilized Pro283 NP variant is not tolerated in a chaperone-depleted host environment at a restrictive temperature , as NP is unable to engage host chaperones to address the Pro283-induced biophysical defect . Remarkably , we observe that even in the presence of selection pressure imposed by MxA that strongly favors Pro283 [9] , the fitness of Pro283 NP is still contingent on the host’s chaperone levels and biophysical environment . Based on these results , we expect that host chaperones can impact the accessibility of adaptive viral protein variants far beyond NP and influenza , as amino acid substitutions are largely destabilizing [6] , and many viral proteins are known to engage host chaperones [34 , 40 , 41] . Moreover , previous work has revealed ( 1 ) that viral evolution is fundamentally constrained by protein stability [4 , 10] and ( 2 ) the role of the Hsp90 chaperone in viral replication across numerous viral families [41–44] . Thus , the evolutionary trajectories of diverse viral proteins are likely to be influenced by numerous host chaperones [18 , 19] . Since our work suggests that host chaperones preferentially rescue biophysically defective viral protein variants , as more data accumulate in this field , we may eventually be able to predict how chaperones will impact fitness in a rational manner based on protein variant biophysical properties . Further , our infections in the presence of the MxA restriction factor demonstrate that host chaperones can mediate the accessibility of escape variants irrespective of competing selection pressures . These data raise the possibility of antiviral therapeutic adjuvants targeting host chaperones that inhibit the development of antiviral resistance by constraining the accessible mutational landscape . We further observe that temperature critically influences the fitness of viral variants , with most variants suffering fitness costs at elevated temperatures that mimic fever conditions and/or the body temperatures of birds and small mammals [39] . Thus , fever and host-switching events may impose selection on viral variants that hampers adaptation . Based on our findings here , the nature of such selection is likely to be strongly influenced by host-specific differences in chaperone network compositions . We anticipate that these phenomena extend far beyond the host–pathogen interface and apply to protein evolution more broadly . Previous work by Lindquist and others suggested that the Hsp90 chaperone can potentiate and buffer genetic variation in endogenous proteins [45–49] . Here , the impact of the host chaperone environment on NP variant fitness is driven predominantly not by Hsp90 ( see S10 Fig ) , even though NP does engage this chaperone [50] , but instead by inhibition of HSF1 modulating the composition of a complex network of multiple protein folding and quality control factors . Moreover , our work shows experimentally that chaperones have the largest effect on the fitness of biophysically defective protein variants , a result that may help to explain extensive prior work with Hsp90 in which the potential biophysical mechanism of effects on protein evolution have not been experimentally evaluated . HSF1-regulated host proteostasis network components may modulate NP evolution directly—for example , by impacting NP–chaperone interactions—or indirectly , perhaps by perturbing levels of endogenous chaperone client proteins with antiviral properties . Our data support the former case , as we find that destabilized NP variants are particularly sensitive to chaperone depletion . Nonetheless , we would not rule out possible contributions from secondary effects . Deciphering between primary and secondary effects will first require identifying the individual components of the intricate protein folding network that are primarily responsible for modulating NP fitness , followed by systematic elimination of potential downstream effectors . Whether primary or secondary , the evolutionary implications of the protein folding network clearly extend well beyond Hsp90 and also play critical roles in evolution at the host–pathogen interface . Finally , this work provides experimental evidence for the longstanding hypothesis that chaperones buffer the fitness cost of biophysically destabilized protein variants [45 , 51 , 52] . Experimental validation of this concept in metazoan cells for the first time , to the best of our knowledge , has significant consequences for understanding the constraints on protein evolution , which have so far focused on inherent biophysical properties of proteins [2 , 10] . For specific destabilized adaptive variants , fitness has been attributed to compensatory stabilizing mutations elsewhere in the protein structure [10 , 53] . Cases of this idiosyncratic epistasis mediating pathogen adaptation have motivated efforts to determine the pervasiveness of compensatory mutations [10] . Our data reveal that the constraints on protein evolution are still more complex , establishing that protein variant fitness is constrained not just by inherent stability but also by the cellular environment in which the protein folds .
The following plasmids were used to generate the A/Aichi/2/1968 influenza virus: pHWAichi68-NP [54] , pHWNan95-PB2 [54] , pHWNan95-PB1 [54] , pHWNan95-PA [54] , pHW184-HA [55] , pHW186-NA [55] , pHW187-M [55] , and pHW188-NS [55] . For NP recombinant expression and biophysical studies , a pET28b ( + ) expression vector encoding monomeric A/Aichi/2/1968 NP with an R416A mutation ( which prevents RNA binding ) , deletion of residues 2–7 , and a C-terminal 6×-His tag was used [10] . A lentiviral vector containing a FLAG-tagged human MxA or inactive MxA ( T103A ) sequence under a CMV promoter with a GSG linker ( DYKDDDDKGSG ) at the C-terminus was used for generation of the MxA-expressing MDCK cell line [9] . Downstream of MxA , the plasmid contained an internal ribosome entry site ( IRES ) followed by an mCherry reporter gene to assist the selection of stable single-colony cell lines . Antibodies used were as follows: rat monoclonal anti-FLAG ( Agilent; 200474 ) , mouse monoclonal anti-β-actin ( Sigma; A1978 ) , rabbit polyclonal anti-Hsp90 , rat monoclonal anti-Hsp70 , and rabbit monoclonal anti-Hsp40 from Cell Signaling ( 4877 , 4873 , and 4871 , respectively ) . IRDyes 800CW goat anti-rat , 680LT goat anti-mouse , and 800CW goat anti-rabbit secondary antibodies were obtained from LI-COR ( 926–32219 , 926–68020 , and 926–32211 , respectively ) . For the MDCKdn-cHSF1 cell line construction , MDCK cells were originally purchased from the American Type Culture Collection ( Manassas , VA , United States ) and validated as MDCKs by STR profiling ( Science Exchange ) . Cells were cultured at 37°C in a 5% CO2 ( g ) atmosphere in DMEM ( CellGro ) supplemented with 10% fetal bovine serum ( FBS; CellGro ) and 1% penicillin/streptomycin/glutamine ( CellGro ) . The parental MDCK cells were transduced first with lentivirus encoding a blasticidin-resistant tetracycline repressor and then with lentivirus encoding a zeocin-resistant , tetracycline-inducible dn-cHSF1 construct [23] . Heterostable cells expressing the tetracycline repressor and the dn-cHSF1 construct were then selected using 8 μg/mL zeocin and 4 μg/mL blasticidin . Single colonies were generated by serial dilution in 96-well plates , expanded , and then selected based on functional testing of HSF1 inhibition [23] using qPCR ( S2B Fig ) , as previously described . For MDCKdn-cHSF1-MxA and MDCKdn-cHSF1-MxA ( T103A ) cell line construction , MDCKdn-cHSF1 cells were engineered to constitutively express human MxA or the inactive MxA-T103A mutant [9] by transducing with lentivirus encoding FLAG-tagged MxA variants . At 72 h post transduction , single colonies were generated by serial dilution in 96-well plates . Wells with clonal transduced cells were identified as single clusters of cells expressing mCherry , expanded , and then characterized as described below . Two independent plasmid mutant libraries previously generated [24] from the pHWAichi68-NP template plasmid were used to create mutant viral libraries by transfecting a coculture of 25 , 000 MDCK-SIAT1 and 300 , 000 HEK 293T cells , as previously described [24] . For each library , cocultures in eight 6-well plates were transfected to maintain library diversity , and transfection supernatants were combined to generate the input mutant viral libraries . Viruses were generated and grown in WSN media . Infectious titers of viral libraries were determined by a tissue culture infectious dose ( TCID50 ) assay . Briefly , 10-fold serial dilutions of viruses ( in technical triplicates ) were prepared in 96-well plates and incubated with 5 × 103 MDCK-SIAT1 cells/well for 72 h at 37°C . The wells were then scored for cytopathic effects , and viral titers were calculated using a Reed-Muench Calculator , available at https://github . com/jbloomlab/reedmuenchcalculator . Two plasmid libraries were used to generate two biological replicate viral libraries , one of which was used twice to perform 2 technical replicates of the deep mutational scanning ( S3A Fig ) . MDCKdn-cHSF1 cells were plated in 15-cm plates at a density of 6 × 106 cells/dish and treated with 0 . 1% DMSO or 1 μg/ml doxycycline for 24 h at either 37°C or 39°C . Deep mutational scanning was also performed in cells treated with an Hsp90 inhibitor ( 100 nM 17-AAG; 90 min pretreatment ) . Then , 5 × 106 infectious virions ( as determined by a TCID50 assay ) from each viral library were used to infect four 15-cm plates from each condition at an MOI of 0 . 1 virions/cell ( the cells expanded to approximately 12 . 5 × 106 cells per plate by the time of infection ) . In addition , one 15-cm plate at both 37°C and 39°C was either mock-infected ( negative control ) or infected with wild-type virus . For infection , the cellular growth media were replaced with WSN media containing a mutant virus library , wild-type virus , or no virus for mock infection . After 2 h , the inoculum was replaced with fresh WSN media containing 0 . 1% DMSO , 1 μg/mL doxycycline , or 100 nM 17-AAG . At 48 h post infection , the viral supernatant was harvested , centrifuged at 1 , 000 × g for 5 min to remove cell debris , and stored at −80°C . Viral RNA was extracted from the infectious supernatant using a Viral RNA Mini Kit ( Qiagen ) and reverse transcribed using the AccuScript High Fidelity 1st Strand cDNA Synthesis Kit ( Agilent ) using 5′-BsmBI-Aichi68-NP and 3′-BsmBI-Aichi68-NP primers ( primer sequences in S1 Table ) . At least 106 NP molecules were PCR-amplified for preparation of the sequencing libraries , as previously described [27] ( S3B Fig ) . The resulting amplicons were sequenced on an Illumina HiSeq 2500 in rapid run mode with 250-bp paired-end reads ( S3C and S3D Fig ) . dms_tools ( http://jbloomlab . github . io/dms_tools/ ) [56] was used to align reads to the Aichi NP reference sequence , count amino acid variants across NP , and calculate the differential selection for each variant between 2 selection conditions , as previously described [9 , 28] ( S3 and S4 Data ) . NP variants that reproducibly exhibited the most positive or negative differential selection in the deep mutational scan were selected for pairwise competitions ( P283M , P283A , P283G , P283L , P283S , P283T , S353L , S353F , N334H , D34N , H82N , and a synonymous control P283P ) . The NP mutants were generated by introducing point mutations into the pHWAichi68-NP plasmid using the QuikChange II XL Site-Directed Mutagenesis Kit ( Agilent ) . Technical difficulties with specific site-directed mutagenesis reactions prevented generation of the P283M , P283A , and S353L mutant plasmids . The remaining 9 mutant plasmids were generated successfully and used to produce the corresponding mutant viruses by transfecting a coculture of 25 , 000 MDCK-SIAT1 and 300 , 000 HEK 293T cells , as previously described [55] . The resultant viruses were titered using a TCID50 assay . For each competition , 100 , 000 cells/well of MDCKdn-cHSF1 , MDCKdn-cHSF1-FLAG-MxA , or MDCKdn-cHSF1-FLAG-MxA ( T103A ) cells were plated in 12-well dishes and treated with 0 . 1% DMSO or 1 μg/mL doxycycline for 18 h at either 37°C or 39°C . Cells were infected with a 1:1 mixture of wild-type and mutant viruses at an MOI of 0 . 1 virions/cell in triplicate under conditions identical to that of the deep mutational scanning experiment . After 2 h , the inoculum was replaced with fresh WSN media containing either 0 . 1% DMSO or 1 μg/mL doxycycline . At 48 h post infection , infectious supernatants were harvested , centrifuged at 1000 × g for 5 min to remove cell debris , and stored at −80°C . Viral RNA was extracted from the infectious supernatant using the QIAamp Viral RNA Mini Kit , and at least 106 NP molecules were reverse transcribed using the SuperScript III Reverse Transcriptase ( Thermo Fisher Scientific ) with 5′-BsmBI-Aichi68-NP and 3′-BsmBI-Aichi68-NP primers ( S1 Table ) . The amplicons were visualized on a 1% analytical agarose gel to verify amplification of the NP gene ( 1 . 5 kb ) . The dsDNA was purified using 1 . 5× AMPure XP beads ( Beckman Coulter ) and quantified using a Quant-iT PicoGreen Assay ( Life Technologies ) . Illumina NexteraXT sequencing libraries were prepared using a Mosquito HTS Liquid Handler ( TTP Labtech ) and sequenced on an Illumina MiSeqv2 in 2 runs of either 40-bp single-end or 150-bp paired-end reads . To call sequence variants , reads were aligned to the Aichi NP reference sequence using bwa mem ( v . 0 . 7 . 12-r1039 ) ( arXiv:1303 . 3997v2 ) with flag–t 16 , and sorted and indexed bam files were generated using samtools ( v 1 . 3 ) [57] . These bam files were processed using samtools mpileup with flags–excl-flags 2052 , -d 30000000 , and the same Aichi NP reference sequence used for mapping [58] . For pairwise competitions in the absence of MxA , mutant allele frequencies were normalized to wild-type allele frequencies for each sample , and the resulting values were used to calculate the differential selection [28] ( Fig 2 , S8 Fig , and S5 Data ) . For pairwise competitions in the presence of wild-type or inactive MxA , the change in mutant allele frequencies is reported ( Fig 4C and S8 Data ) . Two sets of simulations were performed for NPs with the sequence of the human H3N2 variant ( Fig 3A and 3B ) . In one set , NP residue 283 was Pro ( this system is termed Pro283 hereafter ) . In the other set , residue 283 was Ser ( this system is termed Ser283 hereafter ) . The initial structures of both systems were prepared using the comparative modeling software RosettaCM [59] , with the structures of H1N1 influenza A virus NP ( PDB ID: 2IQH [36] ) and H5N1 NP ( PDB ID: 2Q06 [37] ) used as templates with equal weights . The first 20 amino acids in the N-terminal region , whose 3D coordinates are missing in the template structures , were considered flexible and also of minimal impact to the region near residue 283 . Therefore , they were removed in the following simulation . In the threading procedure , the target NP sequence was aligned with the templates [60] and assigned coordinates from the template PDB structures ( 2IQH and 2Q06 ) . The helix formed by residues 278–286 of 2IQH and 2Q06 was removed to allow RosettaCM to construct this region without influence from the templates . This region ( residues 278–286 ) , along with all the other regions missing in the template PDBs , was patched in the hybridization step . During hybridization , RosettaCM generated hybridized structures that contained pieces from each of the threaded structures , providing more accurate comparative models that were energetically favorable . Additionally , RosettaCM used fragments and minor ab initio folding to fill in residues not previously aligned with any template sequences during the threading process . A total of 1 , 000 models for each NP system were created , and the best-scoring model without a disulfide bond was used as the initial structure for further MD simulations . Five runs of MD simulations for each NP system were carried out using GROMACS with the oplsaa/tip4p force field [61 , 62] . The N-terminus of the initial structure of the MD simulation was capped with an acetyl group , while the C-terminus was free ( ending with COO– ) . This structure was energy minimized in a vacuum and then immersed in the center of a cubic box containing preequilibrated water molecules with an edge of 12 nm . The system was electrostatically neutralized by adding 11 Cl−ions . The solvated system was further energy minimized to remove any bad contacts . The solvated NP then underwent 2 stages of equilibrations . The first stage of equilibration consisted of a 50-ps isochoric–isothermal ( NVT ) simulation at 300 K and a subsequent 50-ps isobaric–isothermal ( NPT ) simulation at 300 K and 1 bar . During the first stage of equilibration , the NP-heavy atoms were restrained by a harmonic potential with a force constant of 1 , 000 kJ mol–1 nm–2 to equilibrate the solvent molecules and adjust the density . The second stage of equilibration consisted of an additional 100-ps NVT simulation at 300 K without any restraints to equilibrate the whole system , followed by a 100-ns NPT production simulation at 300 K and 1 bar . The V-rescale thermostat was coupled to both the NP and solvent separately , with coupling time constants of 0 . 1 ps . The pressure was maintained using the Parrinello-Rahman barostat with a coupling time constant of 2 . 0 ps and isothermal compressibility of 4 . 5 × 10−5 bar-1 . The leapfrog algorithm with a time step of 2 fs was used for dynamics evolution . All bonds involving hydrogen were constrained using the LINCS algorithm . All neighbor searching , electrostatic interactions , and van der Waals interactions were truncated at 1 . 0 nm . Electrostatics were treated using the particle mesh Ewald ( PME ) summation with a Fourier spacing of 0 . 12 nm and an order of 4 . A long-range dispersion correction for energy and pressure was applied to account for the 1 . 0-nm cutoff of Lennard-Jones interactions . Five 100-ns trajectories were produced for the Pro283 and Ser283 systems , respectively . The trajectories between 50 ns and 100 ns were used for analysis ( S6 Data ) . The P283S , P283A , and P283L amino acid substitutions were introduced into the wild-type influenza A/Aichi/2/1968 NP in a pET28b ( + ) expression vector using the QuikChange II XL Site-Directed Mutagenesis Kit ( Agilent ) . This NP construct contained an R416A mutation and deletion of residues 2–7 to obtain nonaggregated , RNA-free NP in a CD-compatible buffer , as previously described [10] . Mutagenized plasmid DNA was isolated using the E . Z . N . A . Plasmid Mini Kit I ( Omega ) . For bacterial expression , BL21 ( DE3 ) chemically competent cells were transformed with 1 μL of purified plasmid and incubated overnight on LB-kanamycin agar plates . Colonies were used to inoculate 50-mL LB-kanamycin cultures overnight . Then , 10 mL of starter cultures were used to inoculate 1 L LB-kanamycin cultures , which were shaken at 37°C until reaching an OD600 of 0 . 3–0 . 6 . Cultures were then chilled on ice and induced with 500 μM IPTG ( Sigma ) overnight at 20°C . Cells were then pelleted , Dounce homogenized , and lysed by sonication in 50 mL of lysis buffer ( 50 mM sodium phosphate at pH 8 . 0 , 500 mM NaCl , 0 . 5% Triton X-100 , 10 mM imidazole , 1 mM PMSF , and 0 . 1 mg/mL MgCl2 ) . Cells were sonicated for 2 min ( 30% amplitude , 10 s on , 10 s off; Branson Digital Sonifier ) . Lysates were then clarified for 30 min at 10 , 000 × g at 4°C . Clarified lysates were passed through 0 . 45-μm filters . His-tagged NP variants were then incubated on Ni-NTA ( Millipore ) columns for 60 min at 4 °C and washed with Ni-NTA Wash Buffer ( 50 mM sodium phosphate at pH 8 . 0 , 300 mM NaCl , and 20 mM imidazole ) . Proteins were eluted using Ni-NTA Elution Buffer ( 50 mM sodium phosphate at pH 8 . 0 , 300 mM NaCl , and 250 mM imidazole ) . Eluates were dialyzed overnight into analysis buffer ( 20 mM sodium phosphate at pH 7 . 0 with 300 mM NaF ) using 3 . 5-kDa molecular mass cutoff SnakeSkin dialysis tubing ( Fisher Thermo Scientific ) . Dialyzed proteins were concentrated using Amicon Ultra 3K MWCO filters ( Millipore ) and further purified over a size exclusion column ( Bio-Rad ENrich SEC 650 ) ( S9A Fig ) . For circular dichroism analysis ( S9B Fig ) , proteins were diluted to 5 μM in analysis buffer ( quantified by A280 with a BioTek-Take3 micro-volume plate using a molar extinction coefficient of 56 , 600 M–1 cm–1 ) . Thermal melts ( S9C Fig ) were performed at a scan rate of 2 °C per min , maintaining each temperature for 5 min before measurement of ellipticity at 209 nm . Tagg values were obtained , as the thermal denaturation of NP was irreversible and resulted in aggregation and precipitation ( Fig 3C and S9C Fig; S7 Data ) . All circular dichroism analyses were performed on a Jasco J-1500 Circular Dichroism Spectrophotometer with a 1-mm QS quartz cuvette ( Hellma ) . Deep mutational scanning was performed in biological duplicate with 2 technical replicates of one of the biological replicates ( S3A Fig ) . MxA and MxA ( T103A ) protein expression in MDCKdn-cHSF1-FLAG-MxA and MDCKdn-cHSF1-FLAG-MxA ( T103A ) cells , respectively , were evaluated in biological duplicate ( S2E Fig ) . All other experiments were performed in biological triplicate , with replicates being independent experimental entireties ( i . e . , from plating the cells to acquiring the data ) . Correlation between deep mutational scanning replicates was determined by linear regression using GraphPad Prism software , reporting R2 ( S3E Fig ) . Site differential selection values from deep mutational scanning ( Fig 1B ) were tested for significance of deviation from zero ( wild-type behavior ) using a one-sample t test in GraphPad Prism . The raw p-values were adjusted for multiple comparison using the Benjamini-Hochberg procedure [63] , setting an acceptable FDR at 0 . 05 ( p . adjust function in R ) . Analysis of variance ( ANOVA ) was performed on all differential selection values across selections normalized to the Basal 37 °C condition using a nested ANOVA framework ( accounting for replicates ) , modeling treatment/temperature as a fixed effect and the replicate as a random effect ( lme function/RMLE in the R statistical environment , followed by ANOVA computation anova . lme with sequential adjustment ) . Post hoc pairwise comparisons were performed by general linear hypotheses testing using Tukey's method ( as implemented in glht , in the multcomp R package ) comparing all the means with single-step adjustment for multiple comparison . The p-values for pairwise comparisons against HSF1i 37°C versus Basal 37°C were 6 . 148 × 10−6 for Basal 39°C versus Basal 37°C and 9 . 196 × 10−8 for HSF1i 39°C versus Basal 37°C and 0 . 71 for comparison between HSF1i 39°C versus Basal 37°C and Basal 39°C versus Basal 37°C ( Fig 1B ) . Differential selection values from pairwise competitions ( Fig 2B and S8 Fig ) were tested for significance of deviation from zero ( wild-type behavior ) using a one-sample t test in GraphPad Prism with FDR correction ( S8 Fig ) . For pairwise competitions in the presence of MxA , the significance of deviation from the input Ser283 frequency was determined using a one-sample t test in GraphPad Prism followed by FDR correction; Basal 37 °C Inactive MxA ( t = 5 . 916 , df = 2 ) ; Basal 37°C Active MxA ( t = 29 . 01 , df = 2 ) ; HSF1i 37°C Inactive MxA ( t = 18 . 24 , df = 2 ) ; HSF1i 37 °C Active MxA ( t = 17 . 91 , df = 2 ) ; Basal 39°C Inactive MxA ( t = 4 . 924 , df = 2 ) ; Basal 39 °C Active MxA ( t = 0 . 5877 , df = 2 ) ; HSF1i 39°C Inactive MxA ( t = 18 . 98 , df = 2 ) ; HSF1i 39 °C Active MxA ( t = 12 . 29 , df = 2 ) . ANOVA was performed for competitions in presence of MxA as described above; individual p-values for all pairwise comparisons using Tukey all-pair comparisons method are provided in S9 Data . For RNA-seq , log2 fold changes , p-values , and Benjamini-Hochberg-adjusted p-values ( ADP ) are reported for all expressed protein-coding genes in S1 Data . For MD simulations , each of the five 100-ns simulations was considered an independent replicate , and the %-time spent in an α-helical conformation was transformed using ln ( P/ ( 1–P ) ) before t tests were conducted to satisfy the prerequisite assumptions of normality ( Fig 3B ) ; 278 ( t = 0 . 95655 , df = 7 ) ; 279 ( t = 0 . 52331 , df = 6 ) ; 280 ( t = −0 . 46542 , df = 6 ) ; 281 ( t = −1 . 21825 , df = 6 ) ; 282 ( t = −3 . 64906 , df = 7 ) ; 283 ( t = −2 . 83812 , df = 8 ) ; 284 ( t = −2 . 83606 , df = 8 ) ; 285 ( t = 0 . 17240 , df = 7 ) ; 286 ( t = 0 . 34125 , df = 8 ) . For apparent melting temperatures determined by circular dichroism , melt curves were performed in biological triplicate , the average and SEM are reported , and the significance of deviation from wild type was evaluated by a Student t test; Ser ( t = 19 . 59 , df = 4 ) ; Ala ( t = 23 . 94 , df = 4 ) ; Leu ( t = 5 . 644 , df = 4 ) . | Viruses , such as influenza , evade the host immune response by mutating frequently . However , these adaptive amino acid substitutions are often biophysically deleterious and can thus increase the propensity for viral proteins to misfold and hamper viral replication . Host protein folding factors called chaperones interact extensively with viral proteins , like influenza nucleoprotein , and are thus poised to potentiate the fitness of biophysically defective , adaptive variants . Here , we directly test this hypothesis by quantitatively profiling the mutational tolerance of influenza nucleoprotein in host cells with reduced chaperone levels . We find that chaperones indeed increase the accessibility of destabilized adaptive nucleoprotein variants , with an especially strong effect at fever-like temperatures . We observe that the destabilized Pro283 nucleoprotein variant , which is universally conserved across human influenza strains and enables evasion of the Myxovirus resistance protein A ( MxA ) innate immunity restriction factor , is rendered unfit in a chaperone-depleted host environment . Together , these data show that host chaperones critically impact viral adaptation and may serve as targets for antiviral therapeutic adjuvants . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"organismal",
"evolution",
"medicine",
"and",
"health",
"sciences",
"cellular",
"stress",
"responses",
"pathology",
"and",
"laboratory",
"medicine",
"influenza",
"pathogens",
"cell",
"processes",
"microbiology",
"orthomyxoviruses",
"viruses",
"nucleoproteins",
"rna",
"vir... | 2018 | Destabilized adaptive influenza variants critical for innate immune system escape are potentiated by host chaperones |
Environmental factors during early life are critical for the later metabolic health of the individual and of future progeny . In our obesogenic environment , it is of great socioeconomic importance to investigate the mechanisms that contribute to the risk of metabolic ill health . Imprinted genes , a class of functionally mono-allelic genes critical for early growth and metabolic axis development , have been proposed to be uniquely susceptible to environmental change . Furthermore , it has also been suggested that perturbation of the epigenetic reprogramming of imprinting control regions ( ICRs ) may play a role in phenotypic heritability following early life insults . Alternatively , the presence of multiple layers of epigenetic regulation may in fact protect imprinted genes from such perturbation . Unbiased investigation of these alternative hypotheses requires assessment of imprinted gene expression in the context of the response of the whole transcriptome to environmental assault . We therefore analyse the role of imprinted genes in multiple tissues in two affected generations of an established murine model of the developmental origins of health and disease using microarrays and quantitative RT–PCR . We demonstrate that , despite the functional mono-allelicism of imprinted genes and their unique mechanisms of epigenetic dosage control , imprinted genes as a class are neither more susceptible nor protected from expression perturbation induced by maternal undernutrition in either the F1 or the F2 generation compared to other genes . Nor do we find any evidence that the epigenetic reprogramming of ICRs in the germline is susceptible to nutritional restriction . However , we propose that those imprinted genes that are affected may play important roles in the foetal response to undernutrition and potentially its long-term sequelae . We suggest that recently described instances of dosage regulation by relaxation of imprinting are rare and likely to be highly regulated .
Animal models in multiple species have confirmed that early life represents a critical window of phenotypic plasticity , highly responsive to maternal behaviour , stress , metabolism and nutrition ( reviewed by [1] ) . Epigenetic mechanisms , “the structural adaptation of chromosomal regions so as to register , signal or perpetuate altered activity states” [2] are fundamentally involved in the specification of cellular phenotype . We and others have hypothesised that a compromised in utero environment may impinge upon the epigenetic apparatus with lasting consequences for gene expression and development . Changes in DNA methylation and histone modifications at putative regulatory regions correlating with the altered expression of genes implicated in phenotypic development have been observed in a number of animal models of early life compromise [3]–[8] . Such epigenetic modifications are hypothesised to contribute to the stable maintenance of phenotype long after exposure to the environmental insult . The impact of the early life environment has been observed to extend over multiple generations in both human populations and animal models ( for example [4] , [9]–[11] ) . Several potential mechanisms of such non-Mendelian phenotypic inheritance can be considered . For example , transmission via the maternal line often , though not always , involves the recapitulation of the initial environmental trigger , as with the heritability of maternal reproductive behaviour [3]–[4] , [12] . However , paternal transmission of environmentally induced phenotypes has also been documented [13]–[14] , [15]–[17] . This strongly implicates intergenerational epigenetic inheritance because rodent males only contribute to the future generation through the sperm . However , which epigenetic mechanism ( s ) are responsible remains unknown . Transgenerational epigenetic inheritance of DNA methylation has been demonstrated through both maternal and paternal lineages at the Avy and AxinFu murine alleles , formed by the insertion of IAP elements into or near to endogenous genes [18]–[19] . Furthermore , maternal gestational diet affects methylation at these loci in both the offspring and grand-offspring [20] . It is hypothesised that endogenous loci which have an inherent epigenetic vulnerability to environmental conditions may behave similarly to Avy and AxinFu and may play an important role in the developmental origins of health and disease . Imprinted genes , which are functionally mono-allelic in a parental-origin specific manner and subject to multiple layers of epigenetic control of expression , have been hypothesised to be particularly vulnerable to environmental perturbation [21] . Imprinted genes have been shown to regulate the development of key metabolic organs and have therefore been proposed as good candidates to play a role in the developmental origins of health and disease ( reviewed by [22] ) . Furthermore , as germ cell epigenetic reprogramming of imprinting control elements occurs at least partially in utero , it has been postulated that deregulation of this process may be involved in phenotypic inheritance by the next generation . However , it has also been hypothesised that the converse may instead be true: given the dependence of imprinted gene dosage on multiple layers of epigenetic regulation , imprinted gene expression may be more tightly safeguarded in the face of environmental perturbations during development and any mechanism inducing the action of the canonically repressed allele would be highly regulated [23] . Proper investigation of these hypotheses requires the analysis of how the expression of imprinted genes , as a class , responds to environmental challenge relative to the whole transcriptome and compared to other functionally related gene sets . Our aim therefore was to investigate the role of imprinted gene expression in an established murine model of developmental programming . Specifically we aimed to assess imprinted gene expression in the context of the transcriptome to test whether imprinted genes , as a class , are more or less susceptible than bi-allelically expressed genes to perturbation in expression resulting from gestational undernutrition . We have previously reported that the F1 offspring of dams subjected to 50% caloric restriction during the last week of gestation have a phenotype of low birth weight associated with early-life adiposity , altered pancreatic function and progressive glucose intolerance [24] . In this model , both paternal and maternal inheritance of glucose intolerance to the F2 generation is observed in the absence of any further environmental perturbation [15] . Candidate-based qPCR and microarrays were employed to assess the contribution of genomic imprinting to the developmental origins of health and disease in the F1 and F2 generations of this model .
Since expression of most imprinted genes diminishes towards term and during early postnatal life ( our observations , [25] ) , expression was assessed at E16 . 5 , see Figure 1A . Transcriptome analysis of E16 . 5 liver of control ( C ) and in utero undernourished ( UN ) F1 animals demonstrated 765 genes with significantly perturbed expression in UN liver ( false discovery rate , FDR , q<0 . 05 following Benjamini-Hochberg correction for multiple testing ) . Of the affected genes , 383 were up-regulated and 382 downregulated . Power to resolve a 1 . 5 fold change in expression was estimated to be 99% . In E16 . 5 placenta , 304 genes were significantly affected in UN conceptuses ( FDR q<0 . 05 ) . Of the affected genes , 170 were up-regulated and 134 downregulated . Power to resolve a 1 . 5 fold change in expression was estimated to be 75% . Over 70% of all imprinted genes are represented on these arrays . Of the imprinted genes on the array , the majority were found to be expressed in F1 E16 . 5 placenta and liver ( 78% and 54% respectively ) . A single imprinted gene , Grb10 , was identified as being significantly affected in E16 . 5 F1 UN liver . Similarly , IMPACT was the only imprinted gene affected in E16 . 5 UN placenta . Although power was estimated to be relatively high , imprinted genes are rare and it is conceivable that modest changes in expression of multiple imprinted genes may play a significant role in F1 phenotype but would fall beneath the multiple-testing correction threshold for significantly altered gene expression . Gene Set Enrichment Analysis , GSEA , can be employed to investigate the expression of a priori defined gene sets [26] . GSEA did not identify significant enrichment of the imprinted gene set in either the hepatic or placental transcriptional profile of C or UN samples ( normalised enrichment score ( NES ) 1 . 31 and FDR 0 . 81 in UN liver; NES 0 . 84 and FDR 0 . 88 in UN placenta ) . In agreement with this , there was no difference in the ROC area under the curve between imprinted and randomly permuted gene sets for F1 E16 . 5 liver and array data ( Figure S1A ) . This suggests that imprinted genes as a class are not particularly susceptible to expression perturbation following in utero undernutrition . This analysis could be confounded if the foetal transcriptional response to starvation was not fully developed by E16 . 5 , four days into maternal undernutrition , or if cellular heterogeneity and/or inter-individual variation meant that the experimental design did not have sufficient power to detect biologically relevant changes in gene expression . The adult transcriptional response to starvation promotes a switch from glucose to fatty acid and ketone body utilisation . Gene ontology analysis of the F1 hepatic transcriptome data using both DAVID [27]–[28] and Ingenuity Pathway Analysis tools demonstrated enrichment of terms relating to metabolic function among the genes up-regulated in undernourished foetal liver; these included “PPAR signalling pathway” , “Lipid metabolism” , “Fatty acid metabolism” , “Lipid transport” , “The role of Nuclear Receptors in lipid metabolism” ( Figure 1B , 1C ) . Gene set enrichment analysis corroborated these data: gene sets associated with regulation of the key metabolic transcription factors FXR and LXR , “Lipid catabolic process” , “Fatty acid oxidation” , “Lipid transport” , “Starch and sucrose metabolism” and “Oxidoreductase activity” were enriched in undernourished F1 liver ( Table 1 ) . This suggests that the hepatic metabolic transcriptional response to maternal nutritional deprivation is well developed at E16 . 5 , and validates E16 . 5 as an appropriate time point to assess the role of imprinted genes in this process . These data also demonstrate that biologically relevant changes in gene expression patterns are successfully distinguished from the background “noise” , and that the foetal hepatic response to starvation involves an upregulation of lipid catabolism , similar to that of the adult . To compare the response to in utero undernourishment of imprinted genes versus that of other , functionally related gene sets , all genes on the array were ordered according to their FDR q-value of expression change . The distribution of imprinted genes in this FDR ordered list was plotted and compared to a randomly selected group of genes; a group of housekeeping genes expected to be protected from expression change , and a positive control of genes expected to have altered expression in undernourished tissues . As a positive control of a group of genes expected to have altered expression in F1 undernourished foetal liver , a gene set was curated from the literature documenting the adult hepatic transcriptional response to starvation [29]–[32] . As no comparable studies have been done in placenta , a group of genes identified as enriched by gene ontology analysis using the IPA tool , genes involved in the post-transcriptional modification of RNA , were used as a positive control . If imprinted genes , as a class , are more susceptible to in utero undernourishment , they would be expected to resemble those genes involved in the starvation response . Conversely , if imprinted genes , as a class , are protected from environment-induced transcriptional perturbation , they would be expected to resemble housekeeping genes . However , imprinted genes most closely resemble the randomly selected gene group in both liver and placenta ( Figure 1C , 1D ) . This suggests that imprinted genes are neither more susceptible to , nor protected from transcriptional perturbation induced by in utero nutrient restriction in the F1 generation . In order to expand the number of individuals assessed and the number of tissues interrogated , expression of a group of candidate imprinted genes was measured by quantitative PCR in four tissues: liver , muscle , placenta and brain at E16 . 5 , according to tissue-specific imprinting patterns ( Figure 2B–2G ) . Altered dosage of these imprinted genes has previously been shown to perturb foetal and placental growth and affect the development of metabolic organs and axes , with long-term implications for adult metabolism ( reviewed in [22] ) . These tissues were chosen as their development has been shown to be sensitive to early life conditions , susceptible to imprinted gene dosage , and to be critical to metabolic health . In brain , muscle and placenta , imprinted gene expression at E16 . 5 is largely stable , although brain expression of Cdkn1c and Snrpn is significantly but subtly reduced in undernourished animals , and placental expression of Peg3 ( also called Pw1 ) is increased as shown in Figure 2B , 2C , 2G . However , in E16 . 5 liver H19 , Igf2r , Zac1 , Peg3 and Grb10 are significantly up-regulated by maternal undernutrition ( Figure 2D ) . Grb10 is expressed from the maternally-inherited allele in the liver , but from the paternally-inherited allele in the brain , and these two transcript classes utilise alternative transcriptional start sites [33] . The increase in Grb10 expression in F1 UN E16 . 5 liver is of the usually expressed maternal-type isoforms ( Figure 2E , 2F ) . Sexual dimorphism is frequently observed in animal models of developmental programming [16] , [34]–[35] . In this model both sexes are affected , although males more frequently than females [24] . While sex had no impact on expression of the majority of imprinted genes , hepatic Igf2r and Zac1 were significantly upregulated only in undernourished females ( Igf2r UN females average fold change ( FC ) = 1 . 82 , unpaired two-tailed t-test P = 0 . 0008 ( 95% CI 1 . 35–2 . 29 ) , UN males FC = 1 . 13 ( 95% CI 0 . 98–1 . 28 ) ; Zac1: UN females FC = 1 . 88 , P = 0 . 001 ( 95% CI 1 . 36–2 . 39 ) , UN males FC = 1 . 15 ( 95% CI 0 . 85–1 . 45 ) ) . In the placenta , the undernutrition induced upregulation of Peg3 was male specific ( UN males FC = 1 . 50 , P = 0 . 02 ( 95% CI 1 . 13–1 . 86 ) , UN females FC = 1 . 04 ( 95% CI 0 . 85–1 . 22 ) ) . Hepatic upregulation of Peg3 occurred in both sexes , demonstrating that sex-dependent effects are tissue-specific . While sex-specific effects may increase the level of background expression variation , our data suggest we have sufficient power to detect them , even when analysing both sexes together . Altered imprinted gene expression can be brought about by canonical transcription-factor mediated mechanisms , or through the loss or relaxation of imprinting . Loss of imprinting results in either the silencing of the normally expressed allele or the activation of the normally silenced allele and is associated with alterations in the epigenetic marks which control allele-specific expression . Increased Peg3 expression was observed in liver and placenta , but not brain or muscle . To assess whether this was the result of tissue-specific changes in imprinting control , we quantified the level of methylation at the Peg3 promoter , a maternally methylated germline differentially methylated region , DMR , by pyrosequencing in brain and liver . Peg3 DMR methylation was at the expected level of ∼50% in control individuals ( Figure 2H , 2I ) and was unchanged in both brain and liver in UN individuals . As we are assaying a heterogenous cell population , it is theoretically possible that a very small subset of cells may have more substantially perturbed methylation . We also cannot assess the relative distribution of methylation between the maternally and paternally inherited alleles , and pyrosequencing cannot readily distinguish between cytosine methylation and hydroxymethylation . However , overall these data suggest that the modulation in Peg3 expression is likely to be through the transcription-factor mediated upregulation of the canonical paternally expressed allele , and not due to a relaxation of imprinting . Peg3 has been implicated in the central regulation of energy balance , and is known to affect maternal gestational nutrient partitioning , in addition to playing a role in reproductive and nurturing behaviour [36]–[37] . PEG3 is highly expressed in the foetal and the adult hypothalamus [36] , [38]–[39] , and alterations in central PEG3 dosage in the F1 generation may have implications for the phenotypic development of the F2 generation . Although no change in brain Peg3 expression was observed at the mRNA level , a clear induction of PEG3 was observed by Western blot in the brains of UN animals ( Figure 2J , 2K ) . The metabolic phenotype observed in the F1 generation is transmitted to the F2 generation in the absence of further dietary compromise [15] , [40] . The reliance of imprinting control on parental-origin specific differential epigenetic marks necessitates the erasure and sex-specific reapplication of these marks during germ cell development ( reviewed in [41] ) . The de novo methylation of imprinting control regions occurs asynchronously in the male and female gametes [42]–[45] . Consequently , maternal nutritional restriction during the third gestational week coincides with the re-acquisition of methylation at imprinting control regions in the primordial germ cells of the male but not the female F1 embryo [42]–[45] . Recent data have suggested that the maternally-methylated primary DMRs are distinguishable as unmethylated islands in sperm , raising the possibility that they are protected from de novo methylation in the paternal gametes [46] . Therefore , we hypothesised that if imprinted genes are susceptible to methylation change in primordial germ cells due to in utero nutritional restriction , this will be most evident in the offspring of F1 generation males . To test this hypothesis we assessed placental and hepatic imprinted gene expression at E16 . 5 in the F2 generation and directly analysed sperm methylation in F1 males . To assess the impact of paternal in utero undernutrition on the expression of imprinted genes in the context of the whole transcriptome , expression was assessed by microarray in the E16 . 5 liver and placenta of F2 foetuses with a control dam and an in utero undernourished sire ( CU ) , and compared to foetuses whose parents had never experienced in utero undernourishment ( CC ) ( Figure 3A ) . In E16 . 5 F2 liver , 1330 genes demonstrated significantly different expression levels between CC and CU animals ( FDR q<0 . 05 following BH correction for multiple testing ) . Of the affected genes , nearly three quarters ( 72% ) were downregulated . Power to resolve a 1 . 5 fold change was estimated to be 87% . In E16 . 5 F2 placenta , only 4 genes demonstrated significantly different expression levels between controls and CU placentas ( FDR q<0 . 05 ) . However , power to resolve a 1 . 5 fold change was estimated to be 64% . The placenta has a greater variety of cell types than the liver and is morphologically plastic in response to foetal and maternal cues [47] . This may have resulted in a greater inter-individual variability of gene expression which may account for the reduced statistical power . Over 78% of all imprinted genes are assayed by these arrays . Of the imprinted genes on the array , the majority were found to be expressed in F2 E16 . 5 placenta and liver ( 94% and 86% respectively ) . Gene ontology analysis did not detect genomic imprinting as significantly enriched in either liver or placenta ( genes with altered expression >1 . 5 fold were used for GO analysis in the F2 placental data set ) . ROC curve analysis in F2 CU liver suggested that imprinted genes are moderately more likely to have a lower FDR than non-imprinted gene sets ( area under the curve was higher in the imprinted gene set compared to 99/100 randomly permuted gene sets ) , see Figure S1A . However , GSEA did not detect any statistically significant enrichment of imprinted genes among the altered expression profile in either E16 . 5 liver or placenta ( NES 1 . 28 , FDR 1 . 00; NES 1 . 00 , FDR 0 . 84 for enrichment in CU group in liver and placenta , respectively ) . This suggests that imprinted genes , as a class , are not particularly vulnerable to expression perturbation in the offspring of in utero undernourished males , and suggests that the re-programming of imprinting control elements in F1 primordial germ cells is unaffected by caloric restriction . As described above for the F1 generation , we compared the distribution of imprinted genes in a FDR ordered list to a randomly selected group of genes , housekeeping genes hypothesised to be protected from changes in expression and gene sets found to be enriched in F2 CU liver and placenta by gene ontology analysis ( IPA ) . The distribution of imprinted genes in both the F2 E16 . 5 hepatic and placental transcriptomes most closely resembles that of the randomly selected gene group ( Figure 3B , 3C ) . These data indicate that , as with the F1 generation , imprinted genes are neither more susceptible , nor protected from changes in gene expression in E16 . 5 liver or placenta . As a preliminary assessment of the role of imprinted genes in all the F2 crosses , expression of a group of candidate metabolically important imprinted genes was measured by quantitative PCR in liver and placenta at E16 . 5 , see Figure 4B . While there is some variability in the hepatic expression of certain imprinted genes in the F2 generation at E16 . 5 , particularly Igf2r , Grb10 , Zac1 and Cdkn1c , differences between groups do not reach statistical significance , see Figure 4B ( one-way ANOVA , Bonferroni's multiple comparison test ) . Thus , we can conclude that imprinted gene expression is not significantly affected in the F2 E16 . 5 liver . Despite the increased cellular heterogeneity of this tissue , expression of imprinted genes in F2 E16 . 5 placenta is generally less variable than that of F2 E16 . 5 liver . Expression of Igf2P0 is significantly increased in CU placentas and Snrpn in UU placentas , as shown in Figure 4C ( one-way ANOVA , Bonferroni's multiple comparison test ) . There was no sexual dimorphism in expression changes of imprinted genes in F2 tissues ( data not shown ) . To quantify the variation in expression attributable to the in utero nutrition of each parent , and to ascertain whether there is any interaction between these two variables these data were re-analysed using a two-way ANOVA . These results , presented in Table 2 and Table 3 , demonstrate that parental in utero undernutrition does not have a significant impact on the expression of the majority of imprinted genes at E16 . 5 in liver and placenta . Expression of five out of the fourteen genes tested in the placenta had a significant component of variation attributable to parental in utero nutrition . Igf2P0 was significantly affected by both maternal and paternal nutrition , while for Snrpn and H19 only paternal nutrition and for Dlk1 only maternal nutrition contributed significantly to expression variation , while there was a significant interaction between maternal and paternal nutrition on Phlda2 expression . In F2 E16 . 5 liver , expression of three out of the twelve genes tested demonstrated a significant component of variation attributable to parental nutrition – paternal nutrition on Cdkn1c and Phlda2 expression and maternal nutrition on Snrpn expression . Across both liver and placenta there was no discernable relationship between which parent's nutrition had a significant effect and whether the gene was expressed from the paternally inherited or maternally inherited allele . It is conceivable that changes in the epigenetic status of imprinted genes in the F1 sperm may result in changes in expression earlier in gestation that are not detected at E16 . 5 , or be erased during pre-implantation methylation re-programming . Consequently , the methylation status of four germline DMRs was quantitatively assessed in F1 sperm by pyrosequencing . No changes in F1 sperm methylation profile were identified in males that had been undernourished in utero . The intergenic germline DMR ( IG-DMR ) of the Dlk1/Dio3 locus and the H19 DMR are paternally methylated ICRs and are hypermethylated in the sperm of both control and in utero undernourished F1 males ( Figure 4D , 4E ) . In contrast , the Peg3 and Snrpn DMRs are normally methylated on the maternally inherited allele in somatic tissues , and are entirely unmethylated in the sperm of both control and in utero undernourished F1 males ( Figure 4F , 4G ) .
In conclusion , we have demonstrated that , at least in this murine model of prenatal undernutrition , the functional mono-allelicism of imprinted genes and their unique mechanisms of epigenetic control of expression do not render them either more or less susceptible to expression perturbation following environmental challenge . Nor is there any evidence that germline reprogramming of ICRs is susceptible to nutritional restriction . However , we propose that the selective dosage modulation of certain imprinted genes plays an important role in the adaptive foetal response to in utero nutritional scarcity .
ICR mice , an outbred strain , were obtained from the Jackson Laboratory . Mice were housed in an OLAW-approved facility , with controlled temperature , humidity , and light-dark cycle ( 07:00–19:00 ) . Protocols were approved by the Joslin Diabetes Centre Institutional Animal Care and Use Committee . “Principles of Laboratory Animal Care” ( http://grants1 . nih . gov/grants/olaw/references/phspol . htm ) were followed . F1 generation ( as described by [24] ) : Virgin female ICR mice ( age 6–8 weeks ) were caged with ICR male mice . Pregnancy was dated with vaginal plugs ( day 0 . 5 ) , and pregnant female mice were housed individually with ad libitum access to Purina 9F ( 9% fat ) chow . On pregnancy day 12 . 5 , female mice were randomly assigned to either control or undernutrition groups; weight did not differ between control and undernutrition mothers prior to pregnancy or at day 12 . 5 . Food intake of undernutrition mothers was restricted to 50% that of controls , calculated each day , from days 12 . 5 to 18 . 5 . After delivery , litter size in both groups was equalized to eight pups per dam by removing both the heaviest and lightest mice in the litter , thus retaining those with birth weight closest to the median . Mothers received chow ad libitum after delivery . Pups nursed freely and were weaned at 3 weeks onto 9F chow ad libitum . F2 generation ( as described by [15] ) : control and undernourished females from the F1 generation were mated at age 2 months with nonsibling control or undernourished males . After confirmation of pregnancy , females were caged individually and fed ad libitum with no dietary manipulation during pregnancy to produce a second , F2 , generation with four experimental groups: CC – both parents are controls . CU – control dam , in utero undernourished sire; UC - in utero undernourished dam , control sire; UU - in utero undernourished dam , in utero undernourished sire . In this study , gene expression was assessed in both the F1 and F2 generations at E16 . 5 . Mice were anaesthetised with intraperitoneal pentobarbital following an overnight fast and tissues were rapidly dissected and snap frozen in liquid nitrogen . To facilitate multiple extractions of DNA , RNA and protein from the same tissue , samples were pulverised in liquid nitrogen and were never allowed to thaw . All kits were utilised according to the manufacturers' instructions , except as noted below: RNA was extracted using Trizol ( Invitrogen ) , with an overnight precipitation step at −20°C . Newly isolated RNA was quantified by spectrophotometric analysis and the quality was assessed by judging the integrity of the 28S and 18S ribosomal RNA bands by electrophoresis through a 1% agarose gel . All samples were treated to remove DNA contamination with DNase ( using the RNase-free DNase kit , Qiagen ) , followed by re-precipitation . cDNA was generated from 1 µg total RNA per sample using the RevertAid H Minus cDNA synthesis kit ( Fermentas ) with random primers . cDNA samples were diluted 1∶20 and a six-point standard curve of two-fold dilutions was prepared from pooled cDNA . The samples and standard curve were aliquoted and stored at −80°C prior to use . Real-time quantitative PCR with SYBR Green was performed with SensiMix ( Quantace ) using primers in Table S1 [33] , [64] . Primers were designed to assay all annotated splice-variants of a gene where possible and were checked for specificity using NCBI nucleotide BLAST and gel electrophoresis of the PCR product . A standard curve made up of doubling dilutions of pooled cDNA from the samples being assessed was run on each plate , and quantification was performed relative to the standard curve . Target gene expression was normalised to the expression of HPRT , the expression of which did not differ between the groups . All primers amplified with an estimated efficiency of between 110% and 80% and there was no evidence of inhibitors present in the reaction . Reactions were carried out on a DNA engine Opticon 2 thermocycler ( MJ Research ) . Liver gene expression was analysed in F1 E16 . 5 mice using Illumina Mouse WG6v2 microarrays . Samples were prepared in three pools per condition , representing a total of fifteen individuals from five independent litters per condition with the sex ratio controlled between conditions . For F1 placental arrays RNA from seven control and eight undernourished samples from independent litters was hybridised to Affymetrix MOE430A arrays . For F2 liver and placental arrays , cRNA from fifteen control conceptuses from three independent litters , and eighteen CU conceptuses from three independent litters , was hybridised to Affymetrix MOE430-2 arrays in five and six pools per condition , respectively . Analysis of microarray data was carried out inside the R-statistical environment ( http://www . r-project . org ) . Illumina arrays were analysed using the Lumi and Limma Bioconductor packages ( www . bioconductor . org ) . Probes not expressed on any arrays were removed from the analysis . Variance-stabilising transformation ( Lin et al . , 2008 ) and loess normalisation was employed . Affymetrix arrays were analysed using the Affy and Limma Bioconductor packages . RMA transformation and normalisation was carried out . For all arrays probes not expressed on any arrays were removed from the analysis , the quality of normalisation was assessed by density plots of intensity and box plots of amplitude . The Benjamini-Hochberg multiple testing correction was applied to control for false discovery rate , FDR , following pairwise comparison . Genes with an FDR q-value of <0 . 05 were considered to be significantly differentially expressed . Power was estimated using the Bioconductor Sizepower package . The distribution of differential expression in terms of FDR for the four arrays is presented in Figure S1B . Gene ontology analysis was carried out using DAVID: Database for Analysis , Visualisation and Integrated Discovery [27]–[28] and Ingenuity Pathway Analysis , IPA , software ( Ingenuity Systems www . ingenuity . com ) . ROC curves were computed for the imprinted gene set for each array ( Figure S1A ) . In order to provide a context for the 4 experimentally-determined ( imprinted ) ROCs , we also generated ROCs from 100 randomized data sets . The randomized data sets were generated by permuting the gene labels with regard to which genes are imprinted and which are not imprinted . We calculated the area under each of these curves ( AUC ) and compared each of the 4 experimental data sets to the 100 randomized data sets in terms of this area . As a negative control for genes that are expected to be protected from expression change , a list of non-metabolic putative housekeeping genes was curated from the literature [65] . Imprinted genes were limited to those that have been experimentally validated ( Figure S2 ) . F1 generation: As a positive control of a group of genes expected to have altered expression in undernourished foetal liver , a gene set was curated from the literature documenting the adult hepatic transcriptional response to starvation [29]–[32] . As no comparable studies have been done in placenta , a group of genes identified as enriched by gene ontology analysis using the IPA tool , genes involved in the post-transcriptional modification of RNA , were used as a positive control . F2 generation: As no comparable studies exist in the literature , gene groups involved in lipid metabolism which were identified as enriched in CU liver and placenta by gene ontology analysis ( Ingenuity ) were used as a positive control . Western blot was carried out essentially as previously described [66] . Briefly , pulverised snap frozen samples were homogenised on ice in RIPA buffer supplemented with Complete protease inhibitors ( Boeringher ) , activated sodium orthovanadate and PMSF . 80 µg of protein was run on a 5–15% SDS-PAGE gradient gel at 50 V overnight at 4°C and transferred to PVDF membranes by electroblotting overnight at 20 V . Following blocking membranes were incubated with primary polyclonal antibody ( rabbit ) α-Peg3 at 1/24 , 000 [67] , followed by a 1/5000 poly-clonal anti-rabbit HRP-conjugated secondary antibody ( Dako Denmark ) and developed using an ECL plus Western Blotting Detection System ( Amersham , GE Healthcare ) . gDNA isolation was carried out by standard organic extraction including an overnight proteinase K treatment at 55°C . DNA was quantified spectrophotometrically and quality was assessed by running 100–500 ng on a 1 . 5% agarose gel . The sex of E16 . 5 embryos was identified by PCR for the sex-chromosome specific genes Zfy and SMCX/Y using the primers in Table S3 . Two month old male mice were sacrificed and sperm collected from the cauda epididymes and vas deferens as described previously [68] , [69] . Extruded sperm and sliced epididymes were suspended in 50 ml of Solution A ( 0 . 75 mL 5 M NaCl pH 8; 2 . 5 ml 0 . 5 M EDTA; H2O to 50 ml ) and rocked on a platform for 10 min to release sperm . Non-sperm tissue was removed by 10 minutes of settling , followed by serial centrifugation at 500× g for 15 min , and 700× g for 10 min . Sperm was harvested by centrifugation at 1100× g for 5 min . 200 µl Solution B ( 0 . 1 mL Tris-HCl pH 8; 0 . 2 ml 0 . 5 M EDTA; 2 ml 10% SDS; 8 ml 100 mM DTT; H2O to 10 ml ) was added , followed by a standard RNAseA and overnight proteinase K treatment at 55°C . DNA was extracted using DNEasy columns . To accurately quantify sperm DNA concentration Quant-iT PicoGreen was used . Sodium bisulphite mutagenesis was carried out on 1 ug of gDNA per sample using the 2-step conversion protocol of the Sigma Imprint DNA Modification Kit . Two samples with no template were run in parallel to confirm contamination had not occurred during bisulphite treatment . Quantification of methylation following bisulphite conversion was carried out by pyrosequencing as previously described [70] . Pyrosequencing primer design was carried out using Qiagen PyroMark Assay Design software 2 . 0 , sequences are given in Table S2 . Following PCR , 2 . 5 µl of each PCR product was run on a gel to ensure specificity of amplification and suitable concentration of product . The biotinylated strand was purified using strepdavidin sepharose high performance beads ( GE Healthcare ) and PyroMark reagents ( Qiagen ) . Pyrosequencing was carried out on a PyroMark MD pyrosequencer ( Biotage ) using PyroMark Gold Qp6 SQA reagents ( Roche ) and quantification of methylated and unmethylated alleles carried out using Pyro Q-CpG 1 . 0 . 9 software ( Biotage ) . | Environmental perturbations during early life are known to affect one's risk of metabolic disease many years later . Furthermore , that risk can be inherited by future generations , although the mechanisms responsible are poorly understood . Imprinted genes are unusual as only one of the two copies is expressed in a parent-of-origin–specific manner . As only one copy is active , imprinted gene dosage has been hypothesised to be uniquely vulnerable to environmental change . Therefore , it has been suggested that imprinted genes may play an important role in the developmental origins of health and disease . Alternatively , the opposite may be true—imprinted genes may be more tightly safeguarded from perturbation . To test these two hypotheses , we analysed the expression of imprinted genes in the context of all active genes in two affected generations of a mouse model of the developmental origins of health and disease . Our data show that imprinted genes as a class are neither more nor less susceptible to expression change , but a subset of imprinted genes may be involved in the adaptation of the conceptus . Furthermore , imprints in the developing germline are not affected and imprinted genes are largely stable in the second generation . This is important , as it is the first time that this hypothesis has been tested in an unbiased fashion . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"developmental",
"biology",
"genomics",
"physiology",
"genetics",
"biology",
"anatomy",
"and",
"physiology",
"genetics",
"and",
"genomics"
] | 2012 | An Unbiased Assessment of the Role of Imprinted Genes in an Intergenerational Model of Developmental Programming |
The transition from hunting and gathering to farming involved a major cultural innovation that has spread rapidly over most of the globe in the last ten millennia . In sub-Saharan Africa , hunter–gatherers have begun to shift toward an agriculture-based lifestyle over the last 5 , 000 years . Only a few populations still base their mode of subsistence on hunting and gathering . The Pygmies are considered to be the largest group of mobile hunter–gatherers of Africa . They dwell in equatorial rainforests and are characterized by their short mean stature . However , little is known about the chronology of the demographic events—size changes , population splits , and gene flow—ultimately giving rise to contemporary Pygmy ( Western and Eastern ) groups and neighboring agricultural populations . We studied the branching history of Pygmy hunter–gatherers and agricultural populations from Africa and estimated separation times and gene flow between these populations . We resequenced 24 independent noncoding regions across the genome , corresponding to a total of ∼33 kb per individual , in 236 samples from seven Pygmy and five agricultural populations dispersed over the African continent . We used simulation-based inference to identify the historical model best fitting our data . The model identified included the early divergence of the ancestors of Pygmy hunter–gatherers and farming populations ∼60 , 000 years ago , followed by a split of the Pygmies' ancestors into the Western and Eastern Pygmy groups ∼20 , 000 years ago . Our findings increase knowledge of the history of the peopling of the African continent in a region lacking archaeological data . An appreciation of the demographic and adaptive history of African populations with different modes of subsistence should improve our understanding of the influence of human lifestyles on genome diversity .
There is archaeological and genetic evidence to suggest that anatomically modern humans originated in a small , isolated population in Africa 150–200 thousand years ago ( Kya ) . Worldwide population radiation then occurred 50–75 Kya [1]–[17] . However , the history of sub-Saharan African populations , which display considerable cultural , linguistic , phenotypic and genetic diversity , remains less clear [18] , [19] . Studies based on multidisciplinary approaches generally indicate that sub-Saharan Africa was re-peopled recently , during the so-called Bantu expansions , extending outwards from a Nigeria-Cameroon homeland and beginning 3–5 Kya . These expansions were accompanied by the spread of Bantu languages , agricultural practices and sedentism , and probably also by iron working [20]–[23] . Most sub-Saharan African populations have now integrated these sociocultural practices , speaking one of the 450 Bantu languages [24] and presenting principally an agriculture-based sedentary lifestyle . However , a few populations did not adopt the lifestyle associated with Bantu expansions and continue to live as mobile groups , with a mode of subsistence based essentially on hunting and gathering . Today , these groups include the Western ( e . g . , Aka , Baka , Bakola ) and Eastern ( e . g . , Efe , Asua , Sua ) Pygmies , the Khoi , the San , the Okiek and the Hadza [25] . The Pygmy populations occupy a vast territory extending west-to-east along the central African belt from the Congo Basin to Lake Victoria . They have a mostly forest-dwelling hunter-gathering lifestyle , specific cultural practices ( honey gathering tools , etc . [26] ) and distinctive physical traits ( e . g . , lowest mean stature of all human populations [27] , [28] ) . Pygmy groups traditionally live in huts , moving regularly from one camp to another , although some groups remain sedentary for some time due to socioeconomic dependence on neighboring farmers . Most Pygmy populations now speak the language of neighboring farming populations , suggesting extensive cultural — and possibly genetic — exchanges between the two groups [26] , [27] , [29]–[34] . Two main groups of Pygmy populations , each including different ethnic groups , are currently recognized: the “Western Pygmies” inhabiting the western part of the Central African rainforest corresponding broadly to the Congo Basin , and the “Eastern Pygmies” living in the easternmost part of the Central African belt close to the Ituri rainforest and Lake Victoria . Despite the extensive similarity in their modes of subsistence , cultural practices and distinctive phenotypic traits , Western and Eastern Pygmies clearly display both linguistic and genetic ( at least for mtDNA and some protein markers ) differentiation: the resemblance between each of the two Pygmy population groups and local farming populations is greater than that between the two Pygmy groups [8] , [27] , [35] , [36] . Despite the large body of ethnological and linguistic data collected for these populations , little is known about the prehistory , population dynamics and past interactions between African farmers and Pygmy hunter-gatherers . Indeed , our understanding of the past peopling of Central Africa is limited by the virtual absence of human remains in its acidic soils [21] . In addition , the differences in the mode of subsistence of these two groups and the complex interactions between them raise several questions: which historical and demographic events led to the divergence between the ancestors of present-day farmers and Pygmies ? Have the recent Bantu expansions associated with the spread of farming been responsible for the divergence of these two groups of populations ? Or , were these populations already genetically — and possibly ecologically — differentiated before the agricultural revolution in Africa ? How has the size of the populations of these two groups changed since they started to diverge ? Furthermore , how did Western and Eastern Pygmy populations , which today show geographic separation , linguistic differentiation and distinctive genetic features , acquire their shared specific cultural and phenotypic traits ? Did these two groups initially have a common ancestry but subsequently split apart , or do they reflect convergent cultural and genetic adaptation to the rainforest ? We addressed these questions by first considering the demographic characteristics of the agricultural , Western Pygmy , Eastern Pygmy population groups ( i ) to determine how these three population groups separated over time ( i . e . , branching order of the phylogenetic tree ) and ( ii ) to estimate the time at which these population groups separated and the levels of subsequent gene flow between them . We generated a large multilocus resequencing dataset for five agricultural and seven Pygmy populations dispersed over the African continent . We then compared the ∼7 . 8 Mb of diploid sequences obtained with a large number of simulations exploring various demographic and branching scenarios , to identify the models best fitting the observed data . We then estimated , with the approximate Bayesian computation ( ABC ) method [37] , population separation times and levels of gene flow between these populations under an isolation-with-migration ( IM ) framework — a realistic model assuming that populations diverge and subsequently experience gene flow . The model best fitting our data involves early divergence of the ancestors of farming populations and Pygmy hunter-gatherers ∼60 , 000 years ago , followed by a split of the Pygmies' ancestors into the Western and Eastern Pygmy groups ∼20 , 000 years ago . This study thus improves our understanding of the ancient history of the ecologically and culturally diverse populations of sub-Saharan Africa .
We first investigated whether our sampled populations constituted different genetic entities , by clustering individuals as a function of their genotypes for all autosomal and X-linked regions , using the STRUCTURE program [40] . When we specified that the data corresponded to only two groups ( K = 2 ) , Pygmy groups and AGR populations were separated into two different clusters ( Figure 2A ) . This suggests that the genetic structure of African agricultural and Pygmy populations is correlated with their modes of subsistence . However , WPYG and EPYG groups further separated into two clusters at K = 3 , revealing a certain degree of genetic differentiation between the two groups of Pygmy populations . The model with four clusters , which is the most probable given the data ( P ( K = 4/data ) = 75 . 8% ) , further partitioned the group of farmers into those inhabiting the Central African belt and those located in South-East Africa . No other cluster was found for K values higher than 4 ( Figure 2A ) . Overall , our results indicated that the three ethnologically recognized population groups — agricultural populations , Western Pygmies and Eastern Pygmies — corresponded broadly to different genetic entities . However , STRUCTURE analysis revealed that some of the populations within each of these three population groups displayed considerable admixture or genetic differentiation ( Figure 2A ) . Regardless of the value of K considered , three populations had large proportions of individuals with multiple memberships: the Bakola Pygmies from Cameroon and the two populations of Twa Pygmies from Rwanda . This observation confirms the admixed status of the Bakola Pygmies inferred from 28 autosomal microsatellites [41] , indicating substantial levels of gene flow from neighboring farmers . With respect to the two populations of Twa Pygmies , they clearly clustered with South-East African farmers for K = 4 , consistent with these Pygmy groups being admixed ( some anthropologists describe them as “Pygmoids” ) , and with the complete shifting of their cultural practices towards those of neighboring agricultural populations [27] . Furthermore , the STRUCTURE analysis for K = 4 separated Mozambicans from the other agricultural populations ( Figure 2A ) . This suggests the presence of fine-scale population structure in the AGR group , despite the very low and non significant levels of differentiation between AGR populations , on the basis of the FST statistics ( Table S2 ) . Admixture or fine-scale population structure within each of our three population groups ( i . e . , AGR , WPYG and EPYG ) may affect historical and demographic inferences [42] . We therefore conducted all subsequent analyses on a pruned population dataset . This “filtered population dataset” excludes individuals with ancestry in other populations , or populations that appear to be differentiated at K = 4 within each population group . The excluded individuals mostly corresponded to Bakola Pygmies , Twa Pygmies and Mozambicans ( Figure 2B , Text S1 ) . Only the results obtained with this filtered population dataset are discussed . However , we explored the effect of this filtering on our inferences , by also carrying out all analyses with the entire population dataset ( the “composite population dataset” ) , which includes the admixed/structured individuals/populations ( Text S1 , Figure S1 , S2 , and S3 , Tables S3 , S4 , S5 , S6 , and S7 ) . As departures from nonequilibrium demography ( e . g . , population growth or shrinkage ) have been shown to influence the estimation of population separation times and levels of gene flow [43] , [44] , we first assessed the demographic history of each population group ( AGR , WPYG and EPYG ) : we determined the simplest demographic model best fitting the observed within-population variation data for each population group , using a number of diversity indices and neutrality statistics summarizing the data ( Table 1 ) . The patterns of variation observed within the AGR group were characterized chiefly by an excess of low-frequency variants ( Figure 3 ) , as attested by the significant negative values obtained for some neutrality tests for autosomes and mtDNA ( Table 1 ) . The variance of the Tajima's D statistic was also significantly lower across autosomal regions in the AGR group ( Table S4 ) , this pattern being a specific signature of population growth [45] . These observations suggest the occurrence of population growth among the ancestors of present-day farmers . As all the farming populations studied here speak Benue-Congo languages ( including Bantu languages ) , the signatures of population expansion and the low levels of differentiation ( Table S2 ) observed among AGR populations may result from Bantu expansions spreading the farming lifestyle throughout sub-Saharan Africa over the last ∼5 Kya [20]–[23] . None of the classical neutrality tests used detected significant departures from the constant-sized population model for the WPYG and EPYG groups ( Table 1 , Figure 3 ) . However , the occurrence of gene flow between populations with different demographic regimes may dilute the signals of departure from nonequilibrium demography detected by neutrality tests ( e . g . , the signature of a bottleneck among Pygmies is erased by gene flow with the expanding AGR populations , introducing low-frequency variants into the Pygmy gene pool ) . We identified the demographic model best fitting the Pygmy data by comparing the within-population summary statistics of WPYG and EPYG ( Table 1 ) with simulated summary statistics under constant-population size and bottleneck models , in the presence of various levels of gene flow with an expanding AGR population ( Figure 4 , Table S8 , Materials and Methods for details ) . The genetic diversity of both WPYG and EPYG fitted significantly better with models assuming a bottleneck in the Pygmy population accompanied by high levels of gene flow with the AGR population than with a model of a constant-sized Pygmy population with negligible gene flow with the AGR population . A bottleneck beginning 2 , 500–25 , 000 years ago with an 80% decrease in population size , followed by a recovery starting 125 years later with a size increase of between 100% and 400% ( Figure 4 ) , fitted the WPYG data significantly better than the constant-sized population model ( P = 0 . 04 , see Materials and Methods ) . For the EPYG group , a bottleneck starting 250–2 , 500 years ago with a 90 to 95% decrease in population size ( Figure 4 ) fitted the observed genetic diversity significantly better than the constant-sized population model ( P<0 . 01 ) . Population structure models could also theoretically fit the PYG data , in the presence of gene flow with AGR populations . However , the occurrence of population structure in PYG populations alone is unlikely because ( i ) our analyses considered a pruned population dataset excluding admixed populations ( Figure 2B ) and ( ii ) the influence of population structure within WPYG populations is probably negligible because within-population neutrality statistics for each WPYG population individually were always positive ( Text S1 ) . Altogether , our adjustment for the demographic regime of each population group revealed the occurrence of population growth in AGR populations and bottlenecks in both the WPYG and EPYG groups . The sequence of the divergence events underlying the current differentiation of Western Pygmy , Eastern Pygmy and agricultural groups remains unclear . All Pygmy groups share idiosyncratic cultural and phenotypic traits , but substantial linguistic and genetic differentiation between Pygmy groups is also observed [8] , [27] , [35] , [36] , [46] . These observations call into question the postulated common origin of African “Pygmy” populations . Indeed , if Western and Eastern Pygmy groups share a more recent ancestry with their respective agricultural neighbors than with each other , then they may have acquired their shared specific traits by convergence rather than by shared ancestry . Various models can be put forward to explain the current levels of differentiation between these three different groups: ( i ) the A-WE model , involving an ancient divergence between the ancestors of the AGR and PYG groups , followed by a split of PYG ancestors into the WPYG and EPYG groups; ( ii ) the W-AE model , in which the most ancient split is that between the ancestors of the WPYG and AGR groups; ( iii ) the E-AW model , in which the most ancient divergence is that between the ancestors of the EPYG and AGR groups , and ( iv ) the AWE model , in which all populations diverged simultaneously ( Figure 5 ) . To discriminate between these four models , we calculated several between-population summary statistics for all pairs of populations , including FST , the proportion of shared mutations , the proportion of low-frequency shared mutations , and the mean frequency of shared mutations ( Table S5 , Materials and Methods ) Twenty autosomal regions were simulated 1 , 000 , 000 times under the four possible IM models ( Figure 5 ) with IM parameters ( times of divergence , migration rates ) drawn from large , flat prior distributions ( Table S9 ) . As the specific demographic history of each population group may influence the inference of the branching history , we incorporated into our simulations ( Table S9 ) the demographic model identified for each population group most compatible with their observed within-population summary statistics ( Table 1 ) . The mean between-population summary statistics across the 20 simulated regions were then compared with the observed statistics for the 20 autosomal regions ( Table S5 , Materials and Methods ) . The proportion of low-frequency shared mutations and the mean frequency of shared mutations were found to be non informative: their mean values were similar across the four IM models simulated ( data not shown ) . By contrast , FST and the proportion of shared mutations varied considerably between IM models . These two statistics were therefore systematically considered in the sets of summary statistics used for the best-fit approach ( Materials and Methods ) . Independently of the set of summary statistics used , the A-WE model always gave the highest proportion of small distances between the simulated and observed datasets ( Ψ0 . 5 ) , and was therefore identified as the most probable model given the data ( Figure 6 ) . We then investigated whether this result was sensitive to ξ — the threshold at which distances between simulated and observed statistics are considered to be “small” ( Materials and Methods ) . We observed a highly significant negative correlation between ξ and the proportion of small distances Ψξ generated by the A-WE model ( r2 = 0 . 969 , P = 0 . 00014 ) : the smaller ξ , the better the simulations fitted the observed data , and the greater the enrichment of the A-WE model in these simulations . This analysis thus clearly supports our conclusion that the A-WE model is the most probable , given the autosomal data . Unlike autosomal , X-linked and Y-linked regions , mtDNA displayed strong differentiation between Western and Eastern Pygmies ( Table S5 ) , an observation at odds with the A-WE model . Several lines of evidence suggest that sex-biased gene flow , ancient maternal population structure and/or stronger genetic drift have contributed to the high levels of mtDNA differentiation observed today between the two Pygmy groups ( Text S1 for details ) . More generally , genetic drift has probably been greater among PYG populations for all genomic compartments , because the PYG Ne is smaller than the AGR Ne , potentially leading to higher levels of differentiation between the two PYG groups than between each PYG group and the AGR group . Indeed , when simulating the 20 autosomal regions under the AWE model , in which the three populations diverge simultaneously , greater mean differentiation was observed between the two PYG groups than between the PYG and AGR populations ( data not shown ) . Consequently , a more recent divergence between the two Pygmy groups ( than between the PYG and AGR groups ) is required , both to compensate for the stronger genetic drift among PYG populations and to generate the observed lower level of differentiation of autosomal regions between the two PYG groups . Taken together , our analyses , which explored a wide range of models and parameter values ( Table S9 ) , clearly support the hypothesis of a recent common origin of Western and Eastern African Pygmies . We then investigated the time scale of the various events characterizing the branching history of AGR , WPYG and EPYG populations , by estimating IM parameters under the validated A-WE model . The coestimation of population separation time and gene flow levels between two populations is generally difficult because low levels of differentiation may result from either a recent splitting of populations with low subsequent gene flow or from an ancient split with high subsequent gene flow [47] . Several methods have been developed for confident estimation of IM parameters , provided that some fixed differences between diverging groups are observed ( i . e . , species or subspecies ) [48]–[50] . These methods are also limited to an IM model with only two populations , or to constant-sized populations . The application of two of these methods to our dataset — IMa and mimar [49] , [50] — provided no evidence of chain convergence despite good mixing of the Markov chains ( Text S1 ) , probably due to the low overall levels of differentiation between the PYG and AGR groups ( i . e . , no fixed differences observed between the two groups ) . We therefore sought to coestimate these parameters under the ABC framework [37] . We obtained non-flat unimodal posterior distributions for all IM parameters ( Figure 6 ) , using different informative summary statistics ( Materials and Methods ) . We assessed the accuracy of these estimations , by estimating IM parameters for randomly chosen simulations as if they were empirical data , but with known actual parameter values . In ∼95% of cases , the known parameter values were within the 95% confidence interval of parameter estimates ( Table 2 ) , indicating that estimated confidence intervals were accurate . Our estimations indicated that the ancestral effective population size of the African groups here studied was 11 , 402 individuals ( 95% CI: 7 , 670–15 , 653 ) ( Table 2 ) . This ancestral population pool started to diverge , eventually generating the current agricultural and Pygmy populations , 56 Kya ( 95% CI: 25 . 8–130 . 5 ) . The subsequent split of the ancestors of Pygmies into the present-day WPYG and EPYG groups was estimated at 21 . 9 Kya ( 95%CI: 14 . 2–66 . 3 ) . Finally , our estimates for the levels of gene flow between WPYG and EPYG , between WPYG and AGR and between EPYG and AGR populations were 4 . 4×10−4 , 1 . 8×10−4 and 2 . 4×10−5 , respectively . As previously mentioned , all analyses ( adjustment of the internal demographic regimes of each population group , the branching model and ABC estimation of IM parameters ) were also performed with the “composite population dataset” , which includes the admixed/structured individuals/populations ( Text S1 , Figures S1 , S2 , and S3 , Tables S3 , S4 , S5 , S6 , and S7 ) . The results obtained for this entire-population dataset were remarkably similar to those obtained with the pruned population dataset: the best-fitting branching model of populations was the same ( i . e . , the A-WE model , Figure S3 ) and the estimates of population separation times were very similar ( Table 2 , Figure S3 ) . However , estimates of gene flow between population groups were consistently lower for the filtered population dataset , which excludes admixed individuals/populations . Thus , the pooling of populations with different proportions of admixed individuals had no effect on the estimation of population separation times . This highlights the reliability of the ABC approach for estimating population divergence by properly adjusting for the different levels of gene flow between populations . The implications of our estimates are important for broader issues in African prehistory , although they must be interpreted carefully because of their large confidence intervals ( Table 2 ) . The finding that the ancestors of AGR and PYG populations diverged ca . 60 Kya is consistent with our recent single-locus estimation based on the mtDNA diversity of African farmers and Western Pygmies [36] . Most of the large waves of population expansion and migration , both within and out of Africa , have been dated at ca . 50–80 Kya , based on several genetic markers [1]–[17] . It has been suggested that these early population movements within and out of Africa may have been triggered by rapid environmental changes . During this period , sub-Saharan Africa witnessed a major episode of climatic change: a sharp oscillation towards a drier climate , with annual rainfall decreasing by up to 50% [51] . These early population expansions may also have been fuelled by increases in the carrying capacity of some human groups associated with radical changes in technology with the emergence of more complex hunting equipment and large-scale movements of high-quality stone and imported shell ornaments [16] . The environmental changes occurring at this time therefore seem to have favored a drastic increase in the complexity of the technological , economic , and social behavior of certain African groups , potentially leading to a major demographic expansion of these groups in competition with other , adjacent groups [16] . In this context , our estimated date of the initial divergence between the ancestors of present-day farmers and Pygmies implies that this period was characterized not only by major human movements , but also by early episodes of population differentiation within the African continent . Our evidence for a separation of the ancestors of Western and Eastern Pygmy groups ca . 20 Kya is also consistent with a previous mtDNA study , dating the time of separation of these two Pygmy groups to at least 18 Kya [52] . These estimates coincide with another period of major climatic change , the Last Glacial Maximum , which led to a massive retreat of tropical forests in Central Africa [53]–[55] . Our genetic results thus support the anthropological hypothesis that the ancestors of present-day forest specialists — Western and Eastern Pygmies — began to diverge at the same time as the rainforest retreated into refugia , ∼20 Kya [26] . However , the split of Pygmy populations into two pockets corresponding to forest refugia did not totally prevent the occurrence of gene flow between Western and Eastern Pygmy groups ( Table 2 ) . Finally , our estimates of gene flow between each group of Pygmies and agricultural populations yielded contrasting values , with levels of gene flow between WPYG and AGR populations three to seven times higher than those between EPYG and AGR populations ( Table 2 ) . This result , together with those obtained with protein markers [27] , mtDNA [8] , [36] and autosomal microsatellites [41] , [46] , indicates that ( i ) substantial gene flow has occurred between Western Pygmies and agricultural populations , possibly during a period before the strong social barriers currently separating these two groups became established [29] , [32] , [33] , [41] , [56] , and ( ii ) the Eastern Mbuti Pygmies ( i . e . , the EPYG group in the filtered population dataset ) have probably been among the most isolated Pygmy populations of sub-Saharan Africa . Our multilocus resequencing analyses , supported by simulation-based inferences , increase our knowledge of the peopling history of the African continent by revealing that: ( i ) Western and Eastern Pygmies share a recent common ancestry , indicating that their shared specific traits , such as hunting and gathering in rainforest ecosystems and short stature , were acquired by shared ancestry rather than by convergence , and ( ii ) the agricultural revolution associated with Bantu expansions is not responsible for the population differentiation currently observed between farmers and Pygmy hunter-gatherers , suggesting that the ancestors of these two populations had a hunting and gathering lifestyle but possibly in different , specific ecological habitats ( e . g . , forest and savanna ) . The distribution of lithic industries in the Middle Stone Age points to the existence of hunter-gathering groups in the open savanna environment of Central Africa [21] . This , together with the observation that Bantu migrations followed savanna passages [21] , supports the notion that the mode of subsistence of the ancestors of farmers was savanna-based hunting and gathering . The null model of selective neutrality provided by this study will also prove useful for the identification of genetic variants contributing to complex diseases and for the detection of genomic regions targeted by natural selection . In particular , a detailed study of the genome-wide footprints of local positive selection in African farmers and Pygmy hunter-gatherers , integrating the demographic model determined in this study , should facilitate robust identification of the population-specific adaptive responses of these two human groups to their different climatic , pathogenic and nutritional environments . These studies should help to decipher the potential genetic basis of the population-specific traits characterizing these ethnic groups , such as the short mean stature of the Pygmies . More generally , an appreciation of the demographic and adaptive history of these populations will improve our understanding of the influence of human lifestyles on genome diversity in terms of both health and disease .
Sequence variation was surveyed in DNA samples from 12 sub-Saharan African populations . The panel included 118 samples from five agricultural populations ( Yoruba from Nigeria [N = 31] , Ngumba from Cameroon [N = 16] , Akele from Gabon [N = 16] , Chagga from Tanzania [N = 32] and Mozambicans [N = 23] ) , 71 samples from four Western Pygmy populations ( Bakola from Cameroon [N = 16] , Baka from Cameroon [N = 15] , Baka from Gabon [N = 16] and Biaka from the Central Africa Republic [N = 24] ) , and 47 samples from three Eastern Pygmy populations ( Mbuti from the Democratic Republic of Congo [N = 24] and Twa from southern [N = 8] and northern [N = 15] Rwanda ) ( Figure 1 ) . The Biaka , Mbuti , Yoruba , and Chagga samples are subsets of samples described in ALFRED ( http://alfred . med . yale . edu/alfred/index . asp ) under sample UID numbers SA000005F , SA000006G , SA001805O , and SA000487T , respectively . All sampled individuals were healthy donors from whom informed consent was obtained . This study was approved by the Institut Pasteur Institutional Review Board ( n° RBM 2008 . 06 ) . The 24 independent regions sequenced here represent a total sequence length of 32 . 75 kb per individual ( mean sequence length per region of 1 . 31 kb ) . We selected 20 non coding , independent autosomal regions ( Table S1 ) to decipher the genetic architecture of AGR and PYG populations . The regions were selected ( i ) to be at least 200 kb away from any known or predicted gene or spliced EST ( distance determined by inspection of the hg18 UCSC genome assembly ) ; ( ii ) not to be in linkage disequilibrium ( LD ) with any known or predicted gene or spliced EST ( as determined by inspection of the LD levels observed in the four HapMap populations , release 16 ) ; ( iii ) not to be in LD with each other and ( iv ) to have a region of homology with the chimpanzee genome ( November 2003 release ) . We also selected two X-linked regions based on the same criteria , together with two linked regions on each arm of the Y chromosome and one mtDNA region selected in a previous study [57] ( Table S1 ) . The two Y-linked regions were considered as a single region in all analyses . All non coding regions were sequenced with two different primers . All sequencing reactions were run on automated capillary sequencers ( ABI3130 and ABI3730 ) . PCR and sequencing primers and protocols are available upon request . Samples from Mozambique and Rwanda underwent whole-genome amplification before PCR amplification and resequencing . Sequence alignment and SNP detection were carried out with Genalys v . 3 . 3b [58] . In addition , all ABI base-called sequences and genotypes were visually inspected by two independent investigators . All singletons were confirmed by reamplification and resequencing . No false singleton was observed . Less than 0 . 1% of genotypes were left as missing data . We reconstructed haplotypes with PHASE v . 2 . 1 [59] , using five independent runs with different seeds for each of the 22 recombining regions . For X-linked regions , we specified in PHASE that the phase of male haplotypes was known . All runs gave very similar reconstructions . Cryptic relatedness was assessed using the RELPAIR program v . 2 . 0 . 1 [39] . We divided our population samples into two geographic areas: Western Africa ( populations 1–4 and 8–10 in Figure 1 ) and Eastern Africa ( populations 5–7 and 11–12 in Figure 1 ) . We tested cryptic relatedness only between individuals coming from the same geographic area . We considered a pair of individuals as cryptically related when the likelihood of their inferred relationship was >1 , 000 higher than the likelihood of no cryptic relatedness between them . Twenty individuals were excluded based on this criterion: 1 G . Baka , 3 Bakola and 6 Biaka Pygmies , and 1 Yoruba , 3 Akele and 6 Mozambican farmers . Genetic membership of populations was inferred with STRUCTURE v . 2 . 1 software [40] , using the “correlations” and “admixture” models , with and without prior information about populations , 1 , 000 , 000 burn-in steps and 1 , 000 , 000 Monte Carlo Markov chain replications . We excluded the Y-linked and mtDNA regions from the STRUCTURE analysis because this program accepts only diploid loci . We recoded the 20 autosomal and two X-linked regions as microsatellites , considering each haplotype as an allele of a single multi-allelic locus . For each prior K value ( 2 , 3 , 4 and 5 ) , we ran 20 independent runs with different seeds and found likelihoods to be stable across runs . We focused on several aspects of our resequencing dataset , including classical diversity indices ( nucleotide diversity π , Watterson's estimator of theta θW and haplotype diversity Hd ) , neutrality statistics ( Tajima's D , Fu & Li's D* , Fu's Fs and their mean and variance across regions ) and population differentiation statistics ( pairwise FST ) . All these statistics , the observed site frequency spectra and those expected under a constant population size model , as well as the significance of FST values , were obtained with DnaSP v . 4 . 10 . 9 [60] . Novel summary statistics were also developed to capture particular aspects of the genetic data: the proportion of shared mutations between populations , the proportion of low-frequency shared mutations and the mean frequency of shared mutations , which were defined as follows . Consider S mutations segregating in populations i and j . Then is the number of segregating sites in population i , the number of segregating sites shared between populations i and j and the number of shared segregating sites between populations i and j with a relative frequency in merged populations lower than f . Then and . We used coalescence simulations ( i ) to assess the statistical significance of observed neutrality statistics and their means and variances across autosomal regions and ( ii ) to determine which models and parameters best fitted our empirical data . Simulations were performed using coalescent theory , as implemented in SIMCOAL v . 2 . 1 . 2 [61] , and using mutation rates ( μ ) and effective population sizes ( Ne ) drawn from gamma distributions ( Table S10 ) , as in previous studies [17] , [62] . The mean mutation rates of autosomal , X- and Y-linked regions were calculated from human-chimpanzee divergence , assuming that the two species diverged 6 million years ago [63] and a generation time of 25 years . For mtDNA , we used the synonymous mutation rate calculated in a previous study [14] . For all genomic regions , the number of mutations for the observed and simulated data was found to be similar ( data not shown ) . For each independent genomic region , the statistical significance of the neutrality statistics in each population group was assessed by comparing observed values with 100 , 000 values obtained from simulations of a sample , the size of which corresponded to that of the tested population sample , under a neutral model of evolution , assuming a constant population size and no recombination ( only ∼0 . 5% of haplotypes at autosomal regions showed evidence of recombination ) . The statistical significance of means and variances of neutrality statistics across the 20 autosomal regions was assessed by simulating 100 , 000 sets of 20 independent regions under the same assumptions . Models were tested by simulating 100 , 000 and 250 , 000 datasets under each demographic and IM models respectively , with model parameters randomly drawn from prior distributions ( see section below ) . For both the adjustment of the demographic regimes of each population group and the assessment of the branching history of population groups , the simulated model that best fitted our autosomal data was defined as that giving the highest proportion of small distances ( Ψξ ) between the simulated and observed summary statistics , S′ and S . These distances were measured by calculating the normalized metric D ( S′ , S ) [64] , and D ( S′ , S ) was considered to be small when lower than ξ = 0 . 5 . This flexible statistical framework , which is based on comparisons between simulations and observed data , makes it possible to test complex models with fluctuations in effective population size , population separation times and gene flow , without estimating the real likelihood of the data ( ξ = 0 ) , which would be unfeasible given the complexity of the data and the models . The tested demographic and IM models were all simulated with prior distributions of model parameters ( Tables S6 , S7 , S8 , S9 ) . We assessed whether a given model fitted the empirical data significantly better than another model , by resampling 100 times 10 , 000 simulations of each model , calculating for each model Ψ0 . 5 and estimating the P-value using a chi-square test comparing the proportion of small distances between the simulated and observed data , generated by each of the two models . The final P-value is the mean of the P-values obtained across the 100 resampling sets . For all model testing procedures , only the autosomal dataset was considered . Before estimating levels of divergence and gene flow between populations , we determined a demographic scenario best accounting for the observed within-population summary statistics of our three population groups ( AGR , WPYG and EPYG ) . We did not aim to identify a best-fitting model for the demographic regime of AGR populations , because historical [20] , [21] , [23] , linguistic [22] and previous genetic studies [8] , [36] , [65] , [66] strongly suggest that these populations have indeed undergone expansion . For our filtered population dataset of AGR individuals , we considered a single , recent population expansion , with the time of onset and exponential growth rate drawn from flat prior distributions ( time of onset: 5–7 . 5 Kya; growth rate: 0 . 005–0 . 01 ) . Simulated summary statistics ( S , π , Tajima's D and Fu & Li's D* ) under this demographic expansion were similar to the observed statistics for the AGR group ( data not shown ) . For Pygmy populations , we compared the empirical summary statistics obtained for the WPYG and EPYG population groups ( Table 1 ) with summary statistics for 3 , 000 , 000 simulations , considering 33 models of a constant-sized population or bottlenecks , varying in intensity , timing and duration ( Figure 4 , Table S8 ) . We considered this population to have experienced varying levels of gene flow with an expanding population ( Table S8 ) presenting mean summary statistics similar to those observed in the AGR population group ( Table 1 ) . The number of polymorphisms S , π , Tajima's D and Fu & Li's D* observed in the two PYG groups were chosen as the summary statistics for comparisons between simulated and observed data . This adjustment of the demographic regime of each population group was also performed for the composite population dataset ( Text S1 , Figure S2 , Tables S3 and S6 ) . We then investigated the branching history of the three population groups ( AGR , WPYG and EPYG ) , considering the previously described population-specific demographic models for each population group ( Table S9 ) : a model of a population expansion for AGR , a model of bottleneck with recovery for WPYG , and a model of bottleneck for EPYG . We tested four different models potentially accounting for the current genetic differentiation of the three population groups ( Figure 5 ) , using large flat prior distributions for separation time and migration rate parameters , except that the time of the oldest divergence was necessarily constrained by the time of the latest divergence ( Table S9 ) . We simulated 250 , 000 sets of 20 unlinked autosomal regions for each of the four IM models ( Figure 5 ) . We selected several summary statistics to discriminate between the confounding effects of divergence and gene flow on genetic variation: the proportion of mutations shared between populations , the proportion of low-frequency shared mutations , the mean frequency of shared mutations , and pairwise FST ( Text S1 , Figure S5 , Table S5 ) . We tested several combinations of statistics summarizing the within- and between-population genetic diversity ( data not shown ) . Finally , we used a set of statistics that included S , π , Tajima's D , Fu & Li's D* for each population group and pairwise FST and for each pair of population groups . This procedure ( i . e . , incorporation of the demographic characteristics of each population group into the estimation of their branching order ) was also applied to the composite population dataset ( Text S1 , Figure S3 , Tables S5 and S7 ) . Parameter estimation was based on the autosomal data alone . We estimated parameters under the best-fitting IM model ( i . e . , the A-WE model; Figure 6 ) , by comparing our empirical data with 250 , 000 simulations of 20 independent regions under the A-WE model , using large flat prior distributions for separation time and migration rate parameters , except that the time of the oldest divergence was necessarily constrained by the time of the latest divergence ( Table S9 ) . We then used the ABC method , which generates posterior distributions of the parameters of interest deduced from parameter values of simulations satisfying the D ( S′ , S ) <ξ criterion ( see previous section and [37] for more details ) , with ξ chosen so that only 5 , 000 of 250 , 000 simulations are retained [17] . For the ABC procedure , we used the following summary statistics: S , π , Tajima's D , Fu & Li's D* for each population group and pairwise FST and for each pair of population groups . This method was demonstrated to be accurate by estimating IM parameters for 100 simulated datasets for which the IM parameters were specified . Known parameter values were then compared with the 95% confidence interval ( CI ) for the ABC estimates of the parameter considered . Accuracy was estimated as the proportion of comparisons for which the known values were within the 95% CI for the estimated parameters . This procedure ( i . e . ABC estimation of IM parameters ) was also applied to the composite population dataset ( Table 2 , Text S1 , Figure S3 ) . | The central African belt represents a key region for understanding recent changes in human history and modes of subsistence because the largest group of hunter–gatherers of Africa , the Pygmies , still inhabits this region and coexists with neighboring agricultural populations . However , the understanding of the peopling history of equatorial Africa is hampered by the rapid disintegration of fossil remains in the rainforest's acidic soils . When archaeology fails , population genetics can reconstruct the history of populations from their present-day genetic variation . We generated a large resequencing dataset in different farming , Western Pygmy , and Eastern Pygmy populations dispersed over the African continent . By means of simulation-based inferences , we show that the ancestors of Pygmy hunter–gatherers and farming populations started to diverge ∼60 , 000 years ago . This indicates that the transition to agriculture—occurring in Africa ∼5 , 000 years ago—was not responsible for the separation of the ancestors of modern-day Pygmies and farmers . We also show that Western and Eastern Pygmy groups separated roughly 20 , 000 years ago from a common ancestral population . This finding suggests that the shared physical and cultural features of Pygmies were inherited from a common ancestor , rather than reflecting convergent adaptation to the rainforest . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"evolutionary",
"biology/human",
"evolution",
"genetics",
"and",
"genomics/population",
"genetics"
] | 2009 | Inferring the Demographic History of African Farmers and Pygmy Hunter–Gatherers Using a Multilocus Resequencing Data Set |
It is now recognized that molecular circuits with positive feedback can induce two different gene expression states ( bistability ) under the very same cellular conditions . Whether , and how , cells make use of the coexistence of a larger number of stable states ( multistability ) is however largely unknown . Here , we first examine how autoregulation , a common attribute of genetic master regulators , facilitates multistability in two-component circuits . A systematic exploration of these modules' parameter space reveals two classes of molecular switches , involving transitions in bistable ( progression switches ) or multistable ( decision switches ) regimes . We demonstrate the potential of decision switches for multifaceted stimulus processing , including strength , duration , and flexible discrimination . These tasks enhance response specificity , help to store short-term memories of recent signaling events , stabilize transient gene expression , and enable stochastic fate commitment . The relevance of these circuits is further supported by biological data , because we find them in numerous developmental scenarios . Indeed , many of the presented information-processing features of decision switches could ultimately demonstrate a more flexible control of epigenetic differentiation .
The capability of cells to present different stable expression states while maintaining identical genetic content plays a significant role in differentiation , signal transduction and molecular decision-making . These epigenetic phenotypes are partly associated to changes in genomic structural features , including several types of chromatin and DNA modifications [1] . Alternatively , it is also believed that in some cases they are induced by the action of underlying genetic regulatory circuits , exhibiting a positive feedback loop configuration . Recent studies have experimentally confirmed this latter prediction , both in natural and synthetic systems , e . g . , [2] , [3] , which originated back in the early days of microbial molecular genetics [4] and systems theory [5] . A positive feedback topology is nevertheless not sufficient to generate distinct epigenetic states . In addition , the circuit should display some degree of nonlinearity , i . e . , sigmoidality , on its constituent interactions [6]–[8] . This sigmoidal behavior is typical of many molecular interactions and endows these genetic modules , now interpreted as dynamical systems , with multistability , i . e . , the possibility to find the system in alternative steady states under conditions in which all its biochemical parameters are fixed . These equilibria define the different stable expression states regulated by means of the molecular loop . How does the particular structure of a given positive feedback influence its function ? Considering multistability as the most prominent attribute of these architectures , one could argue that genetic design does not really matter , as soon as sigmoidal interactions are achieved in some effective way . Careful analysis of some of the recent experimental reports seems to indicate the contrary . Two general patterns can be suggested . First , positive feedback loops at the core of more complex regulatory networks generally consists of simple structures controlling cell fate decisions . This is normally associated to two complementary expression states , i . e . , bistability ( like the p42–Cdc2 system involved in Xenopus oocytes maturation [9] , or the bacteriophage λ genetic switch [10] ) , but three states is also been recently discussed [11] , [12] . In comparison , loops relevant to signal transduction , or more broadly to conditions where complex biochemical information-processing is required , are commonly constituted by many components [13]–[15] . These architectures can even regulate the plain presence of mono or bistability , e . g . , as a function of the time the activating stimulus is applied [16] . The proposed patterns lead to a set of interesting questions . When is multistability , understood as at least three possible expression states , relevant in differentiation as opposed to bistability ? Which simple feedback loop architectures can produce it , and what biologically relevant parameters do we need to quantify in order to predict these behaviors ? Moreover , we can also ask to what extent complex feedback topologies are necessarily required for multifaceted information-processing and for the execution of elaborated developmental programs . To address some of these issues , we first investigate the number of available expression states of two minimal complementary systems—a two-component mutual-activation and mutual-inhibition circuits—whose constituents are autoregulated ( Figure 1A and 1B ) . Autoregulation is a pervading feature of many eukaryotic master regulators [17] , [18] , a property usually believed to merely impart stability to the associated regulations [18]–[20] . In this context , we show how it determines , in competition with the crossregulatory interactions , the ranges of mono , bi and multistability . Second , we discuss how these topologies facilitate two main switch classes involving transition between expression states in bistable and multistable regimes . We term these classes progression and decision switches , respectively . While multistable decision switches have been recently suggested as a rationale to explain co-expression of antagonistic master regulators in lineage specification scenarios [21] , [22] , we show that these switches not only appear in mutual-inhibition architectures . More importantly , we analyze several complex information-processing tasks in decision switches , such as signal strength [23] and duration [24] discrimination , stochastic fate control [25] , [26] and flexible discrimination [27] , [28] . This demonstrates that simple architectures can indeed show rich computations , an ability related to the presence of multistability and thus to autoregulation . We revisit this relation next to highlight how autoregulation compensates or amplifies transient expression differences between circuit components to induce robust coexisting stable expression states . We close by discussing the implications of these findings , and the specific architectures considered , in various cellular contexts .
To describe dynamical aspects of genetic regulatory networks one can generally adopt two contrasting strategies . The first one is to collect all available molecular information about the regulatory interactions of a specific system . This allows to present the putative regulatory network involved in the mechanism under study , which can then be quantitatively described by using a particular mathematical formalism , e . g . , a set of ordinary differential equations . While this method can be helpful to describe the dynamics of a very specific system , it usually incorporates a degree of complexity that can sometimes hide the key dynamical aspects , and molecular players , determining network behavior ( with the additional drawback that new molecular agents could always be discovered and thus modify both network topology and dynamics ) . An alternative approach is to propose simplified mathematical models based on a number of realistic assumptions . Simple models help in the identification of basic design principles , might act as effective descriptions of more complex circuits and , as we see below , can actually correspond to extant regulatory modules found in several biological scenarios . These models also circumvent the lack of quantitative molecular details required in the more specific studies . We follow here this second approach . We thus introduce a two-component mathematical model to analyze the dynamical behavior of the mutual-activation/mutual-inhibition topologies ( see equations in Materials and Methods , Figure S1 , Figure S2 , and also Text S1 for further details ) . In this model , autoregulatory and crossregulatory interactions between components were represented by Hill equations . This is a widely used approximation as molecular interactions are usually known to behave in a sigmoidal fashion [6] . Indeed , similar simplified models have been used to describe the coexistence of several expression states in specific cell fate systems , such as those involved in hematopoiesis [11] , [12] , [22] or embryonic stem cell differentiation [29] ) . In our case , we present this model as part of a general framework in relation to a broad number of biological scenarios ( see Table 1 ) , and fully characterize the type of information-processing features these circuits exhibit and their potential significance for a more flexible control of epigenetic differentiation . What specific biological features determine multistability ? We address this question by identifying a minimal set of biological determinants able to characterize circuit behavior . This analysis also helps us to highlight some unexpected features of the relation between module structure and epigenetics , and to introduce two main types of switches associated to the circuit dynamics . What type of signal processing enables the presented switches ? In the following , we show the potential of decision switches to robustly discriminate several characteristics of biochemical stimuli , e . g . , strength , duration , timing , etc . These capacities offer flexible control of epigenetic expression , far beyond that attributed to bistable ( progression ) switches . The preceding section discussed several new computational features associated to the presence of multistatibility in these systems . We also argued before how multistatibility is linked to autoregulation , a connection that we now further elaborate . Autoregulation favors multistability by either amplifying or compensating transient differences in expression between the circuit constituents . This modifies the type of steady states usually found in two-component switches; ( low , low ) or ( high , high ) expression for mutual-activation and ( low , high ) or ( high , low ) expression for mutual-inhibition , but see also [35] . Specifically , amplification habilitates activation switches with asymmetric expression states , while compensation induces the presence of a third symmetric equilibrium state in inhibition switches . This dual role requires autoregulation to be strong enough and dominant over crossregulation ( ρ>ν , σ<1 ) ( Figure S7 ) . How does autoregulation-based amplification work ? We consider a mutual-activation circuit in an initial ( transient ) asymmetric expression state , i . e . , x0≠y0 , with x0 , y0 , being the concentration of each circuit component in non-dimensional units . We plot the time evolution of the species concentration ( Figure 7A ) and the probabilities of occupation of the associated binding sites—by the corresponding autoregulatory and crossregulatory species ( Figure 7C ) . Despite the initial asymmetry , both components reach the same equilibrium expression ( or present a monomodal distribution around this value , inset Figure 7A , when considering noisy gene expression [32] ) . The activation of the species with higher initial concentration leads to the increase in expression of the second component , mediated by the crossinteractions . This effect ultimately balances the probability of occupation of each binding site by their own species , which filters out the initial concentration disequilibrium ( Figure 7C ) . Could the initial asymmetry be amplified , so that an asymmetric steady state is favored ? When crossactivation is weaker ( Figure 7B–D , ν = 2 ) , the autointeraction of the species with smaller initial concentration does not become active . Autoregulation is only effective then on the species with higher concentration ( Figure 7D ) . The circuit amplifies in this way the initial differences allowing the coexistence of symmetric and asymmetric equilibria ( we now find three peaks in the distributions obtained by considering noisy gene expression , inset Figure 7B ) . Moreover , autoregulated-based compensation , by comparison , avoids expected unbalances in mutual-inhibition switches working in regimes with hardly active crossinteractions ( small σ's ) , since the autoregulation on both species dominates ( Figure S8 ) . Multistability is however generally not expected in circuits exhibiting relatively strong mutual activation , as this might imply unrealistically strong autoregulation ( Figure 2 and Figure S7 ) . What would be then the influence of autoactivation ? We suggest that it provides switches with a more flexible behavior . For instance , modulation of its strength can enlarge the bistable region , making the module respond to some external signals that would not be sensed otherwise ( Figure S9 ) . Autoregulation may change also the combinatory of signals to which a mutual-activation switch responds ( Figure S10 ) . How biologically relevant is the general framework that we have proposed ? We investigated how these topologies enable different epigenetic regimes ( the phenotypic map ) , characterized the molecular switches driving transitions between them , and discussed several information-processing features exhibited by these switches . To put these ideas in a biological context , we now first identify the presence of these regulatory architectures in specific scenarios , and then discuss several signal discrimination features that have been analyzed in these and related settings . Indeed , we found a number of differentiation programs controlled by circuits constituted by two molecular regulators in a mutual-activation or mutual-inhibition configuration ( Table 1 ) . The main components of these circuits are usually master regulators , i . e . , transcription factors involved in the determination of cellular fate , which exhibited autoactivation ( a prevalent characteristic of master regulators , see Table S1 ) . Interactions between module constituents are sometimes direct , normally transcriptional , or indirect , mediated by other molecular species . Recently studied examples include the embryonic stem-cell master regulators Oct4 , Sox2 , and Nanog [36]–[38] . These factors establish mutual-activation architectures between them—to maintain pluripotency—or mutual inhibition circuits , in combination with additional elements , to induce specific developmental fates [29] , [39] , [40] . Common instances of the latter involve Cdx2 , promoting differentiation to trophectoderm [37] , or Gata4/Gata6 linked to endodermal differentiation [39] , see also [40] . Similarly , various stages of hematopoietic lineage specification are driven by modules exhibiting these architectures . In this situation , the presence of a third expression state , or priming state , is currently under inspection [12] , [21] , [22] , [41] . In particular , genes in various lymphoid lineages are coexpressed at low levels in common lymphoid progenitors ( CLPs ) and in common myeloid progenitors ( CMPs ) [42] , [43] . Specialization to different cell types from these common lineages proceeds through sequential steps where some genes are silenced and other activated [44] , [45] . This is the case in B and T cell development ( from CLP ) [19] , [46] , [47] as well as in the macrophage/neutrophil decision ( from CMP ) —where a graded decision switch was proposed [11] with a similar architecture to the ones discussed here , but see additionally [12] . As in the previous case of embryonic stem cell differentiation , genes common to several lineages , i . e . , at the top of the regulatory hierarchy , control the expression of other transcription factors involved in more specific lineage commitment . In summary , the collected scenarios in Table 1 suggest that the presence of mutual-activation circuits correlate with differentiation as a progression , while mutual-inhibition topologies appear mostly when alternative decisions from a precursor ( priming ) cellular state are made ( Figure 1 ) . What sort of signal discrimination is found in these contexts ? The influence of signal attributes in various developmental scenarios hints at the possibility that more elaborated signal processing could be at work . One example of this influence is the role of signal strength in thymocite differentiation [23] , [48] . In particular , CD4/CD8 T-cell fate commitment is determined by the strength of the T-cell receptor signal , with strong and weak signals favoring either the CD4 and CD8 lineages , respectively [48] . Another interesting case of processing of stimulus strength is morphogen gradient interpretation [49]–[51] . In Xenopus mesoderm formation , activin , a member of the TGF-β family , acts on downstream genes in a concentration-dependent manner , with high concentrations inducing expression of the transcription factor Gsc and low concentrations activating the Xbra transcription factor , both regulating each other in a double negative feedback loop [51] . Similarly , a gradient of Shh signaling can be read by complementary complementary pairs of homeodomain proteins that cross-inhibit each other in a cell autonomous manner [52] , specifying neural differentiation in the spinal cord . Finally , processing of signal duration has also been shown to be important in both T-cell fate commitment [53] and morphogen signalling [50] . The latter case is a good example of a more elaborated processing task in which Shh interpretation integrates both strength and duration of a signal to control differential gene expression . Moreover , the precise temporal expression programs exhibited by genes involved in cell differentiation suggests that discrimination between signals at different times could be also important in these situations . For instance , during pancreas development ( Table 1 ) one of the key factors promoting endocrine development ( Foxa2 ) is already expressed in the first stage of development ( the gut endoderm ) but other essential factor ( Sox9 ) appears later in the first pancreatic progenitor cells [54] . Signal processing ultimately leads to a fate decision . Two models , not necessarily exclusive , have been proposed . In a first model ( sometimes termed ‘instructive’ or ‘selective’ regulation [25] ) , an external signal imposes the specific fate by activating or repressing a particular set of genes . This probably corresponds to the more standard scenario of how signaling determine fates . An alternative model is that in which a given fate is stochastically chosen among different pre-existing programs . An open issue in this latter model , and one that partially answers our analysis , is how the proportion between alternative stochastic fates is regulated in a population . What specific molecule is critical for this particular physiological response ? This question is usually asked when a given cellular behavior comes under scrutiny . The search for such master regulators is specially relevant in the context of differentiation , where they become lineage specification factors , whose expression , or the lack of it , is associated to distinct cell fates . This approach , however , does not seem to be sufficient anymore . Indeed , an increasing number of studies confirm the view that regulators do not work in isolation , and that we need to study them as parts of genetic control circuits to properly recognize their function [9] , [11] , [17] , [19] , [37] . Even though the molecular components of such control circuits are obviously diverse , their architectures do exhibit two main unifying attributes . First , they represent a relatively simple positive loop structure , and second , this structure is constituted by interactions with a degree of sigmoidality ( threshold-like action ) that enables circuits to exhibit bistability [6]–[8] . Does the coexistence of more than two expression states lead to a fundamentally different type of regulation and signal processing ? If so , how can we determine multistability and to what extent is this feature linked to more complex loop architectures ? To analyze these issues , we characterized the function of two-component circuits with the use of mathematical models . An additional property in these systems is that their main constituents are autoregulated . We made use of the phenotypic map , a parameter space characterizing the patterns of expression associated to these modules . Transitions between expression states were then considered to be induced by two major switch classes . A progression switch corresponds to a transition in which at most two expressions states should be available . Alternatively , a decision switch needs of three expression states , one before and two after the decision . Both types correspond to distinct bifurcations of the system equilibria [30] . This analysis also highlighted the fundamental role that autoregulation plays in these designs . Specifically , autoregulation in the mutual-inhibition circuit favors multistability , and thus decision switches , while it provides mutual-activation switches with more flexibility and enlarged stimulus reaction ( but only two coexisting expression states , i . e . , progression switch ) . We examined a number of scenarios where master regulators and their interactions have been experimentally uncovered . We identified several architectures corresponding to the analyzed circuits , i . e . , constituted by two principal autoregulated molecular agents in a mutual-activation/mutual-inhibition topology . Our study also provided an elaborated rationale of why master regulators largely exhibit autoregulation . In addition , we correlated mutual-activation/mutual-inhibition switches with differentiation as a progression or decision , respectively . These theoretical arguments helped thus to unify a wide range of biological data , and present progression and decision switches as fundamental design principles in the control of epigenetic differentiation . We specially studied the abilities of these modules to respond and monitor stimuli , with special emphasis on decision switches . This revealed a series of findings . First , decision switches are able to elicit richer responses to differential signal parameters , like strength or duration , enhancing signal specificity [14] , [24] , [55] . Second , decision switches provide a circuit-based explanation to stochastic , but biased , cell fate determination [25] , [26] . Identical cells can exhibit heterogeneous responses when experiencing similar external cues . These switches originate stochastic differentiation when an external stimulus is able to unstabilize the current expression state of an homogeneous population . The remaining system equilibria are then potentially reachable to each member of the population in a stochastic manner , e . g . , due to biochemical noise [32] . The population distribution of gene expression can be further modulated by any asymmetry presented in the signal or the circuit main characteristics . Last , decision switches appear as a module able to process delayed signals . In particular , we showed how this switch can implement two-interval discrimination tasks . This capability allows cells to adapt to varying environments by holding the history of a previous condition in a kind of short-term memory ( working memory ) . Cells would modify their identities by comparing the first conditions with those found in a second environment . Some of these discrimination performances are similar to the one found in cortical circuits in monkeys , where neural network models of mutual inhibition with recurrent self-excitation have been hypothesized to mediate these decisions [27] , [28] , which emphasizes the presence of similar dynamical principles in circuits underlying apparently non-related biological functions [34] . The presence of a state of working memory also offers an alternative mechanism to the dynamics of transient gene expression . While most studies focused on the role of DNA structural modifications to transform these short-term states into stable long-term memories , [56] , [57] for two recent examples , decision switches would accomplish a similar function by means of feedback regulation .
To derive the mathematical models used in this study , we consider all biochemical reactions involved in transcription regulation and expression of two interacting genes ( dimerization reactions , binding/unbinding of transcription factors to promoters , transcription , translation and degradation , see Text S1 for details ) . Separation of time scales and standard quasi-steady state assumptions lead to the following model for the time evolution of the two gene products: ( 1 ) Here , x , y describe protein concentrations in non-dimensional form . We introduce two types of parameter sets . One is linked to the activation/inhibition strengths in units of basalproduction , i . e . , ρi , νi , μi with i = {x , y} . Specific ranges of ν determine the various modules under study , i . e , mutual inhibition with ν∈[0 , 1 ) , or mutual activation with ν>1 . The second parameter group is associated to the protein threshold values required for an interaction , i . e . , its response , to become active . Specifically , the σ's quantify the ratio of response threshold of each link , e . g . , ratio of binding affinities in the case of transcriptional interactions . In addition , αi and δi correspond to the basal production and degradation rate , respectively . We further assume a symmetric parametric regime ( parameters equal in both species , e . g . , ρ = ρx = ρy ) , and that σxy = 0 , i . e . , non-cooperative interactions ( to see the role of cooperativity see Figure S3C and S3D ) . Finally , we scale the time by the corresponding protein degradation rate which results in an average production rate , a = α/δ ( note also that larger Hill coefficient implies a higher degree of non-linearity but qualitative conclusions hold , Figure S11 ) . These parameters can be quantified experimentally , and represent a minimal set of biologically relevant features able to characterize circuit behavior . Phenotypic maps are obtained numerically by sampling the parameter space with different initial conditions and letting the system to reach all available steady states . Moreover , stochastic gene expression is simulated by using the Gillespie's algorithm in most cases ( taking into account mRNA dynamics , see Equation 7 in Text S1 ) . For the specific situation of signal discrimination , a Langevin method was implemented to reduce computational time ( Text S1 for details ) . | An essential attribute of living cells is the capacity to select among various alternatives when confronted with external or internal cues . These decisions can be directly linked to survival , as the sporulation/competence choice in the bacterium Bacillus subtilis , or be involved in the establishment of developmental programs from pluripotent stem cells . How are these decisions controlled at the molecular level ? Recent studies have identified the presence of a few master regulator genes whose activity is crucial to drive cells to a particular fate . These genes usually exhibit autoregulation and appear in combination with a second molecular partner constituting a minimal regulatory circuit . Here , we investigate how such two-component architectures endow cells with more than two epigenetic states . This ability not only enhances the number of potential developmental outcomes , in a given context , but also drastically increases cell signal processing . We show how these control modules can implement a number of complex computational tasks such as discrimination of stimulus amplitude , duration , or relative timing . Similar aspects have been discussed in relation to neural dynamics in cortical circuits , which suggests the use of equivalent computational strategies and circuit design in the control of biological decision-making . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"computational",
"biology/systems",
"biology"
] | 2008 | Multistable Decision Switches for Flexible Control of Epigenetic Differentiation |
Comparisons of levels of variability on the autosomes and X chromosome can be used to test hypotheses about factors influencing patterns of genomic variation . While a tremendous amount of nucleotide sequence data from across the genome is now available for multiple human populations , there has been no systematic effort to examine relative levels of neutral polymorphism on the X chromosome versus autosomes . We analyzed ∼210 kb of DNA sequencing data representing 40 independent noncoding regions on the autosomes and X chromosome from each of 90 humans from six geographically diverse populations . We correct for differences in mutation rates between males and females by considering the ratio of within-human diversity to human-orangutan divergence . We find that relative levels of genetic variation are higher than expected on the X chromosome in all six human populations . We test a number of alternative hypotheses to explain the excess polymorphism on the X chromosome , including models of background selection , changes in population size , and sex-specific migration in a structured population . While each of these processes may have a small effect on the relative ratio of X-linked to autosomal diversity , our results point to a systematic difference between the sexes in the variance in reproductive success; namely , the widespread effects of polygyny in human populations . We conclude that factors leading to a lower male versus female effective population size must be considered as important demographic variables in efforts to construct models of human demographic history and for understanding the forces shaping patterns of human genomic variability .
Many studies have demonstrated large differences between males and females in the forces of evolution , i . e . , mutation , recombination , selection , gene flow , and genetic drift . For example , mutation rates are often higher in males while females tend to have higher rates of recombination [1] . While the effects of sex-biased mutation and recombination have been directly estimated through genetic studies , we know very little about the extent to which sex-specific differences in gene flow and genetic drift have shaped patterns of variation at the level of the genome . For mammals , it is well known that females and males do not exhibit symmetrical behavior with respect to mating and dispersal practices . For instance , the typical mammalian system is characterized by polygyny ( a mating practice in which a minority of males sire offspring with multiple females ) and female philopatry ( the tendency for females to breed at or near their place of origin ) [2] . The development of sex-specific markers in humans has been instrumental in providing insights into the effects of sex-specific demographic processes . Contrasting patterns of diversity on the mitochondrial DNA ( mtDNA ) and non-recombining portion of the Y chromosome ( NRY ) have been interpreted to reflect sex-specificity in the rate and scale of migration and in effective population size [3]–[5] . However , these patterns could also reflect different molecular properties of these two haploid systems , differential selection , or stochasticity in the evolutionary process [5] . Unlike mtDNA and the NRY , the autosomes and X chromosome undergo recombination and contain numerous evolutionarily independent loci . Additionally , selection only affects those loci that are closely linked to selected sites . Consequently , different patterns of neutral polymorphism associated with the X chromosome and autosomes may be more directly ascribed to demographic differences between females and males . Under standard models of DNA sequence evolution [6] , the level of neutral polymorphism expected at equilibrium is governed by the product of Ne ( the effective population size ) and the mutation rate . Since males carry only one X chromosome , the ratio of the X chromosome effective population size ( Nx ) to the autosomal effective population size ( Na ) is expected to be ∼0 . 75 in simple models of a randomly mating population with equal numbers of breeding males and females ( i . e . , neutral models ) . Equivalently , if we correct for any differences in mutation rates across chromosomes , the X chromosome should have roughly 75% of the genetic diversity of the autosomes . However , under more complicated models the ratio of X to autosomal diversity levels can vary considerably [7] . For example , in populations with a female-biased sex-ratio , X-linked diversity will be higher than 75% of autosomal diversity [8] , while in populations that have undergone recent population bottlenecks X-linked diversity will generally be less than 75% of autosomal diversity [9] , [10] . In addition , if directional selection typically operates on mutations that are at least partly recessive , standard theory predicts that levels of diversity at linked neutral sites will be differentially affected depending on the chromosomal mode of inheritance . For advantageous recessive mutations , hemizygosity in males leads to a higher fixation rate on the X chromosome relative to the autosomes . This in turn will lead to less variability on the X chromosome relative to the autosomes due to the increased prevalence of genetic ‘hitchhiking’ [11]–[13] . In contrast , widespread purifying or background selection should reduce diversity on the autosomes more so than on the X chromosome [11] . In this paper , we analyze DNA sequence data that were collected by Wall et al . [14] for the purpose of testing models of human demographic history . In particular , we analyze data from the X chromosome and autosomes to examine the role that sex-specific processes have played in shaping genomic patterns of variability . We consider several alternative models that could lead to a skew in the ratio of X chromosome to autosomal diversity . Our sequence database includes 40 intergenic regions ( 20 on the X chromosome and 20 on the autosomes ) , each of which encompasses ∼20 kb of DNA ( Figure 1 ) . The sequenced regions were chosen from intergenic/non-coding ( i . e . , putatively non-functional ) regions of medium or high recombination ( r≥0 . 9 cM / Mb ) to minimize any potential confounding effects of natural selection ( see [14] for details ) . These data are also well-suited for testing the role of demographic processes in influencing patterns of diversity because all sites are resequenced in each individual , and multiple diverse human populations are represented in our survey ( i . e . , Biaka from Central African Republic , Mandenka from Senegal , San from Namibia , French Basque , Han Chinese and Melanesians from Papua New Guinea ) . We also utilize the recently available orangutan genome to obtain more accurate estimates of the underlying mutation rate for each of the regions studied .
We analyze a total of ∼210 kb of DNA sequence representing 40 loci from the X chromosome and autosomes from each of 90 humans and three great apes , or a total of ∼18 . 9 Mb [14] . Table 1 provides basic summary statistics for nucleotide diversity in six human populations , as well as the ratio of diversity to human-orangutan sequence divergence . We also use levels of divergence between humans and orangutan ( see Methods ) to estimate mutation rates for each region ( Table S1 ) , and then estimate relative effective population sizes of the X chromosome and autosomes ( Nx / Na ) based on observed levels of diversity ( θW ) [14] . We find that this ratio is higher than expected in all six populations , ranging from 0 . 85 in the San to 1 . 08 in the Basque ( Figure 2 ) . When we use levels of divergence between humans and chimpanzees to estimate mutation rates for the autosomal and X-linked regions , we obtain similar results . For instance , X/A diversity ratios ( e . g . , π/D in Table 1 ) using chimpanzee and orangutan divergence are highly correlated for the six human populations ( r2 = 0 . 95 , P = 0 . 001 ) ( data not shown ) . We also obtain similar π/D values when we subsample the human dataset to standardize the number of autosomes and X chromosomes ( Table S2 ) . To test whether the observed ratios are significantly different from 0 . 75 , we employ a maximum-likelihood method to estimate confidence intervals . Our method uses a population genetic model ( i . e . , the coalescent ) to account for the inherent uncertainty in estimating diversity and divergence rates from sequence data . Figure 2 shows 95% confidence intervals for Nx / Na . For three out of six populations ( Basque , Melanesians and Mandenka ) , the 95% confidence intervals for the ratio of X-linked and autosomal effective population sizes does not include 0 . 75 ( p = 0 . 001 , 0 . 005 and 0 . 030 for the Basque , Melanesians and Mandenka , respectively ) . One interpretation of these results is that there is strong evidence for an unequal female and male Ne in at least three of our six populations , with estimates of the breeding sex ratio ( i . e . , the effective size of females to males ) ranging from 2 . 1 in the San to 12 . 5 in the Basque . If the observed differences in nucleotide variability on the X chromosome and autosomes are caused by long-term ( demographic ) processes , then the estimates of Nx / Na presented in Figure 2 will be highly correlated due to shared population history . When we use the intersection of all six confidence intervals ( 0 . 87–1 . 02 ) to estimate the range of Nx / Na values that are consistent with the data from all six populations , we estimate the range of the breeding sex ratio to be 2 . 4–8 . 7 . We also note that even with a conservative Bonferroni correction , a 1∶1 breeding sex ratio is rejected in two out of six populations . We also employ a separate method for estimating the breeding sex ratio in each population that does not allow for intra-locus recombination but does permit independent mutation rates across loci ( see Methods ) . This method produces similar results to those described above , with estimates of the ratio of female to male effective population size ranging from 1 . 8 in the San to 14 . 0 in the Basque ( Table S3 ) . We interpret this as additional evidence that the unusual patterns observed in our data are real and require explanation .
While directional selection on recessive beneficial mutations is expected to lead to more frequent hitchhiking and lower diversity on the X chromosome compared with the autosomes , linked negative selection on the X chromosome and autosomes ( background selection ) predicts the opposite pattern [11] , [27] . Because recessive deleterious mutants are maintained at lower frequency and removed from populations more quickly on X chromosomes than on autosomes , neutral alleles on X chromosomes are less likely to be linked to a deleterious mutant compared with neutral alleles on autosomes . Thus , all else being equal , background selection should leave X chromosomes more polymorphic than autosomes at linked , neutral sites after correcting for expected differences in population size between X chromosomes and autosomes [12] . Because the effects of background selection are expected to be stronger ( i . e . , reduce local Ne ) in chromosomal regions with lower rates of recombination , we did not a priori believe that background selection would be a significant factor because our experimental design focuses on intergenic DNA in regions of moderate to high recombination [14] . To further explore the potential effects of background selection we assume an average number of deleterious mutations per generation of 4 [28] and use equation 15 in Hudson and Kaplan [29] to estimate the ratio of observed to expected polymorphism . We find this ratio to be 0 . 934 , which suggests that background selection is unlikely to reduce autosomal diversity by more than 6 . 6% relative to X-linked diversity . We note that this estimate is conservative in that it ignores the effects of background selection on the X chromosome [30] , [31] . Thus , it seems unlikely that background selection alone can explain our results . We also point out that alternative selection-based models involving the greater accumulation of sex-antagonistic polymorphisms on the sex chromosomes [32] may be viable . Historical changes in population size ( such as founder effects and bottlenecks ) also might have differential effects on loci with different modes of inheritance [9] , [10] , [33] . Using a simulation approach , we test three plausible models of recent demographic history that incorporate a recent population bottleneck and/or recent population growth . For example , we test a model incorporating 100-fold exponential growth from a constant effective population size of 104 , a bottleneck model with a 100-fold reduction in size followed by instantaneous recovery to an ancestral effective size of 104 , and a model incorporating the aforementioned bottleneck followed by 100-fold exponential growth ( see Methods for details ) . For all parameters tested , the effects on expected relative levels of diversity are minor and in the direction towards reduced X-linked polymorphism ( Table 2 ) . Recently , Pool and Nielsen [16] used an analytical approach to examine the effect of changing population sizes on the expected coalescence time for a pair of sequences with different effective population sizes . They showed that population size reductions can lead to particularly low X-linked diversity , whereas population growth can elevate X-linked relative to autosomal diversity . We employ Pool and Nielsen's [16] model ( which is similar to the bottleneck model described above ) , substituting parameters that are reasonable for human demographic history ( see Text S1 for details ) . When we examine the effects of such a bottleneck over a range of times in the past , we do not find that the expected X/A diversity ratio shifts much above 0 . 75 ( Figure S1A ) . When we search for combinations of parameters that yield X/A diversity ratios and levels of nucleotide diversity similar to those that we observe , we find that ancient bottlenecks ( i . e . , older than ∼100 kya ) coupled with population growth , can indeed produce expected X/A diversity ratios as high as 0 . 85 ( Figure S1B ) . However , computer simulations using these same bottleneck and growth parameters yield summaries of the site frequency spectrum that are inconsistent with those that we observe; i . e . , Tajima's D values that are much more negative than those in Table 1 for both the X chromosome and autosomes ( −1 . 55 and −1 . 91 , respectively ) . There are a number of sex-biased evolutionary forces acting within human populations that are known to have differential effects on loci with different modes of inheritance . A demographic process that may lead to a skew in X-linked versus autosomal diversity is differential migration rates for males and females in a structured population . To explore the effects of sex-biased migration on ratios of X/A diversity , we simulate a two-deme island model with different rates of male and female migration . First , we simulate a symmetric model with only a single sex ( females ) migrating . We assume effective population sizes and migration rates that produce FST values that are similar to those observed in human populations ( i . e . , autosomal FST ∼0 . 12; [14] ) . Because females are exchanging demes at the same rate , it is not surprising that this model yields X/A diversity ratios that are close to those expected under panmixia ( i . e . , 0 . 75 ) ( Table S4 ) . Second , we simulate a model in which one deme sends out females and the other deme sends out males at the same rate . The results indicate that when the X-linked diversity exceeds the value expected under panmixia in one deme , the other deme always shows a deficit of X-linked diversity ( Table S4 ) . If we assume that the six populations that we sampled here evolve independently according to this two-deme model , the probability of observing excess polymorphism on the X chromosome for all six populations would be at most 1/64 ( P<0 . 016 ) . These results are consistent with Laporte and Charlesworth's [34] simulations showing that sex-biased migration only weakly skews levels of X/A diversity unless populations are strongly subdivided . Therefore , we believe that population structure is unlikely to generate a bias towards increased diversity on the X chromosome in all populations , but could contribute to differential bias among populations . A higher variance in male reproductive success over that in females due to sexual selection is also expected to inflate the ratio of X-linked to autosomal polymorphism . In populations with age structure , an additional contribution to the variance in net reproductive success can be caused by the stochastic nature of survival during the reproductive phase and by differences in fertility among individuals in different age classes [26] . However , demographic factors of this kind ( e . g . , lower male survival during adult life or delayed male versus female age of maturity ) are unlikely to have a major effect on the relative effective population sizes of X-linked and autosomal loci [26] . In contrast , an excess variance of male reproductive success over Poisson expectations can have large effects: With an extremely high variance in male fertility relative to female fertility , the ratio Nx / NA approaches 1 . 125 [7] , [26] , [34] . A number of evolutionary forces may be responsible for increasing the effective population size of X-linked versus autosomal loci . Under reasonable parameters for human populations , our results suggest that background selection , changes in population size , and sex-specific migration in a structured population may each have a minor effect in increasing the ratio of X-linked to autosomal polymorphism over that expected under neutral models . While it is possible that multiple processes acting together might lead to a major effect ( i . e . , on the order of what is observed here ) , we hypothesize that a higher variance in male versus female reproductive success can by itself explain most of the observed increase in effective population size of the X chromosome . The human mating system is considered to be moderately polygynous , based on both surveys of world populations [35] , [36] and on characteristics of human reproductive physiology [37]–[39] . The practice of polygyny , in both the traditional sense and via ‘effective polygyny’ ( whereby males tend to father children with more females than females do with males—a common practice in many contemporary western cultures [40] ) , would tend to increase the variance in reproductive success among males . In other words , when more men than women in any generation fail to have any children , and more men than women have very large numbers of children , autosomal Ne is reduced relative to that of the X chromosome . While polygyny may be the most important factor influencing the ratio of X-linked to autosomal diversity , we point out that this process by itself is unlikely to account for all the patterns of nucleotide polymorphism observed here ( e . g . , the frequency spectrum as summarized by Tajima's D in Table 1 ) . Future theoretical work examining the joint effects of multiple demographic processes ( e . g . , sex-biased bottlenecks in which populations are founded by more females than males ( e . g . , [21] ) and experimental research ( e . g . , aimed at refining estimates of the ratio of X-linked to autosomal neutral polymorphism in additional populations ) will increase our understanding of how the different forces of evolution influence variation on the autosomes and X chromosome .
The DNA samples used in this study come from the CEPH Human Genome Diversity Panel [41] , the YCC collection [42] , and established collections in the Hammer lab ( see [14] for details ) . The regions used for sequencing were selected to minimize any potential confounding effects of natural selection . Specifically , we identified 40 different intergenic ( i . e . , putatively non-functional ) regions of ∼20 Kb in length with to medium to high recombination ( r≥0 . 9 cM/Mb ) [43] . While the genome-wide average recombination rate ( mean±SE ) for autosomes and the X chromosome are ∼1 . 29±0 . 018 cM/Mb and ∼1 . 25±0 . 091 cM/Mb , respectively [43] , the average recombination rate for our autosomal and X-linked loci are 2 . 18±0 . 16 and 2 . 29±0 . 23 , respectively . Each region was at least 50 Kb ( 100 Kb for the autosomes ) away from the nearest gene; within each region , we gathered ∼4–6 Kb of sequence data from 3 or 4 discrete subsections that spanned most of the distance of each region ( locus trio ) . For more details on the sequenced regions and the sequencing strategy , see Wall et al . [14] . See Table S5 for the number of alleles sequenced at each locus . We used a maximum-likelihood framework for estimating the effective population size for the X chromosome ( denoted by Nx ) and for the autosomes ( Na ) . We did this separately for each of the six study populations . We tabulated the number of segregating sites and the number of fixed differences ( between human and orangutan ) for each locus , and then used coalescent simulations [44] to estimate the probability of these observations as a function of the population size and the mutation rate . Similar results were obtained when the chimpanzee was used as an outgroup . We assumed an average generation time of g = 25 years for all human generations since the human most recent common ancestor ( MRCA ) and an average generation time of g = 20 years for all generations between the human MRCA and the orangutan sequence . We fixed the human–orangutan split time at 15 million years ago and assumed an ancestral human–orangutan population size of 40 , 000 for the autosomes and 30 , 000 for the X chromosome . The mutation rate was assumed to be constant per base pair , but different for the X ( μX ) and the autosomes ( μA ) . For a specific population , let SAi denote the number of segregating sites in the i-th autosomal locus and let SXj denote the number of segregating sites in the j-th X-linked locus . Similarly , let DAi and DXj denote the number of fixed differences between human and orangutan at the i-th autosomal and the j-th X-linked locus respectively . We correct DAi and DXj for multiple hits using the Jukes-Cantor model [45] . Then , the likelihood we are interested in is ( 1 ) Our basic strategy is to consider a grid of μA , μX , NA , and ( NX / NA ) values and to use Monte Carlo coalescent simulations to estimate ( 1 ) for each grid point . In particular , μA and μX are incremented in units of 0 . 1×10−8/bp per generation , Na is incremented in units of 500 , and ( Nx / Na ) is incremented in units of 0 . 05 . At each locus we generate 105 ancestral recombination graphs ( ARGs ) of n = 9–34 human sequences ( corresponding to the sample size for the actual data ) and one orangutan sequence , reproducing both the actual lengths sequenced and the gaps between the sequenced segments . These ARGs have a recombination rate that is constant per base pair , with the rate estimated from the deCODE map [43] , assuming an effective population size of 12 , 500 . Next , we tabulated the total branch lengths of branches that would lead to segregating sites or fixed differences . For any particular set of parameter values {μA , μX , Na , and ( Nx / Na ) } , it is straightforward to calculate the expected number of segregating sites and fixed differences under the infinite-sites model . Denote these by ES and ED respectively . Then , the probabilities in ( 1 ) follow from the Poisson distribution , andfor the autosomal loci orfor the X-linked loci . Note that the same set of simulations is used to estimate probabilities for a locus over all grid points simultaneously . This added computational efficiency comes at the cost of assuming ρ / bp ( for the ARGs ) is the same across all different values of Na and Nx . Simulations assume a constant population size and no population structure for each human population . The results are somewhat robust to specific demographic assumptions ( see below ) . Denote the female effective population size by Nf and the male effective population size by Nm . We use two separate approaches for estimating the breeding sex ratio α = Nf / Nm . First , we use a method of moments approach to obtain point estimates of α . Define ß = Nx / Na . From standard population genetics [8] , Substituting and rearranging terms leads toWe then substitute the point estimates for ß obtained above to generate point estimates for α . The second method to estimate the breeding sex ratio is a likelihood-based approach similar to the method for estimating Nx / Na described above . As before , we use maximum-likelihood to obtain a point estimate ( of α ) and likelihood-ratio tests to estimate 95% confidence intervals , separately for each of the six populations . In this approach , we assume no recombination within loci , free recombination between loci , and no variation in coalescence times of lines in the ancestral human-orangutan population across the genome . Unlike the previous method , we assume that the mutation rates are not constant across loci . Denote the mutation rates at the i-th autosomal locus and the j-th X-linked locus by μAi and μXj , respectively . Using the same notation as before , the desired likelihood isSince each locus is independent , we can simply maximize the likelihood over μ , Na and α separately for each locus . For the divergence terms the probability is Poisson distributed:where t is the number of generations since the human-orangutan split and No is the effective population size of the ancestral human-orangutan population for the locus in question . For the polymorphism terms , we utilize an exact expression that is available for the standard coalescent without recombination [46]:where Ne is the effective population size of the locus and k is the sample size . Note that no simulations are necessary for calculating likelihoods . Point estimates for α ( as well as 95% CI ) are shown in Table S1 . Results on the performance and robustness of the various estimation methods used are described in Text S1 . To test whether alternative demographic models might influence the observed ratio Nx / Na , we considered simple models that incorporated a population bottleneck and/or recent population growth . These simulations assumed Nx = 7 , 500 , Na = 10 , 000 , g = 25 years , θ = ρ = 0 . 001 / bp in the ancestral population , n = 32 for the autosomes and n = 16 for the X chromosome . Our growth only model assumed that a population of constant size began growing exponentially at various times in the past ( i . e , 10 , 15 , and 20 kya ) , expanding to a size 100-fold larger than the ancestral population . Our bottleneck only model assumed that an ancestral population underwent a 100-fold decrease in size at various times in the past ( i . e . , 10 , 20 , 30 , and 40 kya ) before instantaneously recovering its original size . In all cases the bottleneck lasted for 40 generations . We also considered a modification of the bottleneck model where the population grows exponentially at various times ( i . e . , 10 , 15 , and 20 kya ) after recovering from the bottleneck described above ( for cases where the onset of the bottleneck was 20 and 40 kya ) . For each parameter combination , we simulated 104 replicates of a 5 Kb region , and tabulated θ̅x/θ̅a . We then considered simple two-deme island models to test the effects of sex-biased migration rates on θ̅x/θ̅a . Each population experiences a per-generation migration rate of 3–9×10−5 and an effective population size of 104 . We test symmetric migration models in which females and males migrate equally between demes and in which only females or only males migrate between demes , as well as asymmetric models in which females and males migrate in opposite directions between demes . We performed 10 , 000 simulations for each model ( Table S2 ) . | Like many primate species , the mating system of humans is considered to be moderately polygynous ( i . e . , males exhibit a higher variance in reproductive success than females ) . As a consequence , males are expected to have a lower effective population size ( Ne ) than females , and the proportion of neutral genetic variation on the X chromosome ( relative to the autosomes ) should be higher than expected under the assumption of strict neutrality and an equal breeding sex ratio . We test for the effects of polygyny by measuring levels of neutral polymorphism at 40 independent loci on the X chromosome and autosomes in six human populations . To correct for mutation rate heterogeneity among loci , we divide our diversity estimates within human populations by divergence with orangutan at each locus . Consistent with expectations under a model of polygyny , we find elevated levels of X-linked versus autosomal diversity . While it is possible that multiple demographic processes may contribute to the observed patterns of genomic diversity ( i . e . , background selection , changes in population size , and sex-specific migration ) , we conclude that an historical excess of breeding females over the number of breeding males can by itself explain most of the observed increase in effective population size of the X chromosome . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"evolutionary",
"biology/human",
"evolution",
"evolutionary",
"biology/sexual",
"behavior",
"genetics",
"and",
"genomics/population",
"genetics",
"evolutionary",
"biology/genomics"
] | 2008 | Sex-Biased Evolutionary Forces Shape Genomic Patterns of Human Diversity |
Hepatitis B virus ( HBV ) infection of hepatocytes begins by binding to its cellular receptor sodium taurocholate cotransporting polypeptide ( NTCP ) , followed by the internalization of viral nucleocapsid into the cytoplasm . The viral relaxed circular ( rc ) DNA genome in nucleocapsid is transported into the nucleus and converted into covalently closed circular ( ccc ) DNA to serve as a viral persistence reservoir that is refractory to current antiviral therapies . Host DNA repair enzymes have been speculated to catalyze the conversion of rcDNA to cccDNA , however , the DNA polymerase ( s ) that fills the gap in the plus strand of rcDNA remains to be determined . Here we conducted targeted genetic screening in combination with chemical inhibition to identify the cellular DNA polymerase ( s ) responsible for cccDNA formation , and exploited recombinant HBV with capsid coding deficiency which infects HepG2-NTCP cells with similar efficiency of wild-type HBV to assure cccDNA synthesis is exclusively from de novo HBV infection . We found that DNA polymerase κ ( POLK ) , a Y-family DNA polymerase with maximum activity in non-dividing cells , substantially contributes to cccDNA formation during de novo HBV infection . Depleting gene expression of POLK in HepG2-NTCP cells by either siRNA knockdown or CRISPR/Cas9 knockout inhibited the conversion of rcDNA into cccDNA , while the diminished cccDNA formation in , and hence the viral infection of , the knockout cells could be effectively rescued by ectopic expression of POLK . These studies revealed that POLK is a crucial host factor required for cccDNA formation during a de novo HBV infection and suggest that POLK may be a potential target for developing antivirals against HBV .
Despite the availability of effective vaccines for more than three decades , hepatitis B virus ( HBV ) still persistently infects 240 million people worldwide [1 , 2] . Antiviral therapies with nucleos ( t ) ide analog inhibitors of HBV reverse transcriptase dramatically reduce the viral load , significantly improve the liver function and lower the incidence of liver failure and hepatocellular carcinoma , but fail to cure the viral infection [3 , 4] , due to the persistence of covalently closed circular ( ccc ) DNA in the nuclei of infected hepatocytes [5–8] . Hence , better understanding the molecular mechanisms underlying the formation and maintenance of cccDNA is critical for development of novel therapeutics to cure chronic HBV infection . HBV is the prototype member of Hepadnaviridae family and contains a relaxed circular ( rc ) partially double-stranded DNA genome with its DNA polymerase covalently linked to the 5’ terminus of minus strand [9] . While the minus strand of the rcDNA is completely synthesized with a redundant overhang at the 5’ end , the plus strand is incompletely synthesized , leaving a 3’ terminal gap of variable length [9] . HBV replicates its DNA genome via reverse transcription of an RNA intermediate , the pregenomic ( pg ) RNA [10] . Briefly , HBV entry into hepatocytes is mediated by its host cellular receptor human sodium taurocholate cotransporting polypeptide ( NTCP ) [11–14] . Upon entry into the cytoplasm of hepatocytes , rcDNA in the nucleocapsid is transported into the nucleus and converted into an episomal cccDNA , which is assembled into a minichromosome to serve as the template for the transcription of viral mRNAs [15 , 16] . Following the synthesis of viral proteins , viral DNA polymerase binds to a stem-loop structure ( termed epsilon ) within the 5’ region of pgRNA to initiate its packaging into nucleocapsids where the pgRNA is reverse transcribed to progeny rcDNA [17] . The progeny “mature” rcDNA-containing nucleocapsids can be either enveloped and secreted out of the cell as virion particles or might be redirected into the nucleus to amplify the cccDNA pool [18–20] [21] . Thus , the formation and intracellular amplification of cccDNA plays a central role in the establishment and maintenance of persistent infection . Biochemically , conversion of rcDNA to cccDNA requires the removal of the viral DNA polymerase and RNA primer from the 5’-terminus of minus strand and plus strand DNA , respectively; filling in the gap in plus strand DNA , trimming and ligating the ends of both strands . Although it is speculated that all those reactions are most probably catalyzed by host cellular DNA repair enzymes , identification of the cellular proteins responsible for cccDNA formation has thus far only achieved limited success . For instance , tyrosyl-DNA phosphodiesterase-2 ( Tdp2 ) , a cellular enzyme responsible for cleavage of tyrosyl-5' DNA linkages formed between topoisomerase II and cellular DNA [22] , can release covalently linked RT from the 5’ end of minus-strand DNA in vitro [23 , 24] , and has recently been shown to cleave the tyrosyl-minus strand DNA linkage of HBV . However , Tdp2 gene knockout only slows down the formation of duck hepatitis B virus ( DHBV ) cccDNA from intracellular amplification pathway , but does not inhibit HBV cccDNA formation in HBV infection of HepG2-NTCP cells [25 , 26] . In addition to rcDNA , cccDNA can also be formed from double stranded linear DNA ( dslDNA ) [27] , which is derived from in situ priming of plus strand DNA synthesis [28] . Interestingly , Ku80 , a component of non-homologous end joining DNA repair pathway , has been reported to play an essential role in the synthesis of DHBV cccDNA from dslDNA , but not rcDNA [29] . Completion of plus strand DNA synthesis , or “filling the gap” in the plus strand of rcDNA , is essential for cccDNA synthesis . Studies of DHBV and woodchuck hepatitis virus ( WHV ) , two hepadnaviruses distinct from human HBV but readily cultivable in vitro , showed that viral DNA polymerase inhibitors did not prevent cccDNA formation in the infection of primary hepatocytes of ducks and woodchucks , implying that viral DNA polymerase may be dispensable , while cellular DNA polymerase activity is required for the completion of plus strand DNA synthesis [30–33] . Moreover , continued treatment of primary tupaia hepatocytes and HepaRG cells with viral polymerase inhibitors during HBV infection did not inhibit HBV cccDNA formation [34 , 35] , further suggesting that cellular factor ( s ) play an important role . Here we set out to identify the DNA polymerase ( s ) that complete ( s ) the plus strand DNA synthesis required for HBV cccDNA formation . In HepG2-NTCP cells infected with HBV that is deficient in core protein production , we unambiguously demonstrated that viral DNA polymerase activity is not required for cccDNA formation in a de novo infection . Instead , a focused RNA interference loss-of-function screening identified POLK as a crucial cellular polymerase supporting HBV infection . Both knockdown and knockout of POLK impaired the conversion of rcDNA to cccDNA , which could be rescued by ectopic expression of POLK . Our findings thus suggest that POLK is a key host factor required for cccDNA formation during a de novo HBV infection , and therefore , a potential target for therapeutic intervention of chronic hepatitis B .
To investigate the molecular mechanism of HBV cccDNA formation , we first determined the kinetics of cccDNA synthesis in wild-type HBV infected HepG2-NTCP cells . Using Southern blot analysis , we examined HBV cccDNA in Hirt extracts of infected cells at various time points post infection . The identity of the cccDNA , which migrated at the position of 2 . 1kb linear DNA , was confirmed by the band shift to a 3 . 2kb DNA species corresponding to the size of unit-length linear HBV genomic DNA upon digestion by EcoRI , but not HindIII ( S1 Fig ) . As shown in Fig 1A , the protein-free rcDNA species accumulated at 12 and 24 h post infection and cccDNA became detectable at day 2 post-infection , followed by a modest increase in the next 5 days . The appearance of cccDNA was coincident with reduction of the protein-free rcDNA in the infected cells . Consistently , based on the quantitative analysis of cccDNA using a more sensitive real-time PCR assay , cccDNA was detectable at 24 h post-infection , markedly increased in the first 2 days post-infection , followed by a slower increase to approximately 3 copies per infected cell in the next 5 days ( Fig 1B ) . Similar to the kinetics of cccDNA synthesis , the levels of intracellular HBV 3 . 5kb vRNA ( Fig 1C ) as well as secreted HBeAg and HBsAg ( Fig 1D ) gradually increased following HBV infection . Immunostaining revealed that over 60% of HepG2-NTCP cells were HBcAg positive at day 7 post infection ( Fig 1E ) . As expected , HBV preS1 myr-47 lipopeptide ( myr-47 ) completely blocked the viral infection . Together , these results demonstrated that HepG2-NTCP cells support an efficient HBV infection , resulting in readily detectable cccDNA formation , gene expression and secretion of viral proteins . Hence , the HBV infection cell culture system is suitable for identification of viral and host cellular factors required for cccDNA formation during a de novo HBV infection . As stated above , nuclear HBV cccDNA pool is established by direct conversion of rcDNA from incoming virions and supposedly via intracellular amplification pathway from rcDNA in the cytoplasmic progeny nucleocapsids . It is clear that viral DNA polymerase activity is essential for the intracellular amplification of cccDNA , but its role in cccDNA formation from incoming virions , specifically , in filling the gap in the plus strand of rcDNA , remains to be rigorously examined . Providing an unambiguous answer to this question is challenging due to the apparently indistinguishable nature of cccDNA synthesized from the two different pathways . In order to thoroughly determine the role of HBV DNA polymerase in cccDNA formation during a de nove infection , we first produced HBV virion particles containing genomic DNA with a stop codon at the 38th codon ( Y ) of the core gene open reading frame , designated as HBV-ΔHBc . As depicted in S2 Fig , this was achieved by co-transfection of Huh-7 cells with plasmid containing 1 . 05-mer HBV DNA with the desired mutation and plasmid expressing HBV core protein . HBV-ΔHBc virions were harvested from the culture fluid and purified by ultracentrifugation . Due to its inability to produce capsid protein , HBV-ΔHBc infection of cells will not be able to support progeny viral DNA synthesis and formation of cccDNA through the intracellular amplification pathway . Hence , HBV-ΔHBc infection provides a unique opportunity to investigate the role of DNA polymerases in cccDNA formation from incoming virions . HepG2-NTCP cells were infected with a multiplicity of 100 genome equivalents ( mge ) of wild-type HBV and HBV-ΔHBc virions , respectively . Notably , HBV-ΔHBc successfully infected HepG2-NTCP cells and formed cccDNA at a level comparable to that of wild-type HBV on day 7 post infection and persisted during the following two weeks of extended culturing ( Fig 2A ) . Similarly , immunostaining analysis showed that the levels of intracellular HBsAg were similar in both wild-type HBV and HBV-ΔHBc infected cells during the prolonged 21 days of culture ( Fig 2B ) . As expected , HBV-ΔHBc infected cells did not produce HBeAg in the culture supernatant due to the stop codon mutation . Meanwhile , HBV-ΔHBc infected cells secreted a slightly higher level of HBsAg than wild-type HBV infected cells at the indicated time points post infection ( Fig 2C ) . Together , the results indicate that cccDNA in HBV infected HepG2-NTCP cells were mainly synthesized from input viral rcDNA and that the intracellular amplification pathway did not significantly contribute to the establishment of cccDNA pool under this experimental condition . To directly examine whether HBV polymerase activity is required for conversion of viral rcDNA from the input viruses to cccDNA , HepG2-NTCP cells were infected with HBV-ΔHBc in the presence or absence of adefovir ( ADV ) or entecavir ( ETV ) . As shown in Fig 2D and 2E , neither of the compounds prevented cccDNA accumulation in HBV-ΔHBc infected cells . Quantitative analysis indicated that the viral DNA polymerase inhibitor treatment at relatively high concentration reduced cccDNA accumulation by less than 27% . Interestingly , similar ADV and ETV treatment of wild-type HBV infected cells reduced cccDNA accumulation by 28 to 42% ( S3 Fig ) . These results indicate that the activity of HBV DNA polymerase is dispensable in cccDNA formation from incoming virion rcDNA , and re-enforce the notion that intracellular amplification of cccDNA does not play an important role in establishment of cccDNA pool in the HBV-infected hepatoma cells . There are fifteen different DNA polymerases in mammalian cells . They function in genome replication , DNA repair , and translesion DNA synthesis ( TLS ) of damaged DNA . To identify which host DNA polymerase contributes to HBV infection , we conducted a focused siRNA screening by targeting cellular DNA polymerase genes . We assessed HBV infection upon silencing the expression of individual DNA polymerase gene in HepG2-NTCP cells . All siRNA sequences targeting cellular DNA polymerase genes were obtained from previous studies; these individual siRNAs could unequivocally discriminate between the mRNAs of the 15 different polymerases [36] . An siRNA targeting NTCP and a scrambled sequence ( siRNA NC ) were used as positive and negative controls , respectively . The levels of intracellular 3 . 5kb viral RNA as well as secreted HBeAg in culture supernatants were measured by qRT-PCR and ELISA . As shown in Fig 3 , among 15 cellular DNA polymerases , knocking down POLK , POLH or POLL gene expression significantly decreased the levels of intracellular 3 . 5kb viral RNA as well as secreted HBeAg , POLB and POLD2 silencing only led to a modest decrease of intracellular 3 . 5kb viral RNA production . Of note , no significant cytotoxic effect was observed upon silencing any of the host DNA polymerases ( S4 Fig ) . To confirm the specificity of the siRNA in the targeted RNAi screen , we evaluated the efficacy of single and combined dual siRNA-mediated knockdown of POLK , POLL and POLH genes ( S5 Fig and Fig 4 ) , respectively . Western blot analysis or qRT-PCR assay showed that each of the siRNAs specifically reduced the expression of its targeted DNA polymerase gene ( S5A Fig and Fig 4 ) . Among the three cellular polymerases , silencing of POLK exhibited the most dramatic inhibitory effect on HBV infection , with approximately 74% reduction of HBV infection as judged by HBeAg level , which was comparable to the efficiency of silencing NTCP expression that reduced HBV infection by 80% ( S5C Fig ) . Similar extents of reduction were also observed for intracellular 3 . 5kb vRNA ( S5B Fig ) and HBcAg levels ( S5D Fig ) in POLK siRNA transfected cultures . Knock-down of POLL gene also led to a notable decrease of intracellular 3 . 5kb viral RNA and HBcAg . Interestingly , at the same total concentrations of siRNAs , compared with only knock-down POLK gene ( by siRNAs of NC and POLK ) , dual siRNAs-mediated knockdown of POLK and POLL , or , POLK and POLH demonstrated enhanced inhibition of HBV infection , but did not completely abolish HBV infection ( S5E–S5G Fig ) . These results thus suggest that while POLK , POLL and POLH each individually could contribute to de novo HBV infection , POLK clearly plays a critical role . POLK belongs to the Y family of DNA polymerases , which functions in translesion synthesis and nucleotide excision DNA repair . Its enzymatic activity is resistant to aphidicolin ( APH ) and dideoxynucleotides [37] . In line with this , APH treatment did not inhibit HBV cccDNA synthesis and secretion of HBeAg in HBV infection of HepG2-NTCP cells ( S6 Fig ) . The results also imply that APH-sensitive DNA polymerases ( e . g . POLA , POLE ) are not required for HBV infection of HepG2-NTCP cells , which is consistent with the results obtained from siRNA knockdown experiments . To confirm the role of POLK in HBV cccDNA formation , two individual siRNAs targeting different region of POLK gene were transiently transfected into HepG2-NTCP cells . The siRNAs significantly reduced the expression levels of endogenous POLK at day 3 and 5 post transfection ( Fig 4A ) . We found that POLK knockdown led to 70% reduction of cccDNA levels compared to that in control siRNA transfected cells on day 7 post infection ( Fig 4B ) . In line with this observation , the 3 . 5kb vRNA level was also reduced by approximately 60% in POLK knockdown cells ( Fig 4C ) . HBeAg and HBsAg levels in culture supernatants were reduced to less than 30% of that from control siRNA transfected cells ( Fig 4D ) . Immunostaining of intracellular HBcAg also showed significant decrease in POLK knockdown cultures ( Fig 4E ) . To investigate whether POLK plays a similar role in other in vitro HBV infection experimental models , we conducted siRNA-mediated knockdown of POLK in HepaRG cells ( Fig 4F–4H ) and primary tupaia hepatocytes ( PTH ) ( Fig 4I and 4J ) , respectively . Compared to a negative control ( NC ) , differentiated HepaRG cells treated with two individual siRNAs targeting POLK reduced the expression level of POLK , decreased HBV cccDNA synthesis and reduced secretion of HBeAg and HBsAg . No cytotoxic effect was observed at 5 days post siRNA transfection ( Fig 4G ) . Consistent with the results for HBV infected HepG2-NTCP and HepaRG cells , silencing of tupaia POLK by siRNA ( sitsPOLK ) led to a marked decrease of cccDNA synthesis and production of HBeAg and HBsAg on day 7 post infection of PTH . Moreover , similar to infection by wild-type HBV , transfection of sitsPOLK also significantly reduced the levels of intracellular HBV cccDNA and HBsAg in HBV-ΔHBc infected PTHs ( Fig 4J ) . Importantly , siRNA knockdown of POLK did not reduce the infection efficiency of HDV ( S7A Fig ) or an EGFP-encoding VSV-G pseudotyped lentivirus ( VSV-EGFP ) ( S7B Fig ) , demonstrating that POLK has a specific role in HBV infection . Together , these data suggest that POLK is required for de novo HBV infection and depletion of POLK diminishes HBV cccDNA synthesis . In order to further confirm that POLK is responsible for formation of HBV cccDNA , we intended to restore POLK expression in the siRNA transfected cells . To achieve this goal , an expression plasmid of POLK fused with GFP at N-terminus ( GFP-POLK-wt ) was constructed . Silent mutations were introduced to siPOLK1-targeting sequence for expression of the siRNA-resistant POLK mRNA ( GFP-POLK-res ) . HepG2-NTCP cells were transduced with VSV-G protein pseudotyped lentiviruses expressing GFP-POLK-res , GFP-POLK-wt or GFP alone , respectively . The cells were then transfected with siPOLK-1 . Fluorescence microscopic analysis showed that POLK localized in the nuclei , and siPOLK-1 dramatically reduced the expression of GFP-POLK-wt , but not GFP-POLK-res , suggesting that the expression of GFP-POLK-res is indeed resistant to siPOLK-1 ( S8A Fig ) . We next performed HBV infection assay . As shown in S8B–S8D Fig , while siPOLK-1 transfection efficiently reduced HBV cccDNA formation as well as 3 . 5 kb vRNA expression and HBeAg secretion in HepG2-NTCP cells transduced with control lentivirus expressing GFP , the effects of siPOLK-1 transfection on cccDNA formation and function in HepG2-NTCP cells expressing POLK , in particular GFP-POLK-res were significantly attenuated . The results thus suggest that ectopic expression of POLK partially rescued the suppression of HBV cccDNA formation caused by POLK-targeting siRNA . To rigorously determine the function of POLK in HBV cccDNA synthesis , we took advantage of the CRISPR/Cas9 system to generate POLK knockout in HepG2-NTCP cells ( S9 Fig ) . We first created a stable HepG2-NTCP/Cas9 cell line that constitutively expresses Cas9 . The cell line was then infected with lentivirus encoding an EGFP protein and sgRNA targeting exon 2 of polk gene . By monitoring the expression of EGFP , we could assess the sgRNA transduction efficiency . On day 3 post-transduction , EGFP positive cells were sorted and expanded by culturing for additional 10 days . Independent HepG2-NTCP clones with successful polk gene editing ( HepG2-NTCPpolk+/-and HepG2-NTCPpolk-/- ) were identified and used for further studies . Sequencing analysis revealed that the clones have a frame shift in the coding region owing to nucleotide deletions , which resulted in the disruption of intact POLK protein expression . Western blotting analysis confirmed that the expression of POLK protein was reduced in HepG2-NTCPpolk+/- and abolished in HepG2-NTCPpolk-/- clones ( Fig 5A ) . Knockout POLK in HepG2-NTCP cells did not affect the viability of HepG2-NTCP cells . NTCP level remained unchanged as compared to that in parental HepG2-NTCP cells ( S10A Fig ) . Consistently , a functional assay showed that HepG2-NTCPpolk-/- cells were able to uptake [3H]-labeled taurocholate ( S10B Fig ) and supported HDV infection ( S10C Fig ) at an efficiency similar to that of the parental HepG2-NTCP cells . We next carried out HBV infection assay with these stable cell lines described above . The levels of cccDNA were examined by Southern blotting analysis at day 7 post-infection . No visible band of cccDNA could be detected in HepG2-NTCPpolk-/- cells , indicating that lack of POLK impaired HBV cccDNA synthesis ( Fig 5B ) . Quantitative PCR also showed that the amounts of cccDNA decreased by approximately 3- or 4-fold in the cells with reduced ( HepG2-NTCPpolk+/- ) and abolished ( HepG2-NTCPpolk-/- ) POLK expression , respectively ( Fig 5C ) . ELISA analysis showed that secreted HBeAg and HBsAg levels were markedly decreased from HepG2-NTCPpolk-/- cells ( Fig 5E ) . Similarly , depletion of POLK also dramatically reduced intracellular HBcAg , which was closely related to intracellular 3 . 5kb vRNA levels ( Fig 5D and 5F ) . Importantly , a time course analysis using Southern blot assay showed that cccDNA was readily detectable in parental HepG2-NTCP cells at day 2 post infection and modestly increased in the next 4 days . In contrast , cccDNA only became detectable in HepG2-NTCPpolk+/ -cells containing one intact polk allele at day 4 and day 6 post infection in much reduced amounts , compared to that in the parental HepG2-NTCP cells at the same time points . Only very faint cccDNA bands could be detected in HepG2-NTCPpolk-/- cells with both polk alleles disrupted during the same time period ( Fig 5G ) . In agreement with the cccDNA formation results , POLK knockout cells produced low level HBeAg and HBsAg . To confirm the function of POLK in HBV cccDNA synthesis , we restored POLK expression in HepG2-NTCPpolk-/-cells by stable transduction of lentivirus expressing POLK . HepG2-NTCPpolk-/- cell clones with defective endogenous polk but expressing exogenous POLK were established . Production of POLK protein was largely restored in two independent cell clones as demonstrated by Western blot analysis ( Fig 6A ) . The amount of cccDNA upon HBV infection was assessed by Southern blot . Remarkably , HBV infection in these two clones was rescued , and the cccDNA level correlated with the expression levels of POLK in these cells ( Fig 6B ) . In contrast , transduction of HepG2-NTCPpolk-/- cells with a control lentiviral vector did not rescue cccDNA synthesis . Consistently , ELISA analysis for HBeAg in culture supernatant and immunofluorescence staining of intracellular HBcAg also confirmed that restoration of POLK expression in the HepG2-NTCPpolk-/- cells efficiently rescued not only cccDNA formation , but also viral gene transcription and protein expression ( Fig 6C and 6D ) . Considering that knockout of POLK did not completely abolish HBV infection and cccDNA formation , and knockdown of POLL and POLH with siRNA also reduced HBV infection despite at a lesser extent , we further investigated the role of POLL in HBV infection with more rigorous experimental conditions . We accordingly generated POLL knockout cell lines with CRISPR/Cas9 technology . Infection of poll gene edited HepG2-NTCPpoll+/- and HepG2-NTCPpoll-/- cell clones ( Fig 7A ) with HBV demonstrated reduced levels of intracellular cccDNA ( Fig 7B and 7C ) , 3 . 5kb vRNA ( Fig 7D ) and HBcAg ( Fig 7F ) as well as secreted HBeAg ( Fig 7E ) at day 7 post-infection , as compared to that in the parental HepG2-NTCP cells . However , the extent of POLL depletion on cccDNA formation and viral RNA and protein expression was less than that of POLK depletion . Taken together , our results strongly suggest that while POLK , POLL and POLH are each capable of supporting cccDNA synthesis at a different efficiency during a de novo HBV infection , POLK plays a more dominant role under the infection conditions examined in this study .
We demonstrated in this study that cccDNA is formed from incoming virion DNA in HepG2-NTCP cells at as early as 24 h post infection and establishes the pool size of approximately 3 copies of cccDNA per infected cell within a few days of infection . The kinetics of cccDNA accumulation as well as two lines of independent evidence obtained from HBV-ΔHBc infection of HepG2-NTCP cells and viral DNA polymerase inhibitor treatment of wild-type HBV-infected cells strongly support the notion that intracellular amplification does not play a significant role in the establishment of cccDNA pool in the HBV-infected hepatoma cells . This observation is consistent with the findings from HBV infection of primary human hepatocytes and HepaRG cells [38 , 39] , but distinct from DHBV infection of primary duck hepatocytes where significant intracellular cccDNA amplification occurs in a manner regulated by the level of its large envelope protein expression [19 , 20] . However , intracellular amplification of cccDNA has been observed in HepG2-derived cell lines supporting constitutive or inducible HBV replication [40–42] . Developing therapeutics against chronic HBV infection requires better understanding the contribution of intracellular cccDNA amplification in the maintenance of persistent infection , and further investigation of the activity and regulation of this pathway in HBV-infected hepatocytes in vivo is thus warranted . An elegant study by Chisari and colleagues demonstrated that although DHBV deficient for capsid protein expression ( DHBVΔcp ) infected primary duck hepatocytes and produced similar amounts of cccDNA from the incoming virions as did wild-type DHBV , the cccDNA in DHBVΔcp-infected hepatocytes was significantly less efficiently transcribed into viral RNAs , suggesting an important role of capsid protein in DHBV cccDNA transcription [43] . However , our results showed that HBV-ΔHBc infected HepG2-NTCP cells and expressed viral genes at a similar efficiency as wild-type HBV did . The results therefore suggest that the synthesis of HBV capsid protein may not significantly modify HBV cccDNA transcription activity . Interestingly , DHBV capsid protein is structurally distinct from the capsid proteins of mammalian hepadnaviruses [44] and may have HBx-like function in regulation of DHBV cccDNA transcription . However , the possibility that HBV capsid proteins from in-coming virions interact with cccDNA and promote its transcriptional activity cannot be completely ruled out . It has been shown that HBV capsid protein is a structural component of viral cccDNA minichromosome and its binding reduces the nucleosomal spacing of the minichromosome [45] . In addition , it has also been suggested that capsid protein promotes an epigenetic permissive state of HBV cccDNA by binding on CpG islands of cccDNA [46] . Of note , some non-nucleoside analogue compounds targeting capsid protein can dysregulate functional HBV capsid assembly [47–52] . Those capsid assembly effectors may alter the amounts and/or structure of cccDNA-bound capsid protein and consequentially interfere with cccDNA metabolism and function [53] . Intriguingly , interferon-stimulated gene ( ISG ) APOBEC3A seems to have a role in the destruction of cccDNA by direct interaction with HBV core protein [54] . Further investigation on the differential roles of capsid proteins in regulation of cccDNA function should shed light on this aspect of hepadnaviral pathobiology . We showed herein that , similarly to DHBV and WHV , completion of plus strand DNA synthesis during de novo HBV infection of HepG2-NTCP cells is not sensitive to viral DNA polymerase inhibitors , suggesting the reaction is most likely catalyzed by a host DNA polymerase . In support of this hypothesis , by following the fate of viral DNA sequence during conversion of rcDNA into cccDNA , Seeger and colleagues demonstrated that independent of a viral enzymatic activity , a cellular DNA polymerase may fill in the 3’ end of both DNA strands [55] . The observed slight reduction of cccDNA amounts in ADV or ETV treated cells in this study could indicate either a minor contribution of viral DNA polymerase to cccDNA formation or an off-target inhibition of cellular functions required for cccDNA formation . Interestingly , studies of HBV cccDNA biosynthesis via intracellular amplification pathway in HepG2-derived stable cell lines , such as HepAD38 or HepDES19 cells , suggested that deproteinization and uncoating of progeny rcDNA require the completion of plus strand DNA synthesis , which requires viral DNA polymerase activity [41 , 42] . Hence , different from rcDNA in virion particles with various length of incompletely synthesized plus-stranded DNA , the precursor rcDNA for cccDNA synthesis from the intracellular amplification pathway may have a very short gap in plus strand DNA and thus distinct DNA repair enzymes may be recruited to convert the rcDNA to cccDNA . Moreover , it had been shown that a small fraction of cccDNA can be formed from dslDNA via NHEJ DNA repair pathway . The cellular DNA polymerases required for cccDNA synthesis through intracellular amplification pathway and from dslDNA remain to be determined in future studies . While our results demonstrated that POLK plays a critical role in cccDNA formation during de novo HBV infection of cultured HepG2-NTCP , HepaRG and PTHs , we also showed that POLL and POLH play a role in cccDNA formation , albeit at a lesser extent . It is currently not clear whether each of the cellular DNA polymerases plays a redundant or distinct role in de novo cccDNA synthesis . As mentioned above , plus-strand DNA in rcDNA in virions has gaps of heterogeneous lengths . It is possible that depending on the length of gaps , distinct DNA repair complexes containing different repairing DNA polymerases are recruited to fill the gaps with different length . If this is the case , the three DNA polymerases may play non-redundant roles and be involved in conversion of distinct rcDNA precursors into cccDNA . Alternatively , each of the three cellular DNA polymerases may participate in a different DNA repair complex to fill the plus strand gaps , irrespective of their length , but in a different efficiency . These two possibilities will need to be further investigated . POLK plays a functional role in nucleotide excision repair ( NER ) pathway by filling the gap produced upon excision of damaged nucleotides [56 , 57] . The activity of POLK is partially dependent on the growth state of the cells , and reaches maximum activity under conditions of low deoxynucleotide concentration such as in non-dividing cells [56] . A previous study showed treatment of HBV-transfected HepG2 cells with aphidicolin arrested cells in the G1 phase could result in enhancement of cccDNA synthesis [58] . Consistent with this observation , we found that the efficiency of HBV infection closely correlates with the number of G0/G1 phase cells in HepG2-NTCP cultures . Therefore , HBV cccDNA formation may preferentially occur at G0/G1 phase of cell cycle , supporting the notion that HBV infects non-dividing cells , so that cccDNA is formed and stably exists in quiescent hepatocytes [59] . It is thus conceivable that cell cycle-dependent factor ( s ) or protein post translational modification affecting the physiologic state of hepatocytes may regulate the formation of HBV cccDNA . Moreover , because cellular DNA polymerases must work in concert with other DNA repair proteins to restore the structure of damaged DNA , other DNA repair proteins in NER pathway may also play a role in HBV cccDNA formation . For example , it is possible that HBV hijacks cellular endonuclease ( e . g . XPG ) or exonuclease ( e . g . Exo1 ) to cleave the capped RNA primer to leave a free 5’ end of plus strand DNA of rcDNA , and followed by POLK or other cellular DNA polymerase , such as POLL and POLH , to fill the gap using minus strand DNA as a template . Additionally , POLK has been shown to work together with POLD to fill in single stranded DNA gaps [56 , 60] and XRCC1-Lig3 is required for ligation of NER-induced breaks in quiescent cells [61]; hence it will be interesting to test whether those host enzymes are involved in cccDNA formation . In conclusion , taking advantage of highly efficient genetic manipulation of HepG2-NTCP HBV infection system , and in combination with studies using recombinant HBV virus and chemical inhibitors , we rigorously demonstrated that cellular DNA polymerase κ substantially contributes to HBV cccDNA formation in HepG2-NTCP cells . Our findings shed new light on the molecular mechanism of cccDNA formation and may facilitate the development of novel therapeutics to cure chronic hepatitis B .
Human embryonic kidney cell 293T and human hepatoblastoma cell HepG2 were obtained from American Type Culture Collection ( ATCC ) . Human hepatocellular carcinoma cell ( Huh-7 ) was obtained from the Cell Bank of Type Culture Collection , Chinese Academy of Sciences . All the cell lines were maintained in Dulbecco’s Modification of Eagle’s Medium ( DMEM; Life technologies ) supplemented with 10% fetal bovine serum , 100 U/ml penicillin , and 100 μg/ml streptomycin at 37°C in a 5% CO2 incubator unless otherwise indicated . All experiments with HepG2 cells were carried out with cells grown on collagen-coated plates . HepaRG cells were obtained from Biopredic International ( Rennes , France ) and cultured according to the product manual . Differentiated HepaRG cells were obtained following a two-step procedure as previously described [38] . Primary Tupaia hepatocytes ( PTHs ) were obtained and cultured following a method as described previously [11] . Antibodies for human POLK ( A-9 ) and C9 tag ( 1D4 ) were obtained from Santa Cruz Biotechnology . Antibody for human POLL ( EPR7519 ( 2 ) ) was purchased from Abcam . Antibody against GAPDH , HRP-conjugated anti-mouse IgG and HRP-conjugated anti-rabbit IgG were purchased from Sigma . Alexa Fluor 488- and 546-conjugated anti-mouse IgG antibodies were purchased from Life Technologies . Other mouse antibodies to detect the Core protein ( 1C10 ) and surface antigen ( 17B9 ) of HBV , HDV delta antigen ( 4G5 ) were described previously [11 , 62] . N-terminally myristoylated peptide of the N-terminal 47 amino acid residues of pre-S1 domain of HBV strain S472 ( Accession number: EU554535 . 1 ) , myr-47 , was synthesized by SunLight peptides Inc . . Adefovir ( ADV ) , entecavir ( ETV ) , aphidicolin ( APH ) and other general chemicals were purchased Sigma unless otherwise stated . To construct human POLK expression plasmids , a full-length human POLK cDNA was cloned from HepG2 mRNA into a modified pLKO . 1-puro lentiviral vector under the control of a CMV promoter for stable expression in POLK knockout HepG2-NTCP cells . An N-terminal GFP-tagged POLK expression plasmid ( GFP-POLK-wt ) was constructed for investigating POLK subcellular localization and RNAi rescue experiment . For the single siRNA-mediated gene knockdown experiments , siRNA transfection was performed using Lipofectamine RNAiMAX ( Life technologies ) according to the manufacturer’s instruction . HepG2-NTCP cells in 48-well plates were transfected with 5 pmol siRNA and 0 . 5 μl Lipofectamine RNAiMAX in 25 μl Opti-MEM and were challenged with the HBV ( or HDV , VSV as indicated ) at 72 h post transfection . For combined dual RNAi experiments , HepG2-NTCP cells in 48-well plates were transfected with individual siRNAs for indicated two target genes ( each 2 . 5 pmol , total 5pmol ) and 0 . 5 μl Lipofectamine RNAiMAX in 25 μl Opti-MEM and were challenged with the HBV ( or HDV , VSV as indicated ) at 72 h post transfection . The siRNA target sequences used in this study are shown in S1 Table , which have been confirmed to be functional in other previous studies [36] . siRNAs targeting Tupaia polk gene were designed by RNAi designer ( https://rnaidesigner . thermofisher . com/rnaiexpress/ ) . The specificities of siRNAs were confirmed using a BLAST search . The efficiency of gene knockdown was determined by qPCR or Western blot assays . The qPCR primers used for quantification of human POLK , POLH and GAPDH mRNA expression are listed below . Human POLK: qPOLK-F: 5'-CCAGACATCACAACCATTCC and qPOLK-R: 5'-TCAAGGCTTCCAGACTGATG; human POLH: qPOLH-F: 5'-GTGCCAGTTACCAGCTCAGA and qPOLH-R: 5'-AGGTAATGAGGGCTTGGATG; GAPDH: 5’-GAAGGTGAAGGTCGGAGTCA ( forward ) and 5’-TGGAATCATATTGGAACATGT ( reverse ) . Both human POLK and POLH mRNA levels are normalized to the expression level of GAPDH , respectively . For Tupaia POLK mRNA measurements , qtsPOLK-F: 5'-TCACTAGCCAGCAGCTAAGGAAAGC and qtsPOLK-R: 5'-CATGCTCATTGATCCTACAGCAATG were used as qPCR primers . 5’-GTGAAGGTCGGAGTAAACG ( forward ) and 5’-CCATGGGTGGAGTCATACT ( reverse ) were used for Tupaia GAPDH , the mRNA expression level of tsPOLK is normalized by tsGAPDH . All siRNA oligos were synthesized by NIBS biological resources facility . All transfections were conducted in duplicates . Cytotoxic effects of siRNA were examined using alamarBlue reagents ( Life technologies ) . HBV genotype D virus and HDV genotype I virus were produced by transient transfection of Huh-7 cells with the corresponding plasmids as described previously [11] . HBc protein deficient virus ( HBV-ΔHBc ) was generated by co-transfection of Huh-7 cells with a plasmid harboring 1 . 05 copies of HBV genome with a stop codon at the 38th codon ( Y ) of core gene open reading frame , and an intact HBc protein expression vector . Virus stocks were aliquoted and stored at -80°C . HBV and HDV infection assays have been described previously [11] . Briefly , HepG2-NTCP cells were firstly cultured in 48-well plates with DMEM complete medium for 3–4 h , then the cells were cultured with PMM medium for another 20 hrs . The cells were then infected with a multiplicity of 100 genome equivalents ( mge ) of HBV or 500 mge of HDV in the presence of 4% PEG8000 for 24 hrs at 37°C . Cells were maintained in PMM with regular medium changing every other day . For viral inhibition assay , 100 nM myr-47 lipopeptide was pre-incubated with HBV viruses before adding to HepG2-NTCP cells; chemicals were pre-incubated with HepG2-NTCP cells at 37°C for 12 h before virus inoculation . To generate VSV-G protein pseudotyped lentiviral particles expressing EGFP , PLKO . 3G plasmid together with pSPAX2 and pMD2G were co-transfected into 293T cells at the ratio of 4:3:1 ( PLKO . 3G:pSPAX2:pMD2G ) using Lipofectamine 2000 ( Life technologies ) . Forty-eight hours after transfection , supernatants were harvested , cleared by centrifugation and stored at -80°C . Human POLK or gRNA expressing lentivirus was generated using a similar protocol as that described above but replacing the pLKO . 3G plasmid with the pLKO . 1-CMV-POLK-puro or pLKO . 1-gRNA-EGFP plasmid , respectively . Infection of cells with lentiviral pseudovirus was performed as described previously [63] . Virus-related experiments were conducted in a BSL-2 facility at the National Institute of Biological Sciences , Beijing . HBV viral particles were purified using Nycodenz gradient ultracentrifugation of the culture supernatants of Huh-7 cells transfected with plasmid for HBV production . The HBV DNA levels of the virus fractions were quantified using specific primers: 5’-GAGTGTGGATTCGCACTCC ( forward ) and 5’-GAGGCGAGGGAGTTCTTCT ( reverse ) by real-time PCR using SYBR Premix Ex Taq kit ( TaKaRa ) on an ABI 7500 Fast Real-Time system instrument ( Applied Biosystems , United States ) . The viral genome equivalent copies were calculated based on a standard curve generated with known copy numbers . For analysis of HBV 3 . 5kb viral RNA , total RNA from infected cells was extracted by TRIzol reagent ( Life technologies ) , 0 . 5 μg of total RNA was digested with DNase I ( Life technologies ) and reverse-transcribed into cDNA using PrimerScript RT Reagent Kit ( TaKaRa ) in a 20 μl reaction . Real time-PCR analysis was performed using the SYBR Premix Ex Taq on ABI 7500 Fast Real-Time PCR System . HBV 3 . 5kb viral RNA copy numbers were deduced from a standard curve generated from known nucleic acid quantities . The mRNA level of HBV 3 . 5kb viral RNA was normalized to that of GAPDH mRNA . The primers used for each gene examined are listed below . HBV 3 . 5kb viral RNA: 5'-GAGTGTGGATTCGCACTCC ( forward ) and 5'-GAGGCGAGGGAGTTCTTCT ( reverse ) ; GAPDH mRNA: 5’-GAAGGTGAAGGTCGGAGTCA ( forward ) and 5’-TGGAATCATATTGGAACATGT ( reverse ) . For quantification of HBV cccDNA , infected cells were lysed in a lysis buffer ( 20 mM Tris , 0 . 4 M NaCl , 5 mM EDTA , 1% SDS , pH = 8 . 0 ) in the presence of proteinase K ( QIAGEN ) , total DNA was extracted according to a standard phenol-chloroform extraction protocol . 500 ng of total DNA was digested with 0 . 5 μl plasmid-safe adenosine triphosphate ( ATP ) -dependent deoxyribonuclease DNase ( PSAD ) ( Epicentre Technologies ) in 25 μl reaction for 8 h at 37°C to allow removal of linear genomic DNA and HBV replication intermediates ( rcDNAs , single-stand DNAs , linear double-strand DNAs ) . DNase was inactivated by incubating the reactions for 30 min at 70°C . 20 ng of digested DNA was used for quantification of HBV cccDNA , 5'-TGCACTTCGCTTCACCT ( forward ) and 5'-AGGGGCATTTGGTGGTC ( reverse ) were used as HBV cccDNA specific primers , the real-time PCR was performed using the SYBR Premix Ex Taq on ABI 7500 Fast Real-Time PCR System as the following reaction procedure: 95°C for 5 min then 45 cycles of 95°C for 30 s , 62°C for 25 s , and 72°C for 30 s . The amount of HBV cccDNA in a DNA preparation was determined by real-time PCR using a plasmid containing HBV-D genome as the standard . The pool size of HBV cccDNA per infected cell was calculated by quantification of cccDNA copies using digital PCR in the whole cell population and estimation of the number of infected cells by immunostaining of intracellular HBcAg . Selective extraction of HBV cccDNA from HBV infected cells was achieved by a modified Hirt method as previously described [64 , 65] . Briefly , infected cells from one well of 6-well plates were lysed in Hirt lysis buffer ( 10 mM Tris-HCl , 10 mM EDTA , 0 . 6% SDS , pH = 7 . 4 ) for 30 min at room temperature . After adding 5 M NaCl , the cell lysate was vigorously mixed and incubated at 4°C overnight . After centrifugation at 10 , 000 rpm for 30 min at 4°C , the supernatant was extracted twice with saturated Tris-phenol ( pH = 8 . 0 ) and once with phenol:chloroform . The extracted DNA was precipitated with equal volumes of isopropanol at -20°C overnight . The DNA pellet was washed with 70% ethanol and dissolved in TE buffer ( 10 mM Tris-HCl , 1mM EDTA , pH = 8 . 0 ) , and digested with HindIII or EcoRI restriction enzyme ( NEB ) before being analyzed . For detection of cccDNA by Southern blot , the extracted HBV cccDNA sample was subjected to 1 . 2% agarose gel electrophoresis and transferred onto Amersham Hybond-N+ membrane ( GE Healthcare ) . The Hybond-N+ membrane was crosslinked in a UV crosslinker chamber with UV energy dosage at 1200 mJ and followed by being probed with [α-32P]dCTP ( 250 μCi , Perkin Elmer ) -labeled HBV genotype D ( Accession number: U95551 . 1 ) linear full-length genomic DNA . Hybridization was performed in Perfecthyb plus hybridization buffer ( Sigma ) with prehybridization for 1 h at 65°C and overnight hybridization at 65°C , followed by two washes in wash buffer ( 0 . 1×SSC , 0 . 1% SDS ) at 65°C . The membrane was exposed to Carestream X-OMAT BT Film ( XBT-1 , Carestream ) . 100 pg each of 3 . 2kb , 2 . 1kb and 1 . 7kb HBV DNA fragments prepared by PCR amplification of a plasmid containing 1 . 0 copies linear HBV genotype D genome was used as DNA marker . HBeAg and HBsAg from supernatants of HBV infected cells were measured using ELISA kits ( Wantai Pharm Inc . Beijing , China ) by following the manufacturer’s instructions . Supernatants from HBV infected cells were harvested at each time point examined in the various assays and were diluted 2-fold with PMM before ELISA . All experiments were performed in duplicates and repeated at least two times independently . Virus infected cells in 48-well plates were washed three times with pre-cooled PBS and fixed by 4% paraformaldehyde for 10 min , followed by permeablization for 10 min at room temperature with 0 . 5% Triton X-100 . After incubation for 1 h with 3% BSA for blockade of nonspecific binding , primary antibodies were added for incubation for 1 h at 37°C . The bound antibodies were visualized by incubation with secondary antibodies ( Alexa Fluor 488 donkey anti-mouse IgG or Alexa Fluor 546 anti-mouse IgG ) . Images were acquired using a Nikon A1-R confocal microscope or a Nikon Eclipse Ti Fluorescence Microscopy . The POLK rescue cell lines were generated by infection of POLK stable knockout cells with recombinant lentivirus expressing POLK , and cells were selected with puromycin ( Sigma ) . Expression of POLK in the established stable cell lines were verified by Western blot assay . To generate POLK deficient HepG2-NTCP cells , genomic engineering of polk gene was achieved with the CRISPR/Cas9 system as described with the following single-guide ( sg ) RNA target sequences: CTTCTCCTTTGTGCTATCCA ( sgPOLK-1 ) , GATGATCTTCTGCTTAGGAT ( sgPOLK-2 ) . Firstly , stably expressing Cas9 cell line ( HepG2-NTCP/Cas9 ) was generated by transfection of human codon-optimized Cas9 ( hCas9 ) expressing vector ( Addgene ) using Lipofectamine 2000 and selection with Blasticidin ( Calbiochem ) . Then the cells were infected with gRNA expressing lentivirus . After culture for 3 d in vitro , EGFP positive cells were sorted by flow cytometry FACSAria II , and further analyzed by T7 endonuclease I ( NEB ) assay . Genomic DNA sequence of POLK around the gRNA targeting site was amplified using the following primers: cas9-POLK-F: 5'-GTGTCGAACCCCTGAGCTCAGTCAATCT and cas9-POLK-R: 5'-AGGTGAACAGGAACATATACATTATTT . Single clones of sorted cells were obtained by serial dilutions and amplified , verified by sequencing of PCR fragments , and confirmed by Western blot using anti-POLK antibody . POLL deficient HepG2-NTCP cell lines were constructed by following the above mentioned procedure , a 20-bp single-guide sequence ( sgPOLL: CGGGCCCATGTTGTGCGCAC ) targeting DNA within the third exon of poll gene was selected . cas9-POLL-F: 5’-GCTATATGTAGAAGGAAAGCTGTC and cas9-POLL-R: 5’-ACTGGGATCAGCCCACCTACTGG were used as primers for amplification of a region around the gRNA targeting site and the PCR products were further analyzed by T7E1 assay . Individual clones were validated by sequencing of PCR fragments , and confirmed by Western blot using anti-POLL antibody . | HBV chronically infects 240 million people worldwide . Persistent HBV infection relies on stable maintenance of the nuclear form of viral genome , the covalently closed circular ( ccc ) DNA . However , the molecular mechanism underlying the conversion of HBV genomic relaxed circular ( rc ) DNA into cccDNA remains elusive . Our studies reported herein provide unambiguous evidence suggesting that host DNA polymerase κ ( POLK ) is required for repairing the gap of rcDNA and formation of cccDNA in a de novo HBV infection . POLK is thus a potential therapeutic target for treatment of chronic hepatitis B . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"transfection",
"medicine",
"and",
"health",
"sciences",
"liver",
"enzyme-linked",
"immunoassays",
"nuclear",
"staining",
"pathology",
"and",
"laboratory",
"medicine",
"molecular",
"probe",
"techniques",
"gene",
"regulation",
"pathogens",
"dna-binding",
"proteins",
"micro... | 2016 | DNA Polymerase κ Is a Key Cellular Factor for the Formation of Covalently Closed Circular DNA of Hepatitis B Virus |
Most epidemiological and clinical reports on snake envenoming focus on a single country and describe rural communities as being at greatest risk . Reports linking snakebite vulnerability to socioeconomic status are usually limited to anecdotal statements . The few reports with a global perspective have identified the tropical regions of Asia and Africa as suffering the highest levels of snakebite-induced mortality . Our analysis examined the association between globally available data on snakebite-induced mortality and socioeconomic indicators of poverty . We acquired data on ( i ) the Human Development Index , ( ii ) the Per Capita Government Expenditure on Health , ( iii ) the Percentage Labour Force in Agriculture and ( iv ) Gross Domestic Product Per Capita from publicly available databases on the 138 countries for which snakebite-induced mortality rates have recently been estimated . The socioeconomic datasets were then plotted against the snakebite-induced mortality estimates ( where both datasets were available ) and the relationship determined . Each analysis illustrated a strong association between snakebite-induced mortality and poverty . This study , the first of its kind , unequivocally demonstrates that snake envenoming is a disease of the poor . The negative association between snakebite deaths and government expenditure on health confirms that the burden of mortality is highest in those countries least able to deal with the considerable financial cost of snakebite .
Our knowledge of the global medical burden of snakebite is limited to just a few reports based primarily on either hospital records [1] or the epidemiological literature [2] , [3] , and more recently , the latter in combination with WHO mortality data [4] . Despite the nearly universal distribution of venomous snakes ( the South Pole , Greenland , New Zealand and Madagascar being the major exceptions ) , each report concludes that the medical importance of snakebite is greatest in the tropics . The vast majority of snakebite-induced deaths ( Figure 1 ) occur in Asia ( estimates ranging from 15 , 400–57 , 600 deaths pa ) and sub-saharan Africa ( 3 , 500–32 , 100 deaths pa ) [4] . Populations in this geographic zone also suffer the medical burden of the world's neglected tropical diseases ( NTD ) . Importantly , the number of snakebite-induced deaths doubles the NTD mortality figures for this region due to African trypanosomiasis , cholera , dengue haemorrhagic fever , leishmaniasis , Japanese encephalitis and schistosomiasis [5]–[7] . A major distinctive characteristic of the NTDs is that they are globally associated with poverty [8] . In line with the recent WHO categorisation of snake envenoming as a NTD [9] , this analysis was therefore conducted to determine whether the medical burden of snake envenoming is , like the other NTDs , also associated with poverty .
To gain a global perspective of the relationship between poverty and lethal snake envenoming , we entered the now readily available country-specific snakebite mortality data [Supplement 2 , [4]] into Epi-Info ( version3 . 5 . 1 ) software package [http://www . cdc . gov/epiinfo] to populate the global map of snakebite mortality by country ( Figure 1 ) ; this differs slightly from the map presented in the report by Kasturiratne et al [4] which presented the same data by geographic region . Entering this country-specific , rather than regional , data provided us with a sufficient spread of data to have confidence in the statistical validity of plotting the mortality data against appropriate socioeconomic indicators . This analysis was performed on the understanding that national estimates of snakebite mortality were not likely to be as accurate as that for the reported wider regions [4] , since many of the country estimates were necessarily ( because of the lack of available data ) extrapolations from neighbouring countries . From amongst the large variety of socioeconomic data available in the public domain , we acquired country-by-country data for ‘Gross Domestic Product Per Capita , US$’ and ‘The Percentage of the Labour Force in Agriculture’ from the CIA World Factbook database [https://www . cia . gov/library/publications/the-world-factbook/] . The data on ‘Per Capita Government Expenditure on Health , US$’ was taken from the World Health Organisation Statistical Information System ( WHOSIS ) database [http://apps . who . int/whosis/data/Search . jsp] . These three datasets were selected because of the potential link between snakebite mortality and an individual's high-risk agricultural occupation , income and access to healthcare . The Human Development Index ( HDI; a composite indicator that reflects life expectancy , education and literacy and standard of living measured by GDP ) was also examined and countries categorised into Low ( 0 . 1 to 0 . 499 ) , Medium ( 0 . 500 to 0 . 799 ) or High ( above 0 . 800 ) HDI status .
Statistically linear relationships were found between the logarithm of snakebite mortality and the HDI ( Figure 2a ) , the logarithm of Per Capita Government Expenditure on Health ( 2b ) , the Percentage of the Labour Force in Agriculture ( 2c ) and the logarithm of GDP Per Capita ( 2d ) . The strength of each of these relationships was determined using the Pearson correlation coefficient; all four correlations were numerically strong ( range: r = 0 . 571–0 . 651 ) and statistically highly significant ( p<0 . 001 ) .
The World Bank defined the absolute poverty line as the percentage of a country's population living on an income of less than US$2 per day [10] . Maps depicting countries based on this definition of poverty [11] show a remarkably similar profile as the snakebite mortality map ( Figure 1 ) . This relationship between poverty and snakebite mortality is clearly demonstrated by the strong negative correlation between snakebite mortality and both HDI and GDP/capita reported here ( Figure 2 ) . The literature on snake envenoming is , like global poverty , rich with references associating rural agriculture with high incidence of disease and death . Taking West Africa as a detailed example , farmers and children in rural communities are consistently identified as being the highest snakebite risk groups in Senegal [12] , the Gambia [13] , Mali [14] , Cote D-Ivoire [15] , Ghana [16] , Benin [17] , Niger [18] , Nigeria [19]–[21] , Cameroon [22] , Gabon [23] and the Congo [24] . The numerous epidemiological reports conducted in Asia and Latin America similarly emphasise that rural subsistent farming communities in these regions also suffer snakebite as a daily occupational hazard ( Figure 3 ) . It is not surprising therefore that figures for ‘The Percentage of the Labour Force in Agriculture’ are strongly correlated with global snakebite mortality ( Figure 2c ) . The survival of many of the rural poor is dependent upon their non-mechanised , low-cost farming techniques and it is a cruel irony that it is exactly these practices that place them at such high risk of snakebite , and that their feet , legs and hands are the most frequent anatomical sites of snakebite in Africa [25] , [26] , Asia [27] , [28] and Latin America [29] . The tissue necrotic effects of snake envenoming are thought to afflict many more survivors of snakebite than victims who succumb [30] . Detailed community-based DALY/QALY-type assessments of the true burden imposed by the tissue destructive effects of snake envenoming on these communities are urgently required . There are very few reports in the literature examining the socioeconomic impact of snakebite [31]–[33] and none that we could find concerning the long-term psychological effects of snakebite . It is important that these types of studies are undertaken and the results appropriately disseminated to ensure that governmental , non-governmental and international health agencies understand the medical and sociological consequences of snakebite and the implication for the strategic allocation of scarce health resources . As with some other NTDs , effective therapy for snake envenoming is available . Antivenom is immunoglobulin purified from the blood of venom-immunised horses and sheep ( and rarely other animals ) . While conceptually simple , the manufacturing process requires GMP-standard production plants and is reliant upon the husbandry and handling of both horses and venomous snakes . Antivenom is therefore relatively expensive ( eg , US&$100/vial in S Africa ) compared to many other medications used in the tropics . Presumably for this reason , in much of Latin America antivenom manufacture is government-subsidised and antivenom is usually provided free to the patient . This provision and delivery of effective therapy to the at-risk communities may be an important factor in explaining why although snakebite incidence in Latin America is high ( 129 , 000 ) , the mortality rates are low ( 1 . 78%; 2 , 300 deaths ) . Much of Latin America has a high Human Development Index ( Figure S1 ) . In contrast , Sub-Saharan African countries , where snakebite victims are charged commercial rates for antivenom ( when available ) , are at the opposite end of the HDI scale . The strong correlation between global ‘Per Capita Government Expenditure on Health’ and snakebite mortality ( Figure 2b ) , illustrates the tragedy that the countries with the highest burden of snakebite mortality are those with the most limited ability to purchase effective antivenom . There are clearly limitations to this kind of analysis: much of the mortality data is estimated and the indicators , although standard , are still only relatively crude indicators . The limitations are indicated by the anomaly of India , with the highest global snakebite mortality rate but medium Human Development ( Supplementary Figure 1 ) and Human Poverty [11] indices . This is likely to reflect factors such as the huge variation of income within India and might also be explained by complex considerations of population and venomous snake densities , antivenom effectiveness and clinical practices and guidelines [34] . Nevertheless , with only occasional exceptions , our approach is robust in demonstrating the relationship of snakebite to socioeconomic markers of poverty . The clinical effectiveness of most antivenoms means that snakebite is , in WHO parlance , a ‘tool-ready NTD disease’ . This is encouraging: unlike some NTDS , the principles of producing a therapy against snakebite are established , albeit with scope for considerable improvement . As detailed in the WHO Global Plan to Combat NTDs [8] , what is required now to resolve the problems of snake bite is a coordinated effort to ( i ) assess the medical burden ( to determine the scale and location of the therapeutic need ) , ( ii ) where possible , to integrate snakebite initiatives with those currently pursued for other NTDs , ( iii ) strengthen health care systems by appropriate capacity building initiatives , ( iv ) develop communication systems to disseminate ‘disease burden’ information to improve advocacy/public awareness , ( v ) improve access to affordable , effective treatment and ( vi ) establish a framework of implementation and evaluation . The scale of this undertaking is daunting but the recent recognition by the WHO that snakebite is a NTD [9] and the establishment of the Global Snakebite Initiative [7] by the International Society of Toxinology is evidence that that the process has already started and that there is enthusiasm for the task . One of the major hurdles will be to ensure the affordability and effectiveness of antivenom . For most NTDs , once the effectiveness of a treatment has been established , it can be utilised worldwide - which improves commercial economies of scale and encourages widespread pharmaceutical support and drug donation [8] . In contrast , the clinical and geographic effectiveness of antivenom is restricted to the species of snake whose venom was used in its manufacture . This severely limits the implementation of economies of scale and , by inference , the commercial incentives for the involvement of ‘large pharma’ . In some regions such as sub-Saharan Africa , until very recently there was only one source of effective antivenom and a crisis in antivenom supply to the continent [35] , [36] . Reports of this ‘therapeutic vacuum’ attracted the commercial influx of ineffective antivenoms manufactured from venoms from non-African snakes [37]–[41] . Therefore , despite the effectiveness of current antivenoms , there are compelling reasons to encourage research to broaden the geographic efficacy and improve the commercial viability of antivenom therapy . There are encouraging experimental developments in this area , including ‘antivenomics’ [42] and ‘epitope-string immunogen’ [43] , [44] approaches , whose objectives are to maximise the clinical and snake-species efficacy of snakebite serotherapy and minimise manufacturing costs . There are an estimated 20–94 , 000 snakebite deaths annually , predominantly occurring in the rural poor in the tropics . Improved antivenoms , implementation of the WHO-recommended strategies and above all , international recognition of the importance of the problem could help to reduce many of these deaths . | Every year snake envenoming kills more people in the tropics than some of the world's recognised neglected tropical diseases ( NTDs ) , including schistosomiasis and leishmaniasis . While lacking the epidemic potential of an infectious/vector-borne disease , snake envenoming in rural tropical communities has as great a medical mortality , if not morbidity , as the NTDs . The recent categorisation of snake envenoming as an NTD is an important advance that hopefully will result in the wider recognition and allocation of resources , particularly since death from snake envenoming is preventable; antivenom is very effective when the appropriate antivenom is correctly administered . Snake envenoming urgently requires international support to instigate the epidemiological , health education , and effective treatment initiatives that proved so potent in addressing the medical burden of NTDs such as leprosy and dracunculosis . All the global estimates of snake envenoming and deaths from snakebite indicate that mortality is highest in the world's tropical countries . Here we examined associations between the globally available data on ( i ) snakebite-induced mortality and ( ii ) socioeconomic markers of poverty . Our data unequivocally establishes that snake envenoming is globally associated with poverty , a distinctive characteristic of the neglected tropical diseases . | [
"Abstract",
"Introduction",
"Methodology",
"Results",
"Discussion"
] | [
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"infectious",
"diseases/tropical",
"and",
"travel-associated",
"diseases"
] | 2009 | Snake Envenoming: A Disease of Poverty |
Protein conformational changes and dynamic behavior are fundamental for such processes as catalysis , regulation , and substrate recognition . Although protein dynamics have been successfully explored in computer simulation , there is an intermediate-scale of motions that has proven difficult to simulate—the motion of individual segments or domains that move independently of the body the protein . Here , we introduce a molecular-dynamics perturbation method , the Rotamerically Induced Perturbation ( RIP ) , which can generate large , coherent motions of structural elements in picoseconds by applying large torsional perturbations to individual sidechains . Despite the large-scale motions , secondary structure elements remain intact without the need for applying backbone positional restraints . Owing to its computational efficiency , RIP can be applied to every residue in a protein , producing a global map of deformability . This map is remarkably sparse , with the dominant sites of deformation generally found on the protein surface . The global map can be used to identify loops and helices that are less tightly bound to the protein and thus are likely sites of dynamic modulation that may have important functional consequences . Additionally , they identify individual residues that have the potential to drive large-scale coherent conformational change . Applying RIP to two well-studied proteins , Dihdydrofolate Reductase and Triosephosphate Isomerase , which possess functionally-relevant mobile loops that fluctuate on the microsecond/millisecond timescale , the RIP deformation map identifies and recapitulates the flexibility of these elements . In contrast , the RIP deformation map of α-lytic protease , a kinetically stable protein , results in a map with no significant deformations . In the N-terminal domain of HSP90 , the RIP deformation map clearly identifies the ligand-binding lid as a highly flexible region capable of large conformational changes . In the Estrogen Receptor ligand-binding domain , the RIP deformation map is quite sparse except for one large conformational change involving Helix-12 , which is the structural element that allosterically links ligand binding to receptor activation . RIP analysis has the potential to discover sites of functional conformational changes and the linchpin residues critical in determining these conformational states .
Protein dynamics play a critical role in a wide variety of biological processes such as catalysis , substrate recognition and binding , allosteric regulation and protein stability [1] . The dynamic behavior associated with these biological functions can involve motions as subtle as a sidechain displacement to large-scale rearrangements of entire domains . The timescales and magnitudes of protein conformational dynamics have been revealed by NMR , small angle x-ray scattering , electron microscopy , and single molecule fluorescence [1] . While potentially quite powerful , molecular dynamics ( MD ) simulations that seek atomic level explanations and predictions of dynamic behavior are limited by the need to sample high energy , transiently populated states in a computationally practical time period . For example , to date the longest MD simulation spans a microsecond [2] , whereas biologically important dynamic behavior often occurs on the millisecond ( and longer ) timescale . Thus , despite the clear importance of protein dynamics to biological function , there is an equally clear need to improve the computational models that seek to populate infrequent and transient states . Given the time limitations of MD simulations , the feasibility of generating meaningful dynamic information depends critically on the size of the fluctuations . Although small motions such as the gating of the sodium channel [3] can be sufficiently sampled over hundreds of nanoseconds , biologically relevant large-scale conformational changes such as those involved in allosteric regulation [4] , molecular motors moving along their filamentous tracks and polymerases moving along DNA [5] , require timescales of milliseconds or longer . These timescales are usually inferred from kinetic data or more increasingly from direct observations and single molecule FRET [1] . Where the final state of a large-scale motion can be deduced from crystal structures , MD simulations biased by driving potentials along a pre-defined trajectory can be used to identify critical events along the trajectory [6] . However , such biased simulations cannot be used to predict conformational changes from a single crystal structure . In the absence of alternate conformations , various approximation schemes based on contact analysis , such as guassian network models [7] and FIRST [8] , have been devised to generate large domain-level motions of a given protein structure [9] , [10] . There is an important class of protein dynamics that lie between the regime of small fluctuations and large domain motions - motions confined to a single structural element moving independently of the rest of the protein , which cannot be readily modeled with contact-based models . Well-studied examples of these intermediate-scale motions show that they are functionally important: the ligand-binding loop on Triosephosphate Isomerase ( TIM ) fluctuates at a rate of 3×104 s−1 [11] where the closed state stabilizes the ligand for catalysis [12]; fluctuations ( 35 s−1 ) of the Met20 loop of Dihydrofolate Reductase ( DHFR ) [13] are postulated to be the limiting step of catalysis [14]; and structural rearrangements of Helix-12 of the nuclear hormone receptors , which interacts with bound ligand , is a key determinant of the receptor's allosteric activation [15] , [16] . The discovery and modeling of such movable segments are important in understanding the functional dynamics of these and other proteins . As experiments suggest that these motions occur in the microsecond/millisecond range , extraordinarily long MD simulations would be needed to allow the protein to explore the relevant rare fluctuations . To circumvent this practical limit in computation , previous simulations predefined interconversion pathways and applied driving potentials , resorted to high temperatures coupled with manually-chosen backbone constraints to maintain structural integrity [17] , or used coarse-grained representations [18] , [19] , [20] . Motions have also been deduced from normal-mode analysis or quasi-harmonic analysis of nanosecond MD trajectories [21]–[25] . Alternatively , hierarchical loop modeling can identify reasonably accurate low energy conformations of short loops [26]–[28] . As both high temperature simulations and hierarchical loop modeling require prior information about the existence of a flexible region , they are not suitable for prediction purposes . Nevertheless , several systems have been designed to model the local flexibility of a given structure [29]–[32] . Although the flexibility generated by these systems reproduce the NMR S2 parameters of various small protein domains , we show that this flexibility does not provide clear evidence of intermediate-scale loop motions . Here , we propose a new and unbiased approach that is capable of inducing intermediate-scale conformational changes by continually applying a local perturbation throughout a short MD simulation . This method , Rotamerically Induced Perturbation ( RIP ) , was inspired by a perturbation method previously developed in our lab , the Anistropic Thermal Diffusion method [33] , which was designed to probe intramolecular signaling within a protein by simulating flow patterns of kinetic energy . In the ATD method , the protein is first cooled to a very low temperature , and then an individual residue is coupled to a 300 K heat bath . For physically interacting residues , a pathway of heat transfer is induced through the protein . The important concept taken from the ATD method is the idea of applying a local perturbation at a single residue in order to generate a deformation in the structure . To probe larger scale conformational changes , one could imagine simply applying a high temperature “bath” to an individual residue in a protein that has been pre-equilibrated to 300 K , but is otherwise uncoupled from any temperature baths . The applied energy would then be distributed amongst the bond , angle and torsional modes of vibrations in the residue . While this does result in larger perturbations , unfortunately most of the energy is taken up by bond vibrations , which quickly conveys the energy through interconnected covalent bonds along the backbone causing the backbone to unfold at the point of perturbation . In the RIP method , instead of applying a general heat bath to a residue , the perturbation is applied only to the sidechain torsional degrees of freedom , resulting in the rotation of the sidechain χ angles , while the bond lengths remain unperturbed . As this motion is orthogonal to the backbone degrees of freedom; for most residues , the RIP method does not produce significant changes in backbone structure . But for certain residues , the RIP method induces large segments of the protein to move , often by several Ångstroms , in a time period of only 10 picoseconds . As this is a relatively cheap calculation , a global map of deformability can be generated by independently perturbing every residue in the protein . In order to see if the induced perturbations capture information about real proteins , the RIP analysis was applied to five proteins with different dynamics . These include TIM ( Figure 1A ) and DHFR ( Figure 1B ) , both of which possess loops that have been measured to move on the millisecond/ microsecond timescale; α-lytic protease ( αLP ) ( Figure 1C ) , a kinetically stable protein that is known to be extremely rigid , which presumably has no mobile loops; the Estrogen Receptor ligand-binding domain ( ER ) ( Figure 1D ) in which Helix-12 is known to undergo a large conformation change in response to ligand binding; and finally the N-terminal domain of the chaperone HSP90 , which has a large lid that interacts with the active site ( Figure 1E ) . The simulations of these proteins using the RIP method generate deformation maps that characterize the mobility ( and lack of mobility ) of different segments in the protein structure . Analysis of these deformation maps demonstrate that the simulated dynamic properties of the proteins compare favorably to experiment .
Previous efforts to induce local perturbations used generic heat baths to apply the perturbation to an individual residue [33] , [34] . While sufficient for inducing small-scale changes , we wished to explore applicability to larger perturbations . In our experience , the application of a high-temperature heat bath propagates most of the energy along the backbone , which invariably unfolds the backbone at the point of perturbation . In contrast , the Rotamerically Induced Perturbation ( RIP ) circumvents the problem of the backbone unfolding by exclusively perturbing the χ angles , generating motions orthogonal to the backbone degrees of freedom . In the RIP method , the protein is first stripped of ligands and waters , energy minimized , and pre-equilibrated without constraints to 300 K over 10 ps using Amber with GB/SA implicit solvent [35] . To avoid having to modify the Amber source code , the actual RIP calculations are performed by running sets of constant energy 100 fs simulations , which allow the rotamer perturbations to develop , with no other restraints applied . Between each interval , the instantaneous rotational velocity is calculated for each χ angle of the residue being perturbed . The bond and bond angle vibrations are suppressed , and a pure χ rotational velocity of the desired magnitude is applied to the atomic velocities and then the next 100 fs interval begins . Because of the reduced degrees of freedom , imposing a sidechain rotamer kinetic energy of 300 K results in far greater rotational velocities than those in a standard 300 K MD simulation . The direction of rotation is maintained until it exceeds a limit of ±60° from the initial χ angle , upon which the direction of rotation is reversed . This limit restricts the sidechain to exploring the basin surrounding a single rotamer , and allows long sidechains such as MET to effectively explore the landscape of a single rotamer . The resultant motion is that of the sidechain rotating back-and-forth around the initial χ angle . If there are no collisions with other residues , no energy gets transferred . But if there is a collision , a large displacement may be induced in another part of the protein . As the protein is simulated under constant energy during each interval , the transferred energy continues to propagate through the protein , resulting in a slight increase in the overall temperature . In RIP , the rotational velocities are calculated directly from the rotational inertia of the sidechain . Thus we expect the rotational velocities of the χ angles in different sidechains to be different . For instance , at 300 K the phenyl ring in Phe should rotate more slowly about its χ2 angle than would the methyl-group in Ile about its χ2 angle . In order to demonstrate that RIP generates plausible χ angle behavior , we simulated the 17 amino acids that possess sidechains having χ-angles using a standard MD protocol . The amino acids were capped with methyl groups , and then simulated in AMBER with GBSA for 10 ps using a standard thermal bath at 300 K . The average values of the χ angle rotational velocities are then extracted from these trajectories , providing a reference set of rotational velocities for the χ-angles of each amino acid . We then performed the RIP method on the same 17 amino acids . The average rotational velocities were extracted from the trajectories of the RIP simulations , and compared to the standard set of rotational velocities ( Figure 2A ) . The Pearson correlation coefficient is 0 . 84 , which shows that the RIP protocol generates the relative differences of the rotational velocities found in the standard simulations . The fit is better for low rotational velocities , which correspond to the motion of the heavier sidechains . The higher rotational velocities , which deviate more from a straight-line fit , correspond to the small sidechains or the ends of the long sidechains . The differences between a residue perturbed by RIP and a residue regulated with a standard thermostat at 300 K can be shown in greater detail with the results for Ile ( Figure 2B–D ) . As intended by the design of the RIP protocol , the average kinetic energy of the residue perturbed by RIP is equivalent to that in a standard simulation ( Figure 2B ) . The differences can be found in the frequency distributions of the χ angles ( Figure 2C and 2D ) , where RIP induces much higher rotational velocities than does the standard distribution . In the RIP rotational-velocity distributions , χ1 is peaked around 0 . 21 rad ps−1 ( Figure 2C ) while χ2 is much flatter ( Figure 2B ) . This is due to the smaller weight of the methyl group controlled by χ2 , which moves much faster and collides more often with the rest of the Ile amino acid , thus broadening the distribution . Triosephosphate Isomerase ( TIM ) has a ligand-binding loop that can close over the active site of the protein . In different crystal structures , this loop is observed in both an open and closed state [36] and NMR relaxation experiments detect loop fluctuations at a rate of 3×104 s−1 [11] . TIM typically exists as a dimer in which inter-monomer contacts are mediated through a separate dimer-interface loop . The RIP analysis was applied to all residues in the TIM monomer having the open state of the mobile loop ( Figure 1A ) . The simulations were also performed on the closed state of TIM and similar results were generated ( discussed below ) . For each perturbing residue , the response of the protein can be measured by the Cα RMSD of the residues in the protein at the end of the 10 ps simulation . As an example , applying RIP to Glu128 , a residue located near the mobile loop of TIM , results in a Cα RMSD deformation response with one large peak at residue 175 ( Figure 3A ) . By overlaying the 10th ps conformation of the RIP simulation over the crystal structures ( Figure 3B ) , it can be seen that the peak of large conformational change at residue 175 corresponds to the ligand-binding loop at residues 167–177 . There is also a displacement of the helix near Glu128 . A global map of deformability in TIM can be constructed by applying RIP to every residue along the entire length of the protein . Each column in the RIP deformation map represents the 10 ps Cα RMSD response to a perturbation of RIP on the residue with sequential numbering on the X-axis ( Figure 4A ) . To facilitate visualization , the actual response is shown on the map only if the responding residue is perturbed by more than a defined threshold . The threshold was chosen by analyzing the probability distribution ( Figure 3C ) of the Cα RMSD responses in the set of RIP simulations applied to every residue in TIM . This distribution has an average of 1 . 5Å , a σ of 1 . 2 Å . The peak of Cα RMSD at 0 . 8 Å corresponds to the residual motion from the initial pre-equilibration of 300 K . Using a cutoff of 6Å ( mean+3 . 75σ ) results in a sparse global map consisting of mostly contiguous segments of deformation ( Figure 4A ) . Lower thresholds result in noisier maps , showing many single-residue deformations . The first point to note is that there is no systematic response along the diagonal in the RIP deformation map . Residues adjacent to the perturbed residue are not automatically disturbed . This demonstrates the key property of the RIP method: perturbing a sidechain does not systematically disturb the local backbone , unless there is a specific interaction of the perturbed sidechain to the backbone . Consequently , if deformations are observed then they can be directly attributed to the perturbing sidechain . As is evident from the TIM RIP deformation map , perturbing some residues can induce large changes in the protein while others have virtually no effect . The magnitude of the perturbation inducible by a particular residue can be quantitated by counting the number of residues that respond significantly to the perturbation ( above the 6Å threshold ) . Residues capable of inducing significant perturbations will be referred to as structural linchpins ( Figure 4B ) . While structural linchpins are generally larger amino acids , some smaller amino acids also show up . Another way of extracting useful information from the RIP deformation map is to quantify the susceptibility of each residue to local perturbation ( Figure 4C ) by summing the number of above-threshold deformations horizontally across the deformation map ( Figure 4A ) . Unlike the perturbation strength defined above , this local flexibility is a global property based on the entire set of perturbations along the protein chain . In TIM , there are two segments with significant local flexibility ( Figure 4C ) . One of the segments of large local flexibility corresponds to a striking horizontal band of deformations at residue 75 in the deformation map ( Figure 4A ) . This segment corresponds to the dimer-interface loop . As deformation of the dimer-interface loop occurs independently of the location of the applied perturbation , a reasonable interpretation is that the dimer-interface loop is an intrinsically mobile loop in the monomer state . In contrast , there is a scattered band of flexibility in the segment corresponding to the ligand-binding loop at residue 175 . Flexibility in the ligand-binding loop is induced only by the cluster of structural residues that surround it ( Figure 4B ) . This can be interpreted as the ligand-binding loop being conditionally flexible , i . e . , large fluctuations of neighboring residues can readily displace the mobile loop . To illustrate the range of motion induced by the RIP method in the mobile loop of TIM , the entire ensemble of the final conformations of each perturbation is shown ( Figure 5A ) . The range of motion generated by RIP is far greater than the differences between the open and closed conformations of the mobile loop found in the crystal structures ( Figure 1A ) . Dihydrofolate Reductase ( DHFR ) catalyzes the reduction of dihydrofolate by NADP . The protein binds both ligands through the Met20 loop . Crystallography of the enzyme with different ligands has defined three states for the Met20 loop ( Figure 1B ) suggesting that loop dynamics may be functionally important: i ) a closed state where the Met20 loop covers both dihydrofolate and NADP; ii ) an occluded state where the Met20 loop packs against the dihdyrofolate but blocks NADP; and iii ) an open state where the loop may be disordered in solution ( Sawaya and Kraut 1997 ) . Using the RIP analysis , the results of the perturbations can be used to generate a RIP deformation map of DHFR ( Figure 6A ) . From this map , potential structural linchpins can be identified ( Figure 6B ) , and the local flexibility calculated ( Figure 6C ) . The RIP deformation map displays many regions of large conformational changes , dispersed in sequence space . Regions of significant local flexibility ( Figure 6C ) map to the Met20 loop , the adenosine-binding loop , the F–G loop , and the G–H loop . Experiments suggests that all these loops couple to the activity of the Met20 loop [37] , [38] . α-Lytic protease ( αLP ) ( Figure 1C ) is a kinetically stable protein shown by experiment to be highly resistant to unfolding [39] and to have extraordinarily high hydrogen-exchange protection factors ( >1010 ) throughout its large hydrophobic core . The reduced dynamics and absence of mobile segments is believed to be functionally important for minimizing proteolytic destruction [40] . Reflecting these properties , αLP is also found to be extremely stable in standard MD calculations [33] , where loops separating secondary structural elements appear to be firmly held by interaction with the body of the protein . This provides a good test for the RIP method , as any large-scale predictions of flexibility would likely be erroneous . In agreement with the known dynamics of αLP , the deformation map for αLP ( Figure 7 ) is remarkably empty; there are only scattered segments of limited deformation , indicating that αLP is particularly insensitive to single residue perturbations and thus , conformationally very stable . The Estrogen Receptor belongs to a family of nuclear receptors that are ligand-inducible transcription factors [15] . The ligand binds in a hydrophobic pocket in the ligand-binding domain of the Estrogen Receptor ( ER ) that is covered by Helix-12 , so that in the bound state the ligand is completely buried . Not only does Helix-12 need to be displaced for ligand entry and exit , but it is also the key allosteric transducer . Crystal structures of ER with bound agonists and antagonists ( Figure 1D ) reveal that Helix-12 responds to the bound ligand by occupying one of several conformations that serve to either support or block binding of the downstream transcriptional co-activator protein [16] . Clearly , Helix-12 is a functionally-critical mobile segment of the protein although there is no information about the timescale of its motion . The RIP deformation map of ER ( Figure 8A ) calculated in the absence of ligand , shows a very limited set of responding distortions with the exception of a horizontal band at residue 162 . The horizontal band corresponds to a short loop between the α-helices ( Figure 8C ) . Apart from a large conformational change of a long segment at the C-terminus ( residues 232–250 ) , ER appears to be very stable . The large conformational change in the C-terminus is induced by Trp83 , which is the only residue that would qualify as a structural linchpin ( Figure 8B ) . Trp83 has been found to be a highly conserved residue in the nuclear receptor family with the greatest change in the solvent accessible surface area between the apo and holo structures [41] , indicating a possible key role in binding Helix-12 . The conformation induced by RIP on Trp83 reveals a dramatic cooperative motion in Helix-12 , where the α-helix remains intact but moves by 13 Å ( red in Figure 9 ) . Although it doesn't move to the same location found in the crystal structure ( green in Figure 9 ) , perhaps due to the short simulation , we do see a hinge motion about the same pivot point ( blue and green in Figure 9 ) observed in the crystal structures . Notably , this conformational change does not result in a significant distortion along Helix-12 . The ability to generate this kind of large cooperative motion is due in part to the removal of the ligand normally found in the crystal structure for the simulations . In the absence of ligand , Helix-12 covers a cavity and is only weakly bound to the body of the protein . Thus the collisions generated by the perturbation on Trp83 are sufficient to induce the entire helix to move cooperatively . Th N-terminal ligand-binding domain of the chaperone HSP90 [42] consists of a active-site lid that is 30 amino acids in length ( Figure 1E ) . The lid is involved in ATP/ADP binding that is associated with the conformational changes of HSP90 . The motion of the lid corresponds to an intermediate-scale motion of the lid independent of the body of the protein , as revealed by crystal structures of HSP90 in different ligand-bound states . The RIP deformation map of the HSP90 ligand-binding domain ( Figure 10A ) in the ADP-bound state , calculated in the absence of ligand , clearly identifies the lid as a highly flexible region . The conformational ensemble of the perturbed structures show large conformational changes in the lid comparable with the scale of motion found in the crystal structures ( Figure 5D ) . There are other systems that calculate protein flexibility from a crystal structure . To compare RIP to these methods ( Figure 11 ) , we have calculated the flexibility using the Anisotropic Network Model ( ANM ) [43] and CONCOORD [32] . We chose ANM as a representative of contact-based network models that generate domain-level motions , where ANM generates a theoretical B-factor for each residue from the sum of the contribution of the lowest modes of oscillations . We chose CONCOORD as a representative of local instability methods , where CONCOORD calculates an RMSDf fluctuation ( RMSDf is the averaged Cα RMSD over different conformations of the same structure ) at each residue from an ensemble generated from a monte-carlo simulation of the protein using distance constraints derived from a standard force-field . Both methods were chosen largely due to the easy availability of the software implementation . Finally , we compare these flexibilities to various experimental data . For the four proteins with large conformational changes . In the case of TIM , for example , the RMSDf was calculated from the position of the Cα atoms between the open and closed structures ( Figure 11A ) after they were aligned with the weighted superposition program Theseus [44] . The flexibilities calculated from the open conformation of TIM provide different results with respect to the RMSDf . All three measures of flexibility identify the dimer-interface loop ( near residue 65 ) as flexible even though the RMSDf of the dimer-interface loop is negligible , as both the open and closed conformations exist in the same dimer arrangement in the crystal . For the ligand-binding loop ( near residue 165 ) , both RIP and CONCOORD identify elevated flexibility whereas ANM does not . CONCOORD also identifies several other regions as flexible , where there is no corresponding elevated RMSDf values . In contrast , RIP identifies as flexible only the ligand-binding loop and dimer-interface loop . As a further comparison , we show the flexibilities calculated from the closed conformation of TIM ( Figure 11B ) . For the ligand-binding loop , only RIP identifies the loop as intrinsically flexible whereas both ANM and CONCOORD do not . There exists a rich set of NMR measurements of DHFR that can be used to evaluate the calculated flexibilities of DHFR . From the averaged RMSDf between the open/closed/occluded conformations in crystal structures , only the ligand-binding Met20 displays any large conformation change . We find that both CONCOORD and RIP identify the Met20 loop as flexible whereas ANM does not . This can be contrasted with the S2 parameters [22] , which measures the backbone 15N mobility of DHFR on the picosecond-nanosecond timescale . We find that CONCOORD provides the best agreement to the S2 parameters , as would be expected from previous studies [32] ( Figure 11D ) . However , RIP also provides a reasonable match , whereas ANM does not . The similarity between RIP and CONCOORD is surprising as they do not agree in the other proteins . One possibility is that the nanosecond motions of the loops in DHFR correspond to microsecond fluctuations . This can be measured by NMR relaxation exchange factors Rex factors [38] that detect significant backbone fluctuations between distinct states on the microsecond/millisecond timescale . Rex measurements for DHFR identify microsecond dynamics in the Met20 loop , the F–G loop , the adenosine-binding loop and the G–H loop ( Figure 11E ) , suggesting that in the case of DHFR , there is significant overlap between the nanosecond and microsecond dynamics . Coupling between these loops have also been identified in nanosecond MD simulations through cross-correlation and quasi-harmonic analysis [21]–[24] . The large conformational change in Helix-12 of ER , as evident in the large RMSDf values ( Figure 11F ) , provide a difficult challenge for these models of flexibility . There are elevated values of RMSDf in short loops near residue 35 and 165 ( Figure 11F ) , where all three flexibilities finds elevated flexibility . None of the flexibilities identify Helix 12 as a highly flexible region . However , with RIP we can identify one large conformational change by examining the deformability map ( Figure 8A ) , where the conformational change is consistent with the crystal structures ( Figure 9 ) . Both ANM and CONCOORD identify elevated flexibility at residue 195 , which do not correspond to elevated RMSD values . The conformational changes in HSP90 is dominated by the motion of the lid , indicated by the large values of RMSDf at residue 110 ( Figure 11G ) . This large conformational change is not identified by ANM but both RIP and CONCOORD identify the lid region as flexible . One advantage with RIP over CONCOORD is that by examining the conformational ensemble of perturbed structures generated by RIP ( Figure 5E ) , we can see that the generated motions are of the order of the RMSDf observed in the crystal structures . Both CONCOORD and ANM identifies a region of elevated flexibility at residue 60 that does not correspond to any elevated region of RMSDf and the large RMSDf in the N-terminal that is involved with inter-domain interactions does not correspond to any calculated flexibility . In conclusion , we find that the flexibility of RIP identifies only loop motions that correspond to large conformational changes of intermediate-scale motions . In contrast , CONCOORD identifies more regions as flexible , where there is some overlap with the regions identified as flexible by RIP . ANM performs poorly for intermediate-scale motions .
Molecular dynamics ( MD ) is generally accepted to be an accurate representation of biochemical processes on the molecular level [45] . However , there is a practical upper limit of hundreds of nanoseconds in most MD simulations . Such simulations can only explore small motions in a protein structure whereas many physically interesting processes occur on much longer time-scales . These long times are needed to allow large but rare conformational changes to occur . Perturbation techniques can theoretically be used to quickly generate these rare conformational changes , but there have been many obstacles to applying unbiased perturbation techniques to protein structures using MD . Here , we have developed an unbiased local perturbation method that can generate selective large conformation changes in very short MD simulations . RIP has several desirable properties that improve upon previous perturbation methods . First , solvation and electrostatics are well treated by the implicit-solvent GBSA method . Second , by driving sidechain rotamers instead of all local atoms , RIP minimizes local backbone distortions , maintaining secondary structures as intact elements . Most perturbations induced by RIP do not result in large-scale distortions of the protein chain , resulting in a sparse map of deformations . Third , RIP can induce large cooperative motions in coherent segments while preserving their local structure , as for example , in Helix-12 in the ER LBD . Fourth , RIP eschews the need for manual restraints or defined trajectories in generating large motions . RIP can thus be applied to any given protein structure . Fifth , RIP is a relatively inexpensive calculation as large Cα RMSD deviations are generated within a short simulation ( 10 ps ) , allowing a global analysis to be performed in nanoseconds of simulation time . Since the perturbations induced by each residue are independent , the simulations can be readily performed on a parallel cluster . As illustrated here , the goal of RIP is to map regions that are readily perturbed and to help discover potential structural linchpins that may dictate local conformation . For example , if a segment is easily deformed by several different perturbations then it is clear that the interactions that bind the segment to the body of the protein are weak . It is found that the local flexibility map , which averages over the global pattern of conformational changes , clearly identifies the loops that have been experimentally determined to be mobile in both DHFR and TIM on the microsecond/millisecond timescale . Furthermore , as revealed by the αLP calculations , the RIP analysis doesn't spuriously find mobile segments where they shouldn't exist . Importantly , RIP can discriminate between proteins that possess intrinsically mobile loops from those that do not . In comparison , a number of contact-based approximations can deduce large domain-level motions of proteins , such as elastic network models [7] and graph theoretic analysis of the contact network [8] . Essentially contact-based models assume that the network of contacts determine the principal degrees of freedom of the protein . By approximating the contacts as a network of springs , the slow time-scale dynamics of the protein can be deduced from the lowest modes of oscillation of the network of springs . Gaussian network models have modeled the domain-level motions of such large systems as the ribosome [9] and the fluctuations of the capsid of a bacteriophage [10] . It has been shown that the lowest modes of coarse-grain elastic network models reproduce the low frequency modes of more detailed calculations such as normal-mode analysis and quasi-harmonic analysis of MD trajectories over several nanoseconds [46] , [47] . Nevertheless , contact-based models cannot detect intermediate-scale motions such as those generated by RIP . In a study of TIM using elastic network models , it was found that the lowest mode of oscillation involved limited motion of the ligand-binding loop [48] where this motion was entangled with other motions in the rest of the protein . In contrast , the liagnd-binding loop in the crystal structures moves independently of the body of the protein ( Figure 1A ) . Independent loop motions are inconsistent with the assumption in contact-based models of the collective motion of a single network of inter-connected springs . When the ligand-binding loop is in the closed state , it forms contacts with the body of the protein . But when the ligand-binding loop opens , the contacts of the loop to the body are broken , resulting in a fundamentally different networks of contacts . Such motions cannot be produced from contact-based models . Another class of models attempts to identify flexibility through the analysis of local instabilities in a given structure . These models typically generate an ensemble of structures that can be used to calculate instabilities along the protein chain . One approach is COREX that calculates the free-energy of unfolding short segments of a protein structure using an analytical approach [29] . A related approach is the Protein Ensemble Method [30] , [31] that explores local unfolding by randomly generating geometric variations of sections of the backbone . Another approach is CONCOORD [32] that generates alternate conformations by monte-carlo exploration of the atoms constrained by distance constraints derived from a standard MD force-field . The flexibility derived from the ensembles generated by PEM and CONCOORD accurately reproduces the NMR S2 parameters for several small proteins [30] , [32] . The S2 parameters of a small protein have been shown to couple strongly to residual dipolar couplings and B-factors that reflect the low-amplitude fluctuations in the picosecond to nanosecond regime [49] . As the flexibility of PEM and CONCOORD correlates well to the S2 parameters , this suggests that these instability methods are specifically modeling the low-amplitude fluctuations on the nanosecond regime . Whilst there is some overlap between CONCOORD and RIP , the flexibility calculated by RIP misses much of the low-amplitude fluctuations that occur on the nanosecond regime as identified by CONCOORD . Instead RIP mainly identifies intermediate-scale motions that occur on the timescale of microseconds or longer . The overlap occurs for loops such as the Met20 loop in DHFR that are mobile on the nanosecond timescale , as revealed by S2 parameters , and also on the microsecond timescale , as revealed by the Rex factors . Overlaps between CONCOORD and RIP also occur for intrinsically mobile loop , which are loops that fluctuate >6Å independently of perturbations in short timescales . Apart from intrinsically mobile loops , which can be easily identified from the deformation map as a horizontal band of fluctuations , RIP identifies conditionally flexible regions that correspond to microsecond scale motions , such as the ligand-binding loop in TIM in both the open and closed conformations , and the Helix-12 motion in ER . The flexibilities identified by RIP are more likely to reveal functionally significant conformational changes in a protein structure . The ability of RIP to generate large conformational changes of several Ångstroms is not due to its ability to sample the rare fluctuations that might occur over a timescale of microseconds or milliseconds . Indeed , because of the non-equilibrium driving conditions , the RIP simulations do not provide any information on the timescale of the simulated motion . Rather it is due to the ability of local perturbations to efficiently explore the strength of contacts that anchor local protein segments . Conformational changes occur only if the perturbation can break the contacts ( hydrophobic , polar and hydrogen bonds ) that hold these segments to the body of the protein . Although the perturbations are large , as implemented here there is a limit to the extent of perturbation - the overall kinetic energy of the perturbed residue matches that of the same residue equilibrated to 300 K . As such , there is only enough energy to induce conformational changes on segments on the surface of the protein or those near potential packing defects . Importantly this also results in limited distortions within displaced structural elements as in the case of Helix-12 in the ER ligand-binding domain . It is important to note that the conformational changes generated by the perturbations are artificially large in that they result from large collisions arising from χ angle rotations at velocities far above their normal values . As a consequence , the simulated motions show a large variance in conformations ( Figure 5 ) , much more than expected from static snapshots from crystal structures or from NMR ensemble analysis . Although it is unlikely that exactly these perturbed conformations would be generated in a very long equilibrium MD simulation , large RIP-induced conformational changes identify regions where the protein chain can undergo large conformational changes . To obtain more realistic structures , the RIP conformations could be used as starting points for conventional MD simulations or high temperature simulations with manual constraints [50] . The flexible regions identified by RIP may identify potential loops that could be modeled using loop-prediction systems such as PLOP [26] , [27] and ROSETTA/BACKRUB [28] . However , as such systems are limited to 12-residue loops , they cannot explore conformations of larger elements such as Helix-12 in ER or the lid in HSP90 . Intriguingly , the motions generated by RIP in DHFR and ER include examples of coupled motions between different mobile segments and ligand-induced structural changes , suggesting that further development of RIP may result in tools to probe mechanisms of allostery . Another possibility is the analysis of the interaction of mobile loops with binding sites , where alterations in surface loop structures can dramatically alter patterns of ligand binding . RIP could provide a computational mechanism for rapid identification of such potentially relevant loops , which might be particularly important for computational ligand screening . Thus RIP followed by MD or loop modeling could provide an efficient means to generate alternate conformations for computational drug discovery .
The RIP method is implemented as a PYTHON wrapper around the Sander package of AMBER [35] and all analysis code was written in PYTHON . The simulations of the RIP method are run in AMBER , using the PARM96 force-field with an GB/SA implicit-solvent term . To prepare for the simulation , ligands and crystallographic waters are removed from the crystal structure . The structure is then minimized and a Langevin thermometer is applied for a short equilibration at 300 K for 10 ps with a friction constant of 5 ps−1 . The standard protocol for a RIP method lasts for 10 ps , which is long enough for large motions to be generated . At the beginning of the RIP method , the equilibrium value of the χ angles of the residue is stored . The run is then broken up into 100 fs intervals where each interval is simulated at constant energy . Between each interval: ( 1 ) the direction of the rotational velocity of each χ angle is stored; ( 2 ) the atomic velocities of the residue is set to zero; ( 3 ) if the value of the χ angle exceeds 60° of the equilibrium χ value , the direction of the rotational velocity is reversed; ( 4 ) the magnitude of each χ rotational velocity is calculated from the sidechain conformation; ( 5 ) the χ rotational velocities are transformed into into atomic velocities and added to each atom; ( 6 ) the kinetic energy of the residue is scaled to the rotation temperature of 300 K . By scaling the atomic velocities , the kinetic energy of the residue is effectively transfered into the rotational modes of motion . This guarantees that even though the motion is artificially large , the amount of energy in the rotation is not more than would be available for the sidechain at equilibrium , even though this is unlikely to happen . Between the intervals , a Python module translates the AMBER restart files into a Python object , from which the RIP protocol is used to generate new AMBER restart files for the next interval . Finally , the trajectories of all the intervals are spliced into a single trajectory . Since the modifications are made on the velocities , the coordinate trajectories are continuous . In the RIP method , a rotational velocity for each χ angle of a sidechain is calculated at the beginning of every interval . From this rotational velocity , the atomic velocities are generated . To generate the the rotational velocities of the χ angles , each χ angle is assumed to be an independent degree of freedom . Based on the equipartition theorem , each independent χ angle can be assigned an energy E derived from the temperature T . This E is drawn randomly from a Gaussian distribution with mean energy ½kT and standard deviation √ ( ½kT ) . To convert a rotational velocity into an atomic velocity , a frame of reference for the axis of rotation must be chosen . As rotational velocities are only defined relative to the axis of rotation; rotations can occur on either end of the axis , and still give the same rotational velocity . Since the purpose of the RIP method is to minimize the motion of the backbone , only the sidechain atoms on the side of the rotation axis away from the backbone are rotated . Consequently , the rotational inertia of each χ angle , I = Σ mr2 , is calculated as the sum of the moment of inertia of these sidechain atoms , where r is the perpendicular radius of each atom from the χ angle axis of rotation . To convert E into a rotational velocity ω , the equation of rotational energy E = Iω2 is used . This is converted to a tangential velocity v through v = rω . This velocity is applied to the atom along the direction of the tangent to the axis of rotation . The atomic velocities due to each χ angle are then added cumulatively to each atom . However , the different χ angles of the same sidechain do not represent completely independent degrees of freedom . As such , the final atomic velocities are re-scaled such that the total kinetic energy of the sidechain is E = 3/2 nkT where T = 300 K . This scaling only changes the magnitudes of the rotations and preserves the pure rotation around the χ angles . In the analysis of the RIP simulations , rotational velocities of the χ angles need to be extracted from the trajectories . In the generation of rotational velocities , only atoms that are on the side of the rotation axis of the χ angle away from the backbone contribute to the rotational velocity . Therefore , in the extraction of the rotation velocities , only these atoms are considered . For each atom that fits the criteria , the tangential velocity v to the axis is calculated . This v is converted to a rotational velocity by ω = v/r where r is the perpendicular radius from the axis . As the contributions of each atom to the total rotational velocity of a χ angle depends on its moment of inertia , a weighting ( w ) for each atom is calculated from the moment of inertia I = mr2 of the atom . The weighting is given by w = I / Itotal where Itotal is the sum of the I for all the atoms involved in the χ angle . The overall rotational velocity is then given by ωtotal = Σ wω . | Many proteins undergo large motions to carry out their biological functions . The exact nature of these motions is typically inferred from the crystal structures of the protein trapped in different states , which normally constitutes a difficult series of experiments . As molecular dynamics is generally accepted to accurately model the motion of proteins , the promise is that a long enough simulation will generate all the motions of a given protein structure . Unfortunately , current systems run too slowly to simulate all but the smallest motions . To overcome this computational limit , we have developed a molecular-dynamics perturbation method that induces large changes in a protein structure in very short simulation times . The changes correspond to large motions of specific structural elements on the surface of the protein that corroborate well with the canonical motions of several well-characterized proteins . This bodes well for our method to identify , for any given protein structure , structural elements on the surface that might bind drugs , regulate signals , undergo chemical modifications , or become unstructured . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"computational",
"biology/molecular",
"dynamics",
"biophysics/theory",
"and",
"simulation"
] | 2009 | Probing the Flexibility of Large Conformational Changes in Protein Structures through Local Perturbations |
As animals move through the world in search of resources , they change course in reaction to both external sensory cues and internally-generated programs . Elucidating the functional logic of complex search algorithms is challenging because the observable actions of the animal cannot be unambiguously assigned to externally- or internally-triggered events . We present a technique that addresses this challenge by assessing quantitatively the contribution of external stimuli and internal processes . We apply this technique to the analysis of rapid turns ( “saccades” ) of freely flying Drosophila melanogaster . We show that a single scalar feature computed from the visual stimulus experienced by the animal is sufficient to explain a majority ( 93% ) of the turning decisions . We automatically estimate this scalar value from the observable trajectory , without any assumption regarding the sensory processing . A posteriori , we show that the estimated feature field is consistent with previous results measured in other experimental conditions . The remaining turning decisions , not explained by this feature of the visual input , may be attributed to a combination of deterministic processes based on unobservable internal states and purely stochastic behavior . We cannot distinguish these contributions using external observations alone , but we are able to provide a quantitative bound of their relative importance with respect to stimulus-triggered decisions . Our results suggest that comparatively few saccades in free-flying conditions are a result of an intrinsic spontaneous process , contrary to previous suggestions . We discuss how this technique could be generalized for use in other systems and employed as a tool for classifying effects into sensory , decision , and motor categories when used to analyze data from genetic behavioral screens .
Active movement is one of the defining features of animals , and the use of locomotion to search for resources within the environment is likely among the most ancient of behaviors . Observations on motile organisms , ranging in scale from bacteria to whales , indicate that search patterns are structured by a combination of internal processes and external cues [1] , [2] . Sensory systems enable organisms to detect favorable objects at a great distance [3]–[5] and they use this ability to localize resources by either directed motion ( taxis ) or changes in locomotor statistics ( kinesis ) . Prior research suggests that , in the absence of external cues , the animal behavior is generated by internal processes , and that the overall animal fitness is sensitive to the exact characteristics of this internal process ( e . g . , Levy statistics ) [6]–[18]; it has also been questioned whether observed large-scale statistics can give any insight on an internal process that generated the behavior , and whether the internal processes can dominate over stimuli-elicited behavior [19]–[22] . As for the internal processes , these can be divided into truly stochastic sources , and deterministic results of a deliberate , but unobservable , internal mechanism based on internal metabolic/neural states . When observing an intact motile organism , it is not easy to determine which components of its locomotion behavior are triggered by internal processes versus external cues , yet such classification is essential for deciphering the underlying logic of its movement and search behavior . The task is further complicated by the fact that an external observer might not be able to distinguish between truly stochastic processes and the deterministic results of a deliberate , but unobservable , internal mechanism . For example , software pseudo-random number generators produce strictly deterministic sequences , which appear to be random to an external observer who does not have access to the internal state of the system [23] . A major goal of both cell biology and neuroscience is explaining the molecular and cellular bases of these three qualitatively different processes ( sensory-driven , purely stochastic , and deterministically based on internal states ) . If the salient features of the external world are known , it is possible to gain insight into sensory-driven behaviors through the use of sensory-response correlation [24] . The analysis of the internally-driven processes is much more challenging . Given uncertainty in measurement and the inability to perfectly reproduce experimental conditions from trial to trial , variability in the results of behavioral experiments has often been treated as a limit on our ability to measure stimulus-driven behavior . In this view , variability in responses from trial to trial reflects irrelevant components of behavior , which are averaged until the mean—interpreted as the response the animal ideally would have produced—becomes clear [25] . From the opposite perspective , many researchers have attempted to artificially remove all relevant sensory input to an animal and measure behaviors in conditions of sensory deprivation to reveal intrinsic properties , especially the statistical distributions of behaviors [26]–[28] . Although focusing in isolation on either the stochastic [9] , [29] , [26] or the sensory components [30] of search behavior have provided key insights , neither of these extremes is sufficient to capture the full range of processes at play as an animal moves under natural conditions . Attempts to investigate the interaction of internal and external processes include studies of bacteria [31] and nematode worms [32] , [33] , organisms for whom chemicals provide the most salient cues for food search . For larger animals with image-forming eyes , vision may provide another essential cue in search algorithms , because vision is the only sense which allows to perceive remote parts of the environment . Often vision cannot be considered separately from the mechanics of locomotion [34] . Flies are a model of computational efficiency and robustness , to date not equaled by artificial systems , which often seek to imitate nature [35] , [36] . Much is known about fly vision [37] , [38] . Since the pioneering work of Kennedy [39] and Mittelstadt [40] , the behavioral responses of flies to experimenter-defined visual stimuli have been extensively investigated . Electrophysiological recordings have complemented and extended our knowledge of phenomena such as the neural basis of motion detection [41]–[44] and other key aspects of sensory processing , such as receptive field tuning [45] . However , there are many challenges in the identification of neural processing and how it produces complex behavior , especially as regards the characterization of “discrete” behaviors , such as the rapid turns ( “saccades” ) of Drosophila , which are the object of this study . In fact , many studies which offered complete characterization of the animal response are limited to “continuous” behavior , for which they provide linear ( or “linearized” ) models [46]–[49]; this allows using techniques such as linear system identification . Identifying the neural causes for “discrete” behavior involves solving a different set of problems . Firstly , there are the problems of segmentation and classification of behaviors ( including the definition of what “behavior” and “a behavior” are ) , for which it is often necessary the use of nonlinear machine learning methods [50] . Then , there is the problem of building models that can correlate the stimulus with the behavior ( s ) . While it is possible to postulate models that also integrate well with our understanding of lower-level behavior [51] , [52] , it is not clear how such methods can be identified from the data . On the practical side , it is evident that discrete decisions , such as turning decisions , are meant to guide exploration and therefore should be investigated in naturalistic situations . This poses practical problems of tracking the animal position in a large environment , and it also precludes ( at the current level of technology ) the uses of direct neural recording . In fact , comparatively few attempts have been made to correlate parameters of visual stimulus with behavioral responses in unrestrained conditions [53]–[56] . In this work , we present an analysis that can quantitatively discriminate the effect of visual stimulus as opposed to internal processes in the generation of saccades in the fruit fly . Our conclusions are that visual stimulus has a dominant role . One important message of our work is that it is very difficult to identify models of complex behavior that can explain everything , often because insufficient data can be collected . Therefore , it is important to “search for simplicity” [57] , for example by framing the problem as dimensionality reduction , and to use models that a posteriori can justify their assumptions . While we describe this analysis for visual processing in Drosophila , our goal is to construct a general method that can be used for other sensory systems , other animal species , or in the context of genetic screens .
Flies from the laboratory stock derived from 200 wild-caught females were reared on a 16 h:8 h light dark cycle under standard laboratory conditions . Three day old adult female flies were anesthetized with cold and individually housed within centrifuge tubes containing a moist tissue paper . Flies were starved ( but provided with water ) in the tubes for four to six hours before being released into the flight arena . Most flies would immediately begin flying , and we terminated tracking after the fly landed . We then removed each fly with a wand attachment of a vacuum cleaner before introducing another fly . Thus , each recorded trajectory is derived from a fly's initial experience exploring the novel environment . The flight arena was a 2 meter diameter , 80 cm high cylinder ( see Figure 1A ) . 10 cm×10 cm red and green gel filters ( Roscolux ) were attached to the arena in a regular checkerboard arrangement and provided a high contrast visual stimulus to flies near the wall . One meter from the wall ( i . e . , at the center of the arena ) , the angular wavelength of this pattern was ∼11° , and consequently would be twice the inter-ommatidial spacing of a ∼5 . 5° in Drosophila [58] . The particular red and green filters were chosen to have similar infrared transmission to facilitate tracking using cameras outfitted with long ( IR ) pass filters . The arena was illuminated from outside with a circular array of eight 750W Fresnel stage lights pointing towards the arena center . These lights provided both visible and infrared light for fly visual responses and machine vision tracking , respectively . A detailed description of our tracking system may be found in [59] . Briefly , we used 11 cameras ( 6 monochrome Pt . Grey Firefly MV USB cameras and 5 monochrome Basler A602f cameras ) with wide-angle lenses and infrared pass , visible cut filters ( R72 , Hoya Filters ) to view the interior volume of the flight chamber . The cameras were positioned so that a fly within the tracking volume was viewed by 2 or more cameras at any given time , enabling a 3D estimate of its position ( Figure 1Bi ) . The cameras were first calibrated to compensate for image warping non-linearities ( deviations from the pinhole model ) and then the extrinsic and intrinsic parameters describing the pinhole model were found . Flies were tracked with an extended Kalman filter ( EKF ) , in which the motion model was a linear constant velocity model , and fly maneuvering is captured by the stochastic component of the Kalman filter . Because tracking updates occurred at a high rate ( 60 fps ) relative to fly maneuvering , we found this simplification to work well in practice . The 3D estimate of the fly position is recovered by triangulation from the 2D tracking data of each camera , and taking into account the relative uncertainty of each observations . Many species of flies , including Drosophila , exhibit rapid changes in heading as they fly , termed “saccades” [53] . Between saccades , flies tend to maintain an approximately straight course , and saccades account for at least 80% of the total net change in heading during flight [60] . There is little doubt that saccades can be triggered by visual stimuli , but the degree to which visual feedback plays a role in determining the velocity , duration , and amplitude of the resulting turn is unclear . Experiments using a magnetic tether , which permits free rotation about the yaw axis , suggest that flies do not respond to visual feedback during a saccade [61] . On the other hand , Stewart et al . [56] have observed a rebound effect after saccades in free flight , which they suggest is consistent with active optomotor feedback during the maneuver . This discrepancy is not of direct interest here , however , as we deal exclusively with the decision of initiating a saccade . To analyze saccades within a flight trajectory , one should choose a detection algorithm that , given the trajectory data , returns a series of saccade events , possibly with other attributes such as direction , amplitude , velocity , etc . In the past , several detection algorithms have been proposed , each one implicitly using a slightly different definition of saccade , and each one able to compensate for different sources of noise . In practice , large saccades are such distinct events that all algorithms agree with respect to most classifications , but different algorithms may disagree on detection of small saccades . We make sure that our results are robust to the choice of the algorithm , by using two distinct algorithms based on different principles . The two algorithms are described in detail in Text S1 and their source code is available on line . Briefly , the Geometric Saccade Detector ( GSD ) detects saccades from the x-y planar trajectory . The Angular-Velocity based Saccade Detector ( AVSD ) works primarily by considering the smoothed angular heading rather than the planar position . Unless otherwise noted , the statistics shown through the paper are derived using GSD , which is a posteriori shown to be better suited for these particular experimental conditions and equipment . Alternative figures showing the same statistics obtained from the AVSD algorithm are available as part of Text S1 . Figure 2A illustrates the conceptual approach of our analysis . We denote by the animal's physical spatial configuration ( its position and velocity in a fixed reference frame ) . The stimulus is the set of all sensory cues perceived by the animal , and it is a function of both the spatial configuration and the appearance of the world . Whereas is a concrete variable that we can possibly measure , the stimulus and the world are placeholders for things that , in general , are unknown . The actions ( e . g . saccades in our case ) are the external manifestations of the internal neural processing , which depend both on the instantaneous stimulus as well as on , another placeholder variable that represents the animal's internal state ( metabolic states , neural states , etc . ) , and which has dynamics of its own . We assume that it is possible to observe the spatial configuration as well as infer the actions from the observations , but that the internal state is not observable . We make a distinction between obtaining a functional model of an animal's behavior and identifying the underlying neural processes . Obtaining a functional description of behavior means obtaining a model that can predict the actions given the spatial configuration and a description of the world . In principle , we can do this by observing an animal's behavior with enough samples of , and . In general , however , there are a variety of neural models that could produce the same functional model . For example , many behaviors appear to be well-localized in time , suggesting an “action potential” neural model , but the underlying neural model can have very different properties [62] ( in other words , the microscopic explanation might be quite different than what the macroscopic observations suggest ) . The model that we now describe and that we will identify should be interpreted as a purely functional model , which can inform the search for neural models , to make sure that they are compatible with the externally observable free flight behavior . Figure 2B shows the particular model that we use in this paper . It is a particular form of the general model discussed above ( Figure 2A ) . In this model , we propose that the animal's actions can be summarized by the saccade events . We divide the saccade events in two classes: left and right saccades . In principle , one would want to consider additional attributes of the saccades , such as speed , duration , and amplitude . The analysis might also be expanded to consider other easily identifiable events [63] . However , limiting ourselves to a binary characterization of saccades allows us to model the behavior generation as Poisson processes , which offers relatively easy inference . We model saccade generation using rate-variant Poisson processes , i . e . , we assume that , for each class of events , internal and external factors influence a time-varying event rate according to a quantitative relation that we will attempt to identify . The most important assumption of our method ( which can and will be verified a posteriori ) is that , for the purpose of generating the behavior , the high-dimensional output can be compressed down to a low dimensional “feature” . This assumption is implicit in many other previous studies , and it is informed by the knowledge of the underlying neurobiology: the first level of sensory processing in flies and other animals consists in taking a very high-dimensional sensory stream and computing the few behaviorally-relevant features from it . Our only assumption is that this low-dimensional feature exists - we do not assume that we know this feature . However , we can attempt to automatically identify this feature from the observable data . It is important to note that we do not assume to know how this feature is computed from the stimulus . Indeed , the advantage of our method is that it allows identifying this feature based only on the observable behavior , without postulating anything on the sensory processing . Figure 2B also shows explicitly that , in addition to the feature-dependent pathway in our model , other unmodeled processing influences the behavior . The effect of this unmodeled processing will be quantitatively estimated as well . The saccade events are assumed to be generated by a set of interacting Poisson process with variable rate . The index stands for either one of the two classes of events ( L: left , R: right ) . The variable rate is assumed to depend both on the stimulus and the internal state , thus incorporating both random and deterministic effects . We write as the sum of three factors: ( 1 ) where the term is the contribution of the external stimulus through the feature ; the term is the contribution of the internal state ; and the term represents the contribution of a purely random stochastic process that does not depend either on an internal state or the stimulus . By omitting some of the terms in the equation above , one can recover many other simpler models . For example , purely random behavior is obtained by setting . The Poisson processes interact by inhibition . If any process generates an event , then any event generated from that process or any other process for a period of length Δ is ignored . This is meant to model a feature of many fixed action patterns that , once initiated , must run to completion before a different motor program can be initiated . Finally , Figure 2B shows another variable , which we call “reduced configuration” . We define as the subset of the spatial configuration variables that actually influence the stimulus , for a particular class of environments . In general , for a freely flying animal , is at least a 12 dimensional quantity , including the 6 degrees of freedom for position/orientation and the corresponding 6 for velocities ( additional degrees of freedom in the animal spatial configuration would be derived from the positions of body joints , such as the neck and wing positions ) . For particular environments , however , the stimulus is only dependent on a subset of . For example , if the environment is distant enough , then the visual stimulus does not depend on the forward velocity . Therefore , even though the spatial configuration is at least 12-dimensional , actually the stimulus depends on a smaller variable , i . e . , the reduced configuration . We will show that it is possible to identify all unknowns in this model . In particular , we will identify how the feature depends on the reduced configuration and how the rates depend on the feature . Remarkably , it is possible to do this without assumptions on how the feature is computed from the stimulus or how the stimulus depends on the reduced configuration . We only assume to be able to observe the reduced configuration and the generated saccade events . Before describing the method , we first discuss how this model based on rate-variant Poisson processes allows us to represent different functional models . In Figure 3 we illustrate the predictions of four qualitatively different functional models in terms of the observed statistics . On the left side we show the functional model , and on the right we show the expected observed event rates as a function of the feature . This exercise assumes that we know how to estimate the feature , which we will show later . Here we describe what we would expect to find , before embarking on the actual computation of . Figure 3A shows a “hard threshold” model , based on the computation of a single feature , which is then thresholded to obtain the event rate . A Poisson process then generates the events based on this time variant rate . The “stochastic trigger” in the figure masks the fact that there are two processes generating two classes of events , and that these processes are interacting ( see discussion above ) , which is not relevant to the present discussion . If the absolute value of the feature is below a threshold , no event is generated; otherwise , saccades to the left and right are generated at a fixed rate . A large fixed rate would mean that the model is practically deterministic , with a large stimulus feature resulting in a behavioral event with only rare failures . On the right side of the figure , we show the observed event rates as a function of the feature . For this simple model , the observed rates as a function of the feature are straight steps . We remark that our analysis does not assume necessarily that the feature exhibits a hard threshold as in this simple model . We choose this shape merely because it allows visualizing the effect of different sources of noise . In particular , we are interested in understanding the implications of a noise source that acts on the computation of the feature ( sensory noise ) compared to noise that generates behavior in a parallel process independent of the stimulus-computed feature ( decision-making or motor noise ) . Figure 3B shows the effect of measurement noise on the hard threshold model . Random fluctuations in the feature turn the hard threshold into a soft threshold . Figure 3C shows the effect of adding a spontaneous generation process in parallel to the feature pathway . This has the effect of raising the predicted event rate by a constant value , as the parallel process is independent of the feature . A parallel generation process that depended on an unobservable internal state would have the same expected statistics if the internal state is uncorrelated with the feature . This means that a constant baseline event rate that is independent of the feature must be interpreted as the joint contribution of a purely stochastic spontaneous event generation together with a deterministic response based on internal states . It is also important to consider the effect of another unmodeled feature on the event rate statistics , if we only model the dependence of one feature . This stems primarily from practical concerns , because the dimensionality of the feature that it is possible to identify depends primarily on the amount of data available . Therefore , once the dimensionality of the feature is fixed , we need a way to judge whether that dimension is sufficient to describe the behavior . Figure 3D augments the model of Figure 3A with an additional pathway that uses a different feature . In such a case , if we plot the rates versus the feature , we will not find a clear functional dependency , indicating that the feature is no longer sufficient to explain the event rates . Conversely , if we find a clear functional dependency , then we can say that the feature is sufficient to capture the influence of the sensory stimulus on the behavior . This does not imply that is the only behaviorally relevant feature of the stimulus , because there could be other features that are relevant for other behaviors not considered in the analysis . Our identification algorithm , described in the next section , recovers the best one-dimensional feature that explains the event rates . This permits constructing a function in which the experimental event rate is plotted against the feature curve . However , we anticipate that the experimental results , being dependent on experimental data , will have error bars both for dependent and independent variables . Strictly speaking , even if one finds a one-dimensional feature that uncovers a deterministic dependency between feature and rates compatible with the error bars , it is not possible to conclude that there is only one feature , because the effect of a second feature might be masked by the measurement noise . In this sense , our claims that one feature is sufficient is an application of parsimony . In summary , we can identify the contributions of several qualitative factors by plotting the event generation rates as a function of . Measurement noise will soften the curve ( e . g . , a hard threshold is turned into a soft threshold ) . A parallel purely stochastic event generation process has the same effect of a deterministic process based on an internal state uncorrelated with the feature , namely it raises the curve by a fixed baseline rate independent of . If another unmodeled feature influences the behavior , there is not a strict functional dependence between the rates and the feature . We devised a procedure that obtains an estimate of the best one-dimensional feature of the input that predicts the observed event rates . We explain here the basic idea , and provide details in Text S1 . Intuitively , the feature and event rates can be obtained from the spatial statistics of the observed behavioral output . With respect to the discussion so far , the main conceptual step consists in translating the problem from the time to the space domain . So far , we have written the feature as a time-varying quantity . We have also assumed that depends on the stimulus , and that the stimulus depends on the animal spatial configuration , or more precisely , on the reduced configuration . Therefore , we rewrite our model writing instead of . The quantity is a spatial field that we interpret as the feature computed from the typical stimulus experienced at the reduced configuration . We will fit a model of the kind: ( 2 ) where is the average event rate for the -th class ( : left , R: right ) observed at the reduced configuration ; denotes the event generation rates for left and right saccades as a function of the feature , and is constant term that we call baseline event rate . Note the differences with respect to the previous model ( Eq . 1 ) . First , we have written the rates as a function of the reduced spatial configuration instead of time . Moreover , we do not model explicitly the contribution of the internal state . As argued above , given that we cannot measure the unobservable internal states , we cannot distinguish between a purely stochastic contribution and the contribution of an internal state Therefore , the constant term will be an estimate of the joint contribution of the two terms that we cannot distinguish: ( 3 ) where indicates the expected value taken over the whole trajectory . We summarize here the three main phases for estimating from the behavioral data , while leaving the details to Text S1 . First , the reduced configuration space is discretized into spatial cells with a resolution that depends on the amount of data available . For each of these cells , basic statistics are computed , such as the average time spent in each cell , as well as the observed event rates in the cell . One advantage of the algorithm is that these spatial statistics , averaged over the whole trajectory , are intrinsically robust to measurement noise and uncertainty in the event detection algorithm . Next , the event generation rates are computed from the observed rates . Because we assume that the Poisson processes interact with each other , and therefore the statistics of each process cannot be processed separately , and appropriate steps are required to take into account the interaction . Once the average event generation rates are estimated , then we find the feature field that explains both event generation rates , in the sense that there exist two functions and such that the constraint described by equation ( 2 ) holds . Writing the constraint explicitly for each cell , and letting the value of the feature to estimate , we can see that we have a system of constraints of the kind:The generated event rates and on the left side have already been estimated , while both the feature value and and have to be estimated . The constants and can be incorporated as part of and . Note that this can be interpreted as a dimensionality reduction problem , because we have to find one cause ( the feature ) that explains two effects ( left and right event rates ) at the same time . In our case , we solve a relatively simple instance of the problem in which is assumed to be a scalar function . Therefore , the constraints can be algebraically manipulated to obtain a closed form solution , which also takes into account the uncertainty in all the data and provide error bars for the estimated feature . The details are given in Text S1 . Our approach is very generic , and can be extended to scenarios with more than 2 behaviors and more than 1 feature . The feature should be considered a dimensionless quantity of arbitrary scale . In fact , the equations that define it have multiple solutions . For example , suppose that is one solution of the system of equations given by ( Eq . 2 ) . If is any invertible function , then one can verify that is a solution as well . Therefore , once we have obtained a solution for , we can rescale it using any function that we find convenient . In the following , we choose the rescaling function such that is uniformly distributed in the interval .
We tracked 88 flies for a total of 5130 seconds or approximately 1 . 4 hours . Of the total recorded time , we considered only the 4814 seconds of data in which the flight speed exceeded 5 centimeters per second . This threshold on the linear velocity allowed working on tracks for which saccades were easier to detect . We detected a total of 6613 saccades with this criterion , giving an average saccade rate of 1 . 37 saccades per second . We chose a reduced configuration that is two-dimensional . This follows from considering only planar motion ( which reduces the effective degrees of freedom to 3 ) , and using the symmetry of the circular arena ( which reduces the degrees of freedom to 2 ) . An implied assumption ( which can be verified a posteriori ) is that the fly's response is not dependent on the variables not considered in the analysis; for example , even though it is known that flies [64] and other insects [65] use gaze to stabilize vision , there is no gaze variable in our model . This is because the resolution of our measurements is not enough to observe directly the relative pose of head and body , in terms of pitch , roll , or yaw . All components of the spatial configuration that are theoretically relevant for the stimulus , but cannot be measured , are “hidden” states whose contribution is lumped into the constant term in ( 2 ) . The two-dimensional reduced configuration can be parameterized in different ways , the results being independent of the particular parameterization . The primary parameterization that we use for computation uses for coordinates: is the distance to the wall and is the angle that the fly heading forms with respect to the axis of the arena ( Figure 1Ci ) . We chose this parameterization because it corresponds to two behaviorally relevant variables . We preferred the axis angle over other potentially valid representations for the heading ( e . g . , approach angle ) because the representation is not singular , as for any value of . We compute all statistics in the space , but we also use another choice of coordinates to visualize the same data . We rotate the original configuration of the fly around the center of the arena , such that the new coordinates are . These “fly-centric” coordinates are displayed using a top-down view of the arena , in which the fly always points up ( Figure 1Cii ) . The reduced configuration was discretized in a grid with sides of 36 cells ( for ) and 20 cells ( for d ) ( Figure 1Di ) . The angle was discretized in 36 cells of equal size 10 deg . The distance was discretized in 20 unequal intervals ( note the unequal axis in Figure 1Di ) . Intervals for are smaller at the center of the arena and larger near the border , in such a way that each annulus of radius and width had the same area . To compensate for the sparseness of the data , each cell extends 50% into the neighbor's area . Although these choices were somewhat arbitrary , we obtain qualitatively similar results if we vary the number of the cells . Figure 1D shows the distribution of time spent at each point of the arena , and Figure 1E shows the distribution of the detected saccades using the GSD algorithm ( see Text S1 for figures using the saccades detected by the alternative AVSD algorithm ) . As clearly evident in Figure 1Eii , most of the detected saccades correspond to the fly avoiding the walls on the left or on the right . However , those are the configurations where the flies spent more time ( Figure 1D ) . Therefore , we need to normalize this data to see the behavioral patterns . Figure 4A shows the estimated saccade generation function across the reduced configuration space . These rates are obtained by first computing the observed generation rates by averaging the number of saccades ( Figure 1E ) by the time spent in each cell ( Figure 1D ) . Then the rates are obtained from by correcting for an estimated inhibition interval s . Panels B and C show the data separately for left and right saccades ( and ) . The most evident phenomenon is that the fly tends to turn left when the wall is on the right ( and vice versa ) , however , there are many saccades of the opposite direction initiated , even when the turning would orient the fly towards the wall rather than away from it . This is the phenomenon that we want the feature to explain: we want to find the best spatial scalar value such that both and can be written as a function of . Figure 5Ai-ii shows the estimated feature as a function of the reduced configuration c . This is the unidimensional feature that best explains both the left and saccade rates . The estimated feature using the alternative saccade detector is qualitatively similar ( Figure S1 ) . We now have the spatial feature as well as the rates as a function of the reduced configuration and can now plot as a function of ( using as an implicit variable ) . This is shown in Figure 5B , which shows , for each cell , the value of as a function of . Figure 5Bi shows the data as a scatter plot , while Figure 5Bii shows the error bars on the estimated rates at the 95% significance level . The data in Figure 5B indicate the predictive power of the feature . If the feature was perfectly predictive of the event rates , then would be a function of . In this case , taking into account the error bounds on the rates , it is possible to find two functions that predict the event rates in approximately 93% of the environment . More specifically , given a generic cell corresponding to the spatial configuration we find that the predicted event rates and are compatible with the observed rates and at the 95% level of significance . In practice , this means that the data in Figure 5B can be explained by two smooth curves ( and ) that intersect 93% of the confidence intervals corresponding to each spatial configuration . In the remaining cells ( 51 of 720 cells ) , the rates cannot be predicted by this feature alone . Further inspection ( data not shown ) reveals that such points correspond to configurations with the fly pointing directly against the wall at a small distance ( <0 . 3 m ) . Note that being able to predict the rates from the feature does not mean that one is able to predict the direction of each single saccade event . For example , in the middle of the arena the probability of left and right saccade is 50% , and this percentage is perfectly predicted by the feature; however , it is impossible to predict the direction of the single saccade better than chance . As explained before , using only external observations of the animal spatial configuration , we cannot distinguish among the contribution of a purely random endogenous saccade generation process , a deterministic process based on an internal state , any unmodeled features computed from the stimulus , and any unobservable spatial configuration that we cannot observe due to the limited resolution of our instruments . These contributions are lumped together in a baseline saccade rate . By examining the curves in Figure 5Bi we can estimate a baseline event rate of about saccades/sec . By comparing with a maximum estimated event rate of saccades/sec , we can estimate that roughly 90% of the saccades are stimulus-driven in the regions of maximum stimulus . This value depends on the geometry and texture of this particular arena ( e . g . , it would be different if the arena was larger or smaller ) . However , we predict that the baseline rate of saccades/sec that we measure at the center of the arena should be independent of the geometry , as the size and textures of this arena were chosen such that the fly cannot perceive significant visual contrast from the center . We can make some informed guesses for the contribution of the various possible processes by considering circumstantial evidence from other experiments . In tethered flight experiments , deliberately performed in the absence of salient visual stimuli , spontaneous saccade rates are on the order of saccades/sec [66] . If we assume that these values obtained in tethered experiments are a good approximation of an assumed spontaneous generation process in free-flight , then we can account for approximately 75% of the unexplained 0 . 4 saccades/sec as the joint contribution of a random process and unobservable internal states . This leaves roughly 25% of unexplained data , which could possibly be explained by estimating an additional feature , perhaps dependent on components of the spatial configuration that we cannot observe , such as the gaze direction . The contribution of a hypothetical feature is therefore very small with respect to the contribution of the estimated , as could possibly explain about 0 . 1 saccades/sec versus the 4 saccades/sec explained by . We conclude that the saccade behavior of Drosophila that depends on external visual stimulus appears to depend for the most part on only a one-dimensional feature of the stimulus . These conclusions must be limited to the particular experimental condition , as we cannot exclude that more complex environments would elicit more complex responses that require a higher dimensional feature to be explained . However , even in our relatively simple flight environment , our analysis implies that the vast majority of saccades we observed are stimulus-driven and are not due to an internal , stimulus-independent search algorithm ( e . g . Levy flights ) , as has been suggested for Drosophila and many other species [6]–[18] . We have been able to compute the feature field from the observable fly trajectory , without any assumptions on the fly visual processing . Nevertheless , it is interesting to test whether this independently identified feature is compatible with existing models of the first stages of visual processing in flies . In particular , we test the hypothesis of whether the identified feature can be expressed as a linear function of the perceived optic flow . We assume the following generative model for : ( 4 ) where is the optic flow , or angular velocity , at the retinal angle at time , and is a retinal input kernel . The value corresponds to the animal's center front visual field . The function is an arbitrary nonlinear function that we include in the model , because the identification procedure allows us to know only up to a monotone transformation ( i . e . , if is a solution of the constraints system , then also is a valid solution ) . We can characterize the optimal as the solution of an optimization problem: ( 5 ) where the error function is given by: ( 6 ) In this last expression , is the typical optic flow that the animal experiences at the reduced configuration . By solving this optimization problem , we try to best approximate the estimated feature over the whole environment , assuming it can be expressed as a linear function of the optic flow . Unfortunately , we found that this optimization problem is ill posed given our data . In particular , is known only at a discrete set of values ( 720 cells — the density of these is constrained by the finite amount of data that we have ) , and it is quite noisy , whereas the unknowns are of high dimension . Given that the resolution of the fly's visual system is around deg , it makes sense to use at least 70 numbers ( ∼330/5 ) for representing . Furthermore , can be any monotonic nonlinear function . We tried to improve the results by penalizing large values and large spatial variations of ( measured either by the spatial derivatives or ) . The modified error function is: ( 7 ) for and different values of . In general , by varying and , we found a multitude of solutions , all very different from each other , having approximately the same predictive power ( Figure 6 ) . We noticed that for increasing regularization values the estimated linear kernel tended to be shaped as an harmonic function , as illustrated by the kernel obtained by regularizing the second derivative ( ) and using a large value of ( ) , shown in Figure 6D . This kernel is still asymmetric . If we impose that the kernel must be symmetric , we find that the best approximation using one harmonic is: ( 8 ) This kernel and relative feature field is shown in Figure 6E , and it is a good approximation of the feature estimated from the data . We conclude that the identified feature can be expressed as a simple function of the optic flow . However , while obtaining the behaviorally relevant feature from the external observations alone is a well-posed mathematical problem , finding the function that maps the stimulus to the feature is an ill-posed problem , because the set of possible models is of very high dimension compared with the data that we have . Note that these issues are already evident when considering only linear functions of the optic flow , and would be even more pressing if we were to add other nonlinear components to the model that are known to exist in the neural circuits of the fly . Most of the estimated kernels obtained using some form of regularization share a particular feature: is never 0 for corresponding to the back of the animal , but has opposite sign in the front of the animal ( Figure 6B , C , D , E ) . Further investigation shows that these non-zero values in the caudal region are responsible for the two small side lobes that appear in the feature field when plotted in fly-centric coordinates . If the kernel is set to zero in the back , these side lobes disappear . This is apparent by comparing the feature field in Figure 6F ( corresponding to the kernel ) with that in Figure 6E . These results suggest that the optic flow in the back of the animal influences the fly's turning decisions . This response cannot be interpreted as pure obstacle avoidance , given that flies tend to fly forward and obstacles in the back are not expected to represent a threat for collision . For convex environments , the saccades initiated from this response would tend to align the fly's course in parallel to the environment boundaries and the overall result is to follow walls rather than completely avoid them ( similar behavior has been observed in bees [67] ) . Such a visuo-motor system might provide a functional advantage with respect to the balance of collision avoidance and object search . An animal that balances attraction and obstacle avoidance would tend to remain relatively close to interesting visual features , whereas an animal whose primary reflex is to fly away from visual features would tend to find itself in large open areas , far from potential landmarks or food sources . The only way of quantitatively verifying this attraction-deflection hypothesis would be to obtain data from experiments within larger environments with more varied visual features . These results are also compatible with the observation in previous experiments on tethered flies that the optomotor response can be written as the function of a kernel in which the rear and front visual fields give opposite contributions [68] , suggesting that a similar visual feature might be used for both behaviors .
In this paper , we introduced a novel method to obtain an estimate of a low-dimensional feature of the stimulus that best predicts the observable behavioral event generation rates . The feature can be obtained from observable quantities , such as the recording of the trajectory of the animal , without any assumption on the nature of the stimulus and its underlying neural processing . Using this method , we have concluded that most of the saccade events generated by fruit flies exploring a structured laboratory environment are induced by visual stimuli , and that the instantaneous stimulus can be compressed down to a one-dimensional feature , while still being predictive of the event rates in % of the environment . Using this method , it is not possible to distinguish between the contribution of an endogenous random process and a deterministic contribution dependent on an unobservable state . However , we can bound the contributions of these two terms in a baseline saccade rate that we estimate at 0 . 4 saccades/sec , roughly a tenth of the maximum rate . The strength of this method is that the feature can be estimated working backwards from an animal's actions , rather than forward by postulating a model for the stimulus and guessing what is the relevant feature . Once we know , as a second step , it is possible to attempt to fit a parametric representation of neural processing to find the forward function from to , based on other assumptions about sensory processing , though this is not guaranteed to be a well posed problem , as one must optimize over all plausible models compatible with the animal's biology . In this particular case , we have shown that the feature responsible for turning decisions in Drosophila can be written as a linear function of the optic flow , and that the particular linear kernel we obtain is compatible with that identified in tethered conditions , for a particular choice of regularization penalty to make the problem well posed . Conversely , finding from the behavioral data is a well posed and intuitive problem , because it can be understood as a dimensionality reduction problem ( find the one feature that explains multiple behaviors ) . The main advantage of this approach , compared with previous methods , is that it can be applied to freely moving animals , and thus permits asking about responses to naturally important stimuli . Moreover , it does not need any assumption of linearity between some aspect of the stimulus and response , a precondition strongly needed in techniques such as reverse correlation [48] . Even advanced reverse correlation techniques in single sensory neurons [69] are not easy to generalize into models of network functionality that could be used to predict behavior . In the future , this method could be applied to different behaviors of the fruit fly and other animals [70] . The formalization is quite generic , though some generalizations are possible . The algorithm documented in Text S1 assumes that the feature is one-dimensional in order to obtain a closed-form solution . To identify a feature of higher dimension , this must be generalized , for example by using one of the various more computationally expensive dimensionality reduction algorithms in machine learning ( e . g . , [71] ) . In any case , the rate-variant interacting Poisson process model seems apt for modeling many other behaviors ( e . g . , landing , taking off ) that can be reliably localized in time ( i . e . , they have a clear beginning and end ) , and that can be caused by both external and internal causes . Thinking in terms of the feature as a proxy of the stimulus can potentially be useful in understanding how different sensory modalities contribute to the same behavior . The feature is independent of the sensory modality because it is just a function of the animal configuration , and it is a proxy of the typical stimulus perceived at the location , so it could be used to study , for example , the influence of olfaction instead of visual processing on turning behavior , or their interaction , which has been the object of much research [72]–[75][56] . Note , however , that we do have the strong assumption that the stimulus is a constant function of the configuration , so the framework cannot be easily extended to time-varying stimuli . This approach might also be useful to study different behaviors at the same time . Drosophila has a large repertoire of behaviors/reflexes which are stimulus-triggered , such as landing , take-off , chasing mates , and escaping from small targets . In this case , we focused on saccades , and we found the feature encoding the relevant function of the stimulus for saccade decisions . If one repeated the analysis for a different behavior ( e . g . , landing ) , there would likely be another feature , that would be different from . However , if this was repeated for all fly behaviors , one would find that at some point the new identified features would be redundant; for example , in the case of vision , the number of features is upper bounded by the number of upstream signals towards the lobula . Ultimately , this exercise might provide a prediction of whether two behaviors are likely to share the same neural pathways . Potentially , this technique could help in quantifying the behavioral differences of different genotypes . This model makes a distinction between the feature and the event generation rate functions . Whereas z is assumed to be correlated with computed percepts , might be correlated more with the motor functions . This distinction could be used to obtain insight regarding the function of genetic manipulations such as a screen in which populations of neurons are “silenced” with a hyperpolarizing ion channel or synaptic release blockade . For example , if a modified animal gives the same feature but modified rate functions , it would be evidence that the silenced neurons are involved with motor generation rather than with stimulus processing . Consequently , with a large-scale screen [76] , [77] , it might be possible to obtain a classification of phenotypes into sensory , decision making , and motor deficits . Similarly , we could use this feature to quantitatively compare the properties of different species . Another interesting but more substantial extension of this work would be to expand the mathematical formalism to incorporate measurements of neuronal activity into the internal processing structure . This is now done in freely moving worms [78] , [79] , [32] and zebrafish [80]; in adult flies , most neural recording during behavior is being done on fixed flies [81]–[83] , [42] . | Researchers have spent considerable effort studying how specific sensory stimuli elicit behavioral responses and how other behaviors may arise independent of external inputs in conditions of sensory deprivation . Yet an animal in its natural context , such as searching for food or mates , turns both in response to external stimuli and intrinsic , possibly stochastic , decisions . We show how to estimate the contribution of vision and internal causes on the observable behavior of freely flying Drosophila . We developed a dimensionality reduction scheme that finds a one-dimensional feature of the visual stimulus that best predicts turning decisions . This visual feature extraction is consistent with previous literature on visually elicited fly turning and predicts a large majority of turns in the tested environment . The rarity of stimulus-independent events suggests that fly behavior is more deterministic than previously suggested and that , more generally , animal search strategies may be dominated by responses to stimuli with only modest contributions from internal causes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"behavioral",
"neuroscience",
"biology",
"computational",
"biology",
"neuroscience"
] | 2013 | Discriminating External and Internal Causes for Heading Changes in Freely Flying Drosophila |
Hypertension is a leading cause of global disease , mortality , and disability . While individuals of African descent suffer a disproportionate burden of hypertension and its complications , they have been underrepresented in genetic studies . To identify novel susceptibility loci for blood pressure and hypertension in people of African ancestry , we performed both single and multiple-trait genome-wide association analyses . We analyzed 21 genome-wide association studies comprised of 31 , 968 individuals of African ancestry , and validated our results with additional 54 , 395 individuals from multi-ethnic studies . These analyses identified nine loci with eleven independent variants which reached genome-wide significance ( P < 1 . 25×10−8 ) for either systolic and diastolic blood pressure , hypertension , or for combined traits . Single-trait analyses identified two loci ( TARID/TCF21 and LLPH/TMBIM4 ) and multiple-trait analyses identified one novel locus ( FRMD3 ) for blood pressure . At these three loci , as well as at GRP20/CDH17 , associated variants had alleles common only in African-ancestry populations . Functional annotation showed enrichment for genes expressed in immune and kidney cells , as well as in heart and vascular cells/tissues . Experiments driven by these findings and using angiotensin-II induced hypertension in mice showed altered kidney mRNA expression of six genes , suggesting their potential role in hypertension . Our study provides new evidence for genes related to hypertension susceptibility , and the need to study African-ancestry populations in order to identify biologic factors contributing to hypertension .
Genetic studies hold the promise of providing tools to better understand and treat clinical conditions . To achieve the clinical and public health goals of reducing hypertension and its sequelae , and to understand ethnic disparities in the risk for hypertension , there is a need to study susceptible populations for genetic determinants of blood pressure ( BP ) . BP traits are highly heritable across world populations ( 30 to 55% ) . [1–4] Over 200 genetic loci have been identified in genome-wide association studies [5–13] and admixture mapping studies . [14–17] These variants explain approximately 3 . 5% of inter-individual variation in BP . [5 , 7] However , there is still a paucity of studies focused on individuals of African descent . Most of the loci identified in the literature have not been replicated in individuals of African ancestry . [18 , 19] African Americans have higher mean BP , an earlier onset of hypertension , and a greater likelihood to have treatment-resistant hypertension than other ethnic groups . [20–23] Emerging research on Africans shows increasing prevalence of hypertension in urban African communities [24 , 25] which are more Westernized than rural African communities and , so , more closely resemble communities in which African Americans live in the U . S . Hypertension contributes to a greater risk of coronary heart disease , stroke , and chronic kidney disease . [26–30] African Americans experience increased risk of these hypertension-related outcomes [31–34] but the underlying mechanisms , whether environmental exposures or increased genetic susceptibility , are unknown . We hypothesized that additional variants associated with BP can be identified in people of African ancestry; some variants may be African-specific , as has been observed for multiple traits , including kidney disease [35] and metabolic syndrome . [36 , 37] Other variants may be identified in novel loci based on a higher frequency of risk alleles in this population . We used high density imputed genotypes from the 1000 Genomes Project ( 1000G ) to expand the genome coverage of genetic variants so that we could examine the evidence for association with BP traits . Here , we report three novel loci associated with BP which are driven by variants that are common in or unique to African-ancestry populations . Through bioinformatics and experimental evidence of kidney gene expression in mice submitted to angiotensin-II ( Ang II ) induced hypertension , we provide evidence for a key role of these genes in the pathogenesis of hypertension . In addition , our study extends the discovery of BP loci to genes related to kidney and the immune systems , and provides biological relevance for these loci to BP regulation .
Study-specific genomic-control inflation ranged from 0 . 98–1 . 06 ( S3 Table , S1 Fig ) and the linkage disequilibrium ( LD ) score regression intercepts of the single-trait BP meta-analyses calculated by the LD score regression approach ranged from 1 . 02–1 . 04 . [38] These results suggest well-controlled population stratification . The single-trait BP meta-analyses identified several genome-wide significant single nucleotide polymorphisms ( SNP ) at eight loci ( P < 5 . 0×10−8 , systolic BP ( SBP ) : three loci , four SNPs; diastolic BP ( DBP ) : three loci , three SNPs; pulse pressure ( PP ) : three loci , four SNPs; and hypertension ( HTN ) : one locus , one SNP ) , with the EVX1/HOXA locus identified for SBP , DBP and HTN ( S2A–S2D Fig ) . When combining summary statistics for SBP , DBP , and HTN using the multi-trait approach CPASSOC , [39] we identified one locus by the multi-trait statistic SHom ( EVX1/HOXA ) and six loci by SHet ( ULK4 , TCF21 , EVX1/HOXA , IGFBP3 , CDH17 , ZNF746 ) at P < 5×10−8 ( S2E and S2F Fig ) . Note some loci overlap between single-trait and multi-trait findings . We observed 264 variants with P < 1×10−6 for either single- or multi- trait GWAS and these variants were further analyzed by conditional association on the most associated SNPs at each locus ( S4 Table ) . These analyses resulted in 72 independent associations , which included 58 SNPs with minor allele frequency ( MAF ) ≥ 0 . 05 and 14 with low frequency variants ( 0 . 01< MAF < 0 . 05 ) ( S5 Table ) . Among these 72 variants carried forward for trans-ethnic replication , nine variants , all low frequency variants ( MAF<0 . 02 ) , were not available in replication cohorts because they were either monomorphic in the replication population or had a low imputation quality , reducing our replication effort to 63 variants ( S6 Table ) . Eleven independent variants at nine loci were significantly associated with BP traits at P < 1 . 25×10−8 in the combined discovery and replication analyses and are reported in Table 1 . This significance level was determined by adjusting for two independent traits for SBP , DBP , PP and HTN , and two tests of multiple trait analysis . This includes six variants that reached significance level at discovery stage ( P <5 x10-8 ) . Two loci were identified only through multi-trait analyses ( FRMD3 , IGFBP3 ) . Three of these nine loci are novel: TARID /TCF21 , FRMD3 , and LLPH/TMBIM4 ( Fig 2A–2C ) . Four loci ( ULK4 , PLEKHG1 , EVX1/HOXA cluster , and GPR20 ) have been reported in our previous BP GWAS of African ancestry ( S3 Fig ) , [7 , 18] and two loci ( IGFBP3 , CDH17 ) have been reported in multiple-trait analyses of African-ancestry studies ( Fig 2D–2F ) . [39] A composite genetic-risk score using the eleven variants identified accounted for 1 . 89% , 2 . 92% , 1 . 03% and 1 . 08% of the variance for SBP , DBP , PP and HTN respectively . Five of the eleven replicated variants are common in individuals of African ancestry but rare or monomorphic in individuals of non-African ancestry ( rs76987554 , rs115795127 , rs113866309 , rs7006531 , and rs78192203 ) ( Table 1 ) . These five variants were 1 ) either low frequency or common variants in COGENT-BP African-ancestry samples; 2 ) low frequency in 1000G Phase I Integrated Release Ad Mixed-American ancestry ( AMR ) ; and 3 ) monomorphic in 1000G Asian ancestry ( ASN ) or European ancestry ( EUR ) . One common variant was present in only 1000G samples of African ancestry ( rs115795127 at FRMD3 , Table 1 ) . These variants were located at the three novel loci ( TARID/TCF21 , FRMD3 , and LLPH/TMBIM4 ) . Given the differences in allele frequency across continental-ancestry populations , we examined the evidence for selection at each of these loci using iHS , which measures the amount of extended haplotype homozygosity at a given SNP along the ancestral allele relative to the derived allele . [40] The iHS score for rs115795127 was 2 . 7 in African American samples from the Candidate-gene Association Resource ( CARe ) consortium ( see Methods ) , suggesting selection at the FRMD3 locus ( S7 Table ) . We observed two independent genome-wide significant variants at the EVX1/HOXA locus ( P < 1 . 25×10−8 ) . The two variants , rs11563582 and rs6969780 , are in weak LD ( r2 = 0 . 21 ) ( S3A–S3C Fig ) , and the LD pattern suggests that these SNPs are located in two blocks ( S4 Fig ) . SNP rs11563582 is in strong LD with the previously reported SNP in the region ( rs17428741 ) . [18] SNP rs6969780 remained significant when conditioning on rs11563582 ( S4 Table ) , thus demonstrating the presence of allelic heterogeneity at this locus . Two independent variants at ULK4 reached the significance threshold: rs7651190 and rs7372217 ( LD r2 = 0 . 15 ) ( S4E Fig ) . SNP rs7372217 is in strong LD with the previous reported SNP rs1717027 . [18] The association evidence of rs1717027 can be explained by rs7372217 but not by rs7651190 in conditional analysis ( S4 Table ) . Thus , rs7651190 is an independent association at this locus . At the GPR20 locus , our most significant SNP , rs78192203 , is 8kb away and it is not in LD with the published SNP , rs34591516 ( r2 = 0 . 008 , D’ = 0 . 68 in African American CARe participants ) .
We performed functional annotation and cell type group enrichment analysis using the stratified LD score regression approach which uses data from ENCODE and the Roadmap Epigenetic Project , as well as GWAS results while accounting for the correlation among markers . [42] We estimated functional categories of enrichment using an enrichment score , which is the proportion of SNP-heritability in the category divided by the proportion of SNPs . We identified super enhancer ( PEnrich = 5 . 4×10−5 , Enrichment = 5 . 6 for DBP ) , enhancer ( PEnrich = 4 . 8 ×10−4 , Enrichment = 4 . 3 for HTN ) , and H3K27ac ( PEnrich = 3 . 2×10−4 , Enrichment = 3 . 6 for HTN ) significant enrichment ( Fig 3 ) . These results support a role of identified noncoding regulatory regions in BP regulation . In addition , the following cell types showed significant enrichment ( P ≤ 2 . 5 × 10−3 ) : the immune ( PEnrich = 1 . 4×10−9 , Enrichment = 8 . 4 for DBP ) , kidney ( PEnrich = 5 . 4×10−5 , Enrichment = 4 . 8 for DBP ) , and cardiovascular ( PEnrich = 8 . 9×10−5 , Enrichment = 4 . 2 for SBP ) systems ( Fig 3 ) . We next determined the enrichment of variants at the eleven genome-wide significant loci for DNase l hypersensitive ( DHS ) sites in 34 tissue categories from ENCODE . At each locus , we identified variants in r2>0 . 1 with the index variant and calculated causal evidence ( Bayes Factors ) for each variant . We then tested for enrichment in the causal evidence of variants in DHS sites using fGWAS . [43] We found enrichment of blood/immune DHS ( Enrichment = 3 . 1 ) and cardiovascular DHS ( blood vessel Enrichment = 28 . 7 , heart Enrichment = 2 . 0 ) , in addition to DHS in several fetal tissues ( S5 Fig ) . Candidate causal variants at several loci overlapped enriched DHS sites . For example , at the LLPH/TMBIM4 locus , the most likely causal variant , rs12426813 , overlaps a DHS site active in immune ( CD14+ , CD4+ , CD34+ ) , blood vessel ( HMVEC ) , and heart ( HCF ) cells ( S5 Fig ) . To examine whether the eleven significant SNPs are eQTL , we searched the genotype-tissue expression ( GTEx ) pilot database , which includes non-disease human tissue . [42] Among the eleven SNPs , three SNPs have been identified as eQTL: rs6969780 ( HOXA2 ) , rs7651190 ( ULK4 ) , and rs62434120 ( PLEKHG1 ) ( S9 Table ) . SNP rs6969780 is an eQTL for expression of HOXA2 , HOXA7 , HOTAIRM1 , and HOXA5 in multiple tissues , including esophagus , artery , lung , skin , nerve , adipose , skeletal muscle , and stomach tissues . SNP rs7651190 is an eQTL for ULK4 and RPL36P20 in artery , whole blood , thyroid , nerve , esophagus , skeletal muscle , skin , brain , and stomach cells/tissues . SNP rs62434120 is an eQTL for PLEKHG1 in testis tissue . To determine if identified genes are functionally involved in BP regulation in the kidney during hypertension , [44] we quantified gene expression in mice kidneys at baseline and during the hypertensive state induced by Ang II . This hypertensive model was chosen for two reasons: 1 ) to mimic the low plasma renin state , albeit more exaggerated than the level observed , in African-ancestry individuals that has been suggested to reflect the elevated renin-angiotensin system activity at the tissue level in the kidney [45] , and 2 ) maintenance of hypertension in the Ang II model requires activation of the immune system that is implicated in several identified loci . [46 , 47] Kidney gene expressions of the identified genes were compared to age-matched untreated mice after two weeks of Ang II infusion , which increases SBP . For the HOXA locus , we examined the expression of genes that are known to be expressed in the mouse kidney: Hoxa1 ( 2 isoforms ) , 5 , 7 , 9 , 10 ( 2 isoforms ) , and 11 . Among all the genes examined , Tmbim4 was the most abundantly expressed gene in the kidney at baseline . Six genes—Hoxa5 , Hoxa10-1 isoform , Hoxa11 , Tmbim4 , Igfbp3 , and Plekhg1—were significantly differentially expressed in the kidney after Ang II treatment compared to baseline ( Fig 4 ) . Except for Hoxa5 , which showed a significant decrease ( Fig 4A ) , the expression of all these genes increased after the intervention . The expression of six genes—Hoxa1-1 isoform , Hoxa7 , Hoxa9 , Hoxa10-2 isoform , Llph , and Ulk4—were unchanged after Ang II infusion ( Fig 4B ) . The following genes were not expressed in the adult mouse kidney at baseline or after Ang II intervention: Frmd3-1 isoform , Frmd3-2 isoform , Grp20 , Tcf21 , Cdh17 , and Hoxa1-2 isoform .
To date , this is the largest genome-wide analysis of African-ancestry populations to study genetic variants underlying BP traits using dense-coverage imputed genotypes . Our main findings are eleven independent variants at nine loci , significantly associated with BP traits , including three newly identified loci ( TARID/TCF21 , FRMD3 , LLPH/TMBIM4 ) . We also found evidence for additional independent SNP associations in fine-mapping of three previously described loci , ULK4 , EVX1/HOXA , and GRP20 . [18 , 39] The most significant variants at TARID/TCF21 , FRMD3 , GPR20 , and CDH17 are common variants in COGENT-BP African-ancestry participants , but monomorphic or low frequency in non-African-ancestry populations . For example , rs115795127 at FRMD3 is rare in European populations ( MAF = 0 . 0007 ) and absent in East Asian and Hispanic/Latino populations . Therefore , they could not be identified in GWAS of non-African-ancestry populations even when increasing sample sizes . We also show evidence for selection for the variant at FRMD3 , although additional studies should confirm these findings . The African-specific variants were not well tagged by HAPMAP2 data and therefore were not detected in our previous African-ancestry GWAS . [18] Overall , our results suggest additional gain in discovery when using dense imputed genotypes and support a role of population-specific alleles in African and African-admixed populations contributing to BP regulation and hypertension . Furthermore , they support the rationale and the need to study diverse populations in order to more effectively characterize the genetic architecture of BP in populations and the ethnic disparities in hypertension . Functional annotation of our lead variants showed co-localization with annotated elements , including super enhancer , enhancer , and H3K27ac chromatic mapping in immune cells and kidney tissues , which has not been previously reported , in addition to cardiovascular tissues . There was also evidence for regulatory function in these relevant tissues through gene expression regulation ( eQTL ) and through overlaps with DHS in relevant tissues/cells . This evidence was additionally supported by experimental findings of differential expression of six genes ( Hoxa5 , Hoxa10-1 isoform , Hoxa11 , Tmbim4 , Igfbp3 , and Plekhg1 ) in the mouse kidney after HTN induced by Ang II treatment . Overall , our results suggest the functional importance of identified genes in regulating BP in both normal and hypertension states . At the newly identified loci , SNP rs76987554 is an intronic variant in TARID ( TCF21 antisense RNA inducing promoter demethylation ) which has not been previously reported to be associated with BP traits . A nearby gene , TCF21 ( transcription factor 21 ) , is a transcription factor of the basic helix-loop-helix family , which is mainly expressed in the liver , kidney , and heart . TCF21 is involved in epithelial differentiation and branching morphogenesis in kidney development , [48] and was associated with hypertension in a study of individuals of Japanese ancestry . [49] At the chromosome 7 , rs115795127 is an intronic variant to FRMD3 ( FERM domain containing 3 ) which encodes a protein involved in maintaining cell shape and integrity . FRMD3 has been associated with type 1 and type 2 diabetic kidney diseases in different ethnic populations , including those of European , African , and Asian ancestries . [50] The diabetes variant , rs10868025 , is not in LD with rs115795127 in our African American samples or in 1000G EUR samples ( r2 = 0 . 00028 and 0 . 0018 , respectively ) , thus representing an independent association at this locus . At chromosome 9 , the functions of LLPH and TMBIM4 genes in BP regulation are currently unknown . LLPH belongs to the learning-associated protein family and is highly expressed in the immune system and the adrenal gland . TMBIM4 encodes the transmembrane BAX inhibitor motif-containing protein 4 and is highly expressed in whole blood , the immune system , and the adrenal gland . [51] The most significant variant at this locus , rs113866309 , overlaps a DHS in immune , blood vessel , and heart cells . In our experimental model in mice , Tmbim4 gene expression was significantly increased after Ang II-induced HTN . This gene has been shown to inhibit apoptosis[52] and to decrease the efficacy of inositol 1 , 4 , 5-triphosphate ( IP3 ) -dependent release of intracellular Ca2+ . [53] This raises the possibility that the TMBIM4 protein may serve to dampen the effect of Ang II , which activates IP3 in vascular smooth muscle cells through the stimulation of the angiotensin type 1 receptor . [51 , 53 , 54] Therefore , it is possible that in conditions of activated renin-angiotensin system , genetic variants that lower the expression of TMBIM4 may augment BP , whereas genetic variants that increase its expression may attenuate BP . Other genes , such as Hoxa5 , Hoxa10-1 , Hoxa11 , Igfbp3 , and Plekhg1 , were significantly differentially expressed after Ang II-induced HTN in our mice experimental models . The HOXA-cluster has been identified in our previous GWAS of BP in African ancestry and in a recent GWAS of BP in European ancestry[5] though the underlying mechanisms related to BP control are unknown . We identified two independent variants at this locus; further studies are needed to delineate which of the HOXA genes are most likely involved in the association . In our experimental mice model , the Hoxa10-1 isoform had a greater than 20-fold increase in kidney expression during Ang II-induced HTN compared to baseline levels . However , it remains to be determined whether it is an effect of Ang II in hypertension , or a compensatory response to hypertension . Future studies using genetic manipulation in rodents are required to determine whether these changes are specific response related to BP and Ang II or simply a generic response to stress . We identified several additional pathways involved in BP traits , including the GSK3 pathway , which has been reported to influence Wnt-mediated central BP regulation . [55] The Th1/Th2 pathway is involved in the regulation of immune responses[56] and has been linked to hypertension and atherosclerosis . [57 , 58] The role of the immune system in the development of hypertension has been suggested in clinical studies and experimental animal models . [59–64] This includes reports of overlap of genetic variant associations between BP traits and immune-disorders [65] and evidence of enrichment of immune pathways from GWAS of BP . [66] Mutations of SH2B3 , a gene identified in a GWAS of hypertension , have been recently shown to attenuate Dahl salt-sensitivity hypertension through inflammatory modulation . [67] In addition , the actions of Ang II in the pathophysiology and maintenance of hypertension are in part mediated through the activation of the immune system . [46] Our assessment of the clinical implications of identified variants is limited by available data on African-ancestry populations . For example , there are currently no large publicly available GWAS of coronary heart disease or stroke outcomes in African-ancestry populations . It should also be noted that most of our replication cohorts were from populations other than those of African ancestry . Therefore , the power of replication analysis could still be low , which explains why only 11 of 63 variants were successfully replicated . In summary , we report 11 independent variants at nine loci that are potential regulators of BP in our African-ancestry population study . Three loci are new . Identified BP variants are enriched in immune , kidney , heart , and vascular system pathways . Our experimental findings suggest that several of these genes may be involved in the renin-angiotensin pathways in the kidney during hypertension . Further population studies and experimental models are required for a comprehensive assessment of the identified genes across the immune , kidney , and cardiovascular systems . Our study demonstrates the need to further study individuals of African ancestry in order to identify loci and new biological pathways for BP .
Each study followed protocols for phenotype harmonization . For individuals taking anti-hypertensive medications , we added 15 and 10 mm Hg to measured SBP and DBP , respectively , a standard method used in other BP GWAS . [6 , 68] PP was calculated as the difference between SBP and DBP after addition of the constant values . HTN was defined by a SBP ≥ 140 mm Hg , a DBP ≥ 90 mm Hg , or use of antihypertensive drugs . [69] Each cohort was genotyped on either Affymetrix or Illumina genotyping platforms . Pre-imputation quality criteria were applied as described in S2 Table , and included exclusion of individuals with discordant self-reported gender and genetic gender . Imputation was performed using the software MACH-ADMIX , MACH-minimac or IMPUTE2 [70–72] using the Phase 1 integrated ( March 2012 release ) multi-ethnic reference panel from the 1000G Consortium ( http://www . internationalgenome . org/ ) . [73] Autosomal chromosome SNP associations for SBP , DBP , and PP were assessed by linear regression for unrelated data or by the generalized linear mixed-effects model for family data , under the assumption of an additive genetic model . All models were adjusted for age , age2 , sex , and body mass index . Up to ten principal components were included , as needed as covariates in the regression models , to control population stratification . [74 , 75] We used standardized pre-meta-analysis QC criteria for all 21 discovery studies . [76] At the SNP level , we excluded variants with 1 ) imputation quality r2 < 0 . 3 in MACH or <0 . 4 in IMPUTE2; 2 ) the number of informative individuals ( 2×MAF×N×r2 ) ≤ 30; 3 ) an effect allele frequency ( EAF ) difference larger than 0 . 3 in comparison with the mixture of 80% YRI and 20% CEU of 1000G; and 4 ) the absolute regression coefficient ≥ 10 . SNPs that passed the QC were carried forward for inverse variance weighted meta-analyses , implemented in METAL . [77] We applied the CPASSOC software to combine association evidence of SBP , DBP , and HTN . CPASSOC provides two statistics , SHom and SHet , as previously described . [39] SHom is similar to the fixed effect meta-analysis method[77] but accounts for the correlation of summary statistics of the multi-traits and for overlapping or related samples among the cohorts . SHom uses the trait sample size as the weight , so that it is possible to combine traits with different measurement scales . SHet is an extension of SHom , and it can increase the statistical power over SHom when a variant affects only a subset of traits . The distribution of SHet under the null hypothesis was obtained through an estimated beta distribution . To calculate the statistics , SHom and SHet , and to account for the correlation among the traits , a correlation matrix is required . In this study , we used the correlation matrix calculated from the residuals of the three BP traits after adjustments for covariates and principal components . All independent SNPs identified with P < 10−6 ( threshold chosen for suggestive association ) in the discovery stage were carried forward for replication in African-ancestry individuals and in multi-ethnic samples of European Americans , East Asians , or Hispanics/Latinos ( Fig 1 ) . For single-trait analyses , we conducted fixed effect meta-analyses in the replication sets for each of four BP traits ( SBP , DBP , PP and HTN ) , followed by a combined trans-ethnic meta-analysis of each trait . This was followed by a mega-meta-analyses , combining the results of discovery and replication for single traits using fixed-effects meta-analysis . We also performed a multi-trait CPASSOC analysis of SBP , DBP , and HTN in each replication study . Because CPASSOC only generated test statistics SHom/SHet and corresponding P values without effect sizes , we combined the association P values from all four replication populations using Fisher’s method ( http://hal . case . edu/zhu-web/ ) . Finally , we combined the CPASSOC meta-analysis results from the discovery and replication stages using Fisher’s method . For a single trait GWAS discovery analysis , we used genome-wide significant level P = 5 . 0×10−8 . We performed six different analyses , four single trait ( SBP , DBP , PP and HTN ) analyses and two CPASSOC ( SHom and SHet ) analyses for each SNP . For the four single correlated traits ( SBP , DBP , PP and HTN ) , we calculated the number of independent traits using the eigenvalues of the correlation matrix , [78] which resulted two independent traits . Therefore , we counted four independent analyses , which were two independent single traits and two statistics of CPASSOC analyses , and applied an experimental significance level P = 1 . 25×10−8 for claiming a genome-wide significance when combining discovery and replication samples . We should point out that the two CPASSOC test statistics and a single trait statistic are not independent . Thus , the significance level P = 1 . 25×10−8 is conservative . Since a locus may consist of multiple independent signals , we applied approximate conditional analysis implemented in GCTA-COJO[79 , 80] using the summary statistics of SNPs with P < 1 . 0×10−6 from both of the individual trait meta-analyses ( http://cnsgenomics . com/software/gcta/cojo . html ) . The LD among variants was estimated from the five African American cohorts from the CARe consortium . [79] Pathway analysis was performed using the Meta-Analysis Gene-set Enrichment of variant Associations ( MAGENTA ) program ( http://www . broadinstitute . org/mpg/magenta/ ) . [41] Using the summary statistics from the four BP traits and two statistics from CPASSOC , from the discovery stage , we tested whether sets of functionally-related genes are enriched for associations . This method first converts the P values of SNPs into gene scores with correcting for confounders , such as gene site , number of variants in a gene , and their LD patterns , and then calculated a gene set enrichment P value for each biological pathway or gene set of interest using a non-parametric statistical test . The nominal GSEA P value refers to the nominal gene set enrichment P value for a gene set . The database of pathway/gene-sets to be tested include Ingenuity ( June 2008 ) , KEGG ( 2010 ) , GO , and the Panther , signaling pathways downloaded from MSigDB and PANTHER ( http://www . broad . mit . edu/gsea/msigdb/collections . jsp; http://www . pantherdb . org/ ) . [81] We applied the parameters suggested by the authors , which includes the 75th percentile cut off of gene scores , the nominal GSEA P-value < 0 . 01 and the false discovery rate ( FDR ) < 0 . 3 . The enrichment of heritability of genomic regions to different functional categories , including cell type-specific elements , was evaluated using the method of LD score regression ( https://github . com/bulik/ldsc ) . [42 , 82] This method partitioned the heritability from the discovery GWAS summary statistics of four BP traits ( SBP , DBP , PP , and HTN ) while accounting for LD among markers . [42] We calculated enrichment , in functional regions and in expanded regions ( +500bp ) around each functional class , based on functional annotation , using a “full baseline model” previously created from 24 publicly available main annotations that are not specific to any cell type . [42] Enrichment was calculated based on the ratio of explained heritability and the proportion of SNPs in each annotation category . The standard error of enrichment was estimated with a block jackknife to calculate z scores and P values . [42] The multiple testing threshold was determined using the Bonferroni correction while accounting for two independent-trait analyses based on Ji and Li’s method[78] ( P of 0 . 05/[25 classes × 2 traits] ) . We also performed cell-type-specific group enrichment analysis using cell-type-specific annotations from four histone marks ( H3K4me1 , H3K4me3 , H3K9ac , and H3K27ac ) , which corresponded to 220 cell types . We divided the 220 cell-type-specific annotations into 10 groups: adrenal/pancreas , central nervous system ( CNS ) , cardiovascular , connective/bone , gastrointestinal , immune/hematopoietic , kidney , liver , skeletal muscle and other . The analysis characterized cell-type-specific annotations within each group and calculated the enrichment of heritability for each group . [42] We selected sets of variants in LD r2 > 0 . 1 from the eleven replicated variants , and calculated Bayes Factors and posterior causal probabilities for each variant from the effect sizes and standard errors , as previously described . [83] Each distinct variant associated with multiple traits was included in the analysis only once . The genomic annotations of DHS sites for 348 cell types from the ENCODE project were obtained and grouped into cell types associated with 34 tissues ( http://genome . ucsc . edu/ENCODE/cellTypes . html ) . Four gene-based annotations—coding exon , 5-UTR , 3-UTR , and 1kb upstream of transcription start site ( TSS ) —from GENCODE transcripts were also obtained . Variants overlapping each of these annotations were then identified . Using the variant annotations and fGWAS ( https://github . com/joepickrell/fgwas ) , we tested for enrichment of variants across all signals in 38 DHS categories , including in the four gene-based annotations in each model . [43] We used the GTEx pilot database [82] ( http://www . gtexportal . org/home/ ) to identify eQTLs in the successfully replicated SNPs . To evaluate population differentiation and natural selection , using Haplotter , [40] we calculated the integrated haplotype score ( iHS ) in five cohorts of CARe so that we could measure the amount of extended haplotype homozygosity ( http://coruscant . itmat . upenn . edu/whamm/ihs . html ) . Hence , we tested the evidence of recent positive selection at five significant SNPs with differences in allele frequency across continental-ancestry populations . The measures were standardized ( mean 0 , variance 1 ) empirically to the distribution of observed iHS scores over a range of SNPs with similar derived allele frequencies . This method assesses the evidence for selection by comparing the extended homozygosity for haplotypes on a high frequency derived allele relative to the ancestry background . [40] Experiments were carried out in accordance with local and the National Institutes of Health guidelines . The animal protocol was approved by the University of Virginia Institutional Animal Care and Use Committee . Wild-type male mice on the 129S6 background at ~ 3 months of age were used for gene expression analyses . All mice were maintained on a 12-hour light-dark cycle with free access to standard chow and water in the animal facility of the University of Virginia . The hypertension experimental model was induced using Ang II ( Sigma-Aldrich , St . Luis , MO ) delivered at 600 ng/kg/min for 2 weeks via Alzet mini-osmotic pumps ( Durect Corporation , Cupertino , CA , model 2004 ) , as previously described . [84] For gene expression analyses , RNA from kidney tissue was isolated by RNeasy Mini kit ( Qiagen ) and transcribed to cDNA by iScript TM cDNA synthesis kit ( Bio-Rad ) . Real time PCR analyses were performed on iQTM5 Multicolor real time PCR Bio-Rad instruments using iQTM SYBER® Green Supermix . Hprt was used as a reference gene for normalization . Sequences of forward and reversed primers ( FP and RP ) for the gene expression studies are shown in S10 Table . | Hypertension is a global health problem which affects disproportionally people of African descent . We conducted a genome-wide association study of blood pressure in 31 , 968 Africans and African Americans to identify genes conferring susceptibility to increased blood pressure . This research identified three novel genomic regions associated with blood pressure which have not been previously reported in studies of other race/ethnicity . Using experimental models , we also showed an altered expression of these genes in kidney tissue in hypertension . These findings provide new evidence for genes influencing hypertension risk and supports the need to study diverse ancestry populations in order to identify biologic factors contributing to hypertension . | [
"Abstract",
"Introduction",
"Results",
"Pathway",
"analyses",
"suggest",
"enrichment",
"of",
"immune",
"pathways",
"for",
"BP",
"traits",
"Discussion",
"Methods"
] | [
"genome-wide",
"association",
"studies",
"africans",
"medicine",
"and",
"health",
"sciences",
"ethnicities",
"mathematics",
"statistics",
"(mathematics)",
"genome",
"analysis",
"mammalian",
"genomics",
"kidneys",
"research",
"and",
"analysis",
"methods",
"mathematical",
"... | 2017 | Single-trait and multi-trait genome-wide association analyses identify novel loci for blood pressure in African-ancestry populations |
Studies of the relationship between DNA variation and gene expression variation , often referred to as “expression quantitative trait loci ( eQTL ) mapping” , have been conducted in many species and resulted in many significant findings . Because of the large number of genes and genetic markers in such analyses , it is extremely challenging to discover how a small number of eQTLs interact with each other to affect mRNA expression levels for a set of co-regulated genes . We present a Bayesian method to facilitate the task , in which co-expressed genes mapped to a common set of markers are treated as a module characterized by latent indicator variables . A Markov chain Monte Carlo algorithm is designed to search simultaneously for the module genes and their linked markers . We show by simulations that this method is more powerful for detecting true eQTLs and their target genes than traditional QTL mapping methods . We applied the procedure to a data set consisting of gene expression and genotypes for 112 segregants of S . cerevisiae . Our method identified modules containing genes mapped to previously reported eQTL hot spots , and dissected these large eQTL hot spots into several modules corresponding to possibly different biological functions or primary and secondary responses to regulatory perturbations . In addition , we identified nine modules associated with pairs of eQTLs , of which two have been previously reported . We demonstrated that one of the novel modules containing many daughter-cell expressed genes is regulated by AMN1 and BPH1 . In conclusion , the Bayesian partition method which simultaneously considers all traits and all markers is more powerful for detecting both pleiotropic and epistatic effects based on both simulated and empirical data .
Studies in the genetics of gene expression combine gene expression and genotype data in segregating populations to detect loci linked to variations in RNA levels . These loci are referred to as expression quantitative trait loci ( eQTL ) . To date , eQTL studies have been pursued in a number of species ranging from yeast to mouse and human [1]–[3] . A common theme of these studies is to treat thousands of gene expression values as quantitative traits and conduct QTL mapping for all of them . Most eQTL studies are based on linear regression models [4] in which each trait variable is regressed against each marker variable . The p-value of the regression slope is reported as a measure of significance for the association . In the context of multiple traits and markers , procedures such as false discovery rate ( FDR ) controls [5] can be used to quantify family-wise error rates . Despite the success of this type of regression approach , a number of challenging problems remain . First , these methods can not easily assess the joint effect of multiple markers beyond additive effects . Storey et al . [5] developed a step-wise regression method to find eQTL pairs , then Zou and Zeng improved it [6] . This procedure , however , tends to miss eQTL pairs with small marginal effects but a strong interaction effect . There are methods for detecting eptistatic effects without main marginal effects [7]–[8] . However , their applications are limited to a few clinical traits instead of thousands of expression traits due to computational constraints . Second , there are often strong correlations among expression levels for certain groups of genes , partially reflecting co-regulation of genes in biological pathways that may respond to common genetic loci and environmental perturbations [2] , [9]–[11] . Previous findings of eQTL “hot spots” , i . e . , loci affecting a larger number of expression traits than expected by chance , and their biological implications further enhance this notion and highlight the biological importance of finding such gene “modules” . Mapping genetic loci for multiple traits simultaneously is more powerful than mapping single traits at a time [12] . Although for a known small set of correlated traits , one can conduct QTL mapping for the principal components [13] , this method becomes ineffective when the set size is moderately large or one has to enumerate all possible subsets . An alternative approach is to identify subsets of genes by a clustering method , and then fit mixture models to clusters of genes [14] . The eQTL mapping then depends on whether the distance metric used by the clustering method is appropriate , whether the method can find the right number of clusters . We address these issues by modeling the joint distribution of all genes and all markers simultaneously . Under a Bayesian framework , we introduce three sets of latent indicator variables for genes , markers , and individuals , and then systematically infer the association between groups of genes and sets of markers . In this framework , correlated expression traits and their associated set of markers are treated as a module so as to account for epistatic interactions and pleiotropic effects . Parameters of interest are the partitions of genes and markers into modules , and the partition of individuals into different types that correspond to the relationships between expression levels and marker genotypes in a given module . A Markov chain Monte Carlo ( MCMC ) algorithm is designed to traverse the space of all possible partitions . Simulation studies show that the proposed method achieves significantly improved power in detecting eQTLs compared to traditional regression-based methods . A particular strength of our method is its ability to detect epistasis with high power when the marginal effects are weak , addressing a key weakness of all other eQTL mapping methods . We applied our method to a previously described data set consisting of gene expression and genotypes data for 112 segregants from a cross between laboratory ( BY ) and wild ( RM ) strains of S . cerevisiae [15] . In addition to identifying several modules linked to single eQTLs that are consistent with previous reports [1] , [11] , [16] , our method dissected large eQTL hot spots into different modules that correspond to different causal regulators or to primary and secondary responses to causal regulators . In addition , we detected nine modules under the control of two genetic loci . One of these modules corresponds to a previously verified result regarding the interaction between GPA1 and MAT [5] , [16] . another is regulated by both ZAP1 expression and genotype , consistent with previously described results [17] . The other seven modules represent novel findings . Three of these appear to be artifacts of cross-hybridization in microarray experiments; while another exhibits strong epistatic interactions between two loci consisting of many daughter-cell expressed genes that we predict are under the regulation of AMN1 and BPH1 .
We define a module as a set of gene expression traits ( referred to as “genes” henceforth ) and a set of genetic markers ( e . g . , SNPs ) such that the variation of the gene expression traits is associated with the variation of the markers , as shown in Figure 1 . This association between multiple genes and markers is characterized by a latent indicator variable , individual type , conditional on which the trait and marker variables are independent of each other . The individual type latent variable can be viewed as representing a certain combination of markers that induces changes in expressions of a certain set of genes across different individual types . In the simplest case with a single marker , the individual type could correspond to a dominant genetic model , as illustrated in Figure 2A . In this instance , our model is mathematically equivalent to the regression model ( Figure 2B ) . In the case of two markers associated with gene expression traits , there could be two to nine individual types ( various genotype combinations ) . Figure 2C illustrates a case with three individual types: 1 ) high expression values associated with red-colored genotype combinations , 2 ) medium expression values with blue-colored combinations , and 3 ) low expression values with green-colored combinations . The goal of the Bayesian partition method is to simultaneously partition genes and SNPs into modules . The details of the Bayesian partition model are described in the Methods section . To test the effectiveness of our method , we simulated 120 individuals with 500 binary markers and 1000 expression traits in the context of inbred cross of haploid strains . There are eight modules ( summarized in Table 1 ) , each consisting of 40 genes , simulated from different epistasis models based on the linear regression framework , which is different from the posited Bayesian model in our analysis . The genotypic means and frequencies for the two loci used in the simulation are listed in Table 2 . We repeated the simulation 100 times and analyzed the simulated data using two methods: ( 1 ) our Bayesian partition method using parallel tempering [18] with 15 temperature ladders , referred to as BP; ( 2 ) the two-stage regression method of Storey et al [5] , referred to as SR . Details of the simulation and implementation of these two methods are described in the Supplemental Material . As shown from the receiver operating characteristic ( ROC ) curves in Figure 3 , BP achieved a significantly higher power to detect eQTLs compared to SR . For example , allowing for 50 false positives , BP detected more than 500 ( out of 640 ) true gene-marker pairs , whereas SR only detected ∼100 true pairs and became plateaued even with many more false positives allowed . There are likely two reasons for this . First , we modeled the co-regulated genes as a module so that information from all genes in a given module could be aggregated to improve the signal . Multiple trait mapping has proven to be more powerful than single trait mapping [12] in the regression framework . Second , we modeled epistatic interactions explicitly so that markers with weak marginal but strong interactive effects could be detected . The contrast of the performances of these two methods is most prominent when the marginal effect is weak . For example , in modules B , D and H , the rate of true positive detections of SR never exceeded 5% even at the generous FDR threshold of 90% . In modules E , F , and G where the major marker explains more than 70% of the genetic variation , SR detected the major marker in nearly 50% of the simulations at the 50% FDR threshold , but not the minor marker . In contrast , BP performed superiorly and robustly in all eight modules . The module by module comparisons are detailed in the Supplemental Material Text S1 and shown in Supplementary Figure S1 . Figure 4 provides a graphical view of the BP result for another simulated dataset with 120 individuals , 1000 genes , and 500 markers . Four distinct modules , with 60 , 60 , 40 , and 40 genes , and controlled by 3 , 2 , 1 , and 2 markers , respectively ( shown in Supplementary Table S1 ) , are simulated similarly as in the previous example ( more details in the Supplemental Material Text S1 ) . The shape and height of a point represent the most probable module classification and the corresponding maximum posterior probability of a gene . We see that all of the “background” genes were correctly classified according to their highest posterior probabilities . Most genes in the four non-null modules were also correctly classified , other than a very few ones that were classified into the null module , most likely due to their weak signals . BP also correctly identified the truly associated markers of the four modules with high posterior probabilities ( shown in Supplementary Table S2 ) . We applied our Bayesian method to a data set consisting of gene expression and genotypes for 112 segregants from a cross between laboratory ( BY ) and wild ( RM ) strains of S . cerevisiae [15] and detected 29 modules of genes and their associated markers ( Methods ) . Among these 29 modules , 20 are linked to a single eQTL while the remaining nine are linked to two eQTLs . Three of the nine linking to two eQTLs give rise to significant epistatic interactions between the two loci . Twenty-six of the 29 modules significantly overlap ( corrected p-value<0 . 05 ) with at least one of the 13 gene groups previously reported as mapping to eQTL hot spots [11] . We also tested each of these modules for enrichment using GO terms , a yeast knockout compendium [19] , and transcription factor binding sites [20] . At p-value<0 . 05 after multiple testing correction , 21 modules have at least one GO term enrichment; 22 modules overlap with at least one knockout signature , and 13 modules are enriched for at least one transcription factor binding site . The result is summarized in Table 3 and a breakdown result is in Supplementary Table S3 . In contrast , the LOD score distributions of transcripts at the associated markers under the “single-transcript-single-marker” model are shown in Supplementary Figure S2 . Our Bayesian method identifies significantly more weak gene-marker associations than the simple model . These GO enrichments support the biological relevance of different modules detected by our method . Each module is described in detail in the Supplemental Material Text S1 .
We have developed a Bayesian partition model for simultaneously mapping multiple eQTLs for multiple sets of co-regulated genes . Whereas conventional linkage analysis has been widely and successfully applied to the study of one or a small number of traits at a time , our module-based method is suitable for analyzing thousands of phenotypes simultaneously . Both simulation studies and empirical data examples demonstrated that our method is effective for detecting marker interactions , even when no marginal effects could be detected . These improvements in power are a direct result of accounting for the correlation among gene expression traits and assessing the joint effect of multiple eQTLs , including interactions , on these correlated gene sets . One of the main advances in our approach is the introduction of the “individual type” as a latent variable to describe associations between gene expression traits and markers . The individual type latent variable can be interpreted as a classification of individuals according to a combination of phenotypes and genotypes . The underlying mathematical model for this dependence structure is represented as a chain in which the joint distribution for some set of markers influences a set of expression traits via a latent “individual type” variable . After integrating out this latent variable , we observe a direct relationship between the marker and gene expression sets , similar to what would have been obtained from a the traditional regression model in the single-marker , single-gene case ( Figures 2A and 2B ) . However , the advantage over the standard regression in introducing the latent individual type variable is its enabling us to model epistatic interactions and pleiotropy simultaneously . Linkage disequilibrium ( LD ) among adjacent markers is an important feature of the genetic marker data . For individuals produced by the laboratory crosses ( e . g . , F1 and F2 designs ) , the marker dependency can be modeled satisfactorily by a Markov chain . The BP model can easily entertain this modification of the background marker distribution , but the computation time required to run this modified model dramatically increases since we need a forward-summation-backward-sampling algorithm to update the marker indicators ( see Supplemental Material Text S1 for details ) . Another ad hoc strategy to account for the marker correlations without directly modeling them was to first scan all markers and to enumerate those marker pairs with correlations exceeding a given threshold . Then , in the MCMC algorithm , we imposed a mutually exclusive condition for such pairs so that highly correlated marker pairs would not appear simultaneously in any module . We compared the Markov model approach with the ad hoc strategy on a small simulated data sets and a subset of the real data ( data not shown ) . The ad hoc strategy always provided nearly identical results to that of the Markov model with only a fraction of the computation cost . Note that there are also markers that are highly correlated but are not physically linked [26] . In such cases the Markov model actually worked less satisfactorily than the ad hoc approach . Our method shares some similarities to other methods in the literature , but also shows clear distinctions . For example , Lee et al . [17] proposed to simultaneously partition the gene expression and genotype markers . However , their method requires strong priors on the potential regulators , while our method does not . Kendzioski et al . [14] proposed a mixture of markers model to find the eQTLs for multiple gene expression . However , their method separates the gene clustering and eQTL mapping steps , where they first use k-means clustering to identify subsets of genes , and then apply eQTL mapping to the clusters of genes . In addition , their method does not address the epistatic effects . In contrast , gene expression partition and eQTL mapping are modeled jointly in our Bayesian method , and we are able to effectively detect epistasis by using a comprehensive statistical model on both the gene expression and the markers . Our analysis of the yeast data identified 20 modules linked to one eQTL and 9 modules linked to two eQTLs , among which three giving rise to strong epistatic interactions between markers . Some of the modules linked to two eQTLs are consistent with previously reported results [5] , [17] , and we were able to identify more true positive hits along with fewer false positives than previously reported . It is of note that our approach can also be applied to mammalian data and to other quantitative traits data with discrete genetic and environmental covariates . In typical mouse studies , about 2000 SNPs are genotyped and 25 , 000 transcripts are measured , among which about 8000 are significantly differentially expressed [2] . The computation time will be at a similar order of the yeast data analysis . In typical human studies , 650 , 000 SNPs are genotyped and 40 , 000 transcripts are measured . The computation time will dramatically increase . We may , however , restrict our attention to hundreds of SNPs identified as possibly associated with gene expression traits in a human cohort , or/and to fewer expression traits identified as being relevant to diseases of interest [27]–[28] . In this type of scenarios , the input datasets would be roughly equivalent to the yeast data set described herein . Many other such applications can be imagined , We are also improving parallelization implementation . Hopefully , we will be able to appropriately generalize and improve the Bayesian model as well as the MCMC algorithm so that our method can be applied to complete mammalian and other large data sets .
A module is defined in the Results section as a set of gene expression traits ( referred to as “genes” henceforth ) and a set of genetic markers ( e . g . , SNPs ) such that the mRNA expression variation of the genes is associated with the allelic variation of the markers . This association between multiple genes and markers is characterized by a latent indicator variable , individual type , conditional on which the trait and marker variables are independent of each other . The individual type latent variable can be viewed as representing a certain combination of markers that induces changes in expressions of a certain set of genes across different individual types . To formally describe our model , consider a sample with N individuals . Each individual i is measured with G gene expression values denoted as and M marker genotypes denoted as . We assume that the observed data can be partitioned into D nontrivial modules plus a null component . The number of non-null modules , D , is pre-specified by the user and should reflect the user's prior belief in the higher level structure of the data . Every gene g or marker m belongs to one of the D nontrivial modules or the null module , determined by the gene indicator and the marker indicator . For each module , we further partition the N individuals into types denoted by the individual indicators for . Each module may have a different number of individual types as well as different ways of partitioning the N individuals . For example , with a single biallelic marker ( alleles ‘A’ and ‘a’ ) in the module , the module may have two individual types corresponding to genotypes aa vs . Aa or AA ( dominant model ) , or 3 individual types corresponding to genotypes aa , Aa and AA ( additive model ) . We seek module partitions in which expression patterns are similar for all genes , and gene expression variations across different individuals can be explained by the individual types . A cartoon illustration of the partition model is shown in Figure 1 . We model the gene expression traits in module d by an ANOVA model so that each trait value is the sum of the gene effect ( ) , the eQTL effect for individual type k ( ) , the individual effect ( ) , and an error term:where gene g is in module d , k is the individual type of i , and ri and αg are random effects , following independent Gaussian distributions with mean zero . To account for epistasis , we model the joint distribution of all the associated markers of module d , , by a multinomial distribution , whose frequency vector is determined by the individual type k , i . e . , For example , if there are two markers in the module and each has three genotypes , then there are nine combinations of the marker patterns . Thus follows a 9-dimensional multinomial distribution . For the null component , we assume that there is no association between the genes and the markers . The gene expression traits follow a normal distribution and the marker genotypes follow an independent multinomial distribution . To avoid overfitting , we put an exponential prior on the indicator variables to penalize partitions with high complexity:where are the number of genes , markers and individual types in module d , and is the number of genotypes at each marker . We use conjugate priors on the continuous parameters , such as means and variances of the Gaussian distributions and frequency vectors of the multinomials , so that most of these parameters can be integrated out analytically to reduce the complexity of the posterior distribution . The joint posterior distribution of all unknown variables is of the form:where β represents the set of left-over continuous parameters unable to be integrated out analytically . In order to make inference on the eQTL modules from this posterior distribution , we construct a Markov chain Monte Carlo method to traverse the joint space of all unknown parameters . Each Markov chain is randomly initialized , and uses the Gibbs sampler and the Metropolis-Hasting algorithm [18] to update the variables . We implement a split-merge algorithm , which is a special case of the reversible jump MCMC [29] , to update the individual partitions globally . Parallel tempering [30] is used to help mixing the Markov chain . Further details of the modeling and sampling strategies can be found in the Supplemental Material Text S1 . Posterior probabilities are evaluated for each gene and candidate marker set to belong to each module based on the Monte Carlo samples . A threshold is then applied to the posterior probabilities to determine whether a particular gene and marker set should be included in a module . We assembled genotypic and expression data from 112 segregants obtained from a previously described yeast cross between the BY and RM strains of S . cerevisiae [15] . Of the 5 , 740 genes represented on the microarrays in this study , we selected 3 , 662 informative genes as input into the partition algorithm following the same criteria as previously described [10] . We then transformed the gene expression values by first performing quantile normalization [31] to make the distribution of the log-expression ratios for each individual to be the same , and then normalizing each gene so that the mean expression level for each gene was 0 and the standard deviation was 1 . Given that genes in the data set have been previously mapped to 13 distinct eQTL hot spots [11] and that there can be multiple causal factors for a single eQTL hot spot , we set the number of starting modules for our MCMC algorithm to 35∼45 ( 3×13 plus a null model ) to account for these previously identified groups , and to also allow for the detection of new groups as well . For the parallel tempering implementation , we used 30 temperature ladders with almost equal spacing so that the average acceptance probability for exchanges between adjacent chains was roughly 0 . 15–0 . 3 . We ran MCMC sampling for 1 , 000 , 000 iterations in each chain , which took one week of 30 CPUs ( accounting for 30 parallel temperature ladders of the MCMC algorithm ) on a Linux cluster with 2GHz CPUs . The log posterior probability and its auto-correlation curve depicted in Figures S5C and S5D highlight that the Markov chain became stationary after a burn-in period . See Supplemental Material Text S1 for more details . Because markers in the yeast data set are very densely distributed , adjacent markers are almost always highly correlated . After MCMC sampling , markers adjacent to the “truly” linked marker often diluted the posterior probability for the true marker-module linkage . Since a proper Markov chain model for unlinked markers is computationally too expensive to implement ( see Supplemental Material Text S1 ) , we employed a heuristic method to counter this problem . We first specified a window centered at each marker so that markers inside the window are in high LD with the marker at the center . The posterior probabilities of all markers in the window were summed up and regarded as the modified posterior probability of the central marker . The markers with peak probabilities exceeding the given threshold were selected and all other markers in the corresponding windows were masked out . Although we did not explicitly model pleiotropic effects for markers ( i . e . , single markers were not allowed to be associated with expression traits in multiple modules ) , we reported several modules mapped to the same markers in the yeast data set ( See Table 3 and discussions in the Supplemental Material Text S1 ) . The reason for this apparent contradiction is due to the aforementioned moving window approach and the dense distribution of the markers . In other words , if marker m is truly linked to two modules , in computation its adjacent markers can serve as its surrogates so that a subset of these markers are mapped to module 1 , and the remainders mapped to module 2 . Then the use of the moving window method can restore the total probability back to marker m . To test the robustness of our result with respect to the initial parameters , we ran our program using three different numbers of modules , , and , each having three independent runs . Samples from the run with the highest average posterior probability for each value of were used in the subsequent analyses . We chose 0 . 8 as the threshold for the posterior probabilities to determine the module membership for each gene and marker . We observed that more than 70% of the genes were consistently grouped together and mapped to the same markers ( or null module ) in all the runs with different D values . These genes and their associated markers formed the list of 29 modules . | Genome-wide association studies ( GWAS ) have yielded several causal genes for many human diseases . However , the mechanisms underlying how DNA variations affect disease phenotypes have not been well understood in many cases . Gene expression is intermediate between DNA and clinical endpoints . Linking DNA variation and gene expression variation , often referred to as “expression quantitative trait loci ( eQTL ) mapping” , has yielded clues of mechanisms and pathways by which DNA variations impact phenotypes . Because of the large number of genes and genetic markers in such analyses , it is extremely challenging to discover how a small number of eQTLs interact with each other to affect mRNA expression levels for a set of co-regulated genes . We present a Bayesian method to identify genetic interactions and more eQTLs by treating co-expressed genes as a module . Our method provides a tool to study genetic interactions in human disease models . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"computational",
"biology/population",
"genetics",
"genetics",
"and",
"genomics/gene",
"expression",
"genetics",
"and",
"genomics/complex",
"traits",
"computational",
"biology/genomics",
"genetics",
"and",
"genomics/epigenetics",
"computational",
"biology/systems",
"biology",
"... | 2010 | A Bayesian Partition Method for Detecting Pleiotropic and Epistatic eQTL Modules |
Although significant variations in the metabolic profiles exist among different cells , little is understood in terms of genetic regulations of such cell type–specific metabolic phenotypes and nutrient requirements . While many cancer cells depend on exogenous glutamine for survival to justify the therapeutic targeting of glutamine metabolism , the mechanisms of glutamine dependence and likely response and resistance of such glutamine-targeting strategies among cancers are largely unknown . In this study , we have found a systematic variation in the glutamine dependence among breast tumor subtypes associated with mammary differentiation: basal- but not luminal-type breast cells are more glutamine-dependent and may be susceptible to glutamine-targeting therapeutics . Glutamine independence of luminal-type cells is associated mechanistically with lineage-specific expression of glutamine synthetase ( GS ) . Luminal cells can also rescue basal cells in co-culture without glutamine , indicating a potential for glutamine symbiosis within breast ducts . The luminal-specific expression of GS is directly induced by GATA3 and represses glutaminase expression . Such distinct glutamine dependency and metabolic symbiosis is coupled with the acquisition of the GS and glutamine independence during the mammary differentiation program . Understanding the genetic circuitry governing distinct metabolic patterns is relevant to many symbiotic relationships among different cells and organisms . In addition , the ability of GS to predict patterns of glutamine metabolism and dependency among tumors is also crucial in the rational design and application of glutamine and other metabolic pathway targeted therapies .
There are a large number of differentiated cell types in the human body . Even among the cells collectively known as fibroblasts [1] , endothelial [2] and smooth muscle cells [3] , gene expression analysis has identified an unexpected level of positional memory and topographic differentiation . Such functional specialization contributes to the phenotypic variations of many human diseases , including cancer . For example , gene expression analysis of breast cancers has identified five intrinsic subtypes ( luminal A , luminal B , basal , HER2+ , and normal-like ) with unique clinical and histological properties [4] , [5] . The classification nomenclature is based on the putative progenitor cell ( s ) for breast carcinogenesis with properties consistent with derivation from the basal and luminal epithelia arrested at specific differentiation stages or from different mature epithelial cells [4]–[7] . Importantly , these subtype-specific gene expression and phenotypic variations are also observed in many breast cancer cell lines with similar molecular phenotypes [8]–[11] . A number of studies have isolated the different populations of primary epithelial cells to investigate their relevant cellular origins and metabolic features for different breast cancer types [7] , [12] , [13] . Although the cellular origin of luminal and basal-like breast tumor has not been resolved [14] , [15] , cell lineage still appears to confer an important source of patterned heterogeneity to the disease . Although gene expression analysis has yielded important insights into the cellular differentiation and various properties associated with tumors from different cell types , very little is known about the corresponding metabolic phenotypes and nutrient requirements . The processes of oncogenic transformation place energy demands on cancer cells to support proliferation , expansion , and invasion . Dysregulated tumor metabolism is a critical part of oncogenesis and may be targeted for therapeutic benefits [16] , [17] . One prominent example of dysregulated tumor metabolism is “aerobic glycolysis” as recognized by Otto Warburg [18] . Most normal mammalian cells shift to glycolysis for energy generation when oxygen is inadequate for effective oxidative phosphorylation under hypoxia . But tumor cells tend to favor glycolysis even with the availability of oxygen , hence termed “aerobic glycolysis” [19] . Such preferential use of glycolysis leads to vigorous glucose uptake and explains the ability of the tracer glucose analog Fluorine-18 ( F-18 ) FDG to image human cancers in FDG-PET . Such understanding of altered metabolism and nutrient requirement in cancer cells may allow us to exploit these differences for diagnostic and therapeutic benefits . Another aspect of dysregulated tumor metabolism is manifested as altered requirements for amino acids . For example , patients with acute lymphocytic leukemia ( ALL ) benefit from asparaginase treatment as the leukemic cells require large amounts of exogenous asparagine due to a deficiency in this metabolic pathway [20] . Recently , evidence is also accumulating for the essential role of glutamine for cancer cells as a building block for protein synthesis , to supply cellular ATP , as a metabolic intermediate for nucleotide synthesis , and for its anti-oxidative capacity [21] , [22] . Such glutamine dependence or addiction is reflected in the growth restriction and cell death in glutamine limiting conditions . The glutamine addiction is also critical for c-myc-mediated oncogenesis [23]–[25] , linked with glucose requirement [26] , and proposed as an attractive target for therapeutic intervention [22] , [27] . The catabolism of glutamine is initiated by glutaminolysis mediated by two different subtypes of mitochondrial glutaminase ( kidney or liver-type encoded by GLS or GLS2 respectively ) to become glutamate [28] . The intracellular pool of glutamate is a versatile metabolic intermediate that connects with a wide variety of distinct biological processes including synthesis of the anti-oxidant glutathione , amino acid catabolism through transamination , and conversion to α-ketoglutarate as a substrate for the TCA cycle . This process of glutaminolysis by glutaminases has been shown to mediate signaling events [29] , to be coupled with c-myc oncogenesis [25] , and proposed as a critical step in targeting glutamine metabolism [24] , [27] . In some cell types , glutamine can be generated from intracellular glutamate through glutamine synthetase ( GS , encoded by GLUL , glutamate-ammonia ligase ) catalyzing the reverse reaction of the glutaminases . This process is important for removal of ammonia or glutamate depending on the cellular context [30] . While glutaminase is known as an important regulator of glutamine requirement , few studies have focused on glutamine synthetase as a potential determinant of glutamine requirement . Although normal glutamine metabolism is well understood , the genetic parameters and mechanisms of variation in this key nutrient pathway among tumors are largely unknown . Deprivation of glutamine and other amino acids triggers a canonical amino acid response ( AAR ) in most mammalian cells that is measurable by gene expression changes [31] . The free and uncharged t-RNA associated with glutamine deprivation activates a serine/threonine-protein kinase GCN2 which phosphorylates eIF2α and inhibits cap-dependent translation [32] . While reducing the global translation rate , eIF2α phosphorylation also preferentially increases the translation of ATF4 and other mRNAs [31] . The increased level of ATF4 protein triggers the AAR gene expression program characterized by the induction of XBP1 and DDIT3 as an adaptive response to amino acid deprivation . The importance of the AAR is demonstrated by the fact that deficiency of ATF4 compromises the AAR and renders cells susceptible to amino acid deprivation and oxidative stresses [33] . Through the analysis of how different breast cancer cells respond to glutamine deprivation , we have found a dramatic difference in the glutamine requirement among different breast cancer cells which tracks with the luminal versus basal type . These metabolic differences can be explained by cell-type specific expression of glutamine-metabolizing genes and enzymes likely acting in concert with cell type specific oncogenic programs . Therefore , we have provided a series of fundamental building blocks to understand how differentiation is coupled with distinct glutamine utilization in normal and neoplastic breast epithelia . Such an understanding will be relevant to both the mechanistic understanding of metabolic phenotypes and present insights into how best to select subsets of breast cancer patients most likely to benefit from glutamine-targeting therapies .
Many cancer cells require glutamine for survival and proliferation and thus exhibit a phenotype of “glutamine dependence” or “addiction” [22] . To determine whether such phenotypes could be also found in breast cancer cells , we tested how glutamine deprivation affected seven different breast cancer cell lines . Consistent with the idea of glutamine dependence , three cell lines ( BT20 , MDAMB231 , and MDAMB157 ) had significantly reduced growth ( MTT assay , Figure 1A ) and prominent cell death ( trypan blue exclusion assay , Figure 1B ) upon glutamine deprivation for 48 h . Unexpectedly , glutamine deprivation had only modest effects on the growth and viability ( Figure 1A , 1B ) of the other four cell lines ( T47D , BT474 , MCF7 , and MDAMB361 ) indicating relative glutamine independence . When we examined the properties associated with the distinct need for glutamine , we found the cell lines that exhibit glutamine dependence are all of the basal-type whereas the four lines that are more glutamine independent are luminal-type cells ( Figure 1A , 1B ) [34] . As glucose and glutamine are two important energy sources for cancer cells we compared how deprivation of glutamine and glucose affected the growth of these breast cell lines . In the three basal-type cell lines , glutamine depletion had a stronger effect on cell growth than glucose depletion ( Figure 1C ) . In contrast , glucose depletion had a more dramatic influence on cell growth than glutamine depletion in the four luminal cell lines ( Figure 1C ) . These results suggested that there is a consistent variation in glutamine phenotype associated with cell lineage in breast cancers . One important function of glutamine is to serve as an energy source in generating cellular ATP . To determine the relative importance of glutamine to ATP generation in the breast cancer cell lines , we measured ATP in cells grown in media containing either normal levels of glutamine ( 4mM ) or no glutamine for 12 hours . Glutamine deprivation led to a much more significant reduction in ATP generation in the basal-type cells than the luminal-type breast cancer cell lines ( Figure 1D ) . These results further support the concept that glutamine is a more important energy source in basal than luminal breast cell lines . To further analyze glutamine metabolism among different cell types , we measured the consumption of glutamine in the medium and intracellular glutamine levels . When compared with luminal-type cells , the basal cell lines had significantly higher levels of glutamine consumption ( Figure 1E ) and lower intracellular glutamine concentrations ( Figure 1F ) . Collectively , these data strongly support the concept of distinct glutamine metabolism and varying dependence for external glutamine between basal and luminal type breast cancer cells . We hypothesized that such distinct glutamine dependence among basal and luminal breast cancer cell lines may be caused by variable expression of key enzymes involved in glutamine metabolism . Glutamine synthetase ( GS encoded by GLUL – glutamate-ammonia ligase ) and glutaminase ( GLS – kidney form or GLS2 – liver form ) mediate the opposite reaction in the reversible conversion between glutamate and glutamine . GS mediates the capture of an ammonia group by glutamate to synthesize glutamine , while glutaminase catalyzes the breakdown of glutamine to glutamate . We first examined the RNA expression of these genes in a microarray expression set [34] and found that the expression of GLUL ( GS ) was significantly higher in the luminal cell lines . In contrast , the expression of GLS ( glutaminase , kidney ) was higher in the basal lines . While lacking GLS expression , the luminal breast cell lines have a higher level of GLS2 ( Figure 2A ) . We confirmed this cell-type specific differential mRNA expression of GLUL , GLS and GLS2 with real-time PCR ( Figure 2B , 2C , and 2D ) . Differential expression was also found at the protein level as shown by the western blots for GLUL ( GS ) and GLS2 ( in luminal cells ) and GLS ( in basal cells ) ( Figure 2E ) . We next examined whether the expression patterns of GLUL , GLS and GLS2 found in luminal and basal cell lines were also reflected in the respective subtypes of primary human breast cancers . In a breast tumor expression dataset [35] , we found significantly different expression levels of GLUL , GLS and GLS2 in the corresponding luminal ( luminal A and B ) and basal-types of breast tumors ( Figure 3A ) . We also examined the expression levels of these three genes in the same dataset within the 5 intrinsic subtypes [35] and found significantly different expression between the luminal A and basal tumors ( Figure S1 ) . This concordance indicates the differential expression of GLUL , GLS and GLS2 in the luminal and basal-type cancer cell lines reflects similar systematic differences in primary breast tumors . To determine whether differential expression of genes driving glutamine metabolism is an intrinsic cell-lineage phenomenon in the breast , we examined their expression levels in normal non-transformed basal and luminal epithelial cells . Primary luminal and basal breast epithelial cells were separated based on surface expression of EPCAM ( TACSTD1 ) from reduction mammoplasty specimens and gene expression levels were determined by microarray analysis [12] . Analysis of isogenic basal and luminal epithelial cells showed that the mRNA levels of GLUL , GLS and GLS2 exhibited similar cell-type specific expression in normal breast cells ( Figure 3B ) . These findings were also confirmed by real-time PCR ( Figure 3C–3E ) . In addition , expression of the GLUL ( GS ) and GLS proteins showed corresponding luminal and basal-specific expression patterns ( Figure 3F ) . The level of GLS2 protein was below detection levels in both primary epithelial cells ( Figure 3F ) . These results suggest that differential expression of glutamine metabolizing enzymes in cancers may be ascribed to systematic differences in cell lineage observed in normal basal and luminal epithelial cells . Given the well-recognized glutamine dependency of many cancer cells , we investigated the roles of GLUL and GLS2 in the relative glutamine independence of the luminal-type cells . We first treated cells with a GS inhibitor ( L-MS [36] ) for 48 h and measured cell viability under glutamine deprivation . We found that L-MS reduced the survival of the luminal cell lines but had no statistically significant effect over glutamine starvation on all three tested basal cell lines ( Figure 4A ) indicating that GS is involved in the glutamine independence of the luminal cells . Next , we performed genetic experiments to examine the role of specific genes in the glutamine phenotype . Silencing of GLUL ( encoding GS ) in the luminal MCF7 line significantly reduced the RNA and protein expression of GS ( Figure S2A and S2B ) and led to a significant reduction in glutamine independence ( Figure 4B ) . In contrast , similar silencing of GLS2 did not affect survival under glutamine deprivation ( Figure S3 ) . In addition , the ectopic overexpression of GLUL ( verified in Figure S4A , S4B ) in the basal MDAMB231 cells conferred partial glutamine independence by significantly increasing the cell survival under glutamine deprivation ( Figure 4C ) . Taken together , these data suggest that GS expression significantly contributes to the differential glutamine phenotypes observed in breast cancer cell lines . We next investigated potential regulatory mechanisms for the subtype-specific expression of glutamine metabolizing enzymes . During the differentiation of luminal epithelial cells , GATA3 is an important master regulatory transcription factor [37] , [38] . The expression of GATA3 in luminal and basal cells is systematically different as previously noted [11] , [39] . Using real-time PCR , we also demonstrated the cell-type specific expression of GATA3 mRNA in MCF7 ( luminal ) and MDAMB231 ( basal ) cells ( Figure S5A ) . Re-analysis of microarray data of the overexpression of GATA3 in mouse breast epithelial cells [38] shows induction of GLUL and repression of GLS and GLS2 ( Figure 4D ) . These data suggested a role for the lineage factor GATA3 in regulating the luminal and basal-specific expression of GLUL and GLS . We directly tested the role of GATA3 in regulating the glutamine phenotype in breast cancer cell lines . The mRNA and protein levels of GATA3 could be effectively reduced by gene silencing through siRNAs ( Figure S5B , S5C ) . Silencing of GATA3 in MCF7 cells led to significant reduction in GLUL at both the RNA and protein levels ( Figure 4E , Figure S5C ) . Conversely , overexpression of GATA3 in the basal MDAMB231 line ( Figure S5D ) led to a significant upregulation of GLUL ( Figure 4F , Figure S5E ) . Furthermore , the silencing of GATA3 in MCF7 cells reduced the survival under glutamine deprivation ( Figure 4G , Figure S3 ) , and overexpression of GATA3 in MDAMB231 cells increased the resistance to glutamine deprivation ( Figure 4H ) , consistent with a direct role for GATA3 mediated GLUL expression in the glutamine independence of luminal breast cells . In addition , the glutamine independence caused by GLUL ( Figure S6A ) or GATA3 ( Figure S6B ) overexpression in MDAMB231 cells was also abolished with treatment of L-MS ( GS inhibitor ) , indicating of the importance of the catalytic activities of GS . Given the ability of GATA3 to increase the expression of GLUL , we examined the promoter region of GLUL and found two potential GATA3 binding sites at −524 to −518 bp ( region A ) and −200 to −194 bp ( region B ) upstream of the transcriptional start site ( Figure 4I ) . We used chromatin immunoprecipitation ( ChIP ) to test whether GLUL may be a direct downstream target of GATA3 transactivation . Consistent with previous data [40] , the promoters of ESR1 ( estrogen receptor alpha ) , but not albumin , were enriched in the GATA3 ChIP samples . Of the two putative GATA3 binding sites in the GLUL promoter , the distal region A but not the more promoter proximal region B , was significantly enriched in the GATA3 ChIP samples ( Figure 4J ) indicating that GATA3 protein can directly bind to a regulatory region of GLUL suggesting that this gene is a target of the luminal transcription factor and further serving to explain the lineage specific requirement for glutamine . The deprivation of amino acids in mammalian cells leads to the stabilization of the ATF4 ( activating transcription factor 4 ) protein and resulting induction of a canonical gene expression program known as the amino acid response ( AAR ) [41] . The response includes the induction of XBP1 ( X-box binding protein 1 ) and DDIT3 ( DNA-damage-inducible transcript 3 ) which are essential for survival under amino acid deprivation [41] . Given the distinct growth and survival response of luminal and basal breast cells to glutamine deprivation , we used microarrays to compare their transcriptional responses on a global scale . Triplicate plates of MCF7 and MDAMB231 cells were cultured under both control ( 4 mM glutamine/Q4 ) and glutamine-depleted ( no glutamine/Q0 ) conditions for 24 hours . RNA from each plate was interrogated with Affymetrix GeneChip U133-A2 arrays ( results deposited in Gene Expression Omnibus ( GSE26370 ) ) . Gene expression profiles of the 12 arrays were normalized by RMA and the transcriptional changes of glutamine deprivation in both cell types were derived by zero-transformation against the average expression levels of the control samples as performed previously [42]–[44] . Probes sets showing at least two fold changes in at least two samples ( n = 405 ) were selected and arranged by hierarchical clustering according to similarities in expression patterns ( Figure 5A ) . This analysis showed that glutamine deprivation induced a strong gene expression response in MDAMB231 ( MB231 ) but less so in MCF7 cells ( Figure 5A ) . We found that the canonical AAR genes were induced by glutamine deprivation only in MDAMB231 cells ( Figure 5A ) . A previous study showed that glutamine deprivation inhibits cell growth by inducing the tumor suppressor gene TXNIP [29] . This gene was also induced only in the MDAMB231 line . We applied a published gene expression study of histidine deprivation [45] as training data and estimated the degree of AAR using a binary regression model . MDAMB231 but not the MCF7 line exhibited a significantly higher probability of AAR after glutamine deprivation using this approach ( Figure 5B and 5C ) . The stronger amino acid response in the MDAMB231 cells was also confirmed by real-time PCR for XBP1 ( Figure 5D ) and DDIT3 ( Figure 5E ) . These data provide further evidence that glutamine deprivation induces a much dramatic response in the basal cells and a weak response correlating with glutamine independence of the luminal cells . We examined how glutamine deprivation affected different glutamine-metabolizing enzymes and found that GS protein ( Figure 6A ) , but not mRNA ( Figure S7A ) , were significantly induced in MCF7 cells in a dosage-dependent manner . This translational regulation may be an adaptive response to compensate for reduced environmental levels of glutamine . To examine the role of GATA3 in the induction of GS during glutamine deprivation , we compared the GS protein levels under different glutamine levels in MCF-7 transfected with control or GATA3-targeting siRNA . We found that while the silencing of GATA3 reduced the GS levels , there was still significant protein induction during glutamine deprivation ( Figure S7B ) . We also measured glutamine concentrations in glutamine deficient media used to culture MCF7 and MDAMB231 cells and found a significant increase in glutamine levels in medium cultured with MCF7 but not MDAMB231 cells ( Figure 6B ) . Similarly , intracellular glutamine concentrations were increased only in MCF7 but not MDAMB231 cells under glutamine deprivation ( Figure 6C ) . Therefore , the glutamine independence phenotype of luminal cells may be due to the capacity of these cells to synthesize glutamine from intracellular glutamate and other sources in the absence of external glutamine . In normal breast ducts , luminal and basal cells are in close physical proximity . Because of the ability of luminal cells to synthesize glutamine and the requirement of basal cells for glutamine , we next tested the potential for glutamine symbiosis between these two cell types with transwell co-culturing experiments ( Figure 6D–6F ) . We found that the viability of MDAMB231 cells under glutamine deficient media was significantly increased when MCF7 cells were used as a feeder layer ( Figure 6E ) , consistent with observed higher extracellular glutamine levels ( Figure 6F ) . Furthermore , conditioned medium from MCF7 cells was also able to support significantly the growth and viability of the MDAMB231 cells ( Figure 6G–6I ) . We showed above that increased levels of GLUL either by transfection with GLUL or GATA3 makes the MDAMB231 line more resistant to glutamine deprivation ( Figure 4C and 4H ) . We next asked whether this was due to increased synthesis of the nutrient . Intracellular glutamine levels increased dramatically in MDAMB231 cells expressing either GLUL or GATA3 ( 5×104 cells in the upper well ) ( Figure 6J , 6K ) . The effects of GLUL and GATA3 overexpression in MDAMB231 cells on intracellular glutamine levels were blocked with L-MS treatment ( Figure S8A ) . We also showed that the intracellular glutamine levels were reduced in MCF7 with siRNAs targeted to GLUL or GATA3 in medium with normal glutamine level ( Q4 ) or no glutamine ( Q0 ) ( Figure S8B ) . Further , in the co-culture system ( Figure 6L ) , MDAMB231 cells demonstrated increased viability when co-cultured with either GLUL or GATA3 expressing MDAMB231 cells ( Figure 6M ) and this correlated with both increased glutamine levels in the medium ( Figure 6N ) and intracellularly ( Figure 6O ) . These data provide a consistent mechanistic picture of a gene expression program related to nutrient requirements and potential glutamine symbiosis . The expression of GLUL and GLS are inversely correlated in the luminal and basal types of primary breast cancers , cancer cell lines , and primary epithelial cells . This pattern of expression made us investigate whether cross-regulation exists between these two genes that encode enzymes mediating directly opposite chemical reactions . We first used siRNA to silence GLUL in MCF7 cells and observed an increase in GLS mRNA expression ( Figure 7A ) . Further , ectopic over-expression of GLUL in MDAMB231 cells reduced GLS mRNA ( Figure 7B ) . In contrast , similar silencing of GLS did not show any effect on GLUL levels ( Figure 7C , 7D ) . The ability of GLUL overexpression in MDAMB231 to repress GLS was also seen at the protein level with a dose dependent decrease in GLS protein observed with increasing amounts of GS protein from varying levels of transfected GLUL ( Figure 7E ) . These results indicated the ability of GLUL to repress the expression of GLS while GLS had no detectable effect on the level of GLUL . Since GATA3 regulated the expression of GLUL , we tested the role of GATA3 in regulating GLS by silencing and overexpressing GATA3 in MCF7 and MDAMB231 cells , respectively . Silencing of GATA3 in MCF7 cells increased GLS expression ( Figure 7F ) and GATA3 overexpression in MDAMB231 significantly reduced the level of GLS ( Figure 7G ) . These changes in GLS expression regulated by GATA3 were also detectable at the protein level compared with GLUL ( Figure 7H ) . These results are also consistent with GATA3 overexpression in the mouse epithelial cells ( Figure 4D ) [38] . Based on the data presented , we propose that basal and luminal breast epithelial cells exhibit different patterns of glutamine metabolism ( Figure 8 ) . In the luminal cells , GATA3 triggers expression of GLUL and contributes to glutamine independence . Furthermore , GLUL has the ability to repress GLS which would also help to maintain the cell-type specific expression pattern and phenotype . Basal-specific expression of GLS may be maintained by the absence of GATA3 and higher activities of c-myc in the basal type cells [4] , [46] . These findings suggest that glutamine deprivation may be a viable treatment strategy for basal-type breast cancers . In addition , the expression of GLUL in luminal type cancers correlates with the ability to synthesize glutamine from ammonia and glutamate , and therefore describes at the molecular level a type of cancer that is predicted to be more resistant to glutamine deprivation treatment .
The distinct glutamine requirement and differential expression of glutamine-metabolizing enzymes among luminal and basal breast cancers are consistent with our understanding of the genetic circuitry governing breast cancer subtypes and regulation of these glutamine-metabolizing enzymes . For example , the higher GLS level and sensitivity to glutamine deprivation of basal-type breast cancer cells are consistent with a high level of c-myc activity in basal cells [51] , [52] and the recently described role of c-myc in regulating GLS [24] , [25] . The higher levels of GLS are also consistent with the susceptibility to growth inhibition by targeting this enzyme [27] and indicate the essential nature of this metabolic pathway in the basal cells . These results indicate that triple-negative basal-like breast tumors , with few current therapeutic options , are addicted to glutamine and may benefit from glutamine-targeting therapies [22] , [27] . In contrast , the luminal specific expression of GLS2 may reflect the higher p53 ( wild type ) activity in luminal cells [4] , [52] given the ability of p53 to regulate GLS2 [53] , [54] . Our results suggest that GATA3 is directly involved in the transcriptional regulation of GLUL in luminal cells . The spatial and cell type specific expression of GLUL ( GS ) seen in our studies on breast epithelial cells is also observed in several other cellular contexts . In the brain , GS is expressed mainly in glial cells [55] and is responsible for the synthesis of glutamine from the uptake of glutamate secreted by adjacent neurons . Similar spatial division of glutamine degradation and synthesis also occurs in distinct patterns of GS and GLS expression in the liver [56] . Glutamine degradation by GLS occurs in the periportal cells where there is a high glutamine level from the digested nutrients in the gastrointestinal tract . In contrast , the expression of GLUL ( GS ) is restricted to zones of hepatocytes surrounding the central lobular vein with lower glutamine levels [56] . In the renal nephron , GLUL ( GS ) expression is restricted to the straight portion of the proximal tubules and plays an important role in the removal of ammonia [57] . Such physical separation of glutaminase and glutamine synthetase associated with differentiation and nutrient availability coordinate the glutamine synthesis and effective detoxification of ammonia and glutamate . Similar distinct glutamine metabolism in luminal and basal breast epithelial cells also appears to impact tumors derived from these different lineages and opens an additional window into the metabolic phenotypes of this heterogeneous disease . Therapeutic interventions based upon metabolic targets will need to incorporate these systematic differences between tumor subtypes . Under glutamine deprivation , the high mRNA levels of GLUL ( GS ) in luminal cells undergo further protein upregulation to provide glutamine and may also support the glutamine requirement in basal cells in physical proximity . Similar nutritional and metabolic interaction underlies many symbiotic relationships among different organisms and cell types , including the symbiotic nitrogen-fixing root nodules on legumes [58] , the mutualistic symbiosis between bacteria and insects [59] , and the glutamate-glutamine shuttle between neurons and astroglial cells in the brain [55] , [60] . Interestingly , GS plays a critical role in all these diverse examples of metabolic symbiosis . In addition to the inter-cellular exchange of nutrients , the synthesis of glutamine from glutamate and ammonia by GS can also remove the potential toxicity from the accumulation of glutamate ( neurotransmitter ) and ammonia ( nitrogen waste ) . Ammonia from glutaminolysis has been shown to act as a diffusible autocrine- and paracrine substance inducing autophagy [61] . Given the physical proximity between basal and luminal cells in breast ducts , such a reciprocal metabolic relationship may also be relevant under different environmental or growth conditions . When the tissue organization is disrupted in malignancy , glutamine dependence of the basal-type tumors may be exploited to treat this type of aggressive cancers . Our findings strongly suggest that there will be significant variation in response to glutamine-targeting therapies . Among breast cancers , systematic variation in the glutamine consumptive vs . synthetic behaviors seen in the basal and luminal tumors will directly influence this response . Similar heterogeneity may be important in other tumor types as well . Our data also provide evidence that glutamine-targeting therapeutics may be of special clinical utility for the triple-negative basal-like breast tumors with few therapeutic options . Many current glutamine-targeting therapeutics focus on glutaminase [25] , [27] , but the cell-type specific expression of GLS and GLS2 in different tumors indicates the importance of choosing compounds with intended specificity for particular glutaminase activities in the treated tumors . Since GS is a key genetic determinant of glutamine independence in luminal cells , the evaluation of the GS levels in tumors may be helpful in predicting response . In addition to the cell-autonomous variations in the GS expression and response to glutamine deprivation , the efficacy of glutamine-targeting therapies may also be affected by the ability of adjacent non-transformed cells to provide glutamine . It is important to note that GS activities have been reported in fibroblasts [62] and macrophages [63] . The availability of glutamine from other non-tumor cells or blood may reduce the efficacy of glutamine-targeting therapies . Thus , GS inhibition may be combined with glutamine-targeting therapies to further enhance efficacy and reduce resistance , similar to the use of GS inhibitors to sensitize cancer cells to L-asparaginase [64] . With the explosion of genomic data , we have obtained significant knowledge on how genetic dysregulation contributes to tumor heterogeneity in human cancers . Since dysregulated metabolism is an essential part of oncogenesis , similarly detailed knowledge of metabolic profiles may be of equal or greater importance in understanding and treating the disease [65] , [66] .
All breast cancer cells were cultured in DMEM with 4 . 5 g/L glucose , supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin in 5% CO2 . Primary luminal and basal cells were obtained from women undergoing breast reduction for non-malignant conditions and were separated by cell surface binding to the TACSD1 protein of the Ber-Ep4 antibody as described [12] . For the MTT assay , 2 . 5×103 cells in 100 µl of medium were seeded in a 96-well culture plate . After treatments , cell number was evaluated . In brief , 10 µl of MTT ( Sigma M5655 ) ( 0 . 5 mg/ml ) was added to each well , and then the plates were incubated at 37°C for 3 h . The formazan product was dissolved in DMSO , and the absorbance at 570 nm was measured using a microplate reader . To measure viability by direct counting , 2×104 cells were seeded in 12-well dishes and treated with medium containing different concentrations of glutamine for 48 h; the cells were collected and stained with 0 . 4% Trypan Blue . Cells excluding and taking up dye were counted on a hemocytometer under phase contrast microscopy . For glutamine synthetase inhibition , L-MS ( L-Methionine-Sulfoximine , 5 mM , Sigma-Aldrich ) was administered to cells for 48 h . Cells ( 5×103/well of a 96 well dish ) were treated with or without glutamine for 12 h and ATP content was measured in accordance with the protocol of the ATP-Lite luminescent ATP detection assay kit ( Perkin-Elmer ) . Briefly , 100 µl of assay reagent was added to the wells and mixed for 10 min in the dark; intracellular ATP content was measured using a luminescence multi-label counter . The ATP levels were normalized based on cell counts measured by the MTT assay . Cells ( 1×104/well ) in a 24 well plate were cultured for 24 h in medium without phenol-red , medium was collected , and cells were lysed with RIPA buffer ( Sigma-Aldrich ) . Concentrations of glutamine in the medium and in the cell lysate were determined with the glutamine/glutamate determination kit ( GLN-1; Sigma-Aldrich ) . Each sample was divided into two parts; part 1 was measured with glutaminase for transferring the glutamine into glutamate , part 2 was measured directly . Samples were then dehydrogenized to α-ketoglutarate accompanied by reduction of NAD+ to NADH . The amount of NADH is proportional to the amount of glutamate and was measured using a spectrophotometer at 340 nm . A standard curve was determined for each experiment to calculate the concentration of glutamate in samples . Glutamine levels were calculated ( part 1 minus part 2 ) and normalized to total protein levels . The glutamine level of normal culture medium was also measured , and the glutamine consumption was calculated as ( glutamine in normal medium-glutamine in medium after culturing cells ) and normalized to protein level . Proteins were separated by 10–12% SDS–PAGE and transferred to Immobilon-P membranes ( Millipore ) . Membranes were blocked with 5% skim milk , incubated with primary antibodies ( GLUL , G2781 , Sigma; GLS , ab60709 , Abcam; GATA3 , sc269 , Santa Cruz; GLS2 , ab91073 , Abcam; tubulin , 2128 , cell signaling ) , HRP-conjugated secondary antibody ( Perkin-Elmer ) , and detected with the ECL Western blotting reagent ( Amersham ) . MCF7 and MDAMB231 cells were cultured in medium with or without glutamine for 24 h in triplicate . RNAs were collected with MirVana kit ( Ambion ) and hybridized to Affymetrix U133A2 arrays . Probe intensities were normalized by RMA and then the changes of expression by glutamine deprivation ( 0 mM glutamine/Q0 ) were derived by zero-transformation against the corresponding cells grown in glutamine containing medium ( 4 mM glutamine/Q4 ) . Cells were transfected with non-targeting control or synthetic siRNAs targeting GLUL , GLS , GLS2 and GATA3 ( Applied Biosystems ) with lipofectamine 2000 ( Invitrogen ) . For overexpression experiments , empty vector or overexpression constructs for GLUL or GATA3 ( Origene ) were transfected into cells with lipofectamine 2000 for 48 hours before the levels of indicated transcripts and proteins were examined by real-time RT-PCR and western blot . Total RNA was reverse-transcribed to cDNA with SuperScript II reverse transcription kit , then used for real-time PCR with Power SYBR Green PCR Mix ( Applied Biosystems ) and indicated primers ( Table S1 ) , and normalized to β-actin mRNA levels measured in parallel . 10% formaldehyde solution was added to cells to crosslink DNA-protein complexes . Isolated nuclear chromatin extracts were sonicated and incubated overnight at 4°C with either anti-GATA3 ( SC269 , Santa Cruz ) or normal mouse IgG ( SC3878 , Santa Cruz ) . This was followed by incubation with 20 ml of Protein G agarose beads ( Roche ) for 4 hours at 4°C . After extensive washing , DNA fragments were harvested by de-crosslinking the immunoprecipitates . Real time-PCR utilizing SYBR Green master mix ( Applied Biosystems ) was performed to check the enrichment of indicated promoter regions in pull-down samples using primers listed in Table S1 and normalized with albumin . For co-culture experiments , MDAMB231 , MCF7 , or transfected MDAMB231 cells were seeded in minicells ( upper well/5×104 cells ) with 0 . 4 µm pores ( Millipore ) and co-cultured with MDAMB231 ( lower well/1×104 cells ) for 12 , 24 or 48 h in 24-well plates . For conditioned medium experiments , MDAMB231 or MCF7 cells ( 5×104 ) were seeded in a 24-well plate and incubated in medium with or without glutamine for 24 hours and then medium was transferred to new wells containing MDAMB231 cells ( 1×104 ) . In 12 and 24 hours experiments , medium was collected; cells were washed by PBS and then lysed with 100 µl RIPA buffer . Glutamine concentration was measured with GLN-1 ( Sigma ) . In 48 h experiments , cell numbers were counted by trypan blue exclusion assay . All experiments were expressed as mean ± standard deviation ( SD ) with t-test . Statistical significance was calculated by t test , considering p<0 . 05 ( * ) and p<0 . 01 ( ** ) as statistically significant . | Different types of cells have distinct ways of utilizing nutrients and generating energy , thus resulting in distinct nutrient needs . Such cell type–specific metabolic differences are associated with many biological processes and force the symbiosis between different cells and organisms . For example , glutamine symbiosis is a well-recognized phenomenon due to different glutamine synthesis ability . In human cancers , glutamine is also recognized as an important and essential nutrient , termed glutamine addiction . But very little is known about how glutamine addiction varies among different tumors of diverse cellular origins , which hinders personalized therapeutic strategies . Here , we found that basal-type breast cancer cells were sensitive to glutamine deprivation while luminal-type breast cancer cells were not . Luminal cell–specific glutamine independence results from expression of glutamine synthetase conferring the ability to synthesize glutamine . Glutamine synthetase also represses glutaminase and contributes to the maintenance of the polarized expression of glutamine synthetase and glutaminase among breast cancer cells . Collectively , these data illustrate cross-talk between mammary differentiation programs and unique nutrient requirements , which may offer novel therapeutics for basal-type breast cancers . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"physiological",
"processes",
"cancer",
"genetics",
"phenotypes",
"heredity",
"physiology",
"genetics",
"genetics",
"and",
"genomics",
"biology",
"anatomy",
"and",
"physiology",
"energy",
"metabolism"
] | 2011 | Glutamine Synthetase Is a Genetic Determinant of Cell Type–Specific Glutamine Independence in Breast Epithelia |
The nematode Caenorhabditis elegans , with information on neural connectivity , three-dimensional position and cell linage , provides a unique system for understanding the development of neural networks . Although C . elegans has been widely studied in the past , we present the first statistical study from a developmental perspective , with findings that raise interesting suggestions on the establishment of long-distance connections and network hubs . Here , we analyze the neuro-development for temporal and spatial features , using birth times of neurons and their three-dimensional positions . Comparisons of growth in C . elegans with random spatial network growth highlight two findings relevant to neural network development . First , most neurons which are linked by long-distance connections are born around the same time and early on , suggesting the possibility of early contact or interaction between connected neurons during development . Second , early-born neurons are more highly connected ( tendency to form hubs ) than later-born neurons . This indicates that the longer time frame available to them might underlie high connectivity . Both outcomes are not observed for random connection formation . The study finds that around one-third of electrically coupled long-range connections are late forming , raising the question of what mechanisms are involved in ensuring their accuracy , particularly in light of the extremely invariant connectivity observed in C . elegans . In conclusion , the sequence of neural network development highlights the possibility of early contact or interaction in securing long-distance and high-degree connectivity .
The complexity of the nervous system continues to protract efforts to understand its development . The relatively simple invertebrate neural systems have been the subject of intense study in the last few decades [1] , helping to shed light on mechanisms involved in development like axon guidance and molecular cues . We looked at the development of the neuronal network of C . elegans using information from 279 of the 302 neurons ( see methods and [2] ) . Information on embryonic and post-embryonic lineages of the neurons [3] , [4] , [5] , [6] formed the basis of our developmental data . Our work marks the first attempt to computationally and statistically represent the neural development of C . elegans based on available biological data , enabling a spatio-temporal analysis of the developing neuronal network . Graph theory [7] is increasingly being applied to elucidate the function based on the structures of complex networks like the brain [8] , [9] and here we carry out a structural analysis of neuronal networks during different stages of C . elegans development . We observe neuronal growth ( see Video S1 ) , development times of different classes of neurons , and the time windows for establishing short- and long-distance connectivity . It is now known that C . elegans displays a higher than expected wiring cost with the total wiring length of all connections being twice as high as for an optimized network with spatially rearranged neurons [2] , [10] . Instead , processing speed indicated by the number of intermediate neurons in a pathway and made possible by long-distance connections , seems to be a critical constraint [2] . Establishing accurate connectivity , particularly long-distance , is a critical challenge in development . While the role of guidance molecules is well established in correctly wiring neurons , our results suggest that temporal and possible spatial closeness permitting early neuron-neuron interaction could also have an important role to play in the process . It may be conjectured that this would serve to reduce metabolic costs during development and also increase the probability of accurately establishing long-distance connections . Long-distance connections are vital and known to be affected in several neurological disorders in humans [11] .
The growth in the neuronal network was visualized with respect to the spatial positions of neurons in the nematode body . In the development phase , neurons are born in two bursts – a relatively brief embryonic burst lasting around four hours and a longer post-embryonic phase stretching across seventeen hours . These intervals have been sub-divided into smaller time intervals to enable a more detailed visualization of the sequential appearance of neurons during these bursts . In Figure 1 neurons appearing at a particular stage are indicated in black , while previously existing neurons are coloured in gray . In the first stage , at the end of 350 minutes of embryonic growth , neurons that eventually reside in the head , along the ventral cord ( classes DAn , DBn and DDn ) and tail of the nematode have appeared , with neurons in the head constituting the majority . By the end of the embryonic burst nearly all of the neurons in the head have appeared . Ventral cord neurons belonging to the other five classes namely , ASn , VAn , VBn , VCn and VDn as well as the rest of the tail neurons are born in the latter phase , towards the end of the first larval stage . A relevant question is whether the spatial clustering of neurons in the head , body and tail of the worm also relates to their topology . We find that neurons in the head are mostly connected to neurons in the head ( 74% of all connections ) while 46% of connections of neurons in the body are to other neurons also in the body . Within spatial cluster connectivity is least for neurons in the tail at 29% . Figure S1 shows the successive appearance of neurons with respect to their final 3-dimensional positions in the adult body on a colour scale . We plotted a histogram of the differences in birth-times of connected neurons to assess the time interval available for formation of connections ( Figure 2A ) . For comparison , we looked at the outcomes from twenty random networks ( see Methods for details ) . The values from the trials were separated into ten time bins and the mean and standard deviation values were calculated for each of the bins . It can be seen that approximately two-thirds of connected neurons appear less than 200 minutes apart and this proportion is much higher than that seen in random networks . A zoom-in on time differences of up to 500 minutes shows that most connected neurons within this interval are born less than 50 minutes apart ( Figure 2B ) . Neurons appearing in the embryonic burst go on to make most of their connections , approximately 80% , with other neurons born in the same phase , while this figure is around 46% for neurons born in the post-embryonic burst . These values in randomly shuffled networks are on an average 62 . 1% and 38 . 2% , for the embryonic and post-embryonic phases , respectively . This observation could well be attributed to high degree neurons being born in the embryonic phase ( discussed below ) , however a statistical comparison of the ratio of connections made within and outside the temporal cluster for the embryonic and post-embryonic phases to random development , highlights significance with p value less than 0 . 001 in a one sample t-test . Early stages are also crucial for the formation of highly connected neurons ( hubs ) . Figure 3 shows node degree ( the number of connections of a neuron ) with respect to birth times of neurons and their positions along the body of the worm , namely head , body or tail . Neurons that appear early on in development are more likely to have a higher degree than those that are born later . As would be expected , the highest degree is possessed by neurons in the head . There are 112 neurons with 20 or more connections . This number drops sharply to 46 and then to 9 , for degrees more than or equal to 30 and 60 , respectively . Of all neurons with degree 20 or more , two thirds appear before hatching ( 840 minutes ) , and nearly all neurons ( ∼98% ) with more than 30 connections are born before hatching . It needs to be noted however , that correlation of early birth and higher degree is less obvious for neurons with less than 30 connections . To gauge the significance of these results we repeated the test for twenty random networks ( details in Methods ) . For this random connection formation , approximately 74% ( ±5 . 5 ) of neurons with degrees above 30 could appear before hatching , which using two-sided t-test corresponded to a p value of less than 0 . 001 indicating significant difference between actual and random networks . We also analyzed how the connected neighbors of a neuron appeared over time , particularly to examine if longer time windows served to receive connections from late appearing neurons . This was indeed the case: all neurons with a degree of more than 60 were connected to more than one late forming neighbour . However in random networks , this was on average true for 60% of the cases ( based on 20 random observations ) and had a statistically lower connectivity with late appearing neurons , with p = 0 . 001 . Thus availability of time between neuron birth and nematode maturation appears to be important to hub neurons . The time differences of bilaterally paired neurons were also compared for examining symmetry during development . All bilateral pairs of neurons were born within approximately ten minutes of each other . Figure S2 shows the development of motor , sensory and interneurons . More than 80% of sensory and interneurons appear before hatching whereas this figure was found to be lower at around 50% for motor neurons . However , several of these motor neurons were polymodal , functioning also as inter-neurons . Only around 30% of exclusive motor neurons appeared before hatching . During such early development , there is less need for motor control: for example , earliest movements inside the shell do not involve neural coordination and first signs of neural activity are only observed around 30 minutes before hatching [12] . The distance between any two connected neurons was calculated in three-dimensional space to determine the approximate length of the connection between them . Numerically , this was calculated as the three-dimensional Euclidean distance between two connected neurons and provided a useful measure to assess the length distribution of all edges . Although there is little information on the timing of synaptogenesis , we wanted to visualize how many of the neuron pairs forming short- , medium- , or long-distance connections in the adult were present at each stage . Hence , when we use the term ‘connection pair’ - it does not imply synaptogenesis and is merely indicative of the birth of both neurons that will eventually connect . For the adult nematode , out of the 2 , 990 actual connections pairs 391 are long-distance ( ∼14% ) , 298 are medium-distance ( ∼10% ) and 2 , 301 ( ∼76% ) are short-distance ( see Methods for details ) . Figure 4A shows the percentage of short , medium and long-range connection-pairs appearing at each stage of development . By the time of hatching ( 840 minutes ) , approximately 73% of short , 36% of medium and 68% of long-range connections-pairs had appeared , accounting for 69% of total connections pairs . For twenty random networks ( Figure 4B ) on the other hand , 52% ( ±3 . 5 ) , 53% ( ±6 . 7 ) and 51% ( ±5 . 7 ) of short- , medium- , and long-distance connection-pairs respectively , occurred before hatching . In none of the cases did the actual value fall within the data domain of random tests . Further analysis revealed that the observed difference between the percentage of connection-pairs appearing before hatching in real and random systems was statistically significant , with p<0 . 001 in each case . To determine whether there was any relation between the degree of neuron and the proportion of short or long distance connections that they possessed , we computed the Pearson's coefficient for degree versus short and long connections . No significant correlation was found , with the correlation coefficient of degree with short and long connections being −0 . 03 and 0 . 11 respectively . We then segregated the three types of junctions , namely , gap junction , chemical synapse or a combination of both . ( If both gap junctions and chemical synapses existed between any two neurons , then the connection was termed as a combination . ) . The proportion of short- , medium- and long-length connections in each category is listed in Supplementary Table S1 . Figure 5 compares the time of appearance of connection-pairs linked by gap junctions and chemical synapses in the adult , for short and long-distance connection lengths . Around 35% of short- and 27% of long-distance connections in C . elegans are electrically coupled ( Figure 5A ) , together constituting more than one-third of all connections in the nematode . Interestingly , electrically and chemically connected neuron pairs appear at the same rate with approximately 70% of each category appearing before hatching . A bar chart was also plotted to visualize the distribution of long , medium , and short-distance connection-pairs during each of the developmental stages ( Figure S3 ) . Short distance connection-pairs are most abundant during all stages of development . The frequency of short-distance connections is followed by long-distance and then by medium-distance connection-pairs . The neurons of C . elegans have membership in various functional circuits [6] and we analyzed the appearance of connection-pairs in relation to this functional segregation . The percentage of each circuit that had appeared was measured in relation to the birth of connection-pairs . If a connection-pair involved neurons from two different circuits , then it was considered to belong to both circuits . Figure 6 shows how the various circuits appear over time . In the analysis of temporal features it was shown that approximately 80% of sensory and 50% of motor neurons appeared before hatching . Here , more specifically , analysis showed that the connection-pairs associated with circuitry of amphids , motoneurons in the nerve ring and other sensory receptors in the head , which are understandably more relevant to the early life of the worm are born sooner than other circuits like that of egg-laying , not required in the embryonic stages . Hence although we observe connection pairs belonging to all circuits present to various degrees in the earliest embryonic stages , the order of functional precedence may also influence the likelihood of early contact or interaction . The networks at each stage were predicted based on the earliest possible time of synaptogenesis - when both neurons had appeared . Here , the network represents the neurons present at each stage with adult connectivity ( see Methods ) . Our motivation behind this was to present the potential of network analysis in inferring development features like significant periods , from even discrete data . The global network over time was analyzed for topological changes such as number of nodes and edges ( Figure S4A , B ) , clustering coefficient ( Figure S4C , E ) , and characteristic path length ( Figure S4D , F ) . The ratio of actual clustering coefficient to random of more than 2 and actual average path length to random of less than 1 . 5 are considered to signify a small-world network . The adult C . elegans network has already been shown to have small-world characteristics [13] , here we find that all the networks display small-world characteristics with the ratio of actual clustering coefficient to that of random networks being above 4 at all times ( Figure S4C ) , and the ratio of characteristic path length in the actual to the random network being consistently below 1 . 15 ( Figure S4D ) . Although this analysis does not consider the effect of pruning , as the ratios are well beyond the characteristic ratios , the results are likely to be robust for small changes . The random networks created for these calculations were Erdős-Rényi networks which only maintained the total number of neurons and connections without preserving the degree distribution .
The neuronal network of C . elegans , being the only fully characterized connectome to date , has provided the opportunity to observe changes occurring during the course of its neural development . Based on available data , we have extracted time of creation of neurons to capture changes during growth and have identified features of neural development that would be significant in establishing long-range connections and network hubs . Continuous monitoring of synaptic connectivity during development is a steep challenge that goes beyond the current approaches for determining the adult connectome of different species [34] , [35] . We present an alternative , more feasible approach of employing global analysis on networks existing at discrete times of growth . With the aid of recent advances , obtaining connectivity information for these time stages could be possible in the near future . Availability of such data on connectivity will permit more detailed analyses , to give an insight into the structural changes unfolding during development . An interesting question that has been raised here is what mechanism ensures accuracy of wiring in late-forming , electrically coupled long-distance connections , and a clear answer is as yet unavailable . We hope that this work will stimulate further experimental and theoretical work on the network development of neural systems and C . elegans in particular .
We have produced a spatial representation of the neuronal network of C . elegans in three-dimensional space , so that the network resembled the anatomical network as much as possible . Three-dimensional coordinates were based on the two-dimensional spatial information of C . elegans neurons [36] , which were updated with spatial information of three neurons that had been excluded . The data for neurons that did not have associated spatial information were obtained based on the spatial data of corresponding bilateral counterparts . The connectivity details from earlier studies [12] , [16] , [37] published in the Worm Atlas was used for the analysis . The ventral cord neuron VC6 that only makes connections through neuro-muscular junctions was not included here . Three loop connections ( connections of a neuron to itself ) were also excluded; as such connections did not influence our spatial and topological measures . Thus a total of 279 neurons and corresponding 2 , 990 connections were used . This included 1 , 584 uni-directional and 1 , 406 bi-directional connections . Biologically , they represent 672 gap junctions , 1962 chemical sysnapses and 376 connections where both gap junctions and chemical synapses exist between the neuron pairs . The latter were represented as bi-directional edges in the connection matrix . Neuro-muscular junctions were not included in the analysis . The coordinate information represents the positions of the soma of the neurons in three-dimensional space . The third dimension of each of the neurons was obtained by treating the body of the worm as a cylinder , guided by the actual shape of the worm . The third spatial coordinate for any neuron was then calculated as a function of the radius of the body of the worm and its known position along the y-axis , as follows:where , r – radius of the worm , was assumed to be constant along the length of the nematode ( 50 µm ) . This returned a positive value of z , and as many neurons in C . elegans have a left or right orientation , the physiological information available [5] was also used in determining the third coordinate . The three-dimensional coordinate was computed so that the spatial properties were closer to reality . Although as C . elegans has a very high length to diameter ratio , the results are unlikely to be affected even if the data had been two-dimensional . The left and right neurons were differentiated into positive and negative values , while those lying along the dorsal and ventral line had their third coordinate as zero . To trace the growth of the network over time , we used the time estimates provided by Sulston et al . ( 1977 & 1983 ) , in the cell lineage charts . The image files representing the lineage charts were read in and then analyzed to obtain the time of creation of each of the neurons . Based on this information , the neuronal network of C . elegans was obtained at different stages of growth . The margin of error in the embryonic lineage , as published , is 10% and 2% in the post-embryonic lineages . We produced spatial representations of the network at times of 350 , 400 , 500 , 800 , 2000 and 2700 minutes after fertilization . The embryo hatches at around 840 minutes [3] , and the network at that time stage was found to be identical to that at 600 minutes in terms of neurons present . The choice of the time interval between the successive stages was based on the number of neurons appearing at different stages . It can be seen from the postembryonic lineage chart that there are very few neurons being created within first 20 hours after hatching ( 1200 minutes after hatching or 2000 minutes from fertilization ) . We therefore chose 2000 minutes post-fertilization as the next stage of development . As all but two neurons in C . elegans derive from the AB cell lineage , cross lineage comparisons were not performed . Although the network contained chemical synapses connecting one neuron to another as well as gap junctions coupling both neurons , the networks were treated as unweighted and directed as more than half of the connections ( 53% ) were unidirectional . Gap junctions were represented as bi-directional . The node degree included both the incoming and outgoing connections . Random networks created for comparative analysis had the same degree distribution as the actual C . elegans network . Neuron identities were randomly shuffled , so that all quantities estimated such as connection-length , birth-time difference , etc , would be modified . A neuron's identity referred to its spatial position and birth-time . At any given time , a connection-pair existed if both neurons forming that connection ( as in the adult ) were present , without however , implying the formation of a synapse between them . The pair-wise , time-difference in the birth of neurons forming each of the 2990 connection-pairs was determined . The values were then separated into ten bins and mean of the time-differences in each bin was computed for enabling comparisons with random networks . The connection length between two connected neurons was the Euclidean distance between them in three-dimensional space . The lengths at each stage were separated into ten bins of size 0 . 12 mm each , the first three were considered to be short distance ( i . e . shorter than 0 . 36 mm ) , the middle four as medium-range and the final three as long-range connections ( i . e . longer than 0 . 84 mm ) . This classification was used as the first three bins and the final three bins displayed a very high frequency of connections , whereas the intermediate bins were sparsely populated . | Long-distance connections are crucial for information processing in neural systems , and changes in long-distance connectivity have been shown for many brain diseases ranging from Alzheimer's to schizophrenia . How do long-distance connections develop ? Traditionally , connections can be formed over long distances using guidance cues for steering axonal growth . Subsequently , other connections can follow those pioneer axons to a target location . Alternatively , two neurons can establish a connection early on , which turns into a long-distance connection as the neural system grows . However , the relative contribution of both mechanisms previously remained unclear . Here , we study long-distance connection development in the neuronal network of the roundworm C . elegans . We find that most neurons that are connected by a long-distance connection could interact and establish contact early on . This suggests that early formation could be an influential factor for establishing long-distance connectivity , with a hypothetical role in neuronal wiring accuracy . Reducing the need for axonal guidance is also likely to reduce metabolic costs during development . We also find that highly-connected neurons ( hubs ) are born early on , potentially giving them more time to host and establish connections . Therefore , neuron birth times can be an important developmental factor for the spatial and topological properties of neural circuits . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience/neurodevelopment",
"neuroscience/theoretical",
"neuroscience"
] | 2011 | Neural Development Features: Spatio-Temporal Development of the Caenorhabditis elegans Neuronal Network |
Plant primary metabolism is a highly coordinated , central , and complex network of biochemical processes regulated at both the genetic and post-translational levels . The genetic basis of this network can be explored by analyzing the metabolic composition of genetically diverse genotypes in a given plant species . Here , we report an integrative strategy combining quantitative genetic mapping and metabolite‒transcript correlation networks to identify functional associations between genes and primary metabolites in Arabidopsis thaliana . Genome-wide association study ( GWAS ) was used to identify metabolic quantitative trait loci ( mQTL ) . Correlation networks built using metabolite and transcript data derived from a previously published time-course stress study yielded metabolite‒transcript correlations identified by covariation . Finally , results obtained in this study were compared with mQTL previously described . We applied a statistical framework to test and compare the performance of different single methods ( network approach and quantitative genetics methods , representing the two orthogonal approaches combined in our strategy ) with that of the combined strategy . We show that the combined strategy has improved performance manifested by increased sensitivity and accuracy . This combined strategy allowed the identification of 92 candidate associations between structural genes and primary metabolites , which not only included previously well-characterized gene‒metabolite associations , but also revealed novel associations . Using loss-of-function mutants , we validated two of the novel associations with genes involved in tyrosine degradation and in β-alanine metabolism . In conclusion , we demonstrate that applying our integrative strategy to the largely untapped resource of metabolite–transcript associations can facilitate the discovery of novel metabolite-related genes . This integrative strategy is not limited to A . thaliana , but generally applicable to other plant species .
Plants produce a large array of structurally and biologically diverse metabolites . Largely due to the missing underlying biochemistry , the genes encoding metabolite-related enzymes or regulatory proteins are known for only a fraction of the metabolites . With the development of metabolomic and genomic tools , alternative approaches have been successfully applied to identify genes encoding enzymes involved in specific biochemical pathways [1–6] . Metabolite levels can be used as quantitative traits , and quantitative trait locus ( QTL ) mapping of metabolite levels using structured populations facilitates the identification of the genomic regions associated with the metabolic variation [7–9] . However , given the relatively low resolution reached using this approach [10] , the cloning of single causal genes has rarely been achieved . Genome-wide association studies ( GWAS ) , due to the presence of many more meiotic events present in natural populations during historical recombination , allow a more refined QTL resolution [11 , 12] . However , the limitation of GWAS , especially in self-mating biological systems such as Arabidopsis thaliana , lies not only in the generation of false positive genotype‒phenotype associations because of the confounding effects of population structure [13 , 14] , but also in the poor resolution reached if associated SNPs are found in extensive islands of haplotypes in linkage disequilibrium ( LD ) [15–17] . Epistasis and lack of natural variation can also result in a high false-negative rate , wherein loci with previous experimental validation for specific traits are not found in GWAS [17 , 18] . In order to take advantage of both resources , a growing number of recent reports have successfully combined mQTL from bi-parental segregating populations and natural populations to elucidate the biochemical nature of metabolite traits [19–21] . Due to limited segregating allelic diversity in bi-parental segregating populations such as recombinant inbred lines ( RIL ) and introgression lines ( IL ) , the validation of GWAS results is not possible in every case [22] . The combination of both GWAS and bi-parental segregating populations , however , is advantageous in reducing the false-positive associations in GWAS due to the fact that in many cases , even after population structure correction , some individuals might be more related to each other than individuals are related on average [23 , 24] . Aside from genetic evidence , the integration of additional forms of genome-scale data , such as metabolite and transcript data , has been applied to detect metabolite‒gene correlations and to largely reduce false-positive correlations [6 , 25–27] . To date , network analysis has mainly focused on correlations between transcripts and transcripts ( i . e . co-expression networks ) [28] , and correlations between metabolites and metabolites ( i . e . metabolic networks ) [29] . The study of metabolite‒transcript correlations is yet to be fully explored . Detection and elucidation of metabolite‒transcript correlations can yield important clues regarding the consequences of altered environmental conditions on metabolite levels in organismal systems [30] . Although a few pioneering investigations have tried to apply this integrative strategy [6 , 19 , 31–35] , the power of combined results from the two orthogonal approaches , i . e . quantitative genetics and metabolite‒transcript networks , for the elucidation of the genetic architecture of metabolite traits has not been fully exploited . Based on first principles , the overlap of results obtained using these two approaches in parallel should increase their statistical confidence . In order to test this hypothesis , we analyzed 94 primary metabolites in a densely genotyped collection of 314 natural A . thaliana accessions , and used these metabolite levels as phenotypic traits to conduct a GWAS with 200K single nucleotide polymorphisms ( SNPs ) . The resulting metabolite‒gene associations from the GWAS were compared and validated with mQTL which had been described before using two A . thaliana populations ( 429 RILs and 97 ILs ) [8] . In parallel , metabolite‒transcript correlation networks were constructed based on reported transcriptome and metabolome levels of A . thaliana as a function of changing environments [36] . Correlations identified between metabolites and transcripts were applied as an additional and independently derived filtering criterion to further support identified metabolite‒gene associations . Furthermore , we applied a statistical analysis framework to test and compare the performance of all single methods ( GWAS , RIL , IL , and network analysis ) with that of the combined strategy by using precision , recall and F-measure . The results indicate that the combined strategy ( the strategy to predict genes supported by network analysis and at least one mapping approach ) exhibits an overall better performance as compared to the single methods , boasting increased sensitivity and accuracy . Using this integrative strategy ( Fig 1 ) , 92 main metabolite‒gene associations were identified . The validity of the approach was confirmed by analyzing two loss-of-function mutants for two novel genes . In conclusion , this study serves as a proof of concept , demonstrating that by integrating two orthogonal approaches , novel metabolite‒gene associations can be obtained with a robust statistical significance .
Information about the A . thaliana accessions used in this study is provided in S1 Table . 94 metabolic features , comprising 26 amino acids , 23 organic acids , 17 sugars , three amines , four other metabolites with known , and 21 with unknown , chemical structure , were reproducibly detected in rosette material of 314 A . thaliana accessions . Metabolite ID , name , classification , and quantification mass used for the following data analysis are shown in S2 Table . Normalized metabolite data across 314 accessions are shown in S1 Dataset . Those metabolites belonging to one functional class were highly correlated , demonstrated by the fact that ten amino acids , nine sugars , and some organic acids were clustered together , respectively ( Fig 2 ) . The metabolic profiles of the accessions revealed that 37 . 8% of all annotated metabolites were associated with at least one locus at a genome-wide significance level of p ≤ 5 . 01 × 10−6 ( LOD = 5 . 3 ) , calculated by a mixed linear model . This model includes principal components as fixed effects to account for population structure ( commonly called the “Q” matrix ) [37] , and a kinship matrix ( commonly called the “K” matrix ) [38] . In order to test how well the model used in GWAS accounts for population structure and familial relatedness across the accessions , we generated quantile‒quantile ( QQ ) plots for all 94 metabolite traits . We observed that the majority of points in the QQ plot lay on the diagonal line for all the metabolite traits , indicating that spurious associations due to population structure and familial relatedness were largely corrected . The SNPs in the upper right section of the QQ plot deviating from the diagonal were most likely associated with the metabolite traits in the study . The QQ plots for the metabolite traits further discussed here are shown in S1 Fig . In total , 117 distinct SNP‒trait associations , resulting in 617 gene‒metabolite-trait associations , were identified ( S3 and S4 Tables ) . In the following , two representative examples of these associations will be described in more detail . A strong association ( p = 4 . 11 × 10−6 , LOD = 5 . 39 ) between SNP m59466 at the AT2G17265 locus and the metabolite trait homoserine was detected . Gene AT2G17265 encodes a homoserine kinase ( HSK ) that catalyzes the chemical reaction with the substrate L-homoserine to produce O-phospho-L-homoserine ( HserP ) , a compound at the branching point of methionine and threonine biosynthesis [39] . A loss-of-function mutant of this gene results in higher levels of the amino acid homoserine [40] , which is in line with the observation described here . Tyramine was significantly associated with SNP m154079 ( p = 1 . 28 × 10−9 , LOD = 8 . 89 ) ( Fig 3B ) . Lead SNP m154079 and other significantly associated SNPs , are located in locus TyrDC ( L-tyrosine decarboxylase 1 , AT4G28680 ) , which was reported to encode a stress-induced tyrosine decarboxylase [41] . This enzyme catalyzes a dicarboxylic reaction on tyrosine to release CO2 and produce tyramine ( Fig 3A ) . There are nine SNP markers in this gene identified by high-throughput genotyping ( Fig 3C ) . Among these nine SNPs , three SNPs leading to changes in the amino-acid sequence are located in the fifth , tenth , and eleventh exon , respectively . The first polymorphism variant ( T/C , m154077 ) results in a serine-to-proline substitution , the second SNP variant ( A/C , m154081 ) causes a serine-to-arginine exchange , and the third SNP ( C/G , m154082 ) brings about a more subtle substitution , from serine to threonine ( Fig 3C ) . Linkage disequilibrium ( LD ) analysis of the mapped genomic region for the tyramine trait revealed that the three exonic SNPs ( m154077 , m154081 , and m154082 ) are highly and significantly linked with the lead SNP m154079 ( r2 > 0 . 75 , p < 0 . 001 ) ( Fig 3D ) . This finding suggests that they are likely to constitute the functional variation underlying this association . However , it is still difficult to completely exclude other variants surrounding this region . Therefore , we took the nine SNP markers in TyrDC to conduct haplotype analysis for the accessions . These nine SNPs give rise to 19 possible haplotypes , eight of them being informative haplotypes defined by more than two accessions within a haplotype . The haplotypes can be further classified into five main clusters according to the haplotype sequence similarities . Cluster II ( H2 , H3 , H5 , H9 , and H18 ) presents significantly higher levels of tyramine than Cluster I ( H1 , H6 , H12 , and H17 ) , Cluster III ( H4 , H10 and H18 ) , as well as two other minor clusters ( Cluster IV and V ) ( Fig 3E ) . Taken together , both the associated SNPs and the haplotype analysis support TyrDC as a candidate gene controlling tyramine levels . One of the main aims of this study is to discover true and novel metabolite‒gene associations involved in A . thaliana primary metabolism by integrating various quantitative genetics and network approaches . To this end , we compared the GWAS obtained in this study with results reported previously based on the analysis of two A . thaliana bi-parental populations: 429 RIL and 97 IL derived from accessions Col-0 and C24 [8] . Out of the 40 metabolite traits described in the RIL dataset , 32 overlap with those of the GWAS , whereas 50 metabolites overlap between the GWAS and the IL data ( cf . S5 and S6 Tables for the mQTL identified in RIL and IL in [8] , respectively ) . It has been described that in many cases the Bonferroni threshold is too stringent for quantitative gene identification [42] . We therefore decided to test the performance of the GWAS when different LOD thresholds were applied based on the four reference gene lists ( RGL1– RGL 4 ) derived from KEGG metabolic pathways ( see Materials and Methods , section “Procedure setup for determining method performance”; cf . S7 Table ) . GWAS performance using various LOD thresholds was evaluated by three statistics: precision , recall and F-measure . These three parameters , as well as the number of correctly predicted metabolites across all tested GWAS LOD thresholds ( from 2 . 0 to 5 . 3 ) were recorded ( S8 Table ) . The measureable values for these four statistics increased with lower thresholds , but were not changed with LOD thresholds lower than 3 . 0 ( S8 Table ) . Additionally , we tested the metabolite-wise precision for each metabolite when applying LOD thresholds ranging from 3 . 0 to 5 . 3 . As shown in S2 Fig , the metabolite-wise precision was very low when applying relatively low LOD thresholds ranging from 3 . 0 to 4 . 0 , implying that the chance of finding true functional related genes from a relatively large mapped locus is very low . LOD threshold 4 . 5 was selected for further integration with other methods , because it can balance well the trade-off between obtaining more correctly predicted metabolite traits and discovering the causal genes for metabolite traits more precisely . Comparison between different datasets was conducted using both the significant LOD threshold after Bonferroni correction ( LOD = 5 . 3 ) and the suggestive LOD threshold ( LOD = 4 . 5 ) . Common loci obtained by comparing QTL results from the GWAS , RIL , and IL datasets using the two GWAS LOD thresholds mentioned above are listed in Table 1 and S9 Table . One example we would like to point out is the QTL detected for nicotinic acid , located on chromosome 5 , with 41507 bp , supported by GWAS , RIL , and IL results together . Quantitative genetic analysis establishes the association between a locus/gene and a trait ( here: metabolite ) by testing the co-occurrence between trait variants and genetic markers . As an orthogonal , albeit still statistics-based approach , we decided to test the associations of metabolites with transcripts resulting from metabolite‒transcript correlation networks for their power to identify candidate genes involved in the synthesis and/or degradation of a given metabolite . Though this approach has been successfully used in many instances with secondary metabolites [25–27 , 43 , 44] , the comparable investigation of primary metabolites has not been fully explored . Metabolite and transcript data were obtained from a previously published study from our group , in which the metabolomic and transcriptomic responses of A . thaliana towards eight environmental conditions differing in temperature and light intensity were recorded at a high kinetic time-resolved resolution [36] . Significantly changed metabolites across 23 time points in each condition at a significance level of 0 . 05 after multiple correction , together with all 15 , 089 transcripts , were used to construct condition-specific networks ( eight individual networks in total ) . The numbers of primary metabolites and transcripts , as well as the statistically significant Pearson Correlation Coefficient ( PCC ) thresholds derived from permutation test for the individual networks , are shown in S10 Table . Multiple metabolite‒transcript correlations shared across different conditions were detected , suggesting conserved associations , 219 of them being maintained across all eight conditions ( S11 Table ) . These highly robust correlations found between transcripts and primary metabolites indicate conserved/tight regulation in A . thaliana . In order to test the likelihood of these correlations to be functionally significant , all metabolite‒transcript correlations detected by network analysis were compared with the GWAS . The common associations supported by both GWAS and network analysis under the two GWAS LOD thresholds are presented in S12 and S13 Tables , respectively . In the following , we will describe some exemplary results in more detail . Temperature- and light-stress treatments were abbreviated as follows: ( i ) 4°C and darkness ( 4-D ) , ( ii ) 21°C and darkness ( 21-D ) , ( iii ) 32°C and darkness ( 32-D ) , ( iv ) 4°C and normal light ( 4-L ) , ( v ) 21°C and low light ( 21-LL ) , ( vi ) 21°C and normal light ( 21-L ) , ( vii ) 21°C and high light ( 21-HL ) , and ( viii ) 32°C and normal light ( 32-L ) . Network data revealed a conserved and significant correlation between SPMS ( spermidine synthase 3 , AT5G53120 ) and β-alanine . For six conditions , high PCCs were observed ( 21-L , –0 . 61; 21-D , –0 . 75; 4-L , –0 . 78; 4-D , –0 . 87; 32-D , –0 . 64; 21-LL , –0 . 86 ) . Furthermore , this association is in agreement with the GWAS data . SPMS is annotated as encoding a novel spermine synthase and is a paralog of previously characterized spermidine synthases , SPDS1 and SPDS2 [45 , 46] . The protein that SPMS encodes can catalyze the reaction from spermine to spermidine , and thus fuel the subsequent two steps in β-alanine biosynthesis . A robust link between tyrosine and TAT7 ( tyrosine aminotransferase 7 , AT5G53970 ) was observed in five out of eight condition-specific networks ( the PCCs observed were: 4-L , 0 . 68; 21-LL , –0 . 65; 21-L , –0 . 69; 21-HL , –0 . 57; 32-D , 0 . 67 ) . TAT7 encodes a tyrosine aminotransferase as proven by both loss-of-function mutants and an in vitro recombinant protein assay , whereby it was suggested that TAT7 is a tyrosine-specific aminotransferase not involved in tyrosine biosynthesis , but rather in the utilization of tyrosine for other metabolic pathways , e . g . tocopherol biosynthesis [47] . Levels of tyrosine , as a central primary metabolite , can be influenced by many factors . Its profiles observed for the five environmental conditions indicated that temperature may be the more influential element for tyrosine content rather than light intensity ( S3 Fig ) . The correlation between tyrosine and TAT7 is also supported by the RIL dataset . Another strong correlation discovered by the network analysis was between tyrosine and HGO ( homogentisate 1 , 2-dioxygenase , AT5G54080 ) , displaying high positive correlations in three darkness conditions independent of temperature , and in another low-light stress condition: 4-D , 0 . 85; 21-D , 0 . 78; 21-LL , 0 . 74; 32-D , 0 . 73 . The profiles of tyrosine and HGO across 23 time points in these four conditions are shown in Fig 4A . HGO is reported to encode a homogentisate 1 , 2-dioxygenase that can convert homogentisate to malylacetoacetate , and is likely to be involved in tyrosine degradation [48] . A merged network was constructed by combining the four condition-specific networks in 4-D , 21-D , 21-LL and 32-D stress conditions ( Fig 4B ) . In order to represent the most robust correlations with tyrosine , only transcripts that are connected with tyrosine in all four conditions and metabolites that are connected with HGO in at least two conditions are displayed in this zoom-in merged metabolite‒transcript correlation network ( Fig 4B ) . The merged network shows that the majority of associated transcripts belong to functional groups encoding amino-acid metabolism and protein degradation/post-translation/transport/targeting proteins , which is in line with the metabolic pathway for tyrosine . Again , the link between tyrosine and HGO is also supported by the RIL dataset . A major goal of this study was to test the power of integrating results obtained by various quantitative genetics and network approaches for increased robustness and sensitivity . The performance of each single method and of the combined strategy ( network analysis and at least one mapping approach ) was tested by calculating precision , recall and F-measure , widely applied as scoring metrics in pattern recognition and information retrieval [49] , based on different LOD thresholds ranging from 3 . 0 to 5 . 3 . As a comparison set , we built four reference gene lists ( RGL1 , RGL 2 , RGL 3 , and RGL 4 ) for all the metabolites shown in the different datasets based on KEGG metabolic pathway [50] ( see Materials and Methods , section “Procedure setup for determining method performance”; cf . S7 Table ) . As shown in S4 Fig for precision , S5 Fig for recall , and Fig 5 for F-measure , the combined strategy performs better than any other single method based on RGL2 ( LOD ranging from 4 . 5 to 5 . 3 ) , RGL3 ( LOD ranging from 3 . 5 to 5 . 3 ) and RGL4 ( LOD ranging from 3 . 8 to 5 . 3 ) , except in the case of RGL1 , in which the network approach performs better than the combined strategy ( LOD ranging from 3 . 4 to 5 . 3 ) , indicating that the network approach is superior to the combined strategy with regard to providing information about genes directly linked to the metabolite ( neighbor transcripts ) . It is however important to note that the combined strategy performed better when applying the two selected LOD thresholds ( significant threshold 5 . 3 and suggestive threshold 4 . 5 ) in this study based on RGL2 to RGL4 . In order to test whether the combined strategy has a better prediction ability of true associations as compared to random methods , we applied a randomization test in which we shuffled the related genes for all the annotated metabolites in the combined dataset , and obtained the permuted F-measure by comparing the shuffled related gene list with the four reference gene lists . After 10 , 000 iterations , the actual F-measure was compared with the permuted F-measure 10 , 000 times and an empirical p-value was estimated . Table 2 shows the actual F-measure , permuted F-measure , and p-values when applying LOD thresholds 5 . 3 and 4 . 5 . The results suggest that all the actual F-measures are significantly higher than the permuted ones , which means that the combined strategy using both significant and suggestive LOD thresholds performs significantly better than the randomized method . The metabolite-wise precision is another important determinant parameter allowing us to compare the performance of different methods . Therefore , the metabolite-wise precision was calculated and compared across all the individual methods and the combined strategy . The comparison between methods for metabolite-wise precision based on all four reference gene lists and applying both significant and suggestive LOD thresholds ( 5 . 3 and 4 . 5 ) is shown in S6 and S7 Figs . When applying LOD threshold 4 . 5 , the metabolite-wise precision of the combined strategy is significantly higher than that of any other single methods based on RGL3 and RGL4 ( combined strategy and network analysis: p-values are 0 . 029 and 0 . 050 based on RGL3 and RGL4 , respectively ) , and showing the highest trend in the combined strategy based on RGL2 . When using LOD threshold 5 . 3 , the metabolite-wise precision of the combined strategy shows a trend higher than any other single method's based on RGL3 and RGL4 . Overall , the results indicate that the combined strategy of integrating quantitative genetics and network analysis can largely improve the power of detection of true metabolite‒gene associations involved in A . thaliana primary metabolism . All associations between genes and primary metabolites detected by GWAS were cross-validated with the results from network analysis and from metabolic QTL results from RIL and IL populations using the two GWAS LOD thresholds described above . All associations supported by the four datasets are summarized in S14 and S15 Tables , showing the overall comparison based on the two GWAS LOD thresholds evaluated . Fig 6 represents the overall chromosomal distribution of 76 selected candidate genes in 92 main associations resulting from this study . Among them , 86 associations are supported by at least two of the approaches . One chromosomal hotspot supported by GWAS , network analysis , and QTL from IL population becomes immediately evident . It is located on chromosome 4 , from 8231017 bp to 8366653 bp , and was previously reported to be related to biomass , resistance to a broad range of pathogens from different phyla [51] , and to general metabolic activity [8] . Additional detailed information for candidate associations discussed in the text is listed in Table 3 . Validation of all associations disclosed is beyond the scope of this study . As a proof of concept , we focused on two promising candidate genes to experimentally validate our strategy and results . The first candidate gene is HGO ( AT5G54080 ) , associated with tyrosine in our analysis ( Fig 4A and 4B ) . Although the function of HGO was partly elucidated [48] , genetic evidence based on mutant analysis to explore its metabolic roles in A . thaliana is still lacking . Therefore , a knockout line ( SALK_027807 ) for HGO was grown in parallel with wild-type Col-0 plants under control ( 21-L ) and stress ( 32-D ) conditions ( due to tyrosine showing dramatic accumulation in 32-D , the latter was chosen as the representative stress condition; see S8 Fig ) , whereupon both lines were subjected to GC‒MS metabolomic analysis . As evident from Fig 4C , tyrosine increased in both Col-0 and hgo plants under 32-D condition as compared to normal condition , in agreement with our previous report [36] . More importantly , however , we observed that tyrosine levels in the hgo mutant were significantly higher as compared to wild-type plants under normal condition ( p = 0 . 002 ) , and had increasing trends in the hgo mutant plants under stress conditions ( 32-D ) ( Fig 4C ) . These results are in line with the involvement of HGO in tyrosine degradation and confirm the usefulness of integrating information from network analysis and quantitative genetics approaches . The second example concerns the gene AGT2 ( alanine:glyoxylate aminotransferase 2 , AT4G39660 ) . The association between AGT2 and β-alanine is supported by both GWAS and network analysis . In the networks , this association displays high positive correlations under four conditions ( 21-D , 0 . 818; 21-LL , 0 . 649; 21-L , 0 . 849; 32-D , 0 . 613 ) , representing very robust correlations between AGT2 and β-alanine . In the GWAS , β-alanine mapped to a locus spanning 41 kb on chromosome 4 . We considered three candidate genes encoding metabolic enzymes enclosed in this locus ( AT4G39640 , gamma-glutamyl transpeptidase 1 , GGT1; AT4G39650 , gamma-glutamyl transpeptidase 2 , GGT2; AT4G39660 , AGT2 ) . AGT2 is the only one supported also by network analysis , for which reason we selected it as the most promising candidate gene related to β-alanine . There are seven SNP markers in AGT2 , five of them showing significant associations with β-alanine . Notably , one of the SNPs ( m160527 , position 18406944 bp on chromosome 4 ) can result in amino-acid substitution from proline ( non-polar ) to serine ( polar ) with the nucleotide variant from cytosine ( C ) to thymine ( T ) . This suggests that SNP m160527 could be the causative SNP in AGT2 . Based on sequence homology , AGT2 was annotated as a putative alanine:glyoxylate aminotransferase . An attempt to functionally characterize AGT2 , using an in vitro enzymatic assay , did not identify the enzyme as an alanine aminotransferase [55] . To date , the function of AGT2 still remains unknown . Recently , Wen et al [34] also found that the close homolog of AGT2 in maize ( ZM01G05170 ) strongly mapped to β-alanine; this finding was further validated by their linkage analysis and eQTL ( expression QTL ) results . Therefore , we conducted a phylogenetic analysis on AGT2 and its homologs from A . thaliana and from other plant species to explore the evolutionary history of this gene in plant taxa ( S9 Fig ) . The first feature detected is the presence of at least two clusters , including sequences from both monocots and dicots , confirming that AGT2 belongs to a multigene family . Interestingly , AGT2 clustered together with the maize sequence ZM01G05170 reported by Wen et al [34] , indicating that AGT2 is the strict ortholog to the characterized enzyme in maize . In order to test for the role of AGT2 in β-alanine metabolism , two independent loss-of-function lines ( SALK_003381 and SALK_035035 ) for AGT2 , plus wild-type plants , were grown under normal ( 21-L ) and stress ( 32-D ) conditions ( 32-D was selected as a representative stress condition because β-alanine strongly accumulated under this stress; see S10 Fig ) . β-alanine significantly increased in Col-0 plants under stress condition comparing with plants grown under control condition ( p = 8 . 15E-13 ) ( Fig 7 ) , in agreement with previous observations [36] . More importantly , however , both KO plants displayed a very strong increase in β-alanine independent of the growth condition ( Fig 7 ) ( statistical significance levels by pair-wise comparison: SALK_003381_N & Col-0_N: 8 . 15E-13; SALK_035035_N & Col-0_N: 8 . 15E-13; SALK_003381_S & Col-0_S: 1 . 50E-12; SALK_035035_S & Col-0_S: 1 . 60E-12 ) . These results thus suggest that AGT2 is involved in β-alanine metabolism , reinforcing the utility in combining network and quantitative genetics analyses .
Metabolites are the terminal products of cellular regulatory processes , and their levels can be regarded as the ultimate responses of biological systems to environmental changes in a given genetic background , and thus serve as a link between subtle genotypes and visible phenotypes [56] . The genetic regulation of primary metabolites ( essential for the viability of the cell ) and secondary metabolites ( required for the viability of the organism in the environment ) is different . This derives from the fact that secondary metabolites are highly specific for particular genotypes , while primary metabolites are synthesized through common pathways and influenced by multiple and complicated factors [57] . Here , a GWAS strongly suggests polygenic regulation of primary metabolism in A . thaliana , owing to the fact that the individual metabolite traits mapped to multiple loci ( each primary metabolite was mapped to 1 . 4 and 3 . 3 loci on average when applying the significant/suggestive LOD thresholds 5 . 3 and 4 . 5 , respectively ) , which is in agreement with previous studies [32 , 58] . The centrality and complexity of primary metabolism in A . thaliana makes it difficult to detect the true genetic‒metabolic relationships by a single method [6] . Within this study , we integrated GWAS based on a collection panel of 314 natural A . thaliana ecotypes , metabolite–transcript correlation network analysis for eight different environmental conditions based on data in [36] , and mQTL results from two structured populations ( RIL and IL; [8] ) . In order to test the validity of the combination of the two orthogonal approaches ( quantitative genetics and network analysis ) in comparison to each single method , we generated a statistical framework using four reference gene lists based on KEGG metabolic pathways ( Materials and Methods , section “Procedure setup for determining method performance” ) . The performance of the different methods was evaluated and compared by precision , recall and F-measure , widely applied in pattern recognition and information retrieval [49] . We observed improved performance of the combined strategy ( the strategy to predict genes supported by at least one mapping approach and network analysis ) based on three out of four reference gene lists we applied ( S4 and S5 Figs and Fig 5 ) . Although the combined strategy did not perform better than network approach based on RGL1 , this indicates that network analysis outperforms the quantitative genetics methods in detecting enzymes directly linked to a given metabolite . Still , the combined strategy exhibited an overall better performance . Furthermore , the performance of the combined strategy was confirmed by permutation test ( Table 2 ) . Taken together , the statistical framework that we applied here illustrates that the combined strategy increases the sensitivity and robustness of candidate gene discovery . Using the resulting metabolite–transcript associations , we identified connections between primary metabolites and structural genes that were previously reported to take part in the biosynthesis of the respective metabolites . For instance , the association between homoserine and AT2G17265 ( HSK ) supported by GWAS per se [40]; nicotinic acid and AT5G14760 ( L-aspartate oxidase , AO ) supported by GWAS , RIL , and IL results [54]; glycine and AT1G62800 ( aspartate aminotransferase 4 , Asp4 ) [52] , nicotinic acid and AT5G14780 ( formate dehydrogenase , FDH ) [53] supported by all four datasets , illustrating the validity and feasibility of our combined strategy . Our integrative strategy offers a valuable tool not only for addressing previously reported primary metabolite‒gene associations , but also for discovering novel and under-explored candidate associations/genes involved in the regulation of A . thaliana primary metabolism . We found a strong association between tyramine and TyrDC ( AT4G28680 ) in GWAS ( Fig 3B ) . Analysis of SNPs leading to amino-acid substitution ( Fig 3C ) , LD analysis ( Fig 3D ) , and haplotype analysis ( Fig 3E ) supported TyrDC as the most prominent candidate gene for the metabolic trait tyramine . TyrDC was previously shown by enzymatic assay to encode a protein that catalyzes the conversion of tyrosine to tyramine [41]; our GWAS further provides genetic evidence for the gene annotation . Another two candidate genes that are also involved in tyrosine metabolism were discovered by network analysis , both of them being supported by the RIL dataset as well ( Fig 8A ) . TAT7 ( AT5G53970 ) , encoding a tyrosine aminotransferase whose products are 4-hydroxyphenylpyruvate ( 4-HPP ) and L-glutamate [47] , is linked to tyrosine in five conditions from the network analysis . HGO , previously shown to convert homogentisate to malylacetoacetate using in vitro enzymatic assays [48] , is connected with tyrosine in four conditions from our network analysis . Using knockout lines , we further verified the function of HGO ( Fig 4C ) in tyrosine degradation . With the current knowledge on tyrosine synthesis and catabolism pathway , we could simultaneously identify three key genes in tyrosine degradation ( Fig 8A ) . The detection of all these three critical genes manifests the strength of the integrative strategy based on the guilt-by-association principle [59] . We observed the strong correlation between tyrosine and TAT7 in five conditions , showing negative correlations in three of them ( 21-LL , 21-L , and 21-HL; the common feature is 21°C ) , and two positive correlations in the stress conditions 4-L and 32-D ( S3 Fig ) . One of the possible explanations for the flip of correlations for the same metabolite–transcript pair is that metabolic reactions , especially in primary metabolism , are regulated on different levels , and the metabolic fluxes are constantly changing when plants are exposed to various environmental stresses . It seems that a feedback loop regulation might control TAT7 expression in a temperature-dependent manner . This is not always reflected in the actual tyrosine levels under different physiological conditions . In the present study , we could identify two candidate genes ( SPMS , AT5G53120 and AGT2 , AT4G39660 ) involved in the β-alanine metabolic pathway , both supported by GWAS and network analysis ( Fig 8B ) . In plants , three predicted pathways for β-alanine biosynthesis have been reported , including uracil degradation , polyamine oxidation , and propionate catabolism , but only the last enzyme in the uracil degradation pathway was studied in detail [60] , leaving β-alanine metabolism in plants largely unexplored . The first candidate gene we identified is SPMS , reported to catalyze the conversion from spermine to spermidine by elongation of the polyamine chain [45 , 61] . Notably , β-alanine can be produced by spermidine within the subsequent two reaction steps . Although SPMS has already been well characterized before , this example clearly demonstrates the power of the integrative strategy for detecting biochemically relevant associations between genes and metabolites that are not directly linked in a pathway ( Fig 8B ) . We also identified a strong association between β-alanine and AGT2 . Based on sequence homology , AGT2 was annotated as a putative alanine:glyoxylate aminotransferase . Plant leaf peroxisomes are hypothesized to contain at least four aminotransferase activities , including Ser:glyoxylate aminotransferase ( SGT ) , Glu:glyoxylate aminotransferase ( GGT ) , Ala:glyoxylate aminotransferase ( AGT ) , and Asp:glyoxylate aminotransferase ( AspAT ) [62 , 63] . Animals possess two structurally distinct types of AGTs: AGT1 and AGT2 . Previous kinetic analysis of A . thaliana AGT1 suggested that this protein mainly uses the substrates Ser and glyoxylate with SGT activity , while the function of AGT2 remained obscure [55 , 64] . We further tested this association using two independent knockout lines of AGT2 . Both lines showed remarkable accumulation of β-alanine in comparison with wide-type plants , both in control and in stress conditions ( Fig 7 ) , supporting the association between β-alanine and AGT2 . In A . thaliana , AGT2 shows sequence homology to AGT3 ( AT2G38400 ) and PYD4 ( AT3G08860 ) . Interestingly , PYD4 is predicted to have β-alanine aminotransferase activity . Additionally , in maize , β-alanine mapped to a genetic locus harboring the homolog gene ( ZM01G05170 ) of AGT2 , which was further supported by linkage analysis and eQTL results [34] . In our phylogenetic analysis ( S9 Fig ) , AGT2 clustered together with its maize homolog ZM01G05170 reported by Wen et al [34] , suggesting that both genes maintain the same function . PYD4 ( AT3G08860 ) clustered in a separate branch among sequences from other dicots before the speciation event ( S9 Fig ) . Taking all the above evidence together with our findings using network analysis , GWAS and analysis of knockout lines , we can conclude that AGT2 might be involved in β-alanine metabolism , but its decisive role as a β-alanine aminotransferase still needs to be confirmed by biochemical assays . It seems that AGT2 , PYD4 and AGT3 , together with the maize homolog ( ZM01G05170 ) , are part of a large gene family of β-alanine aminotransferases , conserved both in monocot and dicot plants ( S9 Fig ) . Nowadays , GWAS is steadily becoming a common practice to identify the underlying genetic loci determining a plethora of phenotypic traits , but causal-gene identification still remains an obstacle . To overcome this , we present here a strategy based on the combined use of GWAS , metabolite–transcript correlation network analysis , and linkage mapping using structured populations , facilitating candidate association selection and providing functional and biological insight into A . thaliana primary metabolism . We demonstrate , using statistical analysis , that the combined strategy outperforms the single methods . Based on hypotheses generated by this comprehensive strategy , the functions of two novel genes were validated by transgenic methods . Our results illustrate that the integrative strategy described here offers an invaluable tool for advancing our knowledge of A . thaliana primary metabolism , a tool that can be applied to other plant species for functional elucidation of unknown genes . To our best knowledge , it is the first report to apply this combined strategy with all the above potent sources to cross-validate and prioritize candidate associations involved in A . thaliana primary metabolism .
Metabolite extraction and derivatization from A . thaliana leaves using GC‒MS were performed as described by Lisec et al [70] . The GC‒MS data were obtained using an Agilent 7683 series auto-sample ( Agilent Technologies , http://www . home . agilent . com ) , coupled to an Agilent 6890 gas-chromatograph‒Leco Pegasus two time-of-flight mass spectrometer ( Leco; http://www . leco . com/ ) . Identical chromatogram acquisition parameters were applied to those previously used [36] . Chromatograms were exported from LECO CHROMATOF software ( version 3 . 34 ) to R software . Ion extraction , peak detection , retention time alignment and library searching were obtained using the TargetSearch package from Bioconductor [71] . Day-normalization and sample median-normalization were conducted; the resulting data matrix was used for further analysis . Metabolite intensity data after transformation and normalization were used for ANOVA to test the significance levels of metabolite changes in knockout and Col-0 plants under normal and stress conditions , following by correction for multiple comparisons using the “p . adjust” function in R ( http://www . r-project . org/ ) . Subsequently , pair-wise comparison was conducted by the Tukey HSD tests using the “TukeyHSD” function in R . Target A . thaliana protein sequences in this study were extracted from The Arabidopsis Information Resource ( TAIR , https://www . arabidopsis . org/ ) . The sequences of all biochemically characterized alanine aminotransferases and AGT-like proteins from other species were extracted from NCBI ( http://www . ncbi . nlm . nih . gov/ ) and PLAZA 3 . 0 ( http://bioinformatics . psb . ugent . be/plaza/ ) . Amino-acid sequences were aligned using the CLUSTALW ( version 1 . 83 ) program . A maximum likelihood tree was constructed using MEGA 7 . 0 software with all default parameters . | Primary metabolites are key elements in plant growth and development . Our partial understanding of their biosynthesis and regulation derives mostly from biochemical and genetic modification experiments . The recent generation of large-scale genome-wide data , along with the advances in mass-spectrometry techniques , allows us to treat metabolite levels as quantitative traits and to link them to genomic information , resulting in the identification of so-called metabolic quantitative trait loci ( mQTL ) . These mQTL contribute to the discovery of new biosynthetic and regulatory elements that control the plant's metabolic landscape . Low mapping resolution , however , normally limits discovery to one causal gene per locus . Here , we utilize a complementary strategy to support the identification of casual genes by genetic mapping . We measured metabolite levels in 314 A . thaliana accessions , then used genome-wide association mapping to identify mQTL . We next used previously published results from a time-course stress study to construct metabolite‒transcript correlation networks . Integrating data from both approaches enabled us to select candidate genes linked to specific metabolites . We finally validated two of the novel gene‒metabolite associations using knockout lines . We demonstrated that by using an integrative strategy , we can validate previously characterized gene–metabolite associations , and most importantly , identify novel associations between metabolites and genes . The combined quantitative genetics and metabolite‒transcript networks that we present here can be applied to other organisms and fields of research . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome-wide",
"association",
"studies",
"chemical",
"compounds",
"metabolic",
"networks",
"brassica",
"enzymology",
"organic",
"compounds",
"model",
"organisms",
"tyrosine",
"metabolites",
"network",
"analysis",
"genome",
"analysis",
"enzyme",
"metabolism",
"amino",
"aci... | 2016 | Combined Use of Genome-Wide Association Data and Correlation Networks Unravels Key Regulators of Primary Metabolism in Arabidopsis thaliana |
Enhanced protein aggregation and/or impaired clearance of aggregates can lead to neurodegenerative disorders such as Alzheimer’s Disease , Huntington’s Disease , and prion diseases . Therefore , many protein quality control factors specialize in recognizing and degrading aggregation-prone proteins . Prions , which generally result from self-propagating protein aggregates , must therefore evade or outcompete these quality control systems in order to form and propagate in a cellular context . We developed a genetic screen in yeast that allowed us to explore the sequence features that promote degradation versus aggregation of a model glutamine/asparagine ( Q/N ) -rich prion domain from the yeast prion protein , Sup35 , and two model glycine ( G ) -rich prion-like domains from the human proteins hnRNPA1 and hnRNPA2 . Unexpectedly , we found that aggregation propensity and degradation propensity could be uncoupled in multiple ways . First , only a subset of classically aggregation-promoting amino acids elicited a strong degradation response in the G-rich prion-like domains . Specifically , large aliphatic residues enhanced degradation of the prion-like domains , whereas aromatic residues promoted prion aggregation without enhancing degradation . Second , the degradation-promoting effect of aliphatic residues was suppressed in the context of the Q/N-rich prion domain , and instead led to a dose-dependent increase in the frequency of spontaneous prion formation . Degradation suppression correlated with Q/N content of the surrounding prion domain , potentially indicating an underappreciated activity for these residues in yeast prion domains . Collectively , these results provide key insights into how certain aggregation-prone proteins may evade protein quality control degradation systems .
Protein misfolding disorders involve the conversion of native proteins into non-native , deleterious forms . Some misfolded proteins form highly ordered amyloid aggregates , stabilized by intermolecular cross-β sheets . Once formed , these aggregates can convert remaining soluble proteins to the aggregated form via a templated misfolding mechanism [1] . Harmful aggregates must be prevented , sequestered , disassembled , or degraded by cells to prevent disruption of essential cellular functions . Enhanced protein aggregation or impaired clearance of aggregates can lead to neurodegenerative disorders such as Alzheimer’s Disease , Parkinson’s Disease , Amyotrophic Lateral Sclerosis ( ALS ) , and Huntington’s Disease ( for review , see [2–9] ) . Prion diseases represent a unique sub-class of protein misfolding disorders in which protein aggregates are infectious . Prions can arise de novo through protein misfolding events that convert native proteins into the infectious form , or may be acquired through environmental encounter with the infectious form [10] . Although first described in mammals , a number of prion proteins were later found to occur in budding yeast [11 , 12] . Saccharomyces cerevisiae has been used extensively as a model organism to study prions [11 , 13] . Discovery and characterization of the first two yeast prion proteins , Ure2 and Sup35 , revealed that both proteins contain remarkably glutamine/asparagine ( Q/N ) rich prion domains [12 , 14 , 15] . The prion domains also contain relatively few charged and hydrophobic residues . Scrambling experiments demonstrated that the ability of Ure2 and Sup35 to form prions is largely dependent on the amino acid composition of the prion domains , rather than the primary amino acid sequence [16 , 17] . Methods for scanning the yeast proteome for additional proteins with similar compositional features resulted in successful identification of new yeast prions [18–20] . To date , nine yeast proteins have been demonstrated to form aggregation-mediated prions [12 , 18 , 21–27] . The majority of these proteins also contain prion domains with high Q/N content and low charged/hydrophobic content . Examination of the human proteome with more sophisticated composition-based search algorithms revealed a number of human proteins with “prion-like domains” ( PrLDs ) , defined as domains that compositionally resemble yeast prion domains [5 , 28] . Many of the top candidates ( including TDP-43 [29 , 30] , FUS [31 , 32] , EWSR1 [33 , 34] , TAF15 [34–36] , hnRNPA1 [37] , hnRNPA2B1 [37] , and TIA1 [38] ) have been implicated in protein misfolding disorders . In addition to containing PrLDs , aggregates formed by these proteins are thought to spread throughout an individual in an infectious prion-like manner along a neuroanatomical path that parallels the progression of pathological symptoms [3 , 39] . Furthermore , the PrLDs from hnRNPA1 and hnRNPA2B1 are able to support prion activity when substituted in place of the portion of the Sup35 prion domain that is responsible for nucleating prion activity [37 , 40] . Although composition-based algorithms have been reasonably effective at identifying candidate yeast prion proteins and potential disease-associated human PrLDs , these algorithms are less effective at predicting the aggregation propensity of these domains or the effects of mutations [41] . One limitation of these methods is that while they assess the frequency with which amino acids occur in prion domains , this frequency may not reflect the importance of each amino acid in prion formation . To address this knowledge gap , we previously used a quantitative mutagenesis method to score the prion propensity of each amino acid in the context of a Q/N-rich prion domain [42] . Interestingly , although the yeast prions tend to be strikingly Q/N-rich , both glutamine and asparagine were found to have neutral prion propensity scores [42 , 43] . Instead , many of the non-aromatic hydrophobic amino acids ( I , M , and V ) and the aromatic amino acids ( F , W , and Y ) were observed to have strong prion-promoting effects , implicating these amino acids as key nucleators of prion aggregation [42 , 44 , 45] . We then used these prion-propensity scores to create PAPA , a prion prediction algorithm optimized for yeast prion domains [46 , 47] . PAPA is reasonably effective at predicting the prion propensity of Q/N-rich domains , as well as the effects of mutations on prion activity [45 , 47 , 48] . However , although the composition of the human PrLDs resembles the composition of yeast prion domains , the human PrLDs tend to be less Q/N-rich and contain a higher percentage of serine and glycine ( for review , see [41] ) . Therefore , it is likely that prediction methods developed for yeast prion domains may not be optimized for human PrLDs . To understand how amino acid context ( i . e . the starting composition ) of PrLDs affects amino acid prion propensities , we sought to determine the prion propensity of each amino acid in the context of two model glycine ( G ) rich human PrLDs from hnRNPA1 and A2 . As in the context of Q/N-rich yeast prion domains , we found that aromatic amino acids were strongly prion-promoting in the context of these G-rich PrLDs . However , contrary to their effect in Q/N-rich yeast prion domains , the non-aromatic hydrophobic amino acids were not strongly prion-promoting; instead , they served as a signal for targeted degradation of the G-rich PrLDs . This suggests that aromatic amino acids may have the unique capacity to increase the aggregation propensity of prion or prion-like domains while avoiding efficient detection by protein degradation systems . Furthermore , Q/N residues strongly inhibited degradation of the G-rich PrLDs , suggesting that they may help prevent degradation of prion and prion-like domains . Indeed , many of the same sequences that led to degradation in the context of the G-rich PrLD had no effect on turnover of a Q/N-rich prion domain . These results broaden our understanding of the proteostatic regulation of aggregation-prone proteins , and shed light on the role of Q/N residues within prion domains .
The core PrLDs from the human RNA-binding proteins hnRNPA1 and hnRNPA2 were chosen as model substrates to examine the sequence requirements for aggregation within G-rich PrLDs . Both proteins contain a C-terminal G-rich PrLD . Mutations in these domains cause ALS and multisystem proteinopathy in humans , increase their aggregation propensity in vitro , and cause muscle degeneration when the proteins are expressed in Drosophila [37 , 48 , 49] . We previously used a yeast prion system to examine the effect of mutations on the aggregation propensity of the hnRNPA1 and A2 PrLDs [37] . The yeast prion protein Sup35 contains three functionally distinct domains: an N-terminal prion domain that is necessary and sufficient for formation of prion aggregates; a C-terminal functional domain , which is involved in translation termination; and a highly charged middle domain [15 , 50 , 51] . The first 40 amino acids of the prion domain , referred to as the nucleation domain ( ND ) , are very Q/N-rich and are responsible for nucleating prion aggregates [40] . We therefore replaced the Sup35 ND with the core PrLD from hnRNPA1 and hnRNPA2 to test whether these PrLDs could support prion activity [37] . These fusion proteins allowed us to use well-established Sup35 prion detection assays to probe the relationship between amino acid sequence and aggregation activity for the hnRNPA1 and hnRNPA2 PrLDs . Formation of [PSI+] , the prion form of Sup35 , can be assayed by monitoring nonsense suppression of the ade2-1 allele in the presence of tRNA suppressor SUP16 [52] . ade2-1 mutants are unable to grow on medium lacking adenine ( SC-ade ) , and grow red on medium containing limited adenine ( YPD ) due to accumulation of a pigment derived from the substrate of the Ade2 enzyme; [PSI+] formation results in a low level of read-through of the ade2-1 premature stop codon , allowing for growth on SC-ade , and formation of white colonies on YPD . The fusion proteins showed a number of hallmarks of mutation-dependent prion activity [37 , 48] , including; 1 ) spontaneous formation of ADE+ colonies , and an increase in ADE+ colony formation upon PrLD overexpression; 2 ) curability of the ADE+ phenotype by 4mM GuHCl , a treatment that cures [PSI+] [53]; 3 ) transmission of the phenotype by cytoduction; and 4 ) the formation of microscopically-visible foci in ADE+ cells and the absence of foci in ade- cells . Furthermore , the in vitro amyloid propensity , the formation of visible foci , and the frequency of appearance of the ADE+ phenotype could be influenced in a predictable manner with rationally-designed mutations derived from an established prion propensity scale . These studies provide strong evidence that the fusion proteins form prions . However , some Sup35 mutants with modified prion domains can show similar nonsense suppression that is not due to Sup35 prion formation [54] . Therefore , we took additional steps to confirm the prion activity of the fusion proteins . Prion maintenance requires continuous expression of the prion protein . To provide additional evidence that the hnRNP-Sup35 fusions form canonical aggregation-mediated prions in yeast , we induced prion formation by overexpressing the A2 D290V PrLD in cells expressing hnRNPA2-Sup35 ( D290V ) as the sole copy of Sup35 [37] . Two independent [PRION+] colonies were transformed with a Sup35 plasmid lacking the prion domain ( Sup35MC ) and passaged in the absence of selection for the A2-Sup35 plasmid . In both cases , the prions formed by the A2-Sup35 fusions were cured upon loss of the A2-Sup35 expressing vector , and only rarely spontaneously reappeared upon re-introduction of the vector , indicating that the prion phenotype required expression of the A2 PrLD to be maintained ( Fig 1A ) . Since the hnRNP-Sup35 fusions contain a portion of the native Sup35 prion domain , we also examined whether the prions formed by the A2 PrLD were transferable to wild-type Sup35 , which could suggest that the remainder of the Sup35 prion domain ( rather than the A2 PrLD ) was predominantly responsible for prion activity . Co-expression of wild-type Sup35 with A2 D290V suppressed the prion phenotype . After passaging these cells in the absence of selection for the A2-Sup35 plasmid , 23 out of 24 isolates that had not maintained the A2-Sup35 plasmid showed a [prion-] phenotype ( Fig 1B ) , while the final isolate exhibited an atypical yellow phenotype with small colonies that resembled neither a [PRION+] or a [prion-] phenotype ( potentially indicating contamination ) . This suggests that prions formed by the A2-Sup35 fusions were not sufficient to structurally convert Sup35 to a prion state . Collectively , these results indicate that the hnRNP-Sup35 fusions form canonical , PrLD-dependent prions . We previously developed a method to quantitatively score the effects of mutations on Sup35 prion activity [42 , 44] . We replaced 8-amino acid segments of the prion domain with a random sequence , generating libraries of mutants . Each mutant was expressed as the sole copy of Sup35 in the cell . Randomly mutagenized libraries were plated onto medium lacking adenine to select for mutants that maintained the ability to form [PSI+] . This method was applied to various regions of wild-type and scrambled Sup35 , including the Sup35 nucleation domain [42 , 44] . Therefore , to examine how the sequence requirements for aggregation differ between Q/N-rich and G-rich PrLDs , we repeated this method , mutating the hnRNPA1-Sup35 and hnRNPA2-Sup35 fusions ( herein referred to as A1-Sup35 and A2-Sup35 respectively; Fig 2A ) . As targets for mutagenesis , we selected segments with a mixture of predicted aggregation-promoting , aggregation-inhibiting , and neutral amino acids , near the site corresponding to a region previously mutagenized in Sup35 ( Fig 2B ) . Spontaneous [PSI+] formation is typically a stochastic and very rare event , occurring at a rate of less than 10−6 per generation [55] . By contrast , mutations that reduce Sup35 activity without causing prion aggregation will result in a constitutive ADE+ phenotype . Thus , to detect rare prion formation events from among a library of mutants , it is necessary to first eliminate mutants that have a constitutive ADE+ phenotype ( Fig 2A ) . In previous screens with wild-type or scrambled Sup35 , such constitutive ADE+ mutants were relatively rare , comprising ~5% of screened isolates [42 , 44] . Unexpectedly , for the mutagenized A2-Sup35 and A1-Sup35 fusions , approximately 30–40% of the isolates were able to grow in the absence of adenine . These ADE+ isolates were not cured by treatment with 4mM GuHCl , suggesting that the growth on SC-ade resulted from non-prion-based inactivation of the hnRNP-Sup35 fusion proteins . As observed for [PRION+] isolates , replacement of the A2-SUP35 plasmid with a plasmid expressing Sup35MC results in loss of the ADE+ phenotype . However , in contrast to [PRION+] strains , when plasmids expressing A1- or A2-Sup35 mutants were isolated from representative strains with a constitutive ADE+ phenotype and shuffled back into the parent strain , the ADE+ phenotype spontaneously re-appeared ( S1 Fig ) . This indicates that the phenotype results from loss of activity of the A2-Sup35 fusion protein , not from mutations in other cellular proteins or from classical PrLD-dependent prion propagation . Finally , while prion formation was associated with increased levels of insoluble A2-Sup35 protein , the constitutive ADE+ mutants did not contain substantial amounts of insoluble A2-Sup35 protein ( S2 Fig ) . Therefore , we sought to determine the basis of Sup35 inactivation among these isolates . We sequenced the mutagenized region of the A1/A2-SUP35 gene from randomly selected ade- and ADE+ isolates to determine whether specific sequence features were correlated with the ADE+ phenotype . For each amino acid , an odds ratio was calculated ( Eq 1 ) , representing the degree of over- or under-representation of the amino acid among ADE+ isolates ( Table 1 ) . For both libraries , each of the non-aromatic hydrophobic amino acids ( I , L , M , and V ) were over-represented among ADE+ isolates , while glutamine , asparagine , and each of the charged amino acids ( D , E , K , and R ) were under-represented ( Table 1; Fig 3 ) . Individually , not all of these biases reached the standard threshold of statistical significance ( p < 0 . 05; Table 1 ) . Grouping amino acids of similar physical properties can increase statistical significance by effectively increasing sample sizes . When considered as a group , the biases for hydrophobic amino acids , against charged amino acids , and against Q/N were each statistically significant in both libraries ( P<0 . 01 in all cases; Table 1 ) . One possible explanation for the ADE+ phenotype is that the hnRNP-Sup35 fusions could be poorly expressed or rapidly degraded , causing a decrease in steady state levels of the fusion proteins . To test this possibility , four representative A2-Sup35 isolates that exemplified the amino acid biases among the ADE+ library were selected for comparison with randomly selected isolates from the ade- library . The ADE+ and ade- phenotypes originally observed for these isolates were confirmed by spotting onto SC-ade , YPD , and YPAD ( Fig 4A ) . Previous studies suggest that an ADE+ phenotype is observed when steady-state Sup35 levels drop below about 40% of wild-type [56] . In synthetic complete medium , all four ADE+ isolates had steady-state A2-Sup35 levels that were less than 40% of wild-type , while three of four isolates from the ade- library had steady state A2-Sup35 levels above this threshold ( S3A Fig ) . When cells were shifted to medium lacking adenine , A2-Sup35 levels dropped for all eight strains , but showed lowest levels for the four ADE+ isolates . Furthermore , as a group , steady state protein levels for ADE+ isolates were significantly lower ( p < 0 . 001 ) than the grouped protein levels for ade- isolates in both synthetic complete and adenine-deficient synthetic complete media . Exposed hydrophobic patches are known in some cases to trigger protein degradation [57 , 58] . Therefore , we hypothesized that the lower average expression levels seen among the ADE+ isolates might be due to increased degradation . Cycloheximide ( CHX ) globally inhibits translation by preventing translocation of the ribosome along mRNA , providing a convenient tool to assay protein turnover [59] . After treatment with CHX , the fusion proteins within ADE+ isolates were rapidly degraded ( Fig 4A ) . Three of the four ADE+ isolates contained little or no detectable A2-Sup35 by 2 . 5 hours after addition of CHX , while the fourth showed a substantial decrease in A2-Sup35 levels over the 5 hour timecourse ( Fig 4A ) . By contrast , A2-Sup35 levels remained relatively stable or decreased only slightly over a period of 5 hours after addition of CHX for all of the ade- isolates , as well as for the wild-type A2-Sup35 fusion ( Fig 4A ) . These results suggest that hydrophobic amino acids trigger degradation of the A2-Sup35 fusions . Interestingly , random mutagenesis of the Sup35 prion domain yielded very few isolates with the degradation phenotype in the initial screen [44] , suggesting that the Sup35 prion domain can buffer the effects of degradation-promoting peptides . Indeed , when the degradation-promoting 8-amino acid sequences from the A2-Sup35 library were substituted into the corresponding region of the Sup35 prion domain , each of the proteins resulted in phenotypically ade- cells ( Fig 4B ) , and maintained steady-state Sup35 levels well above the 40% of wild-type ( S3B Fig ) . Furthermore , none of the peptides accelerated the degradation rate of Sup35 over 5 hours ( Fig 4B ) . Therefore , while the A2 PrLD is susceptible to the degradation-promoting effects of hydrophobic amino acids , the Sup35 prion domain can mask these effects and resist degradation . The ubiquitin-proteasome system is one of the main protein recycling pathways in eukaryotic cells . MG-132 , a commonly used proteasome inhibitor , is effective in yeast lacking the pleiotropic drug resistance 5 gene ( pdr5Δ ) . To assess whether degradation of the A2-Sup35 proteins occurs via the proteasome , PDR5 was deleted from the genome , and the turnover of the A2-Sup35 proteins was assessed in the presence or absence of MG-132 . Pre-treatment with MG-132 for 1 hour prior to addition of CHX resulted in nearly complete stabilization of the degradation-prone A2-Sup35 fusions over the 5 hour timecourse ( Fig 5 ) . This result suggests that the ADE+ phenotype is due to enhanced turnover of the A2-Sup35 fusion proteins via the ubiquitin-proteasome system . Since degradation-promoting sequences failed to cause degradation of Sup35 ( Fig 4B ) , we reasoned that our previous dataset from random mutagenesis of Sup35 [44] would contain some peptide sequences that did not cause degradation in the context of the Sup35 prion domain , but would promote degradation of the A2-Sup35 fusion protein . To identify potential degradation-promoting sequences , each peptide from the library was scored by summing the log-odds ratios from Table 1 for the eight amino acids in the mutagenized region . Three sequences predicted to promote degradation ( i . e . , sequences enriched in non-aromatic hydrophobic residues , with few charged or Q/N residues ) were selected from the dataset . When substituted into A2-Sup35 , all three predicted degradation-promoting peptides led to enhanced turnover of A2-Sup35 and characteristic degradation phenotypes , albeit to varying degrees ( Fig 6A ) . All three strains appeared light pink on YPD , and growth on SC-ade correlated qualitatively with the degree of degradation conferred by each peptide . Additionally , two sequences predicted to have no effect on A2-Sup35 turnover ( i . e . , sequences enriched in charged and polar residues ) were chosen from the same dataset as controls . When substituted into A2-Sup35 , neither peptide enhanced degradation , and both strains displayed the associated ade- phenotypes ( Fig 6A ) . By contrast , four of the five peptides substituted into the Sup35 prion domain had little effect on turnover and resulted in the characteristic ade- phenotype ( Fig 6B ) , while the fifth showed modest degradation and only a weak ADE+ phenotype . These results demonstrate that the compositional biases originally observed in the ADE+ libraries are sufficient to predictively categorize sequences as degradation-promoting or degradation-inhibiting . The sequences obtained through random mutagenesis are heterogeneous with respect to composition and sequence . To more rigorously define the minimum number of non-aromatic hydrophobic residues required to accelerate the rate of degradation or prion formation , hydrophobic content was progressively increased in WT A2-Sup35 and WT Sup35 . Valine , leucine , and methionine ( the hydrophobic residues most over-represented in the A2-Sup35 ADE+ library ) were inserted in an alternating fashion adjacent to the region targeted for random mutagenesis ( Figs 2B and 7 ) . As few as two hydrophobic residues were sufficient to slightly increase turnover of A2-Sup35 , as indicated by western blot and the characteristic ADE+ phenotype ( Fig 7A ) . Three hydrophobic residues further accelerate A2-Sup35 degradation , and four to seven hydrophobic residues caused almost complete loss of A2-Sup35 by 2 . 5 hours after the addition of CHX . Two or fewer hydrophobic residues inserted into Sup35 resulted in uniform ade- phenotypes , whereas three or more hydrophobic residues resulted in the appearance of white sectors , which are classical indications of prion formation ( Fig 7B ) . Strikingly , the degree of sectoring increased in a dose-dependent fashion as hydrophobic content increased . Elimination of the [PIN+] prion did not affect the degradation of A2-Sup35 or stability of Sup35 upon insertion of hydrophobic residues ( S4 Fig ) . To more accurately quantify the frequency of ADE+ colony formation by each mutant , serial dilution of each mutant was plated on SC-ade , starting from a higher density than originally assayed . Fewer than two hydrophobic residues in A2-Sup35 resulted in minor growth only at high cell density , whereas two or more hydrophobic residues resulted in robust growth even at very low cell density ( Fig 7C ) . Treatment with GuHCl did not alter the color phenotype on YPD ( Fig 7D ) , suggesting that the ADE+ growth was not due to prion formation . By contrast , three or more hydrophobic residues in Sup35 resulted in a progressive increase in the frequency of ADE+ colonies , consistent with the progressive increase in sectoring observed on YPD for these mutants ( Fig 7E ) . Treating the cells with GuHCl reverted the ADE+ phenotype to an ade- phenotype ( Fig 7F ) , confirming that growth on SC-ade was due to the formation of bona fide prions . These results were not unique to these specific positions within Sup35 and A2-Sup35 . We made additional hydrophobic insertions one-quarter and three-quarters of the way through the Sup35 ND and the hnRNPA2 PrLD ( positions 10 and 30 for Sup35; positions 11 and 33 for A2; Fig 2B ) . As with the original hydrophobic insertions ( Fig 7A ) , insertions at both additional positions in the A2 PrLD resulted in increased degradation , although the effects of insertion were weaker at position 33 ( Fig 7G ) . Likewise , Sup35 was far more resistant to the degradation-promoting effects of hydrophobic amino acids at both positions , although modest degradation was observed when six hydrophobic amino acids were inserted at position 10 ( Fig 7H ) . It is possible that physical interactions between the Sup35ND and the remainder of the Sup35 sequence or with native Sup35 binding partners are responsible for the apparent stability of the Sup35 ND . However , insertion of hydrophobic residues in the A2 PrLD alone fused to GFP resulted in a progressive increase in degradation rate ( Fig 8A , top ) , whereas insertion of hydrophobic residues in the Sup35 ND had no effect on degradation ( Fig 8B , top ) . Nearly identical trends were observed for FLAG-tagged version of the A2-Sup35 and Sup35 NM domains ( Fig 8 , bottom ) . Collectively , these results demonstrate that the Sup35 ND can mask the degradation-promoting effects of hydrophobic residues , and that this effect is not dependent on the remainder of the protein . The features promoting degradation of the A2 PrLD are consistent with previous studies indicating that the degradation machinery recognizes exposed hydrophobic segments , and that there is a strong correlation between the sequence features that promote aggregation and degradation [57 , 58] . We were interested in whether this correlation is absolute , or whether there are sequence features that can promote aggregation of the G-rich PrLDs without promoting degradation . Our A1- and A2-Sup35 fusions provide a useful system for comparing the sequence requirements for degradation versus aggregation . To determine whether specific sequence features could promote prion aggregation without triggering degradation , isolates with an initial ade- phenotype were plated onto medium lacking adenine to screen for the ability to spontaneously form prions ( Fig 2A ) . [PRION+] isolates were confirmed by curing with GuHCl , and the mutagenized A1/A2-SUP35 gene in each was sequenced . Sequences from each library were pooled , and the prion propensity scores for each amino acid were determined , as described previously ( [42 , 44]; Eq 4 ) . Interestingly , while both non-aromatic and aromatic hydrophobic residues were strongly prion-promoting within the Q/N-rich Sup35 ND [42 , 44] , only aromatic amino acids were significantly over-represented among [PRION+] isolates for the A2-Sup35 and A1-Sup35 libraries; non-aromatic hydrophobic residues were approximately equally represented among [PRION+] and ade- isolates ( Table 2 ) . Furthermore , Q/N residues were significantly under-represented among A2-Sup35 [PRION+] isolates , although their effects were mixed among A1-Sup35 [PRION+] isolates . Together , these results suggest that a hitherto unappreciated property of aromatic amino acids is the unique ability to promote protein aggregation of prion and prion-like domains , while avoiding detection by the degradation machinery . Indeed , while there is a statistically significant ( P = 0 . 008 by Spearman rank analysis ) correlation between the prion propensity ( as scored by PAPA ) of each amino acid and its propensity to promote degradation ( Fig 9 ) , there are five amino acids which have substantially lower degradation propensities than would be predicted by their prion propensities: the three aromatic amino acids , glutamine , and asparagine . Strikingly , these amino acids are all overrepresented among yeast prion proteins . While both aromatic and non-aromatic hydrophobic amino acids strongly promote prion formation [42] , candidate prion domains with prion activity tend to contain more aromatic residues and fewer aliphatic residues than candidate prion domains with no detectable prion activity [44] . Likewise , although serine , glycine , threonine , glutamine , and asparagine each promote intrinsic disorder and have similar prion propensities [42] , Q/N residues are far more common among yeast prion domains . Collectively , these results suggest a possible explanation for the amino acid biases observed among yeast prion domains . Many components of protein quality control systems act specifically to antagonize protein aggregation . Therefore , proteins that form observable protein aggregates must possess mechanisms to avoid or outcompete antagonistic proteostasis machinery . Yeast prion domains tend to favor amino acids that promote aggregation while being poorly recognized by the degradation machinery . These results may also provide an explanation for Sup35’s resistance to degradation . Q/N residues were among the lowest scoring amino acids in the degradation libraries . The human PrLDs and the Sup35 ND differ most notably in their Q/N content; the Sup35 ND contains a much higher percentage of Q/N-residues , while the A1 and A2 core PrLDs are more G-rich . This suggests the simple hypothesis that the high Q/N-content of the Sup35 ND may protect highly aggregation-prone features from recognition by components of the proteostasis machinery . To test this hypothesis , two of the degradation-prone members of the A2 library and their Sup35 counterparts were chosen as initial substrates for mutagenesis . To examine the relationship between Q/N content and degradation , we mutated some or all of the Q/N’s in the Sup35 nucleation domain to G’s ( Fig 10A ) . Similarly , we mutated some or all of the G’s in the A2 PrLDs to Q/N . The rate of degradation of Sup35 correlated with Q/N-content in a dose-dependent manner . Partial substitution of Q/N-residues for G’s significantly increased the turnover rate of each Sup35 derivative and resulted in the emergence of the ADE+ phenotype ( Fig 10B; S5 Fig ) . Substitution of the remaining Q/N’s for G’s further enhanced the rate of Sup35 degradation . Partial or full substitution of G’s for Q/N’s in the A2 PrLD resulted in a modest , albeit statistically significant increase in stability for one of the two mutagenized PrLDs . However , no stabilizing effect was observed for the second mutagenized PrLD , suggesting that other sequence features of the A2 PrLD besides Q/N content must contribute to its sensitivity to degradation . Therefore , in addition to their role in prion formation , Q/N residues help the Sup35 prion domain resist degradation by intracellular anti-aggregation systems .
Protein misfolding is a selective challenge faced by all cellular life . Misfolded proteins can result in proteotoxicity , either through loss-of-function of the native protein or through a toxic gain-of-function of the misfolded species . To address these selective challenges , eukaryotic cells possess extensive proteostasis machinery , which constitutively act to procure and maintain pools of natively folded proteins . The proteostasis machinery broadly consists of three main systems: 1 ) the protein chaperone network , which aids in nascent protein folding as well as the re-folding of partially or fully denatured proteins , 2 ) the ubiquitin-proteasome system , and 3 ) the autophagy system , which together aid in the destruction of aged , terminally misfolded , or aggregated proteins ( for review , see [60 , 61] ) . Despite the constant surveillance of protein quality control systems , numerous diseases result from misfolding and aggregation of proteins . Additionally , a variety of proteins form functional aggregates that are involved in the regulation of various cellular processes [62–64] . Therefore , understanding how the proteostasis machinery detects misfolded proteins , and how some aggregation-prone proteins evade this detection , may provide insight into both functional and pathogenic aggregation . One way through which proteostasis network components achieve specificity for misfolded proteins is by recognizing patches of solvent-exposed hydrophobicity [65–73] . Hydrophobic patches are generally buried in the interior of folded proteins [74] , so exposed hydrophobicity can act as a signal of protein misfolding . Additionally , there is a strong correlation between hydrophobicity and aggregation propensity [75] , so recognizing exposed hydrophobicity would seem to be an effective mechanism to recognize aggregation-prone misfolded proteins . One well-characterized example that uses this mechanism is the yeast E3 ubiquitin ligase San1 , a nuclear protein involved in the ubiquitin-proteasome degradation system [76] . San1 is a largely disordered protein that is particularly adept at targeting toxic misfolded proteins for degradation [77] , primarily by recognizing exposed hydrophobic residues in substrates [58] . Interestingly , San1 recognition of these substrates tends to correlate with their insolubility , demonstrating the effectiveness of targeting hydrophobicity to prevent protein aggregation [57] . It should be noted that degradation of the A1- and A2-Sup35 fusions was independent of San1 ( S6 Fig ) , so additional work will be required to identify the cellular factors responsible for recognition and degradation of these PrLDs . Identifying these factors and studying the mechanisms by which they recognize aggregation-prone proteins may help explain mechanistically how aromatic amino acids can promote aggregation without triggering PQC degradation . Although our data is generally consistent with the idea that exposed hydrophobic amino acids promote recognition by the proteostasis machinery , our results provide some additional unexpected insights . First , in contrast to what has been proposed for San1 , we show that aggregation propensity and recognition by the proteostasis machinery can be uncoupled in a composition-dependent manner: aromatic amino acids within the G-rich hnRNP PrLDs increase aggregation propensity without substantially enhancing recognition by the proteostasis machinery . Second , the ability of the proteostasis machinery to recognize hydrophobic patches was highly context dependent: the Q/N-rich Sup35 ND had the inherent capacity to mask otherwise degradation-promoting amino acids . These results highlight an important point related to the proteostasis of prion and prion-like domains . While it is sometimes useful to broadly categorize certain amino acids as “aggregation-promoting” or “degradation-promoting” , the effects of these amino acids may vary from protein to protein depending on the larger sequence context within which they are found , and on the interactions between these domains and cellular proteostasis factors . Since both the short sequence features and the surrounding context play such important roles in aggregation and degradation , future examination of the extent to which these heuristics apply to other aggregation-prone proteins and in other organisms would be interesting . While PolyQ regions reportedly resist degradation by the proteasome [78] ( although this too remains quite controversial [79 , 80]; for review , see also [81 , 82] ) , Sup35 has a roughly average half-life in vivo [83] suggesting that it is not inherently unusually resistant to degradation . Some evidence indicates that certain fragments of the Q/N-rich Sup35 prion domain exhibit a high rate of turnover [84] , and the Sup35 prion domain can be proteolytically cleaved [85] , indicating that the degradation and proteolytic systems are not incapable of processing the Sup35 prion domain in vivo . However , our results illuminate a principle fundamentally distinct from inherent stability–namely , sequences capable of potently inducing degradation in the G-rich PrLDs are , in some way , protected from the proteostasis machinery by surrounding Q/N-rich regions . Therefore , Q/N residues may potentiate the aggregation of prion domains , in part , by protecting aggregation-prone features from the proteostasis machinery . Although increased degradation and increased aggregation are not necessarily alternatives [84 , 86] , our results suggest that the composition of the Sup35 prion domain allows it to resist degradation , while maintaining the ability to form prions . While the exact mechanism by which Sup35 resists degradation is unclear , high Q/N-content appears to play an important role . The Sup35 prion domain and the A1/A2 PrLDs are each predicted to be intrinsically disordered . However , the Sup35 prion domain is thought to form a collapsed but disordered structure [87] , which may hide hydrophobic patches from the proteostasis machinery . High Q/N content may help mask hydrophobic patches by promoting a collapsed but disordered structure , or by shielding hydrophobic amino acids within these structures . Alternatively , rather than preventing the initial recognition of hydrophobic patches by the proteostasis machinery , Q/N residues may inhibit a downstream step in the subsequent events leading to degradation . Interestingly , the Q/N content of Sup35 is relatively well-conserved across independent Saccharomyces cerevisiae strains and between different yeast species [88 , 89] . Although high Q/N content within the N-domain may be maintained by selection for multiple reasons , it is possible that the stabilizing effects of Q/N at least contribute to the observed compositional conservation . Other features of the Sup35 prion domain besides Q/N content also seem well-suited to avoid detection by the degradation machinery , potentially explaining why increasing the Q/N content of the A2 PrLD was not sufficient to fully stabilize the PrLD . We previously showed that six amino acids are highly prion-promoting: F , Y , W , I , V , and M [42] . The Sup35 prion domain contains 23 of these highly prion-promoting amino acids , yet all except the initiating methionine are aromatic . Additionally , the prion-promoting amino acids are well-dispersed . There is only one position where two occur adjacent to each other , and almost all have adjacent Q/N residues . Thus , the Sup35 prion domain possesses many features that promote aggregation , yet avoids multiple features that can contribute to degradation . Furthermore , these biases are not unique to the Sup35 prion domain; most other yeast prion domains are also Q/N-rich , and tend to favor aromatic amino acids over non-aromatic prion-promoting amino acids [44] . Numerous labs have made extensive progress in defining how the amino acid sequence of a protein affects its intrinsic aggregation propensity . However , our results highlight that intrinsic aggregation propensity is only a small piece of the puzzle . A more complete understanding of functional and pathogenic protein aggregation requires a clearer view of how amino acid sequence affects interactions with other cellular proteins . Our results provide one unexpected piece to this puzzle , demonstrating that specific sequence features can promote protein aggregation , while simultaneously hiding from the proteostasis machinery .
Standard yeast media and methods were used as previously described [90] , except that YPD plates contained 0 . 5% yeast extract rather than the standard 1% . YPAD for all experiments contained the standard 1% yeast extract , as well as 0 . 02% adenine hemisulfate . Prion curing assays were performed for individual ADE+ isolates by streaking onto YPD with and without 4mM GuHCl , then re-streaking to YPD to test for loss of the ADE+ phenotype . In all experiments , yeast were grown at 30°C . The yeast strains used in this study were YER826/pER589 ( α kar1-1 SUQ5 ade2-1 his3 leu2 trp1 ura3 sup35::KanMx ) and YER1161 ( α kar1-1 SUQ5 ade2-1 his3 leu2 trp1 ura3 sup35::KanMx pdr5::HIS3 ) . pER589 expresses a truncated version of Sup35 lacking the prion domain ( Sup35MC ) as the sole copy of Sup35 in the cell . This plasmid was subsequently replaced by plasmid shuffling in order to assay activity of the full-length , randomly mutagenized hnRNP-Sup35 fusions . The A1-Sup35 and A2-Sup35 fusion libraries were generated in a manner similar to MacLea et al . [44] . Briefly , the N-terminal end and C-terminal end of each gene were amplified from a plasmid containing either the A1-Sup35 fusion or the A2-Sup35 fusion ( pER595 for hnRNPA1 and pER697 for hnRNPA2; [37] ) . Oligonucleotides ( from Integrated DNA Technologies ) were used to re-amplify the respective products and incorporate a 24-nucleotide degenerate region in which each of the four nucleotides has a 25% probability of occurring at the first two positions of each codon , while C , G , and T each have a 33% probability of occurring at the final position of each codon . The N-terminal and mutagenized C-terminal products , which contain complementary segments , were mixed and re-amplified . The final PCR products were co-transformed with BamHI/HindIII-cut pJ526 into YER826 and plated on synthetic complete media lacking leucine ( SC-Leu ) to select for cells containing a recombined plasmid . Individual colonies were then picked and stamped onto media containing 5-Fluoroorotic acid ( 5-FOA ) to select for loss of pER589 . After 5-FOA treatment , cells were transferred to YPAD , YPD , and SC-ade . After three days at 30°C , isolates for which more than 5 colonies appeared on SC-ade were identified and placed in a category ( ADE+ library ) separate from those with fewer than 5 colonies ( ade- ) . Randomly selected representative isolates from both groups were sequenced to generate each library . The odds ratio for the ADE+ phenotype ( ORA ) for each amino acid was determined as follows: ORA=[fD1-fD]/[fN1-fN] ( 1 ) where fD represents the per residue frequency of the amino acid among the isolates that were able to grow on SC-ade , and fN represents the per residue frequency of the amino acid among the naïve isolates ( i . e . , those that were unable to grow on SC-ade ) . Final degradation propensity scores for each amino acid ( DPaa ) were determined as follows: DPaa=ln ( ORA ) ( 2 ) In addition , prion isolates were identified as previously described [42 , 44] ( Fig 1 ) and sequenced to generate the prion library . Briefly , the isolates that were initially unable to grow on SC-ade were pooled from the solid YPAD media and re-plated on SC-ade at approximately 106 and 105 cells per plate . After 3–5 days at 30°C , individual colonies were streaked onto YPD and YPD plus 4mM GuHCl to assay for prion loss . Odds ratios for prion activity ( ORP ) for each amino acid were determined as follows: ORP=[fP1-fP]/[fN1-fN] ( 3 ) where fP represents the per residue frequency of the amino acid among the prion-forming isolates . Final prion propensity scores ( PPaa ) were determined as follows: PPaa=ln ( ORP ) ( 4 ) Cells were diluted to an optical density of 0 . 75 in liquid YPAD media and incubated with shaking at 30°C for 1hr before treatment with CHX , or DMSO for untreated cells . Where applicable , MG-132 was added to a final concentration of 10μg/mL 1 hr prior to addition of CHX . After the treatment period , the optical densities of all cultures were measured . 10mL of the least-dense culture for each strain were harvested . Based on the optical densities , the approximate number of cells harvested for each of the remaining cultures was normalized to the least-dense culture within each unique strain . Cells were pelleted by centrifugation at 3 , 000rpm for 5 minutes at 4°C . Cell pellets were lysed as previously described [57] . 30μL of prepared lysate were loaded onto a 12% polyacrylamide gel , transferred to a PVDF membrane , and probed with an appropriate antibody . Primary antibodies ( all monoclonal ) used in this study were: an anti-Sup35C ( BE4 [91] , kindly made available by Susan Liebman ) , an anti-GFP antibody ( Santa Cruz Biotechnology ) , and an anti-FLAG antibody ( Sigma ) . Blots were quantified using Image Studio Version 5 . 2 . Background-subtracted intensities for all quantified blots can be found in S1 Table . As with the degradation assays , yeast cultures were diluted to an optical density of 0 . 75 in liquid YPAD media and incubated with shaking at 30°C for 1hr before normalizing to the least-dense culture and harvesting . Cell pellets were re-suspended in 200μL of chilled non-denaturing lysis buffer ( 100mM TrisHCl , pH7 . 5 , 200mM NaCl , 1mM EDTA , 5% Glycerol , 0 . 1% Triton-X 100 , and Bond-Breaker TCEP solution ( Thermo Fisher Scientific ) and ProBlock Gold yeast protease inhibitor cocktail ( Gold Biotechnology ) to manufacturer recommendations; adapted from [92] ) , transferred to a round-bottom 2mL tube , and vortexed with a single large glass bead ( ~3mm diameter ) on maximum speed for 10 minutes . Lysates were centrifuged gently ( 700 x g for 5 minutes at 4°C ) to pellet unlysed cells and large cellular debris . 50μL of total lysate sample was mixed with 50μL of denaturing buffer ( 1% SDS , 8M urea , 10mM MOPS pH6 . 8 , 10mM EDTA pH8 . 0 , 0 . 01% bromophenol blue , and ProBlock Gold yeast protease inhibitor cocktail ( Gold Biotechnology ) to manufacturer recommendations ) . 100μL of remaining lysate was centrifuged at 16 . 3k rpm for 15 minutes at 4°C to pellet protein aggregates . The supernatant was removed and mixed 1:1 with denaturing buffer ( soluble sample ) . The remaining pellet was resuspended in an equal volume of a 1:1 mixture of non-denaturing:denaturing buffer . Samples were boiled for 5 minutes , then centrifuged at 12 , 900 x g for 5 minutes before loading . Original strains were transformed with a covering plasmid expressing a version of Sup35 lacking the prion domain ( Sup35MC ) and a URA3 selectable marker . Transformants were passaged on SC-Ura until loss of the original plasmid expressing the A2-Sup35 fusion . After loss of the plasmid , strains were re-transformed with the original A2-Sup35 fusion plasmid and the URA3 covering plasmid counter-selected on 5-fluoroorotic acid ( FOA ) . Color phenotypes for each strain were compared by streaking onto YPD . | Protein aggregation is associated with a variety of diseases , including Alzheimer’s disease and Amyotrophic Lateral Sclerosis . Cells possess a number of factors that can recognize aggregation-prone protein features and prevent aggregation . One common way this is achieved is through the pre-emptive degradation of aggregation-prone proteins . While considerable progress has been made in understanding how the amino acid sequence of a protein relates to intrinsic aggregation propensity , little is known about how aggregation-prone proteins avoid intracellular anti-aggregation systems . We used a genetic screen in yeast to define sequence features of aggregation-prone domains that lead to degradation or prion aggregation as it occurs in the context of eukaryotic protein quality control factors . Unexpectedly , we found that only a subset of aggregation-promoting amino acids could effectively stimulate degradation of an aggregation-prone domain . Furthermore , this degradation-promoting effect could be suppressed by classical prion domain features . Our results highlight the complex interplay between pre-emptive protein degradation and protein aggregation , and implicate the unusual composition of yeast prion domains in preventing their degradation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"sequencing",
"techniques",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"prions",
"organic",
"compounds",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"amino",
"acids",
"molecular",
"biology",
"techniques",
"research",
"and... | 2018 | Sequence features governing aggregation or degradation of prion-like proteins |
Oleaginous microalgae are promising feedstock for biofuels , yet the genetic diversity , origin and evolution of oleaginous traits remain largely unknown . Here we present a detailed phylogenomic analysis of five oleaginous Nannochloropsis species ( a total of six strains ) and one time-series transcriptome dataset for triacylglycerol ( TAG ) synthesis on one representative strain . Despite small genome sizes , high coding potential and relative paucity of mobile elements , the genomes feature small cores of ca . 2 , 700 protein-coding genes and a large pan-genome of >38 , 000 genes . The six genomes share key oleaginous traits , such as the enrichment of selected lipid biosynthesis genes and certain glycoside hydrolase genes that potentially shift carbon flux from chrysolaminaran to TAG synthesis . The eleven type II diacylglycerol acyltransferase genes ( DGAT-2 ) in every strain , each expressed during TAG synthesis , likely originated from three ancient genomes , including the secondary endosymbiosis host and the engulfed green and red algae . Horizontal gene transfers were inferred in most lipid synthesis nodes with expanded gene doses and many glycoside hydrolase genes . Thus multiple genome pooling and horizontal genetic exchange , together with selective inheritance of lipid synthesis genes and species-specific gene loss , have led to the enormous genetic apparatus for oleaginousness and the wide genomic divergence among present-day Nannochloropsis . These findings have important implications in the screening and genetic engineering of microalgae for biofuels .
Microalgae represent a promising source of biomass feedstock for fuels and chemicals because many species possess the ability to grow rapidly and synthesize large amounts of storage neutral lipids in a form of triacylglycerol ( TAG ) from sunlight and carbon dioxide . They can be cultivated on non-arable land with non-potable water and waste streams ( e . g . , flue gases and wastewaters ) and thus pose little competition to food crops while providing environmental benefits [1] . However , understanding of the divergence and evolution of oleaginous traits and the underlying evolutionary forces and molecular mechanisms in microalgae remains elusive [2] . Nannochloropsis is a genus of unicellular photosynthetic microalgae in the class Eustigmatophyceae , ranging in size from 2–5 µm and widely distributed in marine , fresh and brackish waters . They are of interest as a potential feedstock for fuels and high-value products because they tolerate broad enivronmental and culture conditions while growing rapidly and producing large amounts of TAG and eicosapentaenoic acid , a high-value polyunsaturated fatty acid [3] . A homologous recombination–based gene transformation system was recently established in Nannochloropsis [4] , making trait improvement in this organism possible for overproduction of biomass or desirable products . Here we present a comparative analysis of six genomes of oleaginous Nannochloropsis spp . that includes two N . oceanica strains ( IMET1 and CCMP531 ) and one strain from each of four other recognized species: N . salina ( CCMP537 ) , N . gaditana ( CCMP526 , which was previously reported [5] ) , N . oculata ( CCMP525 ) and N . granulata ( CCMP529 ) ( Figure 1A; Figure S1; Figure S2; Table S1A , S1B ) . Moreover , for N . oceanica IMET1 , the diversity of transcripts was mapped to support gene prediction by sequencing cDNA libraries using 454-based long reads . Furthermore , transcript dynamics were measured via a two-condition ( control condition and nitrogen starved condition ) , three time-point temporal series of transcriptomes during TAG accumulation using Illumina-based short-reads ( Text S1 ) . Integration of phenotypic , genomic and transcriptomic data across a Nannochloropsis phylogeny provided new insights into the molecular mechanisms driving the diversity and evolution of these oleaginous microalgae .
The genome sizes of the six oleaginous Nannochloropsis species and strains range from 25 . 38 to 32 . 07 Mb ( Figure 1A; Table 1 ) . For strain IMET1 , the nuclear , chloroplast and mitochondria genomes are 31 . 36 Mb , 117 . 5 Kb and 38 Kb , respectively , totaling 31 . 5 Mb . Pulse-field gel electrophoresis on total IMET1 DNA confirmed the genome size and indicated the presence of 22 chromosomes ( Figure S3A , S3B ) . For IMET1 , 9 , 754 , 126 and 35 protein-coding genes were predicted in the nuclear , chloroplast and mitochondrial genomes , respectively ( Table 1 ) . Among the nuclear genes , 93 . 4% ( 9 , 111 ) were covered by mRNA-Seq data ( defined as >80% of the transcribed region mapped by at least 10 reads; Table S1C , S1D , Text S1 ) . These Nannochloropsis genomes are all relatively compact ( Table S2; [5] , [6] ) , much smaller than that of the model green microalga Chlamydomonas reinhardtii ( 121 Mb; [7] ) . The IMET1 genome features a higher coding potential ( 52 . 1% ) than the diatom Thalassiosira pseudonana ( 32 . 7%; [8] ) , which has a similar genome size . Mobile elements can be prevalent in algae [e . g . T . pseudonana harbors 238 long terminal repeats ( LTRs ) totaling 1 . 56 Mb] , but they are rather limited in IMET1 , as only 26 LTRs ( 24 . 3 Kb in total ) , along with several DNA transposons ( 864 bp in total ) , are present in the genome without transposases ( Table S2 ) . The relative paucity of mobile elements appears to be one shared feature of the six Nannochloropsis strains ( Table 1 ) Genomic diversity and divergence defining microalgal genera , species or strains are largely unknown [9] . A whole-genome phylogeny of Nannochloropsis ( Figure 1A ) was constructed from 1 , 085 single-copy-orthologous groups identified from the six genomes , which is consistent with the 18S-based phylogeny ( Figure S2 ) . Among the five Nannochloropsis species , N . granulata and N . oculata have a recent common ancestor and are clustered with the two N . oceanica strains . Among the 1 , 085 single-copy orthologous groups , 628 ( 61 . 7% ) exhibited congruent phylogenies with the whole-genome phylogeny . The mean Ka/Ks of 0 . 08 calculated from these candidate phylogenetic markers in the nuclear genomes was higher than in the chloroplast genomes ( 0 . 031 ) and in the mitochondrial genomes ( 0 . 064 ) . Among these candidate markers , 25 genes exhibited sequence variations large enough to differentiate each of the species and strains ( density of inter-species SNP at 20–40% and intra-species over 1% ) , but allowed for the design of consensus flanking PCR primers ( Dataset S1 ) . Those with the highest resolution included cytochrome P450 , btaA , plastid ribosomal protein S1 and transaldolase etc . , which represent novel phylogenetic markers that are more sensitive than 18S or ITS sequences ( 0 . 16% and 0 . 52% in intra-species SNP density , respectively ) in strain-typing of Nannochloropsis . Between any two genomes among the six Nannochloropsis strains , 35% of protein-coding genes ( ranging from 2 . 6% between the two N . oceanica strains to 66 . 4% between IMET1 and N . salina CCMP537 ) were not found in the other genome on average , despite >98% similarity in full-length 18S rDNA . This places their inter-species genome divergence higher than the green algae studied and their intra-species divergence comparable to E . coli and yeast ( Figure 1B ) . Therefore , the Nannochloropsis pan-genome , as defined by the six strains , consists of at least 38 , 000 protein-coding genes , along with a relatively small pool of Nannochloropsis core genes ( e . g . , 2 , 734 genes in IMET1 ) that are shared by the six strains ( Figure 1C , Text S1 ) . Most ( 93 . 2% ) of these core genes have blast hits in NCBI non-redundant ( NR ) database , of which 94% were functionally annotated . The core genes mostly encode proteins involved in DNA , RNA , and protein synthesis and modification , transporters , signal transduction and central metabolic pathways ( Figure 1D; for functional classification based on molecular function and cellular component , see Figure S4 ) . The accessory genes , referring to those missing in at least one strain , mainly encode ( i ) central metabolism such as carbohydrate , lipid , energy , and nucleotide and amino acid metabolism ( which are overlapped with the core genes ) , ( ii ) secondary metabolism and N-glycan biosynthesis ( which are complementary to the core genes , and ( iii ) unknown functions ( Figure S5 ) . There were 164–1 , 513 genes that were strain-specific among the six genomes . In contrast to the 2 , 734 Nannochloropsis core genes , of which 96 . 7% were supported by our mRNA-Seq reads , 11 . 0% ( 18 ) of the 164 IMET1-specific proteins lacked such supports , suggesting the possible presence of pseudogenes or false positives in gene prediction . Among the IMET1-specific genes with mRNA support , 94 . 5% were putative novel genes without any known homologs ( Blast hits ) in the NCBI NR database . It is possible that some of them might have horizontally transferred from unsequenced species . Among strain-specific genes with functional annotations , most were involved in responses to freezing in N . oculata CCMP525 , N . granulata CCMP529 and N . oceanica CCMP531 . In N . gaditana CCMP526 , transporters were prevalent , while in N . salina CCMP537 , no significant enrichment was found in any processes ( Figure S6 ) . Correlation analysis revealed that the core and accessory genes exhibited different sequence and transcriptional features under the experimental conditions tested ( Text S2 ) . The accessory genes tend to be under lower purifying pressure while lower transcriptional levels ( Text S2; Figure S7; Figure S8 ) , supporting a link between sequence evolution and transcriptional activity [10] , [11] , [12] . To probe the link between the accessory genes and divergence of the genomes , protein-coding genes in the six Nannochloropsis were classified into different groups based on the number of strains in which they were present ( thus those present in all the six strains were part of the Nannochloropsis core ) . The most prominent group included the genes shared by four of the strains , in which the majority ( 97 . 3% ) were found in the phylogenetically closely related species , i . e . , N . oceanica ( two strains ) , N . granulata and N . oculata . The absence of these genes in the other two species explained the small number of Nannochloropsis core genes ( Figure S9A ) . These genes might have been present in the common ancestors of heterokonts and later lost in N . salina and N . gaditana , as >60% of them were found in other heterokonts ( e . g . , diatoms , Ectocarpus and other non-photosynthetic heterokonts such as Phytophthora ) . The functions supported by these genes were similar to those of core genes , with oxidation-reduction , transmembrane transport and protein-related metabolism being dominant . This does not support the presence of functional bias in the gene loss events ( Figure S9B ) . To seek the cause of the structural divergence among the Nannochloropsis genomes , we clustered all encoded proteins based on their sequence similarity . Among sequenced plant and algal genomes , large paralogous groups are common , e . g . , 217 F-box family protein genes in Arabidopsis [13] and 51 Class III guanylyl and adenylyl cyclase genes in Chlamydomonas [7] ) . However , Nannochloropsis spp . appear to have adopted a strategy in which paralogous groups are less biased in size , i . e . , they formulate a large number of relatively small paralogous groups ( Figure 2 , Text S1 ) . There are 4 , 263 , 4 , 325 , and 7 , 171 paralogous groups in Thalassiosira , Chlamydomonas , and N . oceanica IMET1 , respectively . The top 15 largest paralogous groups in each Nannochloropsis genome range in size from two to seven genes , with a median value of three to four ( Figure 2 ) . For example , the largest paralogous group in IMET1 consists of 11 genes ( mainly in metabolic process ) , which is in a sharp contrast with T . pseudonana ( 46 genes; protein modification process; [8] ) , Cyanidioschyzon merlae ( 23 genes; DNA metabolic process; [14] ) and C . reinhardtii ( 150 genes; protein modification process; [7] ) . As genes from different origins might exhibit relatively low sequence conservation and thus fail to formulate a paralogous group , the reduced sizes of paralogous groups in the Nannochloropsis genomes might result from the integration of multiple genome resources , which is consistent with the proposal that heterokonts originated from multiple secondary endosymbiosis [15] . This observation also suggests that strain-specific gene sequence duplication was relatively rare in Nannochloropsis . On the other hand , it is also possible that Nannochloropsis spp . have adapted to their environment via a strategy of frugality in proteome structure , with paralogous protein-coding genes either emerging less frequently or many of them being lost . Among the 8 , 992 homologous groups from the six Nannochloropsis genomes ( by OrthoMCL [16]; based on amino acid sequence similarity ) , 1 , 731 included the genes from all six strains . However , 2 , 312 , 1 , 515 , 1 , 551 and 1 , 653 groups included genes from two , three , four and five of the strains , respectively , and thus were “mosaic” groups as they included genes from only a subset , but not all , of the strains . Furthermore , 230 groups were specific to one of the six strains , with 4 to 151 such groups in each strain . The large number of mosaic paralogous groups ( 7 , 031 or 78 . 2% in total ) could explain the large size of the Nannochloropsis pan-genome , although the numbers of genes and gene groups could be over- or under-estimated due to the presence of alternative splice forms or artifacts of genome assembly . Despite their high structural diversity , each of the six Nannochloropsis genomes exhibits functional features that underlie their oleaginous phenotypes . There is significantly higher gene enrichment for cellular lipid metabolism in each genome than in C . reinhardtii ( Figure 3 , Dataset S2 ) . In all or most of the Nannochloropsis strains , the subcategories of lipid metabolism are enriched , including glycerolipid metabolism , phospholipid metabolism , lipopolysaccharide metabolism and lipid modification . Metabolic pathways enriched in Nannochloropsis also include organic acid metabolism , precursor generation and sulfur compound metabolism . Genes related to stress response , including responses to DNA damage stimulus , DNA repair and cold stress response , were also enriched in several Nannochloropsis strains . However , the number of genes involved in phosphorus metabolism and cellular macromolecule metabolism was significantly lower in each Nannochloropsis strain than in C . reinhardtii ( Figure 3 ) . Thus , the enrichment of gene doses in lipid metabolism pathways and stress response-related pathways appears to be a shared feature of Nannochloropsis genomes and likely underlies their advantageous oleaginous and environmental tolerance traits . In the lipid biosynthesis pathway ( the de novo biosynthesis of fatty acids and TAG ) , a prominent expansion in gene copy number in particular reaction nodes was observed as a shared feature among the six Nannochloropsis strains , despite a genome size only one-fourth of C . reinhardtii . Such enriched genes include those encoding ketoacyl-ACP synthase ( KAS , four to five in each Nannochloropsis strain vs . three in C . reinhardtii ) , acyl-ACP thioesterase ( acyl-ACP TE , five vs . one ) , long-chain fatty acyl-CoA synthetase ( LC-FACS , 11–12 vs . seven ) , phosphatidic acid phosphatase ( PAP , five vs . one ) , and the last two acyltransferases: lysophosphatidyl acyltransferase ( LPAT , seven to eight vs . one ) and diacylglycerol acyltransferase ( DGAT ) ( Figure 4 ) . Multiple copies of KAS proteins were found in each Nannochloropsis strain for the assembly of type II fatty acid synthases . In addition , six bacterial type I fatty acid synthase genes , each with several conserved functional domains , were identified ( compared to only one in C . reinhardtii ) ; phylogenetic analysis revealed that these genes are closely related to polyketide synthases ( Figure S10 ) , yet they might be involved in fatty acid synthesis [17] . Notably , such expansion in gene dose was not ubiquitous along the TAG pathway . For many of the nodes , the gene doses are comparable to those in C . reinhardtii ( Figure 4B ) . These nodes include the acetyl-CoA carboxylase ( ACCase ) , MCAT , KAR , HAD in fatty acid biosynthesis , GPAT in TAG assembly , and other membrane lipid biosynthesis-related enzymes ( such as the MGD and DGD in galactolipid synthesis , SQD in sulfolipid synthesis , BtaA and BtaB in betaine lipid synthesis and EPT in phosphatidylethanolamine synthesis ) . The expansion of gene dose for the selective steps highlights their crucial roles in channeling carbon flux into TAG synthesis and might be considered a “genomic signature” of oleaginousness . To probe the evolutionary forces expanding the TAG biosynthesis gene repertoire in Nannochloropsis , we carried out a phylogenomic analysis to investigate the horizontal gene transfer ( HGT ) events in N . oceanica IMET1 genome ( Text S1; [18] ) . We identified 99 HGT candidates ( Figure S11A; Dataset S3 ) , accounting for 1 . 0% of nuclear genes . Among them , the most abundant functions encoded ( in terms of GO Slim terms in biological process ) included biosynthetic process , small molecule metabolism , cellular nitrogen compound metabolism and lipid metabolism ( Figure S11 ) . HGT appeared to have played an important role in the evolution of oleaginousness loci in these organisms . Totally nine HGT candidates ( 15 . 3% of total lipid biosynthesis genes , much higher than average percentage of HGT presence in nuclear genome ) were inferred in most of the nodes with increased gene doses , such as KAS , enoyl-ACP reductase ( ENR ) , acyl-ACP TE , LC-FACS and PAP ( Figure 4A , Figure S12 , Figure S13 ) . PAP catalyzes the Mg2+-dependent dephosphorylation of phosphatidic acid ( PA ) to yield diacylglycerol ( DAG ) and Pi . Both PA ( via CDP-DAG ) and DAG can enter phospholipid synthesis , and DAG is the direct precursor of TAG . Thus , PAP may control the direction of carbon flux and affect overall cellular lipid synthesis [19] . Five genes encoding PAP enzymes were found in each Nannochloropsis strain: three were conserved in eukaryotes , while the other two were clustered with the bacteria , indicating a bacterial HGT origin ( either one HGT followed by gene duplication or multiple horizontal transfers; Figure S12F , Figure S13F ) . The two horizontally transferred PAP genes exhibited higher transcriptional levels than the eukaryotic ones . The presence of multiple prokaryotic PAP genes suggests complex mechanisms to regulate the substrate preference for the synthesis of various classes and species of lipids . Among the ENR genes in each Nannochloropsis strain , two likely originated by HGT from bacteria into the common ancestor of the six Nannochloropsis strains ( Figure S12C , Figure S13C; suggested by the absence of other heterokonts in the bacterial ENR clade ) , which were then inherited by each of the Nannochloropsis strains . The most prominent example of gene dose expansion is DGAT , which catalyzes the last step of TAG synthesis from DAG and acyl-CoA [20] and includes DGAT-1 and DGAT-2 [21] . There are 12–13 DGAT in each Nannochloropsis strain ( one to two DGAT-1 and 11 DGAT-2 ) , representing the highest dose among known genomes ( Figure 5A ) . In contrast , only six and four DGAT are present in C . reinhardtii and the diatom T . thalassiosira , respectively , and even fewer in some other green algae and heterokonts ( Figure 5A ) . In IMET1 , all the DGAT-1 and DGAT-2 were transcriptionally active [FPKM ( Fragments Per Kilobase of exon per Million mapped reads ) >1 . 0] . Phylogenetic analysis of DGAT from selected bacteria , fungi , algae and higher plants revealed extraordinary evolutionary diversities of all 74 DGAT in the six Nannochloropsis strains ( Figure S14 ) . Several observations were apparent . ( i ) The partition of DGAT-1 and DGAT-2 might have occurred early , likely before the primary endosymbiosis event or even earlier . ( ii ) The copy number of DGAT-1 was lower ( 1–2 ) and less variable than that of DGAT-2 , which is consistent in a wide range of organisms from bacteria to land plants . ( iii ) A similar degree of DGAT-2 dose expansion was observed in all six Nannochloropsis strains ( Figure 4B ) . Moreover , for each of the 11 DGAT-2 identified in each strain , the orthologs in the other five strains were all identified and clustered into a phylogenetic group ( Figure S14 ) ; the sequence identity between orthologous gene pairs was >98% between the two N . oceanica strains , >80% among N . oceanica , N . oculata and N . granulate , and >65% between N . oceanica IMET1 and the outmost N . gaditana . These results suggested the stable inheritance of DGAT-2 genes in Nannochloropsis evolution . In contrast , DGAT-1 might have experienced species-specific gene loss . For example , no counterparts of DGAT-1B in IMET1 were found in N . salina and N . gaditana despite a high degree of conservation of this gene in the other four strains . ( iv ) The 11 DGAT-2 genes in IMET1 exhibited relatively low intra-genome pairwise identity ( averaging 18% ) , and each was grouped into a separate paralogous group with its orthologs from the other five Nannochloropsis strains , indicative of distinct and divergent phylogenetic origins of DGAT-2 in Nannochloropsis . Two of the DGAT-2 genes in IMET1 ( DGAT-2F and DGAT-2D ) exhibited a relatively high protein sequence similarity ( identity at 51% ) , suggesting that the two genes might be derived from a gene duplication event in the Nannochloropsis lineage ( Figure S14 ) . However this individual case of suspected gene duplication cannot account for the expanded dose of DGAT-2 genes in IMET1 . To infer the origin of these genes , a comprehensive phylogenetic analysis was carried out among all species with genomes and ESTs available in several public databases ( Text S1; [22] ) . DGAT-2C showed a phylogeny with strong affiliation with the red algae C . merolae and formed a sister group with those from other chromalveolates ( Figure S15A , Figure S16A ) . The most plausible explanation for such strong links between Nannochloropsis and red algal DGAT is a red-algae derivation of DGAT-2C through endosymbiotic gene transfer ( EGT ) in the secondary endosymbiosis event ( which was proposed as the evolutionary mechanism through which the common ancestor of chromalveolates acquired chloroplasts from a red algae-related endosymbiont [23] ) . On the other hand , four DGAT-2 genes , including DGAT-2A , DGAT-2B , DGAT-2I and DGAT-2G , are clustered with their counterparts from green algae as well as other chromalveolates ( Figure S15B–E , Figure S16B–E ) , suggesting a green algal origin of these DGAT-2 genes . This is consistent with the hypothesis of a green algae related endosymbiont residing in the common ancestor of chromalveolates [15] . Three ( DGAT-2A , DGAT-2I and DGAT-2C ) of the above five red-lineage ( red algae derived ) and green-lineage ( green algae derived ) DGAT-2 genes were predicted to harbor chloroplast targeting signals , supporting their ancestral derivation from the endosymbionts . The higher dose of green- than red-lineage DGAT-2 in each of the Nannochloropsis strains suggests a more significant contribution of the green lineage to the oleaginous traits of modern Nannochloropsis . Furthermore , phylogenetic trees of the other six DGAT-2 genes did not exhibit unambiguous relationships with those from red or green algae and are thus referred to as “unresolved . ” It is possible that several of these genes originated from the secondary host [24] , as four ( DGAT-2D , DGAT-2E , DGAT-2F and DGAT-2H ) of the six genes were predicted to be located in the endoplasmic reticulum ( ER ) or cytosol . Thus , the observed sequence divergence of the 74 DGAT genes ( eight type I and 66 type II ) in the six Nannochloropsis genomes mainly resulted from their diverse origins from the red- or green-algae–related endosymbionts ( through EGT ) and the secondary host ( Figure 5B ) . Phylogenetic evidence also supports a green endosymbiont origin for one gene encoding MCAT ( s00247 . g6828 ) in fatty acid biosynthesis ( Figure S15F , Figure S16F ) . Thus , the diverse evolutionary origin of the Nannochloropsis DGAT-2s and one of the other lipid synthesis genes has underlain their massive genetic pools and likely contributed to the extraordinary capacity for TAG synthesis in present-day strains . In addition to their diverse origins , differentiation in selective pressure appeared to underlie the sequence divergence of DGAT and other members of lipid-related gene families . DGAT genes in Nannochloropsis were generally under strong purifying selective pressure ( Ka/Ks typically under 0 . 1 ) . However , higher Ka/Ks ratios were observed in the red-lineage DGAT-2C of 0 . 11 ( Figure S14 ) . No significant difference in the ratio was found between the green-lineage and secondary-host–originated DGAT-2 . Furthermore , DGAT-2C with the highest Ka/Ks ratio was among the DGAT-2 genes with the lowest transcriptional level under normal growing conditions , while DGAT-2J with the lowest ratio was one of the most transcribed DGAT-2 ( second only to DGAT-2A ) . These findings add further support to the negative correlation between transcriptional level and selective pressure in the evolution of Nannochloropsis genes . Dramatic enrichment of glycoside hydrolase ( GH ) genes accompanied by a reduced pool of glycoside synthase genes ( as compared to C . reinhardtii ) was also observed in each of the Nannochloropsis genomes . C . reinhardtii harbors seven starch synthase genes for starch production and two 1 , 3-β-glucan synthase genes; however , each Nannochloropsis encodes just one 1 , 3-β-glucan synthase gene ( which might convert glucose into the polysaccharide chrysolaminarin or laminarin ) , and no starch synthase genes were found . Conversely , 48–49 GH genes were found in each strain ( Dataset S4 ) , with a gene dose per Mb of genome 6–7 fold higher than that of C . reinhardtii ( 27 GH genes ) . These Nannochloropsis genes were from 13 different GH families , dominated by GH2 ( 13 members ) , GH9 , GH3 and GH1 families with over four members . Surprisingly , there were only three genes in the GH16 family , which specifically hydrolyzes the glycosidic bond of 1 , 3-β-glucan , while GH16 was the dominant group in C . reinhardtii , with five members . In IMET1 , 91 . 7% of the 48 GH genes were transcriptionally active at each of the time points under both N-replete and N-depleted culture conditions ( FPKM>1 . 0; Text S1 ) . Among them , 16 exhibited significant variations at the transcriptional level under N-depleted conditions ( 10 with increased transcription ) , including two members of the GH2_C family and one GH17 gene with a significant increase in transcription ( fold-change >1 . 5 ) from 3 h and 6 h after the onset of N-depletion and one GH9 gene down-regulated under the same conditions . The monosaccharides released from GH that catalyzed hydrolysis of polysaccharides may be used in glycolysis to produce acetyl-CoA and ATP for fatty acid synthesis . Among the 48 GH genes in each strain that were conserved among the six genomes , 16 were inherited from the common ancestor of heterokonts , as their homologs were found in the diatoms Phytophthora and Ectocarpus . Another five GH genes were likely to have originated from bacteria via HGT . Among these , three GH8 genes inferred to be horizontally acquired from cellulose-digesting Clostridium-like bacterium were absent in other sequenced unicellular algae . The remaining 27 GH genes were phylogenetically closest to homologs in animals , insects or multicellular fungi , such as the nine putative cellulase genes that were most similar to those in the nematode Pristionchus , indicating HGT events with donors being Nannochloropsis-like organisms [25] . In addition , N . granulata and N . salina each possessed one strain-specific GH gene that might have been introduced after their speciation .
Microalgae , which are primarily unicellular , aquatic and photosynthetic eukaryotes , are phylogenetically diverse . They are responsible for over 45% of our planet's annual net primary biomass [26] . The Nannochloropsis genomes studied here , one of the first such datasets for microalgae , reveal the nature and degree of genome divergence and dynamics at the strain , species and genus level . They could serve as an initial framework for genome-wide association studies , while the genome-derived nuclear gene markers should be useful for highly sensitive typing of strains . The genomes of the six oleaginous Nannochloropsis strains presented here are of relatively small size and high coding potential and many fewer mobile elements compared to many previously sequenced microalgae [9] . The large size of the Nannochloropsis pan-genome can be partially traced to the large number of mosaic paralogous groups , which further suggests a significant degree of species-specific gene loss during Nannochloropsis evolution . On the other hand , the small core genome size and the large number of mosaic homologous gene clusters among the Nannochloropsis spp . suggest that , as one moves down the tree of life for stramenopiles , the number of shared genes reduces quickly and is replaced by lineage-specific gene gains and losses . The core genes generally exhibit lower Ka/Ks ratio but higher transcriptional levels than non-core genes , suggesting their roles in shaping the evolution of microalgal genes . Our findings , together with observations in yeasts [10] , [27] , revealed a link that is conserved in unicellular eukaryotes in terms of gene function , selective pressure , transcriptional level and gene essentiality . Despite the high sequence diversity of protein-coding genes , the six Nannochloropsis genomes shared a genus-level oleaginousness signature that included enrichment of selective lipid biosynthesis genes and certain glycoside hydrolases that potentially shift carbon flux from storage carbohydrate to TAG synthesis . It is quite remarkable that these gene expansions have occurred despite their significant genome shrinkage relative to other microalgae such as C . reinhardtii . Different mechanisms have underlain the emergence of the signature . Multiple-genome pooling was particularly evident for the 11 DGAT-2 in each strain , which were all transcriptionally expressed during TAG synthesis and apparently originated from at least three ancient genomes: ( i ) the engulfed green algae , ( ii ) the engulfed red algae and ( iii ) the host cell in the secondary endosymbiosis . Chromalveolates include both photosynthetic ( e . g . diatoms and Eustigmatophyceae that include Nannochloropsis ) and non-photosynthetic taxa ( e . g . , Phytophthora ) . The chromalveolate hypothesis suggests that the common ancestor of Chromalveolates originated via an eukaryotic host ( i . e . , the secondary host ) engulfing a red alga ( as the secondary plastid ) in an ancient secondary endosymbiosis event [23] . The presence of a large number of “green genes” in the diatom nuclear genomes has been interpreted as evidence of a cryptic prasinophyte-like secondary endosymbiosis before the red algae intake [15] . Though confounded by potential sampling bias against red algae and artifacts in phylogenetic analysis [28] , this hypothesis was supported by the 172 membrane transporter genes showing potential origins from green or red algae in a relatively strict phylogenomic analysis [22] . Moreover , genomes of the cryptophyte alga Guillardia theta and the chlorarachniophyte alga Bigelowiella natans also contain hundreds of genes with a phylogenetic affiliation to red or green algae [24] . Our search of DGAT-2 in publicly available red algae genomes ( and ESTs ) returned one DGAT-2 each from Cyanidioschyzon merolae , Galdieria sulphuraria [18] and Porphyridium purpureum [29] . The paucity of DGAT-2 in red algal genomes and the distinct features of these genes in the six Nannochloropsis genomes ( the greatly expanded copy number , large pair-wise sequence divergence , rare gene duplication events , and absence of mobile elements or evidence for HGT in each of the DGAT-2 loci ) suggested multiple-genome pooling as the cause for the massive DGAT pool in Nannochloropsis spp . These findings also provided additional support for the existence of a green algae–associated secondary endosymbiosis in the evolutionary history of chromalveolates . Furthermore , among the six Nannochloropsis strains , the inheritance of each DGAT-2 was highly conserved in that no strain-specific duplications or losses were found for any DGAT-2 in each of the six strains , and the genes have been under strong negative selection . In contrast , diatoms such as T . pseudonana ( believed to have also experienced the multiple secondary endosymbiosis [15] ) encode many fewer DGAT-2; only four DGAT-2 were identified , and all were predicted to be from the green algae–related endosymbiont and the secondary host , with none from red lineage . The absence of Thalassiosira genes in certain gene-phylogeny clusters ( e . g . , the red-lineage DGAT-2C ) in the diatom , in contrast to the presence of these genes in Nannochloropsis and many other heterokonts , suggests the loss of DGAT-2 in diatoms . Thus , such strict inheritance and stable maintenance of the large reservoir of DGAT-2 from multiple lineages seem to be Nannochloropsis-specific . It also suggests the essentiality of each DGAT-2 and its possible functional complementarity in the cell . In addition , HGT primarily from bacteria were found in the majority of the gene dose-expanded lipid synthesis nodes and in many glycoside hydrolases . In the red alga Galdieria sulphuraria , 5% of protein-coding genes were acquired from bacteria and archaea via HGT , which forged its adaptation to a thermophilic and metal-rich environment [18] . The HGT events in Nannochloropsis likely reflected an organismal adaptation to a niche that favored oleaginousness and glycoside hydrolysis . Therefore , the multiple-genome pooling and horizontal genetic exchange from bacteria , together with the selective inheritance of lipid synthesis genes and species-specific gene loss , might have underlain the enormous genetic apparatus for oleaginousness and led to the structural divergence and functional conservation observed among present-day Nannochloropsis . In many organisms , other mechanisms such as gene and genome duplications may play an important role in supplying new genetic materials for organismal adaptation [30] and have been frequently proposed as drivers of the emergence of particular traits in bacteria [31] , [32] , fungi [33] , [34] , plants [35] and animals [36] . Thus , the extraordinary origin and evolution of oleaginous traits in Nannochloropsis have important implications in the selection and genetic engineering of such traits in these and other microalgae of economic interest .
All genomic data for this study , including the assembled genomes and mRNA-Seq data , were deposited at NCBI . The BioProject accessions for assembled genomes were: PRJNA202418 for N . oceanica IMET1 , PRJNA65107 for N . oculata CCMP525 , PRJNA65111 for N . granulata CCMP529 , PRJNA65113 for N . oceanica CCMP531 and PRJNA62503 for N . salina CCMP537 . The mRNA-Seq data were deposited at SRA under SRP032930 . Five new Nannochloropsis genomes were sequenced in this work ( Table 1; Table S1 ) . For Nannochloropsis oceanica IMET1 , both shotgun sequencing data and paired-end data with different pair distances from 454 Titanium and Illumina GAIIx were collected . Newbler ( Roche ) was used for initial assembly . Gap-filling and scaffold-building were performed with Illumina data , followed by manual manipulation and sorting of contigs . Genes were predicted by combining the ab initio predictions with predictions based on mRNA-Seq read alignments ( 387K aligned cDNA reads from a Roche 454 Sequencer ) by AUGUSTUS ( v2 . 5 ) . For each of the other four Nannochloropsis strains ( Table S1 ) , paired GAIIx reads were assembled using Velvet with specified insert sizes . The previously published genome sequence of N . gaditana CCMP526 [5] was downloaded from http://Nannochloropsis . genomeprojectsolutions-databases . com/ . Gene models of each of the six genomes were predicted using two different ab initio gene predictors ( AUGUSTUS and GeneID ) followed by a combination of gene models using EVidenceModeler ( EVM ) with a 1∶1 weight ratio . For all strains , predicted protein-coding genes were annotated via searching against three databases: the NCBI NR and KEGG databases by BlastP , and the Gene Ontology database by InterProScan . GO terms were mapped to the GO slim hierarchy proposed by the GO consortium by a customized script ( http://www . bioenergychina . org/fg/d . wang_scripts/ ) . For collecting the transcriptomics datasets underlying TAG production , N . oceanica IMET1 was cultivated in f/2 liquid medium [37] with 4 mM NO3− under continuous light at 50 µmol photons m−2 s−1 . Mid-logarithmic phase algal cells were inoculated in nitrogen-replete and nitrogen-depleted conditions , respectively . Total RNA were collected at 3 , 6 and 24 h after each inoculation and pooled together for full-length cDNA sequencing in 454 Titanium . The data produced were subsequently used for gene prediction . Furthermore , total RNA from each of the aforementioned control ( nitrogen-replete ) and nitrogen-starvation conditions along the time points of 3 , 6 and 24 h after the onset of nitrogen depletion ( six samples under each condition ) were loaded for mRNA-Seq in Illumina GAIIx . Nannochloropsis core genes were identified as the intersections of the five “IMET1 pairwise cores , ” which were obtained by searching IMET1 proteins via BlastP and tBlastN against the proteome and the genome , respectively , of each of the other five strains with an e-value cutoff of 1e-5 and a protein sequence identity cutoff of 80% . Paralogous groups among these six strains were identified by a Markov Clustering algorithm ( OrthoMCL [16] , v . 4 ) with an inflation index of 1 . 5 . PAML ( v . 4 . 4c ) codon substitution models and likelihood ratio tests ( codeml ) were used to estimate the selective pressure . An identical method was applied in the establishment of paralogous groups among other model microalgae . HGT candidates were inferred following the method in the genomic analysis of Galdieria sulphuraria ( Text S1; [18] ) . Phylogenetic trees for each of the putative HGT genes in NEWICK format were deposited in Dataset S3 . The phylogenetic tree for each HGT candidate was manually checked and only accepted when a clear pattern of HGT was observed in both Neighbor Joining ( NJ ) and Maximum Likelihood ( ML ) trees . To deduce the evolutionary origins of lipid biosynthesis-related genes , we first implemented the strategy described in Chan et al . [22] to build a comprehensive database and to construct the homologous groups for each lipid synthesis gene , except that we collected more recently published genomes and EST datasets updated in public databases , including genomes of the red algae G . sulphuraria [18] , Chondrus crispus [38] and Porphyridium purpureum [29] . In the following phylogenetic analysis , phylogenies for the homologous group of each lipid synthesis gene were constructed in MEGA5 by both NJ and ML methods . A gene was inferred to be potentially derived from a green or red algae related secondary endosymbiont when the phylogeny was supported by both NJ and ML trees . For a comprehensive and detailed description of the methods , please refer to Text S1 . | Microalgae are promising feedstock for biofuels . However , the diversity , origin and evolution of oil-producing microalgal genomes in general , and those of their oleaginous traits in particular , remain poorly understood . We present five new genomes of the oleaginous microalgae Nannochloropsis spp . that allow genus- , species- and strain-level genomic comparison . With each Nannochloropsis genome encoding approximately 6 , 562–9 , 915 genes , a core genome of ca . 2 , 700 genes and a large pan-genome of >38 , 000 genes were found . The genomes share key genetic features such as gene dose expansion of selected nodes in lipid biosynthesis pathways . Evidence of horizontal gene transfers , primarily from bacteria , was found in most of these nodes . However , the eleven type II acyl-CoA:diacylglycerol acyltransferase genes ( DGAT-2 ) , the highest gene dose reported among known organisms , likely originated from three ancient genomes of the secondary endosymbiosis host and the engulfed green and red algae; they were strictly vertically inherited in each of the Nannochloropsis spp . Thus , multiple genome pooling and horizontal genetic exchange have underlain the enormous genetic makeup underlying TAG production in present-day Nannochloropsis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genome",
"sequencing",
"plant",
"science",
"functional",
"genomics",
"genome",
"evolution",
"plant",
"evolution",
"genetics",
"plant",
"genetics",
"plant",
"genomics",
"comparative",
"genomics",
"biology",
"genomics",
"evolutionary",
"biology",
"genomic",
"evolution",
... | 2014 | Nannochloropsis Genomes Reveal Evolution of Microalgal Oleaginous Traits |
An increasing number of broadly neutralizing antibodies ( bnAbs ) are considered leads for HIV-1 vaccine development and novel therapeutics . Here , we systematically explored the capacity of bnAbs to neutralize HIV-1 prior to and post-CD4 engagement and to block HIV-1 cell-cell transmission . Cell-cell spread is known to promote a highly efficient infection with HIV-1 which can inflict dramatic losses in neutralization potency compared to free virus infection . Selection of bnAbs that are capable of suppressing HIV irrespective of the transmission mode therefore needs to be considered to ascertain their in vivo activity in therapeutic use and vaccines . Employing assay systems that allow for unambiguous discrimination between free virus and cell-cell transmission to T cells , we probed a panel of 16 bnAbs for their activity against 11 viruses from subtypes A , B and C during both transmission modes . Over a wide range of bnAb-virus combinations tested , inhibitory activity against HIV-1 cell-cell transmission was strongly decreased compared to free virus transmission . Activity loss varied considerably between virus strains and was inversely associated with neutralization of free virus spread for V1V2- and V3-directed bnAbs . In rare bnAb-virus combinations , inhibition for both transmission modes was comparable but no bnAb potently blocked cell-cell transmission across all probed virus strains . Mathematical analysis indicated an increased probability of bnAb resistance mutations to arise in cell-cell rather than free virus spread , further highlighting the need to block this pathway . Importantly , the capacity to efficiently neutralize prior to CD4 engagement correlated with the inhibition efficacy against free virus but not cell-cell transmitted virus . Pre-CD4 attachment activity proved strongest amongst CD4bs bnAbs and varied substantially for V3 and V1V2 loop bnAbs in a strain-dependent manner . In summary , bnAb activity against divergent viruses varied depending on the transmission mode and differed depending on the window of action during the entry process , underscoring that powerful combinations of bnAbs are needed for in vivo application .
Recently identified highly potent broadly neutralizing HIV antibodies ( bnAbs ) are considered as lead components for vaccines and immunotherapeutics ( reviewed in [1–5] ) and extensive characterization of these bnAbs in vitro and in vivo is underway to select the most promising candidates [6] . Proof of activity in animal models , the SHIV rhesus macaque infection model or HIV infection of humanized mice , is considered the most conclusive efficacy testing and is required before application in humans can be considered [7–14] . However , investigations in animal models are currently restricted to only few viral strains , limiting the possibility to judge the in vivo breadth of the bnAbs tested . Assessment of breadth currently only relies on free virus inhibition in vitro , most commonly in the standardized TZM-bl reporter neutralization assay [15 , 16] . Comparison of bnAb activity in vitro and their neutralizing titers in vivo has indicated , however , that required doses in vivo are >100-fold higher [12 , 13 , 17–19] which has been partially attributed to lower tissue concentration of delivered antibodies [11 , 20] , the need to potently elicit antibody-effector functions [21–23] and a reduced activity of antibodies in cell-cell transmission [24–30] . Recent years have broadened our understanding of the HIV infection process and highlighted that the virus has multiple ways of entering and infecting target cells including both , infection by free viruses and direct virus transmission from infected to non-infected cells [31–34] . Cell-cell transmission proved , at least in vitro , to be substantially more efficient than free virus spread [35–41] and may hence foster replication of virus strains with low replicative capacity [42] . While the relative impact of free virus and cell-cell transmission in vivo remains to be defined , intense research efforts have delineated the molecular processes involved in HIV-1 cell-cell transmission ( reviewed in: [43 , 44] ) . Aspects of cell-cell transmission have been studied in a range of experimental setups reaching from visualization of individual virological synapses-the contact area that forms between T cells during viral cell-cell transmission- to replication assays in primary cells [25 , 26 , 30–33 , 38 , 41 , 45–51] . However , all experimental approaches so far share the difficulty in monitoring true cell-cell transmission events by dissecting these from free virus infection and cell fusion . Assay formats are thus commonly tailored to address individual research questions reducing the possibility for direct comparisons between studies . Cell-cell transmission systems described to date vary in respect to donor and target cells , HIV strains and infection systems ( multiple–round and single round infection ) , virus input and readouts used ( reporter assays , direct detection of HIV-1 antigens , single cell and bulk cell analysis ) and these differences are thought in part to have led to contradicting observations in respect to neutralizing antibody and ART activity during cell-cell transmission [25–27 , 46 , 47 , 50–54] . Despite these discrepancies , there is agreement that cell-cell transmission , at least in vitro , is vastly more efficient than free virus spread and leads to multiple infection of target cells [36–38] . Further , higher rates of transferred viruses seem to reduce the efficacy of reverse transcriptase ( RT ) inhibitors [46 , 47 , 50 , 51] . While neutralizing antibodies and entry inhibitors with few exceptions displayed reduced efficacy during cell-cell transmission , the magnitude of the reported activity loss varied across different studies [25–27] , suggesting that a range of factors contribute to the efficacy of neutralizing antibodies during cell-cell transmission and that not all of these factors are captured equally by the different assay systems . Indeed , functional differences of the entry process [49] , the maturation status of the virus during cell-cell transmission [49] and the stage of the entry process blocked by neutralizing antibodies [25 , 30] have been proposed as factors that influence antibody activity during cell-cell transmission . To verify bnAb activity during cell-cell transmission , a widely applicable assay such as the TZM-bl neutralization assay used for screening neutralization against free virus infection would be highly desirable . Preferably , such an assay should allow for high throughput inhibitor screening under physiological relevant conditions , most ideally by studying T cell to T cell transmission . Free virus infection ( e . g . by using single round replicating viruses ) and cell fusion ( e . g . by using donor cells that solely express envelope ) can be readily assessed in vitro . In contrast , monitoring cell-cell transmission while excluding the influence of free virus infection and fusion events remains challenging and is in most assay formats only partially achieved [26 , 27 , 46] . In the present study , we adapted the recently established inGluc reporter system [41 , 55 , 56] to allow for a side-by-side assessment of bnAb neutralization of free virus infection of T cells and cell-cell transmission to T cells using A3 . 01-CCR5 T cells as target cells . Neutralization during cell-cell transmission in this assay proved to perform identical to PBMC-PBMC transmission and allowed us to probe 16 bnAbs targeting the CD4bs , the V3 glycan region , the V1V2 loop and the MPER domain in their activity against 11 HIV-1 strains from subtypes A , B and C . The mechanistic features of bnAb neutralization we explored in the present study may help to explain their in vivo function and to select of the most promising candidate bnAbs for therapeutics and vaccine development . Besides probing the capacity of bnAbs to inhibit cell-cell transmission at comparable breadth and potency as free virus spread , we measured the bnAbs’ ability to access the viral envelope ( Env ) trimer pre- and post-CD4 engagement , as we had previously observed that post-CD4 engagement may be beneficial for maintaining neutralization activity during cell-cell transmission [25] . As we highlight here , bnAbs vary substantially in their capacity to inhibit cell-cell transmission of divergent strains . While we identified bnAbs that retain efficacy to a higher extent , our data reveal that loss of activity during cell-cell transmission cannot generally be predicted by a bnAb’s performance in free virus inhibition . CD4bs- directed bnAbs showed a wide range of activity loss , irrespective of their potency against free virus spread . Unexpectedly , for bnAbs directed to the V3 glycan domain and the V1V2 loop , inhibitory concentrations required for free virus inhibition were inversely correlated with activity loss during cell-cell transmission . This indicates that bnAb features steering high potency to block free virus transmission are not necessarily a driving factor for inhibiting cell-cell transmission . In some cases , V1V2-specific bnAbs displayed even higher activities against cell-cell transmission than free virus infection . Inhibitory activity post-CD4 engagement also differed in a strain- and bnAb-dependent manner and was inversely linked with pre-CD4 attachment activity .
To enable a controlled comparison of bnAb neutralization activity in free virus and cell-cell transmission , we sought to employ an assay format , in which i ) both free virus and cell-cell transmission are studied using Env-pseudotyped viruses to limit the infection to a single round , ii ) target cells used for free virus and cell-cell transmission are identical and yield results comparable to primary T cell ( PBMC ) infection , iii ) the assays are scalable and can be used to evaluate larger antibody and virus panels . Considering the potential of the standardized TZM-bl neutralization assay that utilizes 293-T-produced Env pseudoviruses for infection [57] , co-culture of TZM-bl with 293-T donor cells expressing Env pseudoviruses would appear as an attractive setup to screen and compare neutralization activity during cell-cell transmission . However , co-cultures of TZM-bl and pseudovirus expressing 293-T cells are highly prone to fusion , saturating the attainable luciferase reporter signal through direct transfer of Tat from donor cells to an extent that de novo infection by free virus and cell-cell transmission cannot be reliably quantified simultaneously . While there is certainly interest in exploring cell fusion during the interactions of infected cells in vivo [58] , we here specifically aimed at comparing the sensitivity of virions during free virus and cell-cell transmission using an assay setup that allows to exclusively monitor these two events . To this end , we employed a recently developed reporter virus system [41 , 55 , 56] , specifically tailored to record cell-cell transmission using a readout that is not influenced by cell fusion . The system relies on a reporter pseudotyped vector that contains an intron-regulated Gaussia luciferase gene ( NLinGluc ) . Reverse orientation of the Gaussia luciferase gene in the plasmid and the intron prevent luciferase expression from the transfected vector in the producer cells . Gaussia luciferase is only expressed upon successful infection of the target cell by the reporter viruses carrying the intron spliced genomic viral RNA and its reverse transcription . By this , a reporter signal can only result from an infection by free virus or cell-cell transmission but not from cell fusion events or non-productive virus uptake . While the NLinGluc reporter virus has been designed to distinguish cell-cell transmission from fusion , it needs to be combined with measures that restrict free virus transmission during co-culture if genuine cell-cell transmission needs to be assessed . As recently described [25] , we made use of the fact that for many HIV strains efficient infection of transformed cell lines by free viruses but not cell-cell transmission requires the addition of polycations such as DEAE [16 , 25 , 59–62] . Polycation dependency is most prominent for R5 viruses with a low positive V3 net charge . These viruses have a decreased capacity to overcome the charge repulsion of negatively charged membrane proteins of the viral and cellular membrane which appears to be infection limiting in transformed cell lines [60–63] . Importantly , as previously shown , neutralization efficacy is not affected by the presence or absence of polycations [60–63] . Since T cell to T cell transfer is considered of high relevance for HIV-1 spread in vivo [64] , we chose the T cell line , A3 . 01-CCR5 , as target cells for free virus and cell-cell transmission . In a first step , we verified that free virus infection with Env-pseudotyped NLinGluc reporter viruses , like the commonly used reporter virus NLlucAM [65] which encodes for firefly luciferase , requires DEAE for efficient infection of A3 . 01-CCR5 T cells ( Fig 1A , S1 Fig ) . This ensures that by omission of DEAE during co-culture , we only allow for HIV cell-cell transmission but not free virus infection to occur [25] . While 293-T donor cells transfected with Env-pseudotyped NLinGluc reporter viruses generated no Gaussia luciferase signal ( Fig 1B ) , co-culture of these donor cells with A3 . 01-CCR5 target cells in absence of DEAE yielded Gaussia luciferase activity , indicating cell-cell transmission ( Fig 1B ) . To further verify that the measured reporter activity was indeed due to genuine cell-cell transmission and not influenced by free virus infection or cell fusion , we studied the effect of reverse transcriptase ( RT ) and protease inhibition on cell-cell and free virus transmission . Free virus spread was vulnerable to RT inhibition but not protease inhibition , as virions in the harvested virus stocks are fully matured . In contrast , cell-cell transmission proved fully sensitive to both RT and protease inhibitors as described , excluding a reporter gene transfer by fusion [47 , 51] ( Fig 1C ) . The choice of donor and target cells for cell-cell transmission assays suitable for neutralization screening is not trivial as the selected system should provide the most physiologically relevant information , most ideally by using primary T cells , but at the same time guarantee high intra- and inter-assay reproducibility , high throughput capacity and cost effectiveness . To verify our choice of A3 . 01-CCR5 cells as target cells , we first compared neutralization of free virus infection of A3 . 01-CCR5 cells using the JR-FL-pseudotyped NLinGluc ( Gaussia ) and NLlucAM ( firefly ) reporter viruses . Infection with both virus constructs yielded identical neutralization profiles for all bnAbs tested ( Fig 1D , S2 Fig ) and thus allowed us to use the less expensive and more stable luminescence signal emitting firefly reporter readout for the assessment of free virus neutralization on A3 . 01-CCR5 . The cell-cell transmission system we employed here , utilizes 293-T cells as donor cells as these cells , unlike T cells , can be transfected with high efficiency and are thus commonly used to produce HIV-1 pseudoviruses . As therefore in our cell-cell transmission setup only one of the partners , the A3 . 01-CCR5 target cells , are T cells , we sought to verify if the transmission from 293-T to A3 . 01-CCR5 is comparable to genuine T cell to T cell transmission and yields similar neutralization patterns . To this end , we studied HIV-1 inhibition by a selection of bnAbs during free virus infection of PBMC and cell-cell transmission of PBMC to PBMC , as PBMC are considered the most physiological relevant in vitro cell system available . Of note , assessment of free virus inhibition using a range of Env-pseudotyped NLlucAM ( firefly ) reporter viruses on PBMC and A3 . 01-CCR5 yielded almost identical inhibitory patterns , indicating that A3 . 01-CCR5 cells are indeed a valid substitute for PBMC ( Fig 1E , S3 Fig ) . Inhibition of PBMC to PBMC transmission was studied using JR-FL and JR-CSF infected PBMC as donor cells and rhTRIM5α overexpressing PBMC as target cells in which free virus infectivity is restricted [25] . Infection in the PBMC co-cultures was assessed by monitoring Gag protein transfer from infected PBMC to rhTRIM5α expressing PBMC target cells by flow cytometry as previously described [25] . Most notably , the PBMC-PBMC and the 293-T-A3 . 01-CCR5 neutralization assays yielded identical inhibitory profiles ( Fig 1F , S4 Fig ) , confirming that the 293-T to A3 . 01-CCR5 transmission assay has the capacity to capture the essential components of PBMC-PBMC transmission in respect to antibody accessibility and activity during cell-cell transmission . Based on this , we concluded that A3 . 01-CCR5 T cells , albeit no primary T cells , are valid target cells for the cell-cell transmission studies . A main goal of our study was to assess the efficacy of bnAbs during cell-cell compared to free virus transmission to derive insights into the magnitude of the neutralization activity losses during cell-cell transmission , as this might be important information when using bnAbs as therapeutics or components of vaccine induced immunity . To assess the breadth of bnAb activity during cell-cell transmission , we sought to test a range of genetically different virus strains as so far most cell-cell neutralization studies have included only a comparatively small number of viruses , mostly from subtype B [25–27] . Our virus panel consisted of 11 viruses from subtype A , B , and C and from different disease stages ( transmitted founder ( T/F ) viruses , acute and chronic infection; S1 Table ) . BnAbs that were probed for activity in free virus and cell-cell transmission included bnAbs directed to the CD4bs ( b12 , VRC01 , NIH45-46 , PGV04 , 3BNC117 ) , the V3-glycan region ( PGTs 121 , 125 , 128 , 135 ) , V1V2-glycan dependent PGT145 , PG9 , PG16 ) and the MPER domain ( 2F5 , 10E8 , 4E10 ) ( S2 Table ) . When we assessed the activity loss during cell-cell transmission for the seven subtype B viruses included in our panel , we observed a strong decrease in neutralization capacity during cell-cell transmission for the majority of bnAbs ( Fig 2A , S5A–S5P Fig and S6A Fig ) confirming earlier findings [25–27] . The loss in bnAb activity occurred independent of the epitope region targeted or the specific bnAb analyzed and its extent varied substantially for individual bnAbs against the divergent viruses . Further , the decrease in neutralization activity was not specific to the virus strains tested . As exemplified for strain PVO . 4 , the highest loss in activity was seen for bnAb 3BNC117 with a 34-fold increased 50% inhibitory concentration ( IC50 ) during cell-cell transmission whereas bnAb 2G12 retained its activity at higher levels with only a 7 . 2-fold loss in activity ( Fig 2A , S3 Table ) . Across viruses and bnAbs tested , we observed an overall diverse pattern . Intriguingly , some rare bnAb-virus pairings showed no reduction in activity during cell-cell transmission as exemplified by PG16 neutralization of strain THRO ( Fig 2A ) . This is particular notable as other bnAbs , the CD4bs ( b12 , VRC01 , NIH45-46 and 3BNC117 ) probed in parallel against the same isolate , showed a pronounced decrease in activity during cell-cell transmission ( 100–2000-fold increased IC50; Fig 2A , S3 Table ) . Overall , the individual potencies of the bnAbs against the probed subtype B viruses varied substantially in both , the free virus and the cell-cell transmission setting . To probe if the observed decrease in neutralization activity is subtype-specific or a general feature of HIV-1 cell-cell transmission , we expanded our analysis to investigate viruses from subtype A ( BG505 ) and Subtype C ( ZM53 , ZM109 , ZM214 ) ( Fig 2B , 2C and S6A Fig ) . Free virus and cell-cell transmission of these viruses proved to follow the same pattern as we had observed for subtype B viruses with bnAb potency against individual viruses varying in a wide range for both transmission routes ( Fig 2B and 2C and S6A Fig ) . To compare the extent of change in inhibitory activity during cell-cell transmission , we determined the fold changes in IC50 ( Fig 2D , S6B and S6C Fig ) . Across bnAbs , the median IC50 against all strains was elevated for cell-cell compared to free virus transmission ( Fig 2B and 2C , S3 Table ) , resulting in a 4 . 5-fold loss in activity for bnAb 2G12 to 256-fold loss for bnAb b12 with an overall median fold reduction of 22 . Although few in numbers , some bnAb-virus combinations yielded activity in cell-cell transmission inhibition that closely matched their potency against free virus ( less than 5-fold reduced; Fig 2D ) . These were 2G12 ( with strains JR-FL and JR-CSF ) , PGT121 ( with strains REJO and BG505 ) , PGT128 ( with strain BG505 ) , PGT145 ( with strains BG505 and ZM109 ) , 10E8 ( with strain ZM53 ) and 4E10 ( with strains JR-CSF and ZM53 ) . Most intriguingly , PGT145 neutralized ZM53 better during cell-cell transmission ( 12-fold lower IC50 ) . A similar trend was seen for PG16 against strains DH123 and THRO although there the increases in activity during cell-cell transmission were smaller . Apart from the two MPER bnAbs , all other bnAbs which occasionally yielded identical activity in free virus and cell-cell neutralization had glycan-dependent epitopes centering in or around the V3 loop ( PGT121 , PG128 , 2G12 ) or V1V2 loop ( PGT145 , PG16 ) . In contrast , among CD4bs-directed bnAbs for none of the probed virus strains identical activity during cell-cell transmission was obtained ( S5A–S5E Fig ) . BnAb PGV04 retained neutralization activity during cell-cell transmission to the highest extent , while b12 , the weakest and least broad of the probed CD4bs bnAbs , completely lost activity against several viruses ( S5A Fig ) . Of note , for all epitope classes , the overall median activity loss of the probed bnAbs during cell-cell transmission was in a similar range , irrespective of the virus subtype ( S6B Fig ) . Individual virus strains varied , however , substantially to the extent by which cell-cell transmission affected the inhibitory potentials of bnAbs ( S6C Fig ) . To determine if the reported losses in neutralization activity of cell-cell spread can be predicted by the bnAbs’ performance during free virus inhibition or if they need to be determined individually , we assessed the interdependence of potency in free virus and cell-cell neutralization and the extent of the activity loss inflicted by cell-cell transmission ( Fig 3A and 3B ) . Inhibitory concentrations obtained for free virus and cell-cell neutralization proved to be linked ( Fig 3A ) . Unexpectedly , efficacy in free virus inhibition and the extent of activity loss during cell-cell transmission showed an inverse correlation for V1V2- and V3-directed and to a lesser extent also for MPER-directed bnAbs , but not for bnAbs targeting the CD4bs ( Fig 3B ) . Thus , with the exception of CD4bs-directed bnAbs , a higher loss in neutralization activity during cell-cell transmission was more frequently observed when bnAb activity against free virus transmission was very high , suggesting that features which steer the bnAbs’ ability to potently inhibit free virus spread are not equally decisive in cell-cell transmission . In fact , the opposite appeared to be true for some bnAb-virus combinations: The bnAbs that comparably neutralized a virus strain during both transmission modes were in most cases relatively weak inhibitors of these strains ( 2G12 for JR-FL and JR-CSF , PGT121 for REJO and BG505 , PGT128 for BG505 , PGT145 for ZM53 , PG16 for DH123 and THRO , 10E8 for ZM53 , 4E10 for JR-CSF and ZM53 ) . High potency against free virus activity was only in two cases ( PGT145 for BG505 and ZM109 ) paired with low loss of activity during cell-cell transmission ( Fig 2B and 2D , S3 Table , S5J Fig ) . Collectively , this suggests that antibody features that steer high potency against free virus spread are not the driving factors determining the activity during cell-cell transmission . In summary , the activities of bnAbs during inhibition of free virus and cell-cell transmission proved highly diverse and varied in a strain- and antibody-dependent manner ( Fig 2A–2D , S3 Table , S5A–S5P Fig and S6 Fig ) . None of the 16 probed bnAbs retained the free virus inhibitory activity during cell-cell transmission across all viruses tested , highlighting that the activity loss during cell-cell transmission needs to be considered when determining bnAb breadth and exploring in vivo infective doses for bnAb use in therapy and prevention . Amongst the panel of bnAbs probed against the 11 viruses , the CD4bs bnAb PGV04 proved the most consistent and combined high breadth ( 10 of 11 virus strains inhibited in both transmission modes ) , potency ( median IC50 of 0 . 039 and 0 . 645 µg/ml for free and cell-cell transmission respectively ) and a relatively low loss in activity during cell-cell transmission ( median 13-fold reduction , IC50 values ranging from 7 . 5 µg/ml for BG505 to 33 µg/ml for SF162 ) ( Fig 2B–2D , S3 Table , S5C Fig ) . However , several bnAbs performed in a comparable range: 3BNC117 was closest to PGV04 in performance but displayed a somewhat wider range in activity loss during cell-cell transmission ( median 34-fold IC50 reduction , range 6 . 6 for ZM109 to 237 for THRO ) ( Fig 2B–2D , S3 Table , S5E Fig ) . The loop and glycan specific , gp120-directed bnAbs probed had a generally lower breadth and hence could not be assessed for the entire virus panel . Amongst these , PGT121 combined best breadth , potency and a low loss during cell-cell transmission . This was also true for the MPER-specific bnAb 10E8 which was very consistent in its activity during both transmission modes , neutralizing all 11 probed virus isolates . To validate the results obtained in the cell-cell transmission studies , we included the gp41-targeting fusion inhibitor T-20 ( S2 Table ) as a control in all experiments as it has high activity across different genetic subtypes of HIV-1 [66] and our previous work [25] indicated that T-20 has a high capacity to preserve its activity during free virus and cell-cell neutralization . Since T-20 , like neutralizing antibodies , targets the virus envelope , we reasoned that T-20 may provide a valuable reference for neutralizing activity in free virus and cell-cell transmission . T-20 proved to be the most successful agent in preserving activity during cell-cell transmission across all virus strains tested ( Fig 4 , S3 Table ) . Reduction in activity during cell-cell transmission was less than 10-fold for all 11 probed viruses including strains THRO and ZM214 ( median fold reduction for T-20 in cell-cell transmission of 1 . 6 and 5 . 3 respectively ) which were the most resistant viruses to bnAb neutralization during cell-cell transmission ( S3 Table ) . Thus far , a similarly retained potency during cell-cell transmission has only been reported for cell-directed inhibitors targeting CD4 and co-receptors but not for Env-specific agents [25 , 27] . Considering that T-20 has been in clinical use [67] and extensive in vivo efficacy data are available , our observations highlight that T-20 could be considered as valuable control in future experiments which aim to dissect if a preserved cell-cell transmission activity is required for the in vivo efficacy of neutralizing antibodies and entry inhibitors . A consequence of the activity loss of neutralizing antibodies during cell-cell transmission that needs to be considered is that escape to neutralization may occur more rapidly than via free virus transmission as postulated for reverse transcriptase inhibitors [46] . To determine how much more likely it is that an antibody neutralization resistant variant arose during infection via cell-cell transmission than via free virus spread , we translated this question into a mathematical model ( Materials and Methods ) . To this end , we first derived the mutant occurrence probability for the two pathways , i . e . the probability that an antibody resistant mutant arises depending on the infection pathway . This expression incorporates the measured IC50 and slope values ( S3 Table and S4 Table ) of the bnAb-virus combinations tested in free virus and cell-cell neutralization ( S5A–S5P Fig ) . In a second step , we divided the mutant occurrence probability for cell-cell transmission by the one for free virus transmission . By doing so , the probability that a point mutation arises during reverse transcription , which cannot be experimentally determined , cancels out . This allowed us to determine how much more likely it is that a given mutant arose during cell-cell compared to free virus transmission . Fig 5 depicts the ratios of the mutant occurrence probabilities in dependence of the bnAb concentration . We found that irrespective of the bnAbs’ individual neutralization capacities across a wide range of bnAb concentrations , it is more likely that a neutralization resistance conferring mutation evolves via the cell-cell transmission route ( indicated by the light grey shaded area in Fig 5 ) than via the free virus route ( dark grey shaded area ) . In certain cases ( e . g . for bnAb VRC01 and BG505 , PGV04 and BG505 and ZM214 , PGT121 and ZM214 and PGT125 and JR-FL ) probabilities for mutants to occur via cell-cell transmission were up to 50 times higher . For the analysis , we assumed that the per site mutation probability is the same for cells infected during the cell-cell and free virus pathway . However , as cells infected via cell-cell transmission appear to be more frequently infected with more than one virus [37 , 38 , 46] , the total number of possible mutations that can occur is higher and recombination of viral RNA from different strains is more likely , which would result in higher diversity . In our experimental setup , this should not be a driving factor as we utilized Env-pseudotyped molecular clones . This may , however , play a role in vivo when genetically diverse virus populations are present . We thus tested which influence higher mutation rates in cells infected via cell-cell spread could have on mutant generation . We found that higher mutation rates and diversity in cell-cell transmission would further increase the probability that a mutation arose via cell-cell in comparison to free virus transmission ( S7 Fig ) . Together , these findings support the hypothesis that cell-cell transmission of virus particles can serve as a rescue route from neutralizing antibodies due to the reduced sensitivity of this pathway to neutralization and additionally by fostering the evolution of resistant strains . Depending on the accessibility of their epitopes during entry , bnAbs differ in their ability to neutralize HIV prior to or post-CD4 engagement [25 , 68 , 69] . As discussed previously , the capacity to access the virus post-CD4 interaction may be advantageous for antibodies , particularly in the setting of cell-cell transmission as this may elongate the bnAbs’ window of action [25] . To probe the capacity of bnAbs to interfere with HIV entry pre- and post-CD4 attachment , we first assessed the activity of all 16 bnAbs and T-20 during free virus neutralization of six virus strains from subtypes A , B and C when present during the entire infection period ( Fig 6A ) or solely added post-CD4 attachment ( Fig 6B ) . To measure total neutralization activity ( cumulative bnAb action before and after CD4 attachment; Fig 6A ) , virus and bnAbs were preincubated before the inocula were spinoculated onto A3 . 01-CCR5 target cells to synchronize infection . BnAbs were thus present during the entire infection process and could act both , before and after CD4 attachment . Concentrations of bnAbs were chosen to yield maximal inhibition for all viruses in these experiments ( S5 Table ) . As we previously showed , attachment to cells during spinoculation is highly CD4-driven [25] , hence allowing a synchronized initiation of infection in our experimental design , starting from CD4-bound virions . To assess their post-attachment inhibition potential , bnAbs were added after spinoculation of viruses onto target cells in identical concentrations ( Fig 6B , S5 Table ) . The inhibitory activity measured post-attachment was expressed relative to the total activity of the bnAb over the entire infection period ( Fig 6A ) which was set to 100% ( Fig 6C ) . In line with the enhanced accessibility of the MPER domain following receptor binding and rearrangements [70 , 71] , the MPER-directed bnAbs showed high activities when added post-attachment , yielding 100% inhibition in the majority of cases . The lowest post-attachment activities observed for MPER bnAbs were still comparatively high with levels above 67% observed against JR-FL ( 2F5 79% and 4E10 74% ) and ZM109 ( 10E8 72% and 4E10 67% ) . Consistent with their epitope specificity , CD4bs-directed bnAbs , on the other hand , only reached post-attachment inhibition capacities which were below 50% for most virus strains . Strain ZM53 was the only exception , as most CD4bs-directed bnAbs retained around 60% of their activity . Post-attachment activity of V3 loop-directed bnAbs was comparable to CD4bs-directed bnAbs and for most bnAb-virus combinations below 50% . The V1V2-directed bnAbs displayed a higher strain-dependency: ZM53 , which retained high sensitivity to CD4bs-directed bnAbs at the post-attachment stage , remained fully sensitive to PGT145 and also highly sensitive to PG9 ( 73% ) and PG16 ( 76% ) ( Fig 6C ) . Overall , the variability in post-attachment activity was higher than we previously anticipated , judging from the analysis of smaller antibody and virus panels [25] . Comparing total and post-attachment activity additionally provided insight into the pre-attachment activity of the bnAbs , which was measured as the activity against the unbound virus ( during pre-incubation ) and the initial steps of CD4 engagement ( during spinoculation ) in our experimental setup ( Fig 7A ) . While a low post-attachment activity implies that a bnAb preferentially acts prior to CD4 engagement , the reverse is not necessarily the case as action during both stages , pre- and post-CD4 attachment may occur . Thus , to investigate the pre-attachment activity more precisely , we assessed the potential of bnAbs to neutralize when solely present before CD4 binding has been completed ( Fig 7A ) . To this end , we first pre-incubated virus and bnAbs to allow for bnAb binding to the native spike . Concentrations of bnAbs were chosen to yield maximal inhibition ( S5 Table ) . During subsequent spinoculation of the inoculum onto the target cells , virus binding to CD4 was initiated [25] allowing bnAb binding to epitopes exposed in the CD4-bound stage . Spinoculation was performed at room temperature to permit trimer binding to CD4 while arresting further structural rearrangements and fusion . The excess bnAbs and unbound virus were washed off and the infected target cells cultivated in absence of bnAbs . Efficacy in blocking virus entry was compared to controls in which bnAbs remained in the culture after spinoculation and therefore provided reference values for total neutralization activity covering bnAb action during the entire infection period ( Fig 7A ) . Pre-attachment neutralization efficacy monitored in our assay set-up provides a composite information on the capacity of an antibody to interact with native trimers on free virus and Env spikes that have already bound to CD4 but have not fully undergone Env rearrangements . Activity during these steps can be steered by several factors: i ) the capacity of bnAbs to bind the native trimer with high affinity [72] , ii ) their efficacy to establish irreversible binding and neutralization ( which is not lost upon washout of the bnAb ) [69] , iii ) their capacity to block attachment to CD4 or an early stage of CD4 engagement , and iv ) the capacity of the bnAb to induce a functional decay of the Env trimer [69] . Overall , bnAbs proved highly competent in neutralizing HIV during the pre-attachment phase . Low pre-attachment activity was predominantly observed with the subtype C strains ZM53 , ZM109 and ZM214 , particularly for bnAb PGT121 , the V1V2- and the MPER-directed bnAbs and surprisingly with ZM214 also for T-20 ( Fig 7B ) . CD4bs-directed bnAbs had the highest capacity to neutralize prior to CD4 attachment . Loss of CD4bs bnAb activity upon antibody removal was only observed with subtype C virus strains and to a relatively low extent ( up to 25% ) . In general , high potencies of bnAbs against free virus transmission were linked with high pre-attachment activities while post-attachment activities ( Fig 6C ) and pre-attachment activities proved to be inversely linked ( Fig 7C and S8 Fig ) , highlighting the importance of the capacity to bind the virus pre-CD4 engagement for bnAb efficacy against free virus spread . Of note , inhibitory concentrations for free virus and cell-cell transmission showed a high correlation , indicating that key aspects of the bnAb interaction with the viral Env during both entry processes must nevertheless be shared . Interestingly , the loss in neutralization activity during cell-cell transmission correlated with the pre-attachment activity , highlighting that bnAbs which preferentially act before CD4 engagement are less potent during cell-cell transmission confirming earlier findings [25] . In sum , our analyses suggest that the superior activity of bnAbs against free virus transmission is at least partially steered by their higher potential to access the virus prior to CD4 engagement . This feature seems to be less important for the efficacy in cell-cell neutralization as also highlighted by the activity of MPER-targeting bnAbs and T-20 during cell-cell transmission . The V1V2-directed bnAbs proved to be less active during the pre-CD4 attachment phase of the subtype C viruses ZM53 and ZM109 . Larger antibody and virus panels will be needed to define if this decreased neutralization activity of V1V2 bnAbs is a common feature of subtype C viruses or of the respective bnAbs probed . To obtain further insight , we tested two additional , clonally related V1V2-directed bnAbs , CAP256-VRC26 . 08 and CAP256-VRC26 . 09 , which were isolated from a subtype C superinfected donor [73 , 74] . The two bnAbs were not able to neutralize the subtype B viruses in our panel but inhibited strains BG505 and BG505 N332 ( subtype A ) as well as ZM53 and ZM214 ( both subtype C ) . Quite strikingly , the CAP256 bnAbs retained their activity during cell-cell neutralization at high levels for BG505 , BG505 N332 and ZM53 ( 1 . 4–3 . 9-fold , 3 . 5-7-fold and 8 . 7-10-fold over free virus IC50 ) ( Fig 8A–8C ) . The most surprising results were obtained for strain ZM214 , which was the most neutralization resistant virus during cell-cell transmission in our panel . CAP256-VRC26 . 08 and CAP256-VRC26 . 09 proved to be considerably more potent against ZM214 in cell-cell transmission than free virus transmission ( 11-fold and 24-fold , respectively ) . Of note , ZM214 was not sensitive to any of the other V1V2-directed bnAbs tested ( S5J–S5L Fig ) . Interestingly , the only other case where we found a bnAb to have a clearly superior activity during cell-cell transmission was also observed with a V1V2-directed bnAb and a subtype C virus , namely PGT145 against ZM53 ( S5J Fig ) . In accordance with earlier observations ( Fig 3 ) , the capacity of the CAP256 bnAbs to block cell-cell and free virus transmission at identical levels was not linked to a particularly high inhibitory potency against this strain . In fact , neutralization of ZM214 required the highest CAP256 bnAb concentrations of all four viruses tested ( Fig 8B , S3 Table ) . The subtype C strain ZM53 was the virus strain with the lowest IC50 for free virus and cell-cell inhibition ( Fig 8A and 8B ) . In line with the other V1V2 bnAbs , we observed relatively high levels of post-attachment inhibition activity , which was highest for ZM214 with 80% and 77% for CAP256-VRC26 . 08 and CAP256-VRC26 . 09 , respectively ( Fig 8D ) . For ZM53 , the virus strain most sensitive to CAP256 bnAb free virus neutralization , also the highest pre-attachment activities were observed ( Fig 8E ) . In contrast , neutralization of ZM214 , the least sensitive of the four probed strains in free virus inhibition , was completely abolished during the pre-attachment phase , again indicating that pre-attachment activity is linked to high bnAb potency . In sum , CAP256-VRC26 . 08 and CAP256-VRC26 . 09 differed in their properties from the other probed V1V2-directed bnAbs , PGT145 , PG9 and PG16 , which all failed to neutralize ZM214 and lacked pre-attachment activity against ZM53 .
The progress in the development and characterization of highly potent bnAbs in the recent years has brought new hope for their use in therapeutic settings and the development of vaccine regimens eliciting alike responses [1] . Considering the increasing wealth of identified bnAbs , it will be important to focus these developments on the most promising candidates . This affords extensive in vitro evaluation and efficacy testing in animal models , the classical rhesus macaque SHIV model and increasingly also in humanized mouse models [7–14] . While efficacy testing in animal models is indispensable before clinical use of bnAbs , these models can currently not cover all aspects of human HIV-1 infection . Setting aside the issues of differences in immune systems , disease patterns and virus loads in these models , an inherent limitation are the virus strains used in animal experiments , as only few strains are available for in vivo application , especially in the case of the rhesus macaque model . It has been consistently observed that the in vivo efficacy of neutralizing antibodies is considerably lower than their activity in vitro [12 , 13 , 17–19] . Amongst other factors , the decrease in neutralization activity during cell-cell transmission has been discussed as contributing factor for reduced in vivo efficacy [24–27] . Since breadth of bnAbs has so far only been assessed by free virus neutralization in vitro , differential activity loss of bnAbs in vivo due to a reduced activity in cell-cell transmission may not be fully captured by efficacy testing in animal models due to the limited number of virus strains that can be applied . Hence , if the capacity to potently block cell-cell transmission proves important for in vivo efficacy , tailored in vitro analyses are needed to dissect which bnAbs have the required activity . We therefore focused our study on defining the characteristics of free virus and cell-cell neutralization by bnAbs in vitro to provide additional functional information aiding the future selection of bnAbs for therapeutic and vaccine development . We concentrated our analysis on four criteria of neutralization activity: ( i ) We probed the capacity of bnAbs to inhibit HIV-1 cell-cell transmission at comparable breadth and potency as free virus spread , considering that bnAbs that preserve high activity during cell-cell transmission across divergent strains can be viewed as attractive leads for therapeutic use and vaccine development . ( ii ) We investigated the probability of bnAb resistance to occur during free virus and cell-cell transmission . ( iii ) To derive more insight into the mechanisms of neutralization and the potency in the two transmission modes , we investigated the capacity of bnAbs to neutralize post-CD4 engagement . This was prompted by our previous work [25] , which indicated that the capacity of antibodies to access the viral envelope post-CD4 engagement may be beneficial for maintaining activity during cell-cell transmission . Considering that trimer binding to CD4 initiates the formation of the virological synapse [31] , antibodies that neutralize post-CD4 engagement could indeed benefit from a longer time window of action . ( iv ) Likewise , the capacity of bnAbs to act prior to CD4 engagement may be decisive for their activity . Ideally , a potent neutralizing antibody can be envisaged to have a high on-rate and low off-rate in binding the native Env spike before receptor binding , resulting in virtually irreversibly Env binding , obstruction of receptor binding or viral decay leading to irreversible neutralization [69 , 72 , 75] . Reversibility of binding and thereby neutralization can occur if the antibody has a high off-rate in binding the viral Env protein [69 , 72 , 76] . Rapid clearance before opsonized virions regain their infectivity upon antibody detachment may thus be crucial when binding and neutralization are reversible and potentially could require higher antibody doses that ascertain optimal triggering of effector functions and clearance [21–23] . Due to the complexity of the assays , neutralization of cell-cell transmission has previously only been assessed for a relatively small number of antibodies and HIV-1 isolates [24–30] . An improved pseudovirus-based assay system to study cell-cell transmission allowed us to screen the cell-cell inhibition capacity of a wide range of virus-bnAb combinations . In sum , we probed the sensitivity of 11 virus strains from subtypes A , B and C to a panel of 16 bnAbs targeting the CD4bs , V3 glycan region , the V1V2 loop and the MPER domain during free virus and cell-cell transmission . In accordance with previous reports , we observed an overall decreased neutralization activity of bnAbs to inhibit cell-cell compared to free virus spread with few exceptions . Interestingly however , activities against both transmission routes varied substantially between the different virus strains tested and we could not identify a single bnAb that equally blocked free virus and cell-cell transmission over a broad range of HIV-1 strains . Many of the probed bnAbs performed in a similar range . Against the viruses investigated , the most potently and consistently neutralizing bnAbs for both pathways were the CD4bs-directed bnAb PGV04 and the MPER-directed bnAb 10E8 . Amongst loop-specific bnAbs , which all had genuinely lower breadth , PGT121 was the most effective in combining breadth , potency and a low loss during cell-cell transmission . Intriguingly , while inhibitory concentrations required for free virus and cell-cell inhibition correlated , a high potency against free virus spread did not ensure a lower loss in activity during cell-cell transmission ( Figs 3 , 7C and S8 Fig ) . On the contrary , we identified a negative correlation between the inhibitory concentrations required for blocking free virus spread ( IC50 ) and the loss in activity during cell-cell transmission ( fold change of the corresponding IC50 values ) . We frequently observed that bnAbs that already required high concentrations to neutralize free virus spread maintained similar neutralization activities against cell-cell transmission . Collectively , this suggests that properties that render a bnAb highly potent in neutralizing free virus spread are not equally relevant for inhibiting cell-cell transmission . It has been recently suggested that virus particles in free virus and cell-cell transmission may differ in their maturation status . While free virus might be largely matured , cell-cell transmitted virus may still be in an immature form which potentially could also change the trimer conformation through differential gp41 cytoplasmic tail Gag interactions , also affecting antibody binding [49] . Additionally , immediately upon virus release , virions may carry more intact trimers which rapidly decay over time , adding to the intrinsic differences in trimer content observed across isolates [77] . A factor that likely steers free virus neutralization is the on-rate of the antibody for binding the native trimer before receptor engagement . If the epitope is optimally exposed on the free virus , the antibody can bind rapidly , with high affinity ( and ideally irreversibly ) , resulting in quick and potent neutralization of the free virus . If firm bnAb binding to the native trimer requires substantial conformational rearrangements ( e . g . induced by the bnAb binding itself or CD4- and co-receptor engagement ) , antibody binding and neutralization kinetics of free virus transmission will be slower and likely less efficient . Hence in the latter case , neutralization of free virus and cell-cell transmission may both occur predominantly after trimer binding to CD4 [31] . In summary , we observed that despite their inherent high activity against free viruses , bnAbs can substantially differ in their capacity to block cell-cell transmission , highlighting that the potency in inhibiting free virus spread does not allow to reliably draw conclusions on the neutralization performance during cell-cell transmission . In our study , decreased sensitivity to neutralization during cell-cell transmission proved not to be a characteristic of specific viruses or bnAbs , confirming the need to include a range of viruses when assessing bnAb activity during cell-cell transmission . This was probably best evidenced by the divergent patterns we observed for the three subtype C viruses in our panel . While inhibition of ZM53 and ZM109 across the probed bnAbs required comparatively low IC50 for free virus inhibition , paired with relatively modest activity losses during cell-cell transmission , we observed large differences in capacity for free virus and cell-cell neutralization of ZM214 . The only Env-directed inhibitor that we identified to exert a widely preserved activity in both transmission modes was the fusion inhibitor T-20 [78–80] . Strikingly , while neutralization capacities of bnAbs in free virus and cell-cell transmission varied substantially , the gp41-directed inhibitor T-20 retained its potency in both transmission modes across viruses at comparative levels . Defining the precise properties of T-20 that preserve its activity during cell-cell transmission in future studies will hence be of interest as this may provide guidance on what mechanistic features are required for high efficacy in cell-cell neutralization and potentially also in vivo activity . Of note , as T-20 targets the six-helix bundle formation , a step in virus entry post-receptor engagement , kinetics of T-20 inhibition could indeed be similar for free virus and cell-cell transmission . While T-20 ( 4 . 5 kDa ) is considerably smaller than antibodies ( 150kDa ) , its size alone may not be decisive for its activity during cell-cell transmission . In our previous studies , we demonstrated that both , CD4-directed neutralizing antibodies and the CD4-peptide mimetic CDM47 ( 2 . 9 kDa; [81] ) equally loose potency during cell-cell inhibition [25] . With few exceptions , neutralization of free virus by CD4bs- and V3-directed bnAbs proved to be highly potent during the pre-attachment phase while neutralization by V1V2- and MPER-targeting bnAbs depended also on post-attachment activity for several virus strains . A decreased efficacy during the pre-attachment phase could be caused by a failure of the bnAbs to bind the native trimer , to interfere with attachment to CD4 or to irreversibly neutralize the virus . If virus neutralization remains reversible for an extended time period , this could potentially cause a reduced in vivo efficacy if opsonized viral particles are not rapidly cleared [21–23] , highlighting the need to further evaluate how reversibility of neutralization affects the in vivo efficacy of bnAbs . Interestingly , we found that high potency against free virus was linked with high pre-attachment activity ( Fig 7C and S8 Fig ) while low pre-attachment activity was linked with activity post-CD4 attachment and lower losses during cell-cell transmission . Although it is not possible to generalize our findings for all antibody-virus combinations studied , the reactivity patterns we observed tended to fall into two distinct groups ( Fig 9 ) : BnAbs with high potency against free virus spread show a high activity against the virus prior to CD4-binding , indicating optimal access of the epitope on the native trimer and high affinity binding . BnAbs with lower neutralization activity against free virus spread lack pre-attachment activity , show potent neutralization post-CD4 attachment and retain activity during cell-cell transmission , indicating suboptimal binding to the native trimer and improved access to the epitope post-CD4 engagement . To date , the relative impact of free virus and cell-cell spread in vivo has not been determined . If cell-cell transmission should prove to contribute to viral spread in vivo , the decreased sensitivity of this transmission mode to antibody neutralization may indeed be of concern as it may provide viruses with a possibility to better tolerate antibody pressure and acquire resistance mutations . Here , we investigated if the chances that resistance mutations occur indeed differ for free virus and cell-cell transmission . As our mathematical analysis revealed , cell-cell transmission proved to be substantially more prone to give rise to escape mutants than free virus transmission . This highlights the importance of controlling virus replication via the cell-cell transmission pathway even if the contribution of this transmission mode should proof to occur to a lesser extent than free virus spread in infected individuals . The selection of bnAbs that are developed further for clinical use need to be carefully designed to ascertain in vivo efficacy to the best of our knowledge . Most likely , the increasing data on currently ongoing bnAb efficiency in in vivo testing will give us an improved picture on which features determine the in vivo efficacy of bnAbs . These will probably not only involve potency and breadth against free viruses but also favorable pharmacokinetics of bnAbs [11 , 26] and their capacity to elicit effector functions and to induce rapid virus clearance [21–23] . As highlighted by our analyses , functional characteristics such as the efficacy to neutralize virus during cell-cell transmission and the different stages of the entry process are important features that shape the antibodies’ activity , steer their potency and their potential for mutant selection . Hence , these characteristics should be considered in future when evaluating bnAbs for clinical use . The high variability of bnAb activity across genetically divergent viruses and their varying capacities to block both transmission modes and to differentially neutralize pre- and post CD4 attachment further highlights the need to create powerful combination of bnAbs , either by multi-component vaccines or antibody cocktails [82] in passive immunization to ensure that all mechanistic features required for effective virus control are represented .
Peripheral blood mononuclear cells ( PBMC ) were purified from buffy coats from anonymous blood donations from healthy individuals obtained by the Zurich Blood Transfusion Service ( http://www . zhbsd . ch/ ) under a protocol approved by the local ethics committee . Properties and sources of antibodies and inhibitors used in this study are listed in S2 Table . We thank D . Burton , J . Mascola , M . Nussenzweig , M . Connors , P . Moore and L . Morris for providing antibodies and antibody expression plasmids for this study either directly or via the NIH AIDS Research and Reference Reagent Program ( NIH ARP ) . 293-T cells were obtained from the American Type Culture Collection ( ATCC ) , TZM-bl [16] were obtained from the NIH ARP . A3 . 01-CCR5 cells were described previously [25 , 83] . 293-T and TZM-bl cells were cultivated in DMEM with 10% heat inactivated FCS and 1% Penicillin/Streptomycin . A3 . 01-CCR5 cells were maintained in RPMI with 10% heat inactivated FCS and 1% Penicillin/Streptomycin . Stimulated PBMC were isolated from healthy donors as described [61] and cultivated in RPMI with 10% heat inactivated FCS , 1% Penicillin/Streptomycin and 100 units/ml ( U ) IL-2 . Properties and sources of plasmids encoding the envelope of strains BG505 ( subtype A ) , JR-FL , JR-CSF , SF162 , DH123 , PVO . 4 , REJO , THRO ( all subtype B ) , ZM214 , ZM109 , ZM53 ( all subtype C ) are listed in S1 Table . The T332N envelope point mutation in BG505 was generated by site-directed mutagenesis ( Agilent QuikChange II XL ) according to the manufacturer instructions . The point mutant envelope ( denoted as BG505 N332 ) was sequenced by in-house Sanger sequencing to confirm presence of the desired mutations and absence of unintended sequence changes . For the production of single-round replicating pseudovirus stocks , 293-T cells were transfected with the respective viral backbones and env expression plasmids as described [61] . The following backbone constructs were used: The luciferase reporter HIV-1 pseudotyped vector pNLlucAM [84] and the NL4-3 based pseudotyped vector with integrated inGluc reporter construct ( NLinGluc [56]; a gift from Dr . M . Johnson ) . Infectivity of reporter viruses was quantified by titration of virus containing supernatants on 1*104 TZM-bl or 5*104 A3 . 01-CCR5 in a 1:3 ratio starting from 100 µl virus solution/well . Infection of target cells was assessed by either measuring the Gaussia luciferase signal from the supernatant using the Renilla Luciferase Assay System ( Promega , Madison Wisconsin , USA ) for NLinGluc reporter viruses or by measuring the firefly luciferase activity from the lysed cells using firefly luciferase substrate ( Promega , Madison Wisconsin , USA ) for NLlucAM reporter viruses . Replication competent virus subtype B isolates JR-FL and JR-CSF were propagated on CD8 T-cell-depleted PBMC and titered as described [85] . Free virus inhibition by bnAbs and inhibitors was assessed on A3 . 01-CCR5 cells using Env-pseudotyped NLlucAM reporter viruses in all experiments except Fig 1A , where Env-pseudotyped NLinGluc virus was used . Virus input was chosen to yield virus infectivity corresponding to a firefly activity of around 10’000 relative light units ( RLU ) per 96 well in absence of inhibitors . Viruses and indicated doses of inhibitors were pre-incubated for 1 h at 37°C and added to 5*104 A3 . 01-CCR5 target cells per 96 well in the presence of 10 µg/ml diethylaminoethyl ( DEAE , Amersham Biosciences , Connecticut , USA ) . For free virus neutralization assays on primary T cells ( PBMC ) , virus-inhibitor mix was added to 1 . 5*105 PBMC per 96 well in the presence of 8 µg/ml Polybrene . After 65 h incubation at 37°C , infection was assessed by luciferase production after cell lysis and addition of firefly luciferase substrate ( Promega , Madison , Wisconsin , USA ) . Emitted RLU were quantified on a Dynex MLX luminometer ( Dynex Technologies Inc . , Chantilly , Virginia , USA ) . The inhibitor concentrations causing 50% reduction in viral infectivity ( 50% inhibitory concentration; IC50 ) were calculated by fitting pooled data from two to three independent experiments to sigmoid dose response curves ( variable slope ) using GraphPad Prism . If 50% inhibition was not achieved at the highest or lowest inhibitor concentration , a greater-than value was recorded . For comparison , neutralization assays were also performed with NLinGluc reporter viruses ( Fig 1D , S2 Fig ) with the same experimental setup . Infectivity was assessed by measuring Gaussia luciferase activity in the culture supernatant . Neutralization of HIV-1 cell-cell transmission was analyzed using NLinGluc virus transfected 293-T cells donor and A3 . 01-CCR5 target cells in all experiments , except in Fig 1F and S4 Fig , where also PBMC-PBMC transmission was used . The cell-cell transmission assay utilizes an NL4-3 derived pseudotyped HIV backbone carrying an intron-regulated Gaussia luciferase LTR-reporter construct termed inGluc which is co-transfected with an env plasmid into donor cells [41 , 55 , 56] . The intron and a reverse orientation of the reporter gene prohibit the luciferase expression in the transfected cell . Production of functional Gaussia luciferase requires correct splicing , packaging into viral particles and infection of and expression in the target cells . This allows for a clear separation of cell-cell transmission and cell fusion events [41 , 55] . To exclusively study cell-cell transmission , free virus infectivity was restricted by the omission of DEAE in the infection media as previously described [25] . For assessing neutralization of cell-cell transmission , 293-T cells were transfected with env and NLinGluc plasmids in a 1:3 ratio . 6 h post transfection , 5*103 transfected cells were seeded in 50 µl per 96 well and serial dilutions of inhibitors in 50 µl per 96 well were added . After 1 h incubation at 37°C , 1 . 5*104 A3 . 01-CCR5 target cells in 100 µl RPMI medium were added to the 293-T–inhibitor mix per 96 well . After 65 h of incubation at 37°C , Gaussia luciferase activity in the supernatant was quantified using the Renilla Luciferase Assay System ( Promega , Madison Wisconsin , USA ) according to the manufacturer’s instructions . Neutralization data were analysed with GraphPad Prism as described above . As DEAE omission cannot be used for restricting free virus infectivity in PBMC , rhTRIM5α restriction in the primary target cells was used , as recently described [25] . Briefly , PBMC were transduced with a bicistronic lentiviral GFP and rhTRIM5α expression vector ( [86]; provided by J . L . Riley ) one day after isolation and stimulation with OKT3 and 2 µg/ml CD28 in RPMI with 8 µg/ml Polybrene . Four days after transduction , rhTRIM5α positive cells were detected via the bicistronic GFP expression by FACS and sorted using a FACS AriaIII ( BD Biosciences , New Jersey , USA ) . To analyze the neutralization capacity of inhibitors , stimulated , CD8-depleted PBMC were infected with replication competent virus stocks for five days at 37°C and a MOI around 0 . 01 . PBMC were washed twice to remove free virions and 1*104 PBMC in 50 µl RPMI with 100 U IL-2 per 96 well were pre-incubated with serial dilutions of inhibitors for 1 h at 37°C . For co-culture , 1*104 rhTRIM5α transduced PBMC in 100 µl RPMI with 100 U IL-2 per 96 well were added to the donor cell- inhibitor mix . After three days of co-culture at 37°C , infectivity was assessed by intracellular p24 staining , analyzed with a FACS CyAn ADP ( Beckman Coulter ) . Neutralization data of infected , rhTRIM5α-positive cells were analyzed with FlowJo software ( TreeStar , Oregon , USA ) and GraphPad Prism as described . The inhibitory capacities of inhibitors at a pre- and post-attachment state of NLlucAM reporter viruses on A3 . 01-CCR5 target cells were analyzed as previously described [25] . In short , to test total inhibitory activity covering both , the pre- and post-attachment stage , NLlucAM reporter viruses yielding a firefly activity of around 10’000 relative light units ( RLU ) per 96 well in absence of inhibitors were pre-incubated with inhibitors for 1 h at 37°C . The pre-treated virus was then spinoculated onto 1*105 A3 . 01-CCR5 target cells in RPMI with 50 µM Hepes and 10 µg/ml DEAE per 96 well for 2 h at 1200 g and 23°C . Unbound virus and inhibitors remained with the cells during the subsequent cultivation at 37°C . These conditions provide information of the cumulative activity of an inhibitor before and after attachment , for brevity we refer to it as total activity . To test inhibitory capacity at the post-attachment stage , NLlucAM reporter viruses were first spinoculated onto A3 . 01-CCR5 cells which were then incubated with inhibitors for 1 h at 23°C before raising the temperature to 37°C . All samples were incubated for 65 h at 37°C and infectivity was determined by firefly luciferase production from the lysed cells as described . Total activity samples were set to 100% inhibition and post-attachment inhibition was expressed relative to this value . For assessing neutralization activity only at the pre-attachment step , following pre-treatment and spinoculation , unbound virus and inhibitors were washed off during two washing steps at 450 g for 2 min . Therefore , this condition provides information on how much of the binding and neutralization can occur before finalisation of CD4 engagement and if the affinity of the binding is high enough to sustain washing . The inhibitory capacities measured after washout of bnAbs reflect neutralization that was initiated prior to receptor engagement . Pre-attachment activity is expressed in relation to the total inhibitory activity which was set to 100% . To compare the probabilities that a mutant variant arises via the free virus pathway versus the cell-cell pathway for a certain antibody concentration , we first derive an analytical expression for these probabilities , pM ( c , type ) , where c is the antibody concentration and type is either cell-cell , cc , or free- virus , fv . This probability is the product of the probability that a mutation M arises given cell infection , I , and the probability that infection happens: pM ( c , type ) =P ( M|I , c , type ) ×P ( I , c , type ) =P ( M|I , c , type ) ×P ( I|c , type ) ×P ( c , type ) The last factor is one . The probability that a mutant arises , given a cell is infected via the pathway type , P ( M | I , c , type ) , does not depend on the concentration of antibodies , c as all mutations arise during reverse transcription in the infected cell in a process that is not affected by the antibodies . The probability that a mutant arises , given infection in the cell-cell pathway , is denoted P ( M | I , c , cc ) = μcc and the same probability for the free virus pathway P ( M | I , c , fv ) = μfv . Despite the fact that point mutations occurring via reverse transcription should happen with the same frequency independent of the two pathways , μcc and μfv can differ due to recombination . Recombination is more likely to happen , when a cell is infected with more than one virion , which might be more frequent in cells transfected via the cell-cell route [37 , 38 , 46] To include the possibility of differing point mutation rates , let μcc = δμfv , with δ > 0 . If δ ranges between 0 and 1 , the probability that a mutant arises in a cell that was infected via the cell-cell pathway is lower than that for a free virus infected cell . If δ > 1 , the probability that a mutant arises via the cell-cell pathway is higher than for the free virus pathway . The latter parameterization is biologically more likely due to the reasons mentioned above . The probability that a cell becomes infected given a certain antibody concentration , c , is: P ( I|c , type ) =π ( 1−%inhib ( c , type ) /100 ) π is the probability of cell infection without any antibody and %inhib ( c , type ) is the inhibition determined in in vitro infection experiments , %inhib ( c , type ) =imaxcmcm+IC50m By dividing the probability that a mutation arises via the cell-cell pathway by that via free virus infection we obtain: γ ( c ) =pM ( c , cc ) pM ( c , fv ) =δ ( 1−imax , cccmcc ( cmcc+IC50ccmcc ) −1 ) 1−imax , fvcmfv ( cmfv+IC50fvmfv ) −1 For this analysis , we used the IC50 and slope values ( S3 and S4 Tables ) obtained by inhibition data analysis performed with GraphPad Prism . The formula are implemented in the R statistical analysis software [87] . Statistical analyses ( correlation analyses according to Spearman using the untransformed data sets ) were performed using GraphPad Prism . | When selecting broadly neutralizing antibodies ( bnAbs ) for clinical application , potency and breadth against free viruses are vital , but additional features may be needed to ensure in vivo efficacy . Considering that HIV-1 can utilize free virus and cell-cell transmission to infect , the efficacy of neutralizing antibodies in vivo may depend on their ability to block both pathways . While breadth and potency of bnAbs against free viruses have been intensely studied , their precise activity during cell-cell spread remains uncertain . Our analysis of the cell-cell neutralization capacity of a large selection of bnAbs against a spectrum of HIV-1 strains revealed that while bnAbs showed an overall decreased activity during cell-cell transmission , losses varied substantially depending on bnAb and virus strain probed . Although bnAbs occasionally retained activity during cell-cell transmission for individual viruses , this ability was rare and generally not associated with a high potency against free virus spread . Notably , neutralization of free virus but not cell-cell transmission was linked with the activity of bnAbs to inhibit prior to CD4 engagement , highlighting the functional differences of the processes . Since no single bnAb combines the entire range of mechanistic features anticipated to support in vivo efficacy , our study adds further evidence that combinations of bnAbs need to be considered for human application . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Capacity of Broadly Neutralizing Antibodies to Inhibit HIV-1 Cell-Cell Transmission Is Strain- and Epitope-Dependent |
Taiwan is an island located in the south Pacific , a subtropical region that is home to 61 species of snakes . Of these snakes , four species—Trimeresurus stejnegeri , Protobothrops mucrosquamatus , Bungarus multicinctus and Naja atra—account for more than 90% of clinical envenomation cases . Currently , there are two types of bivalent antivenom: hemorrhagic antivenom against the venom of T . stejnegeri and P . mucrosquamatus , and neurotoxic antivenom for treatment of envenomation by B . multicinctus and N . atra . However , no suitable detection kits are available to precisely guide physicians in the use of antivenoms . Here , we sought to develop diagnostic assays for improving the clinical management of snakebite in Taiwan . A two-step affinity purification procedure was used to generate neurotoxic species-specific antibodies ( NSS-Abs ) and hemorrhagic species-specific antibodies ( HSS-Abs ) from antivenoms . These two SSAbs were then used to develop a sandwich ELISA ( enzyme-linked immunosorbent assay ) and a lateral flow assay comprising two test lines . The resulting ELISAs and lateral flow strip assays could successfully discriminate between neurotoxic and hemorrhagic venoms . The limits of quantification ( LOQ ) of the ELISA for neurotoxic venoms and hemorrhagic venoms were determined to be 0 . 39 and 0 . 78 ng/ml , respectively , and the lateral flow strips were capable of detecting neurotoxic and hemorrhagic venoms at concentrations lower than 5 and 50 ng/ml , respectively , in 10–15 min . Tests of lateral flow strips in 21 clinical snakebite cases showed 100% specificity and 100% sensitivity for neurotoxic envenomation , whereas the sensitivity for detecting hemorrhagic envenomation samples was 36 . 4% . We herein presented a feasible strategy for developing a sensitive sandwich ELISA and lateral flow strip assay for detecting and differentiating venom proteins from hemorrhagic and neurotoxic snakes . A useful snakebite diagnostic guideline according to the lateral flow strip results and clinical symptoms was proposed to help physicians to use antivenoms appropriately . The two-test-line lateral flow strip assay could potentially be applied in an emergency room setting to help physicians diagnose and manage snakebite victims .
Envenoming resulting from snakebites is a significant public health issue in many regions of the world , particularly in tropical and subtropical countries and some poor rural communities [1] . An estimated 1 , 800 , 000–2 , 700 , 000 envenoming cases and 81 , 410–137 , 880 associated deaths occur each year globally owing to snakebite [2] . The regions with the highest burden are South Asia , Southeast Asia , and Africa [2 , 3] . Administration of antivenom is the standard treatment for snake envenomation . In most countries , multiple types of antivenom are clinically available , but uncertainty regarding the appropriate antivenom to use in any given situation remains an important issue . To date , the species responsible for envenomation of snakebite victims referred for medical treatment is initially identified primarily based on the shape of the wound or identification of dead snakes brought to the hospital . Thereafter , the physician monitors local symptoms to confirm which antivenom should be used . However , some clinical symptoms caused by envenomation are similar among species , and non-venomous snakes are often responsible for the patient’s snakebite [4] . Additionally , physicians are often misled by incorrect descriptions of the snake by victims or their family members [5] . Identification of venomous snake species is important for optimal clinical management , because it allows physicians to use the correct antivenom for effective treatment , thereby improving patients’ prognosis and preventing the waste of expensive antivenoms and exposing victims to antivenom-induced adverse reactions [6] . Although identification of snake species is important for the management of snakebite-related injuries worldwide , there are currently no developed standard platforms or guidelines for snakebite diagnosis globally . Detection of venom proteins using antibodies is a simple and effective approach for identifying the species responsible for snakebite . To date , various immunoassays for detecting venom proteins in body fluids have been described [7–13] , including radioimmunoassay [14] , agglutination assays [9 , 15] , enzyme-linked immunosorbent assays ( ELISAs ) [10–12 , 16 , 17] , and fluorescence immunoassays [18 , 19] . In addition to immunoassays , immunology-based biosensors have been explored for detection of snakebite [20 , 21] . ELISAs and lateral flow assays [22 , 23] are arguably the best choice of immunoassays for snakebite identification . ELISAs , the most common and general immunoassays in clinical use , are sensitive to their target at pictogram per milliliter levels [18] . Although the antibodies are relatively costly , ELISA devices and reagents are affordable for routine diagnosis . Compared with ELISAs , lateral flow assays offer advantages in terms of detection time and required equipment: it takes only ~5–20 min to obtain assay results and no supporting instrumentation is needed [24 , 25] . Although lateral flow assays mainly provide qualitative results , their simple design and operation compared with quantitative ELISAs make them the most user-friendly for the public , allowing rapid adoption in rural countries . Snake venoms contain many proteins , and closely related snake species have some of the same or similar venom components , causing cross-reactions in immunoassays applied to detect venom proteins [11 , 12 , 26] . The venom antigens responsible for the observed cross-reactivity would further cause ambiguities and false-positive results in snake species detection [11 , 27] . Hence , the direct use of polyclonal antibodies against whole venoms for snake species detection is inappropriate , and elimination of cross-reactive antibodies is critical for generating an immunoassay with high specificity for discriminating snake species [11 , 12 , 28] . Solving the problem of cross-reaction and improving the specificity of immunoassays might most efficiently be achieved through purification of species-specific antibodies ( SSAbs ) on affinity columns immobilized with venom proteins cross-reactive to the polyclonal antibodies or antisera [11 , 12] . Six venomous snakes—Deinagkistrodon acutus , Trimeresurus stejnegeri , Protobothrops mucrosquamatus , Daboia russelii formosensis , Bungarus multicinctus and Naja atra—are indigenous to Taiwan , a subtropical island in East Asia [29] . Four kinds of antivenom had been produced by the Vaccine Center , Center for Disease Control , Taiwan , to treat envenomation by these six venomous snakes and effectively limit snakebite mortality [30] . Freeze-dried hemorrhagic antivenom ( FHAV ) is used to treat envenomation by T . stejnegeri and P . mucrosquamatus , and freeze-dried neurotoxic antivenom ( FNAV ) neutralizes venom of B . multicinctus and N . atra . Envenomation by the other two snake species is treated by monovalent antivenoms . A population-based study of venomous snakebites in Taiwan from 2005 to 2009 reported a total of 4647 snakebite cases , of which 380 ( 8 . 1% ) received at least two types of antivenoms , mainly because of similarities in the clinical presentations of different snakebites and the inability of some patients to identify the culprit snake [31] . In some studies , such unidentified cases accounted for 12–45% of total cases [32–35] . In addition , according to a clinical survey of antivenom usage in Taiwan , more than 99% of snakebite patients that had received FHAV or FNAV treatment were rescued [36] , indicating that most snakebite cases in Taiwan represent envenomation by T . stejnegeri , P . mucrosquamatus , B . multicinctus or N . atra . Unfortunately , there have been very few efforts to develop sensitive assays for detecting snake venom in Taiwan . Currently , only one ELISA-based blood assay has been developed to detect the N . atra venom , but it is not commercially available [7] , and no laboratory test can be used to identify other types of venoms . In the present study , we designed a workflow to develop immunoassays for snakebite detection based on clinical antivenom usage in Taiwan . We used FHAV and FNAV as resources for purification of hemorrhagic species-specific antibodies ( HSS-Ab ) and neurotoxic species-specific antibodies ( NSS-Ab ) , and applied these two critical reagents to develop ELISAs and lateral flow strip assays . These assays hold the potential for use in identification of snake species responsible for snakebites in Taiwan .
Lyophilized venoms of T . stejnegeri , P . mucrosquamatus , B . multicinctus and N . atra were obtained from the Center for Disease Control , R . O . C ( Taiwan ) . The venoms were collected from several adult specimens , then freeze-dried and stored at -20°C before use . Hemorrhagic venom ( T . stejnegeri and P . mucrosquamatus ) -immunized and neurotoxic venom ( B . multicinctus and N . atra ) -immunized horse plasma were also donated by the Center for Disease Control , R . O . C ( Taiwan ) . The plasma was stored at -80°C before use . For coupling of venom proteins onto Sepharose beads , CNBr-activated Sepharose 4B was swollen in 1 . 0 mM HCl ( pH 3 . 0 ) , then incubated with 10 mg hemorrhagic or neurotoxic snake venoms dissolved in coupling buffer ( 0 . 1 M NaHCO3 pH 8 . 3 ) overnight at 4°C on a round rotator . After washing with coupling buffer , any remaining active sites on beads were blocked by incubating overnight at 4°C with blocking buffer ( 1 . 0 M diethanolamine pH 8 . 0 ) on a rotator . The beads were then alternately washed three times with an acidic buffer ( 0 . 1 M C2H3NaO2 pH 4 . 0 , 0 . 5 M NaCl ) and basic buffer ( 0 . 1 M Tris pH 8 . 0 , 0 . 5 M NaCl ) and packed into a column , The resulting venom affinity columns were equilibrated with binding buffer ( 10 mM Tris-HCl pH 7 . 5 ) and stored at 4°C before use . To purify HSS-Ab , 2 ml FHAV was diluted in 30 ml of binding buffer and the diluted sample was pumped into the neurotoxic venom affinity column at 4°C for 3 h . The flow-through fraction was then pumped into the hemorrhagic venom affinity column at 4°C for another 3 h . The hemorrhagic venom affinity column was then washed with 60 ml binding buffer and 60 ml wash buffer ( 10 mM Tris-HCl pH 7 . 5 , 0 . 5 M NaCl ) . After washing , each affinity column was eluted with 20 ml of acidic ( 100 mM glycine pH 2 . 5 ) or basic ( 100 mM triethylamine pH 11 . 5 ) elution buffer , and eluted fractions ( 1 ml/fraction ) were collected into microcentrifuge tubes containing 100 μl of neutralized buffer ( 1 . 5 M Tris-HCl pH 8 . 0 ) . Finally , all eluted fractions were pooled , concentrated , and exchanged into phosphate-buffered saline ( PBS ) by dialysis overnight . The concentrated antibodies in PBS were diluted with an equal volume of glycerol and stored at -20°C . Similar protocol was used to purify NSS-Ab from 2 ml FNAV , in which the diluted FNAV was passed through the hemorrhagic venom affinity column first , and the flow-through fraction containing NSS-Ab was further purified using the neurotoxic venom affinity column . Snake venom proteins ( 100 ng ) were diluted in 100 μl PBS and coated onto 96-well polystyrene microplates ( Corning , USA ) by incubating at 4°C overnight . The plates were washed six times with 200 μl of PBST ( PBS contain 0 . 1% Tween-20 ) and blocked by incubating with 200 μl of 1% ovalbumin in PBS at room temperature for 2 h . After washing wells six times with PBST , antivenom or purified Ab ( 1 mg/ml ) was serial diluted ( from 1:2000 to 1:16000 ) and added to individual wells , then the plate was incubated at room temperature for 2 h . Wells were again washed six times with PBST , and then alkaline phosphatase-conjugated anti-horse IgG antibody ( Santa Cruz Biotechnology , USA ) was added to each well and the plate was incubated at room temperature for 1 h . After washing six times with PBST , the substrate 4-methyl umbelliferyl phosphate ( 100 μM , 100 μl/well; Molecular Probes ) was added to each well , and fluorescence was measured with a SpectraMax M2 microplate reader ( Molecular Devices , USA ) at excitation and emission wavelengths of 355 and 460 nm , respectively . Snake venom proteins ( 5 μg ) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) , transferred onto PVDF ( polyvinylidene difluoride ) membranes ( Millipore , USA ) , and probed with antivenom or purified Ab . Immunoreactive proteins in PVDF membranes were detected by incubating for 1 h with the appropriate alkaline phosphatase-conjugated anti-horse IgG antibodies ( Santa Cruz Biotechnology , USA ) and visualized using the CDP-Star Western Blot Chemiluminescence Reagent ( PerkinElmer , USA ) . Antibodies were biotinylated using a Lightning-Link biotinylation kit ( Innova Biosciences , USA ) according to the protocol provided by the manufacturer . Briefly , 100 μl of SSAb ( 2 mg/ml ) was mixed with 10 μl of modifier reagent , then added to the tube containing biotinylation powder and incubated for 15 min in the dark . After the biotinylation reaction , 10 μl of quencher reagent was added and the reaction mixture was stored at -20°C until use . SSAb ( 100 μl at 2 mg/ml ) , diluted 1:1000 in PBS , was coated onto 96-well polystyrene microplates . Thereafter , wells were blocked by incubating with 1% bovine serum albumin ( BSA ) in PBS for 1 h , then washed six times with 200 μl PBST . Test samples ( 100 μl ) were added into individual wells and incubated at room temperature for 2 h . After washing six times with PBST , 100 μl of biotin-labeled SSAb , diluted 1:16000 in PBS , was added and plates were incubated for 2 h . Plates were again washed six times with PBST , then alkaline phosphatase-conjugated streptavidin was added and allowed to interact with biotin . The alkaline phosphatase substrate , 4-methyl umbelliferyl phosphate ( 100 μM , 100 μl/well ) , was then added to each well , and fluorescence was measured with a SpectraMax M5 microplate reader at excitation and emission wavelengths of 355 and 460 nm , respectively . Experiments were performed on male 7-wk-old littermate mice ( C57BL/6Narl strain ) . Mice were maintained under specific pathogen-free conditions with a 12:12 h light-dark cycle at a temperature of 22°C and a humidity level of 60–70% . Animals had ad libitum access to food and water . Mice ( n = 3/group ) within a defined weight range ( 20–25 g ) were subcutaneously ( B . multicinctus and N . atra venom ) or intraperitoneally ( T . stejnegeri and P . mucrosquamatus venom ) injected with a precise 0 . 1 ml volume of sterile saline solution containing a minimal lethal dose ( MLD ) of venom . Blood samples from each mouse were collected using a heparinized capillary blood collection system ( Kent Scientific , USA ) 0 . 5 , 1 , 1 . 5 and 2 h after venom injection . Collected blood was centrifuged at 3000 × g for 20 min . The resulting supernatant ( plasma ) was collected into a microcentrifuge tube and stored at -80°C before use . A colloidal gold ( 40 nm ) solution ( REGA Biotechnology Inc . , Taipei , Taiwan ) was adjusted to pH 8 . 0 with 0 . 1 M potassium carbonate . The optimal concentration of SSAb ( 10 mg ) was added to 2 ml of colloidal gold solution and incubated at room temperature for 10 min with gentle mixing . The mixture was blocked by incubating with 0 . 5 ml of 5% BSA in PBS at room temperature for 15 min with gentle mixing , and then centrifuged at 10 , 000 × g at 4°C for 30 min . The gold pellets were suspended in PBST containing 1% BSA , and washed by repeated centrifugation and suspension in the same solution . The final precipitates were suspended in 1 ml PBST containing 1% BSA and stored at 4°C until use . The strips were manufactured by REGA Biotechnology Inc . ( Taipei , Taiwan ) . Nitrocellulose membranes , sample pads , conjugate pads and absorbent pads were all from REGA Biotechnology Inc . Conjugate pads were saturated with HSS-Ab–or NSS-Ab–conjugated colloidal gold , then dried at 37°C for 1 h before assembling . The nitrocellulose membrane was pasted to the cardboard , after which conjugated and absorbent pads were also pasted to the cardboard such that they overlapped with each side of the nitrocellulose membrane by about 2 mm . The sample pad was also laid over the absorbent pad ( 2 mm overlap ) and pasted onto the cardboard . The AGISMART RP-1000 rapid test immuno-strip printer ( REGA Biotechnology Inc . ) was used to dispense HSS-Abs and NSS-Abs ( 2 mg/ml ) onto hemorrhagic and neurotoxic test lines , respectively , and goat anti-horse IgG antibody ( 2 mg/mL ) ( REGA Biotechnology Inc . ) onto the control line on the nitrocellulose membrane . The distance between each line was 5 mm . The strips were prepared and assembled in a low-humidity environment , packaged into an aluminum pouch , and stored at room temperature before use . Patients with suspected snakebite were admitted directly to the Emergency Departments of Taipei Veteran General Hospital , Linkou Chang Gung Memorial Hospital , Chiayi Chang Gung Memorial Hospital or Hualien Tzu Chi Hospital , and did not receive antivenom treatment before being enrolled in this study . After obtaining signed , informed consent forms from patients , 5 ml of blood was collected in SST blood collection tubes ( BD , Franklin Lakes , New Jersey , USA ) and centrifuged at 4°C for 10 min to obtain serum samples . A 100–200 μl aliquot of serum sample was immediately applied to lateral flow strip test in the emergency room , and results were determined by clinical physicians . The remainder of each sample was sent to the laboratory in Chang Gung University and stored at -80°C . All samples were re-analyzed using the lateral flow strip test in the laboratory to confirm emergency room result; samples were also analyzed by sandwich ELISA to measure the concentrations of venom proteins . Each serum sample ( 100–200 μl ) was diluted with 1 volume of reaction buffer ( 100 mM borax , 250 nM polyvinylpyrrolidone ( PVP ) -40 and 1% Triton X-100 ) in a microcentrifuge tube . The strips were directly soaked in the samples , and results were recorded after a 10-min reaction . The Cohen's kappa coefficient ( κ ) statistic [37 , 38] was used to assess the strength of inter-method agreement for diagnosis results . The value of kappa coefficient statistic over 0 . 75 , between 0 . 75 to 0 . 40 , or below 0 . 40 indicates excellent agreement , good to fair agreement , and poor agreement , respectively [39 , 40] . All clinical serum samples were collected and obtained at Taipei Veteran General Hospital , Linkou Chang Gung Memorial Hospital , Chiayi Chang Gung Memorial Hospital or Hualien Tzu Chi Hospital from February 2017 to February 2018 . All study subjects are adult participants and signed an informed consent form approved by the Institutional Review Board ( IRB ) of Taipei Veteran General Hospital ( Approval No: 2017-06-013BCF ) and Linkou Chang Gung Memorial Hospital ( Approval No: 201800098B0 ) permitting the use of plasma samples for this study . Experiments involving the care , bleeding , and injection of mice with various venoms were reviewed and approved by the Institutional Animal Care and Use Committee of Chang Gung University ( Permit Number: CGU14-024 ) . The protocol for mouse studies was based on guidelines provided by the Council for International Organizations of Medical Sciences ( CIOMS ) [41] .
To assess the cross-reactivity among four venoms and two antivenoms , we performed indirect ELISAs and immunoblotting . The results of indirect ELISAs showed that cross-reactivity of FHAV towards B . multicinctus and N . atra venom was very low ( Fig 1A ) ; however , FNAV strongly cross-reacted with T . stejnegeri and P . mucrosquamatus venom ( Fig 1B ) . Cross-reaction signals increased gradually with increases in antivenom concentration , and both antivenoms showed stronger reactivity toward homologous venoms than heterologous venoms . As shown in Western blot profiling data , FHAV primarily cross-reacted with protein bands in the high molecular weight region ( 55–70 kDa ) of N . atra venom ( Fig 1C ) , whereas FNAV cross-reacted with multiple bands in T . stejnegeri and P . mucrosquamatus venoms , predominantly towards protein bands in the 15–25 kDa range in P . mucrosquamatus venom ( Fig 1D ) . A comparison of the protein profiles of the four venoms ( S1 Fig ) showed that , generally , most venom components of these venoms were recognized by the corresponding homologous antivenom . In this study , we used an affinity purification procedure to eliminate cross-reactive antibodies from antivenoms . Heterologous venom-immobilized affinity columns were prepared and used to remove cross-reactive antibodies from antivenoms , after which the remaining antibodies were purified using a homologous venom-immobilized affinity column , yielding SSAbs . SDS-PAGE analysis of the affinity-purified HSS-Abs and NSS-Abs showed a typical pattern of IgG heavy and light chains ( S2 Fig ) . Indirect ELISAs and Western blotting assays were performed to evaluate the specificity of affinity-purified SSAbs , HSS-Abs and NSS-Abs . The results of indirect ELISAs showed that both SSAbs possessed high specificity toward the homologous venoms , and showed significantly decreased cross-reactivity with heterologous venoms compared with the original antivenoms ( Fig 2A & 2B ) . The immunoreactivity of HSS-Ab towards P . mucrosquamatus venom was stronger than that towards T . stejnegeri venom , whereas NSS-Ab preferentially reacted with venom proteins from N . atra compared with those from B . multicinctus . Consistent with the ELISA data , Western blot analyses also showed the high specificity of HSS-Ab and NSS-Ab towards their homologous venoms ( Fig 2C & 2D ) , although NSS-Ab did weakly react with high-molecular-weight proteins ( 55–70 kDa ) in the two hemorrhagic venoms ( Fig 2D ) . Proteins in the high molecular weight region ( 25–70 kDa ) of T . stejnegeri and P . mucrosquamatus venom represented the dominant targets of HSS-Ab ( Fig 2C ) ; in contrast , NSS-Ab mainly recognized lower molecular weight proteins ( <15 kDa ) in the two neurotoxic venoms ( Fig 2D ) . To form sandwich complexes for ELISA measurements , we used HSS-Ab ( or NSS-Ab ) as the capture antibody and biotinylated HSS-Ab ( or NSS-Ab ) as the detection antibody . Antibody concentrations , buffers , and incubation times used for these sandwich ELISAs were optimized based on the ELISA development guide provided by the manufacturer ( R&D Systems , Inc . ) . To determine the sensitivity of sandwich ELISA assays for snake venom detection , we serially diluted the four snake venoms in plasma and measured their reactivity by sandwich ELISA , generating standard curves for each venom ( Fig 3 ) . The limits of detection ( LODs ) of sandwich ELISAs for detecting T . stejnegeri , P . mucrosquamatus , B . multicinctus and N . atra venom were 0 . 39 , 0 . 14 , 0 . 56 and 0 . 23 ng/ml , respectively . In all cases , R2 values of standard curves were greater than 0 . 99 . Taken together , these results suggest that our sandwich ELISA has the potential to identify snake species and quantify venom proteins in body fluids . For further application of this snakebite sandwich ELISA , the four venoms were used as the gold standards for venom quantification , and the LOD value determined as described above was set as the cutoff for detecting each venom . To determine whether snake venoms are still detectable after neutralization by antivenoms , we individually neutralized a fixed amount of venom with serially diluted antivenoms and then performed sandwich ELISAs . ELISA signals produced by 10 ng of T . stejnegeri ( Fig 4A ) and P . mucrosquamatus ( Fig 4B ) venom were completely eliminated by 8–40 nl of FHAV . Similarly , 40–200 and 8–40 nl of FNAV totally blocked ELISA signals derived from 10 ng of B . multicinctus ( Fig 4C ) and N . atra ( Fig 4D ) venom , respectively . These observations show that our sandwich ELISA assays only detects “free” venom proteins , and not antivenom-neutralized venoms . Importantly , they also suggest that our assays are suitable for evaluating the amount of free venom proteins remaining in a snakebite victim , making it possible to determine whether the dosage of antivenom delivered is sufficient to treat the patient . The MLD of each venom was determined using an experimental envenomation animal model . The MLD of T . stejnegeri , P . mucrosquamatus , B . multicinctus and N . atra were 1 . 5 , 3 , 0 . 3 and 0 . 65μg/g , respectively . All mice developed local symptoms within 10–20 min after injection of a lethal dose of venom . As soon as 30 min post injection , all four venoms could be detected by sandwich ELISA in plasma samples from mice injected with venom; as expected , none of the saline-injected control mice showed a positive reaction in these assays ( Fig 5 ) . The plasma concentrations of T . stejnegeri ( Fig 5A ) , P . mucrosquamatus ( Fig 5B ) and N . atra ( Fig 5D ) venom proteins in these mice gradually increased during a 2-h period post injection . In contrast , the plasma concentrations of venom proteins in mice injected with B . multicinctus venom decreased dramatically during this period ( Fig 5C ) . Collectively , these results demonstrate that the newly developed sandwich ELISA can successfully identify and quantify these four Taiwanese snake venoms in vivo . Although the newly developed sandwich ELISA assay exhibited high specificity and sensitivity , the assay time in its current format is too long for use in clinical practice . To reduce the operation time and simplify the platform for snakebite diagnosis , we sought to develop another assay using a lateral flow strip format with two test lines ( Fig 6A ) . To assess the specificity and sensitivity of this lateral flow strip , we tested it on the four venoms serially diluted ( from 500 ng/ml to 5 ng/ml ) in human plasma . The assay was evaluated based on the appearance of a control line , a hemorrhagic test line ( H line ) , or a neurotoxic test line ( N line ) ( Fig 6B ) . All strips showed a visible control line , confirming that all test samples were successfully flowed onto the strips ( Fig 7 ) . An H line was only observed in those strips used to test T . stejnegeri and P . mucrosquamatus venom ( Fig 7A & 7B ) , and the N line appeared only in assays of N . atra and B . multicinctus venom proteins ( Fig 7C & 7D ) . These results indicate that this newly developed strip assay does not exhibit sufficient cross-reactivity to cause ambiguous results . In assays of hemorrhagic venom , the H line was still detectable after reducing the concentration of T . stejnegeri and P . mucrosquamatus venom proteins to 50 ng/ml ( Fig 7A & 7B ) . For neurotoxic venom detection , the N line was still visible when both venom protein levels were reduced to 5 ng/ml ( Fig 7C & 7D ) . Thirty-two victims of snakebite sent to Emergency Departments of the four participating hospitals from May 2017 to February 2018 were enrolled in this study . Among them , eleven patients were excluded because they had been treated with the appropriate antivenom before arrival in the Emergency Department ( n = 9 ) or displayed no symptoms ( n = 2 ) . The serum samples obtained from the remaining 21 cases were analyzed by sandwich ELISA and lateral flow strip assay ( Table 1 ) . The lateral flow strip assay showed 100% ( 5/5 ) specificity and 100% specificity ( 5/5 ) for the detection of neurotoxic envenomation samples . However , the sensitivity for detecting hemorrhagic envenomation samples was only 36 . 4% ( 4/11 ) . We used the kappa statistic to assess the strength of agreement between the two assays , and this analysis indicated good to fair agreement ( κ = 0 . 53 ) between snakebite sandwich-ELISA and lateral flow strip assay ( Table 1 ) . The clinical information of these 21 patients were summarized in Table 2 . Most of the culprit snakes were initially identified by patients’ description or recognition of snake photograph ( 17/21 ) , and 2 of them were definitely confirmed according to the killed snakes brought to the hospital . The aggressor snakes of case 18–21 cannot be identified at scenes of ED . In the laboratory identification , both ELISA and lateral flow strip assay were shown hemorrhagic venom positive results for case 18 and 19 , and venom negative result for case 20 and 21 . All patients were presented with local swelling except case 11 who was initially identified as B . multicinctus envenomation , and no neurologic symptoms appeared in all . Case 16 , 18 and 19 , who were performed surgery , have higher level of venom concentration than other victims . Seven cases with hemorrhagic venom-positive ELISA results appeared with negative lateral flow strip results . The venom concentration of them was ranged from 2 . 2 to 10 . 6 ng/ml , which are lower than other cases detected by lateral flow strip assay . Among them , five cases were shown mild clinical severity , and 2 cases shown moderate severity . Case 11 , 15 , 17 , 20 and 21 have ELISA undetectable venom level . All of them have mild clinical severity that the local swelling restricted in fang mark area , or even did not have local swelling . The sample time after bite for the majority of the victims ( 15/21 ) was ≦3 . 5 h . Overall , there was no significant correlation between the blood venom concentration and sampling time after snakebite or the bitten area according to this small-scale clinical study .
The presence of common antigens in heterologous venoms has been demonstrated to be a major source of bias for the development of snakebite detection assays [26 , 42] . The appearance of widespread cross-reactivity between heterologous snake venoms and polyvalent or monovalent antivenoms considerably hampers the specificity of such assays [11 , 12 , 28 , 43] . Consistent with these previous observations , the current study also found that FHAV and FNAV cross-reacted towards heterologous venoms , as evidenced by the detection of 3–5 protein bands in Western blot analyses ( Fig 2A & 2B ) . However , snake venoms are known to comprise multiple ( 10–100 ) proteins , many of which have the same or similar epitope ( s ) , but with different molecular weights . At present , it is difficult to predict the venom components that contribute to this cross-reactivity . Immunoaffinity purification appears capable of removing antibodies in antiserum that recognize common epitopes of venom components . Even though the identity of the species-specific antigens and common epitopes that contribute to the cross-reactivity remain largely unknown , we were still able to successfully obtain venom protein antibodies with high specificity ( i . e . , low cross-reactivity among different snake species ) . In addition , detection of snake envenomation by monoclonal antibodies generated using a single species-specific venom protein can considerably improve assay specificity [44–47] . However , the sensitivity of these antibodies may not be high enough , because venoms contain numerous protein components and a mAb can only react with a single epitope on its target protein . Moreover , the targeted venom component may become degraded through metabolic processes in biological systems . Thus , the application of monoclonal antibodies to the development of snakebite kits remains a considerable challenge . The promising data shown in the present study suggest that purification of SSAbs from antivenoms could be a feasible and cost-effective strategy for generating effective probes for snake venom detection and species discrimination . Sandwich ELISAs , which have been widely used in snake venom detection and snakebite diagnosis [10 , 11 , 44 , 48] , are capable of measuring venom proteins at the level of a few nanograms per milliliter . In conjunction with the biotin-streptavidin amplification system , the detection limit can be further improved , reducing the lower limit to less than 1 ng/ml [10] . Generally , two different antibodies are used for sandwich ELISA assay development . Because we used the same SSAb as both capture and detection antibody in our sandwich ELISA , the capture SSAbs in the solid phase only occupied one binding site on their cognate antigen molecules . Thus , the detection SSAb was still capable of recognizing the remaining epitopes on the captured antigens . With this approach , how to pair two suitable antibodies to form the sandwich complex for detection is not a concern , making it easy to adapt for snake venom detection . Although the sandwich ELISA assay is time consuming , and thus is likely not the most appropriate assay for use in emergency rooms , it is still a good tool for snakebite epidemiology and prognosis studies . The usefulness of our sandwich ELISA assay was demonstrated by detecting venoms in blood samples from an experimentally envenomed mouse model ( Fig 5 ) . These experiments showed that this assay is capable of identifying the envenoming species and quantifying venom concentrations in blood . Application of this ELISA to the snakebite animal model revealed that concentrations of T . stejnegeri , P . mucrosquamatus and N . atravenom proteins gradually increased in mouse plasma during a 2-h period post-injection; in contrast , the concentration of B . multicinctus venom proteins dramatically decreased over this same time period ( Fig 5 ) . A previous study reported that more than half ( nearly 60–80% ) of B . multicinctus venom components are neurotoxins , including β-bungarotoxin , α-bungarotoxin and γ-bungarotoxin [49] . These bungarotoxins bind to specific receptor ( s ) on presynaptic and postsynaptic membranes , leading to paralysis and neurotoxicity [50–52] . Our findings suggest that , when injected into the victim , these bungarotoxins rapidly interact with specific receptors , and thus are immobilized in the neuromuscular junctions; this , in turn , causes a significant decrease in their bioavailability , accounting for the rapid decrease in their concentration in blood plasma . The lateral flow strip assay is a sandwich-based immunostrip used to rapidly ( 5–20 min ) examine whether target molecules are present in a sample [53] . This type of assay is appropriate for use in snakebite detection and diagnosis , and can offer guidance to physicians in administering antivenom [22 , 23] . Furthermore , the visual diagnosis format of this assay is simple , making it desirable for use in developing countries , where snakebites are most prevalent . However , some factors and sampling conditions may profoundly affect strip assay results . For example , a high concentration of serum proteins and high viscosity of the test sample could interfere with the formation of the red line in test and control zones , and samples containing high concentrations of salt , such as urine , often cause false-positive results . Thus , in some situations , sample pretreatment is required . The lateral flow strip assay developed here has two test lines for discriminating hemorrhagic and neurotoxic snake envenomation in Taiwan . This strip assay successfully detected and identified snake venom in serum samples from snakebite patients . Our small-scale clinical study demonstrated that the lateral flow strip assay is useful for assessing neurotoxic envenomation , exhibiting a sensitivity/specificity of 100% . It is suggested that the newly developed strip assay holds promise for the diagnosis of neurotoxic snakebite . However , the sensitivity of this assay for hemorrhagic envenomation was nearly 40% . ELISA results of these 11 hemorrhagic envenomation samples showed that the T . stejnegeri or P . mucrosquamatus venom protein concentrations in 7 lateral flow strip-negative samples were less than 10 ng/ml ( Table 2 ) , suggesting that this assay is not sensitive enough to detect snakebite cases with low blood concentration of hemorrhagic venom in clinical practice . Although , at this point , we cannot definitively establish the appropriateness of our lateral flow strip assay for precise diagnosis of all clinical snakebites , the combination of clinical symptoms and the results of lateral flow strip could improve the clinical utility of our lateral flow strip , especially in the weak aspect of diagnosis of hemorrhagic snake envenomation . A diagnosis flowchart which composed of clinical symptoms and the result of lateral flow strip was therefore proposed ( Fig 8 ) . Because of the relative high sensitivity and specificity of our lateral flow strip in diagnosis of neurotoxic snake envenoming , cases with negative lateral flow strip results have a great possibility of hemorrhagic snake envenoming when they have developed local tissue swelling . This diagnosis flowchart may further enhance the ability of our lateral flow strip to guide the usage of antivenom . Because only 21 snakebite cases were included , further study using a larger sample set is needed to verify the sensitivity , specificity , stability , and feasibility of this strip assay . Seven of the 21 clinical samples examined in this study showed positive ELISA result but negative on the lateral flow strip test . All of them were identified as hemorrhagic snake envenomation with low venom concentration level accompanying with mild or moderate clinical severity . Even though these patients have been transferred to hospital and sampled nearly within 1–2 hrs , their blood venom concentrations were still lower than the others . It is highly possible that the amount of venom injected into these victims was originally low , which is hard to detect by lateral flow strip assay after dilution in the systemic circulation , and only induced mild clinical symptoms . Despite initial identification of envenoming species is almost the same as the test results in our small-scale study ( Table 2 ) , sometimes , envenoming species identified by patients or their family may mislead the physicians . Take case 8 as an example , this patient was initially identified as P . mucrosquamatus envenomation according to family members’ recognition of the snake pictures , however , both ELISA and lateral flow strip assay showed positive result of neurotoxic snake envenomation , indicating the culprit snake is N . atra . Furthermore , few cases with negative result of both assays may be bitten by non-venomous snakes . There are more than 50 snake species in Taiwan . It is hard for citizens to correctly recognize and distinguish all of them . Bringing the envenoming snake to the hospital , like cases 9 and 13 , is the most reliable way for species identification . In the present study , all five cases ( case 11 , 15 , 17 , 20 and 21 ) with negative quantification of venom displayed the mild clinical severity . These patients may be bitten by non-venomous snakes , or the dry bite . As mentioned above , snakebite victims have the chance to misidentify the envenoming species , and slight swelling usually occurred around the fang mark even if they were bitten by non-venomous snakes . It is one of the reasons leading to the negative results in both assays . In addition , although we did not observe a close relationship between the transcurrent time from the bites to the ER consult and the results of the diagnostic test , 3 of the 5 cases with negative ELISA result had longer transcurrent time . Case 11 , 15 and 21 had their transcurrent time for 14 , 10 . 5 and 34 hrs , respectively . The metabolism time more than 10 hours may allow the venom to be eliminated from patients’ body and resulted in negative test result . The delay in seeking medical help may be another reason leading to the negative test results . On the other hand , case 18 had 12 . 5 hours of transcurrent time , but displayed severe clinical symptoms and positive test results . It is reasonable to assume that the type and amount of venom injected into patients is the main factor to determine the outcome of the test results , and the effect of transcurrent time could be minor . The current study used serum samples from snakebite patients to evaluate the performance of the snakebite lateral flow strip assay . Other types of specimen , such as urine , wound exudate and blister fluid , have been reported as alternatives for venom detection [7 , 12] . The highest amounts of venom proteins ( >100 ng/ml ) are found in wound exudates and blister fluid; thus , venom proteins are more easily detected and measured in these types of specimens [12] . However , cases with blister fluid are very rare; in the current study , only one patient formed blister fluids after envenomation . Wound exudates are easier to obtain than blister fluids , but obtaining untreated wound exudates for pre-clinical trials is another challenge . Because people have been taught to perform first aid when bitten by snakes , snake venom remaining in the wound will typically have been washed out or swabbed out . Furthermore , fang marks have usually clotted by the time victims arrive at the Emergency Department . Thus , although wound exudate maybe the best sample type for venom detection , how to collect good quality samples for survey remains a daunting challenge . Countries in tropical and subtropical regions have various indigenous venomous snake species . Two or more antivenoms are currently available for clinical treatment of snake envenomation . Directly using these antivenoms as a resource for the development of snakebite diagnostic assays could be a cost-effective approach for snakebite management . The use of an affinity purification strategy makes it possible to obtain SSAbs from antivenoms , thereby eliminating cross-reactive antibodies and preventing false-positive results in assays of snake venoms . This approach obviates the need to produce additional polyclonal or monoclonal antibodies , and alleviates concerns regarding whether the antigens targeted by the polyclonal or monoclonal antibodies produced are species specific . SSAbs purified from antivenoms are suitable for use in developing sandwich ELISAs and lateral flow assays for rapid detection of snake venoms . The ability of these purified SSAbs to detect venom in the blood of animal models as well as in blood samples taken from snakebite patients validates the usefulness of this strategy . In conclusion , our data indicate the feasibility of a cost-effective approach ( i . e . preparation of SSAbs from specific antivenoms available in Taiwan ) to develop the snakebite diagnostic assay for discriminating hemorrhagic and neurotoxic snake envenomation in Taiwan . When combining the clinical observation of patient’s symptom , this assay would aid in the clinical decision of the appropriate antivenom to be used where the signs and symptoms of the envenoming did not allow a precise diagnosis by the clinician responsible to treat the envenomed patient . Although our present results are promising , further studies including improvement of detection sensitivity/specificity of the assays and application of the optimized assays to a larger sample set are needed to validate the clinical utility of the assays for snakebite management . | Snakebite is a public health issue that causes life-threatening medical emergencies . Rapid diagnosis of snakebite in the clinic is a critical necessity in many tropical and subtropical countries , where various venomous snakes are common . Venoms from different snake species contain distinct protein components that require treatment with different antivenoms . However , given the similarity in clinical symptoms among some snake envenomations , it is often challenging for physicians to precisely define the snake species responsible for envenomation . Thus , a reliable method or assay for rapidly diagnosing envenoming species is urgently needed . Here , we present a two-step affinity purification procedure for generating species-specific antibodies ( SSAbs ) from antivenom , followed by the development of a sandwich ELISA ( enzyme-linked immunosorbent assay ) and lateral flow strip assay using these SSAbs . This feasible and cost-effective strategy allowed us to develop workable assays for distinguishing between venom proteins from hemorrhagic and neurotoxic snakes in Taiwan . The usefulness of this strategy was demonstrated in the clinic , where both diagnostic assays were shown capable of detecting venoms in blood samples from snakebite patients . Together with the observation of clinical symptoms , the two-test-line lateral flow strip assay is potentially applicable in an emergency room setting to improve snakebite diagnosis and management . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"toxins",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"body",
"fluids",
"tropical",
"diseases",
"geographical",
"locations",
"vertebrates",
"animals",
"toxicology",
"toxic",
"agents",
"animal",
... | 2018 | Development of sandwich ELISA and lateral flow strip assays for diagnosing clinically significant snakebite in Taiwan |
Egress of the malaria parasite Plasmodium falciparum from its host red blood cell is a rapid , highly regulated event that is essential for maintenance and completion of the parasite life cycle . Egress is protease-dependent and is temporally associated with extensive proteolytic modification of parasite proteins , including a family of papain-like proteins called SERA that are expressed in the parasite parasitophorous vacuole . Previous work has shown that the most abundant SERA , SERA5 , plays an important but non-enzymatic role in asexual blood stages . SERA5 is extensively proteolytically processed by a parasite serine protease called SUB1 as well as an unidentified cysteine protease just prior to egress . However , neither the function of SERA5 nor the role of its processing is known . Here we show that conditional disruption of the SERA5 gene , or of both the SERA5 and related SERA4 genes simultaneously , results in a dramatic egress and replication defect characterised by premature host cell rupture and the failure of daughter merozoites to efficiently disseminate , instead being transiently retained within residual bounding membranes . SERA5 is not required for poration ( permeabilization ) or vesiculation of the host cell membrane at egress , but the premature rupture phenotype requires the activity of a parasite or host cell cysteine protease . Complementation of SERA5 null parasites by ectopic expression of wild-type SERA5 reversed the egress defect , whereas expression of a SERA5 mutant refractory to processing failed to rescue the phenotype . Our findings implicate SERA5 as an important regulator of the kinetics and efficiency of egress and suggest that proteolytic modification is required for SERA5 function . In addition , our study reveals that efficient egress requires tight control of the timing of membrane rupture .
Malaria is caused by protozoan parasites of the genus Plasmodium . Sporozoites introduced into the human host by the mosquito vector migrate to the liver where the parasite replicates to produce merozoites . These are released into the bloodstream , initiating the asexual blood stage of the infection in which the parasite goes through multiple rounds of intraerythrocytic replication and host cell destruction , producing gradually increasing parasitaemia that eventually leads to clinical disease . Like many intracellular pathogens , the parasite replicates within an intracellular membrane-bound compartment called a parasitophorous vacuole ( PV ) . Replication is by schizogony , in which formation of a multinucleated schizont occurs before a budding or segmentation process generates the individual daughter merozoites . In Plasmodium falciparum , the most widespread agent of fatal malaria , each intraerythrocytic replication cycle takes ~48 h , with production of 16 or more merozoites per schizont . During schizont development , a number of soluble parasite proteins accumulate within the PV lumen . Amongst the best characterized of these is a set of proteins belonging to the serine rich antigen ( SERA ) family , so named due to the presence of 27 or more consecutive Ser residues within the first-recognized member of the family , variously called 111 K antigen , P126 , P140 , SERA or SERP I but now referred to as SERA5 in P . falciparum ( PlasmodDB ID PF3D7_0207600 ) . A unifying feature of the SERA proteins , first noticed in SERA5 by Higgins et al . [1] then confirmed by x-ray crystallographic determination of the P . falciparum SERA5 central domain by Hodder and colleagues [2] , is their possession of a central domain homologous to papain-like cysteine peptidases ( clan CA , family C1 ) . SERA family members are found in all Plasmodium genomes examined [3] , and whilst the number of genes varies depending on the species , in all cases they fall into two classes: those that encode a Cys residue at the position equivalent to the nucleophilic Cys25 of papain ( Cys-type ) ; and those that possess a Ser codon at this position ( Ser-type ) . Gene disruption analysis of the 9 P . falciparum SERA genes suggested that only two , SERA5 ( Ser-type ) and SERA6 ( Cys-type ) , are important in the haploid asexual blood stage parasites [4–6] , implying crucial roles for SERA5 and SERA6 in this clinically relevant part of the parasite life cycle . Very recent work using conditional mutagenesis has confirmed that disruption of the P . falciparum SERA6 gene is lethal [7] . Release ( egress ) of daughter merozoites from the infected erythrocyte has long been known to be sensitive to cysteine protease inhibitors , including the selective covalent modifier trans-epoxysuccinyl-L-leucylamido ( 4-guanidino ) butane ( E64 ) ( e . g . [8] ) . The resemblance of the SERA proteins to cysteine proteases , together with their subcellular localisation in the PV in both blood stages [4 , 9 , 10] and liver stages [11] , has spurred interest in the possibility of the SERA proteins playing a role in egress . In support of this , disruption of a SERA family member that is highly expressed in mosquito stages of the parasite life cycle produced a defect in release of sporozoites from oocysts , structures on the basal surface of the insect midgut in which sporozoite biogenesis occurs [12] . Additionally , SERA5 , which is the most abundantly-expressed family member in P . falciparum blood stages , was shown in early studies to be subjected to extensive proteolytic processing that coincided temporally with and was dependent upon egress [13–16] , suggesting a link between SERA5 function and egress . Subsequent work in P . falciparum has uncovered the mechanism underlying this proteolytic processing . Minutes before egress , activation of a parasite cGMP-dependent protein kinase called PKG leads to the discharge of specialised merozoite organelles called exonemes which contain a subtilisin-like serine protease called SUB1 [17–19] . Upon secretion into the PV lumen , SUB1 cleaves both SERA5 and SERA6 at 2 or 3 discrete positions , releasing their central papain-like domains [10 , 17] . The available evidence suggests that SERA6 possesses proteolytic activity that is activated by this processing [10] . In contrast , SERA5 has a non-enzymatic role in the parasite , since mutations predicted to abolish catalytic activity ( e . g . substitution with Ala of the putative nucleophilic Ser596 ) had no phenotypic effect , whilst similar approaches failed to obtain viable parasites possessing a disrupted SERA5 gene [6] . Thus there is considerable evidence that SERA5 and SERA6 have important functions in the parasite . However , whether any member of the SERA family has a role in blood stage egress and what that role may be , is unresolved . Here we have used a conditional genetic approach combined with selective pharmacological tools to provide the first experimental evidence that SERA5 regulates egress in asexual blood stages of P . falciparum . Remarkably , rather than acting as a mediator of egress as previously suspected , we show that SERA5 enhances egress efficiency by acting as a negative regulator of the kinetics of egress .
In previous work [20] we described the production of a transgenic P . falciparum 3D7-derived clone called 1G5DiCre ( here abbreviated to 1G5DC ) . These parasites possess a modified , partially recodonised ( chimeric ) SERA5 gene encoding the wild type SERA5 amino acid sequence , followed by a single chromosomally-encoded loxP site and an integrated DiCre expression cassette . The latter drives constitutive expression of two individual , enzymatically inactive domains of Cre recombinase , each fused to a different rapamycin ( RAP ) -binding protein , such that addition of RAP induces heterodimerization of the proteins and Cre recombinase activity [21–23] . SERA5 protein expression by 1G5DC parasites is at wild type levels and the parasites replicate and egress normally in culture . Importantly , production of the 1G5DC clone included a step in which the human dihydrofolate reductase ( hdhfr ) drug resistance marker used to select for the desired homologous recombination event was removed from the genome by DiCre-mediated excision . The 1G5DC parasites are therefore fully sensitive to the antifolate drug WR99210 [20] . This recycling step allowed us to reuse the hdhfr selectable marker in a second gene targeting step ( Fig 1A ) in which a further loxP site was introduced into the parasite genome by targeted homologous recombination at the 3′ end of the upstream SERA4 gene , reconstituting the SERA4 gene whilst effectively floxing the entire chimeric SERA5 locus . Two of the resulting transgenic parasite clones , called floxSERA5-1B6 and floxSERA5-3B6 were selected for further analysis . Examination by diagnostic PCR confirmed the expected genomic architecture ( Fig 1A ) , whilst pulse-treatment of synchronised , newly-invaded ( ring-stage ) parasites with 100 nM RAP for just 1 h resulted in the expected DiCre-mediated excision event , rapidly deleting the entire chimeric SERA5 coding sequence with high efficiency within a single erythrocytic cycle ( Fig 1B ) . To assess the effects of SERA5 gene deletion on SERA5 protein expression , mature floxSERA5-1B6 and floxSERA5-3B6 schizonts were examined by indirect immunofluorescence ( IFA ) at the end of the first erythrocytic cycle ( ~44 h ) following RAP-treatment . No SERA5-specific signal was detectable by IFA in most of the RAP-treated floxSERA5-1B6 and floxSERA5-3B6 schizonts ( Fig 1C ) . Exhaustive microscopic examination showed apparently normal SERA5 expression in ~2% of schizonts ( S1A Fig ) , likely representing a small population of parasites in which gene excision had not occurred . This was confirmed by Western blot analysis of schizonts of the RAP-treated floxSERA5-1B6 and floxSERA5-3B6 populations , indicating an overall >95% reduction of SERA5 expression ( Fig 1D and S1B Fig ) . RAP-treatment had no detectable effect on expression of an unrelated merozoite protein , MSP1 ( Fig 1C and 1D and S1B Fig ) , nor on expression of the SERA4 and SERA6 genes , which flank the SERA5 gene ( S1C Fig ) [3 , 5] . Schizont morphology and merozoite numbers in the ΔSERA5 parasites were normal by light microscopy of Giemsa-stained preparations ( S1D Fig ) , and detailed visual examination of schizont SDS extracts fractionated on Coomassie-stained gels showed no detectable effects of RAP-treatment on the total parasite protein profile except for the noticeable absence of a ~120 kDa species identified by Western blot as full-length SERA5 ( S1E Fig ) . These results were consistent with the genetic data , indicating specific , rapid and efficient DiCre-mediated disruption of SERA5 expression . To establish the effects of SERA5 loss on parasite growth , we compared the replication rates of mock-treated and RAP-treated floxSERA5-1B6 and floxSERA5-3B6 clones , using the parental 1G5DC clone as a control . No morphological or growth differences were evident in the ~44 h immediately following RAP-treatment ( referred to as cycle 0 ) , and all the parasites matured to schizont stage at the same rate ( Fig 1C , S1A and S1D Fig ) with no differences in the number of merozoites produced per schizont ( S1D Fig; also see below ) . However , a ~50% reduction in the number of intracellular parasites was evident in the RAP-treated floxSERA5-1B6 and floxSERA5-3B6 cultures by the middle of the next cycle , and further monitoring into cycle 2 revealed a clear replication defect in these cultures ( Fig 2A ) . Over a more prolonged period , periodic examination of the RAP-treated floxSERA5-1B6 and floxSERA5-3B6 cultures by IFA and diagnostic PCR indicated a time-dependent increase in the proportion of SERA5-expressing parasites in the cultures ( Fig 2B and 2C ) , suggesting that the initially small population of non-excised parasites gradually overgrew the cultures , likely as a result of a selective advantage conferred on them by the replication defect displayed by the ΔSERA5 parasites . However , even after 12 erythrocytic growth cycles ( 24 days ) , ΔSERA5 parasites were still detectable in the RAP-treated cultures ( Fig 2C , lower right-hand panels ) , proving that whilst loss of SERA5 expression severely impacted the rate of parasite replication , it did not completely abolish it . To define the point ( s ) in the erythrocytic cycle affected by loss of SERA5 expression , parallel cultures of synchronous RAP-treated or mock-treated floxSERA5-1B6 or floxSERA5-3B6 mature schizonts at identical parasitaemia ( 11% ) at the end of cycle 0 were allowed to undergo egress and invasion for just 4 h , under either normal static conditions or whilst continuously and vigorously shaken . Following removal of residual intact schizonts by centrifugation over Percoll cushions , the newly-formed ring-stage parasites were enumerated at once by FACS . Subsequent intracellular development ( in static culture ) of these parasites was also monitored by FACS approximately 30 h later , as well as by microscopic examination . As shown in Fig 2D , shaking the cultures during the invasion period increased ring formation in all cases , implying more efficient egress and/or invasion as previously observed by others [24] , but production of rings was consistently reduced in the RAP-treated cultures compared to their mock-treated counterparts . However , once formed , these rings developed similarly over the course of the second intraerythrocytic cycle , irrespective of their provenance . Together with the previous evidence that SERA5-null parasites form morphologically normal schizonts , these results convincingly suggested that the reduction in long-term replication rate associated with loss of SERA5 was primarily or exclusively due to a defect not in intracellular parasite growth , but at the transition between schizont maturation and formation of new ring stage parasites . To quantify the effects of SERA5 loss on long-term parasite growth in the absence of parasites still expressing SERA5 , fresh floxSERA5-1B6 and floxSERA5-3B6 cultures were RAP- or mock-treated then identically diluted to obtain densities of <100 parasites/ml and transferred to flat-bottomed microwell plates . Examination of the plaques visible in these wells after 14–16 days ( which form as a result of localised erythrocyte destruction in the static erythrocyte layers [7] ) confirmed a growth defect in the ΔSERA5 parasites , with mock-treated floxSERA5-1B6 and floxSERA5-3B6 parasites producing significantly larger and more numerous plaques than their RAP-treated counterparts ( Fig 3A ) . Parasites from 2 separate cloning wells of the RAP-treated floxSERA5-1B6 and floxSERA5-3B6 parasites , each of which contained only a single small plaque , were individually expanded . The resulting parasite clones ( called 2F8_ΔSERA5 and C6_ΔSERA5 ) were confirmed as ΔSERA5 by Western blot ( Fig 3B ) . They were then compared to the parental clones ( not RAP-treated ) in growth assays ( Fig 3C ) . The results showed that loss of SERA5 results in a ~28% decrease in replication rate at each erythrocytic cycle . This difference in replication rates was rather less than the ~50% observed in static cultures immediately following SERA5 gene excision ( Fig 2A and 2D ) , suggesting the possibility of some adaptation of the 2F8_ΔSERA5 and C6_ΔSERA5 clones during the ~12 week period required for their cloning by limiting dilution and subsequent expansion . Nonetheless , it could be concluded unambiguously from these results that SERA5 is important for efficient blood stage parasite growth but is not essential for long-term viability in vitro . To gain insights into the defect underlying the reduced rates of new ring formation displayed by ΔSERA5 parasites , highly synchronous preparations of the floxSERA5-1B6 and floxSERA5-3B6 clones were RAP- or mock-treated at ring stage , then examined as they reached the point of egress . To enhance the synchrony of egress in these experiments , the purified mature schizonts ( ~43 h post-treatment ) were transferred for 4–6 h into medium containing one or other of the PKG inhibitors 4-[2- ( 4-fluorophenyl ) -5- ( 1-methylpiperidine-4-yl ) -1H-pyrrol-3-yl] pyridine ( compound 1 ) , or ( 4-[7-[ ( dimethylamino ) methyl]-2- ( 4-fluorphenyl ) imidazo[1 , 2-α]pyridine-3-yl]pyrimidin-2-amine ( compound 2 ) both of which potently but reversibly block parasite development at a stage just preceding egress [18 , 25] . Short-term treatment of schizonts with either compound results in the accumulation of ‘stalled’ highly mature schizonts in which the PVM is intact though often porous [26] . Subsequent wash-out of the inhibitors releases the egress block , allowing the schizonts to proceed to rupture within minutes and facilitating time-lapse video microscopic examination of multiple egress events over a relatively short period [18 , 26 , 27] . As shown in Fig 4A–4C , as well as S1 Movie and S2 Fig , the RAP-treated floxSERA5-1B6 and floxSERA5-3B6 parasites displayed a dramatic egress phenotype characterized by abnormally rapid membrane rupture and inefficient dispersal of the released merozoites , which often initially formed asymmetric ‘clusters’ apparently due to them being transiently trapped within partially ruptured bounding membranes . The ΔSERA5 merozoites generally gradually dispersed from these clusters over a period of several minutes , proving that both the PV and host erythrocyte membranes had undergone rupture ( Fig 4C and S2 Fig ) . Indeed , in many cases , PVM rupture clearly occurred shortly before erythrocyte membrane rupture in both mock-treated and ΔSERA5 parasites . This was indicated by transient rounding-up of the schizonts with swelling of the PV , then sudden loss of differential interference contrast of the cell and increased visibility and motility of the internal merozoites ( S1 Movie ) , as previously observed in both P . falciparum [26 , 28–30] and the zoonotic malarial species P . knowlesi [31] . The total proportion of schizonts that underwent rupture over the imaging periods did not differ between parallel mock and RAP-treated cultures , confirming that egress was not inhibited by loss of SERA5 expression . In both mock and RAP-treated schizonts , egress was often accompanied by sudden lateral displacement of the schizont by as much as 5–10 μm , perhaps indicating rapid expulsion of the cytosolic contents of the cell upon the release of the osmotic pressure proposed by some to be responsible for the pre-egress PV swelling [28 , 32] . The time interval between washing away the PKG inhibitors and initiation of egress in control ( mock-treated ) parasites differed somewhat between the PKG inhibitors , with a rather longer delay in the case of compound 1 ( minimum delay to egress ~13 mins ) compared to compound 2 ( minimum time to egress ~7 . 5 min ) . However , in all cases the ΔSERA5 parasites exhibited significantly accelerated membrane rupture compared to similarly treated controls . This was confirmed by Coomassie-staining and Western blot analysis , which showed more rapid appearance of soluble parasite-derived proteins in the supernatants of RAP-treated floxSERA5-1B6 schizonts released from a compound 2-mediated egress block than in the mock-treated samples ( Fig 4D ) . In contrast , Western blot analysis of the schizonts themselves showed no discernible difference in the rates of proteolytic processing of two established SUB1 substrates , SERA6 and MSP1 ( S3 Fig ) . Importantly , the abortive egress phenotype and formation of merozoite ‘clusters’ was also observed in mature ΔSERA5 schizonts that had not been synchronised by treatment with compounds 1 or 2 , confirming that the phenotype was not an artefact associated with prior exposure to the PKG inhibitors ( S2 Movie ) . Time-lapse microscopic examination of the 2F8_ΔSERA5 and C6_ΔSERA5 clones compared with RAP-treated parental 1G5DC parasites ( S3–S5 Movies ) confirmed the premature egress phenotype in the ΔSERA5 clones and showed that these parasites had not undergone major compensatory phenotypic alterations during the extended culture period ( ~12 weeks ) required for their isolation by limiting dilution cloning and expansion . This phenotype was still evident even after extended continuous culture ( >16 months ) of these parasite clones . It was concluded that loss of SERA5 results in accelerated schizont rupture but defective egress , and that this likely explains the replication defect evident in the ΔSERA5 parasites . Elegant previous studies by Glushakova et al . and others in P . falciparum have demonstrated that , in the seconds just preceding host red blood cell membrane breakage at egress , the membrane becomes permeabilized or ‘porated’ , allowing leakage of cellular contents and the ingress of extracellular molecules [26 , 32 , 33] . The latter includes the F-actin-binding peptide phalloidin which , if introduced into the surrounding culture medium , is able just before egress to access and label the actin protofilaments of the cytoskeleton that underlies the host erythrocyte cell membrane . By allowing schizont rupture in the presence of fluorescent phalloidin , poration can be detected by microscopy as the sudden appearance of a fluorescent signal around the host cell circumference [32 , 33] . In addition , phalloidin binding to the ruptured host erythrocyte cytoskeleton immediately after egress enables visualisation of the state of the fragmented membranes , which generally undergo extensive vesiculation [28 , 32] . Host erythrocyte membrane rupture ( but not PVM rupture ) is selectively blocked by the cysteine protease inhibitor E64 but poration of the membrane is not [26 , 32] , so the combined use of E64 and fluorescent phalloidin provides a simple means of reliably observing the onset and efficiency of poration , since the porated schizonts do not rupture . To examine whether loss of SERA5 expression impacts on poration and fragmentation of the erythrocyte membrane , compound 2-synchronised , mock or RAP-treated floxSERA5-3B6 schizonts were washed and transferred to fresh medium containing Alexa Fluor 488-labeled phalloidin , with or without E64 . Simultaneous time-lapse DIC and fluorescence imaging of the E64-containing cultures showed the gradual appearance of phalloidin-labelled schizonts as the host cell membrane of these cells became porated ( Fig 5A ) . Intriguingly , the appearance of phalloidin labelling in individual schizonts was always temporally coincident with PVM rupture , the latter being clearly evident by DIC imaging as the sudden loss of differential interference contrast of the schizont and increased visibility and mobility of the intracellular merozoites ( S6 Movie ) . Of additional interest , compared to the kinetics of egress in parallel samples lacking E64 which progressed rapidly to rupture with kinetics as described in Fig 4 , phalloidin labelling in the presence of E64 was substantially delayed in both control and ΔSERA5 samples ( compare Fig 5B with Fig 4B ) . This is the first evidence to our knowledge that although E64 does not prevent PVM rupture , it delays the process . Whilst the rate of appearance of phalloidin-labelled schizonts in the presence of E64 was slightly faster in the ΔSERA5 schizont population ( Fig 5A and 5B ) , this was only just significant , in contrast to the clear premature rupture phenotype displayed by the ΔSERA5 parasites in the absence of E64 . The total proportion of schizonts labelled with phalloidin within a ~35 min period following removal of the compound 2 block in the presence of E64 was similar for the ΔSERA5 and control cultures ( Fig 5B ) . Time-lapse DIC examination of E64-containing cultures in the absence of phalloidin detected no significant difference between the timing of PV rupture in the control and ΔSERA5 populations ( S4 Fig ) . In cultures containing phalloidin but lacking E64 , extensive vesiculation of the residual host cell membranes was observed in all cases following egress ( Fig 5C ) . Collectively , these results showed that SERA5 does not play a direct role in mediating host cell membrane poration or vesiculation . However , the fact that in the presence of E64 the rate of progress to PVM rupture and host cell membrane poration in the ΔSERA5 parasites was very similar to that of control parasites , implied a potential link between SERA5 function and the protein target ( s ) of E64 . This point is further discussed below . The genes encoding the Ser-type SERA family members SERA1-5 in P . falciparum lie in a head-to tail tandem array on chromosome 2 [3 , 5] . This , together with the high degree of synteny between SERA loci in all Plasmodium species examined , suggests that the Ser-type SERA genes arose through gene duplication events , the number of which varied between different Plasmodium species [3] . Phylogenetic analyses have suggested that all the Ser-type SERA proteins may mediate similar functions [3 , 5] . However , expression levels vary widely , with both transcriptional and proteomic data indicating that in P . falciparum SERA5 is the most highly expressed , whilst SERA4 is the second most abundantly expressed [4 , 34 , 35] and is dispensable in blood stages [5] . In the light of our success in flanking the SERA5 gene on the 1G5DC background with loxP sites using a single-crossover homologous recombination approach , we decided to employ a modified strategy to simultaneously flox both the SERA5 and SERA4 genes with the aim of examining the consequences of disrupting both genes simultaneously . As shown in S5 Fig , a construct ( pSERA3loxP ) designed to insert a loxP site immediately downstream of the SERA3 gene successfully integrated in the expected manner into the 1G5DC genome . Treatment with RAP of a parasite clone ( called floxSERA4/5-B52 ) harbouring the integrated construct resulted in excision of both the SERA4 and SERA5 genes as shown by diagnostic PCR analysis ( S5 Fig ) . The resulting parasites were cloned by limiting dilution and two clones obtained . Examination by Western blot of mature schizonts of these clones ( called C10_ΔSERA4/5 and G8_ΔSERA4/5 ) confirmed the absence of both SERA4 and SERA5 but unaltered expression of SERA3 and SERA6 , as expected ( Fig 6A ) . To assess the effects of simultaneous loss of both SERA4 and SERA5 on parasite egress , we examined the timing and morphology of egress of the floxSERA4/5-B52 parasites following mock-treatment or treatment with RAP . As shown in S7 Movie and S8 Movie , the ΔSERA4/5 parasites displayed an accelerated but defective egress phenotype identical to that of ΔSERA5 parasites . Growth assays comparing the replication rates of the ΔSERA4/5 clones with the ΔSERA5 clones showed no significant differences in replication rates over the course of 3 erythrocytic cycles ( Fig 6B ) . It was concluded that simultaneous loss of both SERA4 and SERA5 produces a defect in egress no more severe than that resulting from loss of SERA5 alone . The very similar phenotype displayed by the ΔSERA5 and the ΔSERA4/5 parasites , together with the conditional nature of the gene disruption strategy and the apparently unaltered expression of the flanking SERA3 and SERA6 genes in both sets of knockouts , gave us a high degree of confidence that the observed phenotype was a direct result of manipulation of the SERA4/5 loci and not due to unintended genetic alterations in the modified parasites . To confirm this and to investigate the structural requirements for SERA5 function , we sought to rescue the ΔSERA5 phenotype by genetic complementation . For this , we introduced into 2F8_ΔSERA5 parasites a transgene construct ( called pDC2_mC_sgS5 ) designed for episomal expression of the SERA5 gene under the control of endogenous flanking 5’ sequence likely to include the genomic SERA5 promoter ( Fig 7A ) . To reliably identify the transgene product , the SERA5 transgene was modified by the introduction of an internal mini TAP tag sequence incorporating a hemagglutinin ( HA3 ) epitope tag , that we had previously shown does not interfere with SERA5 function in the parasite , despite abolishing the SUB1 site 2 processing site [6] . The pDC2_mC_sgS5 construct also contained a cassette for constitutive expression of cytoplasmic mCherry to facilitate identification of parasites harbouring the construct . In parallel , 2F8_ΔSERA5 parasites were independently transfected with a mutant of the same episome , called pDC2_mC_sgS5mut , in which the SERA5 sequence encoding the site 1 SUB1 processing site [17] , as well as a site that is cleaved by an anonymous cysteine protease referred to as protease X [6] , were modified in a manner designed to block correct cleavage of the transgene product ( Fig 7A ) . A third set of 2F8_ΔSERA5 parasites were transfected with a similar construct that contained the mCherry expression cassette but lacked the SERA5 expression cassette . Transfected parasites were subjected to selection in the presence of blasticidin . IFA and Western blot analysis of the resulting three blasticidin-resistant lines ( called 2F8_ΔSERA5:SERA5wt , 2F8_ΔSERA5:SERA5mut and 2F8_ΔSERA5:mCherry ) showed mCherry expression in all cases ( although not in every parasite; see below ) and confirmed expression of the transgenic SERA5 proteins in the 2F8_ΔSERA5:SERA5wt and 2F8_ΔSERA5:SERA5mut lines ( Fig 7A and 7B ) , as well as the expected SERA5 processing defect in the mutant line ( Fig 7B ) . To examine the effects of transgene expression on the egress phenotype , egress of synchronous mature schizonts of the 2F8_ΔSERA5:SERA5wt and 2F8_ΔSERA5:SERA5mut parasite lines following release of a compound 2-mediated block was visualised by simultaneous live DIC and fluorescence time-lapse imaging as previously . Substantial variation in the levels of mCherry expression was evident in the imaged schizonts , likely due to heterogeneity in episome segregation as previously noted for episomal plasmids in P . falciparum [36 , 37] . However comparison of the fluorescent with the non-fluorescent schizonts in each population clearly showed rescue of the delayed ( Fig 7C ) and abortive ( S9 Movie and S10 Movie ) egress phenotype in the 2F8_ΔSERA5:SERA5wt parasites but not in the mutant line . It was concluded that correct proteolytic processing of SERA5 is likely essential for its function in regulating the kinetics of egress .
Like all obligate intracellular pathogens , the malaria parasite has evolved efficient molecular strategies to exit from its host cell . In this study we have made inroads into our understanding of the regulatory mechanisms underlying blood stage egress in P . falciparum , showing that conditional disruption of the SERA5 gene , or both the SERA4 and SERA5 genes simultaneously , leads to a dramatic egress defect characterised by premature but inefficient membrane rupture . The released daughter merozoites disseminate slowly , apparently due to their being retained transiently within residual , incompletely fragmented bounding membranes . The result is a reduction in invasion of new host erythrocytes . The indistinguishable phenotypes of the ΔSERA5 and ΔSERA4/5 mutants is consistent with the notion of functional redundancy amongst the Ser-type SERA genes and previous evidence that loss of SERA4 alone is tolerated by the parasite [5] . Our study provides two important insights into the molecular mechanisms underlying blood-stage egress in P . falciparum . The first concerns the kinetics of the egress pathway downstream of PKG activation . Our work confirms our previous observations [18 , 26] that there is a delay of several minutes between washing away the PKG inhibitors from compound 1 or compound 2-stalled schizonts , and initiation of egress . Here we quantified this period under the conditions used to a minimum of ~13 mins in the case of compound 1 and ~7 . 5 min for compound 2 . We had previously assumed that this delay to egress is primarily a reflection of the rate of dissociation of the respective inhibitor-PKG complexes , and that egress initiates as soon as release of the bound drug allows reactivation of sufficient PKG activity to trigger egress . However our new data force a re-evaluation of that simplistic assumption , because schizont membrane rupture consistently occurs more rapidly in the ΔSERA5 and ΔSERA4/5 mutants . The premature rupture is clearly PKG-dependent , since it is reversibly blocked by PKG inhibitors . It also involves SUB1 discharge ( Fig 4C ) , so our observations reveal for the first time that the delay to egress observed in wild-type parasites following removal of the PKG inhibitors includes a ‘lag phase’ that is additional to the time required for the parasites to become membrane rupture-competent following PKG activation and SUB1 discharge into the PV ( Fig 8 ) . What is the function of this lag phase and how is it controlled ? The answers to these questions may lie in the second important insight provided by our study . The simplest interpretation of the premature rupture phenotype is that SERA5 ( and also perhaps SERA4 ) is a negative regulator of the egress pathway , acting to maintain the lag phase and thus delay rupture of the PVM and host erythrocyte membranes following SUB1 discharge . The fact that the premature egress phenotype is also associated with a defect in merozoite dissemination suggests that the function of the lag phase is to ensure proper destabilisation of the bounding membranes and/or collapse of the host cell cytoskeleton [26] such that , once they rupture , the merozoites can easily and rapidly escape from the residual membranes . Elegant high speed video microscopy and modelling experiments by Abkarian and colleagues [30] has shown that the scattering of merozoites normally observed at schizont rupture is at least in part due to the elastic properties of the erythrocyte membrane which , upon primary rupture at a single site , undergoes extremely rapid curling and buckling , turning inside-out to literally fling the remaining enclosed merozoites outwards . This demonstrably does not happen in the ΔSERA5 and ΔSERA4/5 mutants . We suggest that in the absence of the SERA5-regulated lag phase , partial membrane rupture occurs before critical intracellular membrane modification events–perhaps those that lead to cytoskeleton collapse [26] or that provide the spontaneous membrane curvature to the erythrocyte membrane thought to be required for the elastic phenomenon described above [30 , 38 , 39]—have had time to go to completion . Crucially , as soon as the membranes break in the mutants , further membrane modification probably ceases due to the premature leakage and dilution of the effector molecules that mediate these modifications . These would normally function in the confined space of the PV lumen and/or host cell cytosol in the critical few minutes between SUB1 discharge and final egress . These effector molecules may include proteases such as SUB1 itself ( which is released into the surrounding medium upon premature rupture of the ΔSERA5 parasites; Fig 4C ) , as well as SERA6 [10] and possibly host cell-derived calpain-I [40] , plus putative pore-forming proteins such as PLP1 [33] , which has been proposed as a candidate for mediating the cysteine protease-independent erythrocyte membrane poration that immediately precedes normal egress [32 , 33] , although recent gene disruption data refute this [41] . The ruptured but incompletely fragmented membranes transiently entrap the merozoites , severely impeding their efficient dissemination ( Fig 8 ) . Given the very short invasive half-life of free P . falciparum merozoites , which lies somewhere between 90–300 seconds at 37°C [29 , 42–45] , it is likely that any delay in the dispersal of free merozoites from the bounding membrane ( s ) of a ruptured schizont impacts on invasion efficiency . Whilst vigorous shaking during egress increased the efficiency of invasion and ring formation , presumably due to more efficient dispersal of the released merozoites , even this could not completely rescue the phenotype . We propose that the egress defect is the primary cause of the poor replication rates observed in the ΔSERA5 and ΔSERA4/5 mutants . However , we cannot rule out other possibilities , including that the ΔSERA5 merozoites may have an intrinsic invasion defect due for example to the absence of surface-bound N- and C-terminal SERA5 processing products , which have been shown to selectively bind merozoites following SUB1-mediated cleavage [9] . Unfortunately , distinguishing between these alternative models is experimentally challenging , and attempts by us to assess the invasive capacity of isolated merozoites from the ruptured ΔSERA5 schizonts were inconclusive . We can only speculate on the molecular mechanisms underlying SERA5 function . As one of the most abundant PV proteins , SERA5 may simply act as an abundant SUB1 substrate or ‘sink’ , competing with SUB1-mediated cleavage of other key substrates in the moments following SUB1 discharge and thus slowing down the kinetics of the egress pathway . Arguing strongly against this is our inability to discern any changes in the rate of processing of the known SUB1 substrates MSP1 and SERA6 in the ΔSERA5 mutants ( S3 Fig ) . An alternative model rests on the fact that SERA5 bears structural resemblance to a cysteine protease [2] and indeed displays peptidase activity when the ‘active site’ Ser596 is replaced by a Cys residue [6] . Catalytically inactive enzyme orthologues , including pseudokinases and pseodoproteases , are common , often thought to have evolved from their catalytically active cognate enzymes [46–49] and sometimes existing in regulatory complexes with them [50 , 51] . SERA5 may likewise act in concert with a related partner enzyme . A good candidate partner is the egress-related enzyme SERA6 , a proposition supported by phylogenetic analyses suggesting that the ancestral gene from which all the Ser-type SERA genes evolved was a Cys-type gene likely corresponding to either SERA6 or SERA7 in P . falciparum [3] . However , exhaustive attempts by us to identify interactions between SERA5 and SERA6 or any other candidate partners by co-immunoprecipitation have been unsuccessful; for example , we can readily obtain chromatographically pure SERA5 from parasite extracts in both full-length [17] and processed form [6] under non-denaturing conditions at neutral pH , suggesting that SERA5 does not form high affinity associations with other soluble parasite or host proteins . A third model for SERA5 function , related to the second , is that SERA5 might act by binding and protecting substrates of one or more egress-related proteases , transiently inhibiting cleavage of those substrates . The binding of SERA5 might be released by SUB1 and/or protease X-mediated cleavage of SERA5 , explaining the importance of processing for SERA5 function . Notably , either of these pseudoprotease roles for SERA5 would be consistent with our observation in this study that in the presence of E64 the rate of progression to PVM rupture was independent of SERA5 expression . This is because if the function of SERA5 is to control the enzymatic rate of a related cysteine protease against critical physiological substrates , then pharmacological inhibition of that cysteine protease would be expected to negate the effects of loss of SERA5 . Elucidating the molecular details of SERA5-mediated regulation will be a key priority of further work . Prior to this work , all attempts to directly disrupt the P . falciparum SERA5 gene using conventional targeted homologous recombination had failed [4–6] , leading to the conclusion that the gene is indispensable for in vitro growth . Our results here , using a robust conditional gene modification system , clearly demonstrate that this is not the case , although we also show that due to their replication defect ΔSERA5 parasites can be rapidly out-competed in vitro by wild type parasites . We suspect that the previous lack of success in obtaining SERA5-null parasites was due to the inefficient nature of homologous recombination in P . falciparum , combined with the technical difficulties of isolating blood stage mutants with a growth defect because of their selective disadvantage in the presence of wild-type parasites . We expect that future work using DiCre and similar efficient conditional approaches will enable the isolation of many malarial mutants lacking expression of genes previously considered to be essential . As in this case , the resulting phenotypes will provide valuable insights into the biology of this important pathogen .
The antifolate drug WR99210 was from Jacobus Pharmaceuticals ( New Jersey , USA ) . Blasticidin , E64 and rapamycin were from Sigma . Rapamycin was used as described previously [20] . The PKG inhibitor compound 1 was a kind gift of David Baker , London School of Hygiene & Tropical Medicine , whilst compound 2 was kindly provided by Dr Simon Osborne , MRC Technology , London NW7 1AA , UK . Stocks of both drugs were stored in dry DMSO at -20°C , and were used throughout at final concentrations of 2 μM ( compound 1 ) and 1 μM ( compound 2 ) . Monoclonal antibody ( mAb ) 89 . 1 , which recognises the merozoite surface protein MSP1 , has been described previously [52] , as has the human anti-MSP1 mAb X509 [53] , rabbit polyclonal antisera to P . falciparum SERA5 [6] and SERA6 [10] , and a rabbit polyclonal antiserum against P . falciparum SUB1 [54] . The anti-SERA5 mAb NIMP . M13 was produced using a previously described method [18] , using B-cells from BALB/c mice ( bred in the pathogen-free animal facility of the Medical Research Council National Institute for Medical Research , Mill Hill , London ) immunised with recombinant full-length SERA5 [6] . Rabbit antibodies specific to relatively non-conserved regions of P . falciparum SERA3 and SERA4 ( anti-SE3N and anti-SE4N ) were kind gifts of Professor Toshihiro Horii , Osaka University , Japan . Asexual blood stages of the DiCre-expressing P . falciparum clone 1G5DC [20] were cultured in an atmosphere of 90% nitrogen , 5% carbon dioxide and 5% oxygen at 37°C in RPMI 1640 medium containing Albumax ( Invitrogen ) supplemented with 2 mM L-glutamine , and synchronised using standard procedures [55 , 56] . Parasite developmental stage and viability was routinely assessed by microscopic examination of Giemsa-stained thin blood films . For transfection of integration constructs , Percoll-enriched synchronised mature schizonts were electroporated with plasmid DNA ( 10 μg per transfection ) using an Amaxa P3 primary cell 4D Nucleofector X Kit L ( Lonza ) as described [27] . Growth medium was replaced ~20 h post transfection with fresh medium containing 2 . 5 nM WR99210 . Once drug-resistant parasites appeared and displayed robust growth ( 2–3 weeks post-transfection ) , they were subjected to repeated cycles of culture for 3 weeks without drug followed by culturing with drug ( drug cycling ) to select for parasites in which integration into the genome had taken place [56] . Transgenic parasite clones floxSERA5-1B6 , floxSERA5-3B6 and floxSERA4/5-B52 were obtained by limiting dilution cloning in round-bottomed wells at a calculated 0 . 1–0 . 3 parasite per well as described [56] . Following RAP-treatment of these parasites to induce gene disruption , clones of ΔSERA5 or ΔSERA4/5 parasites were obtained by limiting dilution in flat-bottomed 96-well microplate wells as described [7] , plating a calculated 10 parasites per well . Only wells containing single plaques were subsequently used . Once established , all transgenic clones except the parental 1G5DC clone were maintained in medium containing 2 . 5 nM WR99210 . For IFA , air-dried thin films of parasite cultures were fixed in paraformaldehyde , permeabilized , then probed with relevant primary antibodies as described previously [10] . Secondary Alexa Fluor 488 or 594-conjugated antibodies specific for human , rabbit or mouse IgG ( Invitrogen ) were used at a dilution of 1:10 , 000 . Samples were stained with 4 , 6-diamidino-2-phenylindol ( DAPI ) for nuclear staining then mounted in Citifluor ( Citifluor Ltd . , UK ) . Images were acquired using a Zeiss Axioplan 2 Imaging system ( Carl Zeiss , Germany ) and AxioVision 3 . 1 software , using identical exposure conditions for all samples being compared . Western blots were prepared and probed as described previously [57] . A second loxP site was introduced into the 1G5DC genome upstream of the SERA5 locus by targeting it to a site just downstream of the SERA4 gene . To do this , plasmid pHH1_S4int_US-loxP was first generated by excising the fragment containing the loxP site immediately upstream of the SERA5 targeting region in plasmid pHH1_PreDiCre_A [20] using BamHI and HpaI . This fragment was ligated into pHH1-ΔSERA4 [4] pre-digested with the same restriction enzymes . To produce the targeting sequences , PCR products were amplified from 3D7 genomic DNA using Phusion HF DNA polymerase ( NEB ) with forward primers S4_F3 or S4_HpaI_F and reverse primer S4_XhoI_R , generating target regions 1 ( ~1 . 4 kb ) and 2 ( ~1 kb ) , respectively . The PCR products were blunt-ended using T4 DNA polymerase ( NEB ) then digested with XhoI and cloned into plasmid pHH1_S4int_US-loxP pre-digested with HpaI and XhoI to remove the internal SERA4 sequence , giving rise to plasmids pSERA4loxPa and pSERA4loxPb , respectively . The plasmids were independently transfected into mature 1G5DC schizonts and WR99210-resistant parasites subjected to drug cycling to select for integration into the genome . The resulting parasite lines were screened by diagnostic PCR using primers S4_F4 plus S4_DS_R1 to detect the unmodified 1G5DC SERA5 locus ( giving rise to a 1985 bp fragment ) and primers CAM5’_R3 plus S4_DS_R1 to detect the integrated locus ( giving rise to a 1941 bp fragment for target fragment 1 , or a 1535 bp fragment for target fragment 2 ) . Limiting dilution cloning of the integrated parasite lines resulted in clones floxSERA5-1B6 ( from integration using targeting fragment 2 ) and floxSERA5-3B6 ( from integration using targeting fragment 1 ) . A transgenic P . falciparum line for simultaneous conditional disruption of the SERA4 and SERA5 genes was generated by introducing a loxP site downstream of the SERA3 locus in the 1G5DC parasite clone , thus flanking both SERA4 and SERA5 genes simultaneously . To do this the SERA3 targeting region was amplified using forward primer S3_F1 and reverse primer S3_R1 to generate a 1 . 4 kb product . This was digested with SnaBI and XhoI and cloned into pHH1_S4int_US-loxP , pre-digested with HpaI and XhoI , generating construct pSERA3loxP . This was transfected into 1G5DC schizonts and drug-selected parasites screened for integration using primers CAM5’_R3 and S3_DS_R1 ( which produce an amplicon of ~1 . 7 kb ) . The presence of the endogenous locus was detected using primers S3_F6 and S3_DS_R1 giving rise to a product of 2 . 0 kb . Limiting dilution cloning of the integrant parasite population generated parasite clone floxSERA4/5-B52 . For both the SERA5 and SERA4/5 gene disruption experiments , following treatment of parasites with RAP ( 100 nM for 1–4 h at 37°C ) the non-excised locus was detected by diagnostic PCR using GoTaq Green ( Promega ) , using primers sgS5_seq4F and hsp86_3’_R1 ( amplicon size ~1200 bp ) whilst the excised locus was detected using primers CAM5’_R3 and hsp86_3’_R1 ( amplicon size 580 bp ) . The ΔSERA5 P . falciparum clones 2F8_ΔSERA5 and C6_ΔSERA5 were selected from RAP-treated conditional knockout floxSERA5-1B6 and floxSERA5-3B6 respectively . The ΔSERA4/5 clones C10_ΔSERA4/5 and G8_ΔSERA4/5 were similarly obtained by limiting dilution cloning of RAP-treated clone floxSERA4/5-B52 . Parasitaemia measurements by FACS were as described previously [6] . Briefly , parasites recovered at various time-points were stained with the fluorescent vital stain hydroethidine . As negative controls , uninfected erythrocytes in culture medium were stained and processed in the same way . Parasitaemia was calculated using the FACSCalibur flow cytometer ( Becton Dickson ) as described previously [58] . Briefly , cultures to be analysed were initially screened using forward and side scatter parameters and gated for erythrocytes . From this gated population , the proportion of HE-stained cells in 100 , 000 cells was determined using the FL2 detector ( 585/42 nm ) . Plaque assays were performed by dispensing parasite cultures in flat-bottomed microplates at a haematocrit of 0 . 75% , as described [7] . Plates were imaged using a high resolution flat-bed scanner 14–16 days after setting up the assays . Plaques were counted by visual examination of the images and plaque size quantified using the Lasso tool in Adobe Photoshop CS6 . Statistical analysis ( linear regression analysis by analysis of covariance and t-test ) was performed using GraphPad Prism 7 software ( CA , USA ) or tools available on the GraphPad calculation website ( http://www . graphpad . com/quickcalcs/ttest1/ ) . When required , parasites from wells containing a single plaque were expanded by transferring initially to round-bottomed microplate wells , before further expansion into culture flasks . For transgenic expression of SERA5 or mutants thereof in SERA5-null parasites , a SERA5 expression cassette under the regulation of its native promoter was initially assembled in pHH1_preDiCre_A [20] . This plasmid already contained the 3’ fragment of a previously described synthetic recodonised SERA5 gene , called SERA5synth [20] . The 5’ region of the SERA5 synthetic gene was excised from N-term-pMK by digesting with Pac1 and blunt ending with T4 DNA polymerase , followed by digestion with SalI . The resulting fragment was cloned into pHH1_preDiCre_A pre-digested with HpaI and SalI giving rise to plasmid pHH1_sg-sera5 . The SERA5 promoter region was then amplified from 3D7 genomic DNA with primers S5_US_F5 and S5_US_R3 using Phusion HF DNA polymerase and cloned into pHH1_sg-sera5 using AflII and SnaBI sites , giving rise to plasmid pHH1_S5-5’_sgS5 . To distinguish the synthetic gene product from that of the endogenous gene , a mini TAP tag was incorporated into the SERA5 coding region just downstream of the P50C ( protease X ) cleavage site [6] by cloning a SbfI and XmaI fragment including the mini TAP tag from pHH1SERA5chimΔP6TAP [6] into pHH1_S5-5’_sgS5 , generating plasmid pHH1_S5-5’_sgS5_mT . To obtain constitutive expression of mCherry from the SERA5 expression plasmid , the SERA5 expression cassette was then excised from pHH1_S5-5’_sgS5_mT using NotI-HF and SnaBI , blunt-ended with T4 DNA polymerase , and cloned into pDC2-mCherry ( a kind gift of Catherine Suarez , The Francis Crick Institute ) pre-digested with BamHI and blunt-ended using T4 DNA polymerase , giving rise to the wild-type complementation plasmid pDC2_mC_sgS5 . Mutagenesis of the SUB1 site 1 cleavage of SERA5 was introduced by carrying out two separate PCR reactions using primers sgS5_5’_F_AflII plus S5_st1II_R , and S5_st1II__F plus PbDT3’_R1 . This was followed by an overlapping PCR reaction in which the two PCR products were mixed and amplification carried out using the external primers sgS5_5’_F_AflII and PbDT3’_R1 . The resulting mutated DNA fragment was cloned into pHH1_S5-5’_sgS5_mT using BglII and AflII sites , giving rise to plasmid pHH1_S5-5’_sgS5/st1_mT . The SUB1 site 2 cleavage site of SERA5 was removed when the mini TAP tag was introduced into the SERA5 sequence [6] . To additionally mutate the protease X ( P50C ) cleavage site , two separate PCR reactions were carried out using primers S5synF1 plus p50_R_New , and p50_F_New plus PbDT3’_R1 . This was followed by an overlapping PCR using primers S5synF1 and PbDT3’_R1 . The resulting mutant amplicon was sub-cloned into pHH1_S5-5’_sgS5_mT using BglII and AvrII giving rise to plasmid pHH1_S5-5’_sgS5/p50+st2_mT . To produce a modified SERA5 gene containing mutations at both the SUB1 sites 1 and 2 as well as the P50C cleavage site , the SUB1 site 1 cleavage site mutation from pHH1_S5-5’_sgS5/st1_mT was cloned into plasmid pHH1_S5-5’_sgS5/p50+st2_mT using XcmI . The orientation of the insert was confirmed by sequencing , then the mutant DNA fragment sub-cloned into plasmid pDC2_mC_sgS5 using SnaBI and NotI giving rise to the final mutant complementation plasmid pDC2_mC_sgS5mut . For introduction of the complementation and control constructs into parasites , mature schizonts of the ΔSERA5 P . falciparum clone 2F8_ΔSERA5 were electroporated as described above with 10 μg of each construct pDC2_mC_sgS5 , pDC2_mC_sgS5mut or pDC2_mCherry and maintained initially in medium containing no drug . Approximately 24 h following electroporation , the parasites were transferred into medium supplemented with blasticidin ( 2 μg ml-1 ) then maintained until vigorous growth ensued . All lines were synchronised before use for Western blot , video microscopy or growth assays . Viewing chambers ( internal volume ~80 μl ) for observation of live schizonts were constructed as described [18] by adhering 22 x 64 mm borosilicate glass coverslips to microscope slides with strips of double-sided tape , leaving ~4 mm gaps at each end . Schizont samples were washed then suspended in warm , gassed complete medium , either alone or supplemented where required with E64 ( 50 μM final concentration ) , Alexa Fluor 488 phalloidin ( Invitrogen; diluted 1:50 from a 200 unit ml-1 stock in methanol ) , and/or Alexa Fluor 647-conjugated wheat germ agglutinin ( Thermo Fisher; diluted 1:1000 from a 1 mg ml-1 stock in phosphate-buffered saline ) . The schizont suspension was introduced into the pre-warmed chamber , the ends were sealed and the slide transferred to a temperature-controlled microscope stage held at 37°C . Images were taken either on a Zeiss Axio Imager M1 microscope equipped with an EC Plan-Neofluar 100x/1 . 3 oil immersion DIC objective and an AxioCam MRm camera , or on a Nikon Eclipse Ni-E widefield microscope fitted with a Hamamatsu C11440 digital camera and Nikon N Plan Apo λ 100x/1 . 45NA oil immersion objective . Images were taken at 5 s intervals over a total of 20–30 min , then annotated and exported as TIFFs , AVI or QuickTime movies using Axiovision 3 . 1 or Nikon NIS-Elements software . Mean fluorescence intensity values of individual mCherry-expressing schizonts were determined from exported raw image files ( TIFF format ) as described previously [18] , using the Lasso tool and Histogram options of Adobe Photoshop CS6 . | Malaria , a disease that kills hundreds of thousands of people each year , is caused by a single-celled parasite that grows in red blood cells of infected individuals . Following each round of parasite multiplication , the infected red cells are actively ruptured in a process called egress , releasing a new generation of parasites . Egress is essential for progression to clinical disease , but little is known about how it is controlled . In this work we set out to address the function in egress of a Plasmodium falciparum protein called SERA5 , an abundant component of the vacuole in which the parasite grows . We show that parasites lacking SERA5 ( or lacking both SERA5 and a closely-related protein called SERA4 ) undergo accelerated but defective egress in which the bounding vacuole and red cell membranes do not rupture properly . This impedes the escape and subsequent replication of the newly-developed parasites . We also show that modification of SERA5 by parasites proteases just prior to egress is important for SERA5 function . Our results show that SERA5 is a ‘negative regulator’ of egress , controlling the speed of the pathway that leads to disruption of the membranes surrounding the intracellular parasite . Our findings increase our understanding of the molecular mechanisms underlying malarial egress and show that efficient egress requires tight control of the timing of membrane rupture . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"parasite",
"groups",
"plasmodium",
"viral",
"transmission",
"and",
"infection",
"microbiology",
"cloning",
"parasitic",
"protozoans",
"parasitology",
"organisms",
"apicomplexa",
"protozoans",
"red",
"blood",
"cells",
"molecular",
"biology",
"techniques"... | 2017 | The Plasmodium falciparum pseudoprotease SERA5 regulates the kinetics and efficiency of malaria parasite egress from host erythrocytes |
The therapy of visceral leishmaniasis ( VL ) is limited by resistance , toxicity and decreased bioavailability of the existing drugs coupled with dramatic increase in HIV-co-infection , non-availability of vaccines and down regulation of cell-mediated immunity ( CMI ) . Thus , we envisaged combating the problem with plant-derived antileishmanial drug that could concomitantly mitigate the immune suppression of the infected hosts . Several plant-derived compounds have been found to exert leishmanicidal activity via immunomodulation . In this direction , we investigated the antileishmanial activity of eugenol emulsion ( EE ) , complemented with its immunomodulatory and therapeutic efficacy in murine model of VL . Oil-in-water emulsion of eugenol ( EE ) was prepared and size measured by dynamic light scattering ( DLS ) . EE exhibited significant leishmanicidal activity with 50% inhibitory concentration of 8 . 43±0 . 96 μg ml-1 and 5 . 05±1 . 72 μg ml─1 , respectively against the promastigotes and intracellular amastigotes of Leishmania donovani . For in vivo effectiveness , EE was administered intraperitoneally ( 25 , 50 and 75 mg/kg b . w . /day for 10 days ) to 8 week-infected BALB/c mice . The cytotoxicity of EE was assessed in RAW 264 . 7 macrophages as well as in naive mice . EE induced a significant drop in hepatic and splenic parasite burdens as well as diminution in spleen and liver weights 10 days post-treatment , with augmentation of 24h-delayed type hypersensitivity ( DTH ) response and high IgG2a:IgG1 , mirroring induction of CMI . Enhanced IFN-γ and IL-2 levels , with fall in disease-associated Th2 cytokines ( IL-4 and IL-10 ) detected by flow cytometric bead-based array , substantiated the Th1 immune signature . Lymphoproliferation and nitric oxide release were significantly elevated upon antigen revoke in vitro . The immune-stimulatory activity of EE was further corroborated by expansion of IFN-γ producing CD4+ and CD8+ splenic T lymphocytes and up-regulation of CD80 and CD86 on peritoneal macrophages . EE treated groups exhibited induction of CD8+ central memory T cells as evidenced from CD62L and CD44 expression . No biochemical alterations in hepatic and renal enzymes were observed . Our results demonstrate antileishmanial activity of EE , potentiated by Th1 immunostimulation without adverse side effects . The Th1 immune polarizing effect may help to alleviate the depressed CMI and hence complement the leishmanicidal activity .
Leishmaniasis , a complex vector-borne parasitic syndrome , is caused by obligate intra-macrophagic protozoan parasites of the genus Leishmania , a member of the order Kinetoplastida . The array of manifestations varies from a self-limiting cutaneous form to a potentially lethal visceralizing infestation of the liver , spleen and bone marrow [1] . Resolution of disease correlates with Leishmania-specific CD4+-type 1 T helper ( Th1 ) and CD8+ T lymphocyte responses with production of interferon-γ ( IFN-γ ) , macrophage nitric oxide ( NO ) and reactive oxygen species ( ROS ) generation [2–3] . Leishmania species exploit discrete mechanisms to elude the cellular immune defenses , such as inhibition of phagolysosomal fusion , and reactive nitrogen species ( RNS ) - and ROS-mediated macrophage microbicidal effects , dampening of cell-mediated immune response via blockade of antigenic peptide display to T cells , impaired secretion of Th1 cytokines , and infiltration of IL-10 producing T regulatory cells [4–6] . The treatment of visceral leishmaniasis ( VL ) is complicated because of intra-macrophagic refuge of the amastigotes , rendering the patient immunodeficient and unable to eliminate the parasites through the natural defense mechanisms [7] . The quandary of VL has been compounded due to concomitant infection in acquired immunodeficiency syndrome ( AIDS ) patients [8] . There is no vaccine available against VL though several are in phase III clinical trials [9] . The existing anti-leishmanial therapy suffers from grave impediments such as drug resistance , compromising efficacy , toxicity , prolonged courses and parenteral routes of administration [10–12] . Hence new drugs for the treatment of VL are imperative . In an ongoing quest for safe and cheap antileishmanial agents , plant-based secondary metabolites are gaining ground [13–15] . The use of plant products as immune-stimulants has a traditional history . Treatment of leishmaniasis with natural or synthetic molecules appears to be dependent upon the development of an effective immune response that activates macrophages and lymphocytes to release their effector molecules [16–17] . A plethora of studies have reported immunomodulation with plant secondary metabolites such aslicarin A , [18] , niranthrin [19] , alkaloid skimmianine [20] , quassin [21] , tannins and structurally related compounds [22] , N-Palmitoyl-S- ( 2 , 3-bis ( palmitoyloxy ) - ( 2RS ) -propyl ) -Cys-Ser-Lys4 hydrochloride ( Pam3Cys ) [23] and linalool component of essential oil [24] . Synergistic antileishmanial and immunopotentiating effects of plant fractions or compounds have also been documented [25] . This may result in enhanced clearance of the parasites coupled with boosting of the depressed immunity associated with active VL . Eugenol ( Fig 1 ) is the major constituent of Syzygium aromaticum . S aromaticum or common clove , is indigenous to tropical America and Australia [26] and is endowed with antibacterial [27] and anti-trypanosomal activities [28] . The extracts from flower buds of S . aromaticum has been reported to display antimalarial efficacy [29] . The immunomodulatory effect of S . aromaticum essential oil has been attributed in augmentation of humoral and cell mediated immune responses [30] . We have previously evaluated the leishmanicidal effect of eugenol-rich essential oil of S . aromaticum against promastigotes and intramacrophagic-amastigotes of L . donovani [31] . However , the poor solubility and high volatility limits its stability resulting in paradigm shift from therapeutic use of most oils to biocompatible emulsifiers . The antibacterial and antifungal activities of eugenol emulsions have been explored [32–33] . Encouraged by the above studies , we evaluated the antileishmanial and immunomodulatory potential of eugenol emulsion ( EE ) against experimental VL in BALB/c mice .
The experiments were performed on female BALB/c mice , aged about 6–8 weeks ( 20–25 g ) while the L . donovani parasites were maintained in Syrian golden hamsters ( 4–6 weeks old ) , after prior assessment and approval of the study protocol ( Ethics Clearance number 459 ) by the Jamia Hamdard Animal Ethics Committee ( JHAEC ) . JHAEC is registered under the Committee for the purpose of control and supervision of experiments on animals ( CPCSEA ) that is registered under Animal Welfare Division of the Central Government of India . The animals were kept in the Central Animal House facility of Hamdard University as per the CPCSEA guidelines . The animals were housed in standard size polycarbonate cages in groups of at least 5 mice ( or 3 hamsters ) per cage , with temperature maintained at 23 ± 1°C , relative humidity ( 55 ± 10% ) , 12:12 h light: dark cycle with ad libitum access to a standard pellet diet ( Ashirwad Industries , Chandigarh , India ) and drinking water . For ex vivo experiments , RAW 264 . 7 cells were procured from National Centre for Cell Science , Pune , India . Undesirable side effects of EE for painful abdominal distention and the resulting distress have been avoided by daily monitoring . We have used 0 . 15 ml for the intraperitoneal injection of EE in mice , that is , below the recommended volume ( 0 . 2 ml ) for intraperitoneal administration of emulsion in mice . Leishmania donovani ( MHOM/IN/AG/83 ) promastigotes were cultured in M199 medium , supplemented with 10% heat-inactivated FBS , 2 mM glutamine , 100 units/ml penicillin G sodium , and 100 μg/ml streptomycin sulphate at 22°C . The parasites were maintained in culture for 4–5 days ( initial inocula being 1X106 parasites/ml ) and the late log or stationary phase promastigotes were harvested after the second or third passage . Murine macrophage-like cells , RAW 264 . 7 were maintained at 37°C in 10% FBS-supplemented RPMI-1640 medium ( pH 7 . 4 ) for 48–72 h in a humidified atmosphere with 5% CO2 . The sub-confluent cultures ( 70–80% ) were split in fresh medium split at 2 x 105 cells ml─1 . Eugenol was procured from Sigma-Aldrich and oil-in-water emulsion ( EE ) was prepared using non-ionic surfactant , Tween 80 ( T-80 ) and water to achieve a final concentration of oil mixture in the emulsion as 200 mg ml─1 . Emulsion was formulated by mixing Eugenol and surfactant , 4% micellar solution was prepared with T-80 and deionized water . Eugenol was poured into the micellar solution through calibrated micropipette under continuous stirring . The formulated emulsion was analyzed for particle size by dynamic light scattering ( DLS ) Nano-S90 ( Nanoseries , Malvern Instruments , UK ) . The samples were measured [34] at 25°C with a fixed angle of 90° . The dosing emulsion was carefully stored in aluminium foil covered-glass vials at 4° C prior to biological assays . Promastigotes ( 2 x 106 cells ml─1 ) were incubated for 72 h at 22°C without or with serial three-fold dilutions of EE , starting from 100 μg ml─1 ( 100 , 33 . 33 , 11 . 11 , 3 . 70 , 1 . 23 and 0 . 41 μg ml─1 ) . The mean percent ( % ) viability was calculated by MTT assay [35] . The inhibitory concentration responsible for 50% reduction in promastigote growth ( IC50 ) , was graphically extrapolated by plotting percent ( % ) viability versus drug concentration [31] . Surfactant , T-80 ( 0 . 00175% , present in 100 μg ml─1 EE ) was used as negative control . To assess the effects of EE on intracellular L . donovani amastigotes , RAW 264 . 7 macrophages ( 1X106 cells ml─1 ) were infected with parasites in late log or stationary phase ( cell/promastigote ratio , 1/10 ) at 37°C for 24 h . Thereafter , the non-ingested parasites were gently aspirated , and the infested macrophages were further incubated with EE ( 0–100 μg ml─1 ) and with negative control ( surfactant ) for 48 h . The cells were then fixed , Giemsa-stained and microscopically evaluated for percent amastigote infectivity . At least 200 macrophages were enumerated per coverslip , and the concentration of EE that reduced amastigote infectivity by 50% ( IC50 ) was calculated . In parallel , the nitrite concentration in the cell culture supernatants was analysed using Griess reaction . Briefly , to the harvested supernatant , an equal volume of Griess reagent ( 0 . 1% N-{1-naphthyl} ethylenediaminedihydrochloride and 1% sulphanilamide in 5% phosphoric acid ) was added . After incubation for 10 min at room temperature ( RT ) , the absorbance values were read at 550 nm , and nitrite concentration was calculated by performing linear regression of a standard curve of sodium nitrite [36] . EE and all the reagents were analyzed for endotoxin , lipopolysaccharide ( LPS ) by chromogenic Limulus Amoebocyte Lysate ( LAL ) kit according to the manufacturer’s instructions ( Pierce , Thermo Scientific ) and were found to be free of LPS ( 0 . 2 ng ml─1 endotoxin ) . To determine the adverse cytotoxic effects of EE , RAW 264 . 7 cells in RPMI 1640 medium were incubated for 48 h at 37°C in a humidified 5% CO2 incubator with increasing concentrations of EE ( 0–200 μg ml─1 ) . Surfactant was used as negative control . MTT assay was used to evaluate cell viability , expressed as a percentage relative to untreated macrophages as control [35] . Six to eight weeks old BALB/c mice were injected with 2×107 stationary phase L . donovani promastigotes in the lateral tail vein . At eight weeks of infection , after confirming the parasite load in three randomly selected animals; the mice were arbitrarily divided into four groups of 10 animals each ( A- D ) . Group A comprised of untreated infected mice ( INF ) ; Group B–Saline administered vehicle control mice ( VC , i . p ) . Groups C , D , E ( EE , i . p . ) , Infected mice that received respectively three doses of EE ( 25/50/75 mg/kg body weight {b . w . } ) each day for 10 days . Group F- treated with Amphotericin B ( AMB , 5 mg/kg b . w . , i . v . , alternately over a ten day period ) , worked as the positive control . Ten days post-treatment , mice were euthanized by carbon dioxide asphyxiation for enumeration of hepatic and splenic amastigote burden that was expressed in terms of Leishman Donovan Units ( LDU ) . LDU was evaluated per organ from giemsa-stained multiple impressions smears [23] as the number of amastigotes per 500 host cells X organ weight ( mg ) . Percent reduction in amastigote load ( % protection ) was calculated as the difference between LDU of infected control and treated mice/ LDU of infected control X 100 . Protection coincided with a drop in hepato-splenomegaly and parasite clearance with respect to untreated infected controls [37] . SLA and FT antigens were prepared from promastigotes in the stationary-phase [38–39] . Briefly , promastigotes in third or fourth passage were harvested , washed four times in cold 1X PBS and resuspended at 2x108 cells ml─1 . The suspension was frozen at ─80°C ( 30 min ) and thawed in a 37°C water bath ( 15 min ) alternately for 6 cycles . For SLA preparation , after ten alternate freeze-thawing cycles , the suspension was centrifuged ( 5250 x g , 4°C , 10 min ) and the supernatant containing leishmanial antigens ( SLA ) harvested . FT and SLA were stored at ─70°C until use and the protein quantitated [40] . DTH response in infected mice subsequent to treatment was evaluated as a hallmark of cellular immunity . Briefly , two days prior to euthanisation , mice were intradermally injected with FT ( 50 μl: 800 μg ml─1 ) in the right footpad and PBS in the left footpad . After 48 h , the thickness of footpads was recorded using vernier calipers and the results expressed as the difference in swelling of the right compared to the contralateral left hind footpad [39 , 41] . The Leishmania-specific serum IgG subclasses were measured through enzyme-linked immunosorbent assay ( ELISA ) . In brief , FT ( 0 . 25 μg/well ) was seeded in the wells of ELISA plates ( Nunc , Roskilde , Denmark ) for 1 h at 37°C . After three washes , blocking was done with 1% BSA for 2 h at RT followed by addition of 1 , 000-fold diluted mice sera . Post-washing , IgG1 and IgG2a isotype-specific goat anti-mouse secondary antibodies ( Sigma Aldrich ) were added and the plates incubated at 37°C for 1 h . After washing , incubation with peroxidase-conjugated rabbit anti-goat IgG as the tertiary antibody ( Sigma Aldrich ) was done at 37°C for 1 h . Post-washing , OPD was added and the absorbance taken on an ELISA plate reader at 490 nm [41] . Culture supernatants of peritoneal macrophages from normal and infected mice following treatment were analyzed for LPS- and SLA-specific nitrite ( NO2 ) levels by the Griess method as described previously [42] . Briefly , to the culture supernatants , equal volume of Griess reagent was added and incubated for 15 min at RT . The optical density ( OD ) was measured at 550 nm using a microplate reader . Dilute solution of sodium nitrite ( NaNO2 ) in culture medium served as a standard . All the reagents and fractions were free of LPS ( 0 . 2 ng/ml endotoxin ) as confirmed by LAL assay . To assess the effect of EE on the proliferation of lymphocytes , single cell suspension from spleens ( 5 × 106 cells ml─1 ) and lymph nodes ( 2 × 106 cells ml─1 ) of treated and untreated mice were seeded into 96-well microplates ( 200 μl/well ) and in vitro stimulated with SLA ( 10 μg ml─1 ) at 37°C for 48 h in a humid-saturated atmosphere containing 5% CO2 . Lymphoproliferation was assessed by enumerating the cells using a hemocytometer [43] . Alternatively , for tracking lymphoproliferation by CFSE dilution , the SLA ( 10 μg ml-1 ) -stimulated lymphocytes ( 5 ×106 cells ml-1 ) from infected , treated and naïve mice were labeled with 1 μM CFSE for 48 h . After washing twice with PBS , the cells were resuspended in PBS and acquired in a BD LSR II flow cytometer to assess the population of cells that underwent proliferation . The contour plots were generated after appropriate gating [44] . Cytokine concentrations ( IL-12 , IL-4 , IL-10 and IFN-γ ) in the serum and splenocyte culture supernatants of differently treated mice were estimated by a multiplex bead-based assay as per the instructions of the manufacturer . [45–46] . Briefly , culture supernatants , serum samples and the cytokine standards were added in equal volumes to antibody-coated capture beads prior to incubation with biotinylated detection antibodies ( anti-mouse ) for 1 h at RT in the dark . The beads were washed with wash buffer ( 400 x g , 4°C , 5 min ) and the supernatant gently aspirated . The beads were washed twice followed by incubation for a period of 1 h at RT in the dark with streptavidin-PE . After performing two additional centrifugation steps as described above , the beads were re-suspended in assay buffer and acquired on a BD LSR II flow cytometer ( Becton Dickinson ) . The data were analysed with BD CBA software based on standard curves generated with recombinant cytokines . CD4+ and CD8+ T cell phenotyping was performed as previously described [47] . Splenocytes from treated and untreated BALB/c mice were co-stained with anti-CD4 FITC and anti-CD8 PE antibodies for 15 min on ice . The cells were then washed and resuspended in PBS for acquisition on a BD LSR II flow cytometer equipped with DIVA software . Intracellular detection of IFN-γ- producing CD4+ and CD8+ T lymphocytes was performed by flow cytometry . Single cell suspensions from spleens of treated and untreated infected mice were in vitro stimulated with 10 μg ml-1 SLA for 24 h and further incubated with Brefeldin A ( 10 μg ml-1 ) for 1 h . After washing with FACS buffer , the cells were co-stained with APC and PE conjugated anti-CD4 and anti-CD8 antibodies , respectively . This was followed by washing , fixing and permeabilization with BD Cytofix/Cytoperm . The cells were then stained with FITC-anti-IFN-γor isotype-matched control monoclonal antibodies ( mAbs ) , and acquired and analyzed on a flow cytometer . The CD4+ and CD8+ T lymphocytes were gated individually and the expression of IFN-γ-producing cells ascertained [41] . Peritoneal macrophages ( 2 × 106 cells ml─1 ) from infected BALB/c mice prior or subsequent to treatment were washed with FACS buffer ( 1X PBS with 1% FBS ) . To quantify the expression of co-stimulatory molecules , 2 × 106 macrophages from each sample were stained with APC-labeled anti-CD80 ( B7-1 ) and PE-Cy7 conjugated anti-CD86 ( B7-2 ) mAbs on ice for 15 min . The cells were washed twice with PBS and flow cytometric acquisition was performed on BD LSR II equipped with DIVA software [48] . Splenocytes from infected BALB/c mice prior or subsequent to treatment were washed with FACS buffer and stained with anti-mouse CD8-APC , CD62L-PE and CD44-FITC ( BD Pharmingen ) for 30 min at 4°C , and then washed and fixed with 2% paraformaldehyde . Cells were acquired on a BD LSR II flow cytometer [49] . Ten days post-treatment , blood was drawn from retro-orbital plexus of naïve , infected and treated BALB/c mice and serum separated . To evaluate the hepatic and renal functions , the serum levels of SGOT , SGPT , ALP , urea and creatinine were measured using commercially available kits ( Span Diagnostics Ltd . ) [39] . All the in vitro experiments were repeated at least twice . The in vivo data are from five mice per group . Statistical analysis was performed using Graph Pad Prism 5 software . P value was calculated using ANOVA with Tukey’s post-test . We considered P values <0 . 05 to be statistically significant . The graphs represent the mean with standard error bars .
The average droplet size and size distribution of eugenol nanoemulsion ( EE ) was found to be 990 . 8±2 . 64 nm and 0 . 23±0 . 01 , respectively ( Fig 2A ) . Low basal levels of NO were detected in the cell-free culture supernatants of infected macrophages , correlating with progression of disease . The NO release from infected macrophages upon subsequent incubation with EE ( 0–100 μgml-1 ) was dose-dependent ( Fig 2D ) and the levels were higher than that produced by normal macrophages . The highest dose of EE ( 100μg ml-1 ) induced 14 . 56±1 . 16 and 20 . 03±3 . 28 μM of NO from normal and parasitized macrophages , respectively . Exposure of RAW 264 . 7 to EE ( Fig 2E ) and surfactant ( Fig 2F ) did not compromise the viability and morphology of the murine macrophage cell line . The intra-peritoneal administration of EE ( 75 mg/kg b . w . ) for 10 consecutive days to 8-weeks infected BALB/c mice caused 87 . 01±5 . 85% ( P< 0 . 001 ) and 86 . 68±5 . 42% ( P< 0 . 001 ) decrease in parasitic load in spleen and liver , respectively ( Fig 3A and 3B ) at 10 days post-treatment . At lower dose of EE ( 50 mg/kg b . w . ) , 65 . 52±4 . 55% ( P< 0 . 001 ) and 61 . 19±7 . 76% ( P< 0 . 001 ) protection were conferred in spleen and liver , respectively . The lowest dose ( 25 mg/kg b . w . ) resulted in more than 45% ( P< 0 . 001 ) fall in hepatic and splenic parasitic burden . AMB ( 5mg/kg b . w . ) induced 92 . 22±4 . 96% and 94 . 88±4 . 25% elimination of parasites from liver and spleen , respectively . A significant reduction in spleen size ( Fig 3C , inset ) was also observed at 75mg/kg b . w . of EE , compared to the infected control . 50% and 25 . 68% reduction in spleen and liver weights , respectively was found with EE at this dose ( Fig 3D and 3E ) , which was comparable to that obtained with AMB ( 52 . 33% and 28 . 41% ) . As an in vivo correlate of cell-mediated immunity , DTH reaction to FT was measured in infected mice at 10 days post-treatment . EE treatment resulted in a significant enhancement ( P< 0 . 001 ) in footpad thickness at 48 h compared with the control group ( 0 . 20±0 . 02 mm ) . A dose related increase in DTH reactivity was observed , with maximum swelling at 75 mg/kg b . w . ( 0 . 37±0 . 035 mm ) followed by 50 and 25mg/kg b . w . ( 0 . 33±0 . 03 and 0 . 26±0 . 01 mm , respectively , Fig 4A ) . Whereas AMB treated mice showed marginal levels ( 0 . 27±0 . 02 mm ) of DTH response . At 10 days post-treatment , mice sera were assayed for FT-specific IgG1and IgG2a isotype levels . IgG1 was detected at significantly ( P ≤ 0 . 01 ) higher levels over IgG2a in infected control animals compared with the treated groups; the IgG2a:IgG1 ratio being 0 . 45 ( Fig 4B ) . The highest IgG2a:IgG1ratio ( 1 . 72 ) was observed in mice treated with 75mg/kg b . w . EE ( P<0 . 001 ) , followed by 50 mg/kg b . w . EE ( 1 . 28 ) , AMB ( 1 . 14 ) and 25 mg/kg b . w . EE ( 0 . 97 ) treatment groups . Host defense against intracellular pathogens including Leishmania is primarily mediated by NO and related RNIs [49] . NO production from peritoneal macrophages of EE treated mice was used to evaluate the effectiveness of EE on macrophage microbicidal activity . Nitrite , the stable end-product of NO metabolism was assessed by Griess reagent . Three-fold higher ( 9 . 54±0 . 77 μM ) nitrite levels were found in mice treated with higher dose of EE ( 75mg/kg/b . w . , P< 0 . 001 ) as compared to infected control ( 3 . 78±0 . 63 μM ) and the response was dose-dependent ( Fig 4C ) . AMB treated mice induced moderate levels of nitrite ( 6 . 26± 0 . 38 μM ) . The immunomodulatory potential of EE was evaluated through lymphoproliferation . Upon microscopic enumeration , a significant ( P< 0 . 001 ) proliferative effect of SLA-stimulated splenocytes as well as lymphocytes was observed at 10 days post-EE treatment . The lymphoproliferative response was dose-dependent with maximum being elicited at 75mg/kg b . w . that was followed by 50 and 25mg/kg b . w . EE treatment In contrast , AMB did not induce significant SLA-specific lympho-proliferation ( Fig 5A and 5B ) . In parallel , the lymphoproliferative potential of lymphocytes following EE treatment was corroborated by CFSE staining . 28 . 16% and 28 . 02% lymphocytes from spleen and lymph nodes , respectively underwent cell division in normal mice . The lymphoproliferative response was the highest in spleen ( 78 . 7% ) and lymph nodes ( 80 . 8% ) upon treatment with EE ( 75 mg/kg . b . w . ) . In untreated infected control group , we observed 29 . 28% and 28 . 14% lymphoproliferation in spleen and lymph nodes , respectively ( Fig 5C and 5D ) . Changes in the levels of classical Th1 ( IFN- γ and IL-2 ) and Th2 ( IL-4 and IL-10 ) cytokines were assessed on day 10 post-treatment . Compared to untreated infected controls , EE ( 75mg/kg b . w . ) induced enhanced serum levels of IFN-γ and IL-2 ( 2871±121 and 6943 . 5±129 . 77 pgml─1 , respectively , P ≤ 0 . 001 ) and significantly lower levels of IL-4 and IL-10 ( 472 . 5±96 . 53 and 182 . 5±30 . 76 pgml─1 , P ≤ 0 . 001 ) ( Fig 6A ) . Whereas , AMB treatment restored the cytokines to normal levels . Similar effect on Th1 and Th2 type cytokines was also observed with the culture supernatants of infected and treated mice ( Fig 6B ) . Mice with established L . donovani infection had low numbers of splenic CD4+ ( 7 . 2% ) and CD8+ ( 4 . 1% ) T cells ( Fig 7 ) that enhanced to 14 . 4% and 7 . 6% , respectively 10 days post-treatment with EE ( 25 mg/kg b . w . ) . The T cell counts were increased with 50 mg/kg b . w . EE treatment ( 17% CD4+ and 9 . 4% CD8+ T cells ) . The CD4+ and CD8+ T cell population was , however , the highest ( 18 . 8% and 10 . 7% , respectively ) at 75 mg/kg b . w . of EE ( Fig 7A ) . These data support a bias towards Th1-driven effector functions and the role of CD4+ as well as CD8+ T cells in protection following EE treatment . CD4+ and CD8+ T cells are the main cellular sources of IFN-γ . In infected mice , low frequencies of IFN-γ-producing CD4+ ( 4 . 38±0 . 35% ) and CD8+ ( 3 . 64±0 . 30% ) T cells were detected which were upregulated by AMB treatment ( 9 . 04±0 . 72% CD4+ , 6 . 89±0 . 36% CD8+ ) . The maximum induction of IFN-γ-secreting CD4+ and CD8+ T cells occurred at 75mg/kg . b . w . of EE ( 21 . 18±0 . 99% , 18 . 51±0 . 80% , respectively ) . While EE at 25mg/kg . b . w . induced 9 . 02±1 . 06% CD4+ and 8 . 41±0 . 72% CD8+ T cells secreting IFN-γ that was comparable to that elicited by AMB ( Fig 7B ) . The co-stimulatory molecules ( CD80 and CD86 ) on APCs are essential for optimal T-APC talk and lymphocyte activation for secretion of effector cytokines . It was found that EE significantly augmented the expression of both CD80 and CD86 on peritoneal macrophages . EE ( 25 mg/kg b . w . ) optimally boosted the number of double positive cells co-expressing CD80 and CD86 ( 27% ) over the naïve ( 15 . 5% ) and infected control ( 12 . 9% ) groups ( Fig 8 ) . The response was dose-dependent with heightened expression at 50 mg/kg b . w . of EE ( 28 . 9% ) . CD80 and CD86 double positive cells were maximally expressed at 75 mg/kg b . w . EE ( 32 . 5% ) . The response in case of AMB treated group was also enhanced up to a moderate level ( Fig 8 ) . Memory T cell generation in the host is indicative of resistance to Leishmania re-infection [50] . The subtle co-expression of CD44 and CD62L ( 10 . 7% ) in the infected control group , was up-regulated ( 16 . 73 . 6% ) upon treatment with EE at 75mg/kg b . w . , in concordance with resolution of disease and generation of central memory cells . AMB had insignificant effect ( 10 . 93% ) on induction of memory T cells ( Fig 9 ) . Liver function ( ALP , SGOT and SGPT ) and renal function tests ( urea and creatinine ) were done 10 days post-administration of EE in naive BALB/c mice ( Table 1 ) as well as infected mice upon treatment ( Table 2 ) . EE treated group showed normal values of serum enzymes ( up to 75 mg/kg b . w . ) , indicating no in vivo toxicity .
The immune system is known to synergistically augment the therapeutic effectualness of anti-parasitic drugs [51] . The administration of immunomodulators in conjunction with conventional chemotherapy to rejuvenate the host immune response has a multitude of benefits and aids in advancing the current therapeutic effectiveness by reducing the dose and duration of treatment and hence the toxicity of the drugs . Thus , antileishmanial drugs that can concomitantly ameliorate the immune suppression of the infected hosts , have a two-prong effect and are sought after [52–56] . Previous report showed that the essential oil of S . aromaticum ( EROSA ) , mediated programmed cell death against L . donovani without affecting the host macrophages . Eugenol was the major constituent in EROSA [31] . This is the first report of therapeutic and immunostimulatory potential of emulsion of eugenol ( EE ) against experimental VL . EE showed profound antileishmanial efficacy , coupled with cell-mediated immunopotentiation without any adverse effect on the host . Furthermore , EE exhibited a concentration-dependent significant antileishmanial activity against L . donovani promastigotes as well as the intramacrophagic amastigotes . Ex vivo studies revealed non-toxicity of EE against murine macrophages even at 100μg ml─1 . In case of L . amazonensis , clearance of both axenic and intracellular amastigotes has been attributed to NO [57] . We observed a significant increase in NO ( 20 . 03±3 . 28 μM ) post-incubation of amastigote-infested macrophages with EE , harmonizing with the previous reports . The EE-induced NO production from infected macrophages strengthened the role of RNS in amastigote death . Intra-peritoneal treatment with EE significantly ( P<0 . 001 ) lowered the splenic and hepatic parasite loads in infected mice to levels similar to that achieved with AMB . This was coupled with restoration of liver and spleen weights to normal levels . Our results warrant further studies with lower doses of EE . Our study corroborates immunomodulation complementing the leishmanicidal effect of EE , apparently through Th1 cytokine-driven and macrophage-mediated mechanisms which are also NO-dependent . Immunostimulatory effects have been reported for several natural products , offering a rational basis for their therapeutic potency . In case of leishmaniasis , an ideal drug is envisaged to have dual effects–to directly and selectively kill the parasites as well as rejuvenate the depressed immunity towards Th1 bias [58 , 55] . During the progression of leishmaniasis , cytokine milieu switches from Th1 to Th2 profile [59–60] . An effective therapeutic intervention against L . donovani entails mounting of a strong Th1 response . To prove that EE has immunomodulatory effect , we analysed the cytokines in serum and culture supernatants of lymphocytes from infected BALB/c mice 10 days post-treatment . Therapy with EE mounted a polarized Th1 response with enhanced IFN-γ and IL-2 secretion , which stimulated microbicidal responses of macrophages leading to NO production that was demonstrated in this study ex vivo as well as in vivo . Both reactive oxygen and nitrogen species have been implicated to contribute to reduction in parasitism ex vivo as well as in vivo , and suppression of either compromises the macrophage-leishmanicidal activity [61–62] . IFN-γ is known to stimulate iNOS2 expression with significant NO secretion , that may have provided impressive levels of protection . In our studies , the antagonistic effect of IFN-γ in reducing Th2-associated cytokines , IL-10 and IL-4 was also found in treated mice . Detailed immunological analysis of infected mice depicted persistent IgG1 levels , probably maintained by IL-4 secretion [63] . Treatment with EE was associated with increased IgG2a , possibly due to up-regulated serum levels of IFN-γ . Plant-derived natural products have been found to exert leishmanicidal activity via modulation of the Th1/Th2 responses . Quassin inhibited L . donovani growth by switching from Th2 ( IL-4 and IL-10 ) to Th1 ( IFN-γ and IL-2 ) cytokine [21] . A similar effect occurred with EE , which enhanced IFN-γ and NO production , coupled with marked reduction in the L . donovani load in infected macrophages . EE induced a marked switch in L . donovani-infected BALB/c mice from disease-promoting to Leishmania-specific disease-resolving humoral as well as cell-mediated immune responses . EE at 75mg/kg b . w . rescued T-cell-anergic conditions , inducing elevated levels of DTH , lymphoproliferation , IL-2 , IFN-γ and NO , and maximally reduced the Th1 suppressive cytokines ( IL-4 and IL-10 ) , in a dose-dependent manner , whereas the immune response was restored following treatment with AMB ( 5mg/kg/b . w ) . VL is associated with impaired immunological responses . Effectual antileishmanial therapy entails robust cellular immune responses in addition to antibodies . Experimental models of VL have shown that CD8+ T cells are instrumental in controlling L . donovani/L . infantum infection , through their ability to secrete IFN-γ and/or their cytolytic activity [64] . Furthermore , CD8+ coupled with CD4+ T cells , are essential to thwart reactivation of murine VL , the present investigation showed that the splenic CD4+ and CD8+ T lymphocyte counts were significantly boosted by administration of EE ( 75mg/kg b . w . ) to L . donovani infected BALB/c mice as compared with untreated infected controls . The therapeutic significance of EE in VL was strengthened by the development of central memory ( CD62Lhigh CD44high ) CD8+ T lymphocytes . T cells are known to interact with APCs and play vital roles in eliminating pathogens resident within macrophages . For optimal activation of naive T cells , the T-cell receptor ( TCR ) interacts with the peptide-MHC complex presented by professional APCs , while the second signal is delivered by co-stimulatory molecules , and the secretion of pro-inflammatory cytokines [65] . The ligation of CD80 and/or CD86 on APCs with CD28 on T cells induces the DCs to secrete IL-6 and IFN-γ for optimal T and B cell activation , proliferation , and differentiation [66] . There are reports indicating down modulation of CD80 and 86 expression in certain diseases [67–68] , including leishmaniasis . In our studies , the expression of both CD80 and CD86 on peritoneal macrophages of L . donovani infected mice were found to be significantly enhanced upon treatment with EE , substantiating their immunomodulatory potency . Our results indicate the potentiality of EE in activating the T cells through upregulation of co-stimulatory signals that help to mount an effective immune response by secreting Th1 cytokines such as IFN-γ and IL-2 . The deterioration in renal and hepatic functions is the major dose-limiting side effect of currently available chemotherapeutic drugs . In the present study , upon administration of EE ( 25 , 50 and 75mg/kg b . w . ) to naïve and L . donovani infected BALB/c mice , the serum levels of SGOT , SGPT , ALP , urea and creatinine were found to be in the normal range , proving absence of hepato- and nephro-toxicity . In conclusion , our findings highlight the in vitro and ex vivo leishmanicidal effect of EE that occurred via an increase in production of NO without adverse effects on mammalian macrophages . On the other hand , treatment of infected mice with EE significantly waned the hepatic and splenic parasite loads with diminution in spleen and liver weights . EE induced enhanced lymphoproliferation , up-regulated co-stimulatory molecules ( CD80/CD86 ) expression on APCs , and resulted in expansion of CD4+ and CD8+ T cell numbers and generation of central memory cells . EE further suppressed Th2 cytokines ( IL-4 and IL-10 ) and stimulated the production of Th1 cytokines ( IFN-γ and IL-2 ) with release of NO from the peritoneal macrophages . Th1-driven immune polarization was also reflected from high IgG2a/low IgG1 ratio and significant elicitation of DTH response . Thus , protection against L . donovani infection in the EE treated animals was due to direct parasite killing as well as the induction of cellular immunity via immunopotentiation as depicted in Fig 10 . | In the absence of vaccines and escalating resistance to the available chemotherapeutic drugs , alternate therapeutic options are urgently needed for visceral leishmaniasis ( VL ) or kala-azar , a systemic and potentially fatal , vector-borne neglected tropical disease , caused by Leishmania donovani . In the present study , we explored the two-prong effect of eugenol emulsion ( EE ) in eliminating the parasites with synergistic boosting of the dampened immune system , characteristic of active disease . Infected BALB/c mice upon intraperitoneal administration of EE , exhibited a significant decline in liver and spleen parasite load with concomitant drop in splenomegaly . Protection in treated mice coincided with Th1 immunopotentiation as was evident from substantial DTH and lymphoproliferative responses , elevated levels of IgG2a over IgG1 isotypes , significant enhancement in IFN-γ producing CD4+ and CD8+ T cells , waning levels of IL-4 and IL-10 , augmented nitrite levels , induction of immunological memory and stimulation of antigen presenting capacity of macrophages , compared to infected mice . The dual antileishmanial and immunostimulatory potential of EE with no adverse toxic effects validates it as an adjunct to chemotherapy that may aid in leishmanicidal activity via ameliorating the depressed cellular immunity . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"innate",
"immune",
"system",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"cytokines",
"spleen",
"immunology",
"tropical",
"diseases",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"developmental",
"biology... | 2016 | Immunotherapeutic Potential of Eugenol Emulsion in Experimental Visceral Leishmaniasis |
Contractions on the descending limb of the total ( active + passive ) muscle force—length relationship ( i . e . when muscle stiffness is negative ) are expected to lead to vast half-sarcomere—length inhomogeneities . This is however not observed in experiments—vast half-sarcomere—length inhomogeneities can be absent in myofibrils contracting in this range , and initial inhomogeneities can even decrease . Here we show that the absence of half-sarcomere—length inhomogeneities can be predicted when considering interactions of the semi-active protein titin with the actin filaments . Including a model of actin—titin interactions within a multi-scale continuum-mechanical model , we demonstrate that stability , accurate forces and nearly homogeneous half-sarcomere lengths can be obtained on the descending limb of the static total force—length relation . This could be a key to durable functioning of the muscle because large local stretches , that might harm , for example , the transverse-tubule system , are avoided .
The isometric active force—length relation of muscle fibres [1] and many muscles [2–4] is composed of strikingly linear segments . Based on this observation , the sarcomere microstructure , and the sliding filament theory [5 , 6] , a geometric model of the filament overlap has been established and validated [1] , which shaped our understanding of muscle structure and functioning . From a mechanical point of view , the negative slope of the force—length relationship at long muscle lengths ( descending limb ) was and is of special interest for muscle physiologists and modellers due to its inherent instability . It has been suggested early [7] that contractions on the descending limb of the force—length relation will result in unstable behaviour leading to the formation of short and long half-sarcomeres in series during fixed-length muscle contractions ( cf . Fig 1 top ) . A number of human and animal skeletal muscles work on the descending limb of the active force—length relation [10] . Moreover , passive forces in single muscle fibres [11–13] and some entire skeletal muscles , e . g . the rabbit extensor digitorum longus , extensor digitorum II , and soleus [3 , 4] or the frog semitendinosus [14] , appear only at lengths that correspond to the descending limb of the active force—length relation [13 , 15] leading to a descending limb in the total force—length relation . Simulation of these muscles is particularly challenging using continuum-mechanical finite element models , which typically superimpose the passive stress tensor with an active stress that includes the active force—length relation [16 , 17] . The descending limb of the resulting total stress—stretch relation ( Fig 1 top ) causes instability of these models , which explains why previous continuum-mechanical models focused on muscles with significant passive forces occurring at short muscle length , i . e . , that have no descending limb in the total force—length relation [17–21] . With respect to the above described approach of superimposing passive and active stress contributions , two aspects require further consideration . First , the energy corresponding to a total stress—stretch relation with a descending limb is described by a non-convex function , cf . Fig 1 ( bottom ) , whereas a unique solution requires convexity . Second , the force—length relation is a static muscle property and , in light of muscle contraction phenomena like force enhancement ( muscle force is increased after stretch compared to the corresponding isometric muscle force at the same final length ) [22–24] , it should not be used as a dynamic one [25] . Thus , the force—length relation should not be interpreted as a hyperelastic stress—stretch relation . Unstable half-sarcomere behaviour on the descending limb can be mitigated by the damping effect of the force—velocity relation [26] and might possibly even be prevented by the interaction of the semi-active protein titin with the actin myofilament [27–29] . By calcium-induced actin—titin binding within the isotropic band of the half-sarcomere , the molecular spring length of titin would dramatically reduce . This would lead to an increased passive force when the ( weaker ) half-sarcomere is lengthened , which may prevent the evolution of a heterogeneous half-sarcomere—length distribution and instability on the descending limb of the force—length relation . Recently ( see [30] ) , we included a biophysical model describing force enhancement based on actin—titin interaction [31] in a continuum-mechanical finite element model [19 , 32] . The ‘sticky—spring’ model [31] assumes that , in the presence of calcium , titin’s PEVK ( rich in proline ( P ) , glutamic acid ( E ) , valine ( V ) , and lysine ( K ) ) region can bind to actin . This leads to increased titin forces during and after active stretch . The resulting titin-induced half-sarcomere-based stresses have been homogenised and added to the continuum-mechanical stress tensor consisting of passive and active stress contributions . We used the proposed model to analyse muscle forces during and after active stretch starting on the plateau and the descending limb of the force—length relation by comparing the results obtained by the models with and without actin—titin interaction . In that contribution , we neglected the force—velocity relation . In this study we demonstrate that actin—titin interactions [27–29] stabilise half—sarcomere lengths during fixed—length contractions and active stretches ( neglecting inertia effects ) on the descending limb of the force—length relation . Moreover , we demonstrate that the force—velocity relation does not stabilise half-sarcomeres in a similar way , as it has been discussed controversially in the past ( cf . e . g . [33 , 34] ) .
Muscles consist of muscle fibres arranged in parallel and these fibres are connected to each other laterally through the extracellular matrix , to which a significant portion of the fibre force can be transmitted [36 , 37] . Thus , not only the arrangement of sarcomeres in series within a muscle fibre , but also their spatial arrangement within a muscle has to be considered to examine half-sarcomere—length inhomogeneities . Likewise , a large number of half-sarcomeres needs to be considered to obtain a statistically meaningful distribution . Consequently , a multi-scale continuum model is chosen in the present study . The multi-scale skeletal muscle model [19 , 32] embeds one-dimensional muscle fibre ( finite element ) meshes into the three-dimensional mesh of the muscle geometry . Note that the spatial locations of the computational half-sarcomeres , that make up the computational muscle fibres , is directly coupled to the actual ( deformed ) configuration of the three-dimensional continuum-mechanical model and hence enables force transmission along adjacent muscle fibres [37 , 38] . The computational muscle fibres are used to compute the propagation of the action potential by solving the discrete monodomain equation [39] . The nonlinear reaction term of the monodomain equation is coupled to a model of the excitation—contraction coupling at the half-sarcomere level , describing the electro-physiology of the membrane , calcium release from the sarcoplasmic reticulum ( SR ) , and cross-bridge cycling , cf . Fig 3 . In previous works [19 , 30 , 32] , we used a complex biophysical model [40] to simulate the excitation—contraction coupling on the cellular level . Since a detailed description of individual processes ( e . g . the activity of single ion channels ) is less important within the scope of this work , we replaced the biophysical model of the excitation—contraction coupling [40] with a phenomenological description of the membrane electro-physiology [41] coupled to a simplified Huxley-type model [42] . The coupling is realised by assuming that the slow variable of the phenomenological electro-physiological model [41] ( a dimensionless representation of the conductance of a slow inward ( repolarising ) current [43] ) behaves qualitatively similar to the calcium concentration in the myoplasm , which is required as input for the cross-bridge dynamics model [42] . Neglecting nearest-neighbour cooperative effects and distortion dependencies , the model predicts a linear force—velocity relation ( see [42] ) . The linear approximation of the force—velocity relation is considered to be sufficient for the expected small range of contraction velocities due to the fact that only fixed-length contractions and quasi-static stretch experiments are considered in this work . The maximum shortening velocity of the muscle is assumed to be vmax = 15 L0/s [3 , 44] , where L0 denotes the fibre length of the resting muscle . The resulting coupled model of the excitation—contraction coupling is solved at each discretisation point of the muscle fibre meshes [19] . The normalised half-sarcomere-based cross-bridge force , γ , and the binding probability of titin to bind to the actin filament , ζ , are computed from γ=A1x1+A2x2A2maxx0andζ=A2A2max . ( 1 ) Therein , A1 and A2 are the number of cross-bridges in the attached pre-power stroke and post-power stroke states , respectively , and x1 and x2 denote the corresponding average cross-bridge distortions . Further , A2max is the number of cross-bridges in the attached post-power stroke state during an isometric tetanic contraction ( stimulation frequency 100 Hz ) , and x0 denotes the cross-bridge distortion induced by the power-stroke under isometric conditions [42] . While the binding probability of titin , ζ , is assumed to be independent of the contraction velocity , the cross-bridge force is a function of the half-sarcomere velocity , i . e . , γ=γ ( l˙hs ) . In detail , γ = 0 indicates passive behaviour , and γ = 1 corresponds to an isometric contraction at a stimulation frequency of 100 Hz . The cross-bridge force is first scaled using the relation between active force and sarcomere length ( fl ) [1] and then homogenised ( TH:γfl→γf¯l ) to be incorporated into the continuum-mechanical stress tensor , which is evaluated at the macroscale using the discrete representation of the three-dimensional muscle geometry . The homogenisation is also performed for the binding probability of titin ( TH:ζ→ζ¯ ) . Solving the balance of momentum at the macroscale , one obtains the contraction-induced deformation of the muscle geometry and the induced reaction forces . The local muscle deformation ( represented by the fibre stretch , λf ) is interpolated ( TI:λflhsref→lhs ) using the shape functions of the three-dimensional finite elements to the discretisation points of the muscle fibre meshes , where they are used to evaluate the microscopic force—sarcomere length relation fl = fl ( lhs ) [1] . In the following , we refer to these interpolated lengths as half-sarcomere lengths , lhs . The resting half-sarcomere length in the undeformed , stress-free reference configuration ( λf = 1 ) is assumed to be lhsref=1 . 0μm [11] . Further , the velocity is determined from the interpolated local muscle deformations , where an average over four time steps of the continuum-mechanical model ( Δtmech = 0 . 1 ms ) is computed to increase the numerical stability . As in our previous work [30] , we additionally solve a half-sarcomere-based force enhancement model [31] ( the ‘sticky—spring’ model ) at the discretisation points of the muscle fibre meshes . The ‘sticky—spring’ model assumes that in the presence of calcium titin’s PEVK region can bind to the thin filament , which reduces titin’s free molecular spring length to its distal Ig ( immunoglobulin ) region . The model distinguishes between titin filaments that are bound to actin , and titin filaments that are not bound to actin . The stress—stretch relations of both unbound and bound titin filaments is directly obtained from fitting experimental data [45 , 46] . In the model , the stress in unbound titin filaments θtitinunbound results directly from this relation , i . e . θtitinunbound=θtitinunbound ( lhs ) . The titin stresses induced by bound titin filaments is calculated by solving force equilibrium between titin’s distal Ig region and the actin—titin interconnection , i . e . θdistIg ( lhs , lhs[0] ) = θPEVK ( lhs , lhs[0] ) . ( For further details , we refer to [31] . ) The ‘sticky—spring’ model assumes that the actin—titin interconnection behaves like a linear elastic spring , where the corresponding spring constant kPEVK is a function of the half-sarcomere length at which titin was bound to actin lhs[0] . The relation kPEVK ( lhs[0] ) was determined from numerical simulations [31] . Consequently , the stress induced by bound titin filaments depends on the actual half-sarcomere length lhs and the half-sarcomere length at which the titin filament was bound to the thin filament lhs[0] , i . e . θtitinbound=θtitinbound ( lhs , lhs[ 0 ] ) . Similar to the active stresses , the resulting titin stresses are homogenised ( TH:θtitin→θ¯titin ) and added to the continuum-mechanical stress tensor . The resulting macroscopic first Piola-Kirchhoff stress tensor reads P=-p ( detF ) FT-1+Ppassive ( F , M ) ++Pactive ( γ¯ , F , M ) +Ptitin ( ζ¯ , θ¯titinF , M ) . ( 2 ) Therein , Pactive and Ptitin are the active and titin-induced stress contributions , where Pactive=γf¯lPmaxI4-1/2FMandPtitin=[ζ¯θ¯titinbound+ ( 1-ζ¯ ) θ¯titinunbound]PmaxI4-1/2FM , ( see [19 , 30] ) . Further , Ppassive denotes the three-dimensional , transversely isotropic , hyperelastic passive stress tensor . Note that Ppassive is a smeared-out representation of the overall passive mechanical stiffness of the muscle including the extracellular connective tissue , passive contributions of the muscle fibres ( e . g . cytoskeletal structures like nebulin or desmin ) and matrix—fibre connectivity , but excludes the passive contribution of the titin filaments . This is realised by fitting the passive material parameters to experimental data from uniaxial compression tests , cf . [19] . Since within the scope of this publication we do not want to simulate a specific muscle we scaled the passive material properties ( compared to [19 , 32] ) in order to obtain the property of a descending limb in the static total stress-stretch relation . Furthermore , ( · ) T indicates a transposed tensor , F denotes the deformation gradient tensor , M=a0⊗a0 is a structural tensor , where a0 denotes a referential unit vector pointing in the muscle fibre direction , I4 = a ⋅ a = Fa0 ⋅ Fa0 is the fourth ( mixed ) invariant of the right Cauchy-Green deformation tensor C = FT F and Pmax is the maximum isometric tension . Finally , p is the hydrostatic pressure that enters the equation as a Lagrange multiplier due to the incompressibility constraint [47] . Note that the active and titin-induced stresses only act along the muscle fibre direction , while rather small [30] cross-fibre contributions are not considered . For material parameters of the titin model and the continuum-mechanical model , we refer to our previous publications ( for the ‘sticky—spring’ model see [31] , for the continuum-mechanical model see [30] ) . The entire model is implemented within the open-source software library OpenCMISS [48] . To avoid influences of the geometry , a generic cubic muscle specimen with initial edge length L0 = 1cm is considered , in which the muscle fibres are aligned with one of the edges of the cube . For the simulations , the muscle specimen is first passively stretched to a certain length L . Then , all embedded muscle fibres are simultaneously activated ( tetanic stimulation , frequency 100 Hz ) in their middle , while keeping the total length of the muscle specimen fixed . Note that individual segments of the simulated muscle are not constrained and can shorten against each other . Active forces are computed by subtracting the passive forces from the total forces obtained at a certain muscle length . For the analysis , four models are compared: ( i ) model MATIFv includes actin—titin interactions and the force—velocity relation , ( ii ) model MATI includes actin—titin interactions but omits the force—velocity relation , ( iii ) model MFv omits actin—titin interactions but includes the force—velocity relation , and ( iv ) model Mxx omits both actin—titin interactions and the force—velocity relation . In the models without actin—titin interaction , the titin-induced stresses do not vanish but describe the stresses induced by the titin filaments when they are not bound to actin , i . e . , Ptitin=θ¯titinunboundPmaxI4-1/2FM . The following stability analysis considers the static equilibrium and thus neglects the force—velocity relation . In this case , γ and ζ coincide , i . e . , ζ = γ , cf . Eq ( 1 ) . Moreover , this analysis does not distinguish between macroscopic and microscopic quantities , i . e . , ( · ) ¯= ( · ) and lhs=λf=I41/2 . Within the theory of hyperelasticity , convexity guaranties the existence of a global minimiser and hence a unique solution . A sufficient condition for the existence of deformations minimizing a given hyperelastic potential W ( F ) = W ( F , adj F , det F ) is the polyconvexity condition: ∂2W∂F⊗∂F· ( H⊗H ) ≥0 , ∂2W∂adjF⊗∂adjF· ( H⊗H ) ≥0 , ∂2W∂detF2≥0 , ( 3 ) where H ≠ 0 denotes an arbitrary second-order tensor [49] . Regarding the stress tensor in Eq ( 2 ) , the incompressible Mooney—Rivlin material and the anisotropic stress contribution making up for the first two terms on the right-hand side of Eq ( 2 ) are known to satisfy the polyconvexity condition ( see Eq ( 3 ) ) cf . [49 , 50] . Further , due to the fact that the active and titin-induced stresses are not conservative , no strain energy can be defined for the last two terms on the right-hand side of Eq ( 2 ) [51 , 52] . Rather than introducing pseudo-energies for these stress contributions , the definition of the first Piola—Kirchhoff stress tensor of the theory of hyperelasticity P=∂W∂F is used to eliminate the energy in Eq ( 3 ) . For the sake of readability , we introduce a short-hand notation for the active and titin-induced stress contributions: P^=S^ ( γ , I4 ) FM=Pactive+Ptitin , ( 4 ) where S^ ( γ , I4 ) denotes a scalar-valued function of the cross-bridge force γ and the fourth ( mixed ) invariant I4 . Then , the active and titin-induced stress tensors satisfy the condition of Legendre—Hadamard ellipticity if ∂P^∂F· ( G⊗G ) ≥0 , ( 5 ) where G ≠ 0 denotes an arbitrary second-order rank-one tensor , i . e . , G = h ⊗ k ∀h , k . For a given cross-bridge force γ* and deformation I4* the ellipticity condition can be written as [52] 2I4*dS^ ( γ* , I4* ) dI4+S^ ( γ* , I4* ) ≥0 . ( 6 ) To evaluate this condition , as a worst case scenario , we assume full activation ( γ*=1 ) , such that S^ ( I4 ) =I4-1/2Pmax[fl ( I4 ) +θtitinbound ( I4 ) ] . ( 7 ) Inserting Eq ( 7 ) into Eq ( 6 ) , and using I4-1/2>0 , Pmax>0 , the sum of the active and titin-induced stress tensors satisfies the rank-one ellipticity condition if dfl ( I4* ) dI4+dθtitinbound ( I4* ) dI4≥0 . ( 8 ) Relation ( 8 ) states that the sum of the slopes of the active force—length relation and the titin-induced stress has to be positive for all muscle lengths . This condition is equivalent to the condition for a monotonically increasing scalar-valued function , due to the one-dimensional nature of the active and titin-induced stresses . While the titin-induced stress θtitinbound is a monotonically increasing function [31] , the active force—length relation [1] possesses a descending limb with negative stiffness . Keeping in mind that also the passive stress , which we omitted from this derivation , adds positive stiffness to the total behaviour , the result of the ellipticity condition can be interpreted as follows: the positive stiffness induced by the passive stresses and the titin-induced stresses have to compensate together for the negative stiffness on the descending limb of the force—length relation .
First , we demonstrate that the newly introduced phenomenological half—sarcomere model of the excitation—contraction coupling is able to reproduce the expected physiological behaviour in response to stimulation . Fig 4 shows modelled isometric force responses due to different stimulation frequencies ( top ) and the model’s force—frequency relation ( bottom ) . It can be observed that the model predicts a physiological twitch shape , twitch summation , and twitch-tetanus ratio . Next , the quasi-static behaviour of the microscopic half-sarcomere model with actin—titin interaction is considered in isolation . To this end , isometric contractions have been carried out at different half-sarcomere lengths , lhs[0] . Moreover , for different initial half-sarcomere lengths , lhs[0] , i . e . , the half-sarcomere lengths at which the activation started , quasi-static ( i . e . the force—velocity relation was omitted ) active stretch experiments have been performed for different active stretch increments , Δlhs = lhs − lhs[0] . The results are summarised in a three-dimensional surface plot in Fig 5 . While the relation between the total stress in a half-sarcomere and lhs[0] ( i . e . , the total isometric force—length relation; thick black line in Fig 5 ) is not monotonically increasing , the total stress increases monotonically with the active stretch increment . In other words , the partial derivative of the total stress with respect to the active stretch increment is always positive , although the static total stress—stretch relation has a descending limb . Having established the behaviour of the microscopic half-sarcomere model , macroscopic whole-muscle simulations are investigated in the following . Fixed-length contractions at different muscle lengths have been carried out using the previously described four models , namely the MATIFv , MATI , MFv , and Mxx models . Fig 6 compares the resulting active force—length relations . Models MATIFv and MATI reproduce the force—sarcomere length relation as proposed in [1] at the macroscopic whole muscle level , while models MFv and Mxx predict similar forces on the ascending limb and the plateau but deviating forces on the descending limb . In fact , there is a striking agreement between the stretch regions corresponding to the descending limb of the total stress—stretch relation and the occurrence of deviations from the expected linear force decrease in the models without actin—titin interaction ( Mxx , MFv ) . This analysis compares results after 2 . 5 seconds simulation time . At this time , transient effects caused by the force—velocity relation have disappeared and models MFv and Mxx predict similar results . At the very end of the descending limb of the total force—length relation and beyond ( L/L0 > 1 . 6 ) , the forces induced by bound titin filaments had to be relaxed by up to 25% to obtain convergence of models MATIFv and MATI . This can be explained by the large stiffness of the bound titin filaments at these lengths ( cf . Fig . B1 ( A ) of [31] ) leading to numerical problems . To further investigate the origin of the different behaviours of the models , Fig 7 shows the distribution of half-sarcomere lengths within the muscle specimen in fixed-length contractions at different muscle lengths after 2 . 5 seconds . While almost homogeneous half-sarcomere—length distributions are predicted by models MATIFv and MATI for all muscle lengths , the models without actin—titin interaction predict homogeneous half-sarcomere—lengths only on the plateau and for very long muscle lengths , and strongly heterogeneous distributions on the descending limb of the total force—length relation . For the simulation with L/L0 = 1 . 5 , Fig 8 ( top ) shows the stresses obtained with models MATIFv and MATI ( blue dot ) and the distributions of half-sarcomere lengths obtained with models Mxx and MFv ( red crosses ) superimposed on the total force—length relation . Under steady-state conditions ( after 2 . 5 s ) , the models without actin—titin interactions ( MFv , Mxx ) predict identical half-sarcomere length distributions that significantly deviate from the expected value of 1 . 5 μm . In detail , models MFv and Mxx predict half-sarcomere lengths ranging from 1 . 0 μm to 1 . 9 μm , cf . Fig 8 ( bottom ) . Summarising the results shown in Figs 6 , 7 and 8 , a homogeneous half-sarcomere length distribution and active forces that are in accordance with the isometric force—length relation can only be obtained on the descending limb of the total force—length relation , when actin—titin interactions are included in the model . Further , it can be observed that the force—velocity relation cannot prevent the formation of half-sarcomere length inhomogeneities ( models MATIFv and MATI , and likewise models MFv and Mxx , predict similar results under steady-state conditions ) . The previous results in this section considered steady-state results after 2 . 5 seconds , when transient effects caused by the force—velocity relation have disappeared . Fig 9 shows that the force—velocity relation delays the development of half-sarcomere—length heterogeneities . Note that the steepness of the force—length relation determines the rate of formation of half-sarcomere length heterogeneities , i . e . a steeper force—velocity relation decelerates the development of heterogeneities . For this analysis , the simulation with L/L0 = 1 . 2 is considered , in which inhomogeneities develop more slowly then at longer muscle lengths . While model Mxx predicts a high coefficient of variation ( CoV; standard deviation/mean*100% ) of the half-sarcomere lengths for the entire simulation time , the heterogeneities in model MFv appear only gradually but eventually reach the same high level . In contrast , model MATI predicts a constantly low coefficient of variation of the half-sarcomere lengths . In order to test if the proposed model can predict the experimentally observed reduction in the half-sarcomere—length heterogeneity after active stretch [9] , the previous model is slightly modified . To obtain an initially inhomogeneous distribution of the half-sarcomere lengths , we either introduced heterogeneous passive material properties or assumed that the maximum number of available cross-bridges per half-sarcomere shows random fluctuations . This was realised by scaling the corresponding material parameters ( either the two isometric Mooney-Rivlin parameters and the two anisotropic parameters , or the maximum isometric force Pmax , for further details see [30] ) in each finite element with a random value from a normal distribution with mean one and a standard deviation of 0 . 025 . For both model variants , the initial passive stretch and the subsequent fixed-length contraction ( phases I and II , Fig 10 ) yielded heterogeneous half-sarcomere lengths . For the model variant with variable passive properties the mean ± standard deviation of the half-sarcomere length distribution were 1 . 05 ± 0 . 01 μm . For the model variant with variable active properties the mean ± standard deviation of the half-sarcomere length distribution were 1 . 05 ± 0 . 03 μm . To verify that the linear approximation of the force—velocity relation is justified , the maximum shortening and lengthening half-sarcomere velocities occurring in these simulations have been analysed . They are in the range of |v| < 0 . 25 vmax . When modifying passive ( Fig 10 left column ) or active ( Fig 10 right column ) material properties , active stretch experiments show qualitatively similar results for total stresses ( Fig 10 top row ) and the evolution of the coefficient of variation ( CoV ) of the half-sarcomere lengths ( Fig 10 bottom row ) in models MATIFv , MATI , and MFv . The transient differences between models MATIFv and MATI during and after the active stretch ( phases III and IV ) are caused by velocity-dependent effects . The differences ( force enhancement ) between the stresses of models MATIFv and MATI compared to the stress of of a fixed length reference contraction ( blue dotted line ) at the same final length after 1 s ( Fig 10 top ) are evoked by actin—titin interactions ( phase IV ) . The simulations of the model variant with distributed active properties show a higher magnitude of ( residual ) force-enhancement . This can be explained by the higher half-sarcomere length heterogeneity for the simulations with this model variant ( cf Fig 10 bottom ) and the resulting higher titin forces . When including actin—titin interactions , i . e . for models MATIFv and MATI , active stretch reduced significantly the CoV of the half-sarcomere lengths ( cf . Fig 10 bottom ) . During phases III and IV , the CoV of the half-sarcomere lengths reduced by approximately 25% compared to its value before the active stretch . This behaviour deviates clearly from simulations with model MFv , where the CoV of the half-sarcomere lengths monotonously increases first moderately and then rapidly to more than 20 . For both model variants , i . e . with either heterogeneous active or heterogeneous passive material properties , the CoV of the half-sarcomere lengths for models MATIFv and MATI is not strictly decreasing and shows for all simulations parts with a positive time derivative ( or at least equal zero ) . This behaviour is a consequence of the model assumption that the mechanical titin properties are more uniformly distributed than the passive and/or active material properties , i . e . in general the CoV of the half-sarcomere lengths decreases when the time derivative of the homogeneous distributed forces is greater than the time derivative of the heterogeneous distributed forces . Moreover , Fig 11 compares the final distribution of the half-sarcomere lengths resulting from the active stretch simulation ( L/L0 = 1 . 05 to 1 . 25 ) of model MATIFv with variable passive material parameters ( cf . Fig 11 left ) and variable active material parameters ( cf . Fig 11 right ) to the distribution of half-sarcomere lengths of a fixed-length contraction at L/L0 = 1 . 25 ( blue ) . For both model variants the half-sarcomere lengths are less heterogeneously distributed after an active stretch than for a fixed length contraction at the same final length . Thus , under the assumption that the material properties of titin are more homogeneously distributed than the passive and/or active material properties , the model can predict and explain , based on actin—titin interaction , the experimentally observed reduction in the half-sarcomere length heterogeneity after active stretch [9] .
Our simulations show that actin—titin interactions can enable stable half-sarcomere operation on the descending limb of the active force—length relation during fixed-length muscle contractions and active stretches . Moreover , it has been demonstrated that the force—velocity relation can delay but not prevent the development of half-sarcomere length heterogeneities . Thus , actin—titin interactions and the force—velocity relation affect the stability of the system differently . It may be speculated that a key function of actin—titin interaction is to enable stable , predictable half-sarcomere operation avoiding large local stretches—that may for example damage the T-tubule system—for muscles working through the entire range of the force—length relation . | Muscle force generation is a complex process depending on muscle length , activation , and other time-dependent properties . Contractions on the descending limb of the force—length relation ( i . e . when the maximum isometric force decreases with increasing muscle length ) are interesting , since this behaviour is expected to result in non-physiological length inhomogeneities of the muscle microstructure . Previously , different mechanisms have been suggested that could have a stabilising effect on the muscle microstructure . However , superposition of several phenomena makes it difficult to separate the influence of a single mechanism in an experimental setup . Within a model , individual phenomena can be knocked out to quantify the influence of a single process . Here we show , using a physiologically motivated model of the whole muscle , that variable mechanical properties of the large protein titin ( controlled by calcium as a second messenger ) can guarantee stability of the muscle microstructure during contractions on the descending limb of the force—length relation . | [
"Abstract",
"Introduction",
"Material",
"and",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"myofibrils",
"classical",
"mechanics",
"muscle",
"tissue",
"skeletal",
"muscles",
"biomechanics",
"muscle",
"contraction",
"simulation",
"and",
"modeling",
"materials",
"science",
"damage",
"mechanics",
"research",
"and",
"analy... | 2017 | A continuum-mechanical skeletal muscle model including actin-titin interaction predicts stable contractions on the descending limb of the force-length relation |
HIV is a highly mutable virus for which all attempts to develop a vaccine have been unsuccessful . Nevertheless , few long-infected patients develop antibodies , called broadly neutralizing antibodies ( bnAbs ) , that have a high breadth and can neutralize multiple variants of the virus . This suggests that a universal HIV vaccine should be possible . A measure of the efficacy of a HIV vaccine is the neutralization breadth of the antibodies it generates . The breadth is defined as the fraction of viruses in the Seaman panel that are neutralized by the antibody . Experimentally the neutralization ability is measured as the half maximal inhibitory concentration of the antibody ( IC50 ) . To avoid such time-consuming experimental measurements , we developed a computational approach to estimate the IC50 and use it to determine the antibody breadth . Given that no direct method exists for calculating IC50 values , we resort to a combination of atomistic modeling and machine learning . For each antibody/virus complex , an all-atoms model is built using the amino acid sequence and a known structure of a related complex . Then a series of descriptors are derived from the atomistic models , and these are used to train a Multi-Layer Perceptron ( an Artificial Neural Network ) to predict the value of the IC50 ( by regression ) , or if the antibody binds or not to the virus ( by classification ) . The neural networks are trained by use of experimental IC50 values collected in the CATNAP database . The computed breadths obtained by regression and classification are reported and the importance of having some related information in the data set for obtaining accurate predictions is analyzed . This approach is expected to prove useful for the design of HIV bnAbs , where the computation of the potency must be accompanied by a computation of the breadth , and for evaluating the efficiency of potential vaccination schemes developed through modeling and simulation .
Vaccination is a medical procedure which has played an essential role in protecting mankind against viral and bacterial infections since the time of Edward Jenner , who developed a vaccine for smallpox over 200 years ago . Although for some diseases caused by viruses , such as measles , a small number of vaccinations provide almost permanent immunity , for other such as influenza , an annual revaccination , which provides only limited protection , is required . Since the measles virus , like the flu virus , undergoes error-prone replication that introduces mutations , it is not clear why the measles vaccination works as well as it does . Recent research [1] suggests that the measles virus remains antigenically monotypic because mutations are almost always lethal , though the reason for that is not known . For HIV , which is the focus of this paper , no approved vaccine exists , although we are now almost forty years into the HIV/AIDS epidemic . As is the case for influenza virus , HIV evolves rapidly so that there exist many different viral strains . Some of them can evade the immune response to a vaccination directed against only a small number of strains . Thanks to the development of antiretroviral therapies [2] , people who are infected by HIV can live essentially normal lives , without succumbing to AIDS . Several years after infection , a small fraction of the HIV infected individuals develop antibodies that are referred to as broadly neutralizing antibodies ( bnAbs ) , i . e . antibodies that are effective against many strains of the virus [3]; an example of a detailed structural study of the germline and mature bnAbs from a single patient is given in Fera et al . [4] . The fact that the immune system can develop bnAbs over time has led to renewed interest in the possibility of developing a vaccine that would be effective against HIV [5 , 6] . The “neutralization breadth” of an antibody is a measure of its ability to recognize and neutralize different variants of the virus . An antibody with a high breadth can recognize effectively many different variants , while a low breadth antibody is more specific . Experimentally , the neutralization breadth of an antibody is evaluated by testing its ability to inhibit a panel of antigens from different clades of the target virus [7] . In what follows we simply write “breadth” when referring to “neutralization breadth” . For HIV , the Seaman virus panel contains 109 representatives of HIV clades A , B , C and circulating recombinant forms [7] . To evaluate the breadth of an antibody , the half maximal inhibitory concentration of the antibody ( IC50 ) is measured for each virus in the panel . The breadth is defined as the fraction of viruses for which the IC50 is less than a given cutoff , usually set to 1 μg/ml . The present paper describes a computational method to estimate the breadth of new HIV antibodies using only the sequence of their heavy and light chains , and the assumption that they form antibody/antigen complexes that are similar to a known crystal structure of an antibody/antigen complex that may not have high sequence homology . Of the known HIV antibodies , we select those that target the CD4 binding site ( CD4bs ) of the HIV Envelope glycoprotein , a binding site used by diverse bnAbs [5] . Because the CD4bs is highly conserved , an HIV vaccine designed to elicit bnAbs that bind to this site would have a high therapeutic potential: bnAbs would likely recognize the conserved core , as well as the variable regions in the neighborhood that are required for the broad-based character to develop [8] . However , a major obstacle to a successful computational design of bnAbs is a lack of accurate methods for computing the antibody breadth . Here , we use machine learning [9] coupled with descriptors obtained from all-atoms models of the antibody/antigen complex to predict the IC50 values for the viruses in the Seaman panel from which the antibody breadth can be estimated . To predict the IC50 we use a supervised artificial neural network [10 , 11] , a multilayer perceptron ( MLP ) ; see Fig 1 . The neural network is trained by backpropagation , where the errors in the outputs are propagated backwards into the artificial network structure to optimize the internal parameters of the model . An essential element for the successful application of artificial neural networks is the availability of a large data set for training the networks [11] . For HIV , experimental IC50 values have been collected in the CATNAP database of neutralizing antibodies [12] , which contains more than 40000 IC50 values; of these , 6179 satisfy the criteria required for our machine learning approach . In the body of the paper , we describe training of the MLP to predict either the actual value of the IC50 ( by regression ) , or whether or not the antibody recognizes the virus ( by classification ) . The outputs of the two artificial neural networks are then used to estimate the breadth of known antibodies . In a Concluding Discussion , we consider the limitation and extension of the methodology described here , as well as its utility for vaccine design .
The first step toward the estimation of the breadth of a given antibody is to evaluate the IC50 for the binding of the antibody with all the antigens in the Seaman virus panel [7] . Here , the IC50 values are estimated using a Neural Network regression model ( an MLP regressor ) , trained over experimental IC50 values . Experimental values are obtained from the CATNAP database , which contains a total of about 40000 IC50 values [12] . The values in the database were filtered to make sure that the amino acid sequences of both the antibody and antigen are available , that at least one crystallographic structure of the antibody bound to a related HIV antigen is known , and that the antibody binds to the CD4 binding site ( CD4bs ) . Upon filtering the database for these attributes , a total of 6179 IC50 values were obtained . Of these , 3864 are reported as exact values , while the remaining 2315 are reported as “greater than” a given cutoff ( e . g . “>50 μg/ml” ) ; see Table 1 . We refer to these as “approximated” values . To train the MLP regression model , only the 3864 exact values are used; the approximate ones were discarded . As inputs for the MLP regressor , we use the values of descriptors of the antibody/antigen bound system obtained from an atomistic model of the complex . A total of 24 descriptors were tested; they fall into four classes: atomic descriptors , protein/protein scoring functions , protein stability scoring functions , and entropy models ( see Methods ) . The procedure for building the model for one antibody/antigen complex and evaluating all the descriptors requires about 13 minutes ( see Methods ) ; for evaluating the values of the subset of 19 chosen descriptors only about 20 seconds are required , once the model has been constructed ( see Methods ) . For training the MLP regressor the available experimental IC50 values are randomly divided in half to obtain a training set and a test set . The training is then performed , and the obtained neural network used to predict the IC50 . The training is repeated 30 times using each time a randomly generated training and validation set , and the results are averaged over the 30 neural network results to estimate the error in the predictions . The correlation between the experimental and computed IC50 is shown in Fig 2 for both the training and the test sets . The Pearson correlation coefficient for the training set is 0 . 686 ( Spearman coefficient is 0 . 682 ) . For the test set , the correlation coefficients decrease to 0 . 410 ( Pearson ) and 0 . 408 ( Spearman ) . This is a significant improvement over the use of individual descriptors , which maximally reach a Pearson correlation coefficient of 0 . 28; see Fig 3 ( last column ) . With these IC50 values , the breadth of known antibodies can be computed and compared with the experimental values . In the CATNAP database , 24 antibodies have a known sequence and bind to the CD4bs with a known crystallographic structure; see Table 1 . The breadth is calculated for all of these using the computed IC50 values for the viruses in the Seaman panel . Specifically , for each virus in the Seaman panel , the IC50 is estimated using the MLP regressor , and the fraction of viruses for which the IC50 is less than 1 μg/ml is counted . Fig 4 ( left ) compares the experimental and computed breadths and their estimated errors ( see Methods ) . For most antibodies the predicted breadth is in reasonable agreement with the experiments , although there is a tendency to overestimate it . The Pearson correlation coefficient is 0 . 800 ( Spearman is 0 . 766 ) indicating a meaningful correlation . The breadth of the antibody CAP257-RH1 is significantly overestimated ( experimental is 0 . 013 , computed is 0 . 61±0 . 30 ) , but this antibody is the one for which only a single exact IC50 value is available , and the computed uncertainty is highest . Given the limited accuracy for the prediction of the IC50 values ( Fig 2 ) , this result is particularly encouraging . It indicates a low sensitivity of the breadth to the actual IC50 values . This is rooted in the definition of the breadth , which does not require the exact value of the IC50 , but only whether the antibody does or does not bind to a given virus . As a further test , all 6179 experimental values are used in the regression model including the “greater than” approximate values . For these , the maximum experimental value is used; e . g . if the IC50 is expressed as >10 μg/ml and >50 μg/ml in two different studies , the value 50 μg/ml is used . The obtained regression model shows an increase in accuracy , in particular an increase ability in ranking different antibodies ( Spearman correlation coefficient increases from 0 . 766 to 0 . 960 ) , but it has a loss in sensitivity for all low-breadth antibodies predicting zero breadth for most antibodies with experimental breadth less than 0 . 4; see Fig 4 ( center . ) As a result , although the accuracy of the slope is improved , the regression line is shifted to the right , yielding worse values . The breadth can be estimated more directly with an MLP classifier , which is trained to predict whether the IC50 is less or greater than the 1 μg/ml cutoff . A higher accuracy in the correlation between experiments and predictions is expected , because a classification is a simpler problem than a regression . The methodology used for the MLP classifier is the same as that employed for the MLP regressor , except that the whole set of 6179 experimental IC50 values is used; i . e . , the 2351 “approximate” values are included . Table 1 lists the antibodies and the number of “exact” and “approximate” IC50 values available for each . The correlation between the experimental and the computed breadth using the IC50 classifier is shown in Fig 4 ( right ) . As expected , the results are significantly better than for the MLP regressor , with the correlation with the experimental breadth increasing from 0 . 800 for the regressor to 0 . 973 for the classifier . Moreover , no outliers are present and the estimated breadths of all 24 antibodies lie on the diagonal with near unit slope . The performance of the MLP classifier is evaluated further by the analysis of its confusion matrix , which contains the percentage of true negative , true positive , false negative and false positive values . As reported in Table 2 , the accuracy in the test set , corresponding to the sum of true positives and true negatives , is 72 . 3±1 . 0 , with an almost equal rate of false positive and false negative of about 13% . Although our primary purpose for developing the MLP classifier is to evaluate antibodies designed to have increasing breadth ( see Concluding Discussion ) , we decided to determine whether it could be used to compute the breadth of putative germlines , as compared to the breadth of the mature antibodies . We consider bnAbs VRC01 , NIH45-46 and 3BNC60 , for which the experimental breadths of the mature antibodies are available , see Table 3 . For the putative germlines no experimental breadths are available , but the expected breadth should be low , as they appears to have no affinity for native HIV antigens [13–15] . The computed breadth values are reported in Table 3 . For the mature antibodies the computed values are in good agreement with the experimental data . This is particularly interesting for the 3BNC60 antibody , for which no published structure of any antibody/antigen complex is currently available . For this reason , 3BNC60 was excluded from the training of the MLP regressor and classifier . However , a bound crystallographic structure is available for its putative germline precursor [16] . Assuming no significant change in the binding pose upon affinity maturation , the germline crystallographic structure was used for both the mature and germline antibody . This is likely to be a valid assumption , as indicated by the VRC01 and NIH45-46 antibodies for which crystallographic structures are available for both mature and germline forms bound to an antigen ( see PDB 4JPK and 3NGB for VRC01 and 5IGX and 3U7Y for NIH45-46 ) . The breadth of antibody 3BNC60 is slightly underestimated ( 0 . 53±0 . 21 vs 0 . 78 ) , but it also has a higher than usual uncertainty . For the putative germlines , the calculated breadths range between 0 . 27 and 0 . 45 . Importantly , in all cases the breadth of the germline is significantly lower than the breadth of the mature antibody . This finding is encouraging , because only mature antibodies were used in the training of the MLP classifier , as no quantitative information for germline binding to HIV viruses is available in the CATNAP database . The MLP classifier is thus biased towards mature antibodies . Finally , we consider the DRVIA7 antibody , which is an immature form of a VRC01-class antibody [17] . Few experimental IC50 values are available for this antibody; they show a breadth of about 0 . 23 . The estimated value by modeling is 0 . 45±0 . 17 , which is again higher than expected , but lower than the VRC01 mature value ( 0 . 79±0 . 08 ) and similar to the germline ( 0 . 45±0 . 23 ) . An important aspect in the present application of machine learning for the prediction of antibody breadth is the choice of descriptors . The 24 descriptors considered can be broadly divided into four categories: atomic descriptors such as buried surface area or number of hydrogen bonds ( 15 descriptors in total ) , protein/protein binding affinity measures ( 4 descriptors: Prodigy , ZRANK , ZRANK2 and DFIRE ) , protein folding scoring functions ( FoldX and two pairwise statistical potentials: RFHA and RFCB ) , and entropy models ( normal mode entropy from two different elastic network models: ENM_EXP and ENM_R6 ) ; see Methods for more details . Their relative correlation coefficients and their correlation with the experimental values are reported in Fig 3 . As expected , some descriptors are highly correlated with each other , such as the number of hydrogen bonds evaluated using different definitions , or the van der Waals energy with the buried solvent accessible surface area . Selecting too few descriptors would risk making the model less general . Consequently , we kept a large number of descriptors , using as discriminant the computational cost needed to evaluate them . From the full set of 24 descriptors , 5 are initially removed due to their computational cost . The remaining 19 descriptors are all very quick to evaluate , about 20 seconds in total for one model , obtaining a 12 . 8x speedup in the calculation of the descriptors with respect to whole set of 24 descriptors . No decrease in accuracy is observed . If too many descriptors are used , there is risk of overfitting . This is avoided here by the very large number of available experimental values . For the MLP classifier 6179 experimental values are available , half of which are used in the fitting . The MLP classifier with 19 descriptors and 10 hidden nodes contains 200 fitting parameters ( see Methods ) , versus the 3089 ( 6179/2 ) experimental values used in the fitting . A ratio between the number of experimental values in the training and number of parameters in the model of ten or greater is suggested [11]; here the ratio is >15 . As supporting evidence , Table 2 shows that the accuracy of the MLP classifier in the training and test sets is similar . We also studied the effects of the number of descriptors used in the MLP classifier on the accuracy of the model as measured by the confusion matrix; see Fig 5 . Each sample in these two plots is a different MLP classifier model fitted over a different set of descriptors . The first graph shows the accuracy of the given model , as measured by the confusion matrix as sum of the percentages of true positive and true negative , while the second one shows the Pearson correlation coefficients obtained when comparing the calculated and experimental breadths . The blue points correspond to models with randomly-chosen descriptors , while the red points represent models obtained by successively removing the descriptor that is most correlated with all other descriptors and less correlated with the experimental IC50 ( see Methods ) . The accuracy starts to decrease significantly when the number of descriptors is less than about 10 . The Pearson coefficient is more robust and starts decreasing at around 7 descriptors . At this point , the maximum cross-correlation ( Pearson coefficient ) between the descriptors is about 0 . 4 . These data suggest that a simpler model of about 7–10 descriptors could be used , which would have ( approximately ) the same accuracy of the 19-descriptors model; see Table 4 . As expected from the earlier analysis , the breadth results converge more rapidly than the accuracy . This raises the question of which descriptors are most important in the prediction of the neutralization ability of an antibody . There are a variety of methods to extract descriptor importance from an artificial neural network , but there is no apparent consensus on which is better , in particular if the descriptors show a relative high level of cross correlation [18] . However , it is possible to extract some information from the correlation of the descriptors with the experimental IC50 values and with the deletion order reported in Table 4 . The protein/protein scoring functions are on average more important than molecular descriptors , in particular the protein folding propensity descriptors ( FoldX ) and the two statistical pairwise potentials for estimating protein stability ( RFHA and RFCB ) . Also , the entropy from elastic network models ( ENM_EXP and ENM_R6 ) plays an important role , with ENM_EXP the last deleted descriptors with a relatively high correlation with the experiments ( 0 . 292 correlation coefficient ) . These highlight the importance of entropy in the binding , and the usefulness of complex protein potentials , with respect to simple molecular descriptors . An issue in machine learning models is how good they are at generalizing to inputs outside their training set [11 , 19 , 20] . In computational studies about HIV neutralization epitopes it is relatively common to have one model trained for each antibody under study [21–23] . By contrast , in this work only one model was trained to predict the neutralization breadth of all antibodies . We note that our training set did contain some experimental data from each antibody under study . This suggests the generalization ability of the model should be examined . Fig 6 shows how the predicted breadth of each antibody changes when an increasing number of experimental data relative to that antibody are included in the training set . At zero , no relevant experimental data are included; e . g . for the VRC01 antibody , at zero no experimental IC50 value of the VRC01 antibody is included in the training set . In these plots , the blue line is the breath computed using all available experimental data in the training set , while the red line is the experimental neutralization breadth . Comparing the plots for different antibodies , about half of them have a good breadth prediction even at zero , and the predicted breadth does not change significantly upon inclusion of more data; see for examples antibodies 1B2530 , 8ANC131 , 8ANC134 , CH103 , VRC01 , VRC07 , VRC-PG20 . For other antibodies the prediction at zero is quite far from the experimental value; e . g . antibodies NIH45-46 , CAP257-RH1 , CH235 . 12 , or N6 . However , all these antibodies converge to the correct experimental value when a relatively small number of experimental values are included in the training set , usually between 10 and 20 , relative to the total number of values available ( see Table 1 ) . The Pearson correlation coefficient between the computed and experimental breadth is 0 . 50 for the classifier with zero experimental data when all but 4 out of the 23 antibodies studied are included in the analysis ( excluded antibodies are N6 , CAP257-RH1 , CH235 . 12 and NIH45-46 , which have the highest error in the computed breadth ) . When the entire set of antibodies is considered , the correlation coefficient is 0 . 14; including just one experimental value per antibody , the Pearson coefficient increases to 0 . 54 , and it reaches 0 . 90 when including only four experimental values . Another point to consider is that the good results obtained for some antibodies are due to their close similarity to other antibodies in the training set; e . g . the 8ANC131/8ANC134 pair and the VRC01/VRC07 pair . These results show that the neural network model is able to predict accurate values for the breadth in instances when a small number of experimental data for the antibody under study , or a close relative , are included in the training set . It is of interest to compare the results from MLP classifier used here with other machine learning techniques . We compared the Neural Network with one hidden layer used in this paper with a Neural Network with two hidden layers , k-nearest neighbors ( kNN ) [24] , random forests ( RF ) [25 , 26] , and Support Vector Machine ( SVM ) [27]; see Methods . All methods produce very similar results , in both the correlation with the experimental breadths and the estimated error in the single breadth values; see Table 5 . This is supporting evidence of the robustness of the results obtained with the MLP . Moreover , it suggests that to improve the results new information about the system needs to be introduced , e . g . in the form of different descriptors .
To the best of our knowledge , we report in this paper the first attempt to estimate the breadth of HIV antibodies by the use atomistic modeling and machine learning techniques . We developed a method based on a Multi-Layer Perceptron ( an artificial neural network ) , which is able to accurately reproduce and predict the breadth of CD4bs targeting HIV antibodies . For the development of such a model , many experimental IC50 values to train the neural networks have to be available . Here we used experimental IC50 values from the CATNAP database [12] . After cleaning and filtering the data , 6179 IC50 values were obtained for the training of the classifier and 3864 for the regressor . These significantly outnumber the number of parameters in the neural network ( 200 ) , reducing the possibility of overfitting . For best results , a small number of experimental values specific for the antibody under study , or a related antibody , need to be present in the training set . For the application of the same techniques to different HIV binding sites or different viruses , such as influenza , sufficient high-quality experimental data would have to be collected and validated . An important concern in any HIV model is the glycosylation of the virus since the HIV Envelope protein is heavily glycosylated . In the CD4bs , the focus of this work , there are at least five glycans ( N197 , N234 , N276 , N363 , and N462 ) that can interact , and interfere , with the antibody binding . One glycan , N276 , is particularly problematic in a comparison between germline and mature antibodies , as it has been observed that mature antibodies , like VRC01 , introduce deletions or mutations to glycine in CDRL1 to avoid clashes and to accommodate the glycan [28 , 29] . This could be a major reason why the current model overestimates the breadth of putative germlines . The tools presented here are important for the computational design and screening of potential HIV antibodies . The usual focus in antibody design is in the optimization of the potency and specificity for a particular antigen [30] . As pointed out in the Introduction , this is only one part of the problem for highly mutable virus like HIV , for which the exposed antigens have a high variability , so that bnAbs are required . Thus , it is important to couple the calculation of the potency with an estimation of the breadth . The present technique for estimating antibody breadth based on IC50 values could prove useful in the computational design of vaccination protocols to elicit HIV bnAbs . Promising computational/theoretical approaches for the design of vaccination strategies include a description of the affinity maturation ( AM ) process [31 , 32] . This is achieved by simulating how antibodies mutate during the AM from the germline precursors to the mature antibodies . One essential step in these simulations is evaluating the breadth of the produced antibodies . Since the method described here requires as inputs only the sequence of the antibody and a template for the bound structure , it is likely to be a useful part of the design process . Moreover , the techniques developed here can , in principle , be applied to other highly variables viruses for which a definition of antibody breadth is important , such as influenza or hepatitis C viruses , if sufficient experimental neutralization data are available for training the models . From preliminary examination of the literature , this appears to be true at least for influenza . An alternative perspective is to consider the evaluation of the breadth from the antibody/antigen binding affinity , see S1 Appendix . This requires an accurate , relatively rapid , method for the calculation of the binding affinity . Research on developing such a methodology is in progress [33] . The use of the “free energy of binding breadth” , rather than the “IC50 neutralizing breadth” , would , in principle , bypass the need for machine learning techniques .
To calculate the breadth of an antibody , the IC50 values for the antigens in the virus panel are required . In this work the experimental IC50 are used to both calculate the experimental breadth and to train a neural network ( see below ) to predict the IC50 values and breadth of new antibodies . Experimental values of the IC50 for different antibody/antigen complexes were obtained from the CATNAP database [12] ( accessed on March 5 , 2018 ) . The IC50 values are presented in the database in two forms . If available , the exact value is reported; if the complex has a low binding affinity , the exact value is not available , and the IC50 is reported to be greater than a given cutoff , e . g . “IC50>50μg/ml” . The full database contains information for 1024 HIV viruses and 334 antibodies , for a total of 23851 exact and 16467 approximate IC50 values ( most antibodies are tested on different viruses ) . Of interest in this work are antibodies binding on the CD4bs . Because our method requires atomistic models of the antibody/antigen complex , IC50 values are selected such that the binding is at the CD4bs ( based on known crystallographic structures of the bound antibody/antigen complex ) , the amino acid sequences of both the antibody and the antigen are known , and at least one crystallographic structure of the complex is experimentally available . After applying these filters , 24 antibodies ( over the 334 available ) are selected , for which a total of 3864 exact and 2315 approximate IC50 values are available; see Table 1 . The protocol described in this work requires building of atomistic models for arbitrary antibody/antigen complexes and computing molecular descriptors to train the neural network models . The first task is to build a full 3D atomistic model of the antibody/antigen complex given only the sequences of the two proteins . A requirement is to have a template of the complex , which is used for homology modeling . We select one template for each antibody , neglecting differences in the antigens . The underlying assumption is that the binding mode of a given antibody to any antigen is very similar and can be approximated to be constant . Moreover , the structure of the antigen is assumed to be conserved over the sequence space . This is not generally true for antigens with significant insertion or deletions in their ( hyper ) variable loops . Because we focus on the CD4bs , the assumption is reasonable . Crystallographic structures of antibody/antigen complexes were used as templates for the 24 antibodies . In most cases , the antibody was not bound to naturally-occurring antigens , but to engineered monomeric ( gp120 ) domains . In these cases , the engineered protein was substituted with the BG505 SOSIP structure ( PDB 5FYJ ) using best-fit RMSD alignment for the superimposition . The PDB codes used for each antibody are reported in Table 1 . Given the sequences of the antigen and the antibody and a template crystallographic structure for the complex , Modeller [34] was used to create a model of the complex . The structure was then refined by CHARMM [35] , including adding missing hydrogen atoms , creating disulfide bonds , and minimizing the all-atom structure by 100 steps steepest descent . The system was then further minimized using OpenMM [36] until the potential energy was changing by less than 1 KJ/mol . The CHARMM36 force field [37] was used for the all-atoms energy minimizations . In the modeling , only one monomer of the Env gp160 trimer was built , the gp41 deleted , and only the variable part of the antibody was kept . Glycosylation was not included . The optimized structure obtained by molecular modeling ( see previous paragraph ) was used for the computation of all descriptors to train the neural networks . 24 molecular descriptors were computed . These are divided in four main classes: atomic descriptors , protein/protein binding scoring functions , protein stability scoring , and entropy models . As “atomic descriptors” we consider descriptors that are simple function of the atomic coordinates . The two most obvious are the electrostatic ( Coulomb ) interaction between the antibody and the antigen ( E_elec ) and the dispersive ( Lennard-Jones ) interaction ( E_vdw ) . These are modeled according to the CHARMM36 force field [37] and computed using OpenMM [36] . To approximate the solvation of the protein , the GBn2 [38 , 39] generalized Born implicit model was used as third descriptor ( E_gbsa ) . No cutoff in either electrostatic or dispersion force was used . Other used descriptors are the buried surface area [40] ( MD_sasa ) upon binding , the number of hydrogen bonds according to Baker & Hubbard [41] ( MD_h1 ) and according to Wernet , Nilsson et al . [42] ( MD_h2 ) definitions , and hydrogen bond energy according to Kabsch & Sander [43] ( MD_h3 ) . These four descriptors were computed using the MDTraj [44] python module . To these , the number of interchain contacts classified according to polarity and charge are added ( a total of six descriptors: IC_AA , IC_PP , IC_CC , IC_AP , IC_CP , IC_AC ) , as well as the apolar and charged non-interacting surface area ( NIS_A and NIS_C ) . These last eight descriptors were computed using Prodigy [45] . More complex descriptors are scoring functions developed for estimating the binding affinity of two proteins . These methods have been developed to score binding modes in protein/protein docking software . Here , they are not used as independent scoring functions ( as originally intended ) , but as descriptors . The used methods are: Prodigy [45] , ZRANK [46] , ZRANKr [47 , 48] , and DFIRE [49] . The third class of descriptors consists of scoring functions optimized to reproduce the folding propensity of a protein or its stability . With these scoring functions , the binding affinity can be estimated as the difference between the “stability” of the complex and that of the separated antibody and antigen . The methods tested are: FoldX [50 , 51] ( sidechains are optimized ) , and two pairwise statistical potentials ( RF_HA_SRS and RF_CB_SRS_OD ) [52 , 53] . The fourth class of descriptors contains two methods used to estimate the entropy change upon binding . Two classical approaches to evaluate the entropy of a macromolecule are the use of the normal mode analysis ( NMA ) or quasi-harmonic ( QHA ) [54] . Neither of these can be used here directly: NMA requires that the structure is at an energy minimum , and QHA requires a large ensemble of structures . Alternatives are Elastic Network Models ( ENM ) [55 , 56] . In these models , the protein is modeled as a set of atoms interconnected by elastic springs , which vibrate around the given input structure . The total energy E of the system is obtained as sum of pairwise potentials , each acting between a pair of atoms whose distance is under a given cutoff: E = ∑ijk ( d0 ) ∙ ( dij−d0 ) 2 . The sum is over all pairs of atoms ij , dij is the distance between them , d0 is the reference distance ( from the input structure ) , and k ( d0 ) is the force constant . k ( d0 ) can take different expressions . The first developed used a constant for each pairs under a given distance cutoff [55] . More accurate models use force constants that depend on the reference distance d0 . In this paper two expressions are tested . In the first ( ENM_EXP ) the force constant decreases exponentially with d0: k ( d0 ) =aexp[ ( d0r ) 2] , with r = 7Å [57] . In the second ( ENM_R6 ) , the force constant decreases proportionally to d0−6:k ( d0 ) =ad0−6 [58] . In both cases an arbitrary proportionality constant a has to be set . Here the value of 1000 is used ( in arbitrary units ) . A 100-fold increase in the force constant caused no significant change in the results . From the energy expression , the Hessian matrix can be calculated , and , upon diagonalization of the mass-weighted Hessian matrix , the normal frequencies are obtained , and from them the harmonic entropy . These calculations are fast and do not require energy minimization as the input structure is considered to be the minimum . A problem with most of the descriptors used in this work is their high sensitivity to the values of the atomic coordinates . To alleviate this problem , six models for each complex are generated with Modeller using random initial seeds , and the descriptors are averaged over the six models . An important factor to consider when designing and choosing descriptors to use in neural networks and regressor models is the computational cost , or timing , needed to compute them . In this work , two are the main steps: first , generate the 3D atomistic model; second , compute all the descriptors . As described above , preparation of the atomistic model consists of three steps . Modeller takes on average 69 seconds to generate one model . CHARMM , used to fix the structure and run a quick minimization takes an insignificant amount of time . A deeper energy minimization with OpenMM takes on average 37 seconds . A total of about 100 seconds are thus needed to generate one model . The time required for each descriptor varies significantly . Most descriptors are computed almost instantaneously , requiring one second or less . There are three exceptions: evaluating the OpenMM-based descriptors ( electrostatic , dispersive and generalized born energies ) takes about 51 second per model , the ENM entropy takes 15 seconds , and FoldX is the slowest descriptor used , requiring 233 seconds on average . Computing all descriptors requires about 5 minutes . Skipping FoldX and OpenMM ( the two most expensive steps ) decreases the time from 5 minutes to 20 seconds . These timings are for one single model . For each antibody/antigen complex six models are used , and the descriptors are averaged . This means that to create the six models 10 . 6 minutes are required , and to get all the descriptors , an additional 30 . 8 minutes are needed , for a total of 41 minutes . Without FoldX and OpenMM , the timing for the descriptors decreases to 2 . 4 minutes , reducing the total time to about 13 minutes . To predict the IC50 values a Multi-Layer Perceptron regressor ( MLP regressor ) is used . This type of artificial neural network takes as input a number of descriptors , parse them in a hidden layer composed , in this case , of ten nodes , and outputs the predicted IC50 value . For the training , the 3864 exact IC50 values are split randomly in half into a training set and a validation set . The total number of parameters Np to fit in the MLP is given by Np = NiNh+NhNo where Ni , Nh , and No are the number of input descriptors , hidden nodes , and outputs . Using 19 descriptors , and requiring one output ( the IC50 value ) , the total number of parameters to fit is 200 . As an empirical rule , the ratio between the available experimental values used in the training and the number of parameters to fit has to be much greater than one . In our case , the number of experimental IC50 values used for the training is 3864/2 , which is 9 times higher than the number of parameters to fit . A logistic activation function is used in the hidden nodes , and the lbfgs solver is used for optimization . Fig 3 shows the cross correlation between all descriptors and the experimental IC50 values . A similar artificial neural network is used to predict whether the IC50 is less or greater than the experimentally-derived 1 μg/ml cutoff , which is used to calculate the antibody breadth . In this case , a Multi-Layer Perceptron classifier ( MLP classifier ) is trained instead of a regressor on the same set of descriptors . The same settings as for the MLP regressor were used , with the only difference that the output is filtered with a logistic function to be between 0 and 1 . Here the total number of experimental values is 6179 . Of these , 3089 are randomly selected and used to train the model . The ratio between the number of experimental values in the training ( 3089 ) and the number of parameters ( 200 ) is more than 15 , decreasing the chance of overfitting . Table 6 compares the composition of the full dataset and a randomly generated set containing 50% of the values . All antibodies are represented in the training set , with the same percentages of binders and not-binders as in the whole set . It was observed that the outputs of both the MLP regressor and MLP classifier would change , in some cases significantly , upon repeated training using different random splitting of the experimental values into training and validation sets , and upon fitting the neural networks using different random seeds . For this reason , we average over 30 independently generated neural networks using each time a new splitting of the training and validation set and a new random seed in the training . The standard deviation between the 30 replicas is used as an estimate of the statistical error . To select which descriptors to use in the neural network , the first criterium was to keep the scoring function as fast as possible , avoiding the more computationally expensive OpenMM energy-based terms and FoldX . Removing these four descriptors does not cause a decrease in accuracy , while reducing the required time by a factor of 12 . 8 . Prodigy was also removed , as it is a linear combination of other descriptors ( no effect is observed in the accuracy after removal ) . This way , the number of descriptors used decreases from 24 to 19 without loss in accuracy . To study the performance of simplified models including fewer descriptors , the models are sequentially simplified by removing the highest correlated descriptor . To do this , the cross-correlation matrix with all descriptors is analyzed to find the pair of descriptors with the highest correlation among themselves; see Fig 3 . The descriptor with the lowest correlation with the experimental IC50 values is removed . The process is repeated until only one descriptor is present . The list of descriptors which are deleted at each step ( and their correlation coefficients ) are reported in Table 4 . All descriptors are normalized to zero mean and unit variance ( standard scaler ) . To verify the robustness of the neural network , the same analyses were repeated using other machine learning methods: k-nearest neighbors ( kNN ) [24] , random forests ( RF ) [25 , 26] , and support vector machine ( SVM ) [27] . First , the MLP were compared with a second MLP containing two hidden layers instead of one . Having observed the same results , it was compared with a kNN ( with distance weighting ) , a RF ( composed of 31 trees ) and a SVM ( with radial basis function kernel ) and no significant improvement was observed with any of them . All machine learning methods ( MLP , kNN , RF , SVM ) and the standard scaler are used as implemented in the scikit-learn [59] python module . | Although we are now almost 40 years into the AIDS epidemics , no approved vaccine for HIV exists . This is due to the high mutability of HIV , which allows it to escape the immune system control . Nevertheless , a few long-infected patients have been able to develop antibodies , called broadly neutralizing antibodies ( bnAbs ) , that have a high neutralization breadth and are effective against a variety of viral strains . The knowledge that bnAbs can develop over time suggests that a HIV vaccine can be found that short-circuits the production of bnAbs . In this study we present a computational approach to estimate the breadth of HIV antibodies . Experimentally , the breadth is the fraction of viruses in a panel that are neutralized by the antibody , where the ability to neutralize is quantified as the half maximal inhibitory concentration of the antibody ( IC50 ) . A method to estimate the IC50 by computer modeling and machine learning is described and used for estimating the antibody breadth . This approach is likely to prove useful in the design of new antibodies effective against HIV and for testing the efficiency of theoretically-designed vaccination protocols . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"crystal",
"structure",
"pathology",
"and",
"laboratory",
"medicine",
"neural",
"networks",
"pathogens",
"immunology",
"condensed",
"matter",
"physics",
"microbiology",
"neuroscience",
"retroviruses",
"virus... | 2019 | Estimation of the breadth of CD4bs targeting HIV antibodies by molecular modeling and machine learning |
Among Caliciviridae , the norovirus genus encompasses enteric viruses that infect humans as well as several animal species , causing gastroenteritis . Porcine strains are classified together with human strains within genogroup II , whilst bovine norovirus strains represent genogroup III . Various GI and GII human strains bind to carbohydrates of the histo-blood group family which may be shared among mammalian species . Genetic relatedness of human and animal strains as well as the presence of potentially shared ligands raises the possibility of norovirus cross-species transmission . In the present study , we identified a carbohydrate ligand for the prototype bovine norovirus strain Bo/Newbury2/76/UK ( NB2 ) . Attachment of virus-like particles ( VLPs ) of the NB2 strain to bovine gut tissue sections showed a complete match with the staining by reagents recognizing the Galα1 , 3 motif . Alpha-galactosidase treatment confirmed involvement of a terminal alpha-linked galactose . Specific binding of VLPs to the αGal epitope ( Galα3Galβ4GlcNAcβ-R ) was observed . The binding of Galα3GalαOMe to rNB2 VLPs was characterized at atomic resolution employing saturation transfer difference ( STD ) NMR experiments . Transfection of human cells with an α1 , 3galactosyltransferase cDNA allowed binding of NB2 VLPs , whilst inversely , attachment to porcine vascular endothelial cells was lost when the cells originated from an α1 , 3galactosyltransferase KO animal . The αGal epitope is expressed in all mammalian species with the exception of the Hominidaea family due to the inactivation of the α1 , 3galactosyltransferase gene ( GGTA1 ) . Accordingly , the NB2 carbohydrate ligand is absent from human tissues . Although expressed on porcine vascular endothelial cells , we observed that unlike in cows , it is not present on gut epithelial cells , suggesting that neither man nor pig could be infected by the NB2 bovine strain .
Caliciviruses are small non enveloped viruses approximately 27–35 nm in diameter with a positive-sense single-stranded RNA genome of 7 . 4 to 8 . 3 kb in size . Based on genomic organization and genetic analysis , the Caliciviridae family is divided into five genera , norovirus , sapovirus , vesivirus , lagovirus and becovirus ( or nabovirus ) , and a sixth genus “recovirus” has been recently proposed [1] . Animal caliciviruses are suspected and confirmed causes of a wide spectrum of diseases including gastroenteritis ( pigs , calves , cats , dogs and chickens ) , vesicular lesions and reproductive failure ( pigs and sea lions ) , respiratory infections ( cats and cattle ) and a fatal hemorrhagic disease ( rabbits and hares ) [2] . Human and animal caliciviruses associated with gastroenteritis belong to the norovirus and sapovirus genera . The genus norovirus ( NoV ) has been divided into five genogroups , genogroups I to V [3] . More recently , a classification including two additional genogroups ( VI and VII ) has been suggested [4] . Human strains are classified into genogroups I , II , IV , VI and VII . Analysis of the complete ORF2 sequences , encoding the capsid protein , of genogroups I and II demonstrates high diversity and at present , 8 clusters have been defined within genogroup I ( GI-1 to GI-8 ) and 19 within genogroup II ( GII-1 to GII-19 ) . Porcine NoV have been classified into GII-11 and recently into two novel genotypes ( GII-18 and GII-19 ) [5] . NoV are also detected in calves . The first bovine NoV strain , Bo/Newbury2/1976/UK , was isolated from calves with diarrhea in the United Kingdom [6] . Later , another distinct genotype of bovine NoV , Bo/jena/78/GER was identified in Germany [7] . They belong to genogroup III [8] , [9] , [10] , in which the Jena virus and the Newbury2 ( NB2 ) are respectively the prototype of genotypes GIII-1 and GIII-2 . Two other enteric bovine caliciviruses have been described , the Newburry agent 1 in the UK [11] and Nebraska strain in the USA [12] . Thus , a fifth genus named becovirus ( or nabovirus ) includes these two bovine viruses since they present significant differences with the other genera of the Caliciviridae family [13] . Human NoVs ( HuNoV ) have been found to recognize histo-blood group antigens ( HBGAs ) , with different strains showing distinct specificities [14] , [15] . HBGAs are complex glycans present on many cell types including red blood cells and vascular endothelial cells , as well as on the epithelia of the gastrointestinal , uro-genital and respiratory tracts . They can also be present in a soluble form in biologic fluids such as saliva and milk . HBGAs are synthesized from a series of precursor structures by stepwise addition of monosaccharide units via a set of glycosyltransferases . According to the CAZY classification ( http://www . cazy . org/ ) , three glycosyltransferases families are involved in their biosynthesis . The GT11 , GT10 and GT6 families which encode α1 , 3/4fucosyltransferases , α1 , 2fucosyltransferases and enzymes related to the A and B enzymes of the ABO system , respectively . Some of the corresponding genes are polymorphic whilst others are expressed in a species-specific manner . In humans , the pleiotropic interaction of alleles at three loci , FUT3 , FUT2 and ABO determines the Lewis , Secretor , and ABO phenotypes , respectively [16] . The corresponding antigens are involved in HuNoV recognition of human digestive tissue and are required for infection [14] , [15] . The GT6 gene family ( ABO family ) comprises three other members which encode α1 , 3galactosyl or N-acetylgalactosaminyltransferases . An α1 , 3galactosyltransferase acts on the type 2 precursor disaccharide ( Galβ4GlcNAc ) to give the αGal epitope , also called Galili antigen ( Galα3Galβ4GlcNAc ) , which is expressed in all mammalian species except hominids since in humans , gorilla and chimpanzee the GGTA1 ( glycoprotein , alpha-galactosyltransferase 1 ) gene has been inactivated by several mutations and is therefore a pseudogene [17] . Another gene of the same family encodes a distinct α1 , 3galactosyltransferase which acts on the glycosphingolipid lactosylceramide to generate the isoglobotrihexosylceramide ( iGb3: Galα3Galβ4Glcβ-Cer ) . The corresponding iGb3 synthase gene does not appear to be functional in humans , although this is actively debated [18] , [19] . Finally , the last enzyme of the GT6 family known at present is an α1 , 3N-acetylgalactosaminyltransferase which acts on the glycosphingolipid called globoside ( Gb4 ) or P blood group antigen to generate the Forsmann antigen ( GalNAcα3GalNAcβ3Galα4Galβ4Glc-Cer ) . This enzyme , the Forsmann synthase , is not active in humans [20] . Recently , our group described other genes of the GT6 family ( GT6m5 to GT6m8 ) . However , they are not present in all vertebrate or mammalian genomes and the enzyme activity of the corresponding proteins have not been characterized as yet [21] . The genetic and antigenic relatedness of human and animal noroviruses suggests the possibility for inter-species transmission as illustrated by the recent detection of sequences close to GII-4 HuNoV in swine and cattle in Canada [22] , [23] . Although animal NoVs have not yet been isolated from human , human infection with NoV related to genogroup III bovine NoV has been suggested by the presence of serum antibodies against bovine GIII-2 among veterinarians in the Netherlands [23] . Moreover the use of phylogenetically conserved cellular receptors appears as another risk factor for cross-species transmission . Attachment of the virus to a host ligand constitutes a first step of the viral infection process , and it has been observed that when the receptor is conserved between several species , these are more likely to be infected by viruses that use the shared receptor [24] . HBGAs can be conserved across high phylogenetic distances as shared epitopes have been found between bacteria , invertebrates , plants and mammal [25] . ABH-related structures have been characterized in the gut of all mammalian species tested so far [26] . The use of such molecules as primary ligands by HuNoV strains as well as by RHDV [27] , a rabbit calicivirus of the lagovirus genus , prompted us to look for the ability of a bovine NoV to recognize HBGAs potentially present on bovine , as well as on porcine and human digestive epithelial cells . Such a shared ligand could help these viruses to propagate between the three species which live in close contact and at high densities in areas of intensive breeding .
We previously demonstrated the binding of VLPs of caliciviruses to carbohydrate epitopes using immunohistochemistry as a starting method [27] , [28] . We now used the same method in order to determine if the bovine NB2 strain of norovirus could similarly bind to a carbohydrate ligand expressed in the gut . Tissue sections prepared from the entire bovine gut were thus incubated with rNB2 VLPs and their binding was detected using antibodies . Specific staining , only visible in the presence of VLPs , was readily observed as shown on Fig . 1 . In the duodenum , rNB2 VLPs attached to the epithelial cells of the crypts of Lieberkühn located at the surface of the mucosa ( Fig . 1A ) , but not to the epithelial cells of the Brünner's glands which are located deeper in the mucosa ( Fig . 1B ) . Unlike what was previously observed for other caliciviruses , binding of the VLPs was not restricted to epithelial cells . It was also observed on vascular endothelial cells and erythrocytes . These are clearly visible surrounding Brünner's glands ( Fig . 1B ) or in the serosa layer ( Fig . 1C ) . In addition , other cell types such as smooth muscle cells were labelled , albeit less strongly . In order to determine if the binding involved recognition of carbohydrate epitopes , tissue sections were pretreated with sodium periodate prior to incubation with the VLPs since periodate oxidation cleaves C-C bonds with vicinal hydroxyl groups as found on sugars . Sialic acid residues are more sensitive to mild oxidation than other sugars . Since many viruses are known to use sialic acids as ligands [29] , bovine duodenal tissue sections were first treated with 1 mM sodium periodate . This treatment did not affect the staining after incubation of the rNB2 VLPs ( data not shown ) . Tissue sections were thus pretreated with 10 mM periodate . At that concentration , the staining was completely lost ( Fig . 1E and 1F ) , suggesting that rNB2 VLPs recognize a neutral glycan structure expressed both on digestive surface epithelial cells and other cell types including vascular endothelial cells . Since other norovirus strains are known to bind to neutral carbohydrates of the histo-blood group family , we next sought to determine the expression of such epitopes throughout the bovine gut in order to relate it to the binding of rNV VLPs . To this aim , a set of antibodies as well as the UEA-I and GS1-B4 lectins were used and we observed that some , but not all , of these reagents clearly labeled bovine gut tissue sections . The A histo-blood group antigen was detected , but not the B antigen . Strong positivity was also observed using reagents that detect H type 2/LeY epitopes . In contrast , very little or no staining was detected using reagents specific for type 1 or type 3 based structures . Only the anti-Lea antibodies stained scattered goblet cells in the duodenum and colon , whereas the anti-H type 1 , anti-H type 3 and anti-Leb did not give detectable specific labeling . Similar to that observed for rNB2 VLPs ligands , the A and H type 2/LeY expression was most intense in the pyloric and duodenal surface epithelia and gradually decreased from distal duodenum to disappear from the distal part of the digestive tract ( Fig . 2 ) . Yet , the expression of these fucosylated structures was observed on epithelial cells only and therefore did not match with that of the VLPs . In addition to A and H histo-blood group epitopes , the αGal epitope detected by either a mAb or the GS1-B4 lectin was detected in the bovine gut . The staining obtained with these reagents matched that of the VLPs since , as with the anti-A and H type 2/LeY , it was maximal in the gastro-duodenal area and absent from the distal part of the gut ( Fig . 2 ) , but more specifically , since the labeled cells and the relative intensities of labeling were the same as those observed following incubation with rNB2 VLPs as described above ( Fig . 1G and 1H ) . The pattern of labeling obtained with the anti-αGal reagents was thus indistinguishable from that obtained with the VLPs , suggesting that the latter might recognize a carbohydrate structure related to the αGal epitope . HBGAs were first characterized on human erythrocytes and it has been previously shown that hemagglutination ( HA ) can be used to define the HBGA specificity of human noroviruses [30] . We thus tested the ability of rNB2 VLPs to agglutinate bovine or human red blood cells . A strong agglutination of bovine erythrocytes was obtained both at 4°C and room temperature , but no agglutination at all was detected using human erythrocytes at either temperature , irrespective of their ABO phenotypes . In contrast , a GII . 4 strain agglutinated human O blood group red blood cells , but not bovine erythrocytes ( Fig . 3A ) . Since HBGAs are also present in saliva , we next assayed the binding of rNB2 VLPs to bovine and human saliva samples . When human saliva samples were assayed , no signal above background was obtained . In contrast , using the same amount of VLPs from a GII . 4 strain , all human saliva samples from secretor individuals were strongly recognized , irrespective of their ABO phenotype . As previously observed for other human GII . 4 strains , saliva samples from nonsecretor individuals were not recognized , showing specificity of the binding ( Fig . 3B ) . On bovine saliva samples , rNB2 VLPs binding was readily detected , albeit with highly variable OD values , some samples giving a strong signal and others a signal at background level only ( Fig . 3C ) . Nearly identical results were obtained with human natural anti-αGal antibodies ( Fig . 3E ) . The binding of the UEA-I lectin and of anti-A and anti-B mAbs were assessed on the same set of samples . No binding of the anti-B was detected ( data not shown ) . As depicted on Fig . 3D and 3F , the binding of the UEA-I lectin and of the anti-A mAb were heterogeneous too , but they were not related to each other , nor to that of the rNB2 VLPs . This indicates that individual bovine saliva samples are polymorphic with regard to the presence of either the A or H antigens . In addition , the lack of concordance between either the A or H antigens expression and the binding of rNB2 VLPs or human natural antibodies shows that individual differences in rNB2 VLPs binding were not due to non specific heterogeneity of the total amount of salivary glycans but were due to a true heterogeneity of the expression of the capsids ligand and of the αGal epitope . Taken together , these results clearly indicate that the ligand recognized by rNB2 VLPs is distinct from A , B or H antigens , but not from the αGal epitope and that it is not present on human red blood cells or in human saliva . In order to define the carbohydrate specificity of rNB2 VLPs , their ability to recognize a set of HBGAs related oligosaccharides was determined using an ELISA-based binding assay . The structure of all the oligosaccharides tested is given in Table 1 . Binding was observed on two structures which share a common terminal galactose in α1 , 3 linkage , contrasting with the binding pattern of the human NV strain ( Fig . 4A ) . None of the other tested oligosaccharides allowed binding above background . One of the αGal-terminated structures recognized by the bovine VLPs is fucosylated on the N-acetylglucosamine residue , generating a Lewis X epitope ( αGal-Le x ) . The presence of this fucose residue partially impairs recognition as shown on Fig . 4B . Thus the preferred structure recognized by rNB2 VLPs among those tested is the trisaccharide Galα3Galβ4GlcNAc . Noticeably , the B blood group antigen which also presents a terminal galactose in α1 , 3 linkage was not recognized , consistent with the lack of agglutination of human B blood group erythrocytes and with the lack of binding to human saliva samples from B secretors . In order to determine if the ligand recognized in bovine saliva and digestive tissues corresponds to the αGal antigen , the effect of α–galactosidase treatment was tested . A saliva sample that allows good binding of rNB2 VLPs was chosen ( sample # 8 ) . The binding of VLPs was completely lost following the enzyme treatment . Activity of the enzyme was controlled using the Galα3Galβ4GlcNAc conjugate and binding to the treated conjugate was similarly lost after α–galactosidase treatment ( Fig . 5A ) . The same treatment was applied to a bovine duodenal tissue section and staining completely disappeared ( data not shown , see below for porcine tissues ) . In addition , human natural anti-αGal antibodies could inhibit the binding of rNB2 VLPs to bovine saliva samples ( Fig . 5B ) . These results indicate that rNB2 VLPs specifically recognize the αGal antigen of the HBGAs family in bovine saliva and digestive tract . Binding of the Galili disaccharide ( Galα3GalαOMe ) to rNB2 VLPs was studied using saturation transfer difference ( STD ) NMR experiments [31] . This technique allows identification and characterization of ligand binding to large receptor proteins and yields binding epitopes of the ligand molecules at atomic resolution . It has been shown lately that the technique is well suitable for the investigation of carbohydrate receptor recognition by caliciviruses [32] . STD NMR spectra of Galα3GalαOMe and of the methyl glycoside of the blood group B trisaccharide in the presence of rNB2 VLPs at a saturation time of 2 s are shown in Fig . 6 . The Galili disaccharide yielded sizable STD effects whereas no response at all was observed for the B-trisaccharide . This is in accordance with the biological assays that could not detect any binding to the B-antigen and underlines the strict specificity of the VLPs for the disaccharide moiety Galα3GalαOMe . STD NMR experiments of Galα3GalαOMe in the presence of rNB2 VLPs with increasing saturation times yielded STD build-up curves for individual protons . Fitting of these data to a single-exponential function ( cf . Materials and Methods ) delivered the binding epitope . It has been shown that strong spin-diffusion within large virus like particles requires the acquisition of build-up curves instead of single point measurements in order to generate reproducible binding epitopes . The binding epitope of Galα3GalαOMe derived from this analysis is shown in Fig . 7 . It is remarkable that the protons around the glycosidic linkage received the largest fraction of saturation . This suggests that the α3-glycosidic linkage is in intimate contact with the VLP binding pocket . Obviously , the Galili disaccharide represents the central recognition element . The αGal is not expressed in apes due to inactivation of the GGTA1 gene encoding the α1 , 3galactosyltransferase . Yet , it has been detected in all other mammalian species tested so far [17] . The lack of agglutination of human red blood cells and of binding to human saliva by rNB2 VLPs is consistent with the inability of humans to synthesize this antigen . In addition , no binding to the human duodenal mucosa was observed , confirming that rNB2 VLPs do not cross react with a human epitope present in the human gut ( Fig . 1I ) . Since pigs are known to be infected by noroviruses , and since they are known to express the αGal antigen , we tested the binding of rNB2 VLPs to the porcine gut by immunohistochemistry . A specific staining was readily observed throughout the digestive tract but surprisingly , it was restricted to non epithelial cells , including the vascular endothelium and to a lesser extent smooth muscle cells ( Fig . 1J ) . Porcine duodenal tissue sections were then treated with the α–galactosidase from green coffee beans . Like on bovine tissue sections , the treatment completely abolished the binding of rNB2 VLPs to these cell types ( Fig . 1K and 1L ) . The distribution of the αGal antigen in porcine digestive tract was then tested using an anti-αGal mAb and the GS1-B4 lectin . We observed that the two reagents gave stainings completely parallel with that obtained with the rNB2 VLPs ( data not shown ) . This indicates that in porcine tissues , the αGal epitope is not expressed by digestive epithelial cells although it is present on other cells types recognized by the bovine viral capsids . Since the lack of αGal antigen in humans is due to inactivation of the GGTA1 gene , we tested whether this event was sufficient to have caused the lack of recognition of human cells by rNB2 VLPs . Transfection of human HEK 293 cells with the functional rat Ggta1 cDNA allowed expression of the αGal antigen as detected by flow cytometry using the GS1-B4 lectin . Control HEK 293 cells that lacked the αGal antigen were barely recognized by rNB2 VLPs . In contrast , a clear binding was observed on the Ggta1 transfected human cells . Inversely , pig vascular endothelial cells spontaneously express the αGal antigen as documented from the labeling by the GS1-B4 lectin . The attachment of rNB2 VLPs to such cells was clearly detected by flow cytometry . However , we observed that pig vascular endothelial cells from a Ggta1 KO pig , which accordingly are not stained by GS1-B4 , were no longer recognized by the bovine VLPs ( Fig . 8 ) . These results confirm that rNB2 VLPs bind to the αGal antigen and that the expression of a functional α1 , 3galactosyltransferase is necessary and sufficient to allow their attachment to mammalian cells .
RNA viruses present a high risk of cross-species transmission since they are over-represented in the list of pathogens known to have crossed the species barrier [33] . This is most likely due to their particularly high mutation rate which allows them to evolve fast , providing rapid adaptation to a new host species . The use of phylogenetically conserved ligands facilitates the crossing of the species barrier . Since among Caliciviruses , human strains of norovirus and the rabbit hemorrhagic disease virus bind to HBGAs and since HBGAs can be phylogenetically conserved , we tested the possibility that the prototype of the bovine NoV , the Newbury2 strain classified in the GIII . 2 cluster , could use such a conserved carbohydrate ligand to bind to bovine , porcine or human digestive epithelial cells . Our results demonstrated that VLPs from NB2 attach to the surface of the bovine duodenal epithelium by recognition of the αGal epitope of the HBGAs family and that this ligand cannot be used to infect either man or pig . This conclusion is based on the following observations: ( 1 ) the tissue distribution of rNB2 VLPs binding sites in the three species perfectly matched that of reagents specific for the αGal epitope which is absent from human and pig duodenal epithelial cells ; ( 2 ) among many immobilized HBGAs-related synthetic oligosaccharides , only those presenting the αGal epitope supported binding of the VLPs ; ( 3 ) α-galactosidase treatment of either tissue sections or bovine saliva completely impaired the binding of rNB2 VLPs ; ( 4 ) STD NMR experiments confirmed recognition of the αGal epitope at the atomic level ; ( 5 ) transfection of human cells with an α1 , 3galactosyltransferase cDNA allowed binding of rNB2 VLPs whilst inversely , their binding to porcine vascular endothelial cells was lost on cells from an α1 , 3galactosyltransferase KO pig . The αGal epitope is structurally related to the histo-blood group antigen B type 2 since both share a terminal galactose in α1 , 3 linkage and the type 2 backbone structure ( Galβ4GlcNAc ) . They only differ by the fucose residue of the B antigen , allowing some reagents such as some anti-B mAbs or the GS1-B4 isolectin to cross-react . Nevertheless , the possibility that rNB2 VLPs could recognize a B blood group epitope in addition to the αGal antigen is very unlikely since it failed to agglutinate human B blood group erythrocytes and to bind to human saliva from B blood group individuals of the secretor phenotype who strongly express B epitopes in their saliva . In addition , at the atomic level it was found by STD NMR experiments that the α3-glycosidic linkage serves as the central recognition element . Any disturbances close to this region such as the addition of a fucose residue in the 2-position of the reducing galactose moiety to yield the B-antigen would therefore impede with the binding process . One may speculate whether other positions that are more remote from the glycosidic linkage may be modified so as to obtain a better binder . Interestingly , we did not detect B blood group reactivity on cow tissue sections or saliva , making less likely the possibility for GIII . 2 strains to evolve toward cross-recognition of the αGal and the B epitopes . Besides the species-specific expression of the αGal antigen on bovine digestive epithelial cells , another species-specific bovine characteristic was evidenced with regard to HBGAs expression . Indeed , we failed to detect HBGAs based on type 1 precursor ( Galβ3GlcNAc ) in bovine tissues or saliva , consistent with the results from earlier structural analyses of O-linked oligosaccharides from bovine salivary mucins which described the presence of type 2-based structures only [34] , [35] . Those studies also failed to detect the αGal epitope . However , it could either be carried by N-linked glycans of salivary glycoproteins or the saliva studied could have originated from animals that did not express the epitope , consistent with our observation that not all bovine saliva samples could support attachment of rNB2 VLPs . Human NoV of either the GI or GII genogroups recognize HBGAs motifs based on both type 1 and type 2 precursors , but show stronger binding to type 1-based structures , particularly at 37°C [36] , [37] . Human small intestine presents HBGAs based on type 1 as well as on type 2 backbones , suggesting adaptation of these strains to their host glycans . Even though cows can express A and H type 2 or Ley antigens in their digestive tract , they may be less sensitive to infection by human NoV strains due to the lack of type 1-based structures . Interestingly , using saliva samples , we observed that similar to man and pig , cows are polymorphic regarding expression of the A and H antigens since some cows did not express either A , H or both . As these polymorphisms were unrelated to the αGal expression , the combination of the A , the H and the αGal polymorphisms is expected to generate eight subgroups of bovine and therefore significant individual variation which may be related to host-pathogens interactions . Regardless , our results do not prove that the αGal ligand is necessary for infection of cows by GIII . 2 strains , but several aspects support that possibility . Various human NoV strains that bind to HBGAs have been shown to infect their host in an HBGA-dependent manner [15] . Likewise , we recently obtained indirect evidence that the binding of RHDV , a lagovirus , to the H type 2 antigen is necessary for infection [38] . Thus , the conservation of the binding ability of an HBGA motif by the Newbury2 strain suggests that it may also be required for infection . Furthermore , histopathological analysis of the lesions of calves experimentally infected with NB2 indicated that they were restricted to the proximal small intestine [11] , which coincides with the main site of expression of the αGal antigen on digestive epithelial cells . Finally , there is no clear evidence that bovine NoV can infect an other species [8] . Such viruses have never been detected in human or porcine samples , suggesting that they do not circulate in those species . This is to be expected if the αGal antigen serves as a receptor for infection . Nonetheless , one study described the presence of anti-GIII . 2 antibodies in the serum of veterinarians in the Netherlands [23] . Bovine NoVs share cross-reactive epitopes with human NoVs [39] , [40] . This cross-reactivity may explain the detection of anti-GIII . 2 in some human serum samples . In absence of cell culture models , it is very difficult to prove that a ligand is truly a receptor . The above described demonstrations that HBGAs can be compulsory ligands have been obtained through the analysis of the effect of their polymorphism on infection [41] , [42] , [43] , [44] . Interestingly , we observed that the rNB2 VLPs salivary binding assay could distinguish between binder and non binder cows . If that polymorphism is also present at the level of digestive epithelial cells , it could be used to experimentally assay the sensitivity of either group to infection by the Newbury2 strain . To date recombinants of bovine NoV and HuNoV have been identified and appear to be of frequent occurrence [4] , [45] , [46] , [47] , [48] . Cattle co-infection by a GIII . 2 bovine strain and a human NoV could thus lead to the emergence of recombinant strains able to infect humans . However , this seems unlikely since as discussed above , the lack of type 1-based HBGAs in cow digestive epithelium may decrease recognition by HuNoVs , and since GI , GII and GIII strains are genetically distant and accordingly , inter-genogroup recombination , although recently observed , should be much less frequent than intra-genogroup recombination [4] , [49] . Recent crystallographic analyses of the capsid protein domain of a GI . 1 and a GII . 4 NoV interacting with oligosaccharides showed that the two strains use distinct binding sites on the capsid protein protruding domain , although they bind to very similar oligosaccharides [50] , [51] . A fucose residue is involved in both instances , although it is more essential to the GII . 4 binding site than to the GI . 1 binding site . Here we showed that the best binder of a GIII . 2 strain is a related carbohydrate structure devoid of fucose and that addition of an α1 , 3-linked fucose to the backbone N-acetylglucosamine impaired recognition . It will thus be interesting to define the mode of recognition of the αGal trisaccharide by the NB2 strain in order to know if its binding site corresponds to one of those already characterized for either GI or GII strains . This knowledge should provide crucial information to understand how NoVs adapt to their host species and evolve to maintain recognition of diverse HBGAs that allow binding to allotypic and/or xenotypic host molecules . The Galili antigen has mainly been studied in the context of xenotransplantation since it was originally observed that humans naturally produce antibodies against it and since these antibodies were shown to be the primary cause of hyperacute rejection of pig xenografts organs in human and hominids [17] . But the reason why the GGTA1 gene has been inactivated in the Hominidaea lineage an estimated ∼28 MA ago is unclear . The loss of the αGal epitope allows the generation of so-called natural antibodies , similar to the generation of anti-A or anti-B natural antibodies in the ABO allogenic system , probably because some bacteria express identical or cross-reactive epitopes . These anti-αGal antibodies can recognize pathogens that carry the xenogenic epitope . Thus , envelopped viruses produced in animal cells that express a functional α1 , 3galactosyltransferase carry the αGal antigen on their envelope glycoproteins and the presence of natural antibodies directed against this epitope in human serum leads to their rapid elimination [52] , [53] . For this reason it has been proposed that inactivation of the GGTA1 gene in the Hominidaea lineage may have allowed escape from a highly pathogenic virus thanks to the natural anti-carbohydrate antibodies [17] . Our observation that an animal pathogen can use the αGal antigen as a ligand additionally suggests that the loss of GGTA1 may have allowed escape from some NoV strains . Clearly , at present GIII NoVs are of moderate or low pathogenicity . Nevertheless virulence evolves and past NoVs may have been much more virulent than present ones . The loss of the αGal ligand may have contributed to escape dreadful past NoV epidemics in hominids . Alternatively , the loss of the GGTA1 enzyme may have been completely independent from NoV infection . In this case , NoVs would have more recently evolved carbohydrate-binding specificities adapted to a broad spectrum of mammalian species . Regardless , the exquisite specificity of the NB2 strain for the αGal epitope is well adapted to its bovine host , but inversely should restrict its possibilities of cross-species transmission .
All animals were handled in strict accordance with good animal practice as defined by the relevant national and/or local animal welfare bodies . The BG-4 anti-H type 1 specific monoclonal antibody ( mAb ) was purchased from Signet laboratories ( Dedham , CA ) and mAbs 19-OLE , 7-LE and 2-25LE were obtained from Dr . J . Bara ( CNRS , Villejuif , France ) . They are an anti-H type 2 showing a slight cross-reactivity with Ley ( unpublished results ) , an anti-Lea and an anti-Leb , respectively . The anti-H types 3 and 4 mAb MBr1 was purchased from Covalab ( Villeurbanne , France ) . The anti-A ( all types ) mAb 9113D10 was obtained from Diagast ( Loos , France ) . The anti-B mAb ED3 was a kind gift from Dr . A . Martin ( CRTS , Rennes , France ) . The anti-αGal 4F102c2 was a kind gift from Dr . A . Bendelac ( Howard Hughes Medical Institute , Chicago , IL ) . The lectin from Griffonia simplicifolia B4 isolectin 1 ( GS1-B4 ) , either peroxidase or fluorescein isothiocyanate ( FITC ) conjugated , which recognizes α1 , 3-linked terminal galactosyl residues , was purchased from EY Laboratories ( San Mateo , CA USA ) and Vector Laboratories ( Burlingame , CA USA ) , respectively . The lectin from Ulex europaeus ( UEA-1 ) peroxidase conjugated , which recognizes H type 2 and Ley was obtained from Sigma ( St Louis , MO ) . Alpha-galactosidase from green coffee beans was purchased from Sigma . The anti-GII and anti-GIII rabbit polyclonal antisera were prepared at the Veterinary School of Nantes by immunizing rabbits with VLPs from GII . 4 ( Dijon 171/96 ) and GIII . 2 ( NB2 ) strains , respectively . Synthetic oligosaccharides as polyacrylamide conjugates were prepared as described previously [54] , [55] . Oligosaccharides coupled to HSA ( human serum albumin ) were obtained from IsoSep AB ( Tulligen , Sweden ) . The structure of all oligosaccharides used is given on Table 1 . The disaccharide Galα3GalαOMe was obtained from Calbiochem . The methyl glycoside of the B antigen has been synthesized enzymatically [56] . The recombinant virus-like particles were prepared by infecting High-five insect cells with recombinant baculoviruses according to a previously described method [57] . The Dijon GII . 4 171/96 strain [58] and the GI . 1 Norwalk strain ( NV ) constructs were kind gifts of Dr . E . Kohli ( University of Burgundi , Dijon , France ) and Dr . X . Jiang ( Cincinnati Children's Hospital Medical Center , Cincinnati , Ohio , USA ) , respectively . Five days post-infection High-Five lysed cells and media were centrifuged at 4500 rpm for 15 min and the supernatants collected and centrifuged at 25 , 000 rpm for 3h30 in a SW28 rotor . The pellets were resuspended in distilled water and submitted to 2 rounds of purification on a sucrose gradient . The gene encoding the capsid protein of the NB2 norovirus cloned into pFastBac vector ( InVitrogen ) was a kind gift of Drs S . Oliver and J . Bridger ( Royal Veterinary College London ) . Competent E . coli DH10BAC cells , containing baculovirus shuttle vector plasmid were used to generate recombinant bacmids according to the manufacturer's instructions ( Invitrogen ) . Bacmids were introduced into Spodoptera frugiperda 9 ( Sf9 ) insect cells by lipofection and recombinant baculovirus were recovered . NB2 VLPs were produced by infection of Spodoptera frugiperda 9 ( Sf9 ) insect cells at a MOI∼5 PFU/cell . VLPs were purified as described previously [59] by double CsCl density gradient centrifugation . The composition of the VLPs was confirmed by polyacrylamide gel electrophoresis with Coomassie blue staining and VLP integrity was monitored by negative stain electron microscopy using 1% uranyl acetate stain . VLPs were stored in CsCl at 4°C . Bovine and porcine tissues samples from the oesophagus to the rectum were obtained from healthy animals autopsied at the National Veterinary School of Nantes . Human gastroduodenal junction samples had been obtained from organ donors before the law 88–1138 of December 20 , 1988 concerning resection of human tissues after death for scientific investigations . Animal tissues were fixed in formalin and human tissues were fixed in ethanol 95% for 48 hours , and paraffin embedded . Sections ( 5 µm ) were rehydrated in graded ethanol and washed in phosphate-buffered saline ( PBS ) . Endogenous peroxidase was inhibited by using methanol/H2O2 0 . 3% for 20 minutes . Sections were then washed in PBS for 5 minutes and covered with PBS/bovine serum albumin ( BSA ) 1% for 30 minutes at room temperature in a humid atmosphere . After washing in PBS , sections were covered with either the primary antibodies ( HBGAs mAbs ) , with the peroxidase-conjugated GS1-B4 or UEA-I lectins at 10 µg/mL , or with rNB2 VLPs at 1 µg/ml , diluted in PBS/BSA 1% and left at 4°C overnight . Sections were then rinsed 3 times with PBS and incubated with either biotinylated anti-mouse immunoglobulin IgG ( Vector Labs , Burlingame , CA ) or with rabbit anti-NB2 . After washing in PBS , the sections were covered with either peroxidase-conjugated avidin ( Vector laboratories ) or with peroxidase conjugated anti-rabbit IgG ( Uptima , Montluçon , France ) for 45 minutes . Reactions were revealed with 3-amino-9-ethylcarbazol , and counterstaining was performed with Mayer's hemalun . Periodate treatment was performed immediately after the endogenous peroxidase quenching step by incubating sections with either 1 mM or 10 mM sodium periodate in 50 mM sodium acetate buffer , pH 5 . 0 , for 30 minutes at room temperature , followed by a 10 minutes incubation with 1% glycine in PBS . Control sections were treated similarly with the same buffer but without sodium periodate . Alpha-galactosidase treatment was performed on some sections by incubation at 37°C with 4 mU galactosidase in 50 mM citrate-phosphate buffer pH 4 . 6 for a total of 18 hours with a renewal after 6 hours . Control sections were incubated in parallel in the same buffer without the enzyme . Following treatments , rNB2 VLPs ( 1 µg/ml ) were added for 1 hour at room temperature and the detection of binding was performed as described above . Bovine blood samples were obtained from the National Veterinary School of Nantes and human blood samples were provided by ABO phenotyped volunteer donors at INSERM U892 ( Nantes , France ) . After collection , whole blood samples were stored at 4°C . Red Blood Cells ( RBC ) were packed in PBS pH 7 . 2 ( without Ca2+ ) by centrifugation for 5 minutes at 2500 rpm . The hemagglutination activity ( HA ) of rNB2 VLPs and GII . 4 VLPs was tested in microtitration plates with V bottomed wells ( Nunc , Roskilde , Denmark ) . Equal volumes ( 25 µl ) of VLPs ( 2 . 5 µg/ml ) serially diluted in PBS and 1% packed RBCs in PBS were mixed and the plates were incubated for 1 hour at either room temperature or 4°C . The HA titer was the reciprocal of the greatest VLPs dilution that did allow sedimentation of the RBCs compared to negative control wells that contained buffer only . The synthetic trisaccharide Galα3Galβ4GlcNAc covalently linked to Sepharose® FF ( Fast Flow 6B , Pharmacia Biotech ) was obtained from Lectinity ( Moscow , Russia ) . A chromatography column ( Biorad Richmond , CA ) was packed with 15 ml of immunoadsorbent and rinsed with 250 ml PBS . Thirty ml pooled human plasma were then passed through the column at a 1 ml/h flow rate . After extensive washing with PBS , bound antibodies were eluted with 20 ml of CH3COOH 0 . 58% in NaCl 0 . 9% , pH 2 . 8 . The eluate was immediately neutralized with 20 ml of 100 mM Tris/HCl , pH 8 . 8 and dialyzed against PBS . The reactivity and specificity of the purified antibodies was then controlled by ELISA on coated PAA neoglycoconjugates . Oligosaccharides as PAA and HSA conjugates were coated at 10 µg/ml or serially diluted onto NUNC Maxisorp immunoplates in 100 mM carbonate buffer pH 9 . 6 by overnight incubation at 37°C in a wet atmosphere . After blocking with 5% defatted dried cow's milk in PBS for 1 hour , VLPs ( 4 . 6 µg/ml ) in PBS 5% milk were added . After incubation for 2 hours at 4°C in the case of oligosaccharides-PAA or at 37°C for oligosaccharides-HSA , rabbit anti rNB2 VLPs serum at 1/1000 dilution in PBS 5% milk was added and incubated for 1 hour at 4°C . Then , peroxidase anti-rabbit IgG ( Uptima ) at a 1/2000 dilution in PBS 5% milk were added and incubated for 1 hour at 4°C . Between each step , the plates were washed 3 times with PBS 5% Tween 20 . The enzyme signals were detected using TMB ( 3 , 3′ , 5 , 5′ tetramethylbenzidine ) as substrate ( BD Bioscience , San Jose , CA ) and then read at 450 nm . Saliva samples were collected from 16 human individuals of known ABO and secretor phenotypes [37] , and from 16 cows from the National Veterinary School of Nantes , respectively . After collection , samples were boiled for 10 minutes and centrifuged for 5 minutes at 13 , 000 g . To assay rNB2 VLPs binding to saliva , microplates were coated with either human or bovine saliva diluted 1/1000 in 100 mM carbonate buffer , pH 9 . 6 and the assay was performed as described above using rNB2 VLPs at a 1 µg/ml concentration and a 1 h incubation at 37°C . To detect HBGAs in bovine saliva , after coating , a blocking step was performed with ELISA Synblock reagent ( Serotec , Kidlington , UK ) for 2 hour at 37°C . Either peroxidase-conjugated UEA-I lectin at 2 µg/ml , monoclonal antibodies at a 1/1000 dilution , or purified human natural anti-αGal antibodies at 50 µg/ml were then added to the wells . Reagents incubations were performed at 37°C for 1 h . Binding was detected either immediately following PBS washings for the UEA-I lectin , or following a 1 h incubation at 37°C with either anti-mouse or human anti-IgG peroxidase conjugates ( Uptima ) , at a 1/2000 dilution . Inhibition of binding of rNB2-VLPs by human natural anti-αGal was performed by mixing VLPs at 1 µg/ml with the purified antibodies at 50 µg/ml for 2 h at 37°C prior to incubation onto the coated bovine saliva samples . After a 1 h incubation at 37°C , binding of the rNB2-VLPs was detected as above . NMR samples contained 0 . 24 µg/µl or 22 . 5 nM VLPs in 23 mM phosphate buffer pH 7 , 154 mM sodium chloride . Assuming a number of 180 monomers and 90 binding sites per capsid , this corresponds to a 4 . 06 µM concentration of monomers , and to a 2 . 03 µM concentration of binding sites . Samples contained 0 . 5 mM carbohydrate ligand , resulting in a ∼1∶250 molar ratio of binding sites to ligand . All experiments were carried out on a Bruker Avance 500 MHz NMR spectrometer equipped with a TCI cryogenic probe . The temperature was set to 282 K . STD NMR experiments [31] were recorded with a 3-9-19 watergate sequence and an inter-scan delay of 25 s [60] . On- and off- resonance frequencies were set at −4 and 300 ppm , respectively [32] . A train of Gaussian pulses with a pulse length of 49 ms , an inter-pulse delay of 1 ms , and an attenuation level of 50 dB was applied for selective saturation of the protein . Spectra of Galα3GalαOMe were recorded with increasing saturation times from 0 . 35 to 4 s , and a total of 64 to 1k scans . The resulting STD build-up curves were subjected to non-linear fitting to a mono exponential equation: with STD being the STD signal intensity at saturation time t , STDmax being the maximal STD intensity at infinite saturation times , and ksat being the observed saturation rate constant [61] . The curve fitting was done with Origin ( Microcal ) and yielded the relative binding epitope [62] . STD spectra of the B antigen were recorded at only one saturation time of 2 s with 816 scans . The complete coding sequence of the Ggta1 gene encoding the rat α1 , 3galactosyltransferase was cloned as previously described [21] and inserted into the PCR3 . 1 eukaryotic expression vector ( InVitrogen , Paisley , UK ) . Human Embryonic Kidney ( HEK ) 293 cells , maintained in D-MEM/F-12 supplemented with 10% ( v/v ) fetal calf serum ( FCS ) 2 mM L-glutamine , free nucleotides ( 10 µg/mL ) , 100 U/ml penicillin and 100 µg/ml streptomycin ( Gibco , Paisley , UK ) were transfected with the rat Ggta1 using lipofectAMIN™ ( InVitrogen ) according to the manufacturer's instructions . Stable transfectants were obtained by selection with 0 . 5 mg/ml G418 ( Gibco ) . Cells were then cultured in the presence of 0 . 1 mg/ml G418 , passaged at confluence after dispersal with 0 . 025% trypsin in 0 . 02% EDTA and routinely checked for mycoplasma contamination by Hoechst 33258 ( Sigma ) labeling . Porcine aortic endothelial cells ( PAEC ) from a wild type pig , and from a Ggta1 knockout pig , were obtained from Dr . B . Charreau ( INSERM U643 , ITERT , Nantes , France ) . The cells were cultured in RPMI supplemented with 10% ( v/v ) FCS , 2 mM L-glutamine , 100 U/ml penicillin and 100 µg/ml streptomycin ( Gibco ) at 37°C in a 5% CO2 humid atmosphere . They were passaged at near confluence as described above . Cells at near confluence were detached by a brief 0 . 025% trypsin/0 . 02% EDTA treatment . Viable cells , 2×105 per well of 96 culture microtiter plates were incubated with either FITC-labelled GS1-B4 lectin at 10 µg/ml in PBS 0 . 1% gelatin or with rNB2 VLPs at 4 . 6 µg/ml in PBS 1% BSA for 45 min at 4°C . In the former case fluorescence analysis was performed immediately following 3 washes . In the latter case , after 3 washes , incubation was performed with the rabbit anti-NB2 serum at a 1/1000 dilution for 30 min . Following 3 washings , a 30 min incubation was then performed with FITC-labeled anti-rabbit IgG ( Sigma ) diluted 1/500 . Finally after 3 more washings , fluorescence analysis was performed on a FACSCalibur ( Becton Dickinson , Heidelberg , Germany ) using the CELLQUEST program . | Noroviruses are a major cause of gastroenteritis in humans and other mammals such as pigs and cows . Various human strains are known to bind complex sugar structures related to ABO blood groups . A single strain does not recognize all people owing to the individual variation in ABO-related antigens . Binding to these molecules is required for infection since individuals lacking the appropriate sugar structures are not infected by a given strain . We now report that a cow-specific strain binds very specifically to the so-called xenoantigen , a sugar motif resembling B blood group antigen , present at the surface of the small intestine of cows . This antigen is absent from all human tissues since the human gene encoding an enzyme required for its synthesis has been inactivated by mutations during evolution of the Hominidaea lineage . Although present in other mammals such as pigs , we observed that this sugar motif is not expressed at the right location to allow infection , that is , the surface of the small intestine . Thus , the cow virus should not infect humans or pigs . Its adaptation to cows would prevent transmission to other species living in close contact with cows such as man and pig . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"virology/virus",
"evolution",
"and",
"symbiosis",
"gastroenterology",
"and",
"hepatology/gastrointestinal",
"infections",
"virology/emerging",
"viral",
"diseases",
"virology",
"infectious",
"diseases/viral",
"infections",
"virology/mechanisms",
"of",
"resistance",
"and",
"susc... | 2009 | The αGal Epitope of the Histo-Blood Group Antigen Family Is a Ligand for Bovine Norovirus Newbury2 Expected to Prevent Cross-Species Transmission |
Transcriptional elongation requires the concerted action of several factors that allow RNA polymerase II to advance through chromatin in a highly processive manner . In order to identify novel elongation factors , we performed systematic yeast genetic screening based on the GLAM ( Gene Length-dependent Accumulation of mRNA ) assay , which is used to detect defects in the expression of long transcription units . Apart from well-known transcription elongation factors , we identified mutants in the prefoldin complex subunits , which were among those that caused the most dramatic phenotype . We found that prefoldin , so far involved in the cytoplasmic co-translational assembly of protein complexes , is also present in the nucleus and that a subset of its subunits are recruited to chromatin in a transcription-dependent manner . Prefoldin influences RNA polymerase II the elongation rate in vivo and plays an especially important role in the transcription elongation of long genes and those whose promoter regions contain a canonical TATA box . Finally , we found a specific functional link between prefoldin and histone dynamics after nucleosome remodeling , which is consistent with the extensive network of genetic interactions between this factor and the machinery regulating chromatin function . This study establishes the involvement of prefoldin in transcription elongation , and supports a role for this complex in cotranscriptional histone eviction .
RNA polymerase II transcribes nuclear protein-coding genes in eukaryotes . After the assembly of the preinitiation complex onto the promoter and the subsequent initiation of transcription , the enzymatic activity of RNA polymerase II enables the synthesis of considerably long RNA molecules in the elongation phase of transcription . During this process , nascent RNA is modified by capping and polyadenylation , exons are defined for splicing , while hnRNP components are recruited , thus ensuring the production of mature mRNAs that can be exported to the cytoplasm [1] . This complexity explains the tight regulation of transcription elongation , which involves a large number of auxiliary factors . Some of these factors contribute to transcription elongation by controlling the cotranscriptional processes affecting nascent RNA [2] or by modulating the processivity of RNA polymerase II by either favoring or preventing its stalling [3] , [4] . When paused , RNA polymerase II can arrest after backtracking [5] . In this case , factors like TFIIS stimulate the RNA cleavage activity of RNA polymerase II to enable it to resume transcription elongation [6]–[8] . Since the physiological substrate of transcription is chromatin , transcription elongation is one of the main processes that contribute to shape the nucleosomal landscape of the genome [9] and the distribution of histone modification marks [10] . Accordingly , a large set of auxiliary factors are involved in facilitating transcription elongation through chromatin [11] . Nucleosome remodeling machines , like RSC , utilize ATP-dependent helicases to destabilize nucleosomes during transcription [12] . Other chromatin factors , like FACT , contribute to the disassembly and reassembly of nucleosomes during transcription elongation in an ATP-independent manner [13] . Other examples are the chromatin remodelers Isw1 and Chd1 , which prevent histone exchange during transcription [14] . The action of these factors is often coupled to specific histone covalent modifications , which are carried out by a different set of auxiliary players; for example , the Rpd3S complex which de-acetylates histone tails [15] , or Set2 , which methylates the K36 residue of histone H3 [16] , and thereby represses histone exchange [17] . One last group of elongation factors controls the phosphorylation state of RNA polymerase II , particularly the carboxyterminal domain of its largest subunit . This CTD domain is the docking platform that allows some of the other above-described elements to act during transcription elongation . Accordingly , the factors affecting CTD phosphorylation control the overall integration of the molecular functions required to complete the transcription cycle [18] . The number of protein complexes in the literature playing specific roles during transcription elongation has increased in the last few years , suggesting that our knowledge on this field is still partial . In order to detect novel factors involved in transcriptional elongation , we performed a systematic genetic screening of a collection of Saccharomyces cerevisiae viable deletions . Using an assay based on the comparison of transcription units whose transcribed region length differs , we detected novel genes that have not been previously linked to transcription elongation . Among them , we found PFD1 encoding a subunit of the prefoldin complex . Prefoldin acts as a co-factor of group II chaperonins in eukaryotes and archaea [19] . Prefoldin forms jellyfish-shaped hexameric complexes consisting of two alpha-type and four beta-type subunits [20]–[22] . It is functionally involved in the cotranslational folding of proteins like tubulin and actin by presenting unfolded polypeptides to cytosolic chaperonin CCT [23] , [24] . All six prefoldin genes are non-essential in yeast , although their deletion is deleterious under conditions that stress microtubules and actin dynamics [25] , [26] . The mutation of prefoldin genes causes embryonic lethality in Caenorhabditis elegans and produces severe cytoskeletal defects in mice that promote their early death [27] . Nuclear functions of prefoldin are scarce in the literature and it has not been previously connected to transcription elongation . In this work we describe our finding of the transcriptional function of prefoldin by showing that it is present in the nucleus , binds chromatin in a transcription-dependent manner and contributes to chromatin dynamics during transcription elongation .
The GLAM ( gene length-dependent accumulation of mRNA ) assay is a useful tool to detect defects in mRNA biogenesis that occur in a gene length-dependent manner . By comparing two transcription units , both encoding the acid phosphatase Pho5 and whose only difference lies in the length of their 3′ UTR , we can easily detect gene mutations that specifically decrease the expression level of the long unit without affecting the short one [28] . We previously made full use of the GLAM assay to establish functional links between already known transcription complexes , like SAGA , the Mediator or the Ccr4-Not complex , and transcription elongation [29] . In order to identify novel elements that play a role in transcription elongation , we carried out a systematic genetic screening by performing GLAM assays using the collection of viable deletion mutants in Saccharomyces cerevisiae . To do this , we introduced plasmid pGAL1-PHO5 and pGAL1-PHO5::lacZ ( Figure S1A ) into all the deleted strains by crossing , sporulating and selecting appropriate markers ( Figure S1B ) . After cultivating all the transformants in microtiter plates , acid phosphatase activity was measured ( Figure S1B ) . In all , 1878 mutants produced reproducible and significant levels of activity with the two plasmids . GLAM ratios were calculated by dividing the phosphatase activity of the cells expressing the long unit by the activity of the cells expressing the short one ( Table S1 ) . Values ranged from 0 . 1 to 1 . 9 and they shaped a Gaussian distribution that peaked between 0 . 6 and 0 . 7 , which was slightly below the 0 . 8 units of the wild-type GLAM ratio ( Figure 1A ) . No specific gene-ontology category was significantly overrepresented among those mutants showing high GLAM ratios , but the mutants scoring between 0 . 1 and 0 . 3 were seen to be significantly enriched in “DNA-dependent transcription , elongation” ( p value 2 . 34 ·10−3 ) and in “gene expression” ( p value 3 . 08 · 10−3 ) ( Figure 1A ) . We repeated the GLAM assays of all these mutants after cultivating them under conditions that ensured exponential growth . Those that were confirmed in this analysis are listed in Figure 1B . The lowest score on the screen was observed with a strain deleted for SET2 , which codes for a histone H3 methytransferase , an enzyme that has been clearly connected to transcription elongation by RNA polymerase II [30] , [31] . Similarly , other mutants in this list , like rtf1Δ ( PAF complex ) , soh1Δ ( Mediator ) , ccr4Δ ( Ccr4-Not complex ) and thp2Δ ( THO complex ) , lacked general transcriptional machinery elements that are directly or indirectly involved in transcription elongation [32]–[35] . A number of genes encoding transcription activators ( Dal81 , Imp2′ ) [36] , [37] or proteins functionally linked to RNA polymerase II ( Hog1 ) [38] were also found . Although the GLAM assay is able to detect any mutant affecting gene expression in a length-dependent manner , mutants related to posttranscriptional aspects , like mRNA stability or translation , were not found . Finally , several mutants with very low scores have never been related to transcription elongation , or even to gene expression ( Figure 1B ) . This study establishes a preliminary link between these genes and mRNA biogenesis , which is to be confirmed by additional studies . We decided to focus on PFD1 , which provided the third most significant effect in the GLAM assay when deleted ( Figure 1B ) . We confirmed the GLAM phenotype of pfd1Δ by Northern blot . The levels of the full-length mRNAs transcribed from GAL1p-PHO5::lacZ were significantly reduced in pfd1Δ as compared to the short mRNA from GAL1p-PHO5 ( Figure 2A ) . The same results were obtained for a second construct , in which the 3′UTR sequence was different ( GAL1p-PHO5::LAC4 ) [28] , indicating that the effect is not sequence specific ( Figure 2A ) . PFD1 encodes one of the components of prefoldin , a heterohexameric complex composed of two alpha subunits ( Gim2 and Gim5 ) and four beta subunits ( Gim1 , Gim3 , Gim4 and Pfd1 ) [39] . We tested the GLAM phenotypes of the mutants lacking the other five subunits . We found that gim1Δ , gim3Δ and gim5Δ also gave GLAM ratios that were significantly below the wild type ( Figure 2B ) . We also analyzed some double mutants . The gim2Δ gim5Δ mutant , lacking the two alpha subunits , gave the same GLAM values as the single gim5Δ mutant ( Figure 2B ) , indicating that only one alpha subunit type is involved in the transcriptional function of prefoldin . We also combined two deletions that exhibit the GLAM phenotype: gim1Δ and pfd1Δ . The resulting double mutant did not give significantly lower GLAM values than the corresponding single mutants ( Figure 2B ) . We conclude that the role of these beta subunits is not redundant and that they cannot substitute each other in the transcriptional function of the complex . Since prefoldin is involved in the cytoplasmic folding of microtubules and actin filaments , we explored the possibility of prefoldin GLAM phenotypes resulting from cytoskeleton defects . To test this , before the assay we treated the cells with benomyl , a powerful inhibitor of microtubules dynamics . We found neither a significant effect of this drug on the GLAM ratios in the wild type ( Figure S2A ) nor a direct correlation in prefoldin mutants between their sensitivity to benomyl and their GLAM phenotype ( Figure S2B ) . Moreover , pfd1Δ , with a low GLAM ratio , showed very mild sensitivity to benomyl ( Figure S2B ) . Similarly , prefoldin mutants exhibited uneven sensitivity to latrunculin A , an inhibitor of actin polymerization ( Figure S2C ) . Whereas pfd1Δ was no more sensitive than the wild type , gim2Δ , which displayed identical GLAM ratios to the wild type , clearly showed hypersensitivity to this drug . These results support a functional contribution of some prefoldin subunits to gene expression which , at least in the case of Pfd1 , might be even more relevant than their role in cytoskeleton folding . The GLAM phenotype of prefoldin mutants suggests the involvement of prefoldin in transcription elongation . The only known connection between prefoldin and transcription elongation derives from genetics since it has been described that gim1Δ , gim3Δ and gim5Δ display a synthetic growth defective phenotype when combined with dst1Δ , the mutant lacking RNA cleavage factor TFIIS [40] . We confirmed this synthetic growth phenotype and extended it to pfd1Δ ( Figure 2C ) . Some transcription elongation mutants , such as dst1Δ , are sensitive to NTP-depleting drugs [41] . It is noteworthy that , although none of the prefoldin mutants was sensitive to high mycophenolyc acid concentrations ( not shown ) , the pfd1Δ dst1Δ double mutant exhibited significantly enhanced sensitivity to very low concentrations of this drug ( Figure 2C ) , and of 6-azauracil ( not shown ) . This synthetic genetic interaction between PFD1 and DST1 was also detected by performing the GLAM assay , as the double pfd1Δ dst1Δ mutant exhibited synergistically lower GLAM ratios ( Figure 2D ) . Altogether , these results support that the synthetic interaction between PFD1 and DST1 takes place in the transcription elongation context . Although it was first described as a cytoplasmic complex [25] , a direct role of prefoldin in transcription would involve its nuclear localization . . We constructed a Pfd1-GFP fusion ( Figure S3A ) and found that it exhibits nucleo-cytoplasmic localization ( Figure 3A ) . Does Pfd1 need active transport to enter and leave the nucleus ? To test this , we used the export thermosensitive mutant xpo1-1 mex67-5 . In this mutant , Pab1-RFP ( used here as a control of the assay ) is excluded from the nucleus at the permissive temperature and accumulates in the nucleus at the restrictive temperature [42] ( Figure 3B ) . As expected from the result in Figure 3A , Pfd1-GFP was not excluded from the nucleus at the permissive temperature , but accumulated at the restrictive temperature , indicating that prefoldin is a substrate of the active mechanisms of nucleo-cytoplamic export ( Figure 3B ) . Similar results were obtained with Gim5-GFP ( not shown ) . The other prefoldin subunits also exhibited nucleo-cytoplasmic localization ( Figure S3B ) . We conclude that the presence of prefoldin subunits in the nucleus is consistent with a role of this complex during transcription elongation . To address whether prefoldin directly regulates transcription by binding to chromatin we constructed a Pfd1-Myc fusion ( Figure S4A ) and carried out chromatin immunoprecipitation ( ChIP ) experiments using anti-Myc antibodies . The obtained results revealed that prefoldin is physically associated with the coding region of actively transcribed genes like ADH1 and PMA1 ( Figure 4A , blue bars ) . The detected level of Pfd1-Myc association with the transcribed region was clearly lower than RNA polymerase II , but was comparable or even higher than well-known elongation factors like TFIIS [43] , or chromatin factors that act during transcription elongation like SAGA [44] , Occupancy was stronger in the 3′ end of the genes tested ( Figure 4A ) . In order to check if this was a general feature of prefoldin binding , we investigated Pfd1-Myc binding genome-wide . Although the overall signal was low , we detected a certain degree of binding of the 3′ gene ends by Pfd1-Myc ( Figure 4B ) . No gene ontology categories were significantly enriched in those genes exhibiting higher Pfd1-Myc occupancy . However , average binding at 3′ was stronger for the 10% genes exhibiting the highest RNA polymerase II occupancy than for the least transcribed decile ( Figure 4B ) . Accordingly the difference between Pfd1 binding in 5′ and 3′ was more significant in highly transcribed genes ( p-value 1 . 1·10−10 after Mann-Whitney test ) than in genes exhibiting poor RNA polymerase II occupancy ( p-value 3 . 9· 10−3 ) . The results above suggest that prefoldin recruitment is influenced by transcriptional activity . To confirm this , we analyzed Pfd1 binding to inducible genes like GAL1 , under active and inactive conditions . We found that Pfd1 binding to GAL1 was transcription-dependent since the ChIP signal in the coding region was found exclusively under culture conditions where GAL1 transcription was active ( galactose-containing medium ) and not under repressive conditions ( Figure 4C ) . Similar results were found for Gim1 , Gim3 and Gim5 , but no significant binding of Gim2 and Gim4 to GAL1 was detected ( Figures S4B–C ) . We also tested the recruitment of Pfd1 to three osmotic stress-responsive genes ( STL1 , CTT1 and GRE2 ) that become transiently induced upon salt treatment [45] . As shown in Figure 4D , Pfd1 did not show any occupancy of these genes under non-stress conditions , whereas Pfd1 was recruited to their coding region only 10 min after the addition of salt and in parallel to RNA polymerase II ( Figure 4D ) . In all the genes tested , Pfd1 occupancy was greater at the 3′ end than at 5′ ( Figures 4A , C , D ) . It has been previously shown that some housekeeping genes undergo strong reduction of RNA polymerase II occupancy in response to osmotic stress [45] . We found that 10 min after salt addition , and again in parallel to RNA polymerase II occupancy , Pfd1 binding to PMA1 and ADH1 diminished ( Figure 4A ) . Altogether , these results support that prefoldin recruitment to chromatin depends on transcription . The difference in transcription-dependent binding among the prefoldin subunits , which correlate with the GLAM phenotype of the corresponding mutants ( Figure 4E ) , suggests the existence of a prefoldin complex variant formed by Pfd1 , Gim1 , Gim3 and Gim5 , which would be involved specifically in transcription elongation . Among these subunits , only Gim5 was of the alpha type . We tested the binding of Pfd1 to GAL1 in a gim5Δ background to find that the ChIP signal in galactose significantly reduced ( Figure 5A ) , indicating that the binding of Pfd1 to transcribed chromatin is stabilized within the prefoldin complex . In contrast , the absence of Pfd1 did not diminish the binding of Gim5 to GAL1 ( Figure 5B ) , indicating that this alpha-type subunit can be recruited on its own to transcribed genes . Pfd1 binding to GAL1 increased with the distance to the transcription start site and reached a maximum between positions 200 and 700 ( Figure 4C ) . Similar results were obtained with Gim5 ( Figure S4C ) . This Pfd1 binding did not match the profile of the RNA polymerase II molecules that are phosphorylated in the Ser5 residues of the CTD domain , but closely followed the profile of Ser2-phosphorylated RNA polymerase II ( Figure 5C ) . In contrast to Ser5 phosphorylation , equally associated with the initiation and elongation steps of transcription , Ser2 phosphorylation exclusively links to transcription elongation [18] . In order to investigate whether Ser2P is involved in the recruitment of prefoldin to transcribed genes , we repeated the experiment in a ctk1Δ mutant , lacking the main kinase of the CTD Ser2 residues [46] . We found that Pfd1 binding to GAL1 was impaired by the ctk1Δ mutation , which occurred in parallel to the reduction in Ser2 phosphorylation exhibited by this mutant ( Figures 5D and S5A ) . However , we were unable to detect a physical interaction between prefoldin and Ser2-phosphorylated RNA polymerase II by co-immunoprecipitation ( Figures S5B , C ) . Altogether , the results we present above support the existence of a specialized nuclear prefoldin capable of acting at transcription sites during elongation , whose recruitment would be favored by the Ser2 phosphorylation of the RNA polymerase II CTD . In order to evaluate the actual importance of prefoldin in transcription elongation across the genome , we measured the effect of PFD1 deletion on the intragenic distribution of elongating RNA polymerase II . We calculated the ratios between elongating RNA polymerase II sitting on the 5′ and the 3′ ends of 377 highly expressed genes in pfd1Δ . We did it by following two different methods: transcriptional run-on , detecting active elongation-competent polymerases; and anti-Rpb3 ChIP , detecting all polymerases , either active or inactive [47] . When all the analyzed genes were averaged , we observed no significant change in the 3′/5′ratios in the pfd1Δ mutant as compared to the wild type by both ChIP and run-on ( not shown ) . In light of the GLAM phenotype of the prefoldin mutants , we compared long genes ( >4 kbp ) with shorter genes . We found that long genes exhibited significantly higher 3′/5′ ratios of total RNA polymerase II ( Rpb3 ChIP ) in pfd1Δ , but this relative enrichment in 3′ did not involve any global variation in their average 3′/5′ run-on ratio ( Figure 6A ) . We also discovered that those genes whose promoter regions contained a canonical TATA box exhibited significantly higher 3′/5′ run-on ratios in pfd1Δ , with no parallel change noted in their 3′/5′ Rpb3 ChIP ratios ( Figure 6A ) . In contrast , TATA-like genes , containing a non-canonical TATA box [48] , did not undergo any significant change in pfd1Δ ( Figure 6A ) . We conclude that prefoldin is required especially for transcription elongation through long and TATA genes , and that the absence of Pfd1 causes alterations in both the intragenic distribution of RNA polymerase II and its tendency to become arrested ( run-on incompetent molecules ) . Since GAL1 is a canonical TATA gene , and as we detected a significant transcription-dependent binding of prefoldin to it , we investigated its transcriptional induction in pfd1Δ . We found a serious delay in GAL1 mRNA accumulation in the absence of Pfd1 . While the wild type accomplished 75% of induction in 30 min , pfd1Δ barely reached 20% ( Figures 6B and S6A ) . This delay in mRNA accumulation was due to a transcriptional defect as it correlated with the slower occupancy of the GAL1 coding region by RNA polymerase II , as measured by ChIP ( Figure 6C ) . We also measured GAL1 induction in pfd1Δ dst1Δ and found an even longer delay in this double mutant , which was unable to reach the wild-type level of maximum mRNA accumulation , even after 150 min ( Figures 6B and S6A ) . In contrast , the single dst1Δ mutant exhibited a wild-type kinetics of induction ( Figures 6B and S6A ) . The fact that the double mutant showed a more serious defect than any of the single mutants confirms the functional cooperation between prefoldin and TFIIS that we detected with the GLAM assays . As mentioned above , CTD phosphorylation is a good marker of active transcription . We analyzed the levels of Ser5 and Ser2 phosphorylation of the RNA polymerase II molecules when transcribing GAL1 in the absence of Pfd1 . We detected no effect of pfd1Δ on the levels of Ser5 phosphorylation ( Figure 6D ) , but we observed that the phosphorylation levels of Ser2 clearly lowered ( Figure 6E ) . This reduction in Ser2 phosphorylation is similar to the effect caused by the deletion of the genes encoding the Paf1 complex , whose involvement in chromatin modifications during transcription elongation is well known [49] . According to the data represented in Figure 6A , the maximal contribution of prefoldin to transcription elongation should take place in long , canonical TATA genes . In fact , the only three TATA genes that are longer than 4 kbp and present in the 5′/3′ array ( YPL082c , YCR089w and YLR342w ) exhibited increased average 3′/5′ ratios at both the total ( Rpb3 ChIP ) and active ( run-on ) RNA polymerase II levels ( not shown ) . We analyzed in detail the effect of pfd1Δ on the transcription of a long transcription unit driven by a TATA promoter . We chose GAL1p-YLR454w , an engineered version of this 8 kbp-long gene fused to the promoter region of GAL1 [50] . The mRNA levels expressed in galactose-containing medium confirmed that GAL1p-YLR454w exhibited greater dependency on prefoldin than GAL1 . The accumulation of YLR454w mRNA was significantly impaired in pfd1Δ and in the double dst1Δpfd1Δ mutant ( Figure 7A ) . We also found a shift in the distribution of total RNA polymerase II towards the 3′ end of the gene in pfd1Δ in relation to the wild type ( Figures 7B , C ) . We detected an even stronger shift of the run-on signal to the 3′ end in the absence of Pfd1 ( Figure 7D ) . These results may well reflect a combination of the transcriptional defects caused by pfd1Δ on long genes and on TATA-containing genes . When we analyzed the effect of dst1Δ on the RNA polymerase II distribution along GAL1p-YLR454w , we found the opposite to be true if compared to pfd1Δ: reduced occupancy of Rpb3-ChIP and run-on signals towards the 3′ end of the gene ( Figures 7C , D ) . The double pfd1Δ dst1Δ mutant showed flat patterns , which is in agreement with an additive and non-epistatic interaction between the two genes . All the mutants tested displayed reduction in the absolute RNA polymerase II levels ( Figures S6B , C ) . This reduction was especially severe in the density of active RNA polymerase II molecules of the double mutant ( Figure S6C ) and was consistent with a more marked defect in transcription elongation when both prefoldin and TFIIS were absent . Could TFIIS help solve those RNA polymerase arrest events caused by the absence of Pfd1 ? To test this hypothesis , we measured TFIIS recruitment to GAL1p-YLR454w in a pfd1Δ background . We observed no enrichment of TFIIS in relation to Rpb3 in pfd1Δ ( Figure S6D ) . So the absence of Pfd1 interferes with elongation but it does not seem to enhance the recruitment of TFIIS . These data are consistent with a role of prefoldin in transcription elongation that is independent of TFIIS action . Increased cryptic transcription , as described for a large set of transcription-related mutants [51] , might explain the accumulation of RNA polymerase II at the 3′end of GAL1p-YLR454w ( Figures 7C , D ) . To test this possibility , we ran a Northern blot analysis of GAL1p-YLR454w . We did not detect different YLR454w mRNA patterns when hybridizing with 5′ and 3′ probes ( Figure S6E ) . Moreover , we did not detect the characteristic 5 . 3 kb cryptic mRNA that is transcribed from YLR454w in other mutant backgrounds [52] ( Figure S6E ) . Finally , we investigated the elongation rate of RNA polymerase II along this transcription unit in pfd1Δ by shifting cells to glucose conditions . By doing so , the GAL1 promoter switched off , so we could follow the last wave of RNA polymerase II across the gene [50] . Two minutes after glucose addition , the pattern of RNA polymerase II occupancy in pfd1Δ and in the double mutant reflected a slower speed of RNA polymerase II as compared to the wild type and to the single dst1Δ mutant . This reduction in speed was particularly visible in the 3′ half of the gene , occupied at this time by the front of the last wave of transcription ( Figures 7E , F ) . This result is consistent with a positive effect of prefoldin in the elongation rate of RNA polymerase II . Curiously , when we measured the distribution of RNA polymerase II 4 min after switching the promoter off , we found a pattern that did not indicate a slower speed , but rather a slightly faster movement of the last subpopulation of RNA polymerase II molecules; that is , the tail of the last wave of transcription ( Figures 7F , S6F ) ( see Discussion ) . Several previous works have reported genetic interactions between the prefoldin subunits and the factors involved in the transcriptional dynamics of chromatin that have never been specifically addressed ( Table 1 ) . For example , negative genetic interactions , involving poorer growth of double mutants than what is expected from the growth of the corresponding single mutants , have been detected between prefoldin subunits and the following chromatin factors: Set2 [30]; H2AZ and the SWR complex involved in the H2AZ/H2A exchange [53]–[55]; the histone deacetylase-containing Set3 and Rpd3 complexes ( Collins et al , 2007 ) ; the histone acetylase-containing NuA4 complex [54] , [55]; the Mediator complex [54]; and the ATP-dependent nucleosome remodeling complexes SWI-SNF [55] , RSC [54] and ISW [54] . Positive genetic interactions , involving stronger growth of double mutants than what is expected based on the growth of the single mutants , were detected between the prefoldin subunits and Chd1 [54] , [55] , the SAGA complex [54] , the PAF1 complex [54] , [55] , the Bre1-Lge1 complex [54] and the Set1 complex [54] . Three of the complexes exhibiting positive interactions with prefoldin cooperate in a sequence of histone modifications events that involves H2B ubiquitination ( PAF1C , Bre1-Lge1 ) and H3-K4 trimethylation ( Set1C ) [56] , [57] . The other two ( SAGA and Chd1 ) copurify and link H3-K4 methylation to histone tails acetylation [58] . The genetic interactions described above suggest the functional relevance of prefoldin in histone dynamics during transcription . We evaluated this hypothesis by measuring histone occupancy on GAL1p-YLR454w . In wild-type cells , occupancy of the whole transcribed region by histones H3 and H4 diminished drastically under activating conditions ( Figure 8A ) , as previously reported [44] . pfd1Δ exhibited only a minor decrease in histone occupancy , when cells were grown in glucose-containing medium , if compared to the wild type ( Figure 8A ) . However , in galactose medium , we observed significantly higher levels of bound H3 and H4 in pfd1Δ than in the wild type ( Figure 8A ) . In the latter , histone occupancy was 3-fold ( 5′ end ) or 5-fold ( the remaining amplicons ) lower in galactose than in glucose . Unlike the wild type , these ratios in pfd1Δ ranged between 1 . 5 and 2 ( Figure 8A ) . Similar results were obtained with histone H2B ( Figure S7A ) . Milder , yet still significant effects of pfd1Δ on histone occupancy in galactose were detected in GAL1 ( Figure S7B ) . We investigated whether the excess H3 detected across GAL1p-YLR454w in pfd1Δ was trimethylated in the K36 residue ( H3K36Me3 ) because this modification is closely associated with transcription elongation [30] . We found that the absolute levels of H3K36Me3 across GAL1p-YLR454w were higher in pfd1Δ than in the wild type ( Figure S7C ) , indicating that at least a fraction of excess H3 accumulated in pfd1Δ had undergone methylation . This fact supports the notion that the higher H3 occupancy detected in pfd1Δ is not just due to lower transcriptional activity . When we normalized H3K36Me3 levels to total H3 , we found it lower in pfd1Δ than in the wild type ( Figure S7D ) , indicating that another fraction of excess H3 is not methylated in K36 . This transcribed unmethylated H3 is likely the consequence of the decreased level of CTD-Ser2 phosphorylation caused by pfd1Δ ( Figure 6E ) , which should affect the recruitment of the methylase Set2 [16] . However , the overall effect of pfd1Δ on H3K36 methylation is quantitatively limited since the global cellular levels of H3K36Me3 did not change significantly in the mutant ( Figure S7E ) . The accumulation of histones that we found across GAL1p-YLR454w in pfd1Δ under activating conditions can be explained by a defect in nucleosome remodeling . To test this hypothesis , we analyzed nucleosome remodeling by measuring the sensitivity of three different regions ( 5′ , middle and 3′ ) of GAL1p-YLR454w to micrococcal nuclease ( MNase ) . We found typical nucleosomal peaks in 5′ and 3′ under repressive conditions ( Figure 7B ) . In these two regions , DNA sensitivity to MNase was significantly greater under activating conditions ( Figure 7B ) . No peaks were detected in either glucose or galactose , in a central region of the gene that does not exhibit positioned nucleosomes ( Figure 7B ) . No significant difference was found between pfd1Δ and the wild type in any of the three regions of the gene . Similar results were obtained for the 3′ region of GAL1 ( Figure S7F ) . We conclude that prefoldin is not required for the initial nucleosome-remodeling step , which destabilize DNA-octamer interactions and facilitate nucleosome mobility upon transcription activation . The higher levels of histone occupancy across the transcribed GAL1p-YLR454w gene in pfd1Δ could be due to increased efficiency in chromatin reassembly or to impaired histone eviction . To test the first possibility , we analyzed the kinetic of histone reassembly after switching the promoter off . H3 ChIP signals were quantified in the central region ( 4 kb amplicon ) of GAL1p-YLR454w at different times after adding glucose to cells exponentially growing in galactose medium . We found no significant difference in the kinetics of H3 occupancy when comparing pfd1Δ and the wild type ( Figure 8C ) . If any , a slightly slower rate was detected in the prefoldin mutant . Since neither nucleosome remodeling nor chromatin reassembly seems to be affected by pfd1Δ , we propose a specific role for prefoldin in histone eviction during transcription elongation . Inverse correlation between histone occupancy and RNA polymerase II density has been reported [59] , [60] . Therefore , the defect in histone eviction that we observed in pfd1Δ might merely be an indirect consequence of suboptimal transcription caused by a different molecular defect . The higher absolute levels of H3K36Me3 ( Figure S7C ) and the normal nucleosome-remodeling pattern exhibited by GAL1p-YLR454w in pfd1Δ ( Figures 8B and S7F ) contradict this hypothesis . Nevertheless , we tested it by measuring H3 and H2B occupancy in dst1Δ and in the double mutant pfd1Δ dst1Δ as both showed significantly lower levels of RNA polymerase II across GAL1p-YLR454w than pfd1Δ ( Figure S6B ) . In dst1Δ , we found much less excess in histone occupancy than in pfd1Δ ( Figures 8D and S7G ) . In the double mutant , which exhibited stronger transcriptional deficiency than pfd1Δ , we observed the same excess in histone occupancy as in pfd1Δ ( Figures 8D and S7G ) . Hence , we found no kind of correlation between the effect of a mutation on the occupancy of the transcribed gene by histones and by RNA polymerase II . What we did observe was a clear correspondence between the pfd1Δ genotype and the retention of histones under activating conditions ( Figures 8D , E and S7G ) . Taken altogether , the above data are compatible with a functional contribution of prefoldin to histone eviction during transcription elongation .
We applied the systematic genetic analysis of yeast deletions to the study of transcription elongation by making full use of the GLAM assay . Genetic analysis is a powerful tool to detect the involvement of specific genes in a given biological function . It proves especially useful to identify unexpected functional links between already known cellular elements and biological processes not related to date . This is the case of the novel role of prefoldin in transcription elongation that we have uncovered . The involvement of prefoldin in the cytoplasmic folding of cytoskeletal components is well established [25] , [26] . Prefoldin-like complexes are also involved in the cytoplasmic assembly of eukaryotic RNA polymerase II [61] . Accordingly , the transcriptional defects of prefoldin mutants might result indirectly from its cytoplasmic role . This does not seem to be the case , as we detected the presence of prefoldin inside the nucleus and its binding to the coding region of expressed genes in a transcription-dependent manner . Moreover , the chromatin binding and transcriptional contribution among prefoldin subunits closely correlate , whereas at least one prefoldin mutant ( pfd1Δ ) displays clear transcriptional defects without exhibiting any significant cytoskeletal dysfunction . Likewise , we detected no impairment in transcription elongation upon treatment with the microtubule-destabilizing drug benomyl . We , therefore , consider that the transcriptional defects exhibited by prefoldin mutants reflect a relevant nuclear role of prefoldin in transcription elongation . Nuclear prefoldin might work together with nuclear actin , which contributes to transcription elongation in metazoa [62][63] . However , recruitment of actin to transcribed loci is maximal in the promoter region and its binding decreases progressively along genes [64] , whereas the prefoldin profile peaks in the 3′ end of genes in parallel to the Ser2-phosphorylation of RNA polymerase II-CTD . Moreover , pfd1Δ is not hypersensitive to latrunculin A , which is able to disrupt actin filaments . This suggests that , even if prefoldin acts on nuclear actin , there is another nuclear role of prefoldin in transcription elongation . Prefoldin subunits have been connected to other nuclear events where actin does not seem to be involved , like the inhibition of the c-Myc transactivation domain [65] and the removal of HIV integrase before viral transcription [66] . The regulated migration of prefoldin from the cytoplasm to the nucleus , where it interacts with DELLA transcription factors , has been recently described in plants [67] . Our experimental results complete this nuclear perspective of prefoldin and establish a functional connection between prefoldin and transcription elongation . A direct action of prefoldin on the elongating form of RNA polymerase II or in RNA polymerase II clearance after termination might explain the recruitment of prefoldin to gene bodies and the accumulation of hypophosphorylated RNA polymerase II at 3′ in its absence . According to this hypothesis , the effect of prefoldin on histone eviction would be indirectly mediated by RNA polymerase II . This would be supported by the described interactions of prefoldin-like complexes with human RNA polymerase II during their cytoplasmic assembly [61] . However , the prefoldin subunits involved in human RNA polymerase II assembly ( Gim2 and Gim4 ) do not exhibit transcriptional phenotypes in yeast ( Figure 2B ) . Moreover , we failed to detect any physical interaction between prefoldin and elongating RNA polymerase II by co-immunoprecipitation ( Figure S5B ) . Although further work will be necessary to elucidate the direct target of prefoldin in the transcription sites , the evidence presented in this article rather indicates prefoldin to be an effector of chromatin dynamics during transcription elongation . This role explains the large set of genetic interactions linking prefoldin and chromatin factors ( summarized in Table 1 ) . We detected defects in co-transcriptional histone eviction in pfd1Δ and we showed that they were not merely due to suboptimal transcriptional activity . In fact , we detected optimal accessibility of micrococcal nuclease to transcribed DNA in the prefoldin mutant , suggesting that the initial steps of cotranscriptional chromatin dynamics are not impaired in the absence of prefoldin ( Figure 8B , S7F ) . Likewise , we did not detect a positive effect of the pfd1Δ mutation on chromatin reassembly after the passage of RNA polymerase II ( Figure 8C ) . Taken altogether , these results are fully compatible with a role of prefoldin in histone eviction during transcription elongation . This would imply that histone eviction is not an automatic consequence of nucleosome remodeling , but an independent step requiring specific factors . In line with this , prefoldin would not be unique . The mutation of other chromatin-related factors that stimulate transcription elongation , like the histone acetyl transferases Esa1 and Gcn5 , also dampen cotranscriptional histone eviction and provoke gene length-dependent defects in transcription [44] , [68] . In this scenario , the accumulation of hypophosphorylated RNA polymerase II across the transcribed region would be the consequence of impaired chromatin dynamics , rather than the primary effect of the prefoldin absence . In turn , decreased CTD-Ser2 phosphorylation would additionally impact chromatin dynamics by reducing the proportion of H3 that undergoes K36 methylation [16] . According to our results , two mutants both causing decrease in the RNA polymerase II occupancy of gene bodies ( dst1Δ and pfd1Δ ) exhibit clearly different levels of bound histones ( Figures 8D , E and S7G ) . This fact is compatible with the existence of alternative modes of chromatin dynamics during transcription elongation , as previously suggested [69][70] . After nucleosome remodeling , histones might be either evicted by specific machinery , which includes prefoldin , or rearranged within the remodeled octamer ( Figure S8 ) . Accordingly , in vitro experiments have shown that RNA polymerase II can transcribe a partially disassembled octamer , the hexasome , without the simultaneous eviction of all histones [71]–[73] . According to our results , these two modes of chromatin dynamics during elongation can operate in any gene but their relative preponderance would be gene-specific depending on characteristics like length and promoter type . The differential effect of pfd1Δ on the front and the tail of the last wave of transcription , as we detected in the GAL1p-YLR454w system ( Figures 7E , F and S6F ) , might reflect these two alternative modes of elongation through chromatin . The existence of alternative modes of dealing with chromatin during elongation predicts negative synthetic phenotypes for those elongation factors that play their role in different pathways . TFIIS and prefoldin indeed show very severe synthetic impairment in all the transcriptional phenotypes that we explored . The preferential participation of TFIIS in the hexasome mode of transcription , where RNA polymerase II would display a greater tendency to backtrack , would explain this negative interaction ( Figure S8 ) . The set of genetic interactions between prefoldin and chromatin factors is similar to that described for the CCT complex; that is , the chaperonin that cooperates with prefoldin in its cytoplasmic function [74] . It is feasible , therefore , that prefoldin and CCT also cooperate in their chromatin task . Another protein chaperon , Hsp90 , also plays an important role in gene transcription by stabilizing negative elongation factor NELF on paused elongating RNA polymerase II [75] . Altogether , these findings situate chaperone-related factors as important players in gene regulation .
We introduced the plasmids of the GLAM system [28] into the library of S . cerevisiae viable deletion mutants using the method previously described for SGA [76] . In short , plasmids pSCh202 ( pGAL1-PHO5 ) and pSCh212 ( pGAL1-PHO5-lacZ ) were transformed into carrier strain BY5563 ( MAT alpha ) . The transformants obtained were then mated to the deletion library ( MAT a ) and sporulated for 7 days . The haploid spores carrying the plasmid ( pSCh202 or pSCh212 ) and the deletion mutation were selected in the appropriate media . The resulting strains were then inoculated in duplicates into 96-well plates containing 200 µl of SGal-Ura medium per well , and were incubated for 48 h at 30°C before assaying acid phosphatase activity as described elsewhere [28] . The GLAM ratio was calculated as the phosphatase activity in the deletion strain carrying pSCh212 divided by the activity in the corresponding strain bearing pSCh202 . The experiment was performed in two independent biological replicates with two technical replicates in each . The strains and plasmids used in this study are listed in Table S2 . All the S . cerevisiae strains utilized were in the BY4741 background , except for the xpo1-1 mex67-5 strain , which was in the W303 background . Strains were grown at 30°C in the indicated media , except for the xpo1-1 mex67-5 mutant , which was shifted from 24°C to 37°C for 4 h . To generate pGMZ2 , a PCR product containing the PFD1 ORF lacking the termination codon and an additional 1 kb upstream of the ORF was obtained by PCR using the oligonucleotides listed in Table S3 and yeast genomic DNA as a template . This product was cloned in pGEM-T easy ( Promega ) , to be then restricted and cloned into YCplac33-yGFP . pGMZ3 , pGMZ4 , pGMZ5 , pGMZ6 and pGMZ7 were generated similarly . In all cases , GFP constructs complemented the null allele to the wild-type extent . Absence of GLAM phenotype and sensitivity to benomyl were checked . To generate Pfd1-Myc5 , the Y07202 yeast strain was transformed with a PCR product obtained using the oligonucleotides listed in Table S3 and pGA2266 plasmid DNA as a template . Transformants were selected in SC-TRP and checked by PCR and Western Blot . Gim1-Myc , Gim2-Myc , Gim3-Myc , Gim4-Myc and Gim5-Myc were generated similarly . Absence of the GLAM phenotype and sensitivity to benomyl were verified . For the growth assays , yeast cultures were diluted to OD600 0 . 5 and serial dilutions ( 1∶10 ) were spotted onto plates . The Latrunculin A-sensitivity assay was performed as previously described [77] . Mycophenolyc acid , benomyl and latrunculin A were purchased from SIGMA . Cells were grown to the mid-log phase in selective SD liquid medium , and were washed and resuspended in TBS buffer . Image acquisition was done either with a Leica DMR microscope equipped with a differential contrast ( DC ) camera , or with a Zeiss LSM 710 laser scanning confocal microscope at 2 µm pinhole aperture . Digital images were processed with Adobe Photoshop CS . Six micrograms of total RNA prepared from yeast cells were subjected to electrophoresis on formaldehyde agarose gels , transferred to Hybond-N filters ( Amersham Biosciences , UK ) and UV cross-linked prior to hybridization at 65°C with a [32P]dCTP-labeled DNA probe . Membranes were exposed in Fuji BAS screens and developed with a FUJIX FLA5100 device . All the values were normalized in relation to the present amount of 18S rRNA . Laemmli-boiled crude extracts were run on a 12% SDS-polyacrylamide gel and transferred to nylon membranes ( Hybond-ECL ) . After blocking with Tris-buffered saline containing 0 . 1% Tween 20 and 5% milk , the following primary antibodies were used: mouse monoclonal anti-Myc ( Santa Cruz Biotechnology , Inc . ) , rabbit polyclonal anti-L1 [54] , rabbit polyclonal anti H3 ( Abcam ) , or rabbit polyclonal anti-H3K36Me3 ( Abcam ) . Finally , peroxidase-conjugated goat anti-mouse or goat anti-rabbit IgG ( both from Bio-Rad ) were used to detect the specific proteins . Yeast nuclear extracts were obtained as previously described in [78] . Immunoprecipitation was performed using IgG Sepharose ( Healthcare ) or protein A Sepharose ( Healthcare ) following the procedure described in [79] . ChIP experiments were performed as previously described [27] . Immunoprecipitations were performed with magnetic beads ( Dynal ) using the following antibodies: Myc ( Santa Cruz Biotechnology ) , HA ( clone 3F10 , Roche ) Rpb3 ( ab81859; Abcam ) , Ser2-P ( ab5095 , Abcam ) , Ser5-P ( ab5131 , Abcam ) , H3 ( ab1791 , Abcam ) , H2B ( ab1790 , Abcam ) , H3 tri methyl K36 ( ab9050 , Abcam ) , H4 ( ab10158 , Abcam ) . DNAs were analyzed by real-time quantitative PCR using SYBR Green Premix Ex Taq ( Takara ) and Light Cycler 480 II ( Roche ) with the primers listed in Table S3 . For Pfd1-Myc ChIP on chip experiments , chromatin immunoprecipitation was performed as above and , after crosslinking reversal , the obtained fragments ( 300 bp approximately ) of enriched DNA were amplified unspecificly and labeled following Affymetrix Chromatin Immunoprecipitation Assay Protocol P/N 702238 . Genomic DNA controls were processed in parallel . After PCR amplification with dUTP , the samples were purified using Qiagen QIAquick PCR Purification Kit ( 50 ) ( Cat . No . 28104 ) . DNA quality and quantity were checked using a NanoDrop ND-1000 Spectrophotometer . 0 . 5 µg of each were used to hybridize GeneChip S . Cerevisiae Tiling 1 . 0R custom arrays . This step was carried out in the Multigenic Analysis Service of the University of Valencia . The obtained CEL archives were normalized and the intensities of the signal were extracted using the TAS ( Tiling Analysis Software ) developed by Affymetrix . The resulting text files were read using R scripts to adjudicate probe intensities to genes . The log2 values of the median intensities of the chosen different group of genes were represented . Run-on assays were performed as previously described [47] with minor modifications . Basically , 50 ml of yeast culture were collected at OD600 0 . 5 by centrifugation . Cells were washed in 5 ml of 0 . 5% sarkosyl solution . Cells were centrifuged and the supernatant was completely removed . The transcription reaction was incubated for 5 min at 30°C . RNA was immediately extracted following the acid-phenol protocol . Slot-blotted membranes were performed as formerly described [47] . Double-strand immobilized probes were obtained by PCR using the primers listed in Table S3 . DNA arrays were produced in the Sección de Chips de DNA-S . C . S . I . E of the University of Valencia as previously described [47] . Briefly , 300-bp-length double strand DNA probes of 377 ORFs were obtained by PCR and printed onto positively charged nylon membranes using a BioGrid robot ( BioRobotics ) . Hybridization with labeled RNA ( Run-on ) or DNA ( Rpb3 ChIP ) samples was performed as described [47] . Shortly , radiolabeled RNA from the run-on was fragmented and denatured prior to hybridization by adding NaOH to the sample . Membranes were exposed in Fuji BAS screens for 5–7 days and were developed with a FUJIX FLA30000 device . Signals were quantified using the Array Vision software , version 8 . 0 ( Imaging Research Inc . ) . Data analysis and normalizations were performed according to [47] . Only those spots with signals 1 . 3 times over the background were considered . In all , 270 genes were successfully analyzed , of which 44 were longer than 4 kbp and 70 contained a canonical TATA box [80] . After quality filters , the log2 of the 3′/5′ratio was obtained . At least three replicas of each experiment were performed . For Rpb3 ChIP on chip purposes , the immunoprecipitated DNA was amplified and radiolabeled following the procedure described in [81] . Yeast spheroplasts and micrococcal nuclease digestions were performed as previously described [82] . DNAs were analyzed by real-time quantitative PCR using SYBR Green Premix Ex Taq ( Takara ) and Light Cycler 480 II ( Roche ) . The chromatin/naked-DNA ratio was normalized to the CDC10 probe described in Table S3 . | Transcription is the biological process that allows genes to be copied into RNA; the molecule that can be read by the cell in order to fabricate its structural components , proteins . Transcription is carried out by RNA polymerases , but these molecular machines need auxiliary factors to guide them through the genome and to help them during the RNA synthesis process . We searched for novel auxiliary factors using a genetic procedure and found a set of potential novel transcriptional players . Among them , we encountered a highly unexpected result: a factor , called prefoldin , so far exclusively involved in the folding of proteins during their fabrication . We confirmed that prefoldin binds transcribed genes and plays an important role during gene transcription . We also further investigated this transcriptional role and found that prefoldin is important for unpacking genes , thus facilitating the advance of the RNA polymerases along them . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | The Prefoldin Complex Regulates Chromatin Dynamics during Transcription Elongation |
Dicer is a large multi-domain protein responsible for the ultimate step of microRNA and short-interfering RNA biogenesis . In human and mouse cell lines , Dicer has been shown to be important in the nuclear clearance of dsRNA as well as the establishment of chromatin modifications . Here we set out to unambiguously define the cellular localization of Dicer in mice to understand if this is a conserved feature of mammalian Dicer in vivo . To this end , we utilized an endogenously epitope tagged Dicer knock-in mouse allele . From primary mouse cell lines and adult tissues , we determined with certainty by biochemical fractionation and confocal immunofluorescence microscopy that endogenous Dicer is exclusively cytoplasmic . We ruled out the possibility that a fraction of Dicer shuttles to and from the nucleus as well as that FGF or DNA damage signaling induce Dicer nuclear translocation . We also explored Dicer localization during the dynamic and developmental context of embryogenesis , where Dicer is ubiquitously expressed and strictly cytoplasmic in all three germ layers as well as extraembryonic tissues . Our data exclude a direct role for Dicer in the nuclear RNA processing in the mouse .
Dicer is a type III ribonuclease that catalyzes the ultimate biogenic step of several distinct classes of small regulatory non-coding RNAs across distinct phyla [1] . These small RNAs function in many aspects of genome and transcriptome regulation . MicroRNAs ( miRNAs ) are the primary Dicer products that are derived from genome encoded hairpin pre-miRNA structures that post-transcriptionally regulate gene expression [2] . In plants , they either pair with perfect or imperfect complementarity to post-transcriptionally regulate target gene expression [3 , 4] . In animals , miRNAs almost invariably pair imperfectly , culminating primarily in transcript destabilization through deadenylation [5–10] . The miRNA pathway is an essential post-transcriptional gene silencing system in both plants and animals . Short-interfering RNAs ( siRNAs ) on the other hand are processed from dsRNA , where one strand is incorporated into an effector Argonaute protein [11 , 12] . In the fission yeast S . pombe , Dicer derived siRNAs direct the establishment and maintenance of centromeric heterochromatin [13] . In plants , siRNAs form an integral part of an adaptive immune system against viral pathogens and transposable elements [14] . This pathway is also present in mammals , where it is active in oocytes regulating gene/transposon expression [15 , 16] , as well as in the soma , where Dicer processes dsRNA viral intermediates generating anti-viral siRNAs [17 , 18] . A defining feature of siRNAs is that they often pair with perfect complementarity to their targets and concomitantly induce their cleavage and destruction [11 , 19] . Dicer-catalyzed generation of siRNAs and miRNAs in mammals occurs in the cytoplasm [20] . Dicer is a large multi-domain protein . In addition to its two type III ribonuclease domains , most animal Dicer proteins contain a helicase and an RNA-binding domain [21] . Indeed , other non-canonical or non-miRNA/siRNA-generating functions for Dicer have recently revealed to be dependent and independent of its endonuclease activity [22–24] . Both human and C . elegans Dicer can function as a RNA binding protein ( RBP ) rather than a nuclease that can be recruited to hairpin-like structures in mRNAs as well as lncRNAs [23 , 24] . Interestingly , Dicer does not process these RNAs but can exert regulatory functions such as recruitment to P-bodies [23] . Mammalian Dicer was initially characterized as an exclusively cytoplasmic protein [25 , 26] , but recent reports have challenged this localization [27] . Dicer was shown to interact with nuclear pore components and engage in nucleocytoplasmic shuttling [28] . An interaction of Dicer with ribosomal DNA repeats was demonstrated , however , a specific function could not be identified [29] . Furthermore , a large fraction of human Dicer was detected in the nucleus of HEK293 cells , where it was shown to cleave dsRNA , failure of which results in cell death due to the accumulation of dsRNA and consequent activation of the interferon response [30] . In both mouse and human cell lines , an additional nuclear role for Dicer was shown in the termination of transcription [31] . R-loops in the vicinity of a terminator induce antisense transcription and the formation of dsRNA that in turn recruits Dicer [31] . Dicer-mediated processing of this terminator-associated dsRNA results in the loading of Argonaute proteins with small RNAs and the subsequent recruitment of G9a and H3K9me2 at terminators , which reinforces RNA polymerase II ( Pol II ) pausing and transcriptional termination [31] . In addition , Dicer was identified as a regulator of alternative cleavage and polyadenylation of pre-mRNA in the nucleus of HEK293 cells [32] . Doyle et al . show that the double-stranded RNA binding domain ( dsRBD ) at the C-terminus of human Dicer has the potential to function as a nuclear localization signal ( NLS ) . However , in full length Dicer , the NLS is masked and Dicer is shown to be cytoplasmic under steady-state conditions [33] . Different relative amounts of nuclear Dicer have been reported , with nuclear-cytoplasmic ratios ranging from approximately 1:4 . 3 [34] to 1:1 . 5 [35 , 36] . Given all the above studies were restricted to cell lines , we sought to explore the extent of nuclear Dicer function in vivo using the mouse as a model system . Surprisingly , we found that Dicer in the mouse is an exclusively cytoplasmic protein , indicating that nuclear RNA processing is not a conserved feature of mammalian Dicer in vivo .
In order to explore the nuclear function of Dicer in vivo , we first sought to determine the relative nuclear-cytoplasmic ratios in primary cell lines and in mouse tissues . The Dicer locus in mouse encodes two transcripts , one that is specific to oocytes and the other that is ubiquitously expressed . The oocyte transcript is driven by a MT-C retrotransposon insertion that is specific to rodents and generates a transcript that encodes a shorter protein lacking the N-terminal DExD helicase domain [37] . This truncated protein has fortuitously higher dsRNA processing activity resulting in the enhanced production of siRNAs for which oocyte development is uniquely dependent upon . To determine the localization of Dicer in vivo , we utilized a recently generated epitope-tagged allele of Dicer , where a Flag-HA-HA tag was placed after the start codon of the ubiquitously expressed transcript , resulting in N-terminally FH-tagged endogenous Dicer ( FH-Dcr ) ( Fig 1A ) [38] . In this study , we used tissues and primary cells derived from mice that were homozygous for the DcrFH allele in order to analyze all Dicer protein expressed in the cell that would not be possible in DcrFH/+ cells/mice . A major advantage of the HA-epitope is that specific antibodies are available that work for multiple applications such as western blotting , immunoprecipitation ( IP ) and most importantly immunofluorescence ( IF ) on tissue sections . Firstly , we generated DcrFH/FH and control wild type primary mouse embryonic fibroblasts ( PMEFs ) , prepared whole cell extracts and performed western blotting using an anti-HA antibody . To define the sensitivity of our Dicer western blotting , we performed a two-fold dilution series of the whole cell extract , from 100% to 0 . 78% . We could detect FH-Dcr in the lane with 1 . 56% of the loaded total whole cell extract , defining this as the detection limit of FH-Dcr for western blotting ( Fig 1B ) . Next , we prepared nuclear and cytoplasmic extracts from DcrFH/FH and wild type PMEFs . For western blot analysis , protein amounts from the same number of cells were used for all fractions . The purity of the respective preparations was confirmed using antibodies recognizing the exclusively cytoplasmic tubulin and nuclear Histone H3 proteins ( Fig 1C ) . Interestingly , endogenous FH-Dcr was only observed in the cytoplasmic fraction . Should the amount of potential nuclear Dicer be below our detection limit , it would amount for less than 1 . 56% of the whole cell extract . However , we cannot exclude the potential loss of nuclear soluble factors in the preparation of the nuclear fraction . Therefore , we applied confocal IF and confirmed the restriction of Dicer to the cytoplasm in the biochemical fractionation experiment ( Fig 1D ) . Next we sought to determine the localization of Dicer in adult mouse tissues by the same methodologies . We selected mouse testis and thymus , as single cell suspensions can be readily made , which is essential for the sub-cellular fractionation protocol . Again , both by western blotting and confocal IF , Dicer was observed to be exclusively cytoplasmic ( Fig 1E–1H ) . Importantly , quantification of confocal microscopy signals determined that the nuclear fluorescence intensity of FH-Dcr is lower than the autofluorescence of the wild type in PMEFs , testis and thymus ( S1 Fig ) . Thus , mouse Dicer in these cell types displays an exclusively cytoplasmic localization . While co-IP data demonstrate that human Dicer interacts with RNA Pol II [30] and engages in multiprotein complexes together with Ago2 , TRBP and TNRC6A in the nucleus [35] , fluorescence correlation spectroscopy suggests the existence of ectopically expressed EGFP-Dicer without any binding partner in the nucleus of human cell lines [34] . We therefore decided to analyze the Dicer interactome through IP coupled to mass spectrometry ( MS ) from PMEFs , testis and thymus whole cell extracts using anti-HA beads . This analysis revealed robust and highly significant interaction with Dicer’s interacting protein TRBP that recruits Ago2 [39] ( Fig 1I ) . Interacting peptides for Ago2 were enriched in the Dicer IP but those for TNRC6A were not . However , we did not observe Dicer interaction with any components of RNA Pol II ( S1–S4 Tables ) . Thus , the interaction of Dicer with RNA Pol II is not a conserved feature of mammalian Dicer . The observation of Dicer residing solely in the cytoplasm of primary mouse cells and ex vivo isolated cells and tissues does not exclude the possibility that a tiny fraction of Dicer shuttles to and from the nucleus . To investigate this possibility , we treated PMEFs with leptomycin B ( LMB ) that is an inhibitor of nuclear export . Cyclin B1 , a protein that is known for nuclear-cytoplasmic shuttling [40] , became enriched in the nucleus of PMEFs after 6 h of LMB treatment ( Fig 2A ) . However , Dicer remained exclusively cytoplasmic as determined by confocal IF ( Fig 2A ) . Biochemical fractionation coupled with western blotting assays also revealed an exclusive cytoplasmic localization ( Fig 2B ) . We therefore concluded that a small percentage of Dicer does not rapidly shuttle to and from the nucleus . In C . elegans , Dicer was identified to localize to the nucleus after ERK-dependent phosphorylation during the oocyte-to-embryo transition [41] . The phosphorylation sites of Dicer are conserved from worms to mammals , and nuclear phospho-Dicer was shown to be present in HEK293T cells upon stimulation with fibroblast growth factor ( FGF ) as well as in the mouse uterus [41] . To understand if FGF signaling can induce Dicer nuclear translocation , we stimulated PMEFs with FGF2 after serum starvation , but could not detect any nuclear Dicer , neither by confocal IF ( Fig 3A ) nor by western blotting after cellular fractionation ( Fig 3B ) . Next we explored the localization of Dicer in the uterus by confocal IF . Consistent with our previous findings , Dicer was again restricted to the cytoplasm in the uterus ( Fig 3C and 3D ) . Dicer and a class of Dicer-dependent small RNAs termed DNA damage response RNAs ( DDRNAs ) also function in DNA repair , however , the subcellular localization of DDRNA processing is not known [42] . Although from our observations thus far DDRNAs are likely made in the cytoplasm , one cannot exclude the possibility that DNA damage signaling results in the translocation of Dicer to the nucleus . We therefore treated MEFs with 20 Gy of irradiation and analyzed Dicer localization 30 min after . The broad accumulation of γH2AX in the nucleus of irradiated cells confirmed extensive DNA damage ( Fig 4A and 4B ) . Both western blotting of Dicer in cytoplasmic and nuclear fractions as well as confocal IF revealed that DNA damage signaling does not alter Dicer’s cytoplasmic localization ( Fig 4A and 4B ) . A caveat of this experiment is that the translocation to the nucleus may be very transitory and we may have missed it with our single 30-min post-irradiation time point . We therefore treated the irradiated cells with 20 nM LMB to block nuclear export and trap any Dicer that may have translocated into the nucleus . Blocking nuclear export did not result in the accumulation of Dicer in the nucleus upon DNA damage ( Fig 4C ) . In summary , DNA damage signaling does not recruit Dicer to the nucleus and therefore DDRNA biogenesis must be a cytoplasmic event . Our experiments would suggest that at least in PMEFs as well as in adult thymus , testis and uterus , Dicer is exclusively cytoplasmic . This is a small representation of adult mouse tissues and embryonic cells , we thus sought to extend our analysis to other tissues and incorporate the demanding and dynamic nature of embryonic development . Therefore , we performed an extensive analysis of Dicer localization across the embryonic day 13 . 5 ( E13 . 5 ) mouse embryo . We chose this stage of mid-gestation development , as all organs/tissues are specified and the entire embryo can also be analyzed by IF on single sagittal sections . Staining of FH-Dcr E13 . 5 sagittal sections revealed as expected Dicer to be a ubiquitously expressed protein , although the levels of expression varied slightly between tissues ( Fig 5A ) . We analyzed derivatives of all three germ layers . Within the endoderm derived fetal lung and liver ( Fig 5B and 5C ) , Dicer was cytoplasmic . The same applied to mesodermal vertebrae ( Fig 5D ) as well as the ectoderm-derived forebrain , root ganglion , lens , and epidermis ( Fig 5E–5H ) . Likewise , extraembryonic tissues such as the placenta featured a solely cytoplasmic localization of Dicer ( Fig 5I ) . Every tissue within the E13 . 5 embryo that was analyzed by confocal microscopy gave the same exclusively cytoplasmic Dicer staining . Although dicer has one described somatic transcript in the RefSeq databases , an alternative transcription start site ( TSS ) that would result in a truncated Dicer protein not possessing the FH-tag could exist . To exclude this unlikely eventuality , we subjected total RNA from wild type and DcrFH/FH E13 . 5 mouse embryos to global 5’ rapid amplification of cDNA ends ( 5’RACE ) followed by high throughput sequencing . Our results demonstrate that in E13 . 5 mouse embryos , the dicer transcript originates from a single TSS ( S2 Fig ) . In summary , we conclude that Dicer is an exclusively cytoplasmic protein during mid gestation mouse embryonic development . Here we show that Dicer in vivo in the mouse is an exclusively cytoplasmic protein , excluding a conserved role for Dicer in nuclear RNA processing . This raises the possibility that Dicer-mediated nuclear clearance of dsRNA and thus restriction of deleterious interferon responses may not be a general feature of mammalian RNA biology . In contrast , the reported function of Dicer in metabolizing terminator-associated R-loop-induced dsRNA and the concomitant recruitment of repressive chromatin marks was shown both in human and mouse cell lines . Should Dicer-induced chromatin changes be central in vivo for the termination of transcription , one would expect to see a fraction or even sizable portion of Dicer in the nucleus . This discrepancy may arise from the fact that cell line observations may not always reflect the mechanism in vivo in an animal model . Given that Dicer was not observed in the nucleus with any of our methods , a physiological relevant role of murine nuclear Dicer at amounts below the detection limit seems unlikely in vivo . In summary , we show that mouse Dicer in vivo both in PMEFs and adult tissues as well as the developing embryo is a cytoplasmic protein restricting its physiological function to cytoplasmic RNA metabolism , binding and regulation .
Mice were bred and maintained at the EMBL Mouse Biology Unit in Monterotondo in accordance with current Italian legislation ( Art . 9 , 27 . Jan 1992 , nu116 ) under license from the Italian health ministry . The DcrFH allele has been described previously [38] . PMEFs were derived from E13 . 5 embryos of DcrFH/+ intercrosses according to standard protocols . Cells were cultured in DMEM supplemented with 12 . 5% fetal calf serum , 2 mM L-glutamine , 1X non-essential amino acids , 100 units/ml penicillin/streptomycin and 100 μM β-mercaptoethanol ( all Gibco ) at 37°C and 7 . 5% CO2 . Anti-HA ( Covance , MMS-101P ) was used at 1:1000 for WB and 1:100 for IF , anti-HA ( Cell Signaling , 3724 ) at 1:500 for IF , anti-α-tubulin ( Sigma , T9026 ) and anti-histone H3 ( Abcam , an1791 ) at 1:1000 for WB , anti-cyclin B1 ( Santa Cruz , sc-752 ) at 1:100 for IF and anti-γH2AX ( Abcam , ab26350 and Bethyl , IHC-00059 ) at 1:1000 for WB and IF . Cytosolic extracts were prepared by incubating cells in CSK buffer ( 0 . 5% Triton X-100 , 100 mM NaCl , 3 mM MgCl2 , 300 mM sucrose , 1 mM EGTA , 1 mM Pipes pH 6 . 8 , protease inhibitors ) for 10 min on ice . Supernatants were recovered after centrifugation at 500 g for 4 min . Nuclear extracts were prepared by pelleting nuclei by centrifugation of cells with increasing speed of 20 g in 30 s intervals from 20 g to 100 g trough a cushion of nuclear isolation buffer ( 20% Ficoll-Paque , 80 mM Tris pH 7 . 4 , 8 mM MgCl2 , 8 mM CaCl2 , 1 . 6% Triton X-100 , 0 . 1% DMSO ) . The nuclei pellet was washed in PBS , resuspended in Laemmli sample buffer and boiled for 10 min at 95°C . Whole cell extracts were generated by resuspending the cell pellet in Laemmli sample buffer and boiling for 10 min at 95°C . Proteins from the same number of cells for every fraction were separated on 5% and 12% SDS-PAGE gels , respectively , and transferred overnight onto a nitrocellulose membrane ( GE Healthcare ) by wet transfer . The membrane was blocked with 3% milk/0 . 1% Tween-20 in PBS , incubated in primary antibody overnight , washed , and incubated with appropriate secondary antibody linked to horseradish peroxidase ( Sigma ) . Proteins were detected using the ECL Western Blotting Detection Reagent ( Amersham ) . Nuclear export was inhibited with LMB ( LC Laboratories ) . PMEFs were treated with 20 nM LMB ( dissolved in ethanol ) for 6 h . Control cells were incubated with same amounts of ethanol instead . PMEFs were grown to 50–60% confluency , starved in medium without FCS for 16 h and then treated with 50 ng/ml FGF2 ( Peprotech ) for 30 min . PMEFs were irradiated for 10 min with a total dose of 20 Gy using a RS 2000 biological irradiator ( Rad Source Technologies ) and then incubated at 37°C for another 30 min . Cells were seeded onto glass coverslips , fixed in 4% formaldehyde for 10 min and permeabilized in 0 . 5% NP-40 for 10 min at room temperature . Blocking was performed in PBG buffer ( 0 . 2% cold water fish gelatin , 0 . 5% BSA ) for 1 h at room temperature , followed by incubation in primary antibody in PBG overnight at 4°C . Tissues were collected and fixed in 4% formaldehyde overnight and embedded in paraffin . Sections of 7 μm thickness were subjected to antigen retrieval using steam vapor for 20 min in antigen unmasking solution ( Vector Lab ) and then permeabilized in 0 . 1% Triton X-100 for 10 min at room temperature . Sections were blocked in 10% donkey serum , 2% BSA and 0 . 1 M glycine ( all Sigma ) for 1 h at room temperature . Incubation in primary antibody was performed in blocking buffer overnight at 4°C . For both cells and tissue sections , appropriate Alexa secondary antibodies ( Invitrogen ) were used at 1:1000 . Hoechst 33342 at 5 mg/ml ( Sigma ) was used to stain nuclei . Leica TCS SP5 confocal microscope was used to acquire z-stacks of four optical sections each of 0 . 5 μm thickness . Images of control and experimental samples were equally processed with ImageJ and Photoshop . Tissues were homogenized by douncing and incubated in IP extraction buffer ( 50 mM Tris pH 8 , 150 mM KCl , 5 mM MgCl2 , 0 . 2% Triton X-100 , protease inhibitors ) for 10 min on ice . After centrifugation at 14 , 000 g for 10 min , supernatants were recovered and incubated with Pierce anti-HA magnetic beads ( Thermo Scientific , 88836 ) for 30 min at 4°C . Beads were washed four times with IP extraction buffer and two times with 50 mM Tris pH 8 , 0 . 2% Triton X-100 and proteins were eluted in 50 mM Tris pH 8 , 0 . 5% SDS for 15 min at 50°C . Biological replicas were trypsin digested as described [43] , desalted using STAGE tips [44] , resuspended in 0 . 1% TFA ( v/v ) and analyzed by nanoLC MS/MS . For the analysis , peptides were separated on an EasySpray ( Thermo Scientific ) 50 cm column coupled to Orbitrap Fusion Lumos ( Thermo Scientific ) mass spectrometer . Raw data were processed using MaxQuant version 1 . 5 . 2 . 8 . Label-free quantitation was performed using MaxQuant LFQ algorithm [45] . Peptides were searched against mouse Uniprot database with commonly observed contaminants ( trypsin , keratins ) . For visualization , LFQ intensities ( output of MaxQuant search ) were imported into Perseus version 1 . 5 . 1 . 6 and processed as described [46] . Experimental protocol and bioinformatics analysis for the global 5’ RACE were performed as described [47] . The global 5’ RACE sequencing data are submitted to the European Nucleotide Archive ( ENA ) under the accession number PRJEB13567 . The original Leica SP5 files are provided under the DOI 10 . 5281/zenodo . 49840 . | Dicer has first been described as the enzyme dedicated to the generation of small RNA fragments from long double stranded RNA . This function of Dicer of playing a central role in the microRNA and short-interfering RNA biogenesis pathways was found to be taking place in the cytoplasm of the cell . However , recent studies reported additional functions of Dicer in the nucleus of human and mouse cell lines , where the protein was proposed to be involved in processing nuclear RNAs as well as in influencing the chromatin state . Consequently , the localization of Dicer within the cell has been highly debated ever since . In this study , we show by biochemical and microscopy techniques that Dicer is only localized in the cytoplasm of embryonic and adult tissues of the mouse . We also exclude the possibility that Dicer only transiently enters the cell nucleus . Our data indicate that nuclear RNA processing is not a conserved feature of mammalian Dicer and highlight the fact that findings from in vitro experiments in cell lines do not always predict the in vivo state within a living organism . | [
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"nuclear",
"staining",
"gene",
"regulation",
"immunology",
"cytoplasmic",
"staining",
"dna",
"damage",
"developmental",
"biology",
"immunoprecipitation",
"dna",
"embryos",
"cellular",
"structures",
"and",
"organelles",
"research",
... | 2016 | Endogenous Mouse Dicer Is an Exclusively Cytoplasmic Protein |
All gammaherpesviruses encode a major glycoprotein homologous to the Epstein-Barr virus gp350 . These glycoproteins are often involved in cell binding , and some provide neutralization targets . However , the capacity of gammaherpesviruses for long-term transmission from immune hosts implies that in vivo neutralization is incomplete . In this study , we used Bovine Herpesvirus 4 ( BoHV-4 ) to determine how its gp350 homolog - gp180 - contributes to virus replication and neutralization . A lack of gp180 had no impact on the establishment and maintenance of BoHV-4 latency , but markedly sensitized virions to neutralization by immune sera . Antibody had greater access to gB , gH and gL on gp180-deficient virions , including neutralization epitopes . Gp180 appears to be highly O-glycosylated , and removing O-linked glycans from virions also sensitized them to neutralization . It therefore appeared that gp180 provides part of a glycan shield for otherwise vulnerable viral epitopes . Interestingly , this O-glycan shield could be exploited for neutralization by lectins and carbohydrate-specific antibody . The conservation of O-glycosylation sites in all gp350 homologs suggests that this is a general evasion mechanism that may also provide a therapeutic target .
Epstein-Barr virus ( EBV ) and Kaposi’s Sarcoma Associated Herpesvirus ( KSHV ) are DNA tumor viruses that provide risk factors for Burkitt's lymphoma , Hodgkin's lymphoma , nasopharyngeal carcinoma , Kaposi's Sarcoma and post-transplant lymphoproliferative disease [1]–[2] . EBV infection has also been associated with multiple sclerosis [3]–[4] . Healthy carriers consistently shed virus in saliva [5] that infects naïve individuals [6]–[7] despite being exposed to virus-specific antibody [8]–[9] . This lack of neutralization contrasts completely with non-persistent mucosal infections such as that of poliovirus [10]–[11] , and implies that gammaherpesviruses have evolved specific antibody evasion mechanisms . Neutralizing antibodies generally target epitopes involved in virion binding or membrane fusion [12] . Targeting of the gB/gH/gL [13]–[16] fusion machinery [17]–[18] seems to be limited by a paucity of good targets [19] and poor immunogenicity [20] . Therefore most studies have looked at binding . The EBV gp350 is an abundant component of the virion envelope that binds to CD21 on B cells [21]–[22] and is a target for antibodies that neutralize B cell infection [23] . However , while EBV lacking gp350 is poorly infectious for B cells [24]–[25] , it infects CD21-negative epithelial cells better than the wild-type [25] , and these may provide a primary target for virions entering naive hosts . Epithelial infection can even be enhanced by gp350-specific antibodies [26] . Therefore the relationship between EBV transmission , gp350 , and gp350-specific antibodies needs further exploration , particularly as gp350 is a candidate EBV vaccine [27]–[28] . Our understanding of EBV and KSHV is limited by their narrow species tropisms . Related animal viruses are therefore an important source of information . Two of the best established experimental models are provided by Murid herpesvirus 4 ( MuHV-4 ) [29] and Bovine herpesvirus 4 ( BoHV-4 ) [30]–[31] . Their homologs of gp350 are gp150 in MuHV-4 [32] , encoded by M7 , and gp180 in BoHV-4 [33] , encoded by Bo10 . While these proteins are diverse in sequence , they seem to be related in function , being involved in both binding to a cellular receptor and in blocking the infection of cells that do not express this receptor [25] , [32]–[33] . It has been proposed that the receptor interaction displaces each homolog to reveal other glycoproteins involved in entry . Thus , a non-essential glycoprotein [24] , [32]–[33] could hide from neutralization some critical epitopes on cell-free virions . To date , the in vivo function of gp350 homologs has only been investigated with MuHV-4 . Surprisingly , gp150-deficient viruses showed only a transient lag in lytic replication in vivo and established normal levels of latency [32] . Gp150 is the most immunogenic MuHV-4 glycoprotein and anti-gp150 antibodies play a predominant role in driving Fc receptor-dependent infection [20] . While gp150 does not have an obvious direct role in cell-binding , BoHV-4 lacking gp180 displays a binding deficit [33] . Therefore this protein may be more closely analogous to gp350 and the KSHV K8 . 1 than is gp150 . Here we investigated the consequences of gp180 deletion for BoHV-4 replication in vivo and neutralization . An important gp180 function seemed to be to block the binding to virions of antibodies that would otherwise neutralize .
We previously described a BoHV-4 strain in which the entire Bo10 ORF was replaced by an eGFP expression cassette [33] . Since expression cassettes can cause in vivo attenuation , we also generated a second Bo10 mutant virus , in which stop codons terminated Bo10 translation 7 amino acids before the end of its predicted signal sequence without any associated deletion ( Figure 1A ) . A revertant strain , called Bo10 STOP Rev , was finally constructed to validate the Bo10 STOP mutant . The predicted molecular structures of the recombinant strains were confirmed by EcoRI restriction mapping and Southern blotting ( Figure 1B ) , and further by DNA sequencing . Immunoblotting with an anti-Bo10-c15 rabbit polyserum [33] established that the Bo10 mutant virions lacked gp180 ( Figure 1C ) . To investigate the importance of Bo10 in vivo , we infected rabbits with the different viral strains as described in the Material and Methods . No rabbit showed clinical disease or noticeable pathology at necropsy 64 days post-inoculation . Host colonization was assayed by quantitative PCR of DNA from peripheral blood mononuclear cells ( PBMC ) over time ( Figure 2A and B ) and from the spleens at 64 days post-inoculation ( Figure 2C and D ) . The Bo10 mutants showed no deficit . We further performed infectious center assays on spleen cells from the WT and Bo10 STOP infected rabbits . Viral plaques were observed in all samples ( Figure 2E ) . No preformed infectious virus was detected in the equivalent freeze-thawed samples ( data not shown ) , so this was latent infection . Thus , we detected no difference in acute replication , latency establishment or reactivation of Bo10-deficient mutants compared to WT or revertant strains . While pathogenesis assays are a useful measure of viral fitness , they do not measure all viral functions . In particular , virion susceptibility to neutralization [33] might not be measured because intra-host dissemination depends mainly on cell/cell virus spread and latency-associated cell proliferation . We therefore further compared the sensitivity of BoHV-4 WT , Bo10 Del , Bo10 Rev and Bo10 STOP strains to neutralization by sera of rabbits infected with the BoHV-4 V . test strain ( Figure 3 ) . WT and Bo10 Rev virions were poorly neutralized . Bo10 Del and Bo10 STOP virions were neutralized much better . In particular , complete neutralization was now possible . Neutralization experiments with eGFP expressing viruses on different cell types confirmed this result , with gp180-deficient virions showing increased sensitivity to neutralization by anti-BoHV-4 serum compared to WT virions ( Figure S1 ) . Thus gp180 seemed to limit virion neutralization . In order to understand how gp180 might protect virion against neutralization , we compared the Bo10 genes of different BoHV-4 isolates [34] . All showed consensus splice donor and acceptor sites that are used in the BoHV-4 V . test strain to generate gp180 [33] . Nucleotide sequences comparison across the entire open reading frames revealed up to 15% inter-strain divergence ( Table S1 ) . Amino acid divergence between American-European and African strains reached 39% , mostly in the N-terminal half of the protein ectodomain ( Figure 4A ) . By comparison , gB , gH and gL differ by <2% between KSHV strains [35]; gB differs by <2% between OHV-2 strains [36]; and ORF71 differs by only 5% between BoHV-4 strains . All gp180s were extremely rich in serine and threonine residues , which accounted for 54 . 3 +/− 0 . 6% of each mature ectodomain ( Figure 4A ) . Asparagine residues accounted for a further 8 . 0 +/− 0 . 6% . Therefore , a conserved feature seemed to be extensive O- and N-linked glycosylation ( Figure 4A–C ) [37] . The Bo10 gene product of BoHV-4 V . test has 122 and 7 potential O- and N-glycosylation sites respectively ( Figure 4 ) [37] . This protein has a predicted molecular mass ( MM ) of 25 kDa but an apparent MM of 180 kDa [33] . To establish the contribution of glycans to the apparent MM , we digested virions lysates with glycanases . We removed high mannose , hybrid and complex N-glycans [38] with PNGase F . We removed O-glycans successively with sialidase A , β1-4 Galactosidase and O-glycanase . While PNGase F did not affect the apparent MM of gp180 ( Figure 5A ) , removing O-glycans reduced it to approximately 20 kDa , consistent with its predicted unglycosylated MM . Therefore , gp180 was extensively O-glycosylated and O-glycans appear to account for most of its mass . To test whether O-glycans protect BoHV-4 against neutralization , we removed N- and/or O-glycans from intact virions before testing their susceptibility to neutralization . As removing glycans itself could affect viral titers ( Figure S2 ) , we expressed the results as a percentage of the number of plaques for each treatment without neutralization . As observed for other viral species [39]–[41] , removing N-glycans increased virion susceptibility to neutralization by immune serum ( Figure 5B ) . Removing O-glycans had a similar effect ( Figure 5B ) . Therefore BoHV-4 uses both N- and O-linked glycans to limit its neutralization . Interestingly , gp180 bears most of the predicted BoHV-4 envelope O-glycans ( 122/155 ) [37] . These results suggest therefore that gp180 O-glycans provide part of a glycan shield for otherwise vulnerable viral epitopes . Our subsequent analysis focused on the identification of the neutralization epitopes hidden by gp180 . The MuHV-4 gp150 seems to form a multiprotein entry complex with gB , gH and gL [42] . We therefore focused on antibodies raised against the BoHV-4 gB , gH and gL . Monoclonal antibodies ( mAbs ) were screened for gB , gH , gL or gH/gL specificity as described in the Material and Methods . Mabs 16 , 29 and 33 recognize gL , gB and the heterodimer gH/gL , respectively ( Figure S3 ) . Mab 35 recognizes gB as previously stated [43] . As with MuHV-4 [32] , infected cell surfaces provide a means of probing antigenic differences between BoHV-4 glycoprotein mutants . We compared cells infected by WT , Bo10 Del , Bo10 Rev , WT BAC , Bo10 STOP and Bo10 STOP Rev BoHV-4 viruses . MAbs 29 and 35 ( recognizing gB ) , mAb 16 ( recognizing gL ) and mAb 33 ( recognizing the gH/gL complex ) all stained cells infected with the Bo10 Del and Bo10 STOP strains better than they stained those infected with wild-type or revertant viruses ( Figure 6 ) . This result was not due to differences in protein expression , as permeabilized cells gave similar staining with each virus ( Figure 6 ) . We then analyzed WT and Bo10 Del virions by immunogold labeling with mAb 35 raised against gB . While binding of gold particles was observed with both strains ( Figure 7A ) , there were statistically more particles on Bo10 Del virions than on WT virions ( p<0 . 001 ) . Representative particules are shown in Figures 7B and S4 . This difference did not reflect a greater gB content of Bo10 del virions , since immunoblotting on the same viral preparations with the same antibody showed equivalent signals between the mutant and the WT ( Figure 7C ) . The different stocks displayed also similar particle/PFU ratios as shown in Figure 7D . Finally , increased accessibility of some epitopes on Bo10 mutant virions was confirmed by immunofluorescence ( Figure 7E ) of virions bound to cell surfaces . The cells were scanned by confocal microscopy with settings unchanged between different viruses stained with the same antibody . Glow pseudo-color analysis established that the staining was stronger when Bo10 was deleted . The difference was particularly evident for mAbs 16 ( anti-gL ) and 33 ( anti gH/L ) ( Figure 7E ) . Together these results established that gB , gL and gH/L epitopes were more accessible on Bo10 mutant virions than on WT or revertant viruses , consistent with gp180 hiding key epitopes from neutralization . We next tested whether mAbs recognizing Bo10 mutants better ( Figure 6 and 7 ) could also neutralize them better than WT or revertant virions . While mAbs 29 , 35 ( anti-gB ) and 33 ( anti-gH/L ) did not neutralize any strain ( data not shown ) , mAb 16 ( anti-gL ) neutralized the Bo10 mutants better in different cell types ( Figure 8A , Figure S5 ) . It was not possible to achieve complete neutralization as it had been with immune sera ( Figure 8B ) . Therefore , gL is likely to be only one of several neutralization targets protected by gp180 or other protection mechanisms exist . However , it was clearly one such target , establishing that the reduction in gL accessibility by gp180 was functionally important . The results obtained above showed that removal of gp180 results in the unmasking of several viral envelope epitopes among which some neutralization targets . To test whether gp180 might also affect BoHV-4 immunogenicity , we compared the humoral immune response induced in the rabbits by the Bo10 STOP strain to that observed with the wild type parental strain ( Figure 9 ) . Over the course of infection , no difference in total anti-BoHV-4 antibody response was observable between the groups of infected rabbits ( Figure 9A ) . However , as the anti-herpesvirus antibody response is often dominated by capsid proteins , some subtle changes could be masked . We therefore investigated specific responses against gB , gH and gL . 293T cells expressing GPI-linked forms of gL , the gB extracellular domain or the gH extracellular domain , were stained with anti-BoHV-4 WT sera or with anti-BoHV-4 Bo10 STOP sera . The results obtained showed that sera of both groups of rabbits stained similarly gB and gH , whereas no detectable gL staining was observed ( Figure 9B ) although specific monoclonal antibodies confirmed cell surface expression of all proteins ( data not shown ) . Finally , we compared the neutralization potential of these sera against WT , Bo10 Del , Bo10 STOP or Bo10 Rev virions . As observed previously for anti-BoHV-4 WT serum , anti-BoHV-4 Bo10 STOP serum neutralized Bo10 mutant viruses better . However , no significative difference in neutralization potential was observable between both groups of serum ( Figure 9C ) . Our results suggest therefore that gp180 deficient virions display enhanced susceptibility to neutralizing antibodies but do not elicit markedly enhanced antibody response in infected rabbits . Thus , antigenicity does not predict immunogenicity . While O-glycans help BoHV-4 to evade neutralizing antibodies , they can potentially be targeted by carbohydrate binding agents , as proposed for other viruses . Gp180 is not essential for BoHV-4 replication , but lectins could still compromise virus entry by steric hindrance . We therefore tested the capacity of jacalin , an O-glycan-specific lectin , to inhibit BoHV-4 infection ( Figure 10A ) . Inhibition was evident for WT and Bo10 Rev virions , whereas Bo10 deleted virions were relatively resistant . Therefore O-glycan-directed neutralization was possible for BoHV-4 and appeared to target mainly gp180 . Another strategy would be to use specific antibodies , much as HIV can be neutralized by an antibody that binds to the high-mannose glycans of its gp120 “silent face” [44] . In animals apart from humans , apes and Old World monkeys , the α1-3-galactosyltransferase enzyme adds a terminal galactose onto glycoproteins and glycolipids in a specific α1-3 linkage to generate the Gal epitope [45] . We have previously shown that human sera consequently exhibit innate BoHV-4 neutralization through complement activation by anti-Gal antibodies [46] . We therefore compared the sensitivity BoHV-4 WT , Bo10 Del , Bo10 Rev and Bo10 STOP virions to anti-Gal dependent neutralization . While complement-containing horse serum supplemented with anti-Gal antibodies neutralized WT and Bo10 Rev virions in a dose-dependent manner , Bo10 Del and Bo10 STOP virions were only slightly affected ( Figure 10B ) . Thus gammaherpesvirus glycan shields are potentially accessible to neutralization by carbohydrate-specific antibodies .
Persistent viruses must evade multiple arms of the host immune response to maintain infectivity [47]–[48] . Gammaherpesviruses are archetypal persistent viruses , and their cytotoxic T cell evasion mechanisms are well-known [49]–[51] . Much less is known about how they evade neutralizing antibodies . Gammaherpesviruses all share a major glycoprotein homologous to EBV gp350 . EBV remains infectious despite the presence of anti-gp350 antibodies in serum and saliva [52]–[55] . Moreover immunization with gp350 fails to reduce either infection rates or virus shedding [27]–[28] . Therefore , we still have much to learn about the interplay between gp350 , gp350-specific antibodies and EBV host entry . For example , the inhibition of B cell infection by gp350-specific antibodies [54] , [56] could have limited relevance to host entry , or even promote it by enhancing epithelial infection [26] . Similarly , antibodies to the MuHV-4 gp150 strongly enhance infection via IgG Fc receptors [57] . Here we showed that BoHV-4 gp180 is dispensable for establishment and maintenance of latency in vivo ( Figure 2 ) , but drastically reduced the susceptibility of BoHV-4 virions to neutralization by immune serum on various cell types ( Figure 3 , Figure S1 ) . Gp180 seemed to hide at least partially several different epitopes on gB , gH and gL ( Figures 6 , 7 and S4 ) , which included neutralization targets ( Figures 8 and S5 ) . Gp180 is extensively O-glycosylated and O-glycans account for most of its mass ( Figures 4 and 5 ) . These results suggest therefore that gp180 O-glycans provide part of a glycan shield for otherwise vulnerable viral epitopes . Since extensive O-glycosylation is a common feature of gammaherpesvirus gp350 homologs , this evasion mechanism may be widely shared . Another common feature of some of these proteins is strong immunogenicity [20] , [23] , [58] . The reason is not fully understood , however if gp180 homologs shield other virion glycoproteins , their location at the viral surface could favor development of an antibody response against them . The substantial gp180 divergence between different BoHV-4 strains ( Table S1 , Figure 4A ) remains to be explained . It is possible that much of the protein does not require a very specific amino acid sequence for its function . Thus , a key feature of the gp350 homologs of different gammaherpesviruses may simply be that they are type I transmembrane proteins with extensive O-glycosylation [33] ( Figures 4 and 5A ) . An importance of glycans for immune evasion has also been hypothesized for gp350 [59] . Similarly , the HIV gp120 [60] uses glycans to provide a “silent face” protected against most antibodies [41] . Thus while neutralization is possible [61]–[62] , this and other mechanisms ensure that it is difficult . The Ebola virus glycoprotein ( EBOV GP ) - again involved in virus binding [63] and a target for vaccine design - is also extensively glycosylated and in this way partially protected against antibody [64] . As with gp180 , different filoviruses show huge glycoprotein diversity but retain the basic protein organization and extensive glycosylation [64] . This immune evasion mechanism appears therefore to be shared by several viral families . While carbohydrates on SIV and HIV envelope proteins can shield these viruses from antibody recognition and neutralization [41] , [65] , it appears that these glycans could also limit the neutralizing antibody response in the context of SIV infection [66] or HIV immunization [67] . We did not observe an increased BoHV-4 immunogenicity in the absence of gp180 ( Figure 9 ) . While most of the epitopes hidden by the gp120 glycan shield are located on gp120 itself [68] , BoHV-4 gp180 protects other virion glycoproteins in trans , rather than simply protecting itself in cis . Studies on different viruses [69]–[70] suggest that protection of a limited number of entry complexes from neutralization is probably sufficient to preserve virion infectivity . In contrast , influence on immunogenicity probably requires covering of all entry complexes in order to render them invisible to the immune system . Gp180 does not hide all the vulnerable epitopes at the viral surface . Indeed , even if gp180 hide most of the epitopes recognized by mAb 16 , 29 , 33 and 35 , some remains accessible at the surface of WT or revertant virions ( Figure 7 ) . Moreover , infected cell debris provides also certainly a source of uncovered antigens . It therefore appears that gp180 influence virion antigenicity but not immunogenicity . Similarly , BoHV-4 gB N-term protects some vulnerable epitopes , but its deletion does not result in an enhanced ability to induce neutralizing antibody responses [71] . Another unusual feature of gp180 was the likely importance of O-linked glycans for viral antibody evasion . In other viruses , most protection against antibody seems to involve N-linked glycans [41] , [72] . While BoHV-4 surface N-glycans are also involved in antibody evasion ( Figure 5B ) , this study strengthens the role of O-glycans in neutralization evasion . O-linked glycans have also been shown to protect MuHV-4 gB N-terminal ectodomain against antibody [71] . Indeed , although MuHV-4 gB N-term confers protection to some neutralization epitopes on gH/L , gB N-term is itself a neutralization target [19] , [71] . However , depending on the host cell , this part of gB can be largely protected against antibody by O-linked glycans [71] . These glycans could also possibly assist in protecting a neutralization epitope on gH/L . In BoHV-4 , gB is the only other described envelope protein that could bear O-glycans . Indeed , gp180 and gB N-term contain respectively 122 and 33 predicted O-glycosylation sites [37] . These two proteins and their glycans could therefore cooperate to render BoHV-4 particularly resistant to neutralization [73] . Similar N-terminal O-glycans occur in the Herpes Simplex virus gC [74] , but do not have a known function . Because N-linked glycans are relatively bulky , O-linked glycans may be better suited to protecting small or linear glycoprotein domains while still allowing protein/protein interactions . It seems with gp180 that protection by O-linked glycans can also be “scaled up” for more extensive protection . Because gp180 is likely to be part of a multi-protein complex , too many N-glycans might disrupt important protein/protein interactions . Another consideration is that glycans can on occasion be targeted by the immune response [75] . In this context , glycan diversity might be useful for a virus , and providing such diversity is a potential function of the BoHV-4 Bo17 gene , which encodes a mucin-type beta-1 , 6-N-acetylglucosaminyltransferase [76] . While glycans offer mainly protection in the natural setting , they can also be artificially targeted for neutralization by carbohydrate binding agents ( CBAs ) [77] . Evading CBAs would require a virus to compromise its glycan shield , thereby promoting neutralization by antibody [78] . As CBAs might be expected to elicit their own antibody response after repeated dosing , thereby attenuating their effect , anti-carbohydrate antibodies might be more useful in long-term settings . This also opens the possibility of vaccination against specific pathogen carbohydrates to target their glycan shields [79] . All together , our results suggest that BoHV-4 gp180 and , by extension , its homologs in other gammaherpesviruses shield the virus from immune recognition . This probably contributes to the ineffectiveness of the antibody response against these viruses .
The experiments , maintenance and care of rabbits complied with the guidelines of the European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes ( CETS n° 123 ) . The protocol was approved by the Committee on the Ethics of Animal Experiments of the University of Liège , Belgium ( Permit Number: 1035 ) . All efforts were made to minimize suffering . Madin-Darby bovine kidney ( MDBK ) ( ATCC CCL-22 ) , Bovine Turbinates ( BT ) ( ATCC CRL-1390 ) , Embryonic Bovine Trachea ( EBTr ) ( ATCC CCL-44 ) , Embryonic Bovine Lung ( EBL ) ( DSMZ ACC-192 ) , Bovine Macrophages ( BOMAC ) [80] , bovine mammary epithelial ( MacT ) [81] and EBL-NLS-Cre [30] cells were cultured in Dulbecco’s modified Eagle Medium ( Invitrogen ) containing 10% fetal calf serum ( FCS ) , 2% Penicillin/Streptomycin ( Invitrogen ) and 1% non Essential amino acids ( Invitrogen ) . Bovine PBMC were prepared as described elsewhere [82] and cultured in RPMI Glutamax Medium containing 10% FCS , 2% Penicillin/Streptomycin ( Invitrogen ) , 1% Essential amino acids ( Invitrogen ) , 1 mM Sodium pyruvate , 25 mM HEPES and 50 µM 2-mercaptoethanol . The BoHV-4 V . test strain initially isolated from a case of orchitis [83] , the BoHV-4 WTeGFP , Bo10 Del and Bo10 Rev strains [33] and the derived recombinant strain cloned as an Bacterial Artificial Chromosome ( BAC ) [30] , were used throughout . The coding sequence for BoHV-4 V . test gL amino acid residues 1-140 was amplified by PCR ( Hi-Fidelity PCR kit , Roche Diagnostics Ltd ) with 5' AvrII-restricted and 3' NotI-restricted primers . Similarly , the coding sequences for BoHV-4 V . test gB amino acid residues 1–725 and gH amino acid residues 1–678 were amplified by PCR with 5' XbaI-restricted and 3' NotI-restricted primers . These PCR products were cloned into the XbaI/NotI sites of pBRAD , thereby attaching a C-terminal glycosyl-phosphatidyl-inositol ( GPI ) membrane anchor [20] , generating gL-GPI , gB-GPI and gH-GPI expression plasmids . The O-glycan specific lectin , jacalin , was purchased from Vector Laboratories . For detection of gp180 on western blotting , we used a rabbit monospecific polyserum raised against the C-term end of the Bo10 encoded protein ( anti-Bo10-c15 ) [33] . For the neutralization experiments , we used sera of 4 different rabbits infected intravenously with 108 PFU of the BoHV-4 V . test strain and collected 63 days post inoculation . The mouse mAb M86 raised against the Galα1-3Gal epitope was purchased from Alexis and used free of sodium azide as previously described [46] . Horse serum was collected as a source of complement . The serum was treated as described previously to preserve complement activity , aliquoted and stored at −80°C [46] . Four mouse mAbs raised against BoHV-4 were also used in the present study [84] . Their specificities were unraveled on 293T cells transfected with the vectors encoding gB-GPI , gH-GPI or gL-GPI . The epitopes depending on the gH-gL heterodimer were reconstituted by co-expressing gH-GPI and gL-GPI . Briefly , transfected 293T cells were fixed and permeabilized in Acetone 95% for 5 min and then stained with the different antibodies in PBS containing 10% FCS ( v/v ) . These antibodies were detected with Alexa 488-coupled Goat anti-mouse IgG-specific antibodies ( Invitrogen ) . Nuclei were counterstained with DAPI ( 4 , 6-diamidino-2-phenylindole ) . Fluorescence was visualized with a Nikon TE-2000 microscope and a Leica CCD camera . We disrupted the BoHV-4 V . test Bo10 coding sequence ( genomic coordinates 65 , 696 to 66 , 595 , Genbank JN133502 ) by introducing stop codons into the coding sequence for the Bo10 signal peptide ( Bo10 STOP ) . BoHV-4 recombinants were produced using BAC cloning and prokaryotic recombination technologies as described before [30] . The V . test BAC G plasmid was used as parental plasmid [30] . The BoHV-4 V . test Bo10 STOP was produced using a two step galactokinase ( galK ) positive/negative selection in bacteria [85] . The first recombination process ( galK positive selection ) consisted to introduce the galK gene into the Bo10 coding sequence ( genomic coordinate 65 , 760 ) resulting in the V . test BAC G Bo10 galK plasmid . Recombination was achieved using the Bo10 galK cassette . It consisted of the galK gene flanked by 50-bp sequences corresponding to Bo10 regions ( 65 , 711-65 , 760 and 65 , 810-65 , 761 of the BoHV-4 V . test strain genome ) . This cassette was produced by PCR using pgalK vector [85] as template and Bo10-fwd-galK 5’agatctgtcatacattcaaattgcatgctttttatattcagcctcgcctgCCTGTTGACAATTAATCATCGGCA 3’ and Bo10-rev-galK 5’ atacggtggtggatgtgctggtgctgttgctggcagttaacccatatataTCAGCACTGTCCTGCTCCTT 3’ as forward and reverse primers , respectively ( galK sequences are indicated in capital letters , Bo10 start codon is in bold ) . The second recombination process ( galK negative selection ) consisted to replace the galK sequence by a Bo10 STOP cassette to generate the BoHV-4 V . test Bo10 STOP plasmid . The Bo10 STOP cassette consisted of a synthetic double strand DNA corresponding to genomic coordinates 65 , 696 to 65 , 831 with the introduction ( genomic coordinate 65 , 761 ) of 36 nucleotides coding for in-frame STOP codons and restriction sites ( Figure 1A ) . These 36 nucleotides do not insert STOP codons in any of the 5 other frames of the genome . The BoHV-4 V . test Bo10 STOP Rev plasmid was produced similarly from BoHV-4 V . test Bo10 STOP plasmid . The first recombination process ( galK positive selection ) was identical to the one described above . The second recombination process ( galK negative selection ) consisted to restore Bo10 to generate a revertant plasmid . This cassette was produced by PCR using BoHV-4 V . test genome as template and Bo10-rec-sens ( genomic coordinates 65 , 183 to 65 , 207 ) and Bo10-rec-rev ( genomic coordinates 67 , 278 to 67 , 257 ) as forward and reverse primers , respectively . Reconstitution of infectious virus from BAC plasmids was obtained by transfection in MDBK cells to obtain Bo10 STOP BAC and Bo10 STOP BAC Rev strains . To excise the BAC cassette , reconstituted viruses were propagated in EBL-NLS-Cre cells expressing Cre recombinase to generate the corresponding excised strain . Southern blot analysis [82] of viral DNA digested with EcoRI was performed with probe corresponding to genomic coordinates 65 , 696 to 66 , 595 of the BoHV-4 V . test genome . BoHV-4 strains grown on MDBK cells were purified as follows . Virions were harvested from infected MDBK cell supernatants by ultracentrifugation ( 100 , 000× g , 2 h ) ; infected-cell debris was then removed by low-speed centrifugation ( 1 , 000× g , 10 min ) . Virions were then centrifuged through a 20 to 50% ( w/v ) potassium tartrate gradient in PBS ( 100 , 000× g , 2 h ) . Virions were recovered from the gradient and finally washed and concentrated in PBS ( 100 , 000× g , 2 h ) . Virions were lysed and denatured by heating ( 95°C , 5 min ) in SDS-PAGE sample buffer ( 31 . 25 mM Tris-HCl pH 6 . 8 , 1% ( w/v ) SDS , 12 . 5% ( w/v ) glycerol , 0 . 005% ( w/v ) Bromophenol Blue , 2 . 5% ( v/v ) 2-mercaptoethanol ) . Proteins were resolved by electrophoresis on Mini-PROTEAN TGX ( Tris-Glycine eXtended ) precast 7 . 5% resolving gels ( Bio-Rad ) in SDS-PAGE running buffer ( 25 mM Tris-base , 192 mM glycine , 0 . 1% ( w/v ) SDS ) and transferred to polyvinylidene difluoride membranes ( Immobilon-P transfer membrane , 0 . 45 µM pore size , Millipore ) . The membranes were blocked with 3% non-fat milk in PBS/0 . 1% Tween-20 , and then incubated with anti-Bo10-c15 rabbit antibodies , mAb 35 or rabbit anti-BoHV-4 polyserum in the same buffer . Bound antibodies were detected with horseradish peroxidase-conjugated goat anti-rabbit IgG pAb or goat anti-mouse IgG pAb ( Dako Corporation ) , followed by washing in PBS/0 . 1% Tween-20 , development with ECL substrate ( GEHealthcare ) and exposure to X-ray film . Specific-pathogen-free New-Zealand white rabbits were used throughout this study . Rabbits were inoculated intravenously with purified stocks of the different viral strains . In one experiment we infected rabbits with WT , Bo10 Del or Bo10 Rev strains ( 108 PFU ) . In a second experiment , rabbits received WT or Bo10 STOP strains ( 107 PFU ) . At the end of the experiment , rabbits were euthanized and a necropsy examination was performed during which the spleen was collected . Blood samples were collected and PBMC were separated by Ficoll ( Ficoll-Paque Plus , GE Healthcare ) density gradient as described previously [86] . Immediately after euthanasia , spleen was removed and half-part of it was homogenized using a tissue grinder ( VWR ) , passed through a stainless steel sieve and washed in FCS-free MEM before further analyses . DNA was purified from the spleen and PBMC using the QIAamp DNA Mini kit ( Qiagen ) . Real-time PCR was performed as described elsewhere [87] . A 103 bp fragment corresponding to BoHV-4 ORF8 was amplified with the forward primer 8startfw ( 5’- CAAATAGTTCATTAGCTGCCTCTCC -3’ ) and the reverse primer 8middlerev ( 5’- TCATCAGTAACAGTTGGAATAGTGG -3’ ) in the presence of the fluorescent probe 5’-FAM-AACACGTCAACA AGCAAGCCATCCACTG-TAMRA-3’ . pGEM-T easy containing BoHV-4 gB ORF was used to establish standard curves . PCR amplifications and fluorescence reactions were carried out in a iCycler system ( Bio-Rad ) under the following conditions: initial activation of the Taq polymerase ( Bio-Rad ) at 94°C for 5 min followed by 50 cycles at 94°C for 1 min , 50 cycles at 51°C for 30 sec and 50 cycles at 72°C for 1 min . Viral detection in spleen cell suspension was assayed by infectious centre assay ( ICA ) as follows . 5 . 105 MDBK cells grown in 6 well cluster dishes ( Becton Dickinson ) were co-cultured for 7 days at 37°C with spleen cells in MEM containing 10% FCS , 2% PS , 0 . 6% CMC and 5 . 10−5M of β-mercaptoethanol ( Merck ) . Cells were then fixed and stained with mAb 35 for indirect immunofluorescent detection of intracellular viral antigen as described previously [33] . Fluorescence was then visualized with a TE2000-S Nikon and a Leica DC300F CCD camera system . All reagents were obtained from New England BioLabs . For SDS-PAGE analysis , samples were denatured in Glycoprotein Denaturing Buffer ( 0 . 5% SDS , 40 mM DTT ) for 10 min at 100°C and then digested for 3 h at 37°C with 250 NEB units PNGase F and/or 250 NEB units of neuraminidase , β1-4 Galactosidase , O-glycanase in G7 reaction buffer ( 50 mM sodium phosphate , pH 7 . 5 ) with 1% NP-40 . Reactions were stopped by the addition of Laemmli sample buffer and proteins were analyzed by immunoblotting as described below . For neutralization assays , N- or O-linked glycans were removed with the same enzymes , but without reduction or denaturation . Thus , intact virions were incubated with the different enzymes ( 3 h , 37°C ) in PBS/5% fetal calf serum buffered to pH 6 . For cell surface staining , cells infected by the different virus strains ( 2 PFU/cell , 36 h ) were washed in PBS and analyzed directly for green channel fluorescence [71] . For intracellular staining , cells were fixed in 1% paraformaldehyde ( 30 min at room temperature ) and then permeabilized with 0 . 1% saponin . Cells were incubated ( 1 h , 4°C ) with the different mAbs specific for BoHV-4 glycoproteins followed by Alexa 633-conjugated goat anti-mouse pAb ( Invitrogen ) . Cells were then washed and analyzed on a FACSAria cytometer ( Becton Dickinson ) . Copper grids of 400 mesh ( Agar Scientific Ltd ) were incubated for 10 min with 2% Alcian blue 8G solution ( Gurr Microscopy Materials , BHD ) to add positive charges . After washing , purified virions ( 108 PFU/ml ) were adsorbed to the grids for 10 min . Non-specific interactions were blocked by incubation of the grids for 15 min in PBS containing 0 . 1% ( w/v ) cold water fish skin gelatin ( CWFG , Aurion ) and 5% ( w/v ) goat serum ( Invitrogen ) . This solution was also used for further incubation and washes . Immunogold labeling was performed by incubation of the grids with mAb 35 as primary antibody for 60 min at room temperature . After washing with PBS and incubation in PBS 0 . 1% CWFG 5% goat serum for 5 min , the grids were incubated with Goat anti-mouse IgG-10 nm gold labeled conjugate ( diluted 1∶50 , BBInternational ) as secondary antibody for 60 min at RT . A final incubation step was performed in 2% uranyl acetate solution for 10 sec ( Agar Scientific ) . Immunogold stained virions were observed using a transmission electron microscope ( FEI , TEcnai Biotwin ) . Micrographs of virion were acquired for at least 20 individual virions per strain . Infected cells ( 20 PFU/cell , 2 h , 4°C ) were fixed in cold Acetone 95% for 5 min on ice . Immunofluorescent staining ( incubation and washes ) was performed in PBS containing 10% FCS ( v/v ) . Samples were incubated at RT for 45 min with the different mAbs raised against BoHV-4 glycoproteins . After three washes , samples were incubated at RT for 45 min with Alexa Fluor 488 or Alexa Fluor 568 goat anti-mouse IgG ( 2 µg/ml; Invitrogen ) . Images were acquired on a Leica TCS SP confocal laser scanning microscope with settings specific for Alexa Fluor 488 or Alexa Fluor 568 . Acquisition settings ( PMT voltage and offset ) were kept identical between slides stained with the same antibodies . Nunc Maxisorp ELISA plates ( Nalgene Nunc ) were coated for 18 h at 37°C with 0 . 1% Tween 20-disrupted BoHV-4 virions ( 2 . 106 PFU/well ) , blocked in PBS/0 . 1% Tween-20/3% BSA , and incubated with rabbit sera ( diluted 1/300 in PBS/0 . 1% Tween-20/3% BSA ) . Bound antibodies were detected with Alkaline Phosphatase conjugated goat anti-rabbit Ig polyclonal antibody ( Sigma ) . Washing were performed with PBS/0 . 1% Tween-20/3% BSA . p-Nitrophenylphosphate ( Sigma ) was used as substrate and absorbance was read at 405 nm using a Benchmark ELISA plate reader ( Thermo ) . Sequence data reported here have been deposited in the GenBank database under the following accession numbers: BoHV-4 V . test Long Unique region ( JN133502 ) . | Herpesvirus transmission between immune hosts implies some kind of antibody evasion . However , the underlying molecular mechanisms remain largely unknown . All gammaherpesviruses encode a major glycoprotein homologous to the Epstein-Barr virus ( EBV ) gp350 . Gp350 binds EBV to B cells and provides a neutralization target . However , despite its immunogenicity , EBV carriers remain infectious . Here we show that the gp350 homolog of the related Bovine Herpesvirus 4 ( BoHV-4 ) , gp180 , and its O-glycans , shield some otherwise vulnerable viral epitopes . Extensive O-glycosylation is common to all gammaherpesvirus gp350 homologs , suggesting that this evasion mechanism is also widespread . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"humoral",
"immunity",
"animal",
"models",
"of",
"infection",
"veterinary",
"diseases",
"viral",
"immune",
"evasion",
"immunity",
"virology",
"viral",
"classification",
"dna",
"viruses",
"veterinary",
"virology",
"biology",
"microbiology",
"veterinary",
"science"
] | 2011 | Antibody Evasion by a Gammaherpesvirus O-Glycan Shield |
It has been suggested that imprinted genes are important in the regulation of sleep . However , the fundamental question of whether genomic imprinting has a role in sleep has remained elusive up to now . In this work we show that REM and NREM sleep states are differentially modulated by the maternally expressed imprinted gene Gnas . In particular , in mice with loss of imprinting of Gnas , NREM and complex cognitive processes are enhanced while REM and REM–linked behaviors are inhibited . This is the first demonstration that a specific overexpression of an imprinted gene affects sleep states and related complex behavioral traits . Furthermore , in parallel to the Gnas overexpression , we have observed an overexpression of Ucp1 in interscapular brown adipose tissue ( BAT ) and a significant increase in thermoregulation that may account for the REM/NREM sleep phenotypes . We conclude that there must be significant evolutionary advantages in the monoallelic expression of Gnas for REM sleep and for the consolidation of REM–dependent memories . Conversely , biallelic expression of Gnas reinforces slow wave activity in NREM sleep , and this results in a reduction of uncertainty in temporal decision-making processes .
Mammalian evolution from the reptile lineage involved important changes in gene regulation and sleep . Many genetic mechanisms play important roles in various electrophysiological and behavioral traits that set the three major states of mammalian life: Wakefulness , Rapid Eye Movement ( REM ) and Non-REM ( NREM ) sleep [1] , [2] . However , here we focus on a particular gene regulation , namely genomic imprinting . Genomic imprinting is an epigenetic mechanism that results in allele-specific expression of some genes according to parental origin and , in vertebrates , is unique to mammals . Clinical observations of neurodevelopmental disorders of sleep suggest a role of genomic imprinting on various measures of sleep [3] , [4] . For example , Prader-Willi syndrome ( PWS ) and Angelman syndrome ( AS ) , both neurodevelopmental syndromes , exhibit opposing imprinting profiles and opposing sleep phenotypes . PWS is associated with maternal duplications/paternal deletions of alleles on chromosome 15q11–13 and is characterized by temperature control abnormalities and excessive sleepiness as well as REM sleep abnormalities [5]–[8] . Conversely , AS is associated with paternal duplications/maternal deletions on chromosome 15q11–13 and is characterized by severe mental retardation and reductions in sleep . The UBE3A gene , that resides in the PWS/AS imprinting region , has been associated to sleep abnormalities . Ube3a deficient mice are characterized by reduced NREM sleep , deteriorated REM sleep , and an increased frequency of waking during the dark–light transition [9] . Interestingly , the serotonin ( 5-HT ) 2A receptors , which mediate aminergic inhibition of REM-on cells in the parabrachialis lateralis region [10] , are primarily expressed from maternal alleles [11] , [12] . The above reviewed findings support the idea that epigenetic regulatory mechanisms , such as genomic imprinting , influence sleep-associated mechanisms . REM and NREM sleep underlie important metabolic , physiological and cognitive processes . REM sleep influences early postnatal developmental behaviors; it facilitates the acquisition of resources from the mother ( i . e . by means of suckling behavior ) and it promotes the release of hormones such as prolactin and oxytocin which are pivotal in the development of attachment behavior [4] . From an evolutionary and behavioral perspective , REM sleep has been associated with adult reproductive success [4] . NREM sleep is associated with more stable metabolic and autonomic responses compared to REM sleep . Furthermore , the presence of REM-like and NREM-like states has been associated with nutritive behaviors in offspring and mother , respectively [4] . Recent advances in functional studies of sleep strongly suggest that memory consolidation benefits from a slow ( <1 Hz ) highly synchronized cortical activity in NREM and from subcortical theta ( 5–9 Hz ) rhythms in REM sleep [13] . Despite these advances in functional understanding of REM and NREM sleep states , genetic and epigenetic mechanisms of sleep-dependent plasticity and memory processing are not currently well understood . Here we report , for the first time , experimental evidence for a role of an imprinted gene , Gnas , in the modulation of REM/NREM sleep physiology . Gnas encodes the stimulatory G-protein subunit Gsα , which is involved in the generation of intracellular cyclic AMP and plays a crucial role in energy expenditure and metabolism by mediating sympathetic effects on many tissues [14] . It is biallelically expressed in most tissues including adult white adipose tissue [15] but it is predominantly maternally expressed and paternally repressed in a subset of tissues such as neonatal brown adipose tissue ( BAT ) [16] , although it is not known if it shows imprinted expression in adult BAT . BAT produces heat by fatty acid oxidation and serves a pivotal thermoregulatory function within the organism [17] . Thanks to the rich presence of mitochondria and by means of the uncoupling protein-1 ( UCP1 ) , BAT is implicated in non-shivering thermogenesis [15] , [18] . Imprinted expression of Gnas is controlled by a cis-acting differentially methylated region ( DMR ) : the Exon1A-DMR [16] . Paternal transmission of a deletion of the Exon1A-DMR ( Gnastm1Jop , hereafter called Ex1a , see Figure 1a ) causes derepression of the normally repressed paternal Gnas allele in imprinted tissues resulting in biallelic Gnas expression and loss of imprinting [16] . We show here , for the first time , that loss of imprinting of Gnas results in specific abnormalities in sleep , cognition and thermoregulation in adult mice .
On paternal inheritance of the deletion , we have observed that in +/Ex1a mice the level of Gnas mRNA expression is increased , compared to littermate controls , in adult BAT ( Figure 1b ) indicating a derepression on the paternal allele due to a loss of Gnas imprinting , as previously reported in newborn mice BAT [16] . Moreover , we show here that Ucp1 mRNA in BAT is significantly higher in +/Ex1a mice compared to littermate controls ( Figure 1b ) . As expected from the observation of increased Ucp1 levels a significantly higher body temperature was found in mutant animals ( Figure 1c ) . The major increase of temperature in +/Ex1a mice compared to littermate controls occurs at the end of the subjective day and at the beginning of the subjective night , when there is a strong urge to sleep ( Figure 1c ) . In order to investigate the sleep-wake profile and the behavioral performance in these mice , we subjected adult +/Ex1a mice and +/+ littermate controls to behavioral and electrophysiological investigation in the home-cage environment . Interestingly , the total REM sleep was significantly reduced in +/Ex1a mice compared to littermate controls ( Figure 2a ) while the total NREM was unaffected between the two groups ( Figure 2b ) . However , at the electrophysiological level , the contribution ( power density ) of delta ( 1–4 Hz ) frequencies , the main synchronized rhythm in NREM sleep , was higher in the +/Ex1a mice compared to +/+ mice ( Figure 3 ) . A 6-hour sleep deprivation protocol triggered , immediately after deprivation , a significantly increase ( rebound ) of REM in +/Ex1a mice , which testifies the need for a homeostatic recovery of REM , but not for NREM sleep , in mutants ( Figure S1 and Figure S2 ) . No differences in body temperature occurred during sleep deprivation and during the following recovery period between the two groups ( Figure S2 ) . Thus , we extended our investigation into specific REM/NREM-dependent behavioral functions . We subjected mice to a classical memory task , the fear conditioning ( FC ) test , which affects REM sleep homeostasis [19] . After the first ( conditioning ) day of the FC protocol , we have observed an increase of REM sleep but not of NREM sleep in +/+ mice ( Figure S1a–S1b ) . The presence of REM and its theta density were significantly higher in +/+ compared to +/Ex1a mice ( Figure S1b ) . We suggest that this lack of REM increase in mutants is the causal mechanism explaining the reduced freezing behavior in +/Ex1a mice , compared to controls , that occurred when the animals were exposed to the same context ( Figure 4 ) the following day . Indeed , the consolidation of fear responses has been previously associated with REM mechanisms that occur during sleep [19] . Because no difference between the two groups was observed in the cue condition ( day 3 ) we reason that the deficit in in +/Ex1a mice is restricted to context-dependent mechanism , which has been previously associated with REM/fear memory consolidation [20] , [21] . A different subset of mice was tested in their home-cage with a cortico-striatal cognitive task , the “Switch-test” [22] . The test requires the animal to decide to nose-poke in different hoppers within the home-cage in response to a short- versus a long-light signal to obtain a reward ( Figure 5a ) . Optimal performance in this task implies that the animal has learnt to distinguish between a short-signal associated with the reward coming at the hopper in one of the two locations and a long-signal associated with the reward coming at the hopper in the other location . This test of cognitive performance assesses whether the animal has an accurate representation of both endogenous/implicit ( the subjective estimation of the signal duration ) and exogenous/explicit ( the ratio between short and long signals ) temporal variables . In one condition ( the “Switch” condition ) all trials resulted in a reward if the animal responded correctly . In a second condition ( the “Probes” condition ) a percentage of trials was never rewarded regardless of the response of the animal . This latter condition was to measure the animal's uncertainty and its cognitive performance during a temporal decision making process [23] . Interestingly , +/Ex1a mice performed better in each phase of the experiment showing higher accuracy and time precision compared to controls ( Figure 5b–5d ) . As expected , this 24-hour home-cage cognitive effort triggered a significant NREM increase and a higher delta power in +/Ex1a mice compared to +/+ mice in both “Switch” and “Probes” conditions ( Figure S1a–S1c ) . Notably , the “Probes” condition , which added a degree of uncertainty to the expectation of obtaining a reward , resulted in an more severe augmentation of NREM sleep respect to the “Switch” condition . REM did not change significantly following both conditions neither in +/+ or +/Ex1a mice ( Figure S1a–S1c ) .
We have shown here clear evidence for an involvement of the imprinted Gnas transcript in the modulation of sleep and sleep-dependent behaviors . Our study demonstrated a close link between sleep and thermoregulation . Temperature control is a well-known mechanism that modulates the expression of REM/NREM sleep in humans and mice [24] . When the thermoregulatory demand increases , REM sleep diminishes in rodents [24]–[26] . Hence , the increase of body temperature in +/Ex1a accounts for the significant reduction of REM sleep that we have observed in these mutants . Moreover , several studies have shown that an increase in body temperature is associated with an increase in NREM sleep propensity [24] , [27]–[29] . Thus , the hyperthermic phenotype is responsible for the REM defect-NREM improvement that we observed in +/Ex1a mice , indicating a link between imprinted Gnas , thermoregulation and REM/NREM sleep [24] . The molecular pathway joining Gnas and thermoregulation in BAT is well understood . Gsα is considered an important mediator of the activity of the sympathetic nervous system ( SNS ) on many functions of BAT , included thermogenesis and energy expenditure [30] . SNS stimulation of BAT leads to thermogenesis . Gsα is a constituent of transmembrane G-protein-coupled receptors that translate adrenergic SNS stimulation , in BAT , to the activation of UCP1 via specific intracellular changes and a signaling cascade that involves the production of cAMP and activation of protein kinase A ( PKA ) . Our results showing increased Gnas - Ucp1 expression are in line with a specific SNS-molecularly mediated cAMP-PKA role in non-shivering thermogenesis and therefore sleep . Indeed , the intracellular activity of cAMP-PKA is inversely related to the urge to sleep . In particular , an increased PKA affects sleep according to the specific cells in which it is expressed [31]–[33] . A reduced REM sleep would , presumably , be associated with an elevated cAMP-PKA activity . We have shown that functional monoallelic expression of Gnas , due to imprinting of this gene , is important , in mice , for the physiology of REM sleep and for the consolidation of fear conditioning contextual memories . Many experimental observations support the idea that hippocampus is crucial for the development of contextual fear conditioning [20] , [34] . However , fear conditioning responses are mediated by a complex interaction within limbic and prelimbic areas and this is also modulated by circadian genes [35] . As Chen and colleagues [36] , have shown that Gnas is not imprinted in the hippocampus , our study implies that this specific cognitive phenotype must be due to imprinting of Gnas in brain regions other than hippocampus . Within this same study it was shown that Gnas is imprinted in the paraventricular nucleus ( PVN ) of the hypothalamus , a brain region that subserves important metabolic functions [36] . The PVN receives important projections from neurons located within the suprachiasmatic nucleus ( SCN ) of the hypothalamus , the master clock for circadian rhythms [37] . The interaction between PVN and SCN is important in orchestrating circadian rhythms and to set proper neuroendocrine responses to stressors [37] . Thus loss of Gnas imprinting , within these structures , can be envisaged as playing a role in both sleep and behavior . The results of our study confirm the idea that REM sleep is fundamental in the consolidation of fear responses [19] , likely , involving a complex network of brain activity . We have also shown that NREM functions are sensitive to Gnas dosage but , in this case , loss of imprinting of Gnas results in a reduction of uncertainty in temporal decision making . This improvement is paralleled by an increased contribution of slow wave activity in NREM sleep . Our result is consistent with the idea that cortical slow oscillations , by modulating synaptic circuits between subcortical and cortical structures , are responsible for the consolidation of daily memories during sleep [13] . In conclusion , in our study , loss of imprinting of Gnas , inhibits REM and primitive REM-linked functions , such as the fear response to a threatening context . Conversely , it enhances NREM physiology and high-level cognitive functions that developed alongside a progressively complex brain . However , from a behavioral point of view an increased precision in interval timing estimation does not necessarily result in a better performance in other behavioral responses . Indeed , if the subject is less certain about the time of the foot-shock , its fear conditioning response may start earlier and stop later [38] , hence resulting in a higher freezing time . Perhaps , evolution has developed a balanced mechanism between temporal uncertainty and fear behavioral responses and then loss of imprinting , in +/Ex1a mice , involves a reduction of freezing because of a higher timing precision . Thus , REM/NREM sleep expression may favor this well-adjusted mechanism . The results of our study indicate a specific role for the imprinted gene Gnas in thermoregulation , which in turn affects REM/NREM sleep and then , cognitive performance . In addition we also reported a novel effect of specific cognitive mechanisms on sleep , by showing that a specific decision making process , the consolidation of an interval timing task , influences NREM sleep . Furthermore , in our mouse mutant model , the particular NREM physiology , exacerbates the effect of cognition on sleep homeostasis . This study attests the relevance of Gnas in brain functions and that loss of imprinting of Gnas affects cognitive processes . Another transcript within the Gnas locus , the paternally expressed transcript Gnasxl , is expressed in specific sleep-related brain areas including the locus ceruleus and cholinergic laterodorsal tegmental nuclei [39] . The activity of neurons in the cholinergic laterodorsal tegmental nucleus is particularly important in the regulation of REM sleep [40] , [41] . Mice with mutations in paternally derived Gnasxl transcripts show phenotypic deficits associated with growth and development , which is a critical stage for REM sleep across many species [4] . Thus Gnasxl may also play a role in sleep .
Initial Ex1a stock mice were produced in MRC-Harwell on 129/SvEv background and then transferred to IIT . In IIT mice were bred and maintained , through paternal inheritance , for several generations on C57BL/6J background , as this is a favorite background for behavioral studies . The genotyping of the mice was conducted following the assay as indicated in [42]: Exon1aF 5′cagtcgcgtcggcaccgcggag3′ and Exon1aR 5′gacgcactcacacgcaaagcag3′ . All the behavioral and electrophysiological experiments , in adult +/Ex1a mice and +/+ littermate controls , were conducted in the home-cage environment . In addition , mRNA expression profiles for Gnas and Uncoupling Protein ( Ucp ) 1 were made in naïve adult mice . Each experiment included 8 male mice ( 10-weeks old ) for each genotype and all procedures were done under the guidance issued by the UK authority ( Project Licence Numbers 30/2526 ) and under the Italian Policy ( licence issued on 19/06/2009 , decreto N°106/2009-B ) . Total RNA was extracted from about 0 . 2 g of snap frozen brown adipose tissue ( BAT ) and homogenized using Pestel with Trizol Reagent ( Invitrogen , Carlsbad , CA ) to isolate total RNA according to the manufacturer's procedure . Q-PCR was conducted essentially as previously described [43] , [44] . Specific primers were: UCP1:f5′-GTCCCCTGCCATTTACTGTCAG-3′ , r5′-TTTATTCGTGGTCTCCCAGCATAG-3′; GNAS: f5′-AGAAGGACAAGCAGGTCTACCG-3′ , r5′-GTTAAACCCATTAACATGCAGGA-3′; β-actin , f5′-GGCACCACACCTTCTACAATG-3′ , r5′-GGGGTGTTGAAGGTCTCAAAC-3′ . Thermal cycling parameters were: denaturation 95°C for 5 min followed by 40 cycles of denaturing- annealing and extending ( 95°C for 15 sec , 60°C for 30 sec and then 70°C for 1 min ) . The results were calculated by the comparative Ct method according to the Applied Biosystems ABI-PRISM-7700 User Bulletin#2 . Each sample was run in quadruplicate to obtain average Ct values and a ΔCt value for the target gene of the same sample , normalizing each sample to β-actin . The expression relative to β-actin was determined by calculating 2−ΔCt . Mean comparison was performed with unpaired Student's t-test . Mice were subjected to a long-term investigation in home-cage environment after the implant of a wireless system ( Data Sciences ) that enables to record electroencephalography ( EEG ) , electromyography ( EMG ) , locomotor activity and body temperature for off-line sleep-stage analysis ( see [19] ) . Automated sleep scoring followed by visual manual inspection was performed using all sleep criteria for mice [19] . We performed Fast Fourier Transform analysis of the EEG signals with SleepSign software . The contributions of EEG frequencies was expressed as power densities in each frequency bin in all NREM and REM sleep epochs ( as described in [45] ) . A 2 week post-surgery period of recovery were given to each mouse to ensure a full recovery of normal sleep . At the end of the recovery period , we started recording all the physiological signals uninterruptedly for 48 consecutive hours ( sleep baseline ) . Then , mice went under 6-hour sleep deprivation ( SD ) and 6-hour recovery period . After an additional 1-week the mice underwent fear conditioning ( FC ) or timing learning ( Switch task ) conditions . | REM and NREM sleep are two distinct stages of the sleeping brain that are involved in the modulation of metabolic , physiological , and cognitive processes . Clinical evidence suggests that epigenetic mechanisms , such as genomic imprinting , play a role in sleep . Here we show that REM and NREM brain states are differentially modulated by the maternally imprinted gene Gnas . In particular , a mutation resulting in loss of imprinted expression of Gnas enhances NREM–dependent physiologic and cognitive functions while repressing REM and REM-linked functions . This is the first experimental demonstration of a specific effect of genomic imprinting on sleep states and their associated effects on cognition . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"cognitive",
"neuroscience",
"behavioral",
"neuroscience",
"cognition",
"genomic",
"imprinting",
"genetics",
"epigenetics",
"biology",
"neuroscience",
"learning",
"and",
"memory",
"genetics",
"and",
"genomics",
"animal",
"cognition"
] | 2012 | Loss of Gnas Imprinting Differentially Affects REM/NREM Sleep and Cognition in Mice |
Intracellular colonization and persistent infection by the kinetoplastid protozoan parasite , Trypanosoma cruzi , underlie the pathogenesis of human Chagas disease . To obtain global insights into the T . cruzi infective process , transcriptome dynamics were simultaneously captured in the parasite and host cells in an infection time course of human fibroblasts . Extensive remodeling of the T . cruzi transcriptome was observed during the early establishment of intracellular infection , coincident with a major developmental transition in the parasite . Contrasting this early response , few additional changes in steady state mRNA levels were detected once mature T . cruzi amastigotes were formed . Our findings suggest that transcriptome remodeling is required to establish a modified template to guide developmental transitions in the parasite , whereas homeostatic functions are regulated independently of transcriptomic changes , similar to that reported in related trypanosomatids . Despite complex mechanisms for regulation of phenotypic expression in T . cruzi , transcriptomic signatures derived from distinct developmental stages mirror known or projected characteristics of T . cruzi biology . Focusing on energy metabolism , we were able to validate predictions forecast in the mRNA expression profiles . We demonstrate measurable differences in the bioenergetic properties of the different mammalian-infective stages of T . cruzi and present additional findings that underscore the importance of mitochondrial electron transport in T . cruzi amastigote growth and survival . Consequences of T . cruzi colonization for the host include dynamic expression of immune response genes and cell cycle regulators with upregulation of host cholesterol and lipid synthesis pathways , which may serve to fuel intracellular T . cruzi growth . Thus , in addition to the biological inferences gained from gene ontology and functional enrichment analysis of differentially expressed genes in parasite and host , our comprehensive , high resolution transcriptomic dataset provides a substantially more detailed interpretation of T . cruzi infection biology and offers a basis for future drug and vaccine discovery efforts .
The kinetoplastid protozoan parasite Trypanosoma cruzi is the etiologic agent of human Chagas disease . This parasite has a complex life cycle that involves hematophagous triatomine insects as vectors for transmission and a broad range of mammalian hosts including extensive domestic animal and sylvatic reservoirs [1] . Epimastigote forms of the parasite proliferate in the midgut of the insect vector and give rise to non-dividing , mammalian-infective metacyclic trypomastigotes that are shed in the feces of blood-feeding triatomine bugs and initiate infection in the vertebrate host . T . cruzi trypomastigotes actively penetrate a wide range of nucleated cell types , become enveloped in an acidified lysosome-like compartment [2] where they receive signals to differentiate into amastigotes [3] . Differentiating parasites gradually escape the lysosomal vacuole [4] and proliferate as amastigotes in the host cell cytosol for 3–5 days ( Fig 1A ) before differentiating back into trypomastigotes ( referred to as tissue or tissue culture trypomastigotes to distinguish these from metacyclic trypomastigotes ) , which are released into the extracellular space/medium upon host cell lysis . Motile trypomastigotes disseminate infection via the lymphatics and bloodstream to distal sites where they undergo further cycles of intracellular multiplication , egress and reinvasion . Thus , at several key points in its life cycle , T . cruzi undergoes developmental reprogramming to adapt to different hosts and variable niches within hosts , however the mechanisms governing these adaptive processes are not well defined . Cellular differentiation is controlled at multiple levels including , for most eukaryotic cells , initiation of gene transcription ( eg . [5 , 6] ) . In trypanosomatids discriminatory mechanisms for the initiation of transcription at individual loci is largely absent . Most protein-coding genes lack promoters and are transcribed as long polycistronic units that are processed into individual mRNAs [7–10] . Consequently , trypanosomes rely on post-transcriptional processes such as mRNA stability , translational efficiency and post-translational modification to coordinate developmental transitions and other adaptive responses encountered throughout their complex life cycles [11–15] . Despite the recent emphasis on mRNA translation efficiency as a primary regulator of protein abundance in trypanosomatids [13 , 16–18] and across eukaryotes more generally [19] , there is strong evidence for the existence of post-transcriptionally generated mRNA regulons in Trypanosoma brucei and Leishmania that coordinate major developmental shifts in these organisms [20–23] . As with other eukaryotes , mRNA stability and translational efficiency are influenced by trans-acting factors ( RNA-binding proteins: RBPs ) that interact with cis-acting regulatory elements in the untranslated regions of trypanosomatid mRNAs ( recently reviewed in [15 , 24] ) . Because trans-acting factors regulate multiple mRNAs in a combinatorial fashion [25 , 26] , it has been challenging to identify cis-acting and trans-acting elements that are associated with the expression of functionally-regulated trypanosomatid genes [27] . However , a growing number of examples link candidate RBP expression levels with the modulation of mRNA subsets ( eg . [15] ) . Indeed , an entire cellular differentiation program was shown to be triggered by the over expression of a single RBP in African trypanosomes [28] . Functional cis-acting elements have been identified in a number of T . cruzi transcripts and associated with the regulation of expression in this organism [29 , 30] including sets of developmentally-regulated [30–33] and functionally-related [34] genes . Although suggestive of the existence of mRNA regulons in T . cruzi , high-resolution transcriptomic data are needed to relate dynamic changes in parasite gene expression to functional adaptation on a global scale . Here , we exploit deep sequencing and informatics approaches to construct high-resolution transcriptome maps for three main T . cruzi lifecycle stages and include the simultaneous capture of parasite and host transcriptional responses during an intracellular infection of human fibroblasts by T . cruzi . With this approach , we gain deeper insights into the biology of T . cruzi with an emphasis on intracellular infection and conclude that transcriptome remodeling is required to alter the ‘blueprint’ upon which major developmental transitions are based .
To capture the global transcriptomic response associated with the establishment and maintenance of intracellular T . cruzi infection , RNA was isolated from low passage primary human foreskin fibroblasts ( HFF ) infected with tissue culture-derived T . cruzi Y strain trypomastigotes , and from mock-infected cells , at 4 , 6 , 12 , 24 , 48 and 72 hours post-infection ( hpi ) ( Fig 1A ) . RNA was also generated from extracellular trypomastigotes and from axenically cultured log-phase T . cruzi epimastigotes for comparative purposes . Two to four independent biological replicates were sequenced for each condition generating 2 . 7 billion high quality reads from 35 samples ( S1 Table ) that were subsequently processed through our RNA-Seq and data analysis pipeline ( S1 Fig ) . Sequence reads generated from T . cruzi-infected cell samples were resolved by mapping pre-processed reads against T . cruzi [35] and human hg19 reference genomes using the Tophat aligner program [36] ( S2 and S3 Tables ) . The well-documented differences in transcriptional regulation between trypanosomes and humans [7 , 8 , 10] were reflected in the distributions of the log2-transformed and size-factor-normalized gene counts for both species ( Fig 1B and S2 Fig ) . As expected , the fraction of total reads mapping to the T . cruzi genome from the mixed host-parasite read pool increased over time as intracellular amastigote replication ensued ( Fig 1A ) . It is worth noting that due to the stringency imposed during mapping ( ≤ 2 mismatches allowed/read ) and the necessity to map T . cruzi Y strain sequences against a heterologous ( CL Brener Esmeraldo ) genome [35] , the depth of coverage of the T . cruzi transcriptome at each stage ( S1 Table ) is most certainly underestimated . Despite this limitation , the demonstrated ability to resolve parasite and human sequences from a mixed read pool , and to obtain a high level of coverage of the T . cruzi transcriptome , bodes well for future transcriptomic analyses of the T . cruzi infection process in vitro and in vivo , particularly as whole genome sequence information for additional T . cruzi strains become available ( eg . [37 , 38] ) . The overall reproducibility and experimental variation between similarly generated independent samples was evaluated with Pearson correlation ( S3 Fig ) and median pairwise correlation analyses for T . cruzi ( S4 Fig ) and human ( S5 Fig ) samples . For T . cruzi , biological replicates corresponding to each of the parasite developmental stages were highly correlated ( S3A and S4 Fig ) , with the intracellular stages ( from 4–72 hpi ) exhibiting greater similarity to each other than to either of the extracellular stages ( trypomastigotes and epimastigotes ) ( S3A Fig ) . The human transcriptome samples also displayed a high level of correlation between biological replicates ( S3B and S5 Fig ) . One exception ( “4hr2” , HPGL0111 ) , identified as an outlier in a more systematic median pairwise correlation analysis ( S6 Fig ) , was removed from downstream analysis ( S3B and S5 Figs ) . To investigate general trends in the data while identifying and quantifying batch effects , principal component analysis ( PCA ) was carried out ( Fig 1C ) as well as hierarchical clustering of all parasite ( S7A Fig ) and human ( S7B Fig ) samples . PCA plots reveal a high degree of similarity between biological replicates for both the T . cruzi and human samples ( Fig 1C ) . For T . cruzi , the extracellular parasite stages ( trypomastigotes and epimastigotes ) were well separated from each other and displayed very tight clustering within each group ( Fig 1C; T . cruzi ) . The intracellular stages grouped according to their maturation status , with nascent amastigotes ( 4 and 6 hr ) clustering together and well separated from the mature replicative amastigote stages ( 24 , 48 , 72 hr ) with the 12 hr amastigotes in between ( Fig 1C; T . cruzi ) . A similar trend is observed for the human data ( Fig 1C; Human ) . Parasite-infected HFF samples are well separated from uninfected samples ( PC2 ) and the early infection time points ( 4–12 hpi ) clustered together ( Fig 1C; Human ) . The later infection time points were more loosely clustered with outliers observed at 48 and 72 hpi ( Fig 1C; Human ) . Notably , the transcriptome of uninfected fibroblasts changed considerably with time in culture ( Fig 1C; Human PC1 ) underscoring the necessity to include mock-infected controls for each infection time course for direct comparison , as we have done here . Consistent with the PCA results , the unsupervised hierarchical clustering of T . cruzi samples ( S7 Fig ) labeled both by biological group and experimental batch date segregated trypomastigote , epimastigote and intracellular amastigote samples into distinct clusters . The partitioning of immature ( 4 , 6 , 12 hpi ) and mature ( 24 , 48 and 72 hpi ) intracellular developmental stages of T . cruzi is also evident ( S7A Fig ) . A similar partitioning of infected human cell samples into early ( 4 , 6 and 12 hpi ) , mid ( 24 and 48 hpi ) and late ( 72 hpi ) time points suggest distinct phases of the host cell response to parasite infection ( S7B Fig ) . To extract biologically meaningful inferences from our expression data , lists of differentially expressed genes ( DEGs ) were constructed from pairwise comparisons of parasite or human expression data ( S4 Table ) . As a large number of genes survived this initial cut-off ( q-value < 0 . 05 ) a second filter ( log2fold-change ≥1 . 0 ) was imposed to identify gene expression changes with the highest potential impact ( S4 Table ) for T . cruzi developmental stages ( S5 Table ) and human fibroblasts at different stages of T . cruzi infection with matched controls ( S7 Table ) . Because of frequent gene duplication and the presence of several expanded gene families in the T . cruzi genome , the parasite DEG data ( S5 Table ) was filtered to remove all but a single representative of each paraloguous group ( S6 Table ) . DEG information for T . cruzi ( S6 Table ) and human fibroblasts ( S7 Table ) were used for downstream Gene Ontology ( GO ) enrichment analysis [39 , 40] and K-means clustering [41] ( S1 Fig ) . Highlights from this collective analysis are presented and discussed in the context of the relevant biology of T . cruzi-host cell interactions in the following sections . Steady-state transcriptome data was generated for three distinct T . cruzi life stages with a more comprehensive analysis of intracellular amastigote development in human fibroblasts ( Fig 1A ) . Focusing on the most disparate samples first , i . e . those derived from distinct developmental stages of T . cruzi: epimastigotes , trypomastigotes and intracellular amastigotes ( 24 hpi ) , we observed ≥2 fold differences in steady state transcript abundance for ~2000 to ~3500 parasite genes ( S4 and S5 Tables ) or between ~1500–2600 after correcting for paralogues ( S4 Table ) . Thus , as a conservative estimate , we find stage-regulated changes in transcript abundance to occur for ~15–30% of the predicted protein-coding genes in the T . cruzi genome contrasting with a previous estimate of >50% derived from comparative microarray hybridization analysis of T . cruzi life stages [42] . Due to the nature of polycistronic transcription in T . cruzi [7–9] few parasite transcripts are expected to exhibit strict stage-specificity ie . detectable in one life stage and undetectable in others ( Fig 1B ) . Despite this , we find a subset of T . cruzi genes ( between ~135 and 350 ) to be over-represented at the transcript level in a single life cycle stage as compared to the other two developmental stages compared in this study ( S8 Table ) . Adding weight to this approach , we observed the expected stage-selective expression of δ- and β-amastins in amastigotes and epimastigotes respectively ( S8 Table ) . With the high proportion of hypothetical genes ( >50% in each gene list; S8 Table ) we were not able to identify significantly enriched GO terms associated with parasite stage-enriched genes . Nonetheless , some general observations offer insight into lifestyle differences among these different parasite stages . For example , trypomastigotes uniquely express a number of protein kinases and an intermediate filament-binding trichohyalin-like protein [43] , which may reflect specialized capacities linked to host cell recognition , signaling and invasion by this parasite life stage [44–46] . Epimastigotes express twice as many genes in a stage-selective manner as the other parasite life stages ( S8 Table ) , many of which encode metabolic enzymes ( eg . pentose phosphate pathway , amino acid metabolism ) or proteins involved in nutrient acquisition ( eg . transporters of hexose sugars , nucleosides , folate/pteridine and amino acids ) . In contrast , amastigotes display elevated expression of cation transporters including ( TcCLB . 509197 . 39 ) which is upregulated early in amastigote development ( by 6 hpi ) with mRNA abundance reaching very high levels by 12 hpi ( log2fold-change 4 . 2; ~20-fold increase ) to become the most highly expressed amastigote gene at 72 hpi ( ~35-fold increase ) ( S5 Table ) as confirmed by qRT-PCR ( S9 Fig ) . Another cation transporter gene ( TcCLB . 507527 . 50 ) that displays preferential expression in amastigotes at 24 hpi ( S5 and S8 Tables ) bears homology to the ferrous iron transporter characterized in Leishmania amazonensis [47] , the expression of which is intimately coupled to amastigote development in this parasite . Thus , selective cation transporter expression , and potentially the need for iron uptake , may be a common feature of amastigote development and intracellular maintenance in both Leishmania and T . cruzi . By virtue of their critical role in regulating gene expression in trypanosomatids [15] , RNA-binding proteins ( RBPs ) represent another class of genes for which parasite stage-selective expression information is of interest . Only the RBPs that exhibit differential expression during the intracellular infection process are highlighted ( S10A Fig ) . T . cruzi trypomastigotes selectively express two RBPs , RBP5 ( TcCLB . 511127 . 10 ) and RBP6 ( TcCLB . 506693 . 30 ) ( S8 Table ) . While RBP5 has yet to be characterized , RBP6 has emerged as a critical regulator of metacyclogenesis in T . brucei [28] . Therefore , stage-specific expression of the orthologuous gene in T . cruzi trypomastigotes suggests that TcRBP6 may exert important stage-specific functions in the non-dividing , cell-invasive forms of this parasite as well . While several RBPs exhibit transient and relatively low level increases in expression in developing amastigotes , a notable exception to this trend is TcCLB . 504005 . 6 ( S5 Table ) , for which mRNA abundance tracks with increasing intracellular amastigote numbers ( S10 Fig ) . It is tempting to speculate that this particular RBP may participate in regulating processes critical to late stage amastigote growth or the next phase of the infection cycle that involves amastigote to trypomastigote conversion starting at ~96–120 hpi . Genome-scale dynamic mRNA expression data for RNA-binding proteins provides new and valuable information regarding the life cycle stage at which these important trans-acting mRNA regulatory factors are likely to act . To gain global insights into the intracellular T . cruzi infection process , a focused transcriptomic analysis was performed to capture dynamic changes in T . cruzi mRNA abundance as the parasite established intracellular residence in mammalian host cells ( Fig 1A ) . Extensive remodeling of the T . cruzi transcriptome was observed within the first 4 hours of trypomastigote invasion of human fibroblasts ( 2790 DEGs , S4 and S5 Tables ) , corresponding to the dramatic shift in environment and initiation of the amastigote differentiation program . More modest changes in T . cruzi transcript abundance occurred as amastigote development and maturation progressed over the next 20 hours of the infection cycle ( 644 DEGs in the 4–24 hpi interval ) ( S4 and S5 Tables ) . Then , once the amastigotes entered into replicative phase of the intracellular infection cycle ( ~22 hpi ) , few additional changes in the steady state transcriptome were detected ( 43 DEGs emerge in the 24–72 hpi interval; S4 and S5 Tables ) . While it is conceivable that a failure to detect additional DEGs at this stage is due to the masking of subtle transcriptome dynamics in asynchronously replicating amastigote populations , an alternative interpretation of this observation is that widespread transcriptome remodeling is only required to launch the T . cruzi amastigote differentiation program . Upon completion of the developmental switch , other important aspects of amastigote biology , such as nutrient acquisition and cell cycle regulation , are likely controlled by mechanisms other than mRNA stability . While we lack definitive data for this prediction , our observations align well with documented global gene expression patterns in related trypanosomatids , T . brucei [12 , 48] and Leishmania donovani [11 , 13 , 23] , where mRNA stability is cited as playing a more prominent role in early parasite development and both translation efficiency and post-translational modification acting as the main regulatory processes that control homeostatic functions [11 , 13 , 16–18 , 23 , 48] . If transcriptome remodeling is required to generate a new blueprint to guide developmental transitions in T . cruzi , it follows that associated changes in morphology and functionality should be reflected in corresponding transcriptomic signatures . The capture of dynamic changes in parasite mRNA abundance over a time course of infection in human fibroblasts provides an opportunity to derive biological inferences based on differential expression patterns and to compare these with known aspects of T . cruzi amastigote biology . Focusing first on the early phase of intracellular infection by T . cruzi where the greatest number of DEGs was detected , we observe several expected features of the trypomastigote to amastigote transition in the transcriptome changes . These include: ( 1 ) rapid downregulation of transcripts encoding major polymorphic surface protein classes ( trans-sialidases , mucins , MASPs and gp63 ) ( S4 and S5 Tables Trypo vs Ama4 ) , some of which have been implicated in host recognition , signaling and immune evasion [45 , 49–51]; ( 2 ) reduced expression of genes involved in flagellar assembly and motility ( within 4 hpi ) ( S5 and S9 Tables ) coincident with dramatic shortening of the single T . cruzi flagellum; and ( 3 ) increased abundance of transcripts encoding the amastigote-specific surface protein , δ-amastin ( S5 Table ) [31 , 52] . Consistent with the plasma membrane remodeling that is expected during trypomastigote to amastigote differentiation , we also observe significantly increased transcript levels for GPI-inositol deacylase ( TcCLB . 510289 . 40 ) [53] , membrane-bound/secreted phospholipase A1 [54] ( TcCLB . 509011 . 90; S5 Table ) and a surface-localized phosphatidylinositoI-phospholipase C ( PI-PLC ) ( TcCLB . 504149 . 160 ) [55] ( S5 Table ) . Although central to parasite plasma membrane remodeling during differentiation [56] , amastigote surface lipases are also well positioned to facilitate breakdown of the parasitophorous vacuole alongside the activity of a secreted hemolysin [57] , analogous to the mechanism of vacuole egress by the intracellular bacterial pathogen , Listeria monocytogenes [58] . Moreover , once T . cruzi amastigotes become cytosolically-localized in mammalian cells there are ample opportunities for parasite surface and secreted/released products to interact with host molecules and to modulate host functions during the course of infection . In this regard , T . cruzi amastigote phospholipase A1 can be considered a parasite-derived effector protein given that its expression is associated with perturbations in host cell phospholipid metabolism [59] and the activation of host protein kinase C [60] . In addition to remodeling at the plasma membrane during the early trypomastigote to amastigote transition , indicators of signaling pathway retooling also emerge in the transcriptome data ( S5 Table; Trypo vs Ama4 ) . For example , a number of predicted protein kinases and phosphatases are differentially expressed in the parasite shortly after trypomastigote entry into mammalian host cells ( S5 Table; Trypo vs Ama4 ) including the previously characterized farnesylated protein tyrosine phosphatases ( TcCLB . 506743 . 130; TcCLB . 506743 . 110 ) [61] . Consistent with the recognized role of cyclic AMP in T . cruzi differentiation processes [62] , we also observe developmental regulation of central components of the cAMP-dependent signaling pathway , such as receptor-type adenylate cyclases ( eg . TcCLB . 511043 . 60; TcCLB . 428999 . 20; TcCLB . 507467 . 10 ) , cAMP-dependent protein kinase A ( TcCLB . 509805 . 10; TcCLB . 506227 . 150 ) and cAMP-dependent phosphodiesterases ( eg . TcCLB . 508277 . 100; TcCLB . 506625 . 80 ) ( S5 Table; Trypo vs Ama4 ) . In the broader context of sensory detection and signal transduction , it is worth noting that ‘ciliary and flagellar motility’ emerges as an enriched GO term associated with mature T . cruzi amastigotes ( ≥48 hpi ) when compared to immature amastigote stages ( S9 Table; eg . Ama4 vs Ama48 ) . This signature is driven by the increase in mRNA abundance for several flagellum-associated protein coding genes in mature amastigotes after their initial decline during the initial stages of amastigote development ( S10B Fig ) . The expression of the calcium-sensing FCaBP family members [63] in intracellular T . cruzi amastigotes suggests that the minimal amastigote flagellum may engage in sensory functions in the intracellular life stages , as suggested for Leishmania amastigotes [64] . As little is known regarding the mechanism ( s ) by which T . cruzi parasites detect and integrate sensory information , particularly for the intracellular mammalian-infective stages , the availability of high-resolution mRNA expression data for T . cruzi intracellular stages ( S5 and S8 Tables ) opens the door to discovery of additional parasite molecules that function in a sensory or signaling capacity including the many predicted protein kinases and phosphatases encoded in the T . cruzi genome [65] that have yet to be characterized . Consistent with a period of rapid remodeling during early amastigote development in which proteins and membranes are expected to undergo extensive turnover , Gene Ontology enrichment analysis identified ribosomal RNA processing ( GO:0006364 ) , protein translation ( GO:0006412 ) and protein folding ( GO:0006457 ) as significantly enriched GO terms associated with nascent and developing amastigotes ( S9 Table; Trypo vs Ama4-Ama12 ) . Despite the fact that nascent intracellular amastigotes will not undergo a first round of replication for another ~20 hours , preparation for the eventuality of cell doubling is already evident in the transcriptome of early amastigote stages . Consistent with the projected increase in nucleic acid synthesis , nucleoside transporters ( TcCLB . 506203 . 10; TcCLB . 506773 . 50 ) and enzymes involved in pyrimidine synthesis ( eg . orotidine-5-phosphate decarboxylase; TcCLB . 508373 . 29 ) and purine salvage are upregulated in immature amastigotes ( S5 Table ) . Genes encoding enzymes in the guanine branch of the purine salvage pathway [66] are selectively upregulated in intracellular amastigotes as compared to trypomastigotes ( eg . guanine deaminase: TcCLB . 504431 . 100; inosine 5' monophosphate dehydrogenase ( IMPDH ) TcCLB . 507211 . 40; TcCLB . 511351 . 9; XPRT; GMP synthase: TcCLB . 508085 . 10; GMP reductase; TcCLB . 506519 . 130 ) . In contrast , enzymes associated with the adenine branch are preferentially expressed in trypomastigotes over amastigotes: eg hypoxanthine-guanine phosphoribosyltransferase: TcCLB . 506457 . 30; adenine phosphoribosyltransferase: TcCLB . 508207 . 74; adenylsuccinate synthetase: TcCLB . 508731 . 60 ) . This observation raises the possibility that flux through the purine salvage pathway is tuned to the different environments encountered by T . cruzi life stages . The anticipated demand for lipid precursors to support membrane synthesis in T . cruzi amastigotes is mirrored by the enrichment of GO functions associated with isoprenoid ( GO:0008299 ) , sterol ( GO:0006696 ) and fatty acid ( GO:0006633 ) synthesis at early stages of amastigote development ( S9 Table ) . Specifically , several enzymes in the mevalonate pathway are upregulated ( Fig 2A ) as are the first two enzymes in the fatty acid synthesis/elongation pathway [67] ELO1 ( TcCLB . 506661 . 30 ) and ELO2 ( TcCLB . 506661 . 20 ) ( Fig 2B and S5 Table ) . Combined , these observations indicate that intracellular T . cruzi amastigotes generate sterols and fatty acids de novo to support replication and membrane homeostasis . Despite its biosynthetic capacity , T . cruzi may opt to scavenge some lipids or precursors from its host cell as seen with other parasites [68–70] . It is currently unclear if or how T . cruzi amastigotes balance de novo synthesis of macromolecular precursors with uptake from the host cell . Intermediary metabolism has been of interest to biochemists in the T . cruzi field for more than fifty years ( e . g . [71 , 72] ) . T . cruzi , like its trypanosomatid relatives , has a partially compartmentalized glycolytic pathway [73] and a non-canonical TCA cycle is predicted [74] . All major T . cruzi life stages exhibit the capacity for oxidative phosphorylation [75 , 76] . Although developmental differences in energy metabolism have been documented for T . cruzi ( e . g . [77 , 78] ) , the specific impact of mammalian host cell colonization on parasite and host bioenergetics remains a poorly understood aspect of the host-parasite relationship . Here , we report dynamic changes in the expression of T . cruzi genes involved in energy metabolism as trypomastigotes establish intracellular infection in mammalian cells ( Fig 3A ) . Highlighted are glycolytic enzymes ( Fig 2C ) and components of the mitochondrial electron transport chain ( Fig 2D ) that exhibit biphasic responses during infection . T . cruzi transcripts encoding glycolytic enzymes were rapidly repressed in nascent amastigotes as compared to trypomastigotes ( 4–12 hpi ) consistent with an earlier report that intracellular amastigotes do not take up hexose sugars [79] . However , transcript levels corresponding to a subset of glycolytic genes rebound in mature T . cruzi amastigotes ( Fig 2C ) , with the emergence of ‘Glycolysis’ as an enriched GO term ( GO:0006096 ) at this stage ( >24 hpi ) as compared to immature amastigotes ( eg . Ama4 ) ( S9 Table ) suggesting that replicative amastigote stages likely retain some capacity for glycolysis inside the mammalian host cell . Genes encoding enzymes involved in mitochondrial oxidative phosphorylation are also upregulated in the intracellular replicative stages of T . cruzi as compared to trypomastigotes ( Fig 2D; Fig 3A ) suggesting that the respiratory capacity differs for these two parasite life stages . A mitochondrial stress test was performed to test this prediction . In line with early comparative studies of mitochondrial respiratory capacity in T . cruzi life cycle stages [75 , 80] , basal respiration ( Fig 3B ) and ATP-coupled respiration ( Fig 3C ) were found to be similar for trypomastigotes and amastigotes . In contrast , the bioenergetic properties of these two parasite life stages diverged significantly at the level of mitochondrial spare respiratory capacity ( SRC ) ( Fig 3D ) , an indicator of the potential of a cell to respond to sudden increases in energy demand [81] . In repeated measurements , isolated intracellular amastigotes displayed measurable mitochondrial reserve capacity whereas trypomastigotes had none ( Fig 3D ) . Furthermore , trypomastigotes failed to maintain ATP levels in the absence of exogenous carbon , whereas homeostatic mechanisms to preserve cellular ATP levels were evident in amastigotes ( Fig 3E ) . We hypothesize that mitochondrial reserve capacity may be important for T . cruzi amastigotes in the context of cell/tissue infection to provide a buffer against environmental stressors such as oxidative stress ( e . g . [82] ) . A key short-term regulator of spare respiratory capacity in cells is cytochrome c oxidase ( mitochondrial complex IV ) [83] . Several cytochrome c oxidase subunits are more highly expressed at the transcript level in mature amastigotes as compared to trypomastigotes ( Fig 2D and S5 Table ) . Differences in expression of this enzyme complex may contribute to observed homeostatic differences and mitochondrial SRC between these life stages . In the related organisms , T . brucei and Leishmania , cytochrome c oxidase subunit expression correlates with mitochondrial respiration rates , ATP production , parasite replication and virulence [84–86] . In this light , it would be interesting to probe the relationship between cytochrome c oxidase activity , mitochondrial spare respiratory capacity and T . cruzi strain-dependent differences in host infectivity and virulence . The main carbon sources that fuel energy production in intracellular T . cruzi amastigotes are not definitively known . It is assumed that glucose is limiting in the host cell cytosol and that intracellular amastigotes rely on uptake of amino acids and fatty acids for energy [42 , 87] . Our transcriptomic data generally support this projected trend with the anticipated increase in expression of fatty acid oxidation genes ( Fig 2E ) and of several amino acid permeases ( S5 and S10 and S8 Fig cluster 5 ) during amastigote development in mammalian cells . In addition , glutamate dehydrogenase ( GlutDH ) was found to be highly expressed in replicative T . cruzi amastigote stages as compared to trypomastigotes ( Fig 2F and S5 Table ) . GlutDH exerts an important anapleurotic function by converting glutamate to α-ketoglutarate that feeds into the TCA cycle to replenish intermediates diverted to biosynthetic functions . T . cruzi has two different glutamate dehydrogenase activities: one NAD+-linked [88] and the other NADP+-linked [89] . Both activities are expressed in epimastigotes [90]; also reflected in our mRNA expression analysis ( S5 Table ) . However , only the NADP+-linked enzyme ( TcCLB . 507875 . 20 ) is upregulated in intracellular amastigotes ( Fig 2F ) . The significance of this finding is unknown but suggests a degree of specialization for these enzymes in the two main replicative T . cruzi life stages . Consistent with higher GlutDH expression in T . cruzi amastigotes , we find that exogenous glutamine drove higher oxygen consumption rates ( OCR ) in isolated amastigotes as compared to trypomastigotes ( Fig 3F ) . Oxidative phosphorylation is inhibited in both parasite life cycle stages by ELQ271 ( Fig 3F ) , an endochin-like quinolone that selectively inhibits mitochondrial complex III activity of apicomplexan parasites over mammalian cells [91–93] . We further show that intracellular T . cruzi amastigote growth is inhibited by ELQ271 in a dose-dependent manner ( Fig 3G ) , whereas growth/viability of human fibroblast host cells is not compromised ( Fig 3G; insert ) as expected [91–93] . Although it is well-established that mitochondrial respiration in T . cruzi is sensitive to the mammalian complex III inhibitor , antimycin A [75] this is the first demonstration of sensitivity of a kinetoplastid protozoan to endochin-like quinolones . These data agree with the recent demonstration that electron transport is a targetable process in intracellular T . cruzi [94] and support the concept that mitochondrial respiratory chain activity may be essential for T . cruzi amastigote proliferation in mammalian cells . As an obligate intracellular T . cruzi life cycle stage , amastigotes must tap into the nutritional resources of their mammalian host cells in order to survive . The ability of T . cruzi to colonize a wide variety of mammalian cell types suggests a high degree of metabolic flexibility and the capacity for rapid adaptation . With the exception of the essential nutrients that T . cruzi is incapable of synthesizing ( eg . purines , pterins ) we have little knowledge of what cytosolically-localized T . cruzi amastigotes extract from their host cells or how energy metabolism is balanced in this parasite life cycle stage . While isotopic tracer experiments are required to make any definitive statements regarding nutrient uptake and utilization by intracellular T . cruzi amastigotes , the dynamic changes observed for core metabolic processes at the transcriptome level is indicative of metabolic remodeling during T . cruzi amastigote development . Similar metabolic retooling has been described in related kinetoplastid protozoan parasites [95–99] . Going forward , it will be critical to understand how T . cruzi amastigote metabolism is wired , how it is couples to host metabolic pathways , the degree of flexibility that exists within these connections and how this can change in the context of the different cell types that T . cruzi colonizes in the human host . Transcriptomic changes induced in mammalian host cells by T . cruzi have been reported in a variety of host cell types and under different experimental conditions [100–110] . Because T . cruzi is capable of infecting most nucleated mammalian cell types , there has been little consistency among these experiments , complicating direct comparison of host transcriptional response data . Here we opted to use human foreskin fibroblasts ( HFF ) as the model host cell type for T . cruzi infection to facilitate comparisons to microarray hybridization studies previously conducted by our group [100 , 107] . As outlined in the Methods section , our experimental approach permitted the capture of both parasite and host transcriptome response information across an infection time course in vitro ( Fig 1A ) . RNA-Seq libraries generated in parallel for mock-infected HFF cultures provided the appropriate controls for each infection time point . As anticipated , some of the previously documented features of the global host transcriptional response to T . cruzi infection [100 , 107] were recapitulated in the present analysis ( S7 Table ) as discussed below . One notable difference , however , relates to the detection of ~450 differentially expressed genes in T . cruzi-infected fibroblasts within the first 4 hpi of infection ( S4 Table ) contrasting sharply with the minimal response previously observed at early parasite infection time points [100] . The enhanced detection capability is likely due to the increased dynamic range and sensitivity achieved with the deep sequencing approach used here . The Gene Ontology enrichment categories associated with the early transcriptome response in T . cruzi infected fibroblasts ( 4–6 hpi ) , while numerous ( S10 Table ) , can be distilled into two main categories: host cell cycle progression and immune response . Among the 288 fibroblast genes that are upregulated ≥2-fold following parasite infection at 4 hpi ( S4 and S7 Tables ) a significant enrichment in functions related to cell cycle progression , mitosis and cell division are observed eg . GO:0000278 ( S10 Table; upregulated ) . Plotting the mRNA expression dynamics for several cell cycle regulators ( Fig 4A ) shows this trend continuing until 24 hpi , after which the expression of host cell cycle genes declines precipitously ( Fig 4A , S8B Fig cluster 2 , and S12 Table ) such that ‘mitotic cell cycle’ becomes an enriched biological process associated with downregulated host genes ( S10 Table; downregulated ) . Overall , these observations coincide with our previous finding that T . cruzi infection pushes host cells toward S-phase in the first 24 hr of the infection cycle , with a subsequent block imposed on host cell cytokinesis at later time points [107] . An innate immune response to T . cruzi infection was also evident in the early transcriptome signature of infected fibroblasts ( S10 Table; upregulated; e . g . GO:0002376 ) with the upregulation of pro-inflammatory cytokine and chemokine genes ( Fig 4B and S7 Table ) as well as type I interferon inducible genes ( Fig 4C and S7 Table ) with different dynamics ( Fig 4C ) . Cytokine/chemokine gene expression peaks at 24 hpi ( Fig 4B ) whereas the type I IFN response ( ie . genes that are expressed downstream of type I IFN receptor activation ) increases gradually over the infection time course ( Fig 4C ) to become the dominant host transcriptomic signature by 72 hpi ( S10 Table; upregulated ) . Differences in the expression profiles of these distinct immune response pathways is presumably related to differences in the mechanism of pathway activation by T . cruzi [111–116] and regulatory processes related to signal amplification [107] . Pro-inflammatory cytokine activation via Toll-like receptor ( TLR ) and cell-intrinsic response pathways is required for host protection against T . cruzi [111 , 113–116] . In contrast , the type I IFN response does not require TLRs for activation in response to T . cruzi [112] and is associated with exacerbation of T . cruzi infection under instances of high parasite load to the detriment of the host [117] , similar to the impact of type I IFNs on the host in a number of other non-viral pathogen infection models ( reviewed in [118] ) . Although not required for host protection , type I IFN ( IFNA6 ) and several IFN-inducible genes ( IFI44L , STAT1 and GBP1 ) emerged in an unbiased RNAi screen conducted in HeLa cells [119] as positively effecting the T . cruzi infection process . This finding raises the unexplored possibility that the host type I IFN response triggered by T . cruzi may be beneficial to the parasite under certain circumstances . Corresponding with increasing intracellular parasite burden is the elevated expression of host genes related to metabolism . Included in this list are several classes of solute transporter ( S7 Table ) and enzymes involved in lipid biosynthesis ( S8B Fig; cluster 4 ) . In fact , most of the genes in the mevalonate pathway are upregulated ≥2-fold between 24–72 hpi ( Fig 4D ) as is SREBP2 ( ENSG00000198911 ) , an important regulator cholesterol homeostasis in mammalian cells [120] . It is tempting to speculate that elevated sterol biosynthesis fuels membrane synthesis in infected host cells to accommodate a steadily increasing intracellular parasite load and/or provides a pool of sterol intermediates to be scavenged by intracellular amastigotes . However , given that cholesterol biosynthesis is an intensely oxygen-consuming process [121] , it is possible that the main function of this late host response to T . cruzi is defense against oxidative stress . While the presence of replicating parasites in the host cell cytoplasm is expected to dramatically impact energy homeostasis in the host cell , energy metabolism does not emerge as an enriched GO function at any time in the intracellular infection cycle ( S12 Table ) . Thus , the cellular response to such perturbations is predicted to occur at the post-transcriptional and post-translational levels . The ability to simultaneously capture host cell and parasite transcriptomes with high resolution and sensitivity sets the stage for the generation of host-parasite interaction networks . The integration of transcriptome data with other types of expression data and with functional information will be instrumental in modeling the critical aspects of the parasite-host interaction and aid in the identification of targetable processes therein . As a limited exercise , we performed an intersection of datasets containing host genes that are upregulated in response to T . cruzi infection ( 24 hpi; S7 Table ) with those previously shown to impact T . cruzi growth in a genome-scale RNAi screen [119] . Within this subset of genes ( S13 Table ) is GCH1 , which encodes GTP-cyclohydrolase 1 , the rate-limiting enzyme in tetrahydrobiopterin ( BH4 ) synthesis . In previous work we demonstrated that siRNA-mediated knockdown of GCH1 expression in host cells impaired intracellular T . cruzi amastigote growth in a manner that was rescued by the addition of dihydrobiopterin [119] . Here we find that GCH1 ( ENSG00000131979 ) is upregulated in T . cruzi-infected fibroblasts ( S7 Table; ≥12 hpi ) at the same time that its negative regulator ( GCHI feedback regulator; ENSG00000137880 ) is repressed ( S7 Table ) . Additionally , the T . cruzi gene encoding pterin-4-alpha-carbinolamine dehydratase ( PCDB1: TcCLB . 503613 . 40 ) , an enzyme involved in biopterin recycling , was also found to be rapidly upregulated in developing intracellular amastigotes ( S5 Table; 4 hpi ) . Together , these observations are consistent with a predicted need to increase flux through the host BH4 synthesis pathway to fuel the growth of intracellular T . cruzi , a pterin auxotroph [122] . With this example , it is possible to see how threads of a functional host-parasite network can emerge from data integration , an important goal of host-pathogen transcriptomic studies going forward . A key feature of our work is the demonstrated ability to parse out human and parasite sequence reads from complex pools generated from T . cruzi-infected cells and to obtain high coverage of both the parasite and host transcriptomes to enable downstream analyses with high statistical confidence . As emphasized here , the simultaneous capture of dynamic changes in host and parasite gene expression over an infection time course provides immediate and new insights into the biology of T . cruzi infection and serves as a unique resource for the construction of high-resolution maps of parasite-host interactions . Moreover , the transcriptome dynamics observed during a major life stage transition in T . cruzi revealed marked similarities between this parasite and its trypanosomatid relatives with respect to the ordering of processes that control global gene expression during differentiation . Specifically , our data support previous observations in T . brucei [12 , 48] and Leishmania donovani [11 , 13 , 23 , 123] indicating that regulation of mRNA levels exerts the greatest impact during the initial phases of a developmental transition in these parasites , whereas downstream mechanisms such as translational efficiency and post-translational modification dominate in the subsequent phases of development and maintenance . While regulation of gene expression in T . cruzi is understood to be multi-layered and complex ( eg . [124 , 125] ) , with translational efficiency playing a key regulatory role [18] , our findings argue for the value of transcriptome data to derive meaningful biological inferences related to parasite biology . Extrapolation of this finding to the many hypothetical genes encoded in the T . cruzi genome , for which we now have dynamic mRNA expression data , has exciting implications for biological discovery in the trypanosome field . Finally , integration of transcriptome information , with emerging proteomic ( eg . [87 , 126–133] ) , functional [119] and metabolic data will , in the near future , create novel opportunities to pinpoint critical processes that govern successful pathogen colonization in the host with links to Chagas disease pathogenesis and increase the potential to identify novel targets for this important neglected disease .
Trypanosoma cruzi Y strain [134] was cultivated by weekly passage in LLcMK2 cells ( ATCC #CCL-7 ) in Dulbecco’s modified Eagle medium ( DMEM ) with 2% fetal bovine serum ( FBS ) , 2 mM L-glutamine , 10 mM HEPES and penicillin-streptomycin maintained at 37°C and 5% CO2 . T . cruzi trypomastigotes released into the supernatants of infected LLcMK2 cells were collected , pelleted by centrifugation ( 1000g , 10 min ) and collected from the supernatant after swimming up from the pellet over a 2–4 hour incubation at 37°C , 5% CO2 . Human foreskin fibroblasts ( HFF ) ( ATCC #CRL-2522 ) were seeded onto 10 cm2 plates or T-25 flasks in complete DMEM ( as above , with 10% FBS ) and grown to 80% confluence ( 48 hr ) before infection . Briefly , HFF monolayers were washed with DMEM-2%FBS ( DMEM-2 ) and either incubated with medium ( mock ) or T . cruzi trypomastigotes for 2h before washing 5 times with PBS and incubation in DMEM-2 at 37°C and 5% CO2 . At the indicated time points ( 4–72 hpi ) , cells were rinsed with ice-cold PBS and lysed directly in Trizol for RNA isolation . T . cruzi epimastigotes , maintained at mid-log phase in liver infusion tryptose medium at 27°C , or purified extracellular trypomastigotes ( >95% pure ) were also used for RNA isolation . To obtain intracellular T . cruzi amastigotes for metabolic studies , infected LLcMK2 ( 60 hpi ) were washed with ice-cold Krebs-Henseleit Buffer ( KHB ) containing 0 . 5mM glucose , and scraped into 1ml ice-cold KHB + 0 . 5mM glucose with a cell scraper . Dislodged cells were collected in 9ml final volume of KHB + 0 . 5mM glucose in a 50ml tube , and centrifuged at 2100g for 10 minutes at 4°C . Pelleted cells were resuspended in 1ml of KHB + 0 . 5mM glucose , transferred to Eppendorf tubes , vortexed for 45s and passed through a 1ml syringe with a 28 . 5G needle 20 times to release amastigotes . Unbroken cells and debris were pelleted at 100g for 1 minute and the amastigote-enriched supernatant was pelleted at 1000g for 10 minutes at room temperature . Amastigotes were resuspended in warm ( 37°C ) KHB + 0 . 5mM glucose at 2x107 amastigotes per ml . RNA was isolated in Trizol reagent , as per manufacturer , and the quality determined using an Agilent 2100 bioanalyzer and quantified by qPCR using a KAPA Biosystems library quantification kit . Standard Illumina protocols were used for mRNA-Seq sample preparation . RNA-Seq libraries were constructed from polyA-enriched mRNAs generated from eight T . cruzi developmental stages: epimastigotes , trypomastigotes , and intracellular amastigotes at 4 , 6 , 12 , 24 , 48 and 72 hrs post-infection ( hpi ) of HFF . Libraries were also constructed from mock-infected HFF at the same time points . For each condition , 2–4 independent biological replicates were sequenced on an Illumina HiSeq1000 . A total of 2 . 7 billion reads from 35 samples were generated from 101 bp paired-ends . The quality of the raw reads was evaluated using the FastQC tool [http://www . bioinformatics . babraham . ac . uk/projects/fastqc/] and one nucleotide was trimmed using the FASTX toolkit ( Hannon Lab , CSHL ) when the mean of the quality score fell below 30 in the last position ( see analysis pipeline S1 Table ) . Tophat v2 . 0 . 8 [36] was used to align all reads to the reference human genome sequence ( hg19 , GRCh37 ) and independently to the of the T . cruzi CL Brener reference genome ( v . 4 . 1 ) [35] Esmeraldo haplotype obtained from the TriTrypDB database ( www . tritrypdb . org ) . Alignment settings allowed 2 mismatches per read and the default -g/—max-multihits parameter of -g = 20 was used for alignments to the human genome . A parameter of -g = 1 was used for T . cruzi to allow reads to map to a single locus in this organism where multi-gene families are abundant . The read counts per coding sequence ( CDS ) were determined using HTSeq [http://www-huber . embl . de/users/anders/HTSeq/] as listed T . cruzi ( S2 Table ) and human ( S3 Table ) genes . Weakly expressed genes , defined as having less than 1 read per million in ‘n’ of the samples , where ‘n’ is the size of the smallest group of replicates [135] ( here n = 2 and 3 for the T . cruzi and human samples , respectively ) were removed from subsequent analyses . Pearson correlation and standardized median correlation analyses , box plots , Principal Component Analysis ( PCA ) and Euclidean distances-based hierarchical clustering approaches were used to evaluate replicates and the relationships between samples across time points and to visualize sample-sample distances . All components of our statistical pipeline , named cbcbSEQ , can be accessed on GitHub ( https://github . com/kokrah/cbcbSEQ/ ) . Samples that did not pass the following quality assessment procedure were removed: for each sample we computed the median pairwise correlation ( mpc ) to all other samples in the dataset ( S5 Fig ) . A standard outlier identification method was then applied to remove samples with low correlation to the other samples: samples were removed if their median pairwise correlation ( mpc ) is less than Q1 ( mpc ) – 1 . 5 IQR ( mpc ) where Q1 ( mpc ) and IQR ( mpc ) are the first quartile and inter-quartile range of the median pairwise correlation across all samples respectively . HPGL0111 was removed from subsequent analyses as a result . A quantile normalization scheme was applied to all samples [86] . Following log2 transformation of the data , Limma [136] was employed for differential expression analyses . Limma utilizes a standard variance moderated across all genes using a Bayesian model and produces p-values with greater degrees of freedom [137] . When appropriate , the Voom module was used to transform the data based on observational level weights derived from the mean-variance relationship . Experimental batch effects were adjusted by including experimental batch as a covariate in our statistical model . Type I error introduced by multiple testing was corrected with q-value [138] . A contrast matrix was used within Limma . To control for expression profile changes in human cells that occur naturally over time in cell culture , normalized-log2-transformed expression values for each gene in ‘uninfected’ was subtracted from ‘infected’ in the paired uninfected/infected HFF samples at each time point . Differentially expressed genes were defined as genes with q-value < 0 . 05 . Each list of differentially expressed T . cruzi genes generated from pairwise comparisons ( S5 Table ) was submitted to a search against itself using FASTA36 [139] . Groups of genes were counted as paralogous when observed with an e-value ≤0 . 0001 and percentage identity ≥80% . The first lexically listed gene in each group was taken as a representative for the group shown in S6 Table . It should be noted that while a large proportion of paralogous genes have been collapsed , a number of truncated ( partial ) genes as well as genes encoding large hypervariable regions ( i . e . MASP ) can only be manually removed , a process more prone to error due to reliance on manual curation . K-means clustering analysis was performed to identify genes with similar expression profiles across different developmental stages of T . cruzi or human host cells with the R function “kmeans” and using the Hartigan-Wong algorithm . Quantile-normalized and batch effect-adjusted expression values were used for clustering and Euclidean distance was computed as the distance metric; 8 partitions were used to generate the clusters following the method of [41 , 140] . Lists of significantly regulated genes resulting from differential expression or clustering analyses were classified into GO functional categories and tested for enrichment using GOSeq , which applies Wallenius approximation to correct the bias of over-detection of differential expression for long and highly expressed transcripts [40] . False discovery rate ( FDR ) was controlled using the Benjamini and Hochberg's procedure [141] . Mitochondrial respiratory capacity was measured using an XFe24 extracellular flux analyzer ( Seahorse Biosciences ) . Briefly , XFe24 assay plates were pre-coated with 30 μl of 7 . 7% Cell-Tak ( Corning ) in 100 mM sodium bicarbonate , pH 8 for 30 minutes , then the wells were washed three times with 0 . 5 ml warm Krebs-Henseleit Buffer ( KHB ) before plating parasites . Isolated T . cruzi ( Tulahuen strain ) [142] trypomastigotes and amastigotes were resuspended in either XF Base Medium ( Seahorse Biosciences ) + 10mM glucose , 2mM glutamine , and 1mM sodium pyruvate or KHB + 0 . 5mM glucose at 2x107 parasites/ml . 2x106 parasites in 100 μl were delivered to each well of a Cell-Tak pre-coated Seahorse XFe24 assay plate and immediately centrifuged at 2056g for 2 minutes . The volume of medium in each well was adjusted to a total volume of 450 μl/well of plating medium . To determine basal respiratory capacity , ATP-coupled respiration , and spare respiratory capacity , drugs from the XF Cell Mito Stress Test Kit ( Seahorse Biosciences ) were resuspended in warm media and injected at 10x their final well concentrations of 2 . 5μM oligomycin , 3μM carbonyl cyanide-4- ( trifluoromethoxy ) phenylhydrazone ( FCCP ) , and a mixture of 1μM antimycin A and rotenone , in order . Results were normalized to parasite DNA based on a quantitative PCR assay of the single copy T . cruzi gene: OSBP ( TcCLB . 508211 . 10 ) . At the end of the run , the media was replaced with 200 μl PBS and the plate was frozen at -20°C . DNA was isolated after thawing the plate using the DNeasy Blood and Tissue Kit ( Qiagen ) . For qPCR , 10 μl of iTaq Universal SYBR Green Supermix ( BioRad ) was combined with 0 . 33 μl isolated DNA , 1 pmol each of primers to amplify a single copy T . cruzi gene OSBP ( F: 5’-CAT CAC CTA CGG CCA CAA GA-3’ , R: 5’-TGC AGT GGA TAC GCA TAC GG-3’ ) , and water for a 20 μl reaction volume . The reaction was run at 95°C for 5 minutes , then cycled 45 times at 95°C for 15 seconds followed by 60°C for 60 seconds . The amount of DNA per Seahorse plate well was calculated by comparing Ct values to a standard curve generated by 1:2 dilutions of 10 ng of T . cruzi DNA . To calculate basal respiratory capacity , the normalized oxygen consumption rates ( OCR ) after antimycin A and rotenone addition was subtracted from the baseline OCR . For ATP-coupled respiration , normalized OCR after oligomycin injection was subtracted from baseline OCR . Spare respiratory capacity was calculated as the difference between normalized OCR after FCCP injection and control wells at the same timepoint without any drug injections . Glutamine ( Gibco , Life Technologies ) and ELQ271 ( generously supplied by M . Riscoe; OHSU ) were resuspended in warm medium and injected at 10x their final well concentrations of 2mM and 10μM , respectively . To measure total ATP levels , T . cruzi trypomastigotes and amastigotes were isolated ( as above ) and resuspended in Krebs-Henseleit Buffer ( KHB ) at 4x106 parasites/ml . Trypomastigotes and amastigotes were plated at 4x105 parasites in 100μl in a separate white 96-well plates ( Corning ) for each time point and incubated at 37°C . Parasites were lysed and ATP levels were determined using the ATPlite assay kit ( PerkinElmer ) , measuring luminescence on a Varioskan Flash plate reader ( ThermoScientific ) . HFF were seeded at 1500 cells/well in 384 well black bottom plates ( Corning ) . At 24h post plating , cells were infected with β-galactosidase expressing T . cruzi Tulahuen strain ( moi 5 ) for 2hr , washed twice , and left in DMEM ( 2% FCS , 2 mM glutamine , 1 mM pyruvate ) . At 18hpi cells were treated with 0 . 3–10 μM of ELQ271 . At 72 hpi HFF viability was measured in a fluorescence-based readout ( CellTiter-Fluor , Promega ) and T . cruzi was measured by luminescence-based readout ( Beta-Glo reagent , Promega ) using an Envision Plate Reader ( PerkinElmer ) as described [32] . The relative infection ( RLU/RFU ) was calculated and normalized to untreated control fitted by non-linear regression to high ( = 100% ) and low ( = 0% ) values using GraphPad Prism software . RNA was isolated from purified T . cruzi trypomastigotes and from infected monolayers ( at the indicated timepoints post infection ) following cell lysis in Trizol reagent and purification using the PureLink RNA Mini Kit ( Ambion ) . DNase-digested ( Turbo DNase , Ambion ) total RNA ( 1 μg ) was converted to cDNA using the iScript ( Bio-Rad ) cDNA synthesis kit according to manufacturer’s instructions . Specific primer pairs to amplify genes of interest in quantitative-RT-PCR reactions were selected on their ability to form single peak in melting curve analysis and verified by sequencing of PCR products . Forward ( F ) and reverse ( R ) primer pairs for are listed below in a 5’ to 3’ orientation . Cation transporter ( TriTrypDB: TcCLB . 509197 . 39 ) F: GAGTGTCATGCTTGAAGTG and R: CGTTAAAAATAAGAGAAAATG; Glutamate dehydrogenase ( TriTrypDB: TcCLB . 507875 . 20 ) F: GAGTACTGCCAGGATTCTC and R: CAAAGCCAAGAAACTTAAG; Fatty acid elongase ( TriTrypDB: TcCLB . 506661 . 30 ) F: GAGGCAACCTGCACATTTAAC and R: GTGTCCATCAACTCAGGAATCT; Fatty acid desaturase ( TriTrypDB: TcCLB . 511073 . 10 ) F: AAGGAACGTGAAGAATCTC and R: AACGGACTTCTCCAGATC; Hypothetical protein , conserved ( TriTrypDB: TcCLB . 509767 . 170 ) F: ATGAAGCTTGCGTTCTCT and R: GGTCACAATAGCCCAGTC; Ribosomal RNA large subunit gamma M1 ( TriTrypDB: TcCLB . 411483 . 20 ) F: TGTGGAAATGCGAAACAC and R: CCCAGGTTTTTGCTTTAATG . Relative mRNA transcript abundance was quantified by SYBR green ( iTaq Universal SYBR Green Supermix , Bio-Rad ) PCR using a StepOnePlus Real-Time PCR Systems ( Applied Biosystems ) . Large subunit ribosomal RNA gamma ( M1 ) : TcCLB . 411483 . 20 ) used as endogenous control . RNA-Seq data are available at the National Center for Biotechnology ( NCBI ) Sequence Read Archive ( SRA ) ( http://www . ncbi . nlm . nih . gov/bioproject ) under Bioproject PRJNA251582 ( accession numbers ranging from SRR1346026-SRR1346052 ) and Bioproject PRJNA251583 ( accession numbers ranging from SRR1346053-SRR1346059 ) . Individual accession numbers are also shown in S1 Table . | In-depth knowledge of the functional processes governing host colonization and transmission of pathogenic microorganisms is essential for the advancement of effective intervention strategies . This study focuses on Trypanosoma cruzi , the vector-borne protozoan parasite responsible for human Chagas disease and the leading cause of infectious myocarditis worldwide . To gain global insights into the biology of this parasite and its interaction with mammalian host cells , we have exploited a deep-sequencing approach to generate comprehensive , high-resolution transcriptomic maps for mammalian-infective stages of T . cruzi with the simultaneous interrogation of the human host cell transcriptome across an infection time course . We demonstrate that the establishment of intracellular T . cruzi infection in mammalian host cells is accompanied by extensive remodeling of the parasite and host cell transcriptomes . Despite the lack of transcriptional control mechanisms in trypanosomatids , our analyses identified functionally-enriched processes within sets of developmentally-regulated transcripts in T . cruzi that align with known or predicted biological features of the parasite . The novel insights into the biology of intracellular T . cruzi infection and the regulation of amastigote development gained in this study establish a unique foundation for functional network analyses that will be instrumental in providing functional links between parasite dependencies and host functional pathways that have the potential to be exploited for intervention . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"viral",
"transmission",
"and",
"infection",
"microbiology",
"parasitic",
"diseases",
"protozoan",
"life",
"cycles",
"parasitic",
"protozoans",
"developmental",
"biology",
"trypomastigo... | 2016 | Transcriptome Remodeling in Trypanosoma cruzi and Human Cells during Intracellular Infection |
The neural crest ( NC ) is a vertebrate-specific cell type that contributes to a wide range of different tissues across all three germ layers . The gene regulatory network ( GRN ) responsible for the formation of neural crest is conserved across vertebrates . Central to the induction of the NC GRN are AP-2 and SoxE transcription factors . NC induction robustness is ensured through the ability of some of these transcription factors to compensate loss of function of gene family members . However the gene regulatory events underlying compensation are poorly understood . We have used gene knockout and RNA sequencing strategies to dissect NC induction and compensation in zebrafish . We genetically ablate the NC using double mutants of tfap2a;tfap2c or remove specific subsets of the NC with sox10 and mitfa knockouts and characterise genome-wide gene expression levels across multiple time points . We find that compensation through a single wild-type allele of tfap2c is capable of maintaining early NC induction and differentiation in the absence of tfap2a function , but many target genes have abnormal expression levels and therefore show sensitivity to the reduced tfap2 dosage . This separation of morphological and molecular phenotypes identifies a core set of genes required for early NC development . We also identify the 15 somites stage as the peak of the molecular phenotype which strongly diminishes at 24 hpf even as the morphological phenotype becomes more apparent . Using gene knockouts , we associate previously uncharacterised genes with pigment cell development and establish a role for maternal Hippo signalling in melanocyte differentiation . This work extends and refines the NC GRN while also uncovering the transcriptional basis of genetic compensation via paralogues .
Development from a single fertilised cell to the complex adult form requires a simultaneously robust and plastic gene regulatory program . The neural crest is a transient pluripotent stem cell population capable of crossing germ layer boundaries and differentiating into highly diverse tissue types while migrating long distances in the developing embryo . The establishment of the neural crest and its subsequent tissue derivatives is specific to vertebrates and has played a fundamental role in their variation and evolutionary success [1–4] . Neural crest cells require a complex combination of external inductive signals such as Wnts , Fgfs , Notch/delta and Bmps ( Fig 1A ) . These extrinsic signals can be considered the first phase of the neural crest gene regulatory network ( GRN ) followed by a second phase of tightly controlled intrinsic gene expression . In this context foxd3 initially promotes neural crest fates by acting as a transcriptional repressor whereas later in development foxd3 promotes neural crest fates as a transcriptional activator [5] . Two other intrinsic signals of fundamental importance for evolution and development of the neural crest that set vertebrates apart from other chordates such as amphioxus and tunicates are the AP-2 and SoxE genes families [6–10] . Mutations in neural crest genes lead to disease in humans , highlighting the importance of this cell population for human health . Animal models faithfully recapitulate these defects demonstrating functional conservation . In humans and mice , mutations in TFAP2A lead to branchio-oculo-facial syndrome presenting as defects in cranial development and cranial closure [11 , 12] . Similarly , mutations in zebrafish tfap2a lead to craniofacial defects in addition to a reduction in melanocytes [13 , 14] . The Tfap2 family arose from a single gene in a chordate ancestor that underwent gene duplication resulting in five family members ( tfap2a , tfap2b , tfap2c , tfap2d and tfap2e ) in zebrafish . Removing combinations of tfap2 family members results in a wide array of phenotypes . For example , the neural crest is completely ablated in tfap2a;tfap2c double homozygous zebrafish whereas there is a dramatic and specific reduction of melanocytes in tfap2a;tfap2e double homozygous zebrafish embryos [15–20] . Furthermore , melanomas , squamous cell carcinomas , most skin and breast cancers and a few cervical and urothelial cancers have strong nuclear immunoreactivity for TFAP2A . [21 , 22] . Haploinsufficiency in one of the AP-2 targets , the SoxE family member SOX10 , results in Waardenburg syndrome; patients exhibit defects in the peripheral and enteric nervous systems and also pigmentation defects [23 , 24] . In zebrafish the known SoxE family members consist of sox8a/b sox9a/b and sox10 . The expression of sox10 is first detectable in premigratory neural crest cells and expression is maintained in certain neural crest linages , for example , glia , but reduced in many other neural crest-derived tissues in zebrafish [25–27] and mouse [28–30] . Following neural crest induction , sox10 plays a vital role in the establishment of non-ectomesenchymal neural crest cells in zebrafish and driving mitfa expression . Knockouts in zebrafish sox10 behave in a recessive manner and lead to the absence of enteric neurons , chromatophores , Schwann cells , sensory neurons and other trunk crest cell types [31 , 25 , 27] . Craniofacial features appear to be largely unaffected in zebrafish sox10 mutants , which is thought to be due to compensation by sox9b in ectomesenchymal neural crest [32 , 33] . Mutations in mitfa lead to the total lack of body melanocytes in zebrafish due to a failure in melanocyte differentiation and as such mitfa is considered to be the melanocyte master regulator transcription factor [34] . Many crucial transcription factors involved in the neural crest GRN have been identified and studied in depth across a number of different species [10 , 35 , 36] and are largely conserved across vertebrates [37] , but many of their downstream targets and interaction partners still remain to be elucidated . For example , AP-2A ChIP-seq analysis using human neural crest cells has identified over 4 , 000 potential AP-2A binding sites and established AP-2A as a chromatin initiating factor [38] . This large number of putative AP-2A downstream targets now requires functional validation . In this work we use transcriptional profiling of zebrafish mutants in genes required at different levels of neural crest induction and differentiation to dissect the GRN downstream of tfap2 . Specifically , by using individual genotyped embryos and many biological replicates , we can identify molecular neural crest signatures before a morphological phenotype arises . Stepwise genetic ablation of tfap2 levels reveals in detail the gene regulatory basis of dose-dependent genetic compensation between two AP-2 paralogues . By analysing the dosage compensation we have identified a subset of genes required to rescue the neural crest . To validate a small subset of novel candidates emerging from this analysis we applied a reverse genetics approach to knock out genes of interest using both ENU and CRISPR/Cas9 mutagenesis [39–41] . Taken together , this work has identified early activation of members of the neural crest GRN and the core gene set underlying genetic compensation of tfap2a or tfap2c perturbations . Our screen has also identified novel downstream neural crest genes and a role for maternal expression of the Hippo signalling member yap1 in the differentiation of melanocytes . All resources are publicly available and we envisage that this will lead to a deeper understanding of neural crest biology .
Neural crest cells can be readily identified as the first somites begin to form , however it is not clear when the neural crest GRN becomes active in the zebrafish embryo . We used a wild-type developmental time course we had published previously [42] encompassing 18 stages from zygote to 5 dpf to identify the specific time points at which relevant transcripts are activated and their expression over time . In addition , the use of single embryos reveals the natural variation across individuals ( Fig 1B–1G ) . In zebrafish , the genome first becomes transcriptionally active between the 1K-Cell and Dome stage [43–45] . A number of early neural crest transcription factors—foxd3 , tfap2a , tfap2c - can be detected at the Dome stage . This is much earlier than neural crest cells are formed , but our data suggest that the top tier of the neural crest GRN has already commenced at these early developmental time points and is expressed in relevant ectoderm-forming regions ( Fig 1B–1D ) [2 , 46] . Their downstream targets sox9b and sox10 begin to be expressed between 75% epiboly and when the first somites appear ( Fig 1E and 1F ) . Both sox9b and sox10 have been shown to be robust markers for premigratory neural crest cells in zebrafish [47] . We first created a catalogue of genes enriched in premigratory and differentiating neural crest cells as a reference set for the subsequent transcriptional analysis of the neural crest mutants . We used Fluorescence-Activated Cell Sorting ( FACS ) on dissociated cells from whole embryos of the sox10:mg line [48] at 22–23 hours post fertilisation ( hpf ) . The transgenic reporter labels neural crest nuclei ( mCherry ) and crest cell membranes ( GFP ) . At 22–23 hpf neural crest cells are migrating in a dorsal to ventral direction and their differentiation is more advanced at the rostral than caudal part of the embryo . We therefore reasoned that this stage would provide us with a comprehensive mixture of neural crest differentiation states . We compared transcripts detected in the transgenic neural crest cell populations to the non-crest cells using DESeq2 to produce neural crest-enriched gene sets ( Fig 1I ) . For comparison to our whole embryo data we aggregated the resulting gene lists from the individual FACS experiments to produce a set of 4995 genes enriched in any FACS neural crest cell population ( Table 1 ) . In order to establish the initiation of the sox10 GRN , a direct transcriptional target of tfap2a;tfap2c[49] , we first created a transcriptional loss of function time course of sox10 and its target , mitfa , by comparing gene expression of homozygous mutants and siblings . Zebrafish sox10t3/baz1 mutant embryos form premigratory neural crest cells in the trunk but these cells fail to migrate and properly differentiate while cranial crest remains largely unaffected [25] . Mutants of the sox10t3/baz1 downstream target mitfa have mostly correctly differentiated neural crest but specifically lack melanocytes of the body while showing mild differences in the numbers of the other two pigment cell types , xanthophores and iridophores [34] . Fig 1J is an overview of all experiments carried out using DeTCT ( differential expression transcript counting technique ) 3’ tag sequencing [50] . Although sox10 is appreciably expressed at the 1–4 somites stage ( ~10 hpf ) ( Fig 1F ) it is only at the 19 somite stage ( 19 hpf ) where we detected a reduction in the abundance of one of its downstream targets , mitfa , in sox10t3/baz1 embryos ( Fig 1J ) . The majority of genes appearing differentially expressed in the sox10 4 somite , 15 somite and 19 somite stages are localised on chromosome 3 , the same chromosome as sox10 ( Fig 1J , Table 1 ) . We also found similar signals for the mitfaw2/w2 mutants at 4 somites , with a very strong enrichment for chromosome 6 at 24 hpf ( Table 1 ) . This enrichment is either due to differential read mapping between the haplotypes owing to the high genetic variation in zebrafish [51] or reflects true expression differences between the two haplotypes , i . e . allele-specific expression [52] . When embryos homozygous for a specific genomic locus , in these cases the areas around the sox10 or the mitfa mutation , are compared to embryos heterozygous and homozygous for the other haplotype , either possibility will cause an enrichment of DE genes on the chromosome carrying the mutation . We next analysed enrichments of terms from the Zebrafish Anatomy Ontology ( ZFA ) associated with differentially expressed genes and plotted all time points with significant enrichments ( S1 Fig ) . As expected , we found a strong and specific melanocyte signal in both mutants across all time points , with sox10 mutants also showing a strong enrichment at 24 hpf for xanthophores and iridophores . By 36 hpf we also found an enrichment for the terms peripheral nervous system and nervous system which is consistent with an established role for sox10 in peripheral nervous system development [25] . Previous data [25] and our developmental time course show that the expression of sox10 begins early , following the establishment of the first neural crest cells at about 4 somites . It is only at the 19 somite stage , however , in which we detect the first molecular signal via the reduction of mitfa transcript , and only at 24 hpf do we see the first ZFA enrichments . Based on the wild-type expression of tfap2a and tfap2c , the morphological double mutant phenotype and the sox10 molecular phenotype we chose three time points , 4 somites , 15 somites and 24 hpf , for the transcriptomic screen of tfap2a;tfap2c mutants . At the 4 somite stage pluripotent neural crest stem cells should be well established based on snail1b expression [33] and detectable with a whole embryo transcriptomic approach . To genetically ablate the neural crest , we created double carrier fish for tfap2a+/sa24445;tfap2c+/sa18857 ( denoted as tfap2a+/-;tfap2c+/- from here on ) alleles , using mutants produced by the Zebrafish Mutation Project ( ZMP https://www . sanger . ac . uk/resources/zebrafish/zmp/ ) [39] . We confirmed the phenotypes previously described in tfap2a;tfap2c depletion experiments [17 , 20] . Double homozygous embryos were indistinguishable from wild-type siblings at the 4 somite stage but were slightly elongated/dorsalised by the 15 somite stage and were clearly discernible by 24 hpf ( Fig 2A and 2B ) . Notably , we also identified a specific pattern of reduction of dorsal tail melanocytes in tfap2a-/-;tfap2c+/- embryos at 48 hpf ( Fig 2C ) in addition to the melanocyte reduction previously noted in tfap2a-/- embryos which demonstrates a dosage effect of tfap2c heterozygosity on tfap2a homozygous mutants ( Fig 2D ) . Conversely , all other genotypic combinations were indistinguishable from their wild-type siblings at 48 hpf with only tfap2a-/- carriers progressing to present craniofacial defects at 72 hpf as previously described [16] . This phenotypic diversity shows that tfap2a and tfap2c do not act in a simple redundant fashion . In light of the observed phenotypes stemming from a dosage effect of tfap2c heterozygosity in tfap2a homozygous mutants our primary aim was to systematically investigate the genetic interactions of tfap2a and tfap2c . We therefore sequenced up to 10 embryos for all 9 genotypes at the three different stages to enable comparison of all genotypic combinations . We crossed double heterozygous tfap2a;tfap2c parents and collected embryos at the three developmental time points as single embryos . Following nucleic acid extraction and genotyping , single embryos were processed and global mRNA transcript levels determined using 3’ tag sequencing ( Fig 1J ) . After quality control and the removal of outlier samples we carried out pairwise analysis using DESeq2 . Our transcriptomic profiling uses whole embryos , therefore tissue-specific gene expression changes tend to be reflected in smaller log2 fold changes than would be expected from tissue dissection or FACS-derived cell populations . However , using high numbers of biological replicates enables us to faithfully detect smaller , but meaningful , effect sizes . We first assessed how the transcriptomes of the different genotypic conditions behaved across time . Comparing the absolute numbers of differentially expressed ( DE ) genes of the four most relevant knockout genotypes over the three developmental time points revealed three major findings ( Fig 2E ) . Firstly , when compared to wild-type siblings , the number of genes differentially expressed in both tfap2a or tfap2c single homozygous embryos is very small in contrast to the double homozygous knockout and the tfap2a-/-;tfap2c+/- mutants indicating genetic compensation . Secondly , despite the severe morphological phenotype of double mutants at 24 hpf the number of DE genes was less than half of that at the 15 somite stage . Conversely , while only beginning to display a mild morphological phenotype at 48 hpf , the tfap2a-/-;tfap2c+/- mutants showed a strong molecular phenotype at 4 and 15 somites , with a longer DE list at 4 somites than the double mutants . This molecular signature was strongly diminished by 24 hpf . Taken together this demonstrates that the complexity of transcriptional changes is not necessarily mirrored in the morphological phenotype , and vice versa . Next we analysed the transcriptional profile of complete ablation of the neural crest in tfap2a-/-;tfap2c-/- knockouts . A role for tfap2a has been previously described in both neural and non-neural ectoderm tissues which lead to the formation of the neural crest , epidermis , and cranial placodes [17 , 54 , 55] . To separate transcripts into subsets specific to the neural crest or the epidermis we filtered the DE genes from the three developmental time points in tfap2a-/-;tfap2c-/- knockouts relative to wild-type siblings with the list of 4995 FACS-identified neural crest genes ( Fig 2F ) . When all genes which appear in at least one of the developmental stages and the neural crest FACS list are analysed together with their associated GO terms , there is an enrichment for pigment cells and melanocytes but no other neural crest subtypes ( magenta box Fig 2F ) . However , zebrafish anatomy enrichment ( ZFA ) returns a strong enrichment for the neural crest ( Fig 2F , S1 Table ) . This finding highlights the current limitations of zebrafish GO annotation which has a bias for genes linked to pigmentation and lacks annotation for genes associated with earlier neural crest states . A relatively small group of 26 genes ( S2 Table ) appearing in all four data sets included tfap2a , sox10 and many keratins . This could potentially signify an epidermal/neural crest precursor cell type which is in the process of committing to one of the lineages . Comparison of the three developmental time points places genes into “early , ” “mid , ” and “later” neural crest-specific groups . Each of these groups ( Fig 2F ) contain numerous examples of previously characterised neural crest-specific genes common across many studied species [37] which helps to validate this approach , but also many unannotated genes or genes previously not associated with the neural crest ( Table 1 ) . The gene lists shared between the different stages but not found in the neural crest FACS data set ( orange boxes Fig 2F ) and their Gene Ontology ( GO ) term annotation revealed an enrichment for epidermal-related terms . Another subset from the 4 somite and 15 somite stages that is not present in the NC-enriched gene list is a group of genes enriched for expression , translation and RNA processing ( blue box Fig 2F ) . Our next question was how the transcript levels of tfap2a and tfap2c , along with three well characterised neural crest-specific genes ( foxd3 , sox10 and sox9b ) , behaved across all nine genotypes and the three developmental stages ( Fig 3A–3O ) . At 4 somites , embryos homozygous for either tfap2a or tfap2c had significantly lower transcript abundances for their respective genes , indicating that nonsense-mediated decay [56] had most likely occurred ( Fig 3A and 3B ) . A genetic interaction is evident in tfap2a-/-;tfap2c+/+ embryos between tfap2a and tfap2c with higher levels of wild-type tfap2c transcripts than in wild-type siblings ( Fig 3B ) while tfap2a is not increased in the inverse case of tfap2a+/+;tfap2c-/- mutants ( Fig 3A ) . This indicates that , by the 4 somite stage , the neural crest GRN is able to detect reduced levels of tfap2a in knockouts and compensation through upregulation of tfap2c is established . Interestingly , at 4 somites and partially at 15 somites , in homozygous tfap2a mutants transcript levels of the three NC genes follow the tfap2c expression pattern . For example , among embryos homozygous for tfap2a , sox9b expression at 4 and 15 somites is highest in tfap2a-/-;tfap2c+/+ embryos , but drops to half in tfap2a-/-;tfap2c+/- embryos ( Fig 3E and 3J ) . This suggests a direct quantitative relationship between tfap2 transcript abundance and that of its targets . By 24 hpf the abundance of tfap2a and tfap2c across the nine genotypes remains much the same as at the previous developmental stages ( Fig 3K and 3L ) . Interestingly , foxd3 and sox9b levels are no longer significantly different in tfap2a-/-;tfap2c-/- embryos , which is suggestive of their exit from the neural crest GRN or initiation of expression in non-neural crest tissues , but levels of sox10 remain strongly reduced in the double mutants ( Fig 3M–3O ) . Also , tfap2a-/-;tfap2c+/- embryos now have levels of foxd3 , sox9b and sox10 comparable to wild type which suggests a general recovery of the neural crest GRN by this stage . These data show that the time point of the strongest molecular phenotype and tfap2c compensation is at around 4–15 somites with the morphological phenotypes beginning to emerge by 15 somites . To further investigate the dose-dependent compensation while also creating a more detailed transcriptomic profile of pluripotent and differentiating neural crest cells , we carried out RNA-seq on tfap2a;tfap2c knockouts at the 15 somite stage . All 9 genotypes were assessed using a total of 90 single embryos . Principal component analysis highlights that tfap2a-/-;tfap2c-/- and tfap2a-/-;tfap2c+/- are most similar on a molecular level in spite of their vastly different morphological phenotypes ( S2A Fig ) . Pairwise comparisons of four different genotypes to their wild-type siblings shows high numbers of genes changing in both tfap2a-/-;tfap2c-/- and tfap2a-/-;tfap2c+/- groups ( S2B Fig , Table 1 ) . The majority of significant genes have reduced transcript levels in double mutants with robust p-values ( S2C Fig ) . The 15 somite 3’ tag sequencing and RNA-Seq data sets showed good correlation of the detected DE genes at an adjusted p-value < 0 . 01 ( S2D Fig ) Hierarchical clustering on the significantly changed genes from the tfap2a-/-;tfap2c-/- versus wild type pairwise comparison and ZFA enrichment placed genes into functional groups . While loss of both tfap2a and tfap2c leads to a reduction in genes involved in neural crest and epidermis development it also leads to an upregulation of genes associated with neural terms ( S2E Fig ) . The 3’ tag sequencing analysis had highlighted that both tfap2a-/-;tfap2c-/-and tfap2a-/-;tfap2c+/- gave the most extensive molecular phenotypes even though tfap2a-/-;tfap2c+/- were morphologically indistinguishable from wild-type siblings at 15 somites whereas tfap2a-/-;tfap2c-/- presented obvious morphological phenotypes by that stage . Hence a single wild-type allele of tfap2c is sufficient to rescue the morphological tfap2a-/-;tfap2c-/- neural crest specification and differentiation phenotype despite the observed effect on the transcriptional level . We were therefore keen to understand which genes are involved and may be required for the rescue of the morphological phenotype . First , we assessed expression of tfap2c in the 15 somites RNA-seq data and found that the levels of tfap2c were significantly higher in tfap2a-/- embryos when compared to wild-type embryos ( Fig 4A ) demonstrating active regulatory compensation rather than redundancy . We then compared the sets of DE genes derived from the pairwise comparisons of wild type with tfap2a-/-;tfap2c-/- and tfap2a-/-;tfap2c+/- . The vast majority of DE genes in the tfap2a-/-;tfap2c+/- condition were also changed in the tfap2a-/-;tfap2c-/- embryos ( Fig 4B ) . Crucially , this set is enriched for genes with a dose-dependent response to successive loss of tfap2a/c alleles where for each gene the log2 fold change in tfap2a-/-;tfap2c+/- is about half that in tfap2a-/-;tfap2c-/- ( Fig 4C ) . This demonstrates that loss of a third tfap2a/c allele affects the neural crest GRN , however the transcriptional changes are not sufficient to derail neural crest specification and differentiation . Together this identifies a core set of tfap2a/tfap2c-responding genes , separate from secondary downstream events caused by differentiation failure and tissue loss . As a single wild-type allele of tfap2c is able to maintain neural crest specification we sought to identify genes that are sensitive to different levels of tfap2c by dissecting the full ablation response ( tfap2a-/-;tfap2c-/- ) using the partial ablation profiles ( tfap2a-/-;tfap2c+/+ and tfap2a-/-;tfap2c+/- ) . To this end we ran four differential gene expression ( DGE ) analyses: double homozygous embryos against embryos with one or two wild-type alleles of tfap2c , and wild-type embryos against tfap2a-/-;tfap2c-/- or tfap2a-/-;tfap2c+/- . Next we overlapped the four lists to produce 14 subsets ( Fig 4D ) . This identified groups of genes that share distinct expression profiles . Subset one contains genes where tfap2a-/-;tfap2c-/- knockout resulted in a mild , but significant , change from wild-type siblings but there is no significant difference between tfap2a-/-;tfapc+/- and tfap2a-/-;tfap2c-/- or wild-type siblings ( Fig 4E ) . For genes in subset three a complete tfap2a-/-;tfap2c-/- knockout resulted in a significant change from wild-type siblings while a single allele of tfap2c was sufficient to return the expression to wild-type levels . An example of this case would be lef1 ( Fig 4F ) . Subset five contained genes that are only partially rescued , for example sox10 , pax7a , ednrab , kctd15a and erbb3b , all of which play vital roles in neural crest development across vertebrates [37] . A single wild-type allele of tfap2c , or even both wild-type alleles , is unable to return expression to wild-type levels but the expression is still significantly different from the tfap2a-/-;tfap2c-/- condition , as exemplified by mmp28 ( Fig 4G ) . Finally , subset seven contained genes that are only rescued by two alleles of tfap2c , such as krt222 ( Fig 4H ) . We next analysed the gene sets for transcription factor motif enrichment using HOMER[57] ( Fig 4I , see Table 1 for figshare link to all significant results ) . The full DGE list for tfap2a-/-;tfap2c-/- against wild type ( Fig 4B ) had 31 enriched known motifs ( q-value < 0 . 05 ) , which included Ap2gamma and Ap2alpha . These motifs also appeared in the top three of 13 enriched motifs for the DGE list for tfap2a-/-;tfap2c+/- against wild type , suggesting that this profile reflects a core set of Ap2 targets . Analysis of the subsets in Fig 4D revealed that only subset five , which contains genes showing dosage sensitivity and partial rescue in tfap2a-/-;tfap2c+/- embryos , had an enrichment for Ap2 targets . Subset three , the set of genes fully rescued by one or two wild-type copies of tfap2c , was enriched mostly for binding sites of zinc-finger domain-containing transcription factors such as KLFs , but also Tead2 , a transcription factor involved in Yap/Taz Hippo signalling[58 , 59] . For a functional gene analysis , we carried out ZFA enrichment on all 14 gene subset lists which yielded significant enrichments for subsets 1–8 ( Fig 4J , S3 Fig ) . Subset three , the genes fully rescued by either one or two alleles of tfap2c , showed the strongest enrichment for terms associated with the neural crest , head and cranial crest and also fin . While fin enrichment may seem nonsensical for a 15 somite embryo , this is due to the fact that many genes annotated for fin development are also involved in craniofacial development . A similar enrichment profile resulted from subset five , the genes where either one or two alleles of tfap2c rescued expression levels to a significant extent , but not completely . By contrast , the two largest subsets , containing genes that change in double homozygous embryos with respect to wild types , but not compared to tfap2a-/-;tfapc+/- , showed a bias towards nervous system and ectoderm enrichment . Crucially , subsets six and seven with genes that failed to be rescued by either one ( subset seven ) or two ( subset six ) wild-type tfap2c alleles , had no or very little neural crest enrichment . This suggests these genes represent tfap2a targets outside of neural crest differentiation . Taken together the enrichment analysis breaks down the full tfap2a/tfap2c knockout response into separate expression classes with different functional profiles . Subsets three and five contain genes that are fully or partially rescued by tfap2c , show the strongest neural crest enrichment and are thus most likely to represent the core of the tfap2 neural crest GRN . Interestingly , only the partial rescue gene list ( subset five ) is enriched for direct Ap2 targets , suggesting that the full rescue list ( subset three ) , which shows the strongest neural crest signature , contains more genes that are further downstream in the network . Next we applied an expression correlation network and Markov clustering approach using Biolayout Express3D [60 , 61] to identify co-expression profiles independent from condition-driven differential expression analysis . We constructed a network graph with genes as nodes and their Pearson correlation as edges from the tfap2a;tfap2c 15 somite RNA-seq dataset and used Markov clustering ( MCL ) to divide the network into discrete sets of co-expressed genes . The network clusters isolated co-expression groups of genes that share a genomic locus ( Fig 5A’ and S4 Fig ) similar to our observation in the sox10 and mitfa mutants , but also identified tfap2a and tfap2c-specific components ( Fig 5A ) within the larger co-expression network . In total 30 clusters containing a total of 600 genes were driven by changes in the tfap2a or tfap2c genotypes ( Fig 5B ) . It is important to point out that in the previous analysis we compared lists derived from pairwise DGE comparisons , whereas MCL clusters genes based on their expression similarity across all samples . Therefore , these clusters might exclude genes that are identified in the DESeq2 analysis because of low expression correlation with other genes , but also include highly correlated genes which did not produce a significant adjusted p-value in the DESeq2 analysis . The unsupervised clustering confirmed the strong signal in the double homozygous fish ( clusters one and two ) and dose-dependent compensation by tfap2c ( cluster three ) . However , in addition it provided increased functional resolution . For example , cluster 17 ( Fig 5E ) was highly specific to neural crest effectors containing the soxE paralogues sox10 and sox9b , the micropthalmia bHLH transcription factor tfec , as well as the Pak4 kinase inhibitor fam212aa , and one uncharacterised gene ( si:ch211-243g18 . 2; ENSDARG00000044261 ) . The differentiation of the neural crest also requires the down-regulation of specific groups of genes , for example to repress a neural fate . Cluster five ( Fig 5F ) contains a collection of soxB family genes ( sox3 , sox19a , sox19b , sox21b ) , one of which ( sox19 ) is one of the earliest CNS markers in vertebrates [47] . Cluster five also includes another example of paralogues of oct-related transcription factors pou3f2b ( Oct-2 ) and pou3f3a , which are associated with controlling CNS development . Cluster 29 ( Fig 5G ) contains a collection of genes ( pax7a , eng2b , mapk12b and enfa2a ) which , based on the midbrain/hindbrain expression patterns of pax7a and eng2b , also suggests a developmental CNS role . All gene lists of individual clusters , along with GO and ZFA enrichments , can be found in ( Table 1 ) . Using many replicates of single , genotyped embryos from the same clutch has allowed us to show how a single allele of tfap2c is sufficient to maintain a minimal neural crest GRN . Based on this we have compiled functional subsets of maintained genes , many of which are still poorly described and previously have never been associated with the neural crest . We have identified multiple cases where gene families or paralogues behave in the same manner , highlighting more potential examples of the compensatory nature of the GRN in general . To validate the association of novel genes with neural crest biology , we next analysed a set of candidates using a knockout approach . Our transcriptional profiles from mutants and FACS enriched neural crest cells contain a large number of novel neural crest candidate genes with poor or no functional annotation ( Table 1 ) . We chose a subset of these for further analysis based on lack of functional annotation and their differential expression across the different datasets ( Table 2 ) . To validate their involvement in neural crest biology , we analysed the expression patterns or screened for knockout phenotypes in zebrafish embryos . For example , transcripts of the gene wu:fc46h12; ENSDARG00000114516 were strongly reduced in several sox10 mutant experiments ( Table 1 ) . At 24 hpf wu:fc46h12 has an identical expression pattern to the xanthophore marker gch2 in wild-type and sox10 mutants ( S5A , S5D , S5G and S5H Fig ) , but diverges at 48 hpf in wild types as wu:fc46h12’s expression domain becomes more specific to a ventral crest population ( S5E and S5F Fig ) , heart and dorsal aorta ( S5E’ and S5F’ Fig ) . The majority of these expression domains are also lost at 48 hpf in sox10 mutants ( S5G and S5H Fig ) . A CRISPR/Cas9 knockout allele , wu:fc46h12sa30572 , was homozygous viable . Maternal-zygotic ( MZ ) mutant wu:fc46h12sa30572 embryos from intercrosses of homozygous females with heterozygous males showed heart oedema at 36 hpf ( S5I and S5J Fig ) . By 5 dpf most larvae formed swim bladders and had grossly normal hearts . A more detailed analysis will be required to ascertain the role of wu:fc46h12 in heart development . Two genes , akr1b1 and cax1 , were differentially expressed in the tfap2a;tfap2c and sox10 mutant data sets ( Table 1 ) . The former , akr1b1 , is ubiquitously expressed [47] and , using CRISPR/Cas9 , we created a premature stop , akr1b1sa30579 . Homozygous akr1b1sa30579 fish develop normally but presented pale xanthophores ( S5K Fig ) . A premature stop in cax1 was already available from the Zebrafish Mutation Project . A previous report has shown expression in xanthophores and upon morpholino knockdown a reduction in neural crest tissues of the jaw as well as xanthophores [62] . While zygotic cax1sa10712 homozygotes present a dulling in the colouring of xanthophores ( S5L Fig ) and rounding up of the normally highly dendritic cells , they lack obvious craniofacial phenotypes . Homozygous cax1sa10712 adults are viable and fertile , but MZcax1sa10712 embryos appear ventralised at somitogenesis pointing to a role for cax1 during early embryonic development ( S5M Fig ) . Expression of the transcriptional regulator yap1 was reduced in double homozygous embryos in our 4 somite tfap2a;tfap2c 3’ tag sequencing data ( Table 1 ) and yap1 was also enriched in neural crest FACS-sorted cells ( Fig 2E ) . We also found that three members of the Hippo signalling pathway , fat2 , lats2 and yap1 , had significant negative log2 fold changes in the 15 somite tfap2a;tfap2c knockout versus wild type RNA-seq dataset ( Fig 6A ) . Furthermore , subset three from the tfap2a/tfap2c analysis was enriched for genes with a Tead2 binding site , an interactor of Yap/Taz transcription factors . These data suggest a very early role for Hippo signalling in neural crest cells . Previous work has identified Hippo signalling as a coactivator of Pax3 , involved in melanocyte gene expression [63] , and in vitro studies suggest that YAP is involved in neural crest fate and migration [64] . To investigate the role of yap1 in neural crest we targeted its first exon using CRISPR/Cas9 and created two alleles , yap1sa25458 and yap1sa25474 , leading to frame shifts and premature stops ( Fig 6B ) . When heterozygous carriers for either yap1sa25458 or yap1sa25474 were intercrossed and embryos raised at 28 . 5°C we found the previously described ocular phenotypes at 48–72 hpf in approximately 25% of embryos [65] , albeit with variable penetrance depending on incubator temperature . We therefore tested whether these two yap1 mutants were temperature sensitive by splitting a single clutch and raising the embryos at two different temperatures . When raised at 24°C , by 5 dpf just under a quarter ( 22 of 116 ) of larvae with normal morphology and a filled swim bladder were homozygous yap1 mutant . By contrast , when raised at 31 . 5°C none of the larvae with a swim bladder ( 108 ) were homozygous mutant for yap1 leaving a 1:2 ratio of 38 wild-type and 70 heterozygous fish ( Fig 6C ) . We raised intercrosses of yap1 carriers for each allele ( yap1sa25458 & yap1sa25474 ) at a permissive temperature of 24°C until 5 dpf then transferred them to our standard fish nursery to test for adult viability of homozygotes . At two months post fertilisation , a subset of these fish seemed smaller than their siblings ( S6A Fig ) . We measured and genotyped intercrosses from both yap1 alleles and confirmed that yap1 homozygous fish were viable , but smaller than their wild-type siblings ( S6B Fig ) . Although zygotic yap1 mutants did not display obvious morphological phenotypes in neural crest cell types , we investigated potential neural crest GRN changes . We intercrossed yap1+/sa25458 carriers , raised them at 28 . 5°C and collected embryos for 3’ tag sequencing at 4 somites , 15 somites and 24 hpf . The transcriptome profiles were normal at 4 somite and 15 somite stages , with the majority of the changed genes on the same chromosome as yap1 ( Fig 1J , Table 1 ) . However , at 24 hpf the early xanthophore pigment cell marker gch2 and wu:fc46h12 , the newly identified pigment marker described above , were significantly reduced in yap1 mutants ( Fig 6D ) . Interestingly , the early epidermis marker padi2 was also reduced ( Fig 6D ) . Previous studies have shown a role for yap1 in very early development of zebrafish and medaka [66–68] . In zebrafish , this precedes zygotic genome activation and thus highlights a role for maternally deposited transcripts . The developmental time course data of yap1 expression confirmed high levels of maternally deposited polyadenylated yap1 ( E-ERAD-475 , www . ebi . ac . uk/gxa/home/ ) . Given the maternal deposition of yap1 transcripts in the egg , we crossed heterozygous male yap1+/sa25458 carriers to homozygous female yap1sa25458 fish and evaluated the resulting MZyap1sa25458 larvae at the restrictive temperature of 31 . 5C . At approximately 30 hpf we observed a strong reduction in the number of melanocytes in roughly half of the embryos . The previously described ocular phenotype [65] was also apparent in addition to a mild pericardial oedema ( Fig 6E ) . These larvae are otherwise morphologically stage matched . To quantify the melanocyte reduction we counted melanocytes in four different sections ( head , yolk , ventral trunk and dorsal trunk ) of each larva and then genotyped them . A significant melanocyte reduction of about 50% in the yolk , ventral tail and dorsal tail was found , with no major difference in the number of melanocytes in the head ( Fig 6F ) . This demonstrates that maternally deposited mRNA is able to rescue a melanocyte phenotype at 30 hpf further highlighting the very early induction of the neural crest GRN .
The neural crest is typically described as being induced at the lateral edges of the neural plate after gastrulation . However , using wild-type developmental time course data we can place the activation of the neural crest transcription factors tfap2a , tfap2c and foxd3 at the Dome stage , which follows zygotic genome activation and precedes gastrulation . In zebrafish , simultaneous loss of tfap2a and foxd3 has been shown to genetically ablate the neural crest [70 , 71] with tfap2a and foxd3 expressed in mutually exclusive compartments of the embryo at the shield stage , midway through gastrulation . The overlap of these expression domains forms the presumptive neural crest [71] . In Xenopus laevis a high degree of overlap exists in the blastula pluripotent GRN and the neural crest GRN with the neural crest retaining the pluripotency of cells in the blastula stage rather than being induced later on in development [72] . Interestingly , the activation of the neural crest marker crestin also coincides with the Dome stage ( E-ERAD-475 , www . ebi . ac . uk/gxa/home/ ) . This suggests that the establishment of the proto-neural crest GRN comes shortly after zygotic genome activation and places its initiation much earlier than previously shown in zebrafish and other vertebrates . This also raises the possibility of maternal mRNAs playing a larger role than previously thought in early neural crest initiation . Nevertheless we do not dispute that the neural crest cell lineage is likely to pass through a well-defined ectodermal state as elegantly shown in single cell studies [73] . In addition to tfap2a;foxd3 loss of function , a combined knockout of tfap2a and tfap2c genetically ablates the neural crest in zebrafish [17 , 71] . In the case of tfap2a;foxd3 , tfap2a is thought to have an activator function whereas foxd3 has been shown to act both as a repressor and an activator [5] . Knockouts of tfap2a fail to form normal jaws and have reduced numbers of melanocytes , but still form neural crest cells . On a transcriptional level , using 3’ tag sequencing , the number of genes which are differentially abundant in the tfap2a or tfap2c mutants alone are modest , 39 and 5 genes respectively at the 15 somite stage ( Fig 2E ) . At the 4 somite stage tfap2c acts in a compensatory manner as its overall abundance is increased by almost 50% in tfap2a-/- embryos ( Fig 3B and Fig 5A ) . By removing a single wild-type tfap2c allele in tfap2a-/- embryos the number of changing genes jumps from 39 to 1152 ( Fig 2E ) , although this extensive change in gene expression is marked morphologically by only a mild decrease in the numbers of melanocytes in the tail at a much later stage . Using RNA-seq at the 15 somite stage increases the total numbers of changing genes detected but the general trends remain much the same . tfap2 family proteins are thought to form both homodimers as well as heterodimers [74] . This stepwise genetic ablation implies that tfap2c does not require tfap2a to initiate the early neural crest GRN and that either homodimers of tfap2c alone or potentially heterodimers with other tfap2 family members are sufficient; however , we do not see upregulation of any other tfap2 genes . Previous work in mouse , frog , chick and fish have identified the important role of AP-2 to initiate the neural crest and drive its downstream targets such as msx1 , sox9/10 and snail . [12 , 75 , 35 , 16] . Most previous analysis of AP-2 downstream targets have taken a gene by gene approach or , in the case of ChIP-seq , return thousands of putative targets . With our large-scale screen our intent is to identify the bulk of relevant AP-2 downstream targets and to make these available for further analysis . For example , Msx1 is a well-defined AP-2 target and our analysis shows transcripts for both paralogues , msx1a and msx1b , are less abundant in the tfap2a/c double knockouts , suggesting that a double knockout is necessary to abolish msx1 function . A further aspect is that AP-2 has been shown to play an important role in epidermis in addition to neural crest . Both of these tissues arise at similar time points from ectoderm , and it is therefore crucial to separate the neural crest from the ectoderm signal . By combining multiple mutant data sets over developmental time along with the neural crest FACS data set we could establish the timing of when different levels of the neural crest GRN begin . Along with a large number of known downstream targets the subsets contain many uncharacterised genes , suggesting a role for these in pigmentation . We can further group genes which are more likely to not be specific to the neural crest but rather involved in epidermis development ( Fig 2F ) . Using the overlaps across the three different time points we have classified groups of genes from an “early” role to “mid” and then “later . ” We appreciate that both tfap2a and tfap2c have broad expression domains and our attempts to separate out neural crest and epidermal-specific signals will have limitations . Due to the shared lineage of neural crest and dorsal epidermis cells , it is also possible that GRN overlap exists before the lineage splits . Single cell studies will be better able to address these issues . We have also further characterised trunk neural crest and melanocyte-specific downstream targets by analysing sox10 and mitfa knockouts . Similar approaches could be used in the future to address different neural crest sub lineages , for example , in sox9b mutants . The 15 somite stage had the highest number of differentially expressed genes in the tfap2a;tfap2c loss of function model and therefore we chose to investigate this stage in more detail using RNA-seq . Using different subsetting approaches we have characterised distinct groups of neural crest genes and also have identified the core neural crest GRN that is maintained via tfap2c . The hierarchical clustered heatmap ( S2E Fig ) highlights an enrichment of neural genes that are increased in the mutant samples . Considering the emerging model that neural crest cells are not actually induced in situ but rather a refinement of pluripotent blastula cells [72] , our data support the notion that not only is the activation of the neural crest GRN important but also the repression of non-neural crest-specific GRNs . Both our 3’ tag sequencing time course and RNA-seq reveal a great disparity between the severity of molecular phenotypes and morphological phenotypes . This data set allows us to identify the genes required to maintain neural crest induction but also what levels of expression are sufficient . RNA-seq analysis of tfap2a;tfap2c knockouts and their siblings revealed an increase of tfap2c mRNA expression in tfap2a mutants at 15 somites . Although not addressed in this study , an interesting question now is: what is the molecular machinery which identifies the need for genetic compensation and how is it carried out ? We find that whereas a single allele of tfap2c is able to rescue the early morphological neural crest ablation phenotype the expression of a core set of downstream effectors cannot be restored to wild-type levels . This separates the morphological phenotype , and its secondary molecular effects , from the primary gene regulatory effect of tfap2 loss of function . We have used this differential behaviour of downstream targets to identify genes which tfap2c is able to return to wild-type levels or to only partially rescue from the tfap2a/c double knockout . This has confirmed known neural crest players but has also added new genes to the neural crest GRN . The genes in subsets three and five ( Fig 4C–4H ) represent a core set of 371 and 162 genes , respectively , of the neural crest GRN required for early neural crest initiation and are most likely to be of high developmental and evolutionary importance . An enrichment for AP-2 transcription factor binding sites in the partially rescued gene subset is consistent with the first tier nature of the genes . Humans are particularly susceptible to haploinsufficient mutations in a number of neural crest-specific genes , including sox10 , leading to Waardenburg syndrome or Hirschsprung disease , whereas this is less the case in zebrafish [76] . sox10+/- fish are adult viable and are phenotypically normal . Based on the developmental timing and clustering behaviour of the soxE family paralogues sox10 and sox9b , there is a good probability that these two genes are able to compensate for each other in early neural crest cells . Similarly , fish with mutations in tfap2c are homozygous viable and tfap2a+/-;tfap2c-/- fish are indistinguishable from their wild-type siblings . By contrast , alterations of TFAP2A , acting in a dominant negative manner , lead to a number of developmental phenotypes in humans . Phenotype-driven forward genetics screens [77 , 78] are very successful in identifying mutations affecting a multitude of processes across the zebrafish lifespan . By contrast , reverse genetics screens have demonstrated , against expectations , that many presumably protein-disrupting mutations fail to lead to an obvious morphological phenotype in the first five days zebrafish of development [39 , 79] . Although more sensitive screening assays across different life stages and conditions are required to identify more subtle phenotypes , a multi-gene loss of function approach may be required to counteract as of yet poorly characterised mechanisms of compensation . Here , using the neural crest as a model , we dissect the relationship between transcriptional robustness and morphological outcomes . Our study has also begun to reveal more evidence of genetic compensation in other paralogous genes . Unsupervised clustering has highlighted that entire gene families clustered together across development [42] and behaved in a similar manner in different genetic combinations in the tfap2a;tfap2c loss of function experiments ( Fig 5B–5E , Table 1 ) . Another example of possible paralogous compensation can be observed in the relatively mild developmental phenotypes of the yap1 knockouts . Recently double knockouts of yap1 and taz ( wwtr1 ) , its paralogue , have shown much stronger early developmental phenotypes and are embryonic lethal [67] . A deeper understanding of genetic and functional paralogues with respect to mutual compensation versus division of function will provide mechanistic insight into gene function evolution . We have identified a reduction in the abundance of some Hippo signalling members in both our 3’ tag sequencing and RNA-seq data sets . Previously , a role for Hippo signalling has been suggested in the neural crest using conditional mouse knockout models and in cell culture [80 , 63 , 64] . However , in the case of the mouse , complete yap1 knockouts are not viable and in human iPS neural crest cell models both YAP1 and TAZ ( WWTR ) require modulation . Yap1 itself is not capable of binding DNA but requires TEAD elements also identified in our studies ( Table 1 gene lists ) . A role for TEADs in both melanocytes and melanoma has been previously documented [81] and placing yap1 downstream of Ap2 signalling adds an interesting aspect to Hippo signalling in melanocytes . In zebrafish we show a role for maternally deposited yap1 in the differentiation of melanocytes , however the effect on other neural crest subtypes remains to be investigated . It is also likely that the yap1 paralogue taz could be playing a compensatory role . Furthermore , transcriptional analysis of non-phenotypic zygotic mutant embryos raised at permissive temperatures shows dis-regulation of several neural crest and epidermis genes . This is a further example of a transcriptional phenotype in the absence of morphological changes . Over the past few years post-embryonic neural crest stem cells have been identified in mouse and zebrafish [26 , 82 , 83] . The temperature sensitive yap1 signalling model described here allows for the conditional inactivation of Hippo signalling and could also be combined with taz heterozygous fish . This would allow for the investigation of Hippo-dependent processes in post-embryonic neural crest stem cells , but also growth , pattern formation and regeneration later in development and in adults . Taken together , we have used transcriptional profiling and stepwise genetic ablation of the neural crest to divide the neural crest GRN into temporal and functional units containing new candidate genes alongside well known factors . The analysis of paralogue compensation separates the morphological neural crest ablation phenotype from the first expression changes to the core tfap2 GRN . We confirm association of previously uncharacterised genes through knockout experiments and demonstrate a role for maternal transcripts in pigment cell development . Future studies of the functional gene clusters described here will help to further refine their role in neural crest development as well as their involvement in human genetic disorders and diseases such as neuroblastoma and melanoma .
Zebrafish were maintained in accordance with UK Home Office regulations , UK Animals ( Scientific Procedures ) Act 1986 , under project licences 80/2192 , 70/7606 and P597E5E82 . All animal work was reviewed by The Wellcome Trust Sanger Institute Ethical Review Committee . Zebrafish were maintained at 23 . 5°C on a 14 h light/10 h dark cycle . Male and female zebrafish from genotyped heterozygous fish carrying mutations were separated overnight before letting them spawn naturally the next day . Fertilised eggs were grown at 28°C and single or multi-allelic phenotyping was carried out as previously described [39 , 84] . The sox10t3 and sox10baz1 alleles were a gift from Robert Kelsh and mitfaw2 was previously a gift from Jim Lister [25 , 34] . Embryos were either morphologically sorted into phenotypically abnormal and normal ( sox10t3/baz1 collected at 28 hpf , 36 hpf and 48 hpf ) or collected blind at the stage of interest . Single embryos were placed individually into the wells of a 2 ml deep well block ( Axygen , Cat number P-DW-20-C-S ) , snap frozen on dry ice and then stored at -80°C . 22–23 hpf embryos were collected from the zebrafish transgenic sox10:mg line which labels neural crest nuclei with mCherry and crest cell membranes with GFP . We observed a delay in the membrane-bound GFP signal causing two separate neural crest populations; one labelled only with the nuclear mCherry marker , and a second labelled both with mCherry and the membrane-bound GFP ( Fig 1H ) . We sorted these two populations separately along with a third non-transgenic population for pairwise differential expression analysis , however for the purposes of this study we pooled the neural crest cell data . We also generated transcriptional profiles of cranial crest and trunk crest separately to capture lowly expressed genes specific to those cell types . We therefore separated heads and tails of embryos from the same stage and isolated individual cranial and trunk neural crest populations from each tissue comprising both mCherry+ and mCherry+/GFP+ cells as well as an unlabelled non-crest population . All cell populations were processed to produce polyA RNA-seq libraries and sequenced . Dissociated cells from about 30–50 larvae were collected for FACS as previously described ( Manoli et al . , 2012 ) . Briefly , embryos were dechorionated using 33 mg/ml pronase ( Sigma ) and pooled either as whole embryos or as pools of heads and tails . The yolks were removed using deyolking buffer ( 55 mM NaCl , 1 . 8 mM KCl , 1 . 25 mM NaHCO3 ) followed by digestion with trypsin-EDTA . Finally , the pellet was resuspended in FACSmax Cell Dissociation solution ( AMS Biotechnology ) and dissociated cells collected by passing the suspension through a 20 μm cell strainer ( Sysmex Partec ) . Using appropriate gating , dissociated cells were sorted into mCherry positive , mCherry and GFP positive and unlabelled non-crest cells on the BD INFLUX . The data was analysed using FlowJo . Sorted cells were collected and lysed in 110 μl of RLT buffer ( Qiagen ) containing 1 μl of 14 . 3M beta mercaptoethanol ( Sigma ) . The lysate was allowed to bind to 1 . 8 volumes of Agencourt RNAClean XP ( Beckman Coulter ) beads for 10 min and RNA was eluted from the beads as per the manufacturer’s instructions . Total RNA was converted into cDNA libraries using the SMART-Seq V4 Ultra Low Input RNA kit ( Clontech ) followed by Nextera DNA Library Prep kit ( Illumina ) as per manufacturer’s instructions . Libraries were pooled and sequenced on Illumina HiSeq 2000 in 75 bp paired-end mode . Frozen embryos were lysed in 100 μl RLT buffer ( Qiagen ) containing 1 μl of 14 . 3M beta mercaptoethanol ( Sigma ) . The lysate was allowed to bind to 1 . 8 volumes of Agencourt RNAClean XP ( Beckman Coulter ) beads for 10 min . The plate was then applied to a plate magnet ( Invitrogen ) until the solution cleared and the supernatant was removed without disturbing the beads . While still on the magnet the beads were washed three times with 70% ethanol and RNA was eluted from the beads as per the manufacturer’s instructions . RNA was quantified using either Qubit RNA HS assay or Quant-iT RNA assay ( Invitrogen ) . Genotyping was carried out as described previously [40] . Briefly , 1 μl of DNA from the extracted total nucleic acid was used to confirm the genotype of each sample using KASP SNP and InDel identification assays ( LGC group ) designed against our allele of interest . The genotyped plates were read on a plate reader ( Pherastar , BMG Labtech ) and 10–12 samples per genotype were selected for making libraries . DeTCT libraries were generated as described previously [50] . Briefly , 300 ng of RNA from each genotyped sample were DNAse treated , fragmented and bound to streptavidin beads . The 3’ ends of the fragmented RNA were pulled down using a biotinylated polyT primer . An RNA oligo containing the partial Illumina adapter 2 was ligated to the 5’ end of the bound fragment . The RNA fragment was eluted and reverse transcribed using an anchored oligo dT reverse transcriptase primer containing one of the 96 unique index sequences and part of the Illumina adapter 1 . The Illumina adapters were completed during a library amplification step and the libraries were quantified using either the BioPhotometer ( Eppendorf ) or Pherastar ( BMG Labtech ) . This was followed by size selection for an insert size of 70–270 bases . Equal quantities of libraries for each experiment were pooled , quantified by qPCR and sequenced on either HiSeq 2000 or HiSeq 2500 . Sequencing data were analysed as described previously [50] . Briefly , sequencing reads were processed with the DeTCT detag_fastq . pl script and aligned to the GRCz10 reference genome with BWA 0 . 5 . 10 . The resulting BAM files were processed using the DeTCT pipeline , which results in a list of regions representing 3’ ends , together with a count for each sample . These counts were used for differential expression analysis with DESeq2 on pairwise combinations of samples . Each region was associated with Ensembl 86 gene annotation based on the nearest transcript in the appropriate orientation . False positive 3’ ends , representing , for example , polyA-rich regions of the genome , were filtered using the DeTCT filter_output . pl script with the—strict option , reducing the number of 3’ ends from 439 , 367 to 53943 . Gene sets were analysed using topgo-wrapper for GO enrichment and Ontologizer for ZFA enrichment . Total nucleic acid was isolated from tfap2a+/sa24445;tfap2c+/sa18857 intercrosses at 15 somites . KASP genotyping was used to identify all nine possible genotypes . Total nucleic acid was treated with DNAseI ( NEB , Catalogue number M0303L ) and 10 replicates per genotype were processed . Ambion ERCC spike-in mix 2 ( Cat . No . 4456740 ) was added to 200 ng RNA according to the manufacturer’s instructions and sequencing libraries were prepared using the Illumina TruSeq Stranded mRNA Sample Prep Kit . Libraries were pooled and sequenced on Illumina HiSeq 2500 in 75 bp paired-end mode . Sequencing data were assessed using FastQC and aligned to the GRCz10 reference genome and Ensembl 86 transcriptome using TopHat2 . Read counts per gene were generated using htseq-count and used as input for pairwise differential expression analysis with DESeq2 . Gene sets were analysed using topgo-wrapper for GO enrichment and Ontologizer for ZFA enrichment . Custom R scripts were used for hierarchical clustering and principal component analysis . Transcription factor motif enrichment was performed using HOMER's findMotifs . pl tool ( v4 . 10 . 3 ) with default settings . The GRCz10 promoter set used was created with HOMER's updatePromoters . pl tool based on RefSeq genes from -2000 bp to 2000 bp relative to the TSS . Count data were clustered using Biolayout Express3D . Graph based network visualization with a Pearson correlation of above 0 . 7 and Markov clustering was carried out using Biolayout Express3D ( https://www . ebi . ac . uk/about/news/service-news/BioLayoutExpress3D ) . Markov clusters were visually inspected and extracted for display as a heatmap using the geneExpr ( https://github . com/richysix/geneExpr ) Shiny App ( Fig 5 ) . Genotyping of embryos and fin clips was performed as previously described [39 , 40] . A schematic of all genes with the positions of their respected mutations can be found in S6 Fig . Previously unpublished alleles used in this study are listed in Table 3 . Zebrafish anatomy enrichment is a similar approach to the widely used Gene Ontology enrichment but instead uses zebrafish anatomical terms linked to zebrafish genes . The enrichment is performed using ontologizer ( http://ontologizer . de/ ) with the ontology from http://ontologies . berkeleybop . org/zfa . obo . ZFIN gene IDs are linked to ZFA terms using http://zfin . org/downloads/phenoGeneCleanData_fish . txt and http://zfin . org/downloads/wildtype-expression_fish . txt and Ensembl IDs are converted to ZFIN IDs using Ensembl . Ontologizer is then run with the Parent-Child-Union calculation and Bonferroni multiple testing correction . RNA DIG-labelled probes were generated from cDNA libraries ( Transcriptor High Fidelity cDNA Synthesis Kit , Roche ) covering all relevant embryonic stages . PCR was performed and then TA cloned using TOPO-TA ( Invitrogen ) . RNA riboprobes were produced using the T7- and SP6-promoter sequence , enabling in vitro transcription of the plasmid using T7- and SP6-RNA polymerase ( Roche ) . All oligonucleotide sequences are listed here: wu:fc46h12_left1:CTGCTGACCTTCACCCTGATTCTG , wu:fc46h12_right1:GGTGTATTGCCTAAAACCCTCAGC wu:fc46h12_left2:ATTGCTGCTGACCTTCACCCTGAT , wu:fc46h12_right2:ATTGCCTAAAACCCTCAGCTTCCA . Creation and identification of CRISPR/Cas9 zebrafish alleles were conducted as previously described using the zebrafish codon optimised double NLS Cas9 [85 , 41] . | Embryonic development is an intricate process that requires genes to be active at the right time and place . Organisms have evolved mechanisms that ensure faithful execution of developmental programmes even if genes fail to function . For example , in a process called genetic compensation , one or more genes become activated in response to loss of function of another . In this work we use the zebrafish model to investigate how two related genes , tfap2a and tfap2c , interact to ensure establishment of the neural crest , a vertebrate-specific cell type that contributes to many different tissues . Losing tfap2a activity causes mild morphological defects and losing tfap2c has no visible effect . Yet when both are inactive , embryos are severely abnormal due to lack of neural crest-derived tissues . Here we show that loss of tfap2a triggers upregulation of tfap2c which prevents the loss of neural crest tissue . However , the genes normally regulated by tfap2a respond differently to tfap2c allowing us to identify the first tier of the Ap2 network and new players in neural crest biology . Our work demonstrates that the expression signature of partial , but morphologically sufficient , genetic compensation provides an opportunity to dissect gene regulatory networks . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"sequencing",
"techniques",
"fish",
"medicine",
"and",
"health",
"sciences",
"vertebrates",
"neuroscience",
"animals",
"somites",
"epithelial",
"cells",
"animal",
"models",
"osteichthyes",
"developmental",
"biology",
"model",
"organisms",
"organism",
"development",
"stem"... | 2019 | The gene regulatory basis of genetic compensation during neural crest induction |
The bacterial enzyme β-lactamase hydrolyzes the β-lactam ring of penicillin and chemically related antibiotics , rendering them ineffective . Due to rampant antibiotic overuse , the enzyme is evolving new resistance activities at an alarming rate . Related , the enzyme's global physiochemical properties exhibit various amounts of conservation and variability across the family . To that end , we characterize the extent of property conservation within twelve different class-A β-lactamases , and conclusively establish that the systematic variations therein parallel their evolutionary history . Large and systematic differences within electrostatic potential maps and pairwise residue-to-residue couplings are observed across the protein , which robustly reflect phylogenetic outgroups . Other properties are more conserved ( such as residue pKa values , electrostatic networks , and backbone flexibility ) , yet they also have systematic variations that parallel the phylogeny in a statistically significant way . Similarly , the above properties also parallel the environmental condition of the bacteria they are from in a statistically significant way . However , it is interesting and surprising that the only one of the global properties ( protein charge ) parallels the functional specificity patterns; meaning antibiotic resistance activities are not significantly constraining the global physiochemical properties . Rather , extended spectrum activities can emerge from the background of nearly any set of electrostatic and dynamic properties .
The bulk of our knowledge concerning protein family evolution has come from comparative analyses of the large body of sequence and/or structure data produced over the last five decades . While this data has been invaluable to our current understanding , sequence and static structural descriptions provide only a narrow glimpse into stability and functional mechanisms . Consequently , there has been a growing interest to include physiochemical and functional details into molecular-evolutionary analyses [1]–[3] . For a complete understanding of these relationships , both conservation and variation must be characterized . Since conservation of function is the ultimate evolutionary driving force [4] , protein orthologs tend to be significantly more similar in function than paralogs , and this functional similarity holds true with increasing sequence divergence as well [5] . Frequently , conserved functional patterns can be explained by conserved physiochemical properties [6] , [7] . The β-lactamase ( BL ) enzyme family provides an excellent mix of preserved and adaptable physiochemical properties that require evolutionary/functional relation interpretation . On the functional aspect , BL enzymes have a chemically diverse set of substrates . Moreover , many BL enzymes can act on the same substrate despite being from evolutionarily distinct outgroups , leading to questions related about the presence ( or absence ) of conserved mechanistic strategies . Antibiotic resistance continues to outpace our ability to bring new antibiotic drugs to market [8] , leading to substantive fears about our continued ability to combat bacterial infections that are currently relatively benign . Central to this growing global health concern is the bacterial enzyme BL , which is produced by some bacteria [9] . BL confers resistance to penicillin and related antibiotics by hydrolyzing their conserved 4-atom β-lactam moiety , thus destroying their antibiotic activity [10] . Bacteria of all species depend on a cross-linked peptidoglycan layer , which preserves cell shape and rigidity . This peptidoglycan layer is primarily composed of alternating β ( 1 , 4 ) -linked monosaccharides , specifically N-acetylglucosamine and N-acetylmuramic acid . The latter is modified by a pentapeptide that always ends with two D-alanine residues . Cross-linking of peptidoglycan units is catalyzed outside the cytoplasmic membrane by cell wall transpeptidase enzymes . In this cross-linking process , a peptide bond is formed between penultimate D-alanine on one chain and pimelic acid ( in Gram-negative ) or L-lysine ( in Gram-positive ) residue on the other . The terminal D-alanine is cleaved off after the linkage is formed with the penultimate residue . β-lactam antibiotics effectively inhibit bacterial transpeptidases , consequently they are often called penicillin binding proteins ( PBP ) . By inhibiting cell wall synthesis , the bacteria become highly susceptible to cell lysis . In response , bacteria have evolved BL enzymes to defend themselves against β-lactam antibiotics . BL has , in fact , evolved from the functional domain of PBP through the acquisition of the new hydrolase activity [11] . The BL enzyme family is broad and is characterized by varying degrees of antibiotic resistance activity . In fact , extended spectrum β-lactamases ( ESBL ) also confer resistance to cephalosporins , which had previously eluded BL hydrolysis [12] , [13] . ESBLs are evolved from traditional BL genes , generally through mutations within the active site [14] , [15] , thus highlighting the critical importance of subtle differences within members of the BL family . To date , more than 470 BL enzymes have been identified and are typically classified into 4 classes ( A to D ) based on sequence similarity [16] . Bush et al . developed a classification scheme for BL proteins based on their functional characteristics [17] . Protein structures belonging to classes A , C and D have similar folds and all have a mechanism that involves a catalytic serine residue , whereas class B enzymes are zinc metalloenzymes that have a distinct fold . In this work we focus on the most clinically relevant class-A family . Comparing a number of different electrostatic and dynamical global properties , we quantify the extent of conservation across the class-A BL family . Our dataset includes twelve structures , each originating from a different bacterial species . We show that – as expected – many of the global properties are qualitatively conserved ( such as residue pKa values , electrostatic networks , and backbone flexibility ) . Additionally , the local active site Ω-loop that is important for substrate recognition and catalysis is consistently established to be marginally rigid . However , some properties visually show large variance , and all properties have quantitative differences to varying degrees . In order to understand the origin of this variation , we quantitatively compare the differences within each property against the evolutionary clustering established by the family's phylogeny . Our results clearly establish that the systematic differences parallel the evolutionary patterns in a statistically significant way . To the best of our knowledge , this report establishes the most comprehensive and statistically robust relationship between physiochemical properties and evolutionary patterns . Going further , we also demonstrate the physiochemical properties parallel in a statically significant way the environmental condition of the bacteria they come from , which is not surprising since environmental segregation is likely related to divergence . Finally , we compare the same set of property differences to antibiotic specificity patterns . With the exception of enzyme charge , no correlations to antibiotic specificity are found , indicating that there is not a simple correspondence between global physiochemical properties and antibiotic specificity . This latter point is particularly alarming because it stresses that new antibiotic resistance patterns can emerge from a large fraction of the known BL enzymes through relatively small changes that do not significantly alter the global properties . Taken together , these results explain the variation within class-A BL physiochemical properties , while simultaneously suggesting new avenues of research regarding the plasticity within antibiotic resistance patterns .
Due to their clinical significance , serine-based class-A β-lactamase proteins are one of the most widely characterized enzyme families . The catalytic mechanism involves acylation of residue Ser-70 at the active site . However , identification of the general base that activates this serine residue has always been a subject of controversy . As such , two distinct residues have been proposed . While one hypothesis suggests that this role is played by the conserved Glu-166 [18]–[21] , the other proposes Lys-73 [22]–[24] . In support of the first hypothesis , crystallographic data and MD studies [21] have suggested the presence of a conserved bridging water molecule that might act as a relay molecule for the transfer of proton between Ser-70 and Glu-166 . Based on other experimental studies involving Glu-166 mutation [24] , [25] , the second hypothesis proposes an unsymmetrical mechanism involving two different general bases , Lys-73 and Glu-166 that carry out acylation and deacylation respectively . Swaren et . al . [26] have argued that substrate binding raises the pKa of Lys-73 , which contributes to lowering of energy barrier for Ser-70 , highlighting the importance of Lys-73 in proton transfer . Conversely , kinetic studies of several Glu-166 mutant enzymes [27] have displayed decreased rates of acylation and deacylation , emphasizing that Glu-166 is more important . Due to this absence of Glu-166 negative charge in mutant proteins , the Lys-73 side chain exhibits a lower pKa shift , acting as an alternate general base in hydrolyzing β-lactam ring [28] . Regardless of which hypothesis is correct , the above studies clearly highlight the importance of both Lys-73 and Glu-166 . Several other residues have also been identified in BL that are catalytically important: Ser-70 being the primary catalytic residue; Lys-73 , Glu-166 , Ser-130 , Lys-234 as secondary catalytic residues . Finally , Asn-136 , Arg-164 , Asp-179 are other important residues that maintain the active site structure ( Figure 1a ) . All of these residues are in spatial vicinity of Ser-70 and affect substrate recognition and catalysis . Detailed sequence and structural comparison across the class-A family has identified similar structural and functional elements that span over active site residues mentioned above [29]–[32] . These conserved elements are SxxK , SDN , ExxLN and KTG . Conservation of important electrostatic properties is a commonly employed mechanism that leads to conserved function [6] , [7] . Figure 2a shows calculated residue pKa shifts ( shifted away from their model values ) across twelve BL proteins . Interestingly , these pKa shifts are mostly conserved , emphasizing a common mechanistic strategy . We further investigate the site-site interactions of residues that have strong electrostatic interactions ( more than 1 kcal/mol ) with the secondary catalytic residues Lys-73 , Glu-166 and Lys-234 ( Figure 1c ) . Remarkably , all conserved electrostatic sites overlap with the four conserved element regions , highlighting the strength of the active site electrostatic forces . All pairwise active site interaction energies are listed in Table 1 . Further , all these sites have a conserved pKa shift . Asp-131 , Glu-166 , Asp-179 and Asp-233 display strong acidic character , whereas Lys-73 and Lys-234 exhibit conserved basic shift in pKa . Lys-73 , which acts as proton extractor from Ser-70 , needs to be deprotonated for acylation . As such , there is a cationic electrostatic microenvironment surrounding Lys-73 , which is created by nearby basic residues Lys-234 and Arg-244 [24] . When Arg-244 is missing ( which is the case in the NMC-A , MFO and G orthologs ) , this role is acquired by Arg-164 as shown in our active site electrostatic networks plot ( Figure 2b ) . Another important feature of BL proteins is the Ω-loop ( comprising of residues 163–178 ) that is involved in substrate recognition . Additionally , the Ω-loop comprises Glu-166 , which is critical for deacylation activity . Our results reveal a strongly conserved acidic behavior within Glu-166 , which activates a water molecule in the vicinity to attack carbonyl carbon of the acyl-enzyme . This ensures a back-delivery of the abstracted proton to Ser-70 γ-O atom , leading to enzyme regeneration [21] . The above results highlight the importance of conserved local electrostatic properties , whereas Figure 3 demonstrates that global electrostatic potential maps can be quite varied across the whole family . For example , the BS3 structure is primarily anionic , whereas the PC1 penicillinase is mostly cationic . Nevertheless , key features within the electrostatic potential maps are visually conserved within evolutionary outgroups . This point is exemplified by the TEM/SHV enzymes that have a conserved anionic patch spanning helices H3 , H4 , H6 and the Ω-loop ( cf . Figure 1b ) ; however , the patch typically missing from structures outside this outgroup . Similarly , other outgroups conserve visual electrostatic features , yet no potential map features visually align with antibiotic activity patterns . Differences within the electrostatic potential maps are not unexpected owing to the sequence and structural variability within the dataset . Pairwise sequence identities range from 27% to 98% , which translates to α-carbon root mean square differences up to 2 . 6 Å . Moreover , the net charge of these twelve enzymes ranges from −6 to +15 ( Table 2 ) . This large structural variation with distinct electrostatic properties raises the question , “How does nature maintain the common functionality of enzymes ? ” Key sequence/structure motifs provide an insight into the description of the underlying conservation . Sequence conserved regions SDN and KTG have a strictly conserved charge of −1 and +1 , respectively , across all twelve BL enzymes . Interestingly , the other two key regions , SxxK and ExxLN , which have variable sites x , are also strictly conserved with a charge of +1 and −2 , respectively . ExxLN lies within the 16-residue Ω-loop ( xRxExxLNxxxxxxxx ) that maintains an overall negative charge ( except PC1 ) ranging from −2 to −4 . The conserved electrostatic properties of key regions range from simple local conservation of charge to complex evolutionary origins of BLs . Conservation of charges at mutable motifs and Ω-loop are achieved through concerted mutations . When there is a charge changing mutation at these important electrostatic regions , there is a charge compensating mutation elsewhere . The preceding sections reveal a rich mixture of conservation and variability within pKa values of important residues , charge , and electrostatic potential maps . Moreover , even when properties are visually conserved , quantitative differences are almost always present . Can the propagation of these differences across the family be explained ? To answer that question , we perform statistical tests ( cf . Methods ) to elucidate the hidden relationships between enzyme sequence and physiochemical properties . Specifically , we test the statistical significance of pattern relationships between the global properties and the evolutionary outgroups defined from the BL phylogeny . That is , are differences within these global properties suppressed within phylogenetic outgroups relative to differences across multiple outgroups ? The answer is yes , we indeed find this to be the case for all electrostatic properties , including electrostatic network pairwise energies , electrostatic network composition , residue charge , and per residue pKa shifts ( cf . Table 3 ) . As such , these results conclusively establish that the observed variations within physiochemical properties , which can at times be extreme , are robustly defined by the phylogeny , thus indicating that variation within these global physiochemical properties is an evolutionary driving force underlying BL divergence . Conversely , variations observed with the local active site Ω-loop do not reflect the phylogenetic clustering because these properties are too conserved based on strict mechanistic requirements imposed on all BL enzymes . Similarly , we divided the BL family into four groups based on the environmental conditions of the bacteria they are from . First , we stratified the dataset based on whether the bacteria is aerobic or facultative anaerobic , and we additionally stratified based on whether the bacterium is gram-positive or gram-negative , which affects the locations of where the bacteria are likely to be found . For example , gram-positive bacteria tend to survive in dry conditions and are found in places like skin or in dust , whereas gram-negative bacteria thrive in aqueous conditions . As before with phylogenetic outgroups , we find that the variations within the electrostatic network pairwise energies , electrostatic network composition , residue charge , and per residue pKa shifts reflect environmental condition in a statistically significant way . Through congruence , the two sets of results clearly indicate that environmental condition has played an important role in BL evolution . This is not new [33] or surprising , but it does represent the first time that physiochemical properties were used to demonstrate the relationship between the two . One of the most attractive features of the BL system is that , in addition to the phylogeny , the family can also be clustered based on antibiotic specificity . Performing the same analyses a third time , but now based on the antibiotic specificity patterns , only one ( charge ) of the electrostatic properties reflects the physiochemical properties in a statistically meaningful way . This indicates that antibiotic specificity patterns are not confined to narrow property ranges , and that the considered properties do not drive the global divergence of the family . This result is interesting and surprising considering the common view that function is the ultimate evolutionary driving force . Moreover , from a public health point of view , this result is alarming because it highlights that new activities can emerge from any global property background . Put otherwise , new antibiotic resistance activities , including those found in ESBLs , are evolutionary easy to achieve because they come about through small changes that do not globally affect structure and the concomitant electrostatic properties ( electrostatic network pairwise energies , electrostatic network composition , residue charge , and per residue pKa shifts ) . The only physiochemical property that reflects the functional patterns in a statistically significant way is residue charge at pH = 7 . 0 , which is consistent with several prior works demonstrating the importance of charge-charge interactions within BL function and specificity . For example , Selzer et al [34] designed new BL proteins by altering surface charged residues that increase association rates . This change in biophysical property leads to changes in long-range electrostatic forces that may even change its functional specificity . Formation and breaking of ionic interactions in directed evolution experiments have also been exploited to design new proteins with distinct substrate activities [35] , [36] . Using the minimal Distance Constraint Model ( mDCM ) , which we have used to characterize dynamic properties across several different groups of related proteins [37]–[42] , we also characterize the extent of dynamical changes in BL . This work is particularly important because only a small number of class-A BL proteins have been studied by NMR and molecular dynamics simulation . As such , little is known about variation and conservation of dynamical properties across the BL protein family . Figure 4a displays the multiple sequence alignment of twelve BL proteins color-coded by flexibility index ( FI ) , which quantifies local flexibility along the protein backbone . Residues colored blue are rigid , whereas the ones colored red are flexible . Figure 4b quantifies the average FI across the complete dataset displaying average FI curve with +/−1 standard deviation . Positive FI values reflect the amount of excess degrees of freedom in flexible regions , and negative values reflect the amount of excess constraints in rigid regions . These results highlight two significant points . First , BL enzymes have a predominantly rigid backbone , and second , this backbone rigidity is conserved across the whole family . Normally , our calculations do not predict structures to be so rigid , but this prediction is consistent with NMR S2 order parameter descriptions [43] . The extent of rigidity is also visible at the N and C termini of BS3 , TEM-1 , SME-1 and SHV-2 . The flexibility/rigidity results of BL proteins presented in Figure 4a are rank ordered based on increasing average rigidity characteristics . Across the alignment , the secondary structure elements appear rigid , whereas intervening loops are flexible ( except the Ω-loop ) . Three flexible regions have been identified as shown in Figure 4c: flexible region 1 at helix H3 , flexible region 2 between H9 and H10 and finally flexible region 3 at H11 . While helix H10 is rigid , it is sandwiched between two flexible regions , meaning it could also have high mobility because the rigid body can “swing” from the flexible hinge in the same way a pendulum swings at a flexible pivot . We point this out because molecular dynamic studies have shown increased mobility in helix H10 upon substrate binding [44] . Mobility within the Ω-loop is thought to be important for substrate recognition and catalysis . Dynamic simulations performed in the past have suggested that the Ω-loop is rigid with order parameters comparable to other secondary structure elements [45] . The authors also illustrate the importance of flexibility at the tip of the Ω-loop , which is important for the opening and closing motion . Interestingly , mDCM results indicate that the Ω-loop is consistently isostatic , that is , marginally rigid along with eight active site residues ( cf . Figure 5 ) . As discussed above , the Ω-loop includes a key catalytic residue , Glu-166 , that performs the deacylation step . Furthermore , deletion of the Ω-loop makes the protein deacylation deficient resulting in the formation of stable acyl-enzyme complexes [46] . The marginally rigid Ω-loop suggests its catalytic importance where rigidity is important for reproducibility in substrate binding , yet also allowing for motion that might be functionally required . The Ω-loop region spans over three out of eight catalytic residues . Except Asn-136 , all catalytic residues exhibit similar isostatic nature even though they occur throughout the BL sequence . In stark contrast to the global variability observed across the BL dataset , the marginal rigidity and electrostatic properties of the active site region are conserved . In most cases , small increases in new activities can be directly attributed to only a handful of active site mutations that sterically allow new substrates to bind [10]; yet active site rigidity is maintained . In fact , this active site rigidity was recently utilized to develop new BL inhibitors using a fragment based drug design strategy [47] . These results support the view that steric and electrostatic complementarity between active site and different antibiotics are primarily responsible for BL resistance activities [48] . Note however that the CTX-M BL enzymes do show increased active site flexibility [49] , while maintaining active site geometrics consistent with the narrow spectrum TEM-1 and SHV-1 enzymes , thereby stressing their mechanistic plasticity within antibiotic resistance activities . Note that the CTX-M structures do not meet our structural quality criteria ( cf . Methods ) ; as such , they are not included in these analyses . In addition to backbone flexibility , the model also calculates a correlation metric called cooperativity correlation ( CC ) that describe pairwise mechanical couplings . As illustrated in Figure 6 , CC between a pair of residues in the native state can be rigidly correlated ( colored blue ) , flexibly correlated ( colored red ) , or uncorrelated ( colored white ) . Taken together , the full CC plot can help elucidate allosteric couplings within structure . In a previous investigation of periplasmic binding proteins [38] , the variability within the cooperativity correlation was explained by differences within the H-bond network . Interestingly , the H-bond network of BL proteins remains conserved ( discussed below ) , yet we observe substantial diversity and richness of CC throughout our dataset . In this way , the results presented here are much closer to our results with thioredoxin [40] , CheY [50] and lysozyme [42] that stress the sensitivity of CC , and thus allostery , to subtle structural perturbations . To further investigate this susceptibility within BL , we again layer the physical descriptions of structure onto the BL phylogeny . As with the electrostatic potential maps , CC properties again cluster in a way that reflects local evolutionary outgroups ( cf . Figure 6 ) . For example , TEM-1 , TEM-52 , SHV-1 and SHV-2 are largely composed of a single rigid cluster , which is consistent with earlier NMR [43] and MD [51] assessments of TEM-1 that indicated it is quite rigid . Carbapenemases SME-1 and NMC-A represent a close evolutionary pair , and thus have similar flexibility properties . Conversely , the L2 cephalosporinase , which belongs to a distinct outgroup , is atypically flexible . As before , the global FI and CC metrics also reflect the evolutionary and environmental condition patterns in a statistically significant way , but not the antibiotic specificities . Comparison of the two penicillinases within the dataset provides an illustrative example of how large the differences within the physiochemical properties can be , even among enzymes sharing antibiotic resistance activities . The backbone of penicillinase PC1 is the least rigid structure characterized , and it also has a very atypical cationic electrostatic surface . However , neither property is shared with penicillinase G . Its surface is primarily anionic and its backbone is significantly more rigid than the average structure . Significant CC differences between the pair are also observed . While large , the differences within the penicillinase pair are not outside ranges established by our whole dataset , especially considering the early evolutionary divergence between PC1 and G established by the phylogeny . Multiple attempts to relate electrostatic and rigidity relationships were made , but all were unsuccessful . Nevertheless , these results clearly demonstrate how systematic differences within electrostatic properties and CC parallel the overall phylogeny across BL enzyme family . Further , it is interesting to note how nature preserves the active site dynamics and their electrostatics properties during evolution . Conservation of function provides the selection bias for proteins to maintain globally similar dynamics while evolving to varying substrate recognition patterns . Table 2 describes the global H-bond statistics showing the number of H-bonds and average total energy across the twelve BL structures . Since the mDCM is in large part based on H-bond networks , it is critical to understand how their variation can affect dynamical properties . H-bond statistics show that the number of H-bonds varies from 495 to 559 , whereas the average H-bond energy ranges from −2 . 86 to −3 . 20 kcal/mol . In our previous studies we have noticed that the number of H-bonds can be trivially explained by the size of the protein [38] . However , due to their relatively constant size , no such correlation is observed here . We also find that the above variations do not trivially predict differences within backbone FI and CC . That is , structures with more H-bonds are not necessarily more rigid than those with fewer . As we have discussed previously [40] , [42] , this observation again stresses that topological considerations get lost in global metrics due to nonadditive nature of the mDCM , which has a considerable effect on the output . We employ a simple but effective approach for comparing H-bond networks by plotting the H-bond density per residue and the H-bond contact maps to visualize essential differences ( Figure 7 ) . There is a rich density of H-bonds at strand β1 , the Ω-loop and β9 , which are conserved throughout the family ( Figure 7a ) . An overlapped H-bond contact map of all the twelve BL structures gives us an insight of regions with strong H-bond interaction , where each pixel is color-coded by H-bond strength ( Figure 7b and 7c ) . The site labeled 1 shows strong interactions between three regions that extend over all key catalytic sites . Similarly , experimental studies [46] have highlighted the importance of strong interactions between ( i . ) Lys-73 and Glu-166 , ( ii . ) Arg-164 and Asp-179 , and ( iii . ) Asn-136 and Glu-166 . In those reports , the authors emphasize that removing any of these interactions can make the enzyme catalytically inefficient , while also disturbing its stability . Site 2 on the contact map highlights a strong interaction network within the Ω-loop region . Based on its location , the network is thus assumed to be important for maintaining functionality . Site 3 illustrates the presence of strong interactions between strands β1 and β9 , which is assumed related to structural stability . Furthermore , strong H-bond interactions are observed at secondary structures as expected . Another interesting observation is that sites 1 and 3 represent the two distinct BL domains ( as defined by SCOP ) , whereas site 2 is at the interface between the two domains and overlaps the active site , thus further stressing the importance of the Ω-loop region . The conserved regions of high H-bond density , which is most pronounced in secondary structure elements and the Ω-loop region , leads to conservation of backbone rigidity . However , local H-bond conservation does not necessarily indicate that their energies are equivalent , which could lead to the observed differences within backbone flexibility and cooperativity correlation . As such , we also compare the H-bond contact maps to the observed property differences to the evolutionary and antibiotic specificity patterns . As with most of the other physiochemical quantities , differences within the H-bond networks again reflect the evolutionary , but not antibiotic specificity patterns . In fact , the relationship between the H-bond contact map and the evolutionary patterns is the strongest relationship ( lowest p-value ) observed . A statistically significant relationship is also observed by looking at the H-bond density of each residue . The BL enzyme family represents an interesting case study in protein family evolution . While conservation of function is the primary driving force in the evolution of most protein families , rampant antibiotic overuse has introduced new pressures leading to new resistance activities that reflect subtle differences within substrate specificity . The bulk of these changes are trivially explained by steric changes within the BL active site [52]; however , it has never been determined if antibiotic specificities are related to global physiochemical properties . We clearly demonstrate that they are not . On the other hand , all of the global properties considered here vary in a systematic way that reflects the family's phylogeny . Physiochemical properties diverged early in the evolution of the family , leading to outgroups with conserved properties therein , and systematic differences between them . Related , stratifying the dataset based on the environmental condition the bacteria they are from also parallels the variations within the global physiochemical properties . Interestingly , differences within local properties at the Ω-loop region do not reflect either because variation is suppressed based on functional requirements . The differences and similarities within two pairs of class-A BLs encapsulate our results . First , consider the PC1 and G pair of penicillinases . The phylogeny clearly indicates that these proteins diverged early in the evolutionary history of the family , yet they have identical antibiotic specificities . In spite of the functional conservation , the evolutionary divergence has led to very different global physiochemical properties , which can be seen most starkly in the global electrostatic potential maps ( Figure 3 ) and cooperativity correlation plots ( Figure 6 ) . Conversely , MFO is from the same evolutionary outgroup as G , but they have vastly different specificities ( MFO has extended spectrum activities and can be classified as a cephalosporinase that can also hydrolyze monobactams ) . Despite substrate specificity differences , the electrostatic potential maps and cooperativity correlations plots are very similar as a consequence of their close evolutionary relationship . This point is particularly noteworthy and cautionary because it suggests that new antibiotic specificities , including extended spectrum activities , can emerge from the background of nearly any set of electrostatic and dynamic properties through local changes that do not significantly alter the global properties .
Additions of hydrogen atoms , residue pKa calculations and intramolecular electrostatic interactions have been performed on energy minimized protein structures using H++ web server [53] . Hydrogen atoms were added and their positions optimized ( MD based ) after calculating ionization states of the titratable residues using Poisson-Boltzmann continuum electrostatics . The server uses MEAD suite of programs , and detailed information of the algorithm can be found here [53] . The salinity and pH conditions are kept consistent with the conditions used in the original DSC experiment , i . e . , 0 . 06M salt concentrations and pH 7 . 0; and a solvent dielectric constant of 80 and an interior protein dielectric of 6 . Residue acidity and basicity changes ( Figure 2a ) are calculated with respect to model pKa values from [54] . The . pqr file generated from H++ containing charge and radii information is fed into APBS [55] to generate electrostatic potential data . The protein is centered on a 65×97×65 grid . The electrostatic potential maps in Figure 3 are displayed at +/−1 . 0 kcal/mol . The DCM is fundamentally based on a free energy decomposition scheme that explicitly accounts for nonadditivity within entropic components [56] . Therein , macromolecular structure is described as an ensemble of network rigidity topological frameworks , where intramolecular interactions are modeled by distance constraints and vertices represent atomic positions . Interactions such as covalent bonds , hydrogen bonds , and local residue conformational states are modeled as a three-dimensional network ( or framework ) of distance constraints . Distance constraints restrict the amount of available degrees of freedom ( DOF ) between adjacent vertices , and each framework is used to describe a set of similar geometric conformations that share a common set of interactions . Distance constraints are associated with a component enthalpy and entropy , and the total enthalpy of a given framework is simply the sum over the set of distance constraints . The free energy of a given framework is calculated by: ( 1 ) where Nint is the number of different types of modeled interactions , ht and σt define enthalpy and entropy of a single distance constraint used to model interaction type t . Nt is the number of times interaction t occurs in a given framework , , and It is the number of independent constraints of type t , where , It is always less than or equal to Nt . However , entropy components are nonadditive due to correlations within the dynamics , thus simple sums result in drastic overestimations of the total entropy . Entropy components are additive only over the set of independent DOF [57] , [58] . The DCM employs efficient network rigidity graph algorithms [59]–[61] to quickly differentiate the independent and redundant constraints . Adding a constraint within a flexible region of the network removes a single DOF , whereas adding a constraint to a rigid region has no entropy affect because all DOF in that region have already been consumed . The network rigidity algorithm recursively adds distance constraints based on their order of entropy ( from smallest to largest ) , rigorously providing the lowest upper bound estimate of the total entropy [62] . Note that a given chemical interaction can be modeled by more than one constraint . For example , torsion force is modeled as one constraint , H-bonds and covalent bonds as five . The free energy of a given protein would simply be based upon the above calculation if thermal fluctuations did not occur . Hence , topological differences arise due to fluctuating interactions , which account for the forming and breaking of weak interactions at equilibrium . Covalent bonds are quenched , meaning they need not be parameterized since the set is uniform across the ensemble . In the mDCM , torsion angle forces are segregated into native and disordered states , and H-bonds can be present or not . Salt bridges are modeled as a special case of H-bonds . For BL , the number of microstates is astronomical ( ∼21850 ) ; as such , the process of solving the mDCM for proteins is based on heterogeneous mean field theory [62] . A free energy landscape is defined by order parameters that specify the number of H-bonds ( Nhb ) and native torsions ( Nnat ) within a given macrostate . The free energy of a given macrostate is given by the free energy functional: ( 2 ) where vnat and δnat correspond to the enthalpy and entropy associated with a native torsion . The corresponding values of vdis and δdis have been fixed in prior works [63] . The total H-bond energy , Uhb , is determined using a modified [50] empirical potential [64] , which the component entropy is linearly related to . When a H-bond breaks , there is an enthalpically compensating interaction with solvent that is described by usol . While not explicitly specified in Eq . 1 , the total conformational entropy , Sconf , is appropriately attenuated by the probability of a distance constraint to be independent to account for nonadditivity . The probability for a distance constraint to be independent is determined by Monte Carlo sampling of topological frameworks that satisfy the order parameters . The mixing entropy term , Smix , arises from the various combinations that can satisfy the order parameters . The hydrophobic interactions are indirectly included in the usol and vnat parameters as discussed in [37] , [65] , i . e . , H-bond formation implicitly accounts for the hydrophobic contacts . Critical to the work presented here , the mDCM provides a large number of mechanical descriptions of structure referred to as Quantitative Stability/Flexibility Relationship ( QSFR ) . Flexibility implies conformational diversity , whereas rigid regions are structurally conserved . The mechanical origins of flexibility and rigidity are directly linked to conformational entropy . Hence , these thermodynamic and mechanical quantities combine to define QSFR . To be precise , the free energy of a protein can be expressed as a function of global flexibility θ , where θ is equal to the average number of independent degrees of freedom divided by the total number of residues . As θ increases , the protein transitions from a folded state to an unfolded state . The ensemble averaged mechanistic or QSFR quantities of a protein are calculated using conformations in the native basin of the protein . Common QSFR metrics include the flexibility index ( FI ) and cooperativity correlation ( CC ) . FI is a local description of backbone dynamics . Positive FI values quantify the number of excess DOF within a region , whereas negative values quantify the number of redundant constraints ( Figure 4b ) . A region is said to be isostatically rigid ( meaning marginally rigid ) when FI = 0 . As described above , CC plots identify all pairwise residue-to-residue couplings across the structure . In both metrics , the presented values represent the Boltzmann-weighted average across the native structure free energy basin . The mDCM is parameterized by finding values of ( usol , vnat , δnat ) that best reproduces the experimental Cp data using simulated annealing method ( Figure 8 ) . We parameterize the model using the Cp curve from B . cereus [66] and the evolutionarily closest structure BS3 . Focusing on our group of twelve class-A BL proteins with well-conserved structures of the same function , we have transferred the three adjustable parameters obtained from above to all the other members , which is an approach we have used previously [39] , [50] . With this fixed parameterization , we have confirmed that mDCM correctly predicts all BL orthologs to have a single peak in Cp and a two state folding/unfolding transition in free energy . Apart from these twelve BLs , an attempt was made to calculate QSFR quantities on three other BL structures ( 1IYS , 1HZO , and 1E25 ) , but their free energy landscapes were not two-state , so they were excluded . This is not to say that having a continuous transition is necessarily wrong; the model has been shown to not give two-state behavior when its inappropriate ( i . e . , met-myoglobin ) [63] . Nevertheless , in the absence of external biophysical characterizations , which to the best of our knowledge do not exist , it is impossible to know if the atypical behavior is real or simply an artifact of pushing the parameterization too far , thus they were excluded . We have consistently demonstrated that while thermodynamic quantities ( i . e . , Tm ) are somewhat sensitive to parameterization and input structure resolution , the mechanical FI and CC quantities are mostly robust to parameter differences . Nevertheless , a single parameter set across the dataset , guarantees that QSFR differences only arise from structural differences . Also , results from our previous works [37] , [39]–[41] have demonstrated that QSFR properties are insensitive to parameterization , and have minimal influence on CC and FI values . As such , the conclusions regarding changes in QSFR properties are robust . In this study , twelve different class-A BL structures are investigated to provide a large evolutionary cross-section for detailed analysis [35] , [67]–[77] , while maintaining a feasible number for data and visual assessment . Our dataset is based on a set of high-resolution BL structures without any internal missing residues . The resolution and R-values of all structures are respectively less than or equal to 2 . 4 Å and 0 . 22 . As provided in Table 4 , three out of twelve structures exhibit penicillinase activity while the rest belong to one of the following classes: broad-spectrum , extended-spectrum , carbapenamase , cephalosporinase or carbenicillinase . Moreover , all enzymes are inhibited by clavulanic acid and their structures are remarkably similar; the pairwise α-carbon root mean square deviation ( RMSD ) ranges from 0 . 73 to 2 . 57 Å ( cf . Figure 9 ) . For expanding sequential coverage , we collect approximately 1100 sequences after searching through the nonredundant protein database using BLASTP [78] . The protein sequence culling algorithm PISCES [79] is employed to filter sequences at 98% mutual sequence identity cutoff . This reduced dataset , which also includes twelve class-A BL protein sequences , is further aligned by MUSCLE [80] followed by phylogenetic tree construction using maximum-likelihood , meaning the phylogenetic tree shown in Figures 3 and 6 is purely derived from sequence information . The twelve BL protein sequences span across the evolutionary tree , which provide a robust structural coverage as well . However , we arrange these twelve BL sequences independent of the larger set , using both sequence and structural information by Protein Align tool in MOE [81] , to achieve better visual comparison across our set . H-bond density for a residue i is defined as: ( 3 ) where is the hydrogen bond energy between residue i and j , and is the number of hydrogen bonds formed between residue i and j . The summation of energies divided by total number of hydrogen bonds provides hydrogen bond density at per residue level ( Figure 7a ) . The hydrogen bond network contact map , shown in Figure 7b , is an overlapped network of all twelve BL proteins . The residue positions on the network follow multiple sequence alignment as described above . As such , identical donor and acceptor residue pair positions across the dataset are achieved for easy visual network assessment of hydrogen bond energies . To determine the statistical significance of our results , we have developed a cluster matching score , S . In the case of the evolutionary patterns , the clusters are defined by the various colors in the phylogeny ( cf . Figure 3 ) , which were determined using a constant cut-level through the rectangular dendrogram . The matching score is calculated using the following equation: ( 4 ) where Rx , y is the correlation between a vector of physiochemical properties associated with protein x and y . The equation has been developed to evaluate both intra-cluster similarities ( all pairs i , j ) and inter-cluster variability ( all pairs i , k ) . That is , random data would not provide a good score because the intra-cluster on the left would be negligible , whereas the inter-cluster term on the right would be negligible if the data were well conserved throughout the family . Conversely , a perfect score , S = 66 , would occur when all Ri , j = 1 and Ri , k = 0 , which is defined by the number of protein pairs in the dataset . For example , in the case of FI , the vectors are defined by all values of FI for residues present across the 12 structures , meaning alignment positions with gaps are ignored . The same is done for the residue H-bond density quantity . Residue charge and residue pKa shifts are very similar , but the only difference is the vector length is only for the subset of titratable residues . In the case of the NxN H-bond network and CC plots , the vector length is N ( N-1 ) /2 . The electrostatic network is also N×N; however , N is a relatively small number based on the residues identified within the active site electrostatic network . Because the size of these vectors is small , we also wanted to consider an alternate Rand Measure ( RM ) [82] that only considers set identity , which we have with good results in alternate work [83] . In this case , the Rx , y correlations are simply replaced with RM , which also scales between zero and one , and gives nearly identical results . Statistical significance of the match scores is determined by comparing the real calculated score , Sreal , to an ensemble of random values , Sshuffle , where the cluster identities have been randomly shuffled . We perform a z-test on each of the property comparing Sreal to the Sshuffle distribution . The corresponding p-values are provided in Table 3 . Statistical significance is assumed if the p-value is less than 0 . 05 , meaning it is highly unlikely to obtain Sreal from the randomized distribution . This indicates that the matching between the physiochemical properties ( i . e . , intra-cluster conserved properties and systematic inter-cluster differences ) and evolutionary groups is statistically significant and represents true sequence/property relationships . Relationships with environmental condition and antibiotic specificity patterns are calculated in the exact same way . The clusters for the former are defined above , whereas the antibiotic specificity clusters are use the Bush and Jacoby classification scheme [17] , [84] ( cf . Table 4 ) . | Comparison of protein sequences and structures sharing function has become a well-established bioinformatics paradigm , leading to countless discoveries related to protein family sequence/structure/function relationships . However , sequence and structure alone provide only crude physiochemical descriptions , thus stressing the need for more sophisticated analyses . In this work , we determine how much dynamical and electrostatic properties vary across the β-lactamase enzyme family . Our results indicate that some properties are mostly conserved across the family , whereas others vary significantly despite the fact that all share the same high-level β-lactamase activity . Despite global variance in some metrics , systematic differences are frequently observed between evolutionary outgroups , indicating that physiochemical properties are simultaneously conserved and variable . As such , these results underscore the richness within physiochemical properties across a protein family and provide insight into how the variations came about . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] | [
"physics",
"statistical",
"mechanics",
"biology",
"computational",
"biology",
"biophysics"
] | 2013 | Variations within Class-A β-Lactamase Physiochemical Properties Reflect Evolutionary and Environmental Patterns, but not Antibiotic Specificity |
The unfolded protein response ( UPR ) is an intracellular signaling pathway that counteracts variable stresses that impair protein folding in the endoplasmic reticulum ( ER ) . As such , the UPR is thought to be a homeostat that finely tunes ER protein folding capacity and ER abundance according to need . The mechanism by which the ER stress sensor Ire1 is activated by unfolded proteins and the role that the ER chaperone protein BiP plays in Ire1 regulation have remained unclear . Here we show that the UPR matches its output to the magnitude of the stress by regulating the duration of Ire1 signaling . BiP binding to Ire1 serves to desensitize Ire1 to low levels of stress and promotes its deactivation when favorable folding conditions are restored to the ER . We propose that , mechanistically , BiP achieves these functions by sequestering inactive Ire1 molecules , thereby providing a barrier to oligomerization and activation , and a stabilizing interaction that facilitates de-oligomerization and deactivation . Thus BiP binding to or release from Ire1 is not instrumental for switching the UPR on and off as previously posed . By contrast , BiP provides a buffer for inactive Ire1 molecules that ensures an appropriate response to restore protein folding homeostasis to the ER by modulating the sensitivity and dynamics of Ire1 activity .
The secreted and membrane-spanning proteins that eukaryotic cells use to sense and respond to their environments and to communicate with other cells are functional only when they attain their proper three-dimensional structures . Folding of these proteins takes place in the endoplasmic reticulum ( ER ) , aided by molecular chaperones . Degradation pathways help to discard misfolded proteins . When cells experience environmental stresses , nutrient depletion , or certain differentiation cues , the ER folding and degradation machineries can become overwhelmed and the cell risks accumulating and secreting malfunctional and potentially harmful proteins [1] . Such conditions of ER stress activate the unfolded protein response ( UPR ) [2] , resulting in an expanded ER [3] , [4] and increased expression of genes encoding ER chaperones , ER associated degradation machinery , and other components of the secretory pathway [5] . As such , the UPR provides a feedback loop that helps cells maintain high fidelity in protein folding and assembly . The UPR plays a fundamental role in maintaining cellular homeostasis and is therefore at the center of many normal physiological responses and pathologies . For example , when the severity of ER stress exceeds the capacity of the UPR to restore homeostasis , mammalian cells commit to apoptosis [2] . Furthermore , the UPR is activated in many cancer cells [6] , [7] , [8] as well as during familial protein-folding and neurodegenerative diseases [9] , [10] . Deficiencies in UPR signaling can also lead to diabetes [11] . Thus , the UPR constitutes an important control module whose core signaling machinery , which is conserved from yeast to humans , proves critical for cell physiology . Misfolded secretory proteins accumulate in the ER lumen . The UPR is initiated in that compartment when the transmembrane sensor molecule Ire1 self-associates and activates its cytoplasmic endoribonuclease domain [12] , [13] , [14] , [15] . Activated Ire1 transmits the signal by removing a non-conventional intron from its mRNA substrates , HAC1 mRNA in yeast and XBP1 mRNA in metazoans , which upon subsequent ligation are translated to produce potent transcriptional activators of UPR target genes [16] , [17] , [18] . Since the Hac1 protein is short-lived ( half-life of ∼2 min ) [18] , [19] , Ire1 activity is the key determinant of the magnitude and duration of the UPR . Despite early clues for Ire1's role as a central UPR regulator , the mechanism by which it senses unfolded proteins remains disputed . One model proposes that Ire1 activity is mainly regulated by the ER-resident chaperone BiP ( Kar2 in yeast ) . In this model , BiP inhibits Ire1 activity by binding to it in the absence of stress . During stress , BiP is titrated away by unfolded proteins , leaving Ire1 free to oligomerize and activate . This model was suggested because immunoprecipitation experiments showed that Ire1 interacts with BiP in unstressed cells and dissociates from BiP under ER stress conditions [20] , [21] , [22] . Site directed mutagenesis of BiP yielded mutants that do not bind to Ire1 [23] , but since they failed to support growth when expressed as the only copy of BiP , they are difficult to interpret mechanistically in view of the many pleiotropic functions of BiP . By contrast , mutants of Ire1 lacking the juxtamembrane segment of its lumenal domain that is responsible for BiP binding retained regulation: mutant Ire1 was inactive in the absence of ER stress and activated in its presence [15] , [22] , [24] , [25] , thus suggesting that BiP release and rebinding are not causal for switching Ire1 on and off . An alternative model of Ire1 regulation postulates that unfolded proteins bind to the lumenal domain of Ire1 , triggering Ire1 self-association and activation of its cytoplasmic effector domains . Support for such activation of Ire1 by direct binding to unfolded proteins stems from structural studies of the Ire1 lumenal domain that revealed a putative peptide binding groove [24] . Mutational probing experiments demonstrated that the residues pointing into the groove are required for signaling [24] . Recently a hybrid , two-step model for UPR regulation has been proposed in which both BiP and unfolded proteins regulate Ire1: initial dissociation of BiP from Ire1 drives its oligomerization , while subsequent binding to unfolded proteins leads to its activation [15] . This model posits that BiP regulates Ire1 oligomerization , yet oligomerization is not sufficient for Ire1 activation . However , in vitro experiments demonstrated that the oligomerization state of the cytoplasmic domains of Ire1 determines the rate of enzymatic activity [12] . Thus , while genetic and biochemical analyses of the UPR have been immensely successful in elucidating many aspects of the UPR's unusual signal transduction mechanism , a coherent model of Ire1 regulation and the involvement of BiP has remained elusive . In this work , we study the UPR as a coordinated homeostatic system by carrying out measurements of the time dynamics of the pathway across a wide range of ER stress levels . Using population-based assays of UPR activity complemented with dynamic dose-resolved flow cytometry and a predictive computational model , we dissect the role of BiP in modulating the sensitivity and duration of the UPR . Specifically , by comparing the wild type UPR to a strain bearing a mutant version of Ire1 that lacks the UPR-specific BiP interaction motif , we show that BiP prevents Ire1 from activating in response to low levels of stress and that it aids in Ire1 deactivation once the stress has been alleviated . Using a single cell Ire1 FRET assay , we provide evidence suggesting that BiP performs these functions by sequestering inactive Ire1 molecules . By buffering Ire1 , BiP ensures that only appropriate levels of stress trigger the UPR and that the duration of UPR induction matches the magnitude of the stress . These data position BiP as a modulator of the dynamic properties of the UPR .
Most UPR studies to date have been carried out under saturating conditions , where induction of protein folding damage surpasses the homeostatic capacity of the UPR and hence remains unmitigated . To position the experimental system in a physiological regime where cells proliferate efficiently when the UPR functions adequately , we probed the response to depletion of the metabolite inositol [26] . In the absence of inositol in the growth media , Ire1 is required for cells to induce the expression of genes required for inositol synthesis as part of the UPR transcriptional program [27] . To monitor UPR induction dynamics following this stimulus , we depleted inositol in a yeast culture and assayed for Ire1 activity as reflected by the splicing of HAC1 mRNA observed on Northern blots ( Figure 1A , see Methods ) . After a lag phase—presumably the time required to exhaust residual inositol stores—HAC1 mRNA splicing reached a maximal level by 120 min , and then declined during an adaptation phase to recover near basal levels by 240 min . Population growth slowed during the induction phase but was restored upon recovery ( Figure S1A ) . Thus , the UPR indeed functions as a homeostat in response to inositol depletion: the lack of inositol triggers activation of the biosynthetic pathway via Ire1 , which initially overshoots and then settles at a new basal level that meets the cells' needs to grow under the new conditions . In this example , our detection of HAC1 mRNA splicing was not sensitive enough to detect a difference between the starting condition and the new basal level . However , blotting for the UPR target INO1 mRNA , which encodes inositol 1-phosphate synthase required for de novo inositol synthesis , demonstrated that the readjusted level at the 240 min time point was elevated compared to the un-induced system ( Figure 1A , right panel ) , as was the expression of a UPR reporter ( Figure S1B ) . To determine whether similar adaptation also occurs after Ire1 activation in response to other modes of UPR induction , we treated cells with DTT , a reducing agent that counteracts disulfide bond formation and thereby induces protein misfolding in the ER . Disulfide bonds are formed through a relay in which ER client proteins are initially oxidized by protein disulfide isomerase ( PDI ) . PDI is in turn oxidized by the FAD-dependent oxidase Ero1 , which is finally oxidized by molecular oxygen [28] . Both PDI and ERO1 are UPR target genes , but since Ero1 directly passes the electrons to molecular oxygen , its abundance limits oxidative capacity . Thus , we reasoned that for moderate amounts of DTT , UPR-mediated induction of ERO1 would compensate for the increased demand for oxidation , allowing Ire1 to deactivate . To test this , we treated cells with a range of DTT concentrations . Cells treated with 5 mM DTT no longer proliferated , indicating the presence of a maximal ER stress beyond which cells can no longer compensate effectively even in the presence of a maximally active UPR ( Figure 1B , black ) . By contrast , cells treated with 2 . 2 mM or 1 . 5 mM DTT continued to proliferate , albeit at rates decreased from control cells ( Figure 1B , purple and green ) . To investigate whether these growth phenotypes correlated with the activation and deactivation of the UPR , we monitored Ire1 activation by measuring HAC1 mRNA splicing as above ( Figure S2 ) . Consistent with the observed growth arrest , Ire1 activation was maximal and sustained in 5 mM DTT ( Figure 1C , black ) : HAC1 mRNA was spliced to its full extent 30 min after DTT addition and splicing was maintained at this high level for the duration of the experiment . By contrast , in cells treated with doses of 2 . 2 mM or 1 . 5 mM DTT , Ire1 deactivation occurred in 4 h and 2 h , respectively ( Figure 1C , blue and green ) . Therefore , under non-saturating DTT conditions , cells show the same transient Ire1 activity that characterized the response to inositol depletion . Furthermore , the duration of that transient response increased along with the magnitude of the stress . To ascertain that the Ire1 activation and deactivation phases are reflective of the regulation of UPR target genes , we measured the expression of a synthetic UPR-regulated GFP transcriptional reporter ( TR ) over time in cells treated with 1 . 5 or 2 . 2 mM DTT ( Figure 1D , E , see Methods ) . In these cells , the TR was induced to dose-dependent plateaus after a lag of approximately 30 min . The lag is consistent with the time required for transcription , translation , and GFP chromophore maturation , while the plateaus reflect the accumulation of the long-lived GFP reporter protein ( half-life >8 h ) . Induction of a natural UPR target promoter , ERO1 , closely matched the response from the synthetic TR ( Figure S3 ) . Therefore , the expression of UPR target genes at any given time is reflected by the rate of GFP production , rather than its abundance . When plotted as a function of the rate of GFP production ( dTR/dt; Figure 1E ) , the TR exhibited activation and deactivation phases at 1 . 5 and 2 . 2 mM DTT that mirrored the dynamics of upstream HAC1 mRNA splicing ( compare Figure 1C and 1E ) . Taken together , the data shown in Figure 1 indicate that under different inducing stimuli , the UPR undergoes induction and adaptation phases that are reflected in the transient splicing activity of its sensor Ire1 . Ire1 activity , in turn , is faithfully transmitted to the system's transcriptional output . To assess whether the activation and adaptation properties of Ire1 are dependent on BiP binding and dissociation , we expressed a mutant form of Ire1 , Ire1bipless , lacking a 51 amino acid segment ( Ire1Δ475–526 , GKSG ) that contains the BiP binding site ( see Methods , Tables 1 , 2 ) . While similar to the Ire1ΔV mutant described in [22] , Ire1bipless retains 10 amino acids defined in the crystal structure of the core lumenal domain [24] that were deleted in Ire1ΔV . As previously reported , wild type Ire1 associated with BiP in a co-immunoprecipitation assay in the absence of ER stress ( Figure 2A , B ) but the association diminished when cells were treated for 1 h with 5 mM DTT ( Figure 2A , B ) . By contrast , no change in the association of Ire1bipless and BiP was observed between stressed and unstressed cells ( Figure 2A , B ) . The residual binding of BiP to Ire1bipless is likely due to non-specific absorption of the notoriously sticky chaperone ( Figure 2A , B ) . As the amount does not change between UPR-induced and uninduced cells , this residual interaction does not reflect a physiologically important regulatory interaction . To determine whether the diminished association between Ire1bipless and BiP impacts Ire1 activation , we measured HAC1 mRNA splicing in wild type cells and cells expressing Ire1bipless grown in the presence and absence of 5mM DTT for 1 h ( Figure 2C ) . In both wild type and Ire1bipless cells , no detectable HAC1 mRNA was spliced in the absence of stress , and splicing was identically induced in the two strains after treatment with DTT . These data refute any model that poses modulation of the BiP•Ire1 association as the exclusive regulator of Ire1 activity . Next , we investigated the subcellular localization of Ire1bipless in the presence and absence of ER stress . In response to ER stress , wild type Ire1 oligomerizes in clusters in the ER membrane that appear as discrete foci in fluorescence microscopy images [14] , [15] . Similar to wild type GFP-tagged Ire1 , GFP-tagged Ire1bipless displayed cortical and perinuclear ER localization in the absence of stress and formed bright foci in cells treated for 1 h with 5 mM DTT ( Figure 2D ) . Quantification revealed that Ire1bipless formed foci of equal magnitude to the wild type protein upon UPR induction . In unstressed cells , however , Ire1bipless displayed a 2-fold increase in the level of clustering compared to wild type Ire1 ( Figure 2E ) , and the foci exhibited considerable cell-to-cell variability ( Figure S4 , see Discussion ) . The increased clustering of Ire1bipless did not apparently lead to activation , since a Northern blot of total RNA from cells bearing Ire1bipless did not show detectable amounts of spliced HAC1 mRNA in the absence of stress ( Figure 2C ) . We considered it possible that splicing occurred at a level below the detection limit of the Northern blot assay . This reasoning is supported by Northern blots for INO1 mRNA , which is a more sensitive indicator of UPR induction as demonstrated above ( Figure 1A , right ) . Indeed , INO1 mRNA was significantly elevated in cells expressing Ire1bipless as compared to cells expressing wild type Ire1 under non-inducing conditions ( Figure 2F ) . Furthermore , there is a notable increase in the basal signal from a UPR reporter in unstressed Ire1bipless cells ( Figure S5 ) . Thus , UPR signaling in Ire1bipless cells is leaky . The propensity of Ire1bipless to form small clusters in the absence of stress prompted us to ask if cells bearing Ire1bipless would be more sensitive than wild type to low levels of stress . To test this notion , we expressed a GFP splicing reporter ( SR ) , in which the first exon of the HAC1 open reading frame is replaced by GFP ( Figure S6A ) . The HAC1 intron represses translation of the mRNA , so GFP is only produced once active Ire1 removes the intron . Using flow cytometry , the SR allowed us to precisely quantify Ire1 activity over time in wild type and Ire1bipless cells . The SR did not compete with endogenous HAC1 mRNA for Ire1 when wild type cells were treated with 5 mM DTT for 1 h ( Figure S6B ) , and similar to the TR , the GFP encoded by the SR decayed with a half-life of >8 h . When wild type cells expressing the SR were treated with increasing concentrations of DTT , the SR was induced to dose-dependent plateaus ( Figure 3A ) , and the rate of GFP production displayed the peak and decline behavior characteristic of the splicing of endogenous HAC1 mRNA ( dSR/dt; Figure S7A ) . Consistent with the data shown in Figure 1 , cells expressing wild type Ire1 were insensitive to DTT at concentrations below 1 . 5 mM as apparent from the absence of SR induction . By contrast , hac1Δ cells were hypersensitive to DTT: they induced the SR to near maximal levels at all doses ( Figure 3B ) , and the rate of GFP production remained high until the reporter saturated ( Figure S7B ) . In the absence of HAC1 , Ire1 activation fails to initiate a transcriptional response , and the stress is never alleviated . Interestingly , Ire1bipless cells showed an intermediate SR phenotype . Ire1bipless cells were more sensitive to DTT than wild type cells , becoming activated at 0 . 66 mM DTT and saturated at 1 . 5 mM DTT ( Figures 3C , S7C ) . These data are consistent with the notion that increased clustering in Ire1bipless cells in the absence of DTT is coupled with sensitization , which allows activation at low levels of stress . To validate that our data are consistent with a model of Ire1 regulation that includes interactions with unfolded proteins and BiP and to provide hypotheses for how BiP could specifically contribute to Ire1 regulation , we built a computational model of the UPR with the following assumptions ( see Text S1 ) . Ire1 can exist in one of three states: ( i ) as a free inactive monomer , ( ii ) as an inactive complex bound to BiP , or ( iii ) as an active complex bound to an unfolded protein ( Figure 4A ) . Further , free BiP can bind to unfolded proteins and either productively aid in their folding or nonproductively dissociate . Unfolded proteins are either reduced or oxidized depending on the redox potential of the ER and must be oxidized in order to fold . In the model , the redox potential is set by the ratio of DTT to Ero1 . When bound to an unfolded protein , the active Ire1 complex initiates the production of the Hac1 transcription factor , which in turn increases the production of BiP and Ero1 to close the UPR feedback loop . To explicitly model the measured experimental output ( GFP fluorescence ) , the active Ire1 complex was set to trigger the production of a simulated SR in addition to producing Hac1 . We extracted available model parameters from the literature and fitted remaining parameters to a subset of the experimental data ( Figure S8 , see Supporting Information for details ) . Using this “wild type” model as a baseline for comparison , we generated a “hac1Δ” model in which no induced production of BiP or Ero1 exists and an “Ire1bipless” model in which the interaction between Ire1 and BiP is disabled ( Figure 4A ) . The functional form of the dissociation of the active Ire1/unfolded protein complex was a modeling choice . Significantly , a model in which this dissociation was assumed to be linear did not reproduce the difference between the wild type and Ire1bipless when the SR time courses were simulated ( Figure S9 ) . Instead , a nonlinear , cooperative dissociation function of the active Ire1-unfolded protein complex was required to recapitulate the data; i . e . , the dissociation rate of the active Ire1-unfolded protein complex must decrease in proportion to the concentration of the active oligomeric complex raised to a power greater than one . Given that Ire1 signals by clustering into foci , this nonlinear dissociation function can be thought of as a consequence of having to disassemble a cooperative enzyme complex ( Figure S10 , see Discussion ) . When simulated with such nonlinear dissociation of the active Ire1 complex , the model robustly recapitulated the DTT titration time course results in wild type , hac1Δ , and Ire1bipless cells ( Figure 3D–F ) . When the SR time course was simulated with the wild type Ire1 model , doses of DTT of 1 . 5 mM and below produced less than 10% activity , 2 . 2 mM DTT produced an approximately half-maximal response , 3 . 3 mM DTT produced a response of approximately 75% of the maximum , and 5 mM DTT produced a near saturating response ( Figure 3D ) . By contrast , simulation of the hac1Δ model produced near saturating responses to all doses , recapitulating the hypersensitivity measured in vivo ( Figure 3E ) . Furthermore , simulation of the Ire1bipless model yielded an intermediate phenotype in which 0 . 66 mM DTT produced 15% activity , and doses of 1 . 5 mM DTT and above saturated the response ( Figure 3F ) . Importantly , this agreement between the model simulations and experimental data was an emergent property of the functional interactions in the system , which arose independently of the choice of parameter values ( Figures S11 , S12 ) . In addition to accounting for the increased sensitivity of Ire1bipless compared to the wild type in the DTT titration time course experiments , our computational model predicted that Ire1bipless should exhibit delayed shutoff dynamics compared to the wild type after DTT is removed ( Figure 4B ) . This prediction can be rationalized in intuitive terms . When DTT is removed , disulfide bonds can form and proteins can mature . Thus the concentration of the ligand for Ire1 activation starts to decrease , and individual Ire1 molecules dissociate from the active oligomer . When wild type Ire1 dissociates , it can either rejoin the signaling complex ( through interaction with an unfolded protein ) , or it can bind to BiP . Therefore , Ire1 deactivation proceeds rapidly since the inactive free form can be sequestered away by binding to BiP . In contrast , Ire1bipless lacks the ability to interact with BiP . Thus , while DTT removal will still prompt the dissociation of Ire1 from the active oligomer as the concentration of unfolded proteins decreases , the inability of Ire1bipless to bind to BiP increases the probability that an inactive Ire1bipless monomer will be recaptured by an unfolded protein and reactivate . As a result , Ire1bipless deactivation would proceed more slowly than that of wild type Ire1 . To test this prediction experimentally , we performed a DTT washout experiment in which wild type and Ire1bipless cells were treated with 5 mM DTT for 1 h to fully activate Ire1 in both strains . Subsequently , DTT was removed by filtration , cells were washed and resuspended in fresh media , and samples were collected over time to assay for HAC1 mRNA splicing by Northern blot ( Figure 4C ) . Additional samples of wild type cells were collected to assay for the association of Ire1 and BiP by immunoprecipitation ( Figure S13 ) . Confirming the model predictions , we found that while Ire1 deactivated after 60 min in the wild type , Ire1bipless retained activity for 120 min . As expected , Ire1 deactivation correlated with re-association with BiP ( Figure S13 ) . These results point to a role for BiP binding in promoting Ire1 deactivation once stress has been alleviated . To pursue the mechanism through which Ire1 deactivation proceeds , we hypothesized that , since Ire1 signals through assemblies of high-order oligomers , BiP binding may sequester breakaway Ire1 monomers , therefore promoting de-oligomerization of active Ire1 complexes . If this were the case , Ire1bipless cells should exhibit slower disappearance of Ire1 oligomers than wild type cells upon removal of stress . To directly test this hypothesis , we co-expressed GFP- and mCherry-tagged versions of Ire1 or Ire1bipless and employed a microscopy-based fluorescence resonance energy transfer ( FRET ) assay [29] to quantify Ire1 self-association ( Figures 5A , S14 , see Methods ) . In an otherwise wild type scenario , the FRET signal displayed a broad dynamic range , from 0 . 01 a . u . ( s . e . m . = 0 . 02 , n = 36 ) in untreated cells in which the Ire1 fluorescence displayed a diffuse ER localization to 0 . 73 a . u . ( s . e . m . = 0 . 06 , n = 41 ) in cells treated with 5 mM DTT for 4 h , in which Ire1 is maximally clustered into foci ( Figure S6B ) . In Ire1bipless cells , the basal FRET signal in the absence of DTT was elevated to 0 . 17 a . u . ( s . e . m . = 0 . 09 , n = 53 ) , but the maximum FRET signal in the presence of DTT ( 0 . 71 a . u . , s . e . m . = 0 . 08 , n = 32 ) was comparable to wild type . As expected , wild type cells displayed transient increases in FRET signal that returned to baseline levels over the course of the experiment after treatment with 2 . 2 or 1 . 5 mM DTT ( Figure 5B , C ) . In contrast , Ire1bipless cells were sensitized and displayed transient increases in FRET signal only when treated with 0 . 66 mM or 0 . 99 mM DTT but showed persistent strong FRET signal when treated with 1 . 5 mM or 2 . 2 mM DTT . These data recapitulate the role of BiP in buffering the Ire1 to low levels of stress ( Figure 3 ) . To assess the role of BiP in the de-oligomerization of Ire1 , we performed a DTT washout experiment and measured Ire1 FRET over time in wild type and Ire1bipless cells ( Figure 5D , E ) . After treatment of both strains with 5 mM DTT for 1 h , we washed the cells in fresh media lacking DTT and imaged the cells over time . Consistent with the deactivation kinetics of wild type and Ire1bipless cells as measured by Northern blot , wild type Ire1 de-oligomerization proceeded rapidly and the FRET signal returned to baseline after 60 min . By contrast , the Ire1bipless FRET signal remained higher than basal levels at 120 min . Taken together , these data indicate that BiP binding to Ire1 contributes to the efficient de-oligomerization of active Ire1 complexes .
When cells experience protein folding stress in the ER , the UPR is activated to increase the ER's folding capacity . For manageable stress magnitudes , the UPR is capable of restoring folding homeostasis . However , if the magnitude of the stress surpasses the capacity of the UPR , yeast cells sustain maximal Ire1 signaling and cease to proliferate ( Figure 1B , C ) . Within the physiological regime of ER stress , the response of Ire1 to moderate DTT inputs ( 1 . 5 mM and 2 . 2 mM DTT , Figure 1C ) displayed transient activation dynamics , followed by adaptation to near basal levels . Interestingly , the duration of Ire1 activity—not the maximal amplitude of its activity—correlated with the magnitude of the stress . Since the Hac1 transcription factor is short-lived , the length of the Ire1 activation pulse should determine the duration of UPR target gene activation by Hac1 [18] , [19] . This in turn determines the volume of the ER and the concentration of ER chaperones , components of the degradation machinery , and other cytoprotective proteins that are produced to combat the stress . This mode of signal regulation in which the duration of the output matches the magnitude of the input is known in engineering as “pulse-width modulation . ” It is widely employed to reduce noise in engineered control systems by transforming an analog signal ( amplitude ) into a digital all-or-none pulse of varying length [30] . Although in principle real-time information about the folding status of the cell could be conveyed exclusively through the interaction of unfolded proteins with Ire1 to determine the duration of UPR induction , we find that BiP plays an important role in modulating the length of the Ire1 activation pulse ( Figures S6A , C , 5B , C ) . Perhaps this modulating role of BiP reflects the necessity for precise tuning of the Ire1 pulse beyond what can be achieved through Ire1 and unfolded proteins alone . Interestingly , it was recently shown that a mutant of mammalian Ire1α shares salient properties with Ire1bipless: it does not bind to BiP , retains ER stress inducibility , and displays increased basal activity [31] . Therefore , it seems likely that the role of BiP in buffering Ire1 oligomerization is conserved in mammalian cells . Moreover , as the transmembrane kinase PERK , which in metazoan cells functions in a parallel UPR signaling branch to Ire1 , shares close sequence homology to Ire1's lumenal domain , lessons learned for Ire1 modulation by BiP are likely to also apply to PERK regulation . Precise tuning , and subsequently the buffering role of BiP , becomes all the more important since the UPR is linked to crucial cell fate decisions such as commitment to apoptosis [32] . The decision to commit to apoptosis might depend directly on the time of exposure to stress or on a thresholding mechanism through which either the extent of cellular damage or UPR machinery are assessed . Both scenarios would translate into an enhanced commitment to apoptosis in the absence of BiP modulation of Ire1 . As detailed above , precision homeostasis in the UPR requires the pathway-specific interaction of Ire1 and BiP . Disruption of this interaction in vivo leads to increased sensitivity to low levels of stress ( “leakiness” ) , coupled to slower deactivation of Ire1 once stress is removed ( Figure 4C ) . By using FRET to measure Ire1 self-association , we found that BiP performs these functions by aiding Ire1 de-oligomerization ( Figure 5C–E ) . In vitro , Ire1 functions as a cooperative enzyme with a Hill coefficient >8 , and the active species are large oligomers [12] . This high cooperativity could translate in vivo to a switch-like response of Ire1 to small changes in the concentration of unfolded proteins . For example , it follows from basic principles of enzyme kinetics that if Ire1 signals in clusters of 16 molecules , a mere 35% increase in unfolded proteins would cause Ire1 to go from 10% to 90% active . In this light , BiP's role as a binding partner that desensitizes Ire1 can be viewed as a gatekeeper that prevents triggering of the Ire1 activation switch following small or transient fluctuations in the local concentration of unfolded proteins . By doing so , BiP works to ensure that Ire1 is only activated when the stress is sufficient to warrant a response , thus improving information quality in the signaling pathway [33] . It is formally possible that in addition to loss of its UPR-specific BiP interaction Ire1bipless retains its ER-stress dependent activation , yet displays altered activation dynamics due to non-native conformational interactions . However , since Ire1bipless oligomerizes and activates in a ligand-specific manner to the same extent as wild type Ire1 , we contend that in the simplest scenario , Ire1bipless , like the previous “bipless” mutant Ire1ΔV [22] , [25] , is a structurally sound molecule that is activated by the same mechanism that activates wild type Ire1 . Though similar to Ire1bipless , Ire1ΔV was not shown to be hypersensitive to DTT or to deactivate after washout with delayed kinetics [22] . However , Ire1ΔV did display hypersensitivity to heat shock and delayed deactivation kinetics in response to ethanol [22] . While the discrepancies between Ire1bipless and Ire1ΔV may be due to differences in experimental resolution , the elevated response of Ire1ΔV to heat shock and ethanol is consistent with the notion that BiP buffers Ire1 to these mild ER stresses . Our study of the intricate UPR dynamics was guided by a computational model which was able to recapitulate our data and generate useful predictions . In the model , BiP serves as a buffer to the pool of inactive Ire1 . By binding to free Ire1 , BiP sequesters the inactive form of Ire1 and both prevents activation at low levels of stress and promotes deactivation once the stress has been overcome ( Figures 3D–F , 4B ) . This mechanism of Ire1 activation in our model contrasts with the two-step Ire1 activation model [15] , in which unfolded proteins first trigger BiP dissociation from Ire1 to induce oligomerization , and subsequently bind to the oligomers to activate signaling . As opposed to separating oligomerization and activation into two steps , our model treats unfolded protein binding as the single activating step; Ire1 is in dynamic equilibrium with BiP and unfolded proteins , and its unfolded protein bound state is active . Thus , BiP dissociation , rather than triggering oligomerization , yields monomeric Ire1 , which can then either bind to an unfolded protein and activate or re-bind to BiP . We note that the small Ire1bipless foci that formed in the absence of stress resulted in increased expression of INO1 mRNA and increased basal levels of UPR reporter fluorescence ( Figures 1A , S5 ) . Thus , we never observed inactive foci , in support of our model that oligomerization and activation occur in the same step . In addition to this different mechanism of Ire1 activation , our model also proposes a mechanism for Ire1 deactivation . Since BiP and unfolded proteins compete for Ire1 , BiP serves as a buffer that allows rapid deactivation of Ire1 as the concentration of unfolded proteins decreases . Finally , in contrast to the static picture of Ire1 activation presented in the two-step model , we present a time-resolved , quantitative model that accurately portrays Ire1 activation in response to any dose of DTT over time in its activation and adaptation phases . While the computational model reflects our current understanding of Ire1 regulation , it is likely to be an oversimplification . Next generation models could easily improve the verisimilitude by including additional ER processes that are not currently represented in the model ( such as glycosylation , ERAD , and BiP's ATP hydrolysis cycle ) or better constraining the model parameters by targeted measurements . Yet even with increasing mechanistic detail the requirement for cooperative Ire1 deactivation is likely to persist ( Figure S9 ) . This feature , modeled as decreasing Hill function of active Ire1 molecules , is consistent with the notion that Ire1 signals through assemblies of high-order oligomers . As Ire1 oligomers grow in size or number , the percentage of Ire1 molecules that have the ability to be deactivated decreases as many molecules become captured inside macromolecular assemblies . Such cooperativity in Ire1 deactivation can be depicted intuitively as a simple steric consequence of Ire1 oligomerization ( Figure S10 ) . Interestingly , this cooperativity can also be invoked to interpret the increased variability in foci formation in the Ire1bipless mutant cells ( Figures S4 and S15 ) . BiP's role can be thought of as a vehicle to help Ire1 traverse the threshold-like inactivation curve . In a wild type cell where focus formation might initiate stochastically , the presence of BiP can accelerate the dissociation of the foci . However , in an Ire1bipless mutant , any stochastically formed focus would be stable for a longer time ( Figure 5C–E ) . If focus dissolution is an all-or-none process , an extreme scenario is one where Ire1 focus formation in wild type and Ire1bipless cells occurs as a pulse train whose low frequency of activation is the same in both populations . However , the duration of each pulse would be longer in Ire1bipless than in wild type cells . This simplified scenario would result in modest differences in foci formation as averaged over the population since the activation probability is itself low . It would nonetheless result in large variations around this average exhibited by individual cells . According to this view , BiP buffering would ensure that activated Ire1 signaling centers assume a more homogeneous size , providing for a consistent input/output relationship and consistent deactivation kinetics . As such , BiP buffering fine-tunes the UPR by filtering noise from the signal transmission process , thereby increasing the information content of the signal and improving the cell's homeostatic control of the ER . This mode of regulation by which a free pool of a protein is buffered by chaperones may be a widely used mechanism in biology . For example , many kinases interact with cytosolic chaperones , and kinase signaling receptors that oligomerize during activation may hence be buffered similarly . Moreover , dynamic protein assemblies , such as clathrin coats or SNARE complexes , utilize chaperone interactions to aid disassembly [34] , [35] . Insights gained from our understanding of the functional consequences of the interaction between BiP and Ire1 may therefore be generally applicable to many other systems , in which protein oligomers have to form and be broken down again in a highly controlled manner .
Reporter constructs and mutant alleles are genomically integrated into wild type or mutant strains . All experiments were conducted in complete , synthetic media ( 2×SDC: yeast nitrogen base , glucose , complete amino acids ) . Ire1bipless is an allele of Ire1 that lacks the 51 amino acid juxtamembrane segment of the lumenal domain . This region is not in the crystal structure of the lumenal domain ( Credle et al . [24] ) . Amino acids 475–526 of Ire1 were removed by 2-step PCR cloning and replaced with a 4 amino acid linker ( Gly-Lys-Ser-Gly ) on an episomal yeast plasmid ( pRS315 ) . The resulting positive , sequenced clone ( Ire1bipless ) was sub-cloned onto integrative plasmids ( pRS305 , pRS306 ) , transformed into Ire1Δ cells ( YDP002 ) , and shown to complement for growth in the absence of inositol . Imaging constructs of Ire1bipless ( GFP- and mCherry-tagged ) were created by sub-cloning from the sequenced plasmid into the integrative wild type Ire1-GFP and Ire1-mCherry plasmids used for the FRET experiments . All experiments except the immunoprecipitations were conducted with genomically integrated Ire1bipless constructs . We cultured cells in 2×SDC media to OD600 = 0 . 4 , collected 50 ml per sample , washed cells in 1 ml 2×SDC and stored pellets at −80°C . Total RNA was extracted by resuspending cells in AE buffer ( 50 mM NaOAc , pH 5 . 2 , 10 mM EDTA in DEPC-treated water ) , adding SDS to 1% and acid phenol ( pH ∼4 ) ( Fisher ) to 50% , and heating at 65°C for 10 min . After spinning out the cell remains , we added chloroform and separated by centrifuging in phase-lock tubes ( 5 Prime ) . We precipitated the RNA with ethanol , washed with ethanol , and finally dissolved in 50 µl DEPC water . RNA samples were quantified by spectrophotometry , added to loading buffer ( 1×E/formamide/formaldehyde/ethidium bromide/bromphenol blue ) , and heated at 55°C for 15 min . Samples were cooled on ice for 5 min and loaded . The gel is 1 . 5% agarose/20% formaldehyde/1×E and is run for 270 min at 100 V . Gels were transferred to nitrocellulose by wicking in 10× SSC for 24 h , and RNA crosslinked with 150 J . Blots were pre-hybridized in Church buffer for 3 h at 65°C , and hybridized overnight with random primer-generated probes from a HAC1 PCR product that incorporated α-32P-CTP using GE ready-to-go beads . Blots were washed in 2× SSC , sealed in plastic , exposed to phosphor-imager screens overnight , imaged with the storm scanner , and quantified with ImageQuant software . We cultured cells bearing the SR or TR at 30°C in 2× SDC in 96 well deep well plates in an Innova plate shaker at 900 rpm . DTT stocks were made fresh from powder stored at 4°C for each experiment , and always 1 M in 10 ml . From this stock kept on ice , we prepared fresh 5× working stocks to start the experiment by diluting DTT in 1 step into 2× SDC to 37 . 5 mM ( 5×7 . 5 mM ) in 10 ml . This 37 . 5 mM working stock was serially diluted by 1 . 5-fold increments ( 6 ml + 3 ml SDC ) 10 times to span the range 0 . 13–7 . 5 mM . Every 2 h throughout the experiment , we repeated the full dilution series from the 1 M stock , making 1× dilution stocks in 2× SDC . To start the experiment , 200 µl of each 5× stock was added to 800 µl cells in the 96 well plates at time 0 . The cells were incubated and shook at 30°C and were sampled every 30 min by 12-channel pipetting 75 µl of each culture into a 96 well microtiter plate . 5 µl of each 75 µl was subjected to flow cytometry analysis using a BD LSR-II equipped with a high throughput sampler , a 488 nm 100 mW laser , FITC emission filter , and FACS DIVA software to compile . fcs files . . fcs files were analyzed in MatLab and/or FloJo . No cuts or gates were applied to cell distributions . Median FITC-A values were calculated for each dose-time point and plotted in ProFit . Errors are calculated from the standard deviation of the median for 3 biological replicates . We constructed the experimental FRET strain by co-expressing Ire1-GFP and Ire1-mCherry in the same cell from the endogenous IRE1 promoter integrated in the genome of an Ire1Δ strain and constructed bleed-through control strains by expressing either Ire1-GFP or Ire1-mCherry integrated alone in the deletion strain . FRET assays were performed using a Yokogawa CSU-22 spinning disc confocal on a Nikon TE-2000 inverted microscope equipped with 150 mW 488 and 562 nm lasers . Cells bearing the reporters were grown in 2× SDC to mid log phase , diluted to OD600 = 0 . 1 , gently sonicated , and 80 µl added to 96 well glass bottom plates coated with concanavalin-A . Cells were allowed to settle for 20 min before imaging . DTT dilutions were prepared as 5× working stocks as in the titration time course experiments , and 20 µl added to wells at time 0 . Cells were imaged at each time point with 3×3 s exposures: 488 excitation/590 emission ( GCh ) , 562 ex/590 em ( ChCh ) , 488 ex/520 em ( GG ) . Images were processed by first identifying cell boundaries and assigning the 16-bit fluorescence images to individual cells using the open-source cell-id software . Background was calculated by the mean intensity of areas in each fluorescent image not assigned to cells and subtracted from the cellular mean intensities to obtain corrected single cell values for GG , ChCh , and GCh . The GCh value is a conglomerate of true FRET signal and fluorescent channel bleed-through from the individual fluorophores . The average GCh values from the single-fluorophore control strains were subtracted from the experimental strain GCh values to obtain final corrected values . FRET was calculated for each cell with the formula: F = GCh/ ( GG*ChCh ) ∧0 . 5 . For each time point at each dose , we obtained images of three different fields of cells , collecting a total of 30–60 cells per dose per time point . Mean values were plotted in ProFit and error bars represent the standard error of the mean . Cells bearing C-terminally HA-tagged Ire1 or Ire1bipless expressed from the IRE1 promoter on 2 micron plasmids were cultured , collected , and stored in the same manner as for the Northern blot analysis . Cell pellets were thawed on ice , resuspended in 1 ml IP buffer ( 50 mM Tris-HCl , pH 8 . 0 , 150 mM NaCl , 1% Triton X-100 , protease inhibitors ) , and subjected to bead-beating ( 5×1 min , with 2 min on ice between iterations ) . Beads and cell debris were centrifuged and the cell free lysate was incubated with anti-HA conjugated agarose beads for 2 h at 4°C . Beads were spun , washed 5× with 1 ml IP buffer , and boiled in SDS-PAGE loading buffer . Samples were run on BioRad ready-gels ( 4%–15% acrylamide , Tris/glycine/SDS ) for 90 min at 35 mA . The proteins were subsequently transferred to Millipore Immobilon PVDF membranes at 220 mA for 45 min . Blots were blocked in 1% casein in TBS ( 10 mM Tris , 150 mM NaCl ) for 30 min , followed by incubation with primary antibodies overnight . The rabbit polyclonal anti-Kar2 was used at 1∶5000 dilution , and the mouse anti-HA was used at 1∶2000 . The next morning , the blots were washed 3× for 10 min with TBS , and then incubated with Li-Cor fluorescently-coupled secondary antibodies , goat anti-mouse 680 and 800 , at 1∶10 , 000 dilution for 30 min . Blots were again washed 3× for 10 min with TBS , scanned with the Li-Cor infrared scanner , and processed with the Odyssey software package . Wild type and Ire1bipless were cultured to OD600 = 0 . 4 in 400 ml 2×SDC at 30° . Cultures were brought to 500 ml and treated with 5 mM DTT for 1 h . Cells were sampled , filtered onto nitrocellulose membranes with 1 µm pores , washed with 100 ml 2×SDC , and then resuspended in 500 ml 2×SDC and returned to 30° incubation and sampled as indicated . For the FRET washout experiment , 1 ml cultures were spun , washed , resuspended , and imaged . | Secreted and membrane-spanning proteins constitute one of every three proteins produced by a eukaryotic cell . Many of these proteins initially fold and assemble in the endoplasmic reticulum ( ER ) . A variety of physiological and environmental conditions can increase the demands on the ER , overwhelming the ER protein folding machinery . To restore homeostasis in response to ER stress , cells activate an intracellular signaling pathway called the unfolded protein response ( UPR ) that adjusts the folding capacity of the ER according to need . Its failure impairs cell viability and has been implicated in numerous disease states . In this study , we quantitatively interrogate the homeostatic capacity of the UPR . We arrive at a mechanistic model for how the ER stress sensor Ire1 cooperates with its binding partner BiP , a highly redundant ER chaperone , to fine-tune UPR activity . Moving between a predictive computational model and experiments , we show that BiP release from Ire1 is not the switch that activates Ire1; rather , BiP modulates Ire1 activation and deactivation dynamics . BiP binding to Ire1 and its dissociation in an ER stress-dependent manner buffers the system against mild stresses . Furthermore , BiP binding accelerates Ire1 deactivation when stress is removed . We conclude that BiP binding to Ire1 serves to fine-tune the dynamic behavior of the UPR by modulating its sensitivity and shutoff kinetics . This function of the interaction between Ire1 and BiP may be a general paradigm for other systems in which oligomer formation and disassembly must be finely regulated . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"biochemistry/protein",
"folding",
"biochemistry/biomacromolecule-ligand",
"interactions",
"biochemistry/macromolecular",
"assemblies",
"and",
"machines",
"biochemistry/theory",
"and",
"simulation",
"computational",
"biology/signaling",
"networks",
"computational",
"biology/systems",
... | 2010 | BiP Binding to the ER-Stress Sensor Ire1 Tunes the Homeostatic Behavior of the Unfolded Protein Response |
In the genome of the biotrophic plant pathogen Ustilago maydis , many of the genes coding for secreted protein effectors modulating virulence are arranged in gene clusters . The vast majority of these genes encode novel proteins whose expression is coupled to plant colonization . The largest of these gene clusters , cluster 19A , encodes 24 secreted effectors . Deletion of the entire cluster results in severe attenuation of virulence . Here we present the functional analysis of this genomic region . We show that a 19A deletion mutant behaves like an endophyte , i . e . is still able to colonize plants and complete the infection cycle . However , tumors , the most conspicuous symptoms of maize smut disease , are only rarely formed and fungal biomass in infected tissue is significantly reduced . The generation and analysis of strains carrying sub-deletions identified several genes significantly contributing to tumor formation after seedling infection . Another of the effectors could be linked specifically to anthocyanin induction in the infected tissue . As the individual contributions of these genes to tumor formation were small , we studied the response of maize plants to the whole cluster mutant as well as to several individual mutants by array analysis . This revealed distinct plant responses , demonstrating that the respective effectors have discrete plant targets . We propose that the analysis of plant responses to effector mutant strains that lack a strong virulence phenotype may be a general way to visualize differences in effector function .
U . maydis is a biotrophic fungal pathogen causing smut disease in maize . To cause disease , haploid cells of compatible mating type need to fuse on the plant surface and develop an infectious dikaryon [1] , [2] . Upon perception of appropriate surface cues [3] , the dikaryon differentiates non-melanized appressoria that penetrate plant cells directly , presumably aided by local secretion of lytic enzymes [4] . During penetration , the host plasma membrane invaginates and encases the fungal hyphae , a feature typical for biotrophs . This establishes an extended interaction interface for the exchange of signals and nutrients [5] , [6] . Initial intracellular biotrophic growth of U . maydis is followed by intercellular growth during later stages of the infection , concomitant with massive proliferation in mesodermal tissue close to the veins . At this developmental stage , huge fungal aggregates form in cavities between plant cells followed by differentiation of ornamented diploid spores [6] . Fungal proliferation coincides with plant cell enlargement and resumption of mitotic divisions [6] . U . maydis can infect and cause symptoms on all above ground maize organs , with the infection staying locally confined . This is in contrast to related smut fungi that show systemic spread throughout the plant but produce symptoms only in the male and female inflorescences [7] , [8] . During initial contact of U . maydis with the plant leaf and presumably triggered by fungal pathogen-associated molecular patterns ( PAMPs ) , a number of plant defense genes are induced . This upregulation disappears during fungal penetration , suggesting that these initial defense responses are actively suppressed by the fungus during the plant colonization stages [9] . Also , several genes associated with suppression of plant cell death are induced . One of these , the maize cystatin CC9 , has recently been functionally analyzed [10] . Silencing of CC9 enhanced maize defense gene expression and upon infection with U . maydis a hypersensitive response was observed . CC9 was shown to suppress apoplastic cysteine protease activity , illustrating that CC9 is a novel compatibility factor for the biotrophic interaction of maize with U . maydis [10] . After plant colonization the most dramatic transcriptional changes in the host affect hormone signaling , induction of antioxidants , secondary metabolism , as well as a block in the transition from a juvenile sink tissue to a mature , photosynthetically active source tissue normally observed during leaf establishment [9] . The latter supports the observation that U . maydis is able to colonize young meristematic maize tissue , but is unable to infect differentiated source tissue [11] . The changes in plant gene expression observed after host colonization are likely to be brought about by secreted fungal effector molecules . The genome of U . maydis encodes about 300 novel secreted effectors that are upregulated during plant colonization and largely lack known InterPro domains . Of these a significant percentage is arranged in gene clusters and deletion of entire clusters can have dramatic effects on virulence [12] , [13] . Genome comparisons with the related smut fungi Sporisorium reilianum and U . hordei revealed that the majority of secreted effectors also exist in these relatives . With respect to conservation , effectors fall in two classes: approximately 34% are highly conserved in all smut fungi sequenced so far [8] and the remainder are poorly conserved , reflecting the arms race with the host . Furthermore , in U . hordei the tight clustering of effector genes seen in U . maydis and S . reilianum is largely disrupted [7] , [8] . Effector genes in U . maydis are plant-induced [12] , and work by Skibbe et al . [14] has revealed that the expression of some effector genes is tissue specific , i . e . is different when U . maydis colonizes seedlings , adult leaves or tassel , and the need for the corresponding effectors may be restricted to the respective tissues . So far the function of only very few of the many novel effectors has been elucidated in the U . maydis/maize pathosystem . Pep1 , a conserved effector is needed for penetration [15] and affects plant defense responses by inhibiting apoplastic plant peroxidases [16] . Pit2 , another conserved effector affects host defense responses [17] through inhibition of apoplastic plant cysteine proteases [18] . Cmu1 is a secreted chorismate mutase that is taken up by plant cells and lowers salicylic acid ( SA ) levels in infected tissue through metabolic priming [19] . Here we describe the beginning of the functional analysis of cluster 19A , the largest effector gene cluster in U . maydis . In a previous study , the entire cluster 19A comprising 23 putative effector genes in a 40 kb genomic region was deleted [12] . Cluster 19A mutants were severely attenuated in virulence and except for some ligula swelling , the mutants rarely elicited tumor formation [12] . In this study , we map the most important effectors for seedling infection in this cluster , and show that most strains deleted for individual effector genes show only minor reductions in virulence but elicit distinct plant responses .
Cluster 19A was originally predicted to encode 23 secreted effectors [12] . The manual reannotation based on comparison with S . reilianum and U . hordei now predicts the presence of 24 effector genes , plus one gene related to a reverse transcriptase ( um05313 ) presumably originating from a retrotransposon and one pseudogene ( um05316 ) ( http://mips . helmholtz-muenchen . de/genre/proj/ustilago ) ( Figure 1A ) . Among the effector genes in cluster 19A , we detect five gene families based on amino acid sequence similarity ( Figure 1A , Figure 2 and Figure S1 ) . Um05294 , Um05295 , Um12302 , Um10553 and Um10554 display between 34–48% similarity at the amino acid sequence level ( Figure 1A ) . Another family ( Figure 1A ) comprises genes um05299 , um05300 and um05301 . The respective proteins show between 34–41% similarity . The two effector proteins encoded by the adjacent genes um05305 and um05306 display 30% amino acid similarity ( Figure 1A ) . A three gene family codes for Um05309 , Um05310 , Um05311 with 35–86% amino acid sequence similarity and the largest family is comprised of Um05312 , Um05314 , Um10557 , Um05317 , Um05318 and Um05319 with 33–51% amino acid similarity ( Figure 1A ) . Outside of cluster 19A the U . maydis genome does not contain paralogs to any of these gene families . The 24 effectors encoded by cluster 19A do not contain recognizable protein domains nor do they display a characteristic spacing of cysteine residues described for several other U . maydis effectors [13] . However , orthologs for most of these genes are found in the genomes of S . reilianum and U . hordei [7] , [8] ( http://mips . helmholtz-muenchen . de/genre/proj/sporisorium/ ) ( http://mips . helmholtz-muenchen . de/genre/proj/MUHDB/ ) ( Figure 2 ) . Published expression data for cluster 19A genes from different U . maydis infected tissues [12] , [14] are compiled in Figure 2 . These studies revealed that except for two genes where expression could not be detected , genes in cluster 19A are differentially induced when different plant organs are colonized ( Figure 2 ) . Furthermore , only three of the cluster 19A genes are downregulated when the central regulator for pathogenic development , the bE/bW complex , is switched off during biotrophic development [20] ( Figure 2 ) . This illustrates , that the individual cluster genes are plant induced but do not appear to be co-regulated . Given the strong virulence phenotype of cluster 19A mutants [12] , we examined whether the cluster 19A mutant is impaired in biotrophic growth . To this end , we compared biotrophic development of the mutant strain SG200Δ19A and the progenitor strain SG200 . To our surprise , SG200Δ19A formed appressoria on the plant surface , proliferated inside plant tissue and differentiated teliospores , at a late time point comparable to infections with SG200 . This explains , why spores had not been detected in the previous study [12] and illustrated that the 19A mutant could undergo biotrophic development and complete the life cycle ( Figure 1B ) . With respect to macroscopic symptoms , tumor formation was abolished and instead chlorotic and necrotic areas became apparent ( Figure 1C ) . In addition , anthocyanin accumulation was completely abolished in mutant infected tissue , while prominent anthocyanin stained regions were observed in leaves infected by SG200 ( Figure 1C ) . To identify the genes contributing to the phenotype of the cluster 19A mutant , a series of strains was generated carrying sub-deletions of cluster 19A . At first , we divided cluster 19A into two parts which we designate 19A-1 ( left half region of cluster 19A ) and 19A-2 ( right half region of cluster 19A ) ( Figure 1A ) and generated respective deletion mutants . When tested for phenotype after seedling infection in comparison to SG200 , SG200Δ19A-1 showed a dramatic reduction of tumor formation and loss of anthocyanin accumulation . The effects were comparable to infections with SG200Δ19A . Conversely , SG200Δ19A-2 was only weekly attenuated in virulence and was able to elicit anthocyanin accumulation ( Figure 1C and Figure 3 ) . To determine possible differences in the efficiency of plant colonization by SG200 and the different derived mutant strains , we analyzed fungal biomass in colonized tissue by quantitative real time PCR using U . maydis peptidylprolyl isomerase ( ppi ) and Z . mays glyceraldehyde 3-phosphate dehydrogenase ( GAPDH ) as reference genes for quantifying fungal and plant biomass , respectively [15] , [21] . At 2 days post infection ( dpi ) , fungal biomass of SG200Δ19A , SG200Δ19A-1 and SG200Δ19A-2 was comparable to the SG200 infection ( Figure 1D , left panel ) . At 6 dpi , however , fungal biomass of SG200Δ19A and SG200Δ19A-1 was lower compared to SG200 ( Figure 1D , right panel ) , suggesting growth defects of these mutant strains at this later time point where massive fungal proliferation inside tumors is observed in SG200 infected tissue . Fungal biomass of SG200Δ19A-2 lacking the right half of cluster 19A was not significantly different from SG200 ( Figure 1D ) , consistent with the weak effect of this deletion on virulence . These results indicated that the major effector genes responsible for the phenotype of the cluster 19A mutant after seedling infection must reside in the left half of cluster 19A . To determine the contribution to virulence of the 14 genes located in the left half of cluster 19A , we divided the region into four parts , 19A-1a , 19A-1b , 19A-1c and 19A-1d making sure that existing gene families were deleted simultaneously ( Figure 3A ) . Respective deletion mutants were generated and assayed for tumor formation and ability to elicit anthocyanin accumulation ( Figure 3B , C ) . Of these four mutants SG200Δ19A-1a and SG200Δ19A-1c did show a small reduction of tumor formation , but this was not statistically significant ( Figure 3B ) . The deletion of 19A-1b and 19A-1d significantly lowered tumor formation with 19A-1d showing the strongest effect ( Figure 3B ) . Anthocyanin accumulation was abolished by the deletion of 19A-1c and was unaffected in infections with the other sub-deletion mutants ( Figure 3C ) . A triple deletion generated by combining 19A-1b , 19A-1c and 19A-1d deletions resulted in a mutant strain ( SG200Δ19A-1bcd ) that was severely reduced in tumor formation and failed to induce anthocyanin , comparable to the deletion strain lacking the left half of cluster 19A ( SG200Δ19A-1 ) . This firmly establishes that region 19A-1a does not measurably contribute to virulence ( Figure 3B ) . Based on the finding that the um05318 gene in U . maydis cluster 19A has three paralogous genes at a syntenic position in cluster 19A of S . reilianum [7] ( Figure 2 ) that contribute to virulence in S . reilianum ( H . Ghareeb and J . Schirawski , personal communication ) , we decided to delete the two rightmost genes in U . maydis cluster 19A ( 19A-2e , Figure 3A ) . This was done also to rule out that the lack of a virulence phenotype when deleting the entire right half ( Δ19A-2 , Figure 3B ) is caused by balancing positive and negative effects of effector gene deletions on virulence [12] . The resulting strain designated SG200Δ19A-2e showed weakly attenuated tumor formation that was comparable to strain SG200Δ19A-2 deleted for the entire right half of the cluster and this was again not statistically relevant ( Figure 3B ) . To identify the major effector genes that contribute most strongly to the virulence phenotype of sub-clusters 19A-1b , 19A-1c , 19A-1d and 19A-2e , we initially generated overlapping sub-deletions and tested them for virulence ( not shown ) . For example , the cluster 19A-1d region was subdivided into a double deletion of um05305/um05306 and a double deletion of um05306/um10556 , respectively . These double mutants were tested for virulence and in this case only the double deletion of um05306/um10556 was affected in virulence , allowing the conclusion that um10556 is the responsible gene ( not shown ) . This was then followed up by single deletions of the genes identified in this approach ( Figure 4A ) . With respect to 19A-1b comprising five related genes designated tin1-1 to tin1-5 , we were unable to identify the respective major individual effector gene ( s ) , as the observed effects of further sub-deletions on tumor formation could no longer be assessed as being statistically relevant ( data not shown ) . This suggested that these five members of the same gene family ( Figure 2 ) contribute weakly but additively to virulence . Complementation experiments where all 5 genes were re-introduced in strain SG200Δ19A-1b revealed that the weak virulence phenotype of the 19A-1b deletion could be complemented ( Figure 4B ) . For the region 19A-1c the deletion of um05302 ( designated tin2 ) showed a comparable reduction in virulence to the deletion of the entire 19A-1c region . In addition , the deletion of tin2 abolished anthocyanin induction ( Figure 4B , C ) . The introduction of a single copy of tin2 into SG200Δtin2 complemented the weakly reduced tumor formation as well as anthocyanin accumulation ( Figure 4B , C ) . The single deletion of um10556 ( designated tin3 ) significantly affected tumor formation ( Figure 4B ) . Also in this case , tumor formation could be restored to a level comparable to SG200 by introducing tin3 in single copy ( Figure 4B ) . After infection , the single gene deletion mutant of um05318 ( designated tin4 ) showed a lower incidence of plants with stronger disease symptoms than SG200 infected plants ( Figure 4B ) . Upon complementation , this disease category was increased compared to SG200Δtin4 and more plants showed heavy tumors and were stunted or dead ( Figure 4B ) . This suggests a weak contribution of tin4 to tumor formation . With respect to um05319 ( designated tin5 ) , the single gene deletion had minor effects on virulence and the reintroduction of the gene did not significantly change the disease scores ( Figure 4B ) . The analysis of single effector gene mutants for cluster 19A revealed in general , that deletions of individual genes ( with the exception of tin3 ) had only minor or statistically non-substantial effects on virulence ( Figure 4B ) , suggesting that the strong virulence defect observed in the entire cluster 19A deletion is due to additive effects and/or concerted action of several effectors . To visualize this and to obtain evidence whether individual effectors target distinct plant processes , we decided to analyze the plant responses to infection by the 19A deletion strain as well as to several single effector gene mutants on the transcriptome level . Maize seedlings were infected by SG200Δ19A , SG200Δ19A-1b , SG200Δtin3 , SG200Δtin4 and SG200Δtin5 . RNA was extracted from infected plant material harvested at 4 dpi , a time point where the individual mutants should not differ in fungal biomass as assessed from the analysis of sub-deletions ( Figure 1D ) . Three biological replicates were prepared and analyzed by Affymetrix maize genome microarrays . For technical reasons , the expression data for plants infected with the SG200Δtin2 mutant strain could not be included in this comparative transcriptome analysis . Maize gene expression profiles of tissue infected with cluster mutant strains were compared to profiles of SG200 infected and mock-treated plants , which had been generated in our previous study on the transcriptional responses of maize to U . maydis and had been used as reference in the analysis of plant responses to pep1 and pit2 effector mutants [9] , [15] , [17] . RMA-normalized microarray data were then subjected to a one-way ANOVA and contrast gene lists were generated using a fold change of ±2 and a corrected p-value of 0 . 05 as cutoffs ( Table S1 ) . Expression of a set of 13 genes differentially regulated after infection with different mutant strains was subsequently analyzed by quantitative real-time PCR ( qRT-PCR ) with RNA from independently generated infected plant material , and this allowed validating the array results ( Figure S2 ) . Compared to SG200 infected samples , 1816 maize genes were differentially regulated in response to SG200Δ19A ( Table S1 ) . A hierarchical clustering of these 1816 genes was performed for the whole data set to visualize the relations between the transcriptional responses to the individual U . maydis strains ( Figure 5A ) . As expected , the maximal distance in gene expression was between SG200 infections and infections by the SG200Δ19A mutant , which caused only very weak disease symptoms and thus displayed the highest similarity to the mock-inoculated plants ( Figure 5A ) . On the other hand SG200Δ19A-1b infections showed highest similarity to SG200 infections , illustrating that the 5 tin1 effector genes have only a weak contribution to plant responses , which is in line with their weak effect on virulence . Profiles of plant responses to strains carrying tin gene deletions clearly discriminated SG200Δtin3 and SG200Δ19A-1b , while responses to SG200Δtin4 and SG200Δtin5 infections were not separated by the hierarchical clustering ( Figure 5A ) , indicating that similar responses were elicited by tin4 and tin5 mutants . We also combined the transcriptional responses elicited by each of the four individual tin mutant strains to be able to compare this to the response elicited by SG200Δ19A , the strain carrying the full deletion of cluster 19A and to reveal contributions of genes not deleted individually . In total , 1513 maize genes were differentially regulated by the four tin mutants compared to SG200 infections , while 1816 genes were differentially regulated after infection with the cluster deletion mutant SG200Δ19A ( Figure 5B ) . Interestingly , a comparison of these combined “responses to individual tin gene deletions” to the SG200Δ19A responsive genes showed only a partial overlap of differentially regulated transcripts ( Figure 5B ) . Of the 726 genes differentially expressed in response to the whole cluster mutant but not detected in the “responses to individual tin gene deletions” 352 genes were induced and 374 genes were repressed ( Figure 5B ) . Among the induced transcripts , particularly biotin synthesis genes were induced specifically in SG200Δ19A infected tissue , while plant cellulose synthesis genes were downregulated after infection by the Δ19A deletion strain ( Table S2 and Figures S3 and S4 ) . In addition , several anthocyanin biosynthesis related genes were downregulated after infection with the cluster 19A mutant while they were not included as differentially expressed in “responses to individual tin gene deletions” ( Table S2 ) . This most likely indicates a contribution of the Tin2 effector to anthocyanin induction , and reflects that plant responses to the tin2 deletion strain could not be included in “responses to individual tin gene deletions” for technical reasons . Of the 1090 “shared differentially regulated plant genes” , 288 genes were induced , while 802 genes were downregulated ( Figure 5B ) . Induced genes comprised pathogen response genes such as PR4 , PR5 and several oxidases , demonstrating an elevated plant defense in response to tin gene and cluster 19A mutant strains ( Table S2 ) . Downregulated transcripts were strongly enriched for genes involved in DNA-metabolism and DNA-modification , particularly histones and DNA-methyltransferases ( Table S2 and Figure S5 ) . This most likely reflects the reduced tumor formation observed in all the mutants compared to SG200 . On the other hand , 423 genes were differentially regulated after infections with the tin mutant strains ( Figure 5B ) but these were not differentially regulated after infection with the cluster 19A mutant . Among these 423 genes , several chitinases and peroxidases were found ( Table S2 ) . To get clues on the possible roles of the tin genes during host colonization we next visualized the differentially regulated plant genes in response to individual tin mutants . This analysis revealed that 476 maize genes were commonly regulated by all four mutants , while 1027 genes only responded to a subset of tin mutants ( Figure 5C ) . The smallest number of specifically regulated maize genes was found after infections with the SG200Δ19A-1b mutant , which lacks the five related tin1 effector genes , i . e . this mutant shared 73% ( 476 out of 651 ) of differentially regulated genes with all other strains . Amongst the 66 plant genes that were specifically regulated after SG200Δ19A-1b infections , four maize endochitinase genes were significantly induced compared to SG200 infected leaves . In addition , transcript levels of two salicylic acid binding proteins and peroxidase-12 , which was found to be involved in the maize apoplastic oxidative burst [16] , were induced ( Table S3 ) . This suggests that tin1 genes modulate basal defenses . The tin3 mutant specifically affected the differential regulation of 70 maize genes ( Table S3 ) . Conspicuously , sucrose synthase and several transcription factors including auxin-response factors were induced suggesting a link to the reduced ability for tumor formation . After infections by tin4 and tin5 mutant strains the majority of differentially expressed maize genes were shared ( Figure 5C ) , which is in accordance with the hierarchical clustering result which places these two strains closely together ( Figure 5A ) . Nevertheless , 167 genes were only differentially regulated by the tin4 mutant and 110 maize transcripts were differentially regulated in infections by the tin5 mutant ( Figure 5C and Table S3 ) . Although tin4 and tin5 mutants displayed an indistinguishable virulence behavior , they elicited distinct molecular responses in maize . This appears to be a general feature and for example also holds for a strain carrying the 19A-1b deletion ( Figure 3B ) where virulence is only weakly attenuated . When the 10 most strongly induced host genes are compared after SG200Δ19A-1b infection , only one gene was among the top 10 genes upregulated after infection with SG200Δtin3 , SG200Δtin4 or SG200Δtin5 , respectively ( Table S1 ) . Gene ontology enrichment analyses were performed for the plant genes upregulated after infection by the three single tin gene deletion strains SG200Δtin3 , SG200Δtin4 and SG200Δtin5 ( Figures S6 , S7 , S8 ) . Differentially enriched functions for the individual effector deletion strains were visualized and in addition genes corresponding to the enriched functions are listed ( Figure 6 and Table S4 ) . This revealed that statistically distinct processes were induced in plants infected by the different mutants . For SG200Δtin3 , oxidoreductases and carbonate dehydratases were enriched , while SG200Δtin4 infection induced plant genes significantly enriched for functions involved in iron ion binding . In SG200Δtin5 infection , on the other hand , lipoxygenases and serine-carboxypeptidases were induced , which were not as highly upregulated in the other mutants ( Figure 6 and Table S4 ) . These results illustrate that plant responses to effector mutants which show no or only weak reductions in macroscopic symptoms can be highly specific and can be used to describe and discriminate mutant phenotypes .
In this communication , we have dissected the largest U . maydis effector gene cluster 19A , identified the most relevant effectors for seedling infection and present evidence , that individual effectors target distinct processes in the host plant . It was published before , that the deletion of cluster 19A abolishes tumor formation [12] . We now show that this dramatic phenotype is not associated with a block in biotrophic development . The cluster 19A mutant was still able to complete the life cycle up to the formation of teliospores . However , massive fungal proliferation observed at later stages in tumor tissue in infections with wild type strains was absent , suggesting that the effectors in this cluster are responsible for tumor induction either directly or indirectly but not for growth per se in the infected tissue . The analysis of plant responses elicited by the whole cluster mutant ( SG200Δ19A ) in comparison to responses to the progenitor strain SG200 revealed that 1816 of the 13 , 339 maize genes represented on the chip were differentially regulated . The analysis of the maize transcriptome changes observed for the individual effector mutant infections revealed that about 60% of these changes were shared by all mutants , suggesting that they are unspecific . Another aspect of the comparative transcriptome profiling was the finding that there is incomplete overlap between the genes affected by the individual mutants and genes altered in their regulation when the entire cluster is deleted . This could indicate that the effects of individual effector deletions cease to be visible when the entire cluster is deleted , i . e . the dramatic phenotype of the cluster deletion might bury the more subtle physiological changes caused by individual mutants . Our study has also revealed that effectors not studied individually because of their undetectable contribution to virulence after seedling infection , may profoundly affect the metabolic activity of the infected tissue . An example is the observed upregulation of biotin biosynthesis after infection with SG200Δ19A which is not observed in any of the tin mutant infections . In Arabidopsis thaliana it has been shown that biotin is critical for suppressing spontaneous cell death [22] . The fact that biotin biosynthesis appears upregulated may indicate a direct involvement of a specific effector for maintaining a certain level of this essential cofactor . Alternatively , biotin upregulation could be a secondary effect that allows the cluster 19A mutant to grow in the infected tissue . Future array analyses with mutants of cluster 19A not studied here with respect to the plant responses they elicit should allow to separate primary and secondary effects and allow to uncover the effector responsible for the regulation of glycolysis and biotin biosynthesis , respectively . Among the common genes differentially expressed after infection with all tin mutant strains , we observe an enrichment of upregulated plant defense genes and downregulation of genes involved in DNA metabolism . This is likely to reflect insufficient suppression of plant defenses due to reduced fungal proliferation ( or the absence of certain effectors ) and reduced plant tumor formation , respectively . Similarly , genes for photosynthesis components were consistently higher expressed after infections with all mutant strains compared to SG200 infections . This is unlikely to reflect an induction of photosynthesis during mutant infections but presumably results from an incomplete shutdown of photosynthesis , usually observed during U . maydis wild type infection [9] . The reduction in plant cell wall biosynthesis gene expression after infection with the cluster 19A mutant , which is not seen in infections with individual tin mutants , likely reflects the reduced ability of the cluster mutant to induce tumors containing enlarged plant cells [15] , while individual tin mutants can still induce tumors . With a minor impact on tumor formation , Tin2 was specifically responsible for anthocyanin accumulation in infected tissue while all other individual mutants showed anthocyanin induction . The transcriptome analysis revealed that several anthocyanin biosynthesis genes were upregulated in all the individual tin mutant strains but not after infection with SG200Δ19A lacking the whole cluster including tin2 ( Table S1 ) . This suggests that Tin2 is directly responsible for inducing these anthocyanin biosynthesis genes . Anthocyanin has been hypothesized to have a protective role against abiotic stresses [23] , and can be induced after biotic stress , although its role here is unclear [24] . While the virulence assays for tin4 and tin5 mutant strains were largely uninformative because of limited assay sensitivity , the transcriptome analysis revealed that tin4 and tin5 mutants elicited a series of plant responses that were mutant specific , but in addition 318 differentially regulated plant genes were differentially regulated by tin4 as well as tin5 mutant strains . Based on the fact that Tin4 and Tin5 share 19% identity and 39% amino acid similarity this could indicate that these effectors are in the process of diversification to different functions ( 167 tin4 specifically regulated transcripts , 110 tin5 specifically regulated transcripts ) while still maintaining some of the original common functions ( 318 commonly regulated transcripts ) . This interpretation would also make sense in view of the fact that um05312 , um05314 , um10557 and um05317 which are also related to tin4 and tin5 ( Fig . S1E ) were not individually deleted in this study . If all of the genes in this family had redundant functions , we would not have expected to see differences in the host responses to the tin4 and tin5 mutants . The gene ontology enrichment analysis showed that several genes involved in iron metabolism/uptake were specifically upregulated when tin4 is missing . Elevated iron availability may directly affect the activity of the respiratory burst oxidase requiring a heme prosthetic group to generate superoxide [25] . Such effects on plant defense have also been described after infections by Erwinia crysanthemi [26] and Blumeria graminis f . sp . tritici [27] . After tin5 mutant infections lipoxygenases and serine-carboxypeptidases were specifically upregulated , and both types of enzymes have been associated with defense [28] , [29] . OsBISCPL1 , a serine-carboxypeptidase from rice , was up-regulated in incompatible interactions between rice and the blast fungus , and was implicated in regulation of defense responses from heterologous expression studies [30] . In S . reilianum three orthologs of tin4 are present ( sr10075 , sr10077 and sr10079; Figure 2 ) . The simultaneous deletion of the neighboring genes sr10073 , sr10075 , sr10077 and sr10079 weakly affected S . reilianum virulence ( H . Ghareeb and J . Schirawski , personal communication ) , similar to what we observe for the tin4 , tin5 double mutant of U . maydis . As in U . maydis , the left half of cluster 19A contributes most strongly to virulence in S . reilianum ( H . Ghareeb and J . Schirawski , personal communication ) . Based on the observation that cluster 19A effector genes of U . maydis are not essential for tumor formation in tassel ( although tumor size in tassel was reduced after infection with SG200Δ19A ) [14] , we consider organ-specificity of effector function more likely to explain this finding . As S . reilianum does not induce tumors in leaves and develops disease symptoms only in the cob and in the tassel , this species may not need effector genes like tin3 for tumor induction in vegetative tissues of the maize plant . Consistent with this is our observation that the U . maydis cluster 19A mutant lacking all 24 effectors can still show biotrophic growth in plant tissue and thus behaves like S . reilianum ( except for the systemic spread ) . Interestingly , two of the Tin4 orthologs of S . reilianum , sr10075 and sr10077 , were recently shown to suppress apical dominance after maize infection ( H . Ghareeb , F . Drechsler , C . Löfke , T . Teichmann and J . Schirawski , personal communication ) . The effect on apical dominance is a late phenotype observed about six weeks after infection of maize plants with S . reilianum , i . e . a time point not covered by our assays . To ascertain whether Tin4 , Sr10075 and Sr10077 have conserved functions it would be interesting to test whether tin4 of U . maydis can complement the apical dominance phenotype of the respective S . reilianum mutants . Tin1-1 to Tin1-5 is a group of weakly related U . maydis effectors , which could not be functionally separated because their individual effects on virulence were too small to be reliably detected . The transcriptome changes of plants infected with a mutant lacking all five related genes revealed specific , strong inductions of endochitinases , SA-binding proteins and the apoplastic peroxidase POX12 . POX12 was recently shown to be inhibited by the U . maydis effector Pep1 leading to a suppression of the PAMP-triggered oxidative burst [16] . In addition , an NBS-LRR class disease resistance gene ( Zm . 3568 . 1 ) that could be involved in PAMP perception , showed transcriptional induction specifically after SG200Δ19A-1b infections . Together , these changes indicate an enhanced defense response against the 19A-1b deletion mutant , which suggests that the Tin1-1 to Tin1-5 effectors contribute to the suppression of basal host immunity . Interestingly , the immune response triggering avirulence factor , UhAvr1 ( UHOR_10022 , Figure 2 ) , of U . hordei is most closely related to the U . maydis effector Tin1-2 and Tin1-3 [31] . With respect to virulence no specific contribution of UhAvr1 could be detected [31] , which may be consistent with the very small contribution to virulence that is seen when all five tin1 genes are deleted . Tin3 is the effector in cluster 19A , which contributes most strongly to virulence . The strong transcriptional induction of two sucrose synthases after infection with the tin3 mutant strain ( as well as after infection with the cluster 19A mutant ) is likely to reflect enhanced photosynthetic activity in contrast to infections with SG200 where the transition from a juvenile sink tissue to a mature , photosynthetically active source tissue is blocked in infected leaves [9] . If an interplay between sucrose and auxin signaling , which was established in A . thaliana [32] also exists in maize; this could explain the observed upregulation of several auxin response factors after infection with these mutant strains . The specific upregulation of a WRKY transcription factor after infection with the cluster 19A mutant as well as the tin3 mutant could indicate elevated defense responses [33] , which are downregulated by Tin3 after infections with wild type strains . Alternatively , this regulatory gene might negatively control cell cycle and/or cell expansion , a feature of U . maydis induced tumors [9] . Another gene exclusively upregulated after infection with strains deleted for tin3 or cluster 19A is cytokinin oxidase 3 , an enzyme involved in cytokinin degradation . Cytokinin oxidases have been shown to restrict cell division and to regulate the sink capacity of kernels [34] . Thus , the downregulation of these activities by Tin3 , presumably after uptake of Tin3 by plant cells , might be necessary for tumor development . The finding of discrete plant responses after infection with individual effector mutants provides important leads for the functional analyses that can now be followed . For example , the predicted changes in hormone levels attributed to Tin3 could be determined from metabolic profiles or directly connected with Tin3 by transiently expressing Tin3 in plants with appropriate reporter gene constructs . The expression of Tin3 in transgenic plants might even allow assessing , whether the predicted effects on photosynthesis are direct or indirect . In more general terms our analyses reveal the power of studying pathogen effector mutants , by combining virulence assays with an assessment of plant responses to these mutants . Such comparisons do not only reveal common plant responses that reflect central processes targeted by the infection but in addition provide specific leads to the function of individual effectors . Furthermore , this approach does not rely on a significant virulence phenotype of the effector mutants studied and may thus be highly useful for the analysis of the vast majority of eukaryotic pathogen effectors that fall into this class [35] , [36] .
U . maydis strains were grown in YEPSL ( 0 . 4% yeast extract , 0 . 4% peptone , 2% sucrose ) with shaking at 28°C at 200 rounds min−1 ( rpm ) , to an optical density ( OD600 ) of 0 . 6–0 . 8 . Cells were centrifuged at 3000 g for 5 min , resuspended in H2O to an OD600 of 1 and used for syringe infection of 7-day-old maize seedlings ( variety Early Golden Bantam , Olds Seeds , Madison ) . At least 3 biological replicates were tested for virulence and disease was scored 12 dpi following described protocols [12] . To demonstrate the statistical differences of disease symptoms in the mutants compared to SG200 each of the symptoms was tested by Student's t-test ( ** p-values<0 . 01 . ) and corrected by Bonferroni correction for multiple testing . The haploid solopathogenic strain SG200 [12] was used for virulence assays and all mutations were introduced into this background . Standard molecular cloning strategies and techniques were applied [37] . All U . maydis strains ( Table S5 ) are derived from the solopathogenic strain SG200 and were generated by a PCR-based gene replacement approach using primers listed in Table S6 or , for complementation experiments , by insertion of p123 derivatives into the ip locus as described [38] . Deletion endpoints are depicted in the respective Figures . Constructs used for complementation contained the respective gene plus the promoter region extending up to the next gene plus the Tnos terminator . All generated constructs were sequenced prior to U . maydis transformation ( Table S7 ) . Isolated U . maydis transformants were tested for integration events in the desired loci by southern analysis . For the complementation constructs , single copy integrations into the ip locus were selected by southern analysis . For 3 strains only derivatives containing two inserts could be obtained , this is marked in the strain list ( Table S5 ) . For quantification of relative fungal biomass in infected maize leaves 7-day-old maize seedlings were infected with SG200 , SG200Δ19A , SG200Δ19A-1 and SG200Δ19A-2 and a section of the third leaf between 1 and 3 cm below the injection site was harvested after 2 dpi and 6 dpi . For genomic DNA extraction leaf material was frozen in liquid nitrogen , ground to powder , and extracted using a phenol-based protocol modified from Hoffman and Winston [39] . The qRT-PCR analysis was performed using an iCycler ( Bio-Rad ) in combination with the Platinum SYBR Green Supermix ( Invitrogen ) . U . maydis biomass was quantified with primers PPI-fw ( 5′-ACATCGTCAAGGCTATCG-3′ ) and PPI-re ( 5′-AAAGAACACCGGACTTGG-3′ ) amplifying the fungal ppi gene . Maize glyceraldehyde dehydrogenase was amplified with primers GAPDH-F ( 5′- CTTCGGCATTGTTGAGGGTTTG-3′ ) and GAPDH-R ( 5′- TCCTTGGCTGAGGGTCCGTC-3′ ) [15] and served as reference gene for normalization . Relative amounts of fungal DNA ( ppi ) were then calculated relative to the amount of GAPDH DNA using the cycle threshold ( Ct ) 2−ΔΔCt method [40] . Three biological replicates were combined and p-values were determined by using Student's t-test ( ** p-values<0 . 01 . ) . To validate the expression data of the microarray experiment 13 maize genes differentially regulated after infection with different mutant strains were subsequently analyzed by qRT-PCR . Infected plant material was generated as described for the microarray experiment and used for RNA extraction with Trizol ( Invitrogen , Karlsruhe , Germany ) . After extraction , the first-strand cDNA synthesis kit ( Invitrogen ) was used to reverse transcribe 3 µg of total RNA with oligo ( dT ) Primers . The qRT-PCR analysis was performed using an iCycler ( Bio-Rad ) in combination with the SYBR Green Supermix ( Invitrogen ) . Primers used for quantification of maize gene transcription levels are listed in Table S6 . Gene expression levels were calculated relative to GAPDH expression levels using the cycle threshold ( Ct ) 2−ΔΔCt method [40] . For the microarray experiments , maize plants ( Early Golden Bantam ) grown under defined conditions in a growth chamber were infected with SG200Δ19A-1b , SG200Δtin2 , SG200Δtin3 , SG200Δtin4 and SG200Δtin5 as described previously [15] . Samples of infected tissue were collected 4 dpi by excising a section of the third leaf between 1 and 3 cm below the injection site . For RNA extraction , material from >20 plants per experiment was combined , ground to powder on constant liquid nitrogen and RNA was extracted with Trizol ( Invitrogen , Karlsruhe , Germany ) . RNA was purified applying the RNeasy kit ( Qiagen , Hilden , Germany ) . Affymetrix maize genome microarrays were performed in three biological replicates , using standard Affymetrix protocols ( Midi_Euk2V3 protocol on GeneChip Fluidics Station 450; scanning on Affymetrix GSC3000G ) . Expression data were submitted to GeneExpressionOmnibus ( http://www . ncbi . nlm . nih . gov/geo/ ) ( Accession Number: GSE48406 ) . Previously published Affymetrix data for SG200 infections [9] ( GEO accession Number: GSE10023 ) and the microarrays performed in this study were analyzed together using the Partek microarray software suite version 6 . 12 . Expression values were normalized using the RMA method . Criteria for significance were a corrected p-value ( per sample ) with a FDR of 0 . 05 and a fold-change of >2 . Differentially expressed genes were calculated by a 1-way ANOVA model using method of moments [41] . For microscopic analysis of different life cycle stages of U . maydis strains SG200 and SG200Δ19A , a section of the third leaf between 1 and 3 cm below the injection site was excised after 1 dpi , 13 dpi and 30 dpi . We used a Zeiss Axiophot with differential interference contrast ( DIC ) optics for microscopic observations . The pictures were taken using a CCD camera ( C4742-95 , Hamamatsu ) . To visualize penetration events , appressoria were stained with calcofluor white ( 100 µg/ml; Fluorescent Brightener 28 , Sigma-Aldrich , Deisenhofen ) for 1 min . Intracellular growing fungal hyphae were stained with chlorazol black E using an established protocol [42] . | In this study , we provide the first step to the functional analysis of the largest gene cluster in the Ustilago maydis genome encoding 24 secreted effectors . While the deletion of the entire cluster dramatically affected tumor formation and abolished anthocyanin induction , only one of the genes had a large contribution to tumor formation , while another effector gene was primarily responsible for the anthocyanin induction . Unexpectedly , the cluster mutant could still colonize plants and complete the life cycle , i . e . behaves like an endophyte . Despite only small contributions to tumor formation , individual effector mutants caused distinct plant responses , suggesting that they affect discrete plant processes . On these grounds we are proposing to use plant responses as a general readout to assess and compare effector gene function . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"mycology",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"host-pathogen",
"interactions",
"medical",
"microbiology",
"microbial",
"pathogens",
"biology",
"and",
"life",
"sciences",
"microbial",
"growth"... | 2014 | Characterization of the Largest Effector Gene Cluster of Ustilago maydis |
Cytomegaloviruses ( CMVs ) establish chronic , systemic infections . Peripheral infection spreads via lymph nodes , which are also a focus of host defence . Thus , this is a point at which systemic infection spread might be restricted . Subcapsular sinus macrophages ( SSM ) captured murine CMV ( MCMV ) from the afferent lymph and poorly supported its replication . Blocking the type I interferon ( IFN-I ) receptor ( IFNAR ) increased MCMV infection of SSM and of the fibroblastic reticular cells ( FRC ) lining the subcapsular sinus , and accelerated viral spread to the spleen . Little splenic virus derived from SSM , arguing that they mainly induce an anti-viral state in the otherwise susceptible FRC . NK cells also limited infection , killing infected FRC and causing tissue damage . They acted independently of IFN-I , as IFNAR blockade increased NK cell recruitment , and NK cell depletion increased infection in IFNAR-blocked mice . Thus SSM restricted MCMV infection primarily though IFN-I , with NK cells providing a second line of defence . The capacity of innate immunity to restrict MCMV escape from the subcapsular sinus suggested that enhancing its recruitment might improve infection control .
Human CMV is a ubiquitous pathogen that causes birth defects and harms immunocompromised hosts [1] . Although adaptive immunity normally prevents disease , adaptive immune priming has not prevented infection establishment [2] , suggesting that this presents a qualitatively distinct challenge , requiring possibly different immune effectors . Analysing early human infection is made difficult by CMV transmission being sporadic and largely asymptomatic . However CMV infections long pre-date human speciation [3] , so different host / virus pairs are likely to share common themes and analogous animal infections can yield key insights . MCMV has particular value for understanding how CMVs work in vivo , as its host provides the standard model of mammalian cell biology . MCMV exploits myeloid cells to spread [4 , 5] , and live imaging shows peripheral to systemic spread via lymph nodes ( LN ) [6] , so LN myeloid cells are likely to be a key target for limiting systemic infection . Host immunity and viral evasion both influence CMV infection outcomes . Which dominates a given context can be hard to predict . Although MCMV has evolved to infect myeloid cells , myeloid phenotypes are diverse [7] and not all support its spread . SSM police the recycling of extracellular fluid from peripheral tissues back to the blood , capturing viruses from the lymph when it enters the LN subcapsular sinus [8 , 9] . Viral replication in SSM potentially delivers an amplified load . Therefore this must be curtailed , or at least slowed sufficiently for other defences to come into play . In addition SSM must protect the FRC that line the subcapsular sinus , and are targeted for example by Ebola virus [10] . While pathogen absorption by SSM can protect down-stream targets such as the blood , it is unlikely to prevent FRC infection as they present a large surface area in the same site [11] . Thus the SSM barrier must be active as well as adsorptive . -In this context SSM would capture viruses to sample rather than cleanse the lymph , and would communicate danger signals , for example to NK cells [12 , 13] . The subcapsular sinus is a prominent site of type I interferon ( IFN-I ) transcription [14] and IFN-I stimulates NK cells [15] , so it could contribute . SSM also communicate directly with B cells [16] . T cell-dependent antibody responses take several days to become effective , but T cell-independent antibody responses might protect SSM and FRC in some settings . The anti-viral functions of SSM have been explored by injecting mice with xenogenic pathogens such as Vesicular Stomatitis Virus ( VSV ) [17] . Footpad-inoculated ( i . f . ) VSV induces IFN-I , which protects neurons against pathological infection [18] . Whether SSM respond strongly to IFN-I is unknown . VSV productively infects both SSM and splenic marginal zone macrophages ( MZM ) , which capture viruses from the blood . Productive MZM infection is associated with a muted response to IFN-I [19] . This protects by inducing neutralizing antibodies . However VSV is unusually susceptible to neutralization by T cell-independent antibodies [20] . Viral evasion is often compromised in xenogenic infections , and the best known human rhabdovirus infection—rabies—spreads predominantly via neurons , so how far VSV in mice constitutes a general case is unclear . MCMV has evolved to infect mice and is not susceptible to T cell-independent antibody responses [21] . Nonetheless SSM restrict its spread [6] . They also restrict Murid Herpesvirus-4 ( MuHV-4 ) , an evasive gamma-herpesvirus [22] . How is unknown . IFN-I may contribute as it has anti-MCMV [23] and anti-MuHV-4 activity [24] , despite viral evasion . After intraperitoneal ( i . p . ) MCMV inoculation , IFN-I signals to NK cells and DC [25] . I . p . MCMV prominently infects the liver , and IFN-I suppresses viral lytic gene expression in in vitro propagated liver cells [26] . However the failure of hepatocytes to spread infection in vivo [27] makes unclear the relevance of liver infection to normal pathogenesis . Herpesviruses normally enter at peripheral sites , whereas i . p . virions reach the blood directly [28] , bypassing SSM . We show that SSM are a key site of IFN-I-mediated defence against MCMV . When IFN-I signalling was blocked , lymph-borne MCMV spread rapidly to systemic sites . NK cells provided a second line of defence but at the cost of tissue damage . Thus , an SSM-centered IFN-I response was crucial to limit MCMV dissemination .
We hypothesized that IFN-I contributes to SSM restricting MCMV infection . We first tracked by live imaging how IFNAR blockade affects MCMV spread . We gave BALB/c mice IFNAR blocking antibody or not i . p . then MCMV-LUC i . f . and imaged them daily for luciferase expression ( Fig 1a ) . Live image signals from untreated infected mice were evident in the feet from day 1 , and in the neck days 4–5 . IFNAR blockade significantly increased foot signals from day 3 and neck signals from day 4 . Plasmacytoid DC ( pDC ) produce IFN-I [29] , and prior pDC depletion with a bst-2-specific antibody also increased live image signals , but it had less effect than IFNAR blockade . This was consistent with genetic pDC depletion having only a modest effect on MCMV spread after i . p . inoculation [30] . Live image signals are comparable between mice for the same organs , but less so between different organs because overlying tissues cause site-dependent signal attenuation . Signals from adjacent organs can also be hard to distinguish . Therefore to understand better how IFNAR blockade affected MCMV passage through LN , we dissected mice 3 and 6 days after i . f . MCMV-LUC and imaged organs ex vivo ( Fig 1b ) . IFNAR blockade increased signals in multiple organs at day 6 , and in popliteal LN ( PLN ) and spleens at day 3 . Depleting pDC also increased luciferase signals , predominantly at day 6 , but had less effect than IFNAR blockade . To correlate luciferase expression with virion production , we measured virus titers in the same organs ( Fig 1c ) . They showed similar trends: IFNAR blockade increased titers in many organs at day 6 and in just PLN and spleens at day 3 . pDC depletion generally had less effect , increasing titers in PLN but not in the feet or salivary glands and only modestly in the liver and spleen . Thus Consistently the PLN appeared to be an important site of IFN-I-mediated anti-MCMV defence . We compared next how IFNAR blockade affected MCMV spread in C57BL/6 mice as an independent strain . BALB/c and C57BL/6 mice were each given anti-IFNAR blocking antibody or not then i . f . MCMV . The C57BL/6 NK receptor Ly49H recognizes directly the MCMV m157 [31–33] . m157 is polymorphic among MCMV strains; many variants do not bind Ly49H [34]; and m157+ virus passage in C57BL/6 mice selects m157- mutants [35]; so m157- MCMV is probably more representative of the viruses that colonize Ly49H+ mice , and in these experiments we used MCMV in which a GFP cassette interrupts m157 ( MCMV-GR; Fig 2 ) . After 3 days IFNAR blockade increased viral titers in the spleens of both mouse strains ( Fig 2a and 2b ) . It significantly increased day 3 PLN titers in BALB/c but not C57BL/6 mice ( Fig 2a and 2b ) . I . f . MCMV reaches the spleen via the PLN [6] , so we also looked earlier for increased PLN infection , and at day 1 observed this in both strains ( Fig 2c and 2d ) . At day 1 spleen infection was low or undetectable and footpad infection did not differ significantly from controls . Therefore IFNAR blockade increased PLN infection independently of footpad infection and before spleen infection , with C57BL/6 mice showing a smaller , less sustained increase than BALB/c mice . MCMV infects multiple cell types . To identify the source of increased virus titers in the PLN following IFNAR blockade , we gave mice anti-IFNAR or not , infected them with MCMV-GR and 1 day later stained PLN sections for viral GFP and lytic antigens ( Fig 3 ) . To focus on common , conserved events rather than increases that might be mouse strain-specific , we analysed the less extensive infection of C57BL/6 mice . IFNAR blockade increased viral GFP and lytic antigen expression around the subcapsular sinus ( Fig 3a , 3b and 3c ) . Most viral GFP+ cells ( approximately 75% ) were viral lytic antigen+ and most lytic antigen+ cells were GFP+ , suggesting that most infection in the PLN was lytic . Higher CD169+GFP+ cell counts indicated that IFN blockade increased SSM infection ( Fig 3a and 3d ) . Higher CD11c+GFP+ cell counts ( Fig 3b and 3d ) were consistent with more SSM infection as SSM express CD11c . More DC infection was also possible . Higher ER-TR7+GFP+ cell counts ( Fig 3c and 3d ) indicated that IFNAR blockade also increased FRC infection . Taken together , IFNAR blockade did not increase early footpad infection ( Fig 2 ) but increased lytic infection in the SSM and FRC of PLN . As a further measure of lytic infection , we stained PLN sections for the MCMV IE1 antigen ( Fig 4a ) . IE1 is not abundant in virions and is expressed in the nuclei of lytically infected cells , so its expression clearly distinguishes lytic infection from antigen uptake . IFNAR blockade significantly increased IE1+ cell numbers in the PLN ( Fig 4b ) . Thus , lytic antigen , IE1 and viral GFP expression all showed IFNAR blockade increasing MCMV infection of the LN subcapsular sinus . To understand why IFNAR blockade did not continue to increase PLN virus titers in C57BL/6 mice after day 1 , we examined day 3 PLN sections ( Fig 5 ) . IFNAR blockade increased viral GFP and lytic antigen expression , as at day 1 . However there was also widespread tissue destruction , with less B cell ( B220 ) and DAPI staining , and disruption of the ER-TR7+ LN architecture ( Fig 5a ) . The relatively low PLN titers at day 3 suggested that these changes were immunopathological rather than a result of -viral cytopathology . NK cells were a possible mediator , and day 1 PLN sections of IFNAR-blocked mice showed a significant increase in NKp46+ cell numbers that was abolished by NK1 . 1+ cell depletion ( Fig 5b and 5c ) . The increase in numbers seemed likely to reflect NK cell recruitment , as 1 day allowed little time for local proliferation . Although IFN-I can recruit NK cells [36] , NKp46+ cell recruitment into MCMV-infected PLN increased when IFN-I signalling was blocked , reflecting presumably a dominant inflammatory effect of higher viral loads ( Fig 5b and 5d ) . NK cell depletion increased virus titers in C57BL/6 PLN at days 1 and 3 , in both IFNAR-blocked and control mice , without significantly increasing titers in footpads or spleens ( Fig 5d ) . As with IFNAR blockade , day 1 PLN sections of NK-depleted mice showed more infected ER-TR7+ cells ( Fig 6a and 6c ) . Therefore both IFN-I and NK cells restricted FRC infection . NK cell depletion increased PLN virus titers as much as IFNAR blockade ( Fig 5d ) but had less effect on viral GFP+ cell numbers ( Fig 6a and 6b ) . Not all GFP+ cells were lytic antigen+ ( Fig 4 ) , so GFP+ /lytic antigen- cells may be poor NK cell targets . NK cell depletion with anti-asialo-GM1 antibody significantly increased day 1 PLN virus titers in both BALB/c and C57BL/6 mice , with a greater effect in BALB/c ( Fig 6e ) , so BALB/c NK cells also helped to control early infection . There was no loss of DAPI staining in dual NK-depleted/IFNAR-blocked PLN of C57BL/6 mice at day 3 ( Fig 6d ) , despite higher virus titers ( Fig 5d ) . This result suggested that NK cells caused the PLN damage . Infected PLN of IFNAR-blocked mice showed extensive expression of the apoptotic cell marker caspase 3 ( Fig 7a ) , which was negligible in uninfected PLN ( Fig 7b ) . Most viral GFP+ FRC were caspase 3+; when NK cells were depleted , most were caspase 3- ( Fig 7a and 7c ) . Therefore NK cell recruitment into the PLN led to apoptosis in infected FRC . Like IFNAR blockade ( Fig 1 ) , SSM depletion accelerates MCMV spread from the PLN to the spleen [6] . Thus we hypothesized that IFN-I production by SSMs restricted PLN infection . To compare SSM depletion with IFNAR blockade , we depleted SSM with i . f . liposomal clodronate , blocked IFNAR , gave both treatments , or gave neither , then infected all the mice i . f . with MCMV ( Fig 8 ) . SSM depletion and IFNAR blockade both increased day 3 PLN and spleen infections without increasing footpad infections ( Fig 8a ) . In PLN they gave similar increases and when combined gave an additive increase . Therefore IFN-I did not come just from SSM—notably pDC depletion also increased virus titers ( Fig 1 ) —and it did not act just on SSM , consistent with its protection of FRC ( Fig 5 ) . The additive increase implied also that IFN-I was not the only SSM defence , consistent with a report of innate effector recruitment via IL-18 [37] . Nonetheless the additive effect was not large , so IFN-I signalling appeared to be a major component of SSM-mediated MCMV restriction . In situ analysis of day 3 PLN showed markedly increased FRC infection after SSM depletion by clodronate loaded liposomes ( “clod”; Fig 8b and 8c ) . However this increased FRC infection recruited relatively few NK cells ( Fig 8d ) . Thus , although NK cells controlled FRC infection , their recruitment did not depend on FRC infection . IFN-I signalling was not essential to recruit NK cells either ( Fig 8d ) . In fact it limited NK cell recruitment and preserved PLN cellularity , presumably by reducing virus loads ( Fig 8c and 8d ) . NK cell recruitment was maximal after SSM depletion plus IFNAR blockade . Thus , non-SSM myeloid cells such as medullary sinus and medullary cord macrophages and DC [38] , which are not depleted by i . f . liposomal clodronate [22] , may also produce the necessary cytokines , and the multiplicity of sufficient signals meant that NK recruitment correlated primarily with total virus load . This recruitment was then associated with a reduction in ER-TR7+ cell infection and a general loss of LN cellularity . While SSM depletion alone increased splenic infection , SSM depletion plus IFNAR blockade gave no increase beyond IFNAR blockade alone ( Fig 8a ) . I . f . liposomal clodronate does not deplete MZM [39] , so here it could increase infection only by increasing seeding , which IFNAR blockade did already , with IFN-I limiting MCMV spread in both the PLN and the splenic MZ . SSM expression of lysM [22] and CD11c [6] allows their MCMV production to be tracked by floxed virus recombination in lysM-cre and CD11c-cre mice . We gave mice anti-IFNAR antibody or not , then i . f . MCMV-GR , which cre switches from GFP to tdTomato expression ( Fig 9 ) . In lysM-cre mice , IFNAR blockade increased MCMV-GR switching in footpads at 3 days ( Fig 9b , % virus switch ) . However the proportion of switched virus in PLN and spleens remained low . Thus , day 3 PLN infection derived from either earlier footpad infection ( Fig 2 ) or inoculated virions [6] . IFNAR blockade did increase PLN virus production by lysM+ cells , because total titers increased . However the % switched showed little change , so virus production by lysM- cells also increased ( Fig 9a ) . In contrast to low switching rates of infectious virus in the PLN and spleen , IFNAR blockade increased the fluorochrome switching of infected cells on tissue sections ( % cell switch; Fig 9b and 9c ) . Infecting CD11c-cre mice gave similar results ( Fig 9d–9f ) , confirming the low virus productivity of SSM and ruling out LysM-CD11c+ DC as the source of unswitched virus in IFNAR-blocked lysM-cre mice . Whereas floxed MuHV-4 [40] and Herpes simplex virus type 1 [41] show substantial switching in lysM-cre and CD11c-cre mice after IFNAR blockade , unswitched MCMV from cre- cells such as FRC seemed to dilute out switched SSM virus regardless of IFNAR blockade . To test this idea further , we gave lysM-cre mice IFNAR blocking antibody , optionally also depleted NK cells , ( to preserve FRC ) then infected the mice i . f . with MCMV-GR . Three days later we recovered virus from footpads , PLN and spleens and quantified fluorochrome switching . On a background of IFNAR blockade , additional NK cell depletion significantly increased virus titers in PLN and spleens ( Fig 10a ) , but it increased virus switching only in footpads . In fact , virus switching decreased as infection progressed from PLN to spleens , indicating an increased contribution of non-myeloid cells to virus production at each stage ( Fig 10b ) . To preserve non-myeloid IFN-I signalling , we infected lysM+/creIFNARflox/flox mice , which inactivate IFNAR specifically in lysM+ cells . These mice , and lysM-cre controls with IFNAR intact , were depleted of NK cells then given MCMV-GR i . f . ( Fig 10c and 10d ) . Virus titers increased in the floxed IFNAR mice , and virus switching , although still low , was significantly more than in the controls . Thus , IFNAR inactivation just in lysM+ cells partly offset the diluting effect of non-myeloid ( FRC ) infection . The specificity of the block is probably limited by a lack of positive feedback through IFNAR reducing IFN-I production by lysM+ SSM in lysM+/creIFNARflox/flox mice [42] , and thereby reducing IFN-I signalling to lysM- FRC . Virus switching was not limited by poor fitness of the switched virus in lysM-cre mice , as i . f . inoculation of a mixture of switched and unswitched viruses gave equal recovery of each from footpads , PLN and spleens ( Fig 10e ) . We observed also equal recovery of mixed switched and unswitched viruses from the liver and spleen after i . p . inoculation , and from the lungs of C57BL/6 mice after i . n . inoculation ( data not shown ) . Nor was the switching insensitive , as i . p . inoculation of lysM-cre mice gave >90% switching in F4/80+ peritoneal macrophages ( Fig 10f ) . Rather MCMV appeared inherently to restrain its lytic replication in myeloid cells and replicate readily in FRC . I . f . MCMV directly infects SSM [6] . If it reached FRC via SSM , IFN-I could protect FRC by limiting SSM infection; but if it reached FRC directly , they must be protected directly . To look for direct FRC infection we gave mice IFNAR blocking antibody or not , then gL- MCMV i . f . ( Fig 11 ) . gL is essential for MCMV membrane fusion , so gL- virus is propagated in gL+ cells . The pseudotyped gL+ virions are infectious but without complementation do not produce infectious progeny . Thus in vivo , gL- MCMV infects just once . The gL coding sequence was disrupted by a β-galactosidase ( βgal ) expression cassette , allowing infected cells to be identified by staining for βgal . Both control and IFNAR-blocked mice had βgal+ cells around the subcapsular sinus ( Fig 11a ) . The latter had significantly more per field of view ( Fig 11b ) . Most βgal+ cells were SSM . However both control and IFNAR-blocked mice also had βgal+ER-TR7+ FRC . A significantly higher proportion of βgal+ cells were ER-TR7+ after IFNAR blockade ( Fig 11c ) . Therefore i . f . MCMV directly infected both SSM and FRC; and IFN-I suppressed FRC infection more than SSM infection . This result , and increased FRC infection after SSM depletion ( Fig 8 ) , argued that FRC infection is normally suppressed by IFN-I from MCMV-exposed SSM .
LN survey extracellular fluid returning to the blood . SSM remove particulate antigens such as viruses . However viruses that replicate in SSM , or bypass them by infecting FRC , mandate additional defences . MCMV , like HCMV , has a broad tropism , encompassing fibroblasts , macrophages and endothelial cells , and lymph-borne MCMV directly infected both SSM and FRC . Thus , the need to contain this infection preceded adaptive immunity . Containment depended primarily on IFN-I , which restricted SSM and FRC infections despite viral IFN-I evasion [42] . Innate immunity is inherently polylithic , and IFN-I was not the sole SSM-based defence: NK cell recruitment was also important . Nor were SSM the sole IFN-I producers at the subcapsular sinus: pDC also contributed . But IFN-I and SSM were major players , and boosting their engagement provides a potential route to better infection control . Most analyses of SSM anti-viral functions have infected mice with model , xenogenic pathogens such as VSV [18] . VSV replicates rapidly in many cell types , including SSM . Limited replication of the mouse-adapted , macrophage-tropic MCMV in SSM was consequently surprising . However filtering macrophages also restrict the macrophage-tropic murine pathogens lymphocytic choriomeningitis virus [43] and murine coronavirus [44] via IFN-I . As IFN-I contributes to myeloid cell homeostasis [45] , incomplete evasion may be a viral compromise necessary to exploit myeloid cells for dissemination and persistence [4] . MCMV replicated preferentially in FRC . These provide LN structure , for example supporting the conduits that take low molecular weight proteins from the afferent lymph to near high endothelial venules for antigen presentation [11] . The conduits are too small to transport virions , but the lining FRC provide a route to the blood , and restricting acute FRC infection was an important part of SSM-mediated host defence . When the SSM , IFN-I or NK cell components of host defence were compromised , FRC seemed to provide the main additional site of viral replication . The greater effects on MCMV of IFNAR blockade than pDC depletion , and of IFNAR blockade plus SSM depletion than SSM depletion alone , implied redundancy in IFN-I production between SSM and pDC . Medullary sinus and medullary cord macrophages [38] and conventional dendritic cells may also contribute , as may resident stromal cells and recruited inflammatory cells . Most pDC are migratory and recruited to inflammatory sites [29] . Thus their response can take time to develop . The early ( day 1 ) inhibition of MCMV infection by IFN-I , including inhibition of viral reporter gene expression in the first cells encountered , was consistent with a key role for resident SSM . SSM depletion possibly under-estimated their contribution , as the ensuing inflammation could recruit other , compensatory IFN-I-producing cells . We envisage that SSM are constitutively in a high state of readiness to mount IFN-I responses to captured virions , whereas immigrant cells function more as a back-up should SSM prove insufficient . NK cell recruitment was another important back-up to the SSM IFN-I response . Whole organ titers have established the importance of NK cells for MCMV control [15 , 31 , 36] and the impact of viral evasion [46] . What infected cell types NK cells target has been less clear . Although lymph-borne virions infected large numbers of SSM , the SSM contribution to new virion production was relatively low . FRC seemed to be the main source . A similar situation occurs in the lungs , where alveolar macrophages are prominent infection targets but most new virus seems to come from type II alveolar epithelial cells [47] . Consequently FRC were an important NK cell target . When IFNAR was blocked , increased FRC infection and NK cell recruitment led to tissue damage . This destructive capacity of NK cells could explain why inflammation around the subcapsular sinus compromises subsequent immune responses [48] . Fig 12 summarizes what we now know of MCMV infection and control at the LN subcapsular sinus . Most virions in the afferent lymph are captured by SSM , but some directly infect FRC . Without IFN-I and NK cells , FRC infection is highly productive . New virions then spread to the liver and spleen via the efferent lymph and blood . Their infections probably follow a similar pattern , with Kupffer cells and MZM taking the place of SSM [49 , 50] . SSM restrict their own infection and that of FRC through IFN-I , and possibly also other pro-inflammatory cytokines such as IL-18 [37]; and they recruit cellular effectors , including NK cells which kill infected FRC . Protection against VSV depends on blocking acute neuronal infection; protection against MCMV depends on reducing its spread to sites of persistence , such as the bone marrow and salivary glands . Thus a key question now is whether recruiting innate effectors more rapidly and in amplified form can reduce long-term systemic CMV loads . IFN-I alone is unlikely to protect against chronic infection . However a capacity to amplify and recruit innate responses at the LN subcapsular sinus , for example through CD4+ T cell-derived IFN-γ [49] , may be an important property of vaccine-induced immunity .
BALB/c , C57BL/6J , LysM-cre [51] , CD11c-cre [52] and IFNARflox/flox [53] mice were maintained at University of Queensland animal units and used when 6–12 weeks old . For myeloid cell-specific IFNAR disruption , LysM-cre mice were back-crossed onto IFNARflox/flox . Virus was injected i . f . ( 106 p . f . u . in 50μl ) under isoflurane anesthesia . For luciferase imaging , mice were given 2mg luciferin i . p . and monitored for light emission by charge-coupled device camera scanning ( Xenogen , IVIS-200 ) . Phagocytic cells were depleted with 50μl i . f . clodronate-loaded liposomes ( http://clodronateliposomes . org ) [39] 1 and 3 days before infection . Depletion was confirmed in uninfected mice by a loss of CD169 expression around the PLN subcapsular sinus [6 , 22] . NK cells were depleted from C57BL/6 mice with anti-NK1 . 1 mAb PK136 ( 200μg i . p . , Bio-X-cell ) and from BALB/c or C57BL/6 mice with anti-asialo-GM1 pAb ( 100μl i . p . , Wako Chemicals ) . Each was given 1 and 3 days before infection and every 2 days thereafter . Cell depletion was >90% effective , as measured by immunostaining of tissue sections with anti-NKp46 antibody . IFN-I signalling was blocked with anti-IFNAR mAb MAR1-5A3 ( 200μg i . p . , Bio-X-cell ) , given 1 day before infection and every 2 days thereafter . pDC were depleted with anti-Bst2 mAb BX444 ( 250μg i . p . , Bio-X-Cell ) , 1 and 3 days before infection and every 2 days thereafter . Cell depletion was 85–90% effective , as measured by flow cytometry for Bst-2 and SiglecH . Experimental groups were compared statistically by two-tailed unpaired Student’s t test unless stated otherwise . All experiments were approved by the University of Queensland Animal Ethics Committee ( Licence numbers 218/15 , 391/15 , 479/15 ) in accordance with the Australian Code for the Care and Use of Animals for Scientific Purposes and regulated under the Queensland Animal Care and Protection Act ( 2001 ) . MCMV strain K181 ( MCK2+ ) was used throughout . All modified viruses were derived from MCMV K181 by homologous recombination . For live imaging we used a derivative ( MCMV-LUC ) that expresses firefly luciferase via autocatalytic release from a fusion protein with the lytic cycle M78 gene product [54] . For floxed reporter gene switching we used MCMV-GR , which has an HCMV IE1 promoter-driven cassette inserted into M157 [6] . The cassette encodes a floxed GFP gene upstream of a nuclear-targeted tdTomato gene . Thus , cre irreversibly switches MCMV-GR from green to nuclear red fluorescence . Switched and unswitched MCMV-GR show no difference in host colonization . MCMV with the essential virion glycoprotein L ( gL ) gene disrupted by insertion of an HCMV IE1 promoter-driven β-galactosidase expression cassette into M115 [6] was propagated on gL-expressing NIH-3T3 cells . All other viruses were propagated on NIH-3T3 cells ( American Type Culture Collection CRL-1658 ) . Infected cells were cleared of cell debris by low speed centrifugation ( 500 x g , 10min ) , then virus was concentrated by ultracentrifugation ( 35 , 000 x g , 2h ) . Infectious virus was plaque assayed on murine embryonic fibroblasts . Cells were grown in Dulbecco’s modified Eagle’s medium supplemented with 2mM glutamine , 100IU/ml penicillin , 100μg/ml streptomycin and 10% fetal calf serum . Organs were fixed in 1% formaldehyde-10mM sodium periodate-75mM L-lysine ( 24h , 4°C ) , equilibrated in 30% sucrose ( 18h 4°C ) , then frozen in OCT . Sections ( 6μM ) were air dried ( 1h , 23°C ) , washed 3x in PBS , blocked with 0 . 3% Triton X-100-5% normal goat serum ( 1h , 23°C ) , then incubated ( 18h , 4°C ) with antibodies to B220 ( rat mAb RA3-6B2 , Santa Cruz Biotechnology ) , CD68 ( rat mAb FA-11 ) , ER-TR7 ( rat mAb ) , CD11c ( hamster mAb N418 ) , SiglecH ( rat mAb 440c ) , β-galactosidase ( chicken pAb ) , Caspase 3 , Lyve-1 ( both rabbit pAb , AbCam ) , CD206 ( rat mAb MR5D3 ) , CD169 ( rat mAb 3D6 . 112 , Serotec ) ; NKp46 ( Rat Mab 29A1 . 4 , Biolegend ) , MCMV IE1 pp89 ( mouse mAb Croma101 ) [55] and MCMV virion antigens ( rabbit pAb , raised in house by subcutaneous inoculation with MCMV K181 propagated in NIH 3T3 cells ) . After incubation with primary antibodies , sections were washed x3 in PBS , incubated ( 1h , 23°C ) with combinations of Alexa 568- or Alexa 647-conjugated goat anti rat IgG pAb , Alexa 647-conjugated goat anti-mouse IgG1 pAb , ( Life Technologies ) , and Alexa 488- goat anti-chicken pAb ( Abcam ) , then washed x3 in PBS , stained with Hoechst 33342 and mounted in ProLong Gold ( Life Technologies ) . Fluorescence was visualised with a Zeiss LSM510 confocal microscope and analysed with Zen imaging software . | Cytomegaloviruses ( CMVs ) infect most people and are a common cause of fetal damage . We lack an effective vaccine . Our knowledge of human CMV is largely limited to chronic infection , which is hard to treat . Vaccination must target early infection . Related animal viruses therefore provide a vital source of information . Lymph nodes are a bottleneck in murine CMV spread from local to systemic infection . We show that viral passage through lymph nodes is restricted by interferons and NK cells . These defences alone cannot contain infection , but boosting their recruitment by vaccination has the potential to keep infection locally contained . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"nuclear",
"staining",
"vesicular",
"stomatitis",
"virus",
"pathology",
"and",
"laboratory",
"medicine",
"spleen",
"pathogens",
"immunology",
"microbiology",
"cytomegal... | 2016 | Type 1 Interferons and NK Cells Limit Murine Cytomegalovirus Escape from the Lymph Node Subcapsular Sinus |
The Gram-negative bacterium Proteus mirabilis is a common cause of catheter-associated urinary tract infections ( CAUTI ) , which can progress to secondary bacteremia . While numerous studies have investigated experimental infection with P . mirabilis in the urinary tract , little is known about pathogenesis in the bloodstream . This study identifies the genes that are important for survival in the bloodstream using a whole-genome transposon insertion-site sequencing ( Tn-Seq ) approach . A library of 50 , 000 transposon mutants was utilized to assess the relative contribution of each non-essential gene in the P . mirabilis HI4320 genome to fitness in the livers and spleens of mice at 24 hours following tail vein inoculation compared to growth in RPMI , heat-inactivated ( HI ) naïve serum , and HI acute phase serum . 138 genes were identified as ex vivo fitness factors in serum , which were primarily involved in amino acid transport and metabolism , and 143 genes were identified as infection-specific in vivo fitness factors for both spleen and liver colonization . Infection-specific fitness factors included genes involved in twin arginine translocation , ammonia incorporation , and polyamine biosynthesis . Mutants in sixteen genes were constructed to validate both the ex vivo and in vivo results of the transposon screen , and 12/16 ( 75% ) exhibited the predicted phenotype . Our studies indicate a role for the twin arginine translocation ( tatAC ) system in motility , translocation of potential virulence factors , and fitness within the bloodstream . We also demonstrate the interplay between two nitrogen assimilation pathways in the bloodstream , providing evidence that the GS-GOGAT system may be preferentially utilized . Furthermore , we show that a dual-function arginine decarboxylase ( speA ) is important for fitness within the bloodstream due to its role in putrescine biosynthesis rather than its contribution to maintenance of membrane potential . This study therefore provides insight into pathways needed for fitness within the bloodstream , which may guide strategies to reduce bacteremia-associated mortality .
The Gram-negative bacterium Proteus mirabilis commonly causes catheter-associated urinary tract infection ( CAUTI ) , particularly in the elderly and in healthcare facilities [1–5] . Consequences of P . mirabilis CAUTI can include infection of the kidneys , urinary stone formation due to bacterial urease ( urolithiasis ) , permanent renal damage , dissemination of bacteria to the bloodstream ( bacteremia and/or sepsis ) , and possibly death [5–9] . In healthcare facilities including nursing homes , CAUTI is the most common cause of secondary bacteremia , which is associated with a one-year mortality rate of 10–13% in most settings , but can be as high as 66% [6 , 7 , 10 , 11] . Prior studies have shown that P . mirabilis is the causative agent in 13–21% of bacteremias experienced by nursing home residents , and the urinary tract is the predominant source of these CAUTI-associated bacteremias [12–18] . There are also increasing reports of antibiotic-resistant P . mirabilis isolates , including production of extended-spectrum β-lactamases ( ESBLs ) and carbapenemases [19–22] , which is problematic as the mortality rate for ESBL-positive P . mirabilis bloodstream infections is significantly higher than that of ESBL-negative P . mirabilis isolates [23–25] . In North America , the percent of imipenem-resistant P . mirabilis bacteremia isolates increased from 0 . 2% to 35 . 3% between 2008 and 2012 [26] . ESBL-positive P . mirabilis isolates also tend to be multidrug-resistant ( MDR ) . For instance , approximately 40% of 405 P . mirabilis clinical urine isolates in Japan were ESBL producers , and roughly 70% of these isolates exhibited concomitant resistance to fluoroquinolones [20] . A recent study of 99 patients with P . mirabilis bloodstream infections highlighted the consequences of MDR infections , as the 21-day mortality rate for individuals with MDR P . mirabilis was 50% , compared to 19% of those with non-MDR strains [27] . Since there are no licensed vaccines against P . mirabilis , identification of novel , non-antibiotic targets of treatment would be advantageous . Many studies have investigated potential virulence factors encoded by P . mirabilis and their contribution to pathogenicity in animal models of UTI and CAUTI [summarized in [5]] , but there have been no direct experimental evaluations of P . mirabilis virulence factors for bloodstream infection . However , some UTI and CAUTI fitness factors also contribute to spleen colonization following infection of the urinary tract , which is indicative of secondary bacteremia . For instance , factors pertaining to defense against antimicrobial peptides are important for fitness in a mouse model of CAUTI , and disruption of the polymyxin resistance gene arnA results in a fitness defect in the urinary tract as well as in the spleen [28] . Nitrogen assimilation pathways also contribute significantly to experimental UTI and CAUTI , with the potential to impact bacteremia . Ammonia is the preferred nitrogen source of P . mirabilis , which encodes two ammonia incorporation systems: glutamine synthetase ( glnA ) and glutamate dehydrogenase ( gdhA ) . Glutamine synthetase ( glnA ) is a critical fitness factor for urinary tract colonization in the CAUTI model and spleen colonization in mice that progressed to bacteremia [28] , and glutamate dehydrogenase ( gdhA ) was previously shown to be important for colonization of the urinary tract and spleen in an ascending model of UTI [29] . We previously generated a genome-saturating library of transposon mutants in P . mirabilis CAUTI isolate HI4320 and utilized transposon insertion-site sequencing ( Tn-Seq ) to identify 629 genes encoding candidate fitness factors for colonization and survival in the catheterized bladder and/or kidneys of infected mice [28] . Many of these genes have the potential to be important for dissemination to the bloodstream from the urinary tract or for survival within the bloodstream . In this study , we applied Tn-Seq to directly identify P . mirabilis fitness factors that contribute to survival within the bloodstream versus those that are important for survival in serum ex vivo .
We previously determined that 34 , 249 transposon mutants are required for 99 . 99% probability of full genome coverage , based on the P . mirabilis HI4320 genome size [28 , 30] . The optimal infectious dose of P . mirabilis HI4320 ( wild type , or WT ) was determined to be 1x107 CFU/ml via intravenous inoculation to achieve spleen and liver colonization in 100% of inoculated mice with no mortality ( S1A Fig ) . This dose results in lesions in some of the mice by 24 hours post-inoculation ( hpi ) and would likely be lethal if allowed to progress , although the majority of mice show no outward signs of disease . To assess potential bottlenecks in the mouse model of bacteremia that could result in decreased recovery of mutants due to factors unrelated to fitness , we performed a competition infection with a P . mirabilis mutant that has a neutral fitness phenotype compared to WT during bloodstream infection . As blood plasma contains only ~3 mM urea [ ( 100- to 1 , 000-fold less than in urine [31 , 32]] , we hypothesized that urease would not provide a significant advantage to P . mirabilis during direct bloodstream infection and that a urease mutant would exhibit a neutral phenotype ideal for bottleneck assessment . We therefore compared a P . mirabilis urease mutant ( ureF ) to the WT strain in the bacteremia infection model ( S1B Fig ) . The mutant achieved a similar bacterial burden as WT in all organs , indicating that it would be suitable for bottleneck assessment . Mice were next inoculated with the ureF mutant and WT P . mirabilis in ratios of 1:1 , 1:1000 , and 1:10 , 000 for bottleneck estimation ( S1C and S1D Fig ) . The median colonization density for all mice was ~2x107 CFU/gram of tissue in the liver , 1x106 CFU/g spleen , and 1x103 CFU/g kidneys , indicating that a sufficient colonization density for Tn-Seq is achieved in the spleen and liver but not the kidneys ( S1C Fig ) . For bottleneck assessment , a competitive index was calculated based on the ratio of ureF:WT from the liver and spleen compared to the input ratio . While there was more variability in the competitive index for mice inoculated with a 1:10 , 000 ratio of ureF:WT than the 1:1 or 1:1 , 000 ratios , the competitive index was not found to be significant for any ratio , indicating that there was not a significant bottleneck ( S1D Fig ) . Based on these data , the use of 5x104 transposon mutants would be suitable for Tn-Seq in the bacteremia model , providing 50x coverage of each mutant for an inoculum of 1x107 CFU . A transposon mutant pool was therefore generated by combining the 5 pools of 10 , 000 mutants each that we previously validated and utilized in the CAUTI Tn-Seq study [28] . A schematic of the Tn-Seq experimental setup is provided in S2 Fig . Ten mice were inoculated intravenously with the P . mirabilis pool of transposon mutants to assess fitness factors for survival within the bloodstream , as measured by recovery from spleens and livers 24 hours post-inoculation . All 10 mice exhibited adequate spleen and liver colonization for analysis ( S3 Fig ) . Concurrently , the transposon mutant pool was also subjected to three in vitro conditions to facilitate identification of fitness factors that are infection-specific , and therefore have defects in the bacteremia screen but are not required for fitness during incubation in serum ex vivo . The in vitro conditions included for this assessment were as follows: 1 ) RPMI medium alone ( labeled “RPMI” ) , 2 ) RPMI medium with 50% heat-inactivated naïve mouse serum ( generated from CBA/J mice , labeled “Naïve” , and 3 ) RPMI with 50% heat-inactivated acute-phase serum ( generated from CBA/J mice 5 hours after intraperitoneal injection with heat-killed P . mirabilis , labeled “APS” ) . Heat-inactivated serum was utilized for these studies to remove heat-labile antimicrobial compounds , allowing genes involved in tolerance of these factors to be retained as potentially infection-specific . Each gene was assigned a fitness index for each condition based on the number of unique insertion-sites within that gene and the depth of reads at each site for a given condition relative to the input samples . Genes were considered to be candidate fitness factors based on having an adjusted P-value <0 . 05 , and a ratio of input/output ≥2-fold . Fitness factors for spleen or liver colonization that were not identified as important for growth in either serum condition in vitro were considered to be infection-specific . The full dataset is provided in S1 Table and an overview of the fitness factors identified for each condition is provided in Figs 1 and 2 , which will be discussed in detail below . A total of 96 genes ( 2 . 5% of the 3747 genes encoded in the P . mirabilis HI4320 genome ) were fitness factors for growth in RPMI in the absence of mouse serum ( S2 Table , and summarized in Fig 1A and 1B ) . RPMI fitness factors were most commonly associated with nucleotide transport and metabolism ( 22 . 4% ) , amino acid transport and metabolism ( 20 . 0% ) , translation , ribosomal structure and biogenesis ( 10 . 6% ) , and cell wall/membrane/envelope biogenesis ( 8 . 2% ) ( Fig 2 ) . 75 of these 96 genes ( 78% ) were also fitness factors in mouse serum ( either naïve or APS , shown in S3 Table and Fig 1A and 1B ) , indicating that they are likely required for growth in RPMI and the presence of serum cannot complement growth . These factors primarily pertained to amino acid transport/metabolism and nucleotide transport/metabolism , and include: glutamine synthetase ( glnA ) , glutamate 5-kinase and glutamate 5-semialdehyde dehydrogenase ( proAB ) , aspartate-ammonia ligase ( asnA ) , pyruvate dehydrogenase ( aceEF ) , purine metabolic genes ( guaAB , purC , purNM , purF , purK , purH , purD , and purA ) , and pyrimidine metabolic genes ( pyrF and pyrC ) . This group also contains a stringent starvation protein ( sspA ) , carbon storage regulator ( csrA ) , and RNA polymerase σ54 ( rpoN ) . 21 genes ( 21% ) were identified as only being important for survival in RPMI but not in the presence of either naïve or acute-phase serum , indicating that components of mouse serum can complement the defects of these mutants in RPMI . Included in this list were genes involved amino acid transport and metabolism ( carAB , thrC , usg , and speB ) , transport of magnesium ( PMI1580 ) and vitamin B12 ( btuC ) , and factors involved in cell wall synthesis ( rfaA , rfaD , rfaF , and wabG ) . Interestingly , there were no fitness factors related to defense mechanisms or lipid transport in the RPMI condition without mouse serum , indicating that these factors are only important in the presence of host components . 138 genes ( 3 . 7% ) were identified as ex vivo fitness factors in 50% mouse serum ( S3 Table and Fig 1A and 1B ) . 121 genes were fitness factors in naïve mouse serum , of which 107 could be assessed based on their cluster of orthologous group function ( COG ) . These genes were primary involved in amino acid transport and metabolism ( 15 . 9% ) , nucleotide transport and metabolism ( 14 . 0% ) , energy production and conversion ( 9 . 4% ) , and translation , ribosomal structure and biogenesis ( 9 . 4% ) , indicative of an increased need for transcriptional and translational machinery during growth in serum as compared to the medium in which the transposon library was generated ( Fig 2 ) . 73/121 ( 60 . 3% ) were also important for growth in RPMI alone , while 48 ( 39 . 7% ) were specific fitness factors for growth in serum . Of these 48 serum-specific fitness factors , 23 ( 47 . 9% ) were only identified as fitness factors in naïve serum , while the remaining 25 were also identified in acute-phase serum . For acute-phase serum ( APS ) , 98 genes were identified as fitness factors ( S3 Table and Fig 1A and 1B ) . 86 of these genes could be assessed based on their COG function , and the most highly represented categories include: nucleotide transport and metabolism ( 14% ) , amino acid transport and metabolism ( 11 . 6% ) , replication , recombination and repair ( 9 . 3% ) , and coenzyme transport and metabolism ( 9 . 3% , Fig 2 ) . Compared to the naïve serum fitness factors , there was a greater proportion of genes involved in replication , recombination , and DNA repair , coenzyme transport , and coenzyme metabolism in the APS . 58 of the 98 genes for fitness in APS ( 59% ) were also important for growth in RPMI alone , while 40 were serum-specific . Of the 40 serum-specific factors , 25 ( 62 . 5% ) were also identified in naïve serum . By excluding the fitness factors important for growth in RPMI , a total of 63 genes were identified as in vitro serum-specific fitness factors ( S4 Table and Fig 1A ) . This group includes members of the chorismate/shikimate pathway ( aroB , aroC , aroE , and aroK ) for synthesis of aromatic amino acids and siderophores , hemin receptors hmuR1 and hmuR2 , the exbBD outer membrane transport proteins along with tonB and tolC , the Rnf electron transport complex ( rnfADEG ) , and numerous metabolic enzymes ( triosephosphate isomerase [tpiA] , glucose-6-phosphate isomerase [pgi] , 6-phosphofructokinase [pfkA] , 6-phosphogluconolactonase [pgl] , and the leucine-responsive transcriptional regulator lrp ) . This category also includes colicin ( cvpA ) , and two components of the “high frequency of lysogenization” locus ( hflCK ) , which has been linked to tolerance of cell membrane and cell envelope stress in other bacterial species [33] . There are also several factors involved in replication , recombination , and repair , such as recA recombinase , DNA polymerase III subunits holD and holC , tyrosine recombinase xerD , and a putative ATP-dependent DNA helicase ( dinG ) . With the in vitro fitness factors characterized , we next analyzed the spleen and liver samples to identify infection-specific fitness factors for experimental bacteremia . In total , 1 , 356 genes ( 36% of the genome ) were identified as candidate fitness factors from spleen samples , 166 of which were also candidate fitness factors from liver samples ( Fig 1B ) . By subtracting out the candidate fitness factors that also exhibited defects in mouse serum ex vivo , there were 1 , 278 candidate infection-specific fitness factors , 1 , 135 of which were specific to spleen samples and 143 were fitness factors for both spleen and liver colonization ( S5 Table ) . There were no liver-specific fitness factors ( as illustrated in Fig 1B ) , indicating that both organs harbor bacteria that have likely undergone the same selection process . In addition to identifying genes with potential fitness defects , Tn-Seq can also reveal if loss of a gene provides a fitness advantage to a bacterium . There were 106 genes that had a fitness advantage in the liver of ≥2-fold compared to the input , but only 2 were statistically significant: a tRNA ( PMIt055 ) and a plasmid gene PMIP18 . In the spleen , 23 genes had a fitness advantage of ≥2-fold compared to the input , but none were statistically significant . Thus , our analysis revealed only 2 genes for which loss may provide P . mirabilis with a fitness advantage during bacteremia . Regarding the infection-specific fitness factors for colonization of both spleen and liver , 71/143 genes were predicted to be contained within operons . We therefore assessed the other genes in each operon for fitness defects in any of the conditions tested ( S5 Table ) . On average , 44% of the genes in each represented operon ( range 7–100% ) were identified as infection-specific fitness factors for both the liver and spleen . This value increases to 78% ( range 33–100% ) for infection-specific fitness factors for the spleen only . Furthermore , only 2/71 operons ( 3% ) contained genes identified as having defects in serum or RPMI in vitro . Thus , this approach appears to accurately identify gene sets that contribute to P . mirabilis fitness during infection but not during incubation in serum in vitro . Of the 143 infection-specific fitness factors for both spleen and liver , 117 ( 82% ) were previously identified as potentially contributing to infection within the urinary tract: 116 were identified as fitness factors during CAUTI infection by Tn-Seq [28] , 29 were identified as being upregulated during ascending UTI [29] , and 8 were identified as fitness factors during ascending UTI [5] . Spleen-specific fitness factors primarily pertained to amino acid transport and metabolism ( 14 . 4% ) , energy production and conversion ( 9 . 6% ) , and carbohydrate transport and metabolism ( 8 . 0% ) functional categories ( Fig 2 ) . Infection-specific fitness factors common to colonization of the spleen and liver pertained to cell wall/membrane/envelope biogenesis ( 16 . 3% ) , amino acid transport and metabolism ( 10 . 6% ) , inorganic ion transport and metabolism ( 7 . 3% ) , and replication , recombination and repair ( 8 . 1% ) ( Fig 2 ) . The fold-change in abundance of mutants representing infection-specific fitness factors in both the spleen and liver ranged from 88 . 8 to 4 . 2 , with 5 of the top 10 fitness factors within the cell wall/membrane/envelope biogenesis category . It is not surprising that survival during bloodstream infection is highly dependent on the integrity of the cell wall , as contact with innate immune defenses is imminent in the bloodstream environment . Examples of infection-specific genes involved in cell wall biogenesis are arnABC , which modifies the charge of LPS and is important for resistance against polymyxin and cationic peptides , as well as the adjacent gene encoded by PMI_RS05085 ( PMI1046 ) that deformylates a component of lipid A . Six members of the waa gene cluster involved in synthesis of the LPS core were also identified as infection specific factors , including three glycosyl transferases ( PMI_RS15630/PMI3163 , PMI_RS15635/PMI3163 , PMI_RS15620/PMI3159 ) . All of these genes were previously implicated as being in vivo fitness factors during both ascending UTI and CAUTI [28 , 34] . In addition , five genes proposed to be involved in capsule biosynthesis were identified as infection specific fitness factors in the spleen ( PMI_RS15765/PMI3188 , PMI_RS15785/PMI3192 , PMI_RS15800-PMI_RS15810/PMI3195-97 ) . The fitness factor with the greatest fold-change in abundance was ompF , an outer membrane porin previously identified as being a fitness factor during CAUTI [28] . PMI_RS16875/PMI3390 , a putative autotransporter about which little is known , was another infection-specific fitness factor that was previously identified as being differentially expressed during ascending UTI as well as a fitness factor for both ascending UTI and CAUTI [28 , 29 , 34] . Notably , the infection-specific fitness factors exhibit a high degree of conservation between P . mirabilis strains . 129 of 143 ( 90% ) infection-specific factors that were identified in strain HI4320 are more than 90% homologous to genes present in 9 other P . mirabilis strains with available complete genome sequences ( Fig 3 and S6 Table ) . Only four genes were completely unique to strain HI4320: a hypothetical membrane protein ( PMI_RS12130/PMI2454 ) , a DNA methyl transferase ( PMI_RS12255/PMI2479 ) , a type II restriction endonuclease ( PMI_RS15540/PMI3141 ) and a putative RNA helicase ( PMI_RS12250/PMI2478 ) . Another seven genes were present in other strains but with less than 90% homology , including RNA polymerase associated protein ( rapA ) , lipopolysaccharide core heptosyltransferase III ( PMI_RS15665/PMI3168 ) , type VI secretion tip protein ( vgrG ) , heptosyl LPS alpha 1 , 3-glucosyltransferase ( waaG ) , vitamin B12 receptor ( btuB ) and two hypothetical proteins PMI_RS15640/PMI3163 and PMI_RS18555/PMIP09 . Thus , there is a remarkable degree of conservation between strains for the genes identified as infection-specific fitness factors in P . mirabilis HI4320 , which has also been observed in the closely related species Citrobacter freundii , with 82% of fitness factors having >90% similarity between isolates [35] . Six mutants were generated in genes identified as infection-specific fitness factors in both the spleen and liver for initial validation of the screen: a polymyxin resistance protein ( arnA , 4/7 genes in this operon were infection-specific fitness factors ) , vitamin B12 transporter ( btuB , not predicted to be part of an operon ) , propanediol utilization protein ( cutC , not predicted to be part of an operon ) , glutamate synthase ( gltB , 2/2 genes in this operon were infection-specific fitness factors ) , arginine decarboxylase ( speA , not predicted to be part of an operon ) , and sec-independent protein translocation ( tatC , 1/3 genes in this operon were infection-specific ) . We also generated mutants for three genes identified as fitness factors in vitro in mouse serum: aspartate-ammonia ligase ( asnA , fitness defect in RPMI and serum ) , colicin ( cvpA , fitness defect in serum ) , and a regulator of the FtsH protease ( hflK , fitness defect in serum ) . Survival in mouse serum in vitro was first assessed to determine if any of the mutants were susceptible to serum killing . These studies utilized naïve mouse serum that had not been heat-inactivated in order to retain heat-labile antimicrobial compounds as a more stringent assessment of survival . All three of the mutants predicted to have defects in vitro recapitulated the Tn-Seq results by exhibiting decreased CFUs compared to WT during the time course ( P<0 . 001 by two-way ANOVA , Fig 4A ) , while none of the candidate infection-specific mutants exhibited growth defects in 50% naïve mouse serum ( Fig 4B ) . Growth in LB medium was next assessed , and only the asnA mutant exhibited a slight defect ( P = 0 . 0095 by two-way ANOVA , S4A and S4B Fig ) . In addition , growth was assessed in the minimal medium PMSM to uncover auxotrophy , and RPMI to recapitulate the Tn-Seq in vitro screen conditions . The candidate infection-specific factors were all able to reach the same level of saturation as WT in PMSM ( S5D Fig ) and RPMI ( S4F Fig ) , although btuB , gltB , speA and tatC all had significant growth delays in RPMI ( P<0 . 001 by two-way ANOVA ) . The asnA mutant was unable grow in PMSM unless supplemented with 10 mM asparagine ( S4C Fig ) and similarly exhibited a severe defect in RPMI ( S4E Fig ) . Unexpectedly , the cvpA mutant was unable to grow in PMSM or RPMI , and the hflK mutant exhibited a defect in RPMI ( P<0 . 007 by two-way ANOVA ) , which was not expected based on the Tn-Seq screen results . Taken together , 7/9 mutants ( 78% ) recapitulated the expected in vitro phenotypes , 1 mutant ( hflK ) recapitulated the expected phenotypes in 3 out of 4 conditions , and 1 mutant ( cvpA ) recapitulated the expected phenotypes in 2 out of 4 conditions . We next assessed the six candidate infection-specific fitness factors by direct co-challenge with WT P . mirabilis during bloodstream infection . Mice were inoculated with a 1:1 mixture of mutant:WT , the infection was allowed to progress for 24 hours , and CFUs of mutant and WT were determined for the liver and spleen of each mouse . The gltB , speA , and tatC mutants all recapitulated the Tn-Seq screen predictions by exhibiting fitness defects in the liver and the spleen ( Fig 5 ) . However , despite having severe defects in the Tn-Seq screen ( 16–41 fold-change in the spleen and 7–23 fold-change in the liver ) , the arnA , btuB , and cutC mutants did not appear to significantly contribute to fitness within the bloodstream during direct co-challenge as none of the mutants were significantly outcompeted by WT in any organ ( Fig 5 ) . This was particularly unexpected for arnA , because disruption of this gene was previously determined to result in a fitness defect in the spleen during CAUTI [28] . Taken together , 3/6 ( 50% ) of the genes chosen for initial validation studies in vivo recapitulated the expected phenotypes . This discrepancy could be due to a combination of factors , including the coverage of TA insertion sites within these genes , the specific gene location where a kanamycin resistance cassette was inserted during generation of TargeTron mutants vs . the transposon insertion sites present in the input pool for Tn-Seq , the fold-change and P value cutoffs used in the analysis , and the 1:1 ratio used for the co-challenge vs . the ~1:50 , 000 ratio present during the Tn-Seq screen . It is also notable that the genes that failed to validate were either not contained within an operon ( btuB and cutC ) , or were within a fairly large operon where only half of the genes were predicted to have fitness defects ( arnA ) . Thus , our initial results clearly highlight the critical importance of validating Tn-Seq results by direct co-challenge , and underscore the utility of exploring the contribution of complete gene operons or functional pathways to pathogenesis . Based on the in vivo defects observed in the tatC , gltB and speA mutants , we therefore chose to investigate the contribution of the twin-arginine translocation system , nitrogen assimilation , and polyamine biosynthesis to P . mirabilis fitness within the bloodstream in greater depth . The twin-arginine translocation ( Tat ) system is utilized by numerous Gram-negative bacterial species to translocate periplasmic proteins and enzymes , particularly those involved in binding redox cofactors [36] . A prior Tn-Seq study in Citrobacter freundii identified the Tat system and putative substrates as important for fitness within the bloodstream [35] . P . mirabilis HI4320 encodes a Tat system ( tatA , tatB , tatC ) with spleen defects ranging from 20–58 fold and liver defects ranging from 11–27 fold . However , only tatC was determined to have a statistically-significant defect ( S5 Table ) . This system was also identified as a candidate fitness factor for kidney colonization during experimental CAUTI [28] , suggesting that it contributes to P . mirabilis fitness during multiple infection types . The Tat-substrate prediction software TatP [37] was used to identify putative Tat substrates encoded by P . mirabilis HI4320 to determine if multiple putative substrates are also likely infection-specific fitness factors . 485 possible Tat substrates were predicted based on potential cleavage sites , 20 of which had clear Tat motifs ( S7 Table ) . Among these 20 candidate Tat substrates , 13 ( 65% ) were candidate infection-specific fitness factors for bacteremia , and one ( sufI ) is a homolog of a predicted Tat substrate that contributed to C . freundii fitness during bacteremia [35] . We therefore sought to further explore the contribution of the Tat system to P . mirabilis fitness in the bloodstream using mutants in tatC and tatA . The tatA mutant was first subjected to the same validation experiments as tatC to determine if both mutants exhibited comparable phenotypes . The tatA mutant had similar growth characteristics to the tatC mutant in vitro , including a pronounced lag phase in LB , PMSM , and RPMI ( P<0 . 05 by two-way ANOVA , S5 Fig ) , and neither mutant exhibited a defect when incubated in naïve mouse serum . During co-challenge vs . WT P . mirabilis in the murine bacteremia model , both tatA and tatC were highly outcompeted by WT in liver and spleen , but the defect for the tatC mutant was more consistent and pronounced than tatA ( Fig 6 ) . This finding is in agreement with the results of the screen , and is consistent with the critical role of TatC in directly interacting with Tat signal peptides . P . mirabilis is well-known for its robust swimming and swarming motility , and mutations that disrupt the Tat system have been shown to perturb motility in Escherichia coli , Yersinia pseudotuberculosis , and C . freundii [35 , 38 , 39] . We therefore investigated the contribution of tatA and tatC to P . mirabilis motility ( Fig 7 ) . Both of the Tat mutants were capable of swarming to a similar extent as WT , but exhibited a minor decrease in the diameter of each swarm ring ( Fig 7A ) . The Tat mutants also both exhibited a dramatic decrease in swimming motility in soft agar ( Fig 7B and 7C ) . Thus , disruption of the Tat system abrogates P . mirabilis swimming motility , but not swarming motility , indicating a potential impact on factors important for chemotaxis rather than synthesis of flagella . Notably , of the 56 genes in the flagella locus of P . mirabilis HI4320 [30] , only 8 ( 14% ) were candidate fitness factors in the spleen during bacteremia , not in the liver . Furthermore , none of the 20 putative Tat substrates with clear Tat motifs were directly related to flagellar biosynthesis , motility , or chemotaxis . It is therefore unlikely that motility contributes substantially to fitness within the bloodstream , and the defects observed for the Tat mutants most likely stem from loss of other secreted substrates , particularly those involved in metabolism . To confirm this hypothesis , we assessed the contribution of flagella to P . mirabilis fitness within the bloodstream by conducting a co-challenge experiment of WT P . mirabilis and a non-motile fliF mutant ( S6 Fig ) . As expected , the fliF mutant did not exhibit a significant competitive defect in the liver or the spleen , indicating that flagella do not contribute to P . mirabilis fitness in a direct bacteremia model . In conclusion , factors secreted through the twin arginine translocation system provide a significant fitness advantage to P . mirabilis during bacteremia that is independent of flagellum-mediated motility . Acquisition of nitrogen is critical for bacterial production of amino and nucleic acids . In general , under conditions where the carbon to nitrogen ratio is low , the GS-GOGAT system ( glnA/gltB ) is utilized to incorporate ammonium via the production of two molecules of l-glutamate , and expression of glnA is activated by the two-component system NtrBC . Conversely , under conditions where the carbon to nitrogen ratio is high , glutamate dehydrogenase ( gdhA ) is favored for nitrogen assimilation through the addition of an amine group onto α-ketoglutarate to generate one molecule of l-glutamate . In the urinary tract , P . mirabilis appears to use one system or the other , but not both [28 , 29] . It is therefore notable that gdhA , glnA , gltB , and ntrB were all identified as fitness factors for spleen colonization ( fold-change 35 . 2 , 19 . 4 , 12 . 7 , and 28 . 3 , respectively ) and gltB also had a defect in the liver ( fold-change 4 . 4 ) ( S5 Table ) . However , glnA was not an infection-specific fitness factor as it was also identified as having a defect during incubation in RPMI alone and in serum . To determine the contribution of these two pathways to growth and survival in the serum environment , we utilized mutants in all four nitrogen assimilation genes . Survival in mouse serum in vitro was first assessed as above , and none of the mutants exhibited growth defects in 50% naïve mouse serum ( Fig 8 ) . Growth in LB was next assessed to determine if any of the mutants exhibit defects in rich medium . Mutants in gltB , gdhA , and ntrB grew similarly to WT , and the glnA mutant exhibited a significant growth delay ( P<0 . 001 by two-way ANOVA ) but reached saturation comparable to WT by 18 hours ( S7A Fig ) . When growth in PMSM minimal medium was assessed , the gdhA mutant exhibited a growth delay ( P<0 . 009 by two-way ANOVA ) but reached saturation comparable to WT , and loss of glnA resulted in glutamine auxotrophy , which could be fully complemented by the addition of 10 mM l-glutamine ( S7B Fig ) . In RPMI , all mutants except for ntrB exhibited significant growth delays , and all except for glnA eventually reached saturation comparable to WT ( S7C Fig ) . To probe the relative contribution of the two nitrogen assimilation pathways to P . mirabilis survival in serum , each nitrogen mutant was cultured independently as well as co-cultured with the WT strain to assess fitness during incubation in naïve mouse serum ( Fig 9A , 9G , 9I and 9K ) . A competitive index was calculated for the ratio of mutant to WT to determine fitness during co-culture ( Fig 9B , 9H , 9J and 9L ) . Consistent with the Tn-Seq results , the glnA mutant exhibited a significant fitness defect during co-culture with WT in vitro ( Fig 9A and 9B ) . Interestingly , this defect could be partially complemented by the addition of either l-glutamine ( Fig 9C and 9D ) or d-glutamine ( Fig 9E and 9F ) , indicating that the defect is likely due to a combination of l-glutamine auxotrophy , defects in peptidoglycan biosynthesis ( d-glutamine ) , and dysregulation of nitrogen assimilation . Loss of gltB or gdhA did not impact fitness in serum during co-culture in vitro ( Fig 9G–9J ) , and loss of ntrB only resulted in a minor defect at 3 hours post-inoculation ( Fig 9K and 9L ) . Thus , glutamine synthetase ( glnA ) is the only nitrogen assimilation factor that appears to contribute to fitness during growth in serum . Selected nitrogen mutants were next cultured in PMSM minimal medium with either glucose or citrate as a carbon source and either ammonium sulfate or l-glutamine as nitrogen sources ( Fig 10 ) . Glutamate dehydrogenase ( gdhA ) was required for optimal growth when ammonium was used as the nitrogen source regardless of carbon source ( P = 0 . 020 for glucose and 0 . 019 for citrate by two-way ANOVA , Fig 10A and 10C ) , and the growth defect of the gdhA mutant was abrogated when l-glutamine was used as the nitrogen source ( Fig 10B and 10D ) . The importance of the GS-GOGAT system was only assessed using gltB , to avoid potential confounding effects from glutamine auxotrophy in the glnA mutant . GOGAT mutants in other bacterial species exhibit growth defects in minimal medium when ammonium is used as the nitrogen source and glucose is used as the carbon source , and their growth defects can be alleviated by switching to a poor carbon source , such as citrate [40–42] . As expected , the gltB mutant exhibited a slight but significant growth defect when glucose was used as the carbon source independent of the nitrogen source ( P = 0 . 029 for ammonium and P = 0 . 034 for glutamine , Fig 10A and 10B ) , and the defect during growth in ammonium was complemented when citrate was the carbon source ( Fig 10C ) . However , citrate was unable to facilitate growth of the gltB mutant when glutamine was provided as the nitrogen source ( P = 0 . 01 by two-way ANOVA , Fig 10D ) . We next determined the contribution of the nitrogen assimilation pathways to fitness during murine co-challenge . All four mutants exhibited fitness defects in vivo in at least one organ ( Fig 11 ) . Notably , glnA , gltB , and gdhA all recapitulated the organ-specific defects predicted by the Tn-Seq results , while ntrB exhibited a defect in the liver but not the spleen , as fitness of the ntrB mutant appeared to follow a bimodal distribution in the spleen . The magnitude of the fitness defects for the nitrogen mutants in vivo in concert with the in vitro defects observed for these mutants indicate that GS-GOGAT is likely the critical pathway for nitrogen assimilation by P . mirabilis within the bloodstream . Thus , the bloodstream likely presents P . mirabilis with relatively low levels of ammonium and poor carbon sources . This is also in agreement with the fact that the concentration of urea in blood is much lower than that in urine , and that urease does not contribute to fitness in the bacteremia model . Polyamines are known to play a critical role in bacterial growth and have been implicated as contributing to cell wall biosynthesis , siderophore biosynthesis , motility , cell signaling , and acid resistance [43 , 44] . Putrescine is a polyamine that can be produced in two ways in P . mirabilis HI4320: 1 ) the decarboxylation of L-arginine to agmatine by arginine decarboxylase ( speA ) followed by the conversion of agmatine into putrescine by agmatinase ( speB ) ; and 2 ) decarboxylation of ornithine to putrescine by ornithine decarboxylase ( speF ) . Putrescine can also be imported via the spermidine/putrescine ABC transporter complex potABCD . All of these genes were identified as infection-specific fitness factors in the spleen ( fold-change 7 . 3–59 . 8 ) , and speAB also exhibited liver defects ( fold-change 13 . 8 and 9 . 9 , respectively , S5 Table ) , although only speA was determined to have a statistically significant liver defect . In addition to its function in putrescine synthesis , speA also contributes to maintenance of membrane potential during arginine catabolism , and both functions have been shown to contribute to colonization of the urinary tract [28 , 44] . To explore the relative importance of arginine degradation and polyamine biosynthesis to fitness during bacteremia , we utilized mutants in speA , speB , speF , and potB . None of the mutants exhibited significant growth defects in LB broth ( S8A Fig ) , and speB was comparable to speA in PMSM , exhibiting a significantly increased lag phase ( P<0 . 001 by two-way ANOVA , S8B Fig ) . In RPMI , all four mutants exhibited increased lag phases , but were ultimately able to reach comparable density as WT at 18 hours ( P<0 . 012 by two-way ANOVA , S8C Fig ) . We next challenged mice with wild-type HI4320 and each of the isogenic mutants as above . Disruption of speA , speB , or potB resulted in a fitness defect in the liver and spleen , indicating that putrescine biosynthesis and import are important for fitness during bacteremia ( Fig 12 ) . The speF mutant did not exhibit a defect , further indicating that P . mirabilis primarily uses the speAB pathway for putrescine biosynthesis , consistent with previous in vitro findings [45] . Notably , the repressor of the arginine biosynthetic pathway ( argR ) was also identified as an infection-specific fitness factor in the spleen ( 33 . 9 fold-change , S5 Table ) , supporting the hypothesis that putrescine biosynthesis is the more critical function of speA than arginine decarboxylation for maintenance of membrane potential . Importantly , the defects in colonization observed during co-challenge with wild-type HI4320 were infection-specific , as co-culture in vitro in 50% mouse serum did not result in differences in CFU of mutant compared to wild-type ( Fig 13A–13F ) . Thus , putrescine import and biosynthesis likely represent infection-specific fitness factors for bacteremia .
Proteus mirabilis HI4320 has been used as a model strain for decades to explore virulence determinants of this unusual bacterial species , particularly for urinary tract infection [2 , 5] . With the availability of the complete genome sequence in 2008 [30] , in vivo transcriptome assessment [29] , and signature-tagged mutagenesis studies [34 , 46 , 47] , much has been learned concerning how this organism adapts to the urinary tract and the virulence factors that contribute to ascending UTI [see [5 , 48 , 49] for review] . Our prior study extended this work by detailing the use of Tn-Seq for identification of fitness factors in a murine model of catheter-associated urinary tract infection ( CAUTI ) , which uncovered genes essential for growth in rich medium and numerous previously unrecognized P . mirabilis fitness determinants , while also highlighting key differences in fitness requirements during ascending UTI versus CAUTI [28] . Several P . mirabilis fitness factors identified through these studies have been tested for contribution to spleen colonization during secondary bacteremia , yet global fitness of P . mirabilis within the bloodstream has never been directly assessed . In the present study , we utilized Tn-Seq to elucidate fitness determinants for bloodstream infection versus factors that contribute to survival in serum in vitro . Through these efforts , we identified 143 candidate infection-specific fitness factors for survival of P . mirabilis within the bloodstream . Nine genes of interest were chosen for an initial validation of the Tn-Seq results: 3 mutants were generated to validate candidate fitness factors for in vitro defects ( asnA , cvpA , and hflK ) and 6 for candidate infection-specific defects ( arnA , btuB , cutC , gltB , speA , and tatC ) . Seven of the nine mutants ( 78% ) recapitulated the expected in vitro results , and 3/6 ( 50% ) recapitulated the expected in vivo fitness defects . Notably , the genes that failed to validate were either not contained within an operon ( btuB and cutC ) , or were within a fairly large operon where only half of the genes were predicted to have fitness defects ( arnA ) . BtuB and CutC were also the only genes pertaining to vitamin B12 transport and choline utilization that were identified as candidate infection-specific fitness factors . Taken together , these data may indicate that arnA , btuB , and cutC were false-positives , which could be due to a variety of reasons including the coverage of TA insertion sites within these genes and the fold-change and P value cutoffs used in the analysis . Alternatively , the failure of these mutants during in vivo validation could pertain to the specific gene location utilized for generation of the TargeTron mutants versus the distinct transposon insertion sites within the genes that exhibited the greatest defects in the initial screen . It is also possible that the mutants only exhibit significant defects when dramatically underrepresented in the input inoculum , such as the ~1:50 , 000 ratio that would be present during the Tn-Seq screen , but not during a 1:1 co-challenge . It is also important to note the myriad of technical limitations of Tn-Seq studies that could result in false hits , and stress the importance of secondary validation of hits from the screen in vivo with clean deletions . One example is the potential under-representation of mutants in a particular gene within the library . The total number of copies of mutants that can be mapped to one gene will be dependent on the chromosomal location ( those closer to the origin have a higher mutation rate due to more copies of sequence near the origin in rapidly replicating cells ) , length of the gene in comparison to number of TA sites ( not all genes have the same number of possible locations for transposon insertion ) , and accessibility of the DNA to transposon mutagenesis ( obstruction of the transposase due to three dimensional structure of the DNA ) . Interruption of a gene with a transposon may also result in a growth delay on LB agar , which would result in fewer copies of a particular mutant when the library is generated compared to those that do not impact growth . Another example of a technical limitation is that our transposon was not designed to prevent polar effects downstream of the insertion site , especially in the last 20% of the gene . Despite these potential limitations , the total validation rate for all mutants assessed under in vitro and in vivo conditions in this study was 12/16 ( 75% ) , which is consistent with other bacteremia Tn-Seq studies ( ranging from 75–86% ) [35 , 50–52] . An additional infection-specific gene ( d-serine ammonia lyase , dsdA ) was also validated as having a fitness defect in the direct bacteremia model in a separate study [53] , bringing the total in vivo validation rate to 10/13 ( 77% ) . It is also notable that the input samples in this study identified 478 genes as essential for growth in rich medium due to lack of insertions , 436 ( 91% ) of which were identified in our prior study [28] . Thus , while one experimental approach involved incubating a single pool of 50 , 000 mutants for 16 hours in LB broth and the other utilized 5 pools of 10 , 000 mutants each incubated separately , the same genes were consistently identified as having a lack of transposon insertions in the input samples . Several of the infection-specific fitness factors identified in this study have been previously implicated as important for P . mirabilis pathogenesis . Tn-Seq of fitness factors in a mouse model of CAUTI identified 116 of our 143 bacteremia hits as contributing to colonization of the bladder and/or kidneys colonization [28] , and eight of these genes were also identified as fitness factors during ascending UTI by signature-tagged mutagenesis [5] . All genes with defects as determined by Tn-Seq that have been previously implicated as having a role in pathogenesis in other studies are demarcated in each of the supplemental tables . Furthermore , transcriptome profiling in the ascending model identified 39 of the infection-specific bacteremia fitness factors as being upregulated during ascending UTI [29] . Altogether , 117 ( 82% ) infection-specific fitness factors for bacteremia have also be implicated as having a role in ascending UTI and/or CAUTI in other genome-wide studies [28 , 29 , 34] . Thus , there may be a core set of fitness factors that P . mirabilis requires for optimal colonization and persistence in a mammalian host regardless of infection route . In addition to the overlap between fitness factors for bacteremia and UTI/CAUTI , seven of the 143 infection-specific fitness factors identified in this screen were previously verified as contributing to secondary bacteremia during co-challenge with the WT strain: gdhA , pta , and cysJ for ascending UTI [29 , 54 , 55] , and arnA , glnA , lon , and argR for CAUTI [28] . Two of these genes ( gdhA and glnA ) both exhibited fitness defects when directly tested in the bacteremia model , while arnA did not . ArnA may therefore play a greater role in fitness of P . mirabilis during dissemination from the kidneys to spleen , or for survival directly within the bloodstream that is not recapitulated by tissue-resident bacterial populations in the bacteremia model . Taken together , these results underscore the critical importance of directly assessing candidate fitness genes identified through Tn-Seq studies , and for considering the importance of the route of inoculation ( e . g . , direct inoculation into the bloodstream , or dissemination to the bloodstream from a secondary route ) . Our study has shed light on metabolic pathways that contribute to P . mirabilis fitness during bacteremia , and highlights key differences between how P . mirabilis adapts to the bloodstream versus the urinary tract . For instance , the urease enzyme is a well-known and critical fitness factor of P . mirabilis , and provides the bacterium with a nitrogen-rich environment through the hydrolysis of urea to ammonium and carbon dioxide . However , in the bloodstream , the urea concentration is 100- to 1 , 000-fold lower than in the urinary tract [31 , 32] . Furthermore , the KM of P . mirabilis urease for urea is only 60 mM , and thus would not catalyze urea hydrolysis well at bloodstream urea concentrations [56] . Consequently , P . mirabilis does not require urease activity for fitness in the bloodstream , and it appears to view the bloodstream as a relatively low-nitrogen environment . Another difference pertains to the role of arginine decarboxylase ( speA ) . In the urinary tract , P . mirabilis appears to predominantly utilize this enzyme for its contribution to tolerance of the mildly-acidic urinary tract and proton motive force [44] , but the role of this gene during bloodstream infection is most likely its contribution to polyamine biosynthesis . Our investigation of P . mirabilis fitness requirements during bloodstream infection also revealed an important role for glucose transport and glycolysis . Several members of the phosphotransferase system were identified as infection-specific fitness factors in the spleen including , ptsG , ptsH , ptsI , treB , nagE , PMI2226 , PMI2982 , and PMI3515 . Several genes involved in glycolysis were also identified as infection-specific fitness factors ( ptsG , gnd , edd , and pykA ) , and genes involved in pyruvate catabolism ( aceEF ) were identified as fitness factors in serum ex vivo and during bacteremia . A previous study in Serratia marcescens also identified ptsI as a fitness factor during bloodstream infection , indicating that glucose transport and metabolism may represent an important fitness factor for survival of other Gram-negative bacteria within the bloodstream [51] . Interestingly , glucose uptake and glycolysis also contribute to P . mirabilis fitness during ascending UTI [29 , 57] , and four genes involved in these pathways were identified as potential fitness factors during CAUTI [28] . Taken together , these results indicate that glucose transport and metabolism are critical for P . mirabilis pathogenesis during bacteremia , and may represent fitness factors for establishment of several different types of infection by this bacterium . Twin arginine translocation ( Tat ) is another fitness factor for multiple types of P . mirabilis infection . In this study , we demonstrate the importance of the Tat system for fitness during bloodstream infection , in addition to previous implications of its involvement during CAUTI [28] . However , it remains to be determined which Tat substrates provide the greatest fitness advantage to P . mirabilis within each infection setting . In E . coli , deletion of tatABC results in a motility defect due to a complete lack of flagellin synthesis [38] . This is not the case for tatA or tatC mutants of P . mirabilis , as they retained flagella-mediated swarming motility despite having lost swimming motility . Furthermore , a fliF mutant that does not produce flagella and is non-motile [34] was able to colonize the liver and spleen to a similar level as WT P . mirabilis , indicating that the defects observed for the Tat mutants stem from loss of motility-independent secreted substrates , such as factors involved in metabolism . Notably , the 143 infection-specific fitness factors identified in this screen are well represented among Proteus species isolates as well as other Gram-negative species that commonly cause bloodstream infections . Specifically , homologs to 14 of the 143 P . mirabilis infection-specific fitness factors were also identified as fitness factors for bacteremia in Acinetobacter baumanii , homologs to 8 were identified in Serratia marcescens , and homologs to 7 were identified in Citrobacter freundii [35 , 51 , 58] . The acr multidrug efflux system was a fitness factor for P . mirabilis bacteremia , and was also identified as contributing to bacteremia in E . coli and S . marcescens [51 , 52] . Similarly , the iron-sulfur cluster transcriptional regulator fur was a fitness factor for bacteremia in P . mirabilis as well as S . marcescens and C . freundii [35 , 51] . There were also several cases where a given gene may not be identified as a fitness factor between species , but other genes within that operon or pathway were identified as fitness factors in the other species . Thus , there are likely shared pathways between Gram-negative bacteria that contribute to survival within the bloodstream . In summary , the use of Tn-Seq as a high-throughput screen has enabled us to investigate the importance of individual genes during P . mirabilis bacteremia , a serious and often fatal complication of CAUTI . By combining assessment of fitness factors for in vivo bacteremia and ex vivo serum survival , we have identified infection-specific fitness factors that contribute to P . mirabilis survival within the bloodstream . Considering that almost 50% of these bacteremia-specific fitness factors have also been implicated as contributing to fitness in the urinary tract , the combined knowledge gained through these studies may uncover core requirements of this multidrug-resistant bacterium for colonization and pathogenesis in a wide range of infection models . These factors would be ideal targets for prevention or treatment of P . mirabilis , particularly in vulnerable populations such as catheterized nursing home residents .
All animal protocols were approved by the Institutional Animal Care and Use Committee ( IACUC ) at the University of Michigan Medical School ( PRO00007111 ) and State University of New York at Buffalo Jacobs School of Medicine and Biomedical Sciences ( MIC31107Y ) , and in accordance with the Office of Laboratory Animal Welfare ( OLAW ) , the United States Department of Agriculture ( USDA ) , and the guidelines specified by the Association for Assessment and Accreditation of Laboratory Animal Care , International ( AAALAC , Intl . ) . Mice were euthanized by inhalant anesthetic overdose ( UM ) or CO2 ( UB ) followed by vital organ removal . Proteus mirabilis HI4320 was isolated in a prior study from the urine of a catheterized patient in a chronic care facility in Baltimore , Maryland [2 , 59] . The P . mirabilis HI4320 transposon mutant library was previously constructed , validated , and successfully utilized in a mouse model of CAUTI [28] . Bacteria were routinely cultured at 37°C with aeration in 5 ml LB broth ( 10 g/L tryptone , 5 g/L yeast extract , 0 . 5 g/L NaCl ) or on LB solidified with 1 . 5% agar . Proteus mirabilis minimal salts medium ( PMSM ) ( 10 . 5 g/L K2HPO4 , 4 . 5 g/L KH2PO4 , 0 . 47 g/L sodium citrate , 1 g/L ( NH4 ) 2SO4 , supplemented with 0 . 001% nicotinic acid , 1mM MgSO4 , and 0 . 2% glycerol ) [60] or RPMI with 2 mM L-glutamine ( Sigma ) . Transposon mutants were cultured in LB containing 25 μg/ml kanamycin ( Sigma ) . Additional P . mirabilis mutants for validation of candidate fitness factors were constructed by insertion of a kanamycin resistance cassette as previously described using the TargeTron system ( Sigma ) , and are listed in S8 Table [61] . Resulting mutants were screened by kanamycin selection and PCR . All primers for generation and verification of mutants are provided in S9 Table . A schematic of the Tn-Seq experimental setup is provided in S2 Fig . Ten mice were inoculated intravenously with the P . mirabilis pool of 50 , 000 transposon mutants to assess fitness factors for survival within the bloodstream , as measured by recovery from spleens and livers 24 hours post-inoculation ( hpi ) . All 10 mice exhibited adequate spleen and liver colonization to be included in the final analysis ( S3 Fig ) . Concurrently , the transposon mutant pool was also subjected to the following in vitro culture conditions , in duplicate , to allow for identification of infection-specific fitness factors: 1 ) RPMI medium , 2 ) RPMI with 50% heat-inactivated naïve mouse serum ( generated from CBA/J mice ) , and 3 ) RPMI with 50% heat-inactivated acute-phase serum ( generated from CBA/J mice 5 hours after intraperitoneal injection with heat-killed P . mirabilis ) . Heat-inactivation of serum was achieved by incubation at 56°C for one hour . Following inoculation , all cultures were incubated statically at 37°C with 5% CO2 for 24 hours . Infection studies were carried out as previously described [62] . For determination of inoculating dose , female CBA/J mice were ( Envigo ) were inoculated by tail vein injection with 100 μl P . mirabilis HI430 suspended in phosphate-buffered saline ( PBS: 0 . 128 M NaCl , 0 . 0027 M KCl , pH 7 . 4 ) to 1x106 , 1x107 , or 1x108 CFU/ml . Mice were euthanized 24 hours post-inoculation ( hpi ) , and organs were harvested into 3 ml PBS . Tissues were homogenized using an Omni TH homogenizer ( Omni International ) , and plated onto LB agar using an Autoplate 4000 spiral plater ( Spiral Biotech ) for enumeration of colonies using a QCount automated plate counter ( Spiral Biotech ) . For bottleneck determination , mice were inoculated with 100 μl containing mixtures of P . mirabilis HI4320 and the ureF mutant ( 1x108 CFU/ml ) . Mice were euthanized and bacterial burden was enumerated as above by plating on plain LB ( total CFUs ) and LB containing kanamycin ( ureF CFUs ) . For the Tn-Seq screen , transposon mutant pools ( 1 ml volume ) were thawed in 9 ml fresh LB with kanamycin and cultured at 37°C for no more than 10 hours . Cultures were centrifuged to pellet , resuspended in PBS , and adjusted to 1x108 CFU/ml . Ten CBA/J mice were inoculated by tail vein injection ( 100 μl of 1x108 CFU/ml for a total inoculum of 1x107 CFU/mouse ) . Mice were euthanized 24 hours post-inoculation ( hpi ) , livers and spleens were harvested into PBS , tissues were homogenized as above , and a 150 μl aliquot was removed and spiral plated for enumeration of colonies . The remaining homogenates were spread plated in their entirety , and colonies were collected , pelleted , and frozen for sequencing . For validation experiments , 5–10 mice were inoculated with a 1:1 mixture of P . mirabilis HI4320 and a mutant of interest , and livers and spleens were homogenized for enumeration of colonies as above . Where indicated , a competitive index ( CI ) was calculated as follows: CI=StrainAoutput/StrainBoutputStrainAinput/StrainBinput Log10CI = 0 indicates that the ratio of the strains in the output is similar to the input , and neither strain had an advantage . Log10CI>0 indicates that strain A has a competitive advantage over strain B . Log10CI<0 indicates that strain B has a competitive advantage over strain A . Sequencing was conducted as described previously [28] . Briefly , genomic DNA was isolated from P . mirabilis in the inputs , serum samples , and the livers and spleens of all mice by hexadecyltrimethyl ammonium bromide ( CTAB ) precipitation [63] . Samples were enriched for transposon insertion junctions as outlined by Goodman et al . [64] . TapeStation analysis was used to confirm concentration and purity , and samples were multiplexed and subjected to V4 single end 50 HiSeq-2500 High-Output sequencing as follows: 1 ) input samples and serum samples ( 2 replicates each ) were multiplexed and sequenced on a single lane; 2 ) 10 spleen samples were multiplexed and sequenced on a single lane; and 3 ) 10 liver samples were multiplexed and sequenced on a single lane . Each lane was spiked with 15% bacteriophage φX DNA to overcome low-diversity sequences . Sequencing was performed at the University of Michigan DNA core facility . The raw sequencing reads are available through the Sequence Read Archive under Study SRP182137: Proteus mirabilis bacteremia TNSeq , and the barcodes associated with each unique sample are provided in S10 Table . Mapping to the P . mirabilis HI4320 chromosome and plasmid sequences ( NCBI accession numbers NC_010054 and NC_010555 ) [30] was conducted using a modification to the Goodman In-Seq pipeline [64] as previously described [28] . Individual genes were only assessed for fitness contribution if the mean of the sum of insertion-site reads was >1000 and the number of insertions in that gene was >5 , to reduce potential over-estimation of fitness factors . The fitness contribution of each gene was then estimated as previously described [28] using an R package called TnseqDiff [65] , which can be installed from the Comprehensive R Archive Network ( CRAN ) . Significant genes for further analysis were selected based on an adjusted P-value <0 . 05 and >2-fold ratio of output over input . To identify homologs of infection-specific fitness factors first , a file was generated containing the FASTA sequences of each fitness factor , then the proteomes of other P . mirabilis strains available on PATRIC were compared to this sequence using the PATRIC proteome comparison tool [66] . For this comparison the sequence identity was limited to ≥10% over a minimum of 30% sequence . Overnight cultures of P . mirabilis mutants were washed once in PBS and diluted 1:100 in growth medium . Where indicated , PMSM was supplemented with 10 mg/mL of l-glutamine , d-glutamine , or l-asparagine . Carbon and nitrogen sources in PMSM were also adjusted as follows: carbon sources ( 0 . 2% glycerol , 0 . 2% citrate , or 0 . 2% glucose ) , nitrogen sources ( 0 . 002% or 0 . 2% ammonium sulfate or l-glutamine ) . A BioTek Synergy H1 96-well plate reader was utilized to generate growth curves . Cultures were incubated at 37°C with continuous shaking , and OD600 readings were taken every 15 min for 18 h . Serum growth curves were performed by inoculating 50% naïve mouse serum ( Innovative Research ) with 1x106 CFU of PBS washed , overnight cultures of either an individual mutant or a 1:1 mixture of wild-type P . mirabilis and a mutant . Inoculated serum was incubated statically at 37°C with 5% CO2 . Aliquots were taken at the time of inoculation and at indicated timepoints , diluted , and plated onto LB agar with and without kanamycin to determine CFUs of mutant and wild-type . Competitive indices for growth in serum were calculated as described above for murine infection studies . The Tat-substrate prediction software TatP [37] was used to identify Tat motifs and probable cleavage sites using the P . mirabilis HI4320 chromosome sequence ( NCBI accession number NC_010054 ) [30] . Swimming motility agar plates ( MOT: 10 g/L tryptone , 0 . 5 g/L NaCl , 3 g/L agar ) were stab-inoculated with an overnight culture of P . mirabilis HI4320 or isogenic mutant . MOT plates were incubated without inverting at 30°C for 18 hours prior to measurement of swimming diameter . Swarm agar refers to LB agar containing 5 g/L NaCl , and swarming was assessed by inoculating 5 μl of an overnight culture of P . mirabilis HI4320 or isogenic mutant onto the surface of a swarm plate , allowing the inoculum to soak in for ~10 minutes , and incubating at 37°C for 18 hours prior to measurement of the diameter of each swarm ring . Significance was assessed using Student’s t-test , two-way analysis of variance ( ANOVA ) with post-hoc multiple comparisons test , and Wilcoxon signed-rank test . These analyses were performed using GraphPad Prism , version 7 ( GraphPad Software , San Diego , CA ) . All P values are two tailed at a 95% confidence interval . | Proteus mirabilis is a common causative agent of catheter-associated urinary tract infections ( CAUTI ) , one of the most prevalent healthcare-associated infections . Multidrug-resistant isolates of P . mirabilis are increasingly common and pose a severe challenge for treatment of CAUTI and secondary bloodstream infections , a common complication of CAUTI . However , there is a knowledge gap regarding the pathogenesis of P . mirabilis during bacteremia . We utilized a library of transposon mutants coupled with next-generation sequencing to perform a genome-wide assessment of the fitness requirements of P . mirabilis during incubation in serum ex vivo compared to during experimental bacteremia . This approach led to identification of a cohort of genes that are specifically important for establishing infection in both the liver and the spleen in vivo , several of which have also been implicated in secondary bacteremia following experimental CAUTI . Further exploration of genes critical for bloodstream pathogenesis could give rise to targets for novel antimicrobial therapies and potentially vaccines . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"spleen",
"pathogens",
"microbiology",
"operons",
"animal",
"models",
"bacterial",
"diseases",
"model",
"organisms",
"experimental",
"organism",
"systems",
"ge... | 2019 | Twin arginine translocation, ammonia incorporation, and polyamine biosynthesis are crucial for Proteus mirabilis fitness during bloodstream infection |
In order to evaluate the role of persisting virus replication during occult phase immunisation in the live attenuated SIV vaccine model , a novel SIVmac239Δnef variant ( SIVrtTA ) genetically engineered to replicate in the presence of doxycycline was evaluated for its ability to protect against wild-type SIVmac239 . Indian rhesus macaques were vaccinated either with SIVrtTA or with SIVmac239Δnef . Doxycycline was withdrawn from 4 of 8 SIVrtTA vaccinates before challenge with wild-type virus . Unvaccinated challenge controls exhibited ~107 peak plasma viral RNA copies/ml persisting beyond the acute phase . Six vaccinates , four SIVmac239Δnef and two SIVrtTA vaccinates exhibited complete protection , defined by lack of wild-type viraemia post-challenge and virus-specific PCR analysis of tissues recovered post-mortem , whereas six SIVrtTA vaccinates were protected from high levels of viraemia . Critically , the complete protection in two SIVrtTA vaccinates was associated with enhanced SIVrtTA replication in the immediate post-acute vaccination period but was independent of doxycycline status at the time of challenge . Mutations were identified in the LTR promoter region and rtTA gene that do not affect doxycycline-control but were associated with enhanced post-acute phase replication in protected vaccinates . High frequencies of total circulating CD8+T effector memory cells and a higher total frequency of SIV-specific CD8+ mono and polyfunctional T cells on the day of wild-type challenge were associated with complete protection but these parameters were not predictive of outcome when assessed 130 days after challenge . Moreover , challenge virus-specific Nef CD8+ polyfunctional T cell responses and antigen were detected in tissues post mortem in completely-protected macaques indicating post-challenge control of infection . Within the parameters of the study design , on-going occult-phase replication may not be absolutely required for protective immunity .
Live attenuated SIV has proven to be a highly effective vaccination strategy in non-human primate ( NHP ) models of HIV/AIDS [1 , 2] , in many cases protecting macaques from detectable superinfection following re-challenge with both homologous and heterologous wild-type SIV administered systemically and mucosally [3–23] . Although safety concerns such as reversion to virulence and recombination with wild-type strains preclude direct application of this vaccine approach in humans , a clearer understanding of mechanisms of pathogenesis and protection may inform the development of more clinically acceptable HIV vaccines . Studies have been performed using vaccine viruses attenuated by genetic disruption of key regulatory genes including nef , vpx , vpr and vif; although the moderately attenuated prototypic vaccine strain SIVmac239Δnef has been used for the majority of studies . Attempts to establish clearly defined immune correlates of protection have not been conclusive , particularly where studies have measured responses in peripheral blood . Indeed , the only robust correlate identified so far is the observation between increasing attenuation of the vaccine virus and decreasing protection [11] . Recently , a detailed comparative study of different attenuated virus strains derived from SIVmac239 concluded that protection was associated with the induction of an effector memory T cell ( TEM ) response and protection of the T follicular helper ( TFH ) cell subset in lymphoid tissue [10] . This association , however , is not definitively established as the mechanism of protection . A crucial property of minimally-attenuated SIV vaccines , which are the most effective , is the widespread distribution of the vaccine virus in multiple lymphoid tissues [22] but the role of occult replication ( i . e . replication in lymphoid tissue when virus is no longer or only intermittently detected in the peripheral circulation ) in the generation of protective immunity is not fully understood . Vaccine virus persistence may result in multiple alterations in the host innate immune system that contribute to protection , in addition to the induction of adaptive immune responses [22] . In the study reported here we have sought to influence occult phase persisting turnover of live attenuated SIV using a novel approach: the conditionally live attenuated SIVmac239Δnef vaccine ( SIVrtTA ) that in vitro is absolutely dependent on the presence of doxycycline ( dox ) to replicate [24 , 25] . Previously , we have shown that SIVrtTA is infectious in Indian rhesus macaques and induced reversible up-regulation of the frequency of global circulating TEM [26] . Here , we report the outcome of an intravenous challenge of two groups of SIVrtTA-vaccinated macaques with wild-type SIVmac239 in comparison with macaques vaccinated with the prototypic SIVmac239Δnef live attenuated vaccine . One group of SIVrtTA vaccinates macaques remained on daily administration of dox , whereas another group received the final dose of dox 8 weeks prior to wild-type virus challenge during the occult phase of virus replication . Protection against detectable infection with wild-type , highly virulent SIVmac239 was observed at various levels; however , the pattern of protection did not associate directly with the experimental treatment protocol , but with the kinetics of vaccine-virus replication in the acute and immediate post-acute period of vaccine viraemia and with vaccine-driven T cell immune responses .
Two groups ( A & B ) of four Indian-derived rhesus macaques were injected intravenously with 5 x 103 TCID50 SIVrtTA vaccine ( genetically engineered from the SIVmac239 backbone as indicated in Fig 1A ) and treated with dox for 6 months followed by a period of 8 weeks without dox ( Group A; E61 , E63 , E65 , E66 ) or treated with dox for 6 months and then maintained on dox ( Group B; E67 , E68 , E70 , E71 ) . A further 4 macaques ( Group C; E73 , E75 , E76 , E77 ) were vaccinated with SIVmac239Δnef for 6 months and four unvaccinated , naïve macaques ( E79-E82 ) were included as challenge controls ( Fig 1B ) . Total SIV gag vRNA profiles are shown for Groups A-C as a continuum of vaccination and wild-type challenge profiles ( Fig 1C ) . As previously reported [26] , the SIVrtTA vaccinates displayed a transient peak in plasma vRNA kinetics which is characteristic for attenuated SIVmac239Δnef with two exceptions: E65 ( Group A ) and E70 ( Group B ) . These animals exhibited a persisting shoulder of ~ 102 vRNA copies/ml to ~100 days post-vaccination . From day 110 post-infection ( p . i ) , prior to removal of dox , E65 plasma vRNA fell below the limit of detection . Plasma vRNA remained stably elevated in E70 , which was maintained on daily dox treatment to the time of wild-type challenge . Another macaque , vaccinated with SIVmac239Δnef ( E76 , Group C ) also failed to completely control viraemia below the limit of detection . Hence , at the time of wild-type SIVmac239 challenge detectable vRNA signals were present in the plasma of vaccinates E65 , E70 ( SIVrtTA ) and E76 ( SIVmac239Δnef ) ( Fig 1 , S1 Fig ) . Challenge outcome was initially assessed by the individual comparison of total SIV gag plasma vRNA profiles for each group ( Fig 1C ) . As expected , all four naive , unvaccinated control macaques challenged with wild-type SIVmac239 ( Group D ) exhibited high plasma vRNA loads ( 1 . 86 x107 mean SIV RNA copies/ml ) by day 14 p . i . which also exhibited a high vRNA steady state ( 106–108 wtSIVmac239 copies/ml ) throughout the 20 week follow-up period . In contrast , a marked vaccine effect was observed in all animals vaccinated with either SIVrtTA or SIVmac239Δnef ( Fig 1C , Groups A and B or C respectively ) . Statistical analyses of suppression of vRNA levels post wild-type challenge were determined using a Kruskal-Wallis analysis with a Dunn’s post-hoc test to determine significance levels . Peak viraemia was statistically significantly suppressed when Groups A , B and C were compared individually to Group D , although the vaccination with SIVmac239Δnef resulted in the most significant outcomes . P values for Groups A-C were , p = 0 . 05 , p = 0 . 05 and p = 0 . 001 respectively at day 14 . When viraemia levels were analysed at day 84 ( steady-state ) significance was retained in Groups A and C ( p = 0 . 012 and p = 0 . 002 respectively ) although , interestingly , significance was lost at day 84 in Group B ( p = 0 . 135 ) . All SIVrtTA vaccinates analysed together ( A and B combined ) exhibited significant differences from Group D challenge controls at peak ( day 14 ) and steady-state ( day 84 ) time-points ( p = 0 . 025 and 0 . 021 respectively ) . E73 , E76 and E77 ( vaccinated with SIVmac239Δnef ) and E65 ( vaccinated with SIVrtTA ) exhibited plasma vRNA levels that remained <100 SIV RNA copies/ml . Macaques E76 ( vaccinated with SIVmac239Δnef ) and E70 ( vaccinated with SIVrtTA ) exhibited plasma viral loads of the vaccine virus higher than 100 SIV RNA copies/ml prior to wild-type SIV challenge , with vRNA viral loads gradually increasing in the 20 week follow-up period . Of the remaining macaques vaccinated with SIVrtTA , with undetectable plasma vRNA on the day of challenge ( E61 , E63 , E66 , E67 , E68 , E71 ) , a significant peak in plasma viremia was detected 14 days after wild-type challenge ( mean 5 . 43 x 105 SIV RNA copies per ml ) , which was partially resolved , but remained between 1 x 103–1 x 106 SIV RNA copies/ml 20 weeks post-challenge . From these initial analyses it was possible to classify macaques into two levels of protection: ( 1 ) complete-protection defined as having no secondary peak after wild-type virus challenge ( E65 , E70: SIVrtTA; E73 , E75 , E76 , E77: SIVmac239Δnef ) . ( 2 ) partially-protected macaques that exhibited a clear secondary peak of viraemia 14 days post SIVmac239 wild-type challenge ( E61 , E63 , E66 , E67 , E68 , E71; SIVrtTA ) . When these data were re-plotted , also as a continuum , two patterns of plasma vRNA profiles were revealed immediately prior to and post wild-type SIVmac239 challenge , reflecting these two general classifications of protection as represented in Fig 2 . Interestingly , in the partially protected group , the secondary spike in vRNA is immediately preceded by a virtual absence in detectable vaccine-virus replication prior to wild-type challenge . By comparison , in the completely protected group , total SIV gag vRNA signals are clearly evident in the same period ( ~100 days ) up to challenge with little perturbations in these levels post-challenge . However , as total plasma SIV RNA levels reveal only part of the overall biomarker of infection picture , discriminatory PCR assays were required to fully evaluate the protection status of each macaque . To discriminate further between superinfection with wild-type virus and recrudescence/persistence of vaccine virus , discriminatory PCR assays were established that selectively detect either vaccine-derived or wild-type vRNA in plasma , total vDNA signals in tissues or cell-associated viral RNA ( CA-RNA ) . From these combined analyses a clear picture of superinfection status emerged with the ability to detect and quantify each viral nucleic acid species in blood and/or selected lymphoid tissues ( Figs 2 and 3; S2 Fig ) . Wild-type SIVmac239-specific vRNA determinations partitioned macaques into completely protected ( E65 , E70 , E73 , E75 , E76 , E77 ) or partially protected ( E61 , E63 , E66 , E67 , E68 , E71 ) as indicated in Fig 3A . There was a highly statistically significant difference between completely protected macaques and naïve challenge controls ( p<0 . 001 at days 14 and 84 post SIVmac239 wild-type challenge ) using a Kruskal-Wallis analysis with a Dunn’s post-hoc test . Partially-protected macaques all demonstrated a spike in plasma vRNA that was unambiguously attributed to establishment of wild-type virus infection which at days 14 and 84 were non-significant ( p = 0 . 095 ) compared to challenge controls , applying the same statistical test as for the completely protected group . Additionally , a broad range of lymphoid tissues was assessed for wild type SIV DNA ( S2 Fig ) . High levels were detected in all tissues in naïve challenge controls , with lower but detectable levels in most tissues in E61 , E63 , E66 , E67 , E68 and E71 , reflecting profiles of plasma viral RNA . No wild type SIV DNA was detected in any tissue from the completely-protected animal vaccinated with SIVrtTA ( E65 ) and a single signal of wild-type SIVmac239 DNA detected in the spleen of E70 . No wtSIVmac239-specific DNA was detected in animals of Group C vaccinated with SIVmac239Δnef . Additional information relating to the ability to detect apparently replication-competent virus , rather than proviral signals , was gained for a number of tissues by measuring CA-RNA concentrations for vaccine and wild-type viruses 20 weeks after SIVmac239 challenge ( Fig 3B ) . Wild-type SIV was never detected by any molecular biomarker of infection in those macaques vaccinated with SIVmac239Δnef ( E73 , E75 , E76 , E77 ) , further confirming the complete protection status of this group . No wtSIVmac239 CA-RNA was detected in macaques E65 and E70 vaccinated with SIVrtTA , compared with high levels of wtSIVmac239 detected in all naïve challenge controls , particularly in the spleen and mesenteric lymph nodes ( MLN ) . Lower levels of wtSIVmac239 CA-RNA were detected in spleen samples from E61 , E63 , E66 , E67 , E68 and E71 and more sporadically from MLN and peripheral lymph node ( PLN ) samples ( Fig 3B ) . Hence , with information gained from virus-specific differential PCR techniques , taking only wild-type SIVmac239 levels as measures of outcome , there was a statistically significant difference in outcome between completely protected macaques and wild-type challenge controls and partially protected and wild-type challenge controls . Although E65 and E70 displayed undetectable signals for wild-type specific plasma and CA-RNA , both vaccinates signalled positive by SIVrtTA-specific RT-PCR , particularly E70 in the plasma , spleen and PLN ( Fig 3B and 3C ) . These data reflect the plasma vRNA signal in E70 post-wtSIV239 challenge which was unambiguously attributable to continuous SIVrtTA replication in the continued presence of dox . Remarkably , SIVrtTA replication did not fluctuate over time in this macaque , nor was it perturbed by administration of the wild-type challenge virus ( Figs 1C , 2 and 3 ) . In this respect , E70 was comparable to macaque E76 ( Group C; SIVmac239Δnef ) that displayed similar continuous viral kinetics post SIVmac239 challenge despite resistance to wild-type superinfection as confirmed by lack of wild-type SIVmac239 RNA signals in either plasma or tissues . Perhaps the most interesting vaccinate of all groups was SIVrtTA-vaccinated macaque E65 , which resisted wtSIVmac239 but displayed highly controlled vRNA kinetics in the later post-acute phase . However , in the absence of dox , four blips of plasma vRNA were noted as determined by total SIV-gag qPCR ( Figs 1C and 2 ) , two prior to wild-type challenge but after dox removal and two blips after wild-type challenge . Analysis of tissues for CA-RNA indicated low , but clearly detectable SIVrtTA in the PLN at termination . Taken together , these data suggest evidence of very low , but persistent replication of SIVrtTA in E65 when there was no , or little , dox present . Moreover , both SIVrtTA vaccinates E65 and E70 had detectable levels of SIVrtTA-specific CA-RNA at termination , many weeks after initial vaccine administration . Extending these observations to Group C vaccinates ( SIVmac239Δnef ) all had some level of residual detectable vaccine virus replication at termination ( Fig 3 ) . Indeed , all 6 completely protected vaccinates signalled positive for the vaccine virus post-mortem in PLN suggesting this to be an important site for virus sequestration , which as well as the spleen represents an important reservoir for the vaccine virus . All vaccinates seroconverted to SIV Gag p27 prior to challenge with wild-type SIV ( S3 Fig ) . Anti-p27 responses were broadly similar amongst all macaques vaccinated with SIVrtTA regardless of dox withdrawal and anti-p27 titres were lower than those in SIVmac239Δnef vaccinates . All fully protected animals , E65 and E70 vaccinated with SIVrtTA and E73 , E75 , E76 and E77 vaccinated with SIVmac239Δnef , showed only minor perturbations in anti-p27 titre after challenge with wild-type SIVmac239 , whereas a marked increase in anti-p27 titres was detected in all other macaques ( S3 Fig ) . In order to address the possibility that mutations arising in SIVrtTA as a result of selection in vivo may have occurred , SIVrtTA RNA recovered from vaccinates was sequenced . For this , plasma vRNA was isolated at several times during the immediate post-acute phase period , when qRT-PCR revealed a vRNA load of >102 SIV RNA copies/ml including where there was the persisting shoulder of prolonged SIVrtTA replication in E65 and E70 SIVrtTA vaccinates . In SIVrtTA , the Tat-TAR transcription activation mechanism has been functionally replaced by the dox-inducible Tet-On gene expression system [24 , 25 , 27] . To achieve this ( 1 ) TAR was inactivated through mutations in the binding sites for Tat and pTEFb , ( 2 ) the gene encoding the dox-inducible rtTA transcriptional activator was inserted at the site of the accessory nef gene and ( 3 ) tet operator ( tetO ) elements to which the dox-rtTA complex can bind were inserted between the NFκB and Sp1 binding sites in the U3 domain of the LTR promoter ( Fig 1A ) . Sequencing of the LTR and leader RNA region of different SIVrtTA RNA samples demonstrated the stable presence of the TAR-inactivating mutations and no additional changes were observed in TAR . The virus also stably maintained the tetO elements but whereas the vaccine strain contained a triplicated NFκB-tetO repeat ( resulting from in vitro evolution; [28] ) , deletion of one of these repeats was frequently observed ( S1 Table; S4 Fig ) . Previous experiments demonstrated that such a deletion slightly reduces the transcriptional activity of the LTR promoter , but does not affect dox-control . In all macaques , a point mutation was observed in the primer binding site ( PBS ) sequence ( T731C ) . This mutation was due to the fact that the SIVrtTA vaccine construct contained a PBS complementary to the infrequently used tRNAlys5 primer for reverse transcription [29 , 30] . As expected , the in vivo replicating virus demonstrated a PBS sequence corresponding to the more frequently used tRNAlys3 primer . Sequencing of the tat gene did not reveal any sequence changes . However , sequence analysis of the rtTA gene identified two non-silent codon changes ( R80W and E191K ) in the E65 samples isolated at 6 weeks after vaccination and later ( S1 Table ) . The E70 sample isolated at 6 weeks after vaccination demonstrated an R80Q change , whereas later E70 samples ( from 14 weeks ) also demonstrated the R80W substitution . We did not identify such rtTA mutations in the other macaques vaccinated with SIVrtTA . The identified amino acid changes had never been observed previously in multiple long-term in vitro evolution experiments with SIVrtTA or with a similar dox-controlled HIVrtTA variant and hence represents a novel finding . Testing the transcriptional activity of the new R80W and E191K rtTA variants demonstrated that the mutations did not increase the background activity in the absence of dox ( no loss of dox control ) nor significantly alter the dox-induced activity ( S5 Fig ) . As both E65 and E70 showed prolonged SIVrtTA replication , the mutations may improve in vivo replication of the virus . Importantly , these results demonstrate that the in vivo replicating virus stably maintains the integrated dox-control mechanism and did not restore the Tat-TAR axis of transcription control . Since TRIM5α status and MHC type may influence vaccine challenge outcome [31 , 32] , the TRIM5α/TRIMcyp status and MHC type of all study macaques was determined ( S2 Table ) . While no direct associations were identified between either MHC or TRIM5/cyp status and outcome it is interesting to note that the two macaques which failed to control the vaccine virus ( E70: SIVrtTA; E76: SIVmac239Δnef ) and were protected from wild-type SIVmac239 did not express any of the major mamu A alleles analysed ( S1 Fig ) . In this study , we could not identify any confounding factors associated with TRIM5α or TRIMcyp genotype . We have previously reported that under replication permissive conditions , during the period when live attenuated virus RNA was essentially below the limit of detection in plasma , the global circulating T effector memory ( TEM ) cell frequency was upregulated [26] . Hence , we were interested to determine if this effect was associated with the degree of protection from superinfection . Comparison of partially and completely protected macaques on the day of challenge revealed that for both CD4+ and CD8+ CD95+ T cells , completely protected macaques had a lower median frequency of TCM and reciprocally a higher median frequency of CD28- CCR7- TEM than partially protected macaques; however , the difference in TEM frequencies between these groups only reached significance in CD8+ T cells ( ρ = 0 . 026; Mann-Whitney rank sum test ) ( Fig 4 ) . Comparison with results from naïve macaques showed that median frequencies of both CD4+ and CD8+ TCM were significantly reduced in completely protected macaques ( ρ = 0 . 003 and ρ = 0 . 008 respectively; Mann-Whitney rank sum test ) . Conversely , the frequencies of CD4+ and CD8+ TEM ( CD28- CCR7- ) were significantly elevated in completely protected macaques compared with naïve macaques ( ρ = 0 . 002 and ρ = 0 . 007; Mann-Whitney rank sum test ) . Despite these differences at the population level exceptions were seen: T cell frequencies for macaque E65 , challenged under conditions of dox withdrawal were similar to those for naïve or partially protected macaques . Conversely , partially protected macaque E67 challenged under dox maintenance had a high frequency of CD28- CCR7- CD8+ TEM and partially protected macaque E71 also challenged under replication permissive conditions had relatively high frequencies of both CD28- and CD28+ , CCR7- CD95+ CD8+ T cells . So , although there was an association between a high frequency of global CD8+ TEM in the circulation on the day of challenge and complete protection , a high TEM frequency alone was not predictive of complete protection status . The phenotype of circulating T cells was examined again at day 130 following superinfection challenge . At this time-point no significant differences were found between partially and completely protected macaques; moreover , global circulating CD4+ memory T cell populations were significantly perturbed ( Fig 4 ) . CD4+ TCM were significantly elevated in superinfection-challenged animals , regardless of protection status compared with frequencies in naïve macaques ( ρ = <0 . 001; Mann-Whitney rank sum test ) . Likewise , comparison of TCM frequencies on the day of challenge with day 130 post-challenge showed significantly elevated frequencies regardless of protection status ( ρ = <0 . 031; Wilcoxon signed rank test ) . The median frequencies of CD4+ CD28+CCR7- ( intermediate ) T cells remained elevated post challenge compared with naïve animals ( ρ = 0 . 003 and ρ = <0 . 001 for partially and completely protected groups respectively; Mann-Whitney rank sum test ) and showed no statistical difference between day of challenge and day 130 post challenge for either group . In contrast , the median frequencies of TEM were significantly reduced 130 days after challenge compared with those in naïve animals; although this was most marked in completely protected animals ( ρ = 0 . 049 and ρ = 0 . 006 , partially and complete protection groups respectively; Mann-Whitney rank sum test ) . Similarly , pairwise comparison of completely protected animals revealed a significant reduction in TEM cell proportions between the day of challenge and 130 days post challenge ( ρ = 0 . 031; Wilcoxon signed rank test ) . Five of 6 partially protected macaques also had lower frequencies at day 130 post challenge ( ρ = 0 . 063; Wilcoxon signed rank test ) . A somewhat different pattern of perturbation in circulating CD8+ T memory cell populations was seen following superinfection challenge ( Fig 4 ) . These changes were again , as for CD4+ T cells , independent of superinfection status . Frequencies of TCM and TEM ( CD28-CCR7- ) were not significantly different from frequencies in naïve animals; whereas CD28+CCR7- cell frequencies were significantly elevated compared with naïve macaques for both partially and completely protected macaques ( ρ = 0 . 014 and ρ = 0 . 007 respectively; Mann-Whitney rank sum test ) and were not significantly different from day of challenge frequencies . Five of 6 completely protected macaques , the exception being macaque E65 , had elevated TCM and reduced TEM ( CD28-CCR7- ) at day 130 post-challenge compared with day of challenge but failed to reach statistical significance ( p = 0 . 063 Wilcoxon signed rank sum test ) . Pairwise comparison of TCM and TEM frequencies at day 130 post-challenge and day of challenge in partially protected macaques showed no significant changes . Thus , the polarisation of circulating CD8+ T memory populations observed in completely protected macaques on the day of challenge was not evident 130 days after superinfection challenge . In order to evaluate the possible influence of SIV-specific T cell quantity and quality on protection status , PBMC were stimulated in vitro with peptide pools from SIV Gag , Rev and Tat and intracellular cytokine staining for IL-2 , IFN-γ , TNF-α and IL-17 was analysed by flow cytometry for CD4+ and CD8+ T cells . The total frequency ( ie mono + bi + tri + quadruple ) of SIV-specific CD8+ T cells was found to be significantly higher in completely protected compared with partially protected macaques on the day of challenge ( ρ = 0 . 041; Mann-Whitney rank sum test ) ; whereas no difference was seen with CD4+ cells ( Fig 5 ) . A similar analysis at day 130 after challenge failed to show a difference between groups for either CD8+ or CD4+ T cells ( S6 Fig ) . It was noted , however , that the frequency of CD4+ T cells was markedly elevated in both groups regardless of protection status when compared with day of challenge and was statistically significantly different for completely protected animals ( p = 0 . 063 for partially protected and p = 0 . 031 for fully protected animals; Wilcoxon signed rank test ) . In only one animal , E61 , were frequencies similar on the two occasions tested ( 3 . 57% and 3 . 51% , day of challenge and day 130 post challenge respectively ) and were largely confined to mono-functionality ( see below ) . Although total frequencies of SIV-specific CD8+ T cells also showed an upwards trend at day 130 after challenge the differences were not statistically significant . Deconvolution of cytokine combinations showed that on the day of superinfection challenge 6/6 completely protected animals had circulating SIV-specific quadruple cytokine expressing CD8+ cells at a frequency of >0 . 02% compared to only 1/6 partially protected macaques ( ρ = 0 . 015; Fisher’s exact test ) . Differences in median frequencies of maximally polyfunctional CD8+ T cells between the groups did not reach statistical significance ( ρ = 0 . 065; Mann-Whitney rank sum test ) due to the outlier E71 ( Fig 6 ) . Similarly , IFN-γ + TNF-α dual positive CD8+ T cells were absent or below 0 . 02% in partially protected animals whereas in the completely protected group 5/6 macaques had frequencies markedly above 0 . 02% ( ρ = 0 . 015; Fisher’s exact test ) with a significantly elevated median frequency ( ρ = 0 . 026; Mann-Whitney rank sum test ) . Significant differences were not seen for any cytokine combination 130 days after challenge ( Fig 7 ) . No significant differences were seen between partially and completely protected groups in the frequencies of circulating SIV-specific CD4+ T cells expressing individual cytokine combinations at either the day of superinfection challenge ( S7 Fig ) or 130 days after challenge ( S8 Fig ) . Although it was not possible to discern protection status-specific differences in circulating CD8+ T cell responses 130 days after wt-challenge , responses in lymphoid tissue may be more informative . Mononuclear cells extracted from mesenteric lymph nodes at necropsy were stimulated in vitro with a pool of Nef unique region-specific peptides . SIVrtTA and SIVmac239Δnef vaccine strains do not produce Nef protein , whereas the SIVmac239 challenge virus expresses full-length Nef . Surprisingly , poly and mono-functional CD8+ T cells were detected regardless of protection status ( Fig 8 ) . Although statistically different frequencies of functional cells were not detected between the groups , there was a trend towards higher reactivity in completely protected animals . Sections of spleen from vaccinates , naïve challenge controls and unchallenged macaques were stained with KK77 monoclonal antibody specific for Nef ( Fig 9 ) . Positive cells were detected in partially-protected SIVrtTA vaccinates as well as fully-protected macaques E65 and E70 . In contrast , macaques of Group C vaccinated with SIVmac239Δnef were indistinguishable from negative controls . Although clearly detectable staining for Nef was present in E65 , the staining pattern was more diffuse with occasionally identifiable foci of positive cells , as distinguished from productively infected macaques which were partitioned into the partially protected group .
The reported breadth and duration of protection conferred in macaques following vaccination with live attenuated SIV has many of the features required of an effective vaccine against HIV/AIDS . Understanding the mechanisms of protection may allow the informed design of intrinsically safe vaccines . Earlier attempts to improve the safety profile of live attenuated SIV by introducing multiple attenuating mutations revealed that the degree of protection was inversely proportional to the degree of attenuation [11] . Hence , it was perhaps not unexpected that SIV clones molecularly engineered to be limited to a single round of replication conferred only limited protection compared with more vigorously replicating attenuated vaccine strains [23 , 33] . The development of SIVrtTA with potential to be temporally modulated for replication in vivo provides a novel tool to further dissect the processes of protection elicited by live attenuated SIV . Previously , we reported this novel virus to replicate in vivo and being fully infectious in rhesus macaques , with the ability to disseminate to lymphoid tissues and elicit a range of immunological responses including reversible changes in the frequency of memory T cell subsets dependent upon the withdrawal of dox [26] . Here , we report that vaccination with SIVrtTA confers protection against homologous wild-type challenge , in some cases similar to the ‘gold-standard’ SIVmac239Δnef vaccine . However , levels of protection were variable . Full or complete protection ( based on absence of a wild-type post-challenge viraemia ) was associated with a prolonged shoulder of persisting SIVrtTA vRNA signal in plasma during the dox-on period rather than the later modulation of replication in lymphoid tissues ( occult replication ) mediated by the administration of dox . This aberrant viraemic profile may be dependent upon intrinsic host factors , for example the availability of alternative secondary receptors , or mutational events in the vaccine virus . Notably , mutations in rtTA which do not affect dox-dependence were detected only in the fully protected macaques and may have contributed to the fitness of SIVrtTA in vivo . Interestingly , a similar virological profile was seen also in one animal vaccinated with SIVmac239Δnef . Despite the replicative fitness cost introduced by the dox-dependent regulatory elements , the remaining animals vaccinated with SIVrtTA demonstrated significant protection from wild-type SIVmac239 challenge , as breakthrough of challenge virus was at lower levels than naive challenge controls with reduced lymphoid virus sequestration . These results support the observation that in the SIV/macaque model , and in common with other live attenuated vaccines , a defining feature of efficacy is related to the ability of the vaccine virus to replicate in the early phases of vaccination and in addition , suggest that limited acute phase replication may be compensated by subsequent persistence . SIVrtTA shows absolute dependency upon dox for its replication in vitro [24 , 25] and as we have shown previously , dox status influences the TEM circulating pool [26] . Nevertheless , we have not formally directly demonstrated that dox status completely controls replication in vivo in all anatomical compartments . Whilst we consider that loss of dox-dependency is unlikely , given the lack of mutations in the known critical sites , future experiments could include challenge of naïve macaques in the absence of administration of doxycycline . Application of discriminatory PCR assays able to unravel the relative contributions of each virus to detectable PCR signals was a critical component of this study . These assays unequivocally established that plasma RNA following challenge of Group C animals , and of fully protected SIVrtTA vaccinates E70 and E65 was vaccine-virus specific . Moreover , this was corroborated by analysis of CA-RNA in lymphoid tissues at the termination of the study . A hallmark of complete vaccine protection appeared to be the persistent replication of vaccine virus in lymphoid tissue . Surprisingly however , only a very low level of vaccine CA-RNA was detected in a single tissue of macaque E75 suggesting there may have been persistence elsewhere such as the gut and/or vaccine generated immunity had cleared infection to limits below detection at least in the tissues examined . The mechanism for the persistent low-level replication of SIVrtTA in the absence of dox in macaque E65 is unknown . As we have reported previously , low levels of vRNA have been detected by in situ hybridisation in small intestine from SIVrtTA-infected rhesus macaques following dox withdrawal [26] . Therefore , we are unable to formally exclude the possibility that dox-dependency in vivo is conditional . It was notable that where breakthrough virus was detected in lymphoid tissues of Group B animals , maintained on dox throughout the experiment , there was no evidence of residual vaccine virus . We reported previously that proviral DNA was detected pre-challenge in the spleen , PLN and MLN of animals maintained on dox although concentrations were lower than in macaques vaccinated with SIVmac239Δnef [26] . Presumably , given the fitness disadvantage , any extant replicating SIVrtTA was displaced by the challenge wild-type virus . Although in this study we were unable to definitively address whether persisting vaccine virus replication in lymphoid tissue is an absolute requirement for complete protection because of the reduced replication of SIVrtTA , the opportunity was available nonetheless to compare T memory cell frequencies and cellular immune responses in partially and completely protected groups . The T memory cell results showed a strong association with protection status , which in most analyses reached statistical significance . The complete loss of these associations when analysis was done 130 days after challenge is striking , particularly in ( 1 ) the polarisation of CD4+ memory T cells toward the TCM phenotype regardless of protection status and ( 2 ) the changes in proportions of CD8+ memory T cells in completely protected animals . Although not reaching statistical significance due to outliers there was a clear trend for reduction in the number of CD8+ TEM with a concomitant increase in TCM . This latter effect probably reflects a reduction in on-going antigen re-stimulation in vivo at this time and/or a redistribution of TEM to tissue compartments . We did attempt analysis in gut tissues taken post mortem; however , cell recovery was poor making interpretation of flow cytometric data unreliable . Despite the reported lack of association between responses detected in the blood and subsequent protection [10 , 12 , 6] , we identified a statistically significant association between high frequencies of global TEM in peripheral blood at the time of challenge and outcome . Moreover , total frequencies of SIV-specific polyfunctional CD8+ T cells were significantly higher in macaques exhibiting complete protection , compared with partially protected macaques , on the day of challenge . Interestingly however , macaque E65 , which demonstrated continuous very low-level replication of SIVrtTA in the absence of dox , failed to show this association , perhaps suggesting that other factors may be associated with complete protection in this animal . Further analysis of cytokine combinations revealed that CD8+ memory T cells with quadruple cytokine staining and cells staining for IFN-γ and TNF-α were present at higher frequency in complete protection compared with the frequencies in partially protected animals . Interestingly , the one macaque that did not have detectable dual-stained CD8+ T cells , E70 , had an exceptionally high frequency of quadruple staining cells . Clearly , this analysis represents only a fraction of the total picture , since proteome-wide expansion of T cells was not performed and only 4 cytokines were analysed . ICS staining for IL-17 was included in the present study since perturbations in CD4+ and CD8+ IL-17-staining cells in both the periphery and mucosal compartments reportedly reflect SIV-induced changes in disease status [34–36] and therefore could be a useful marker particularly in animals that may become dually-infected after challenge with virulent virus ( i . e . may indicate sparing from disease progression ) . Several animals displayed unexpectedly high IL-17 positivity either before or following superinfection challenge . The reasons for this are not known; however , it is worth pointing out that these results were obtained in the context of infection with a novel SIV construct and it is possible that in certain genetic backgrounds this virus stimulates a strongly regulatory T cell phenotype . Analysis of SIV-specific CD8+ T cell frequencies in mesenteric lymph nodes did not reveal a difference between completely and partially protected animals; however , it did reveal evidence of a challenge virus footprint . The Nef-specific T cell responses seen could only be stimulated by wild-type virus challenge . As Nef is not a structural component of the virus , this would require de novo synthesis of Nef in infected cells . The absence of Nef-staining in the spleen of SIVmac239Δnef vaccinated animals is consistent with the notion that the mechanism of complete protection from wild-type virus challenge operates through early clearance of challenge virus; whereas in partially-protected animals T cells may suppress wild-type virus replication rates relative to those in vaccine naïve animals . It was however surprising that a low level of Nef staining was detected in the apparently completely-protected SIVrtTA vaccinated animals . Thus , although by the criteria of RNA detection and Gag-specific antibody responses post-challenge these animals appeared to be completely protected , they should perhaps be considered falling into an intermediate category between completely and partially protected . Clearly , however , these macaques were protected from overt , productive superinfection . In this regard the timing between exposure to wild-type virus and recovery of tissues at autopsy for analysis may be critical . In this study a relatively long period ( 20 weeks ) was allowed to elapse from time of wild-type challenge to autopsy , which is likely to have allowed sufficient time for a response to wild-type virus to be generated but where the virus was no longer detected at termination . In such a scenario , the challenge virus is likely to have been present at some level but which had been subsequently cleared by host T cell responses to wild-type virus infection reflecting previous reports in the literature where much earlier sampling for virus post-challenge ( eg 14 days after wt challenge ) resulted in detection of virus in tissues at necropsy but the overall virological phenotype was that of protection [37] . The likely role of T cells in this protection has been further demonstrated by CD8 T cell depletion experiments where control of the replication of both the challenge and vaccine viruses have been linked to a CD8 T cell response [38 , 39] . Recently reported detailed analysis of immune responses and deep sequence characterisation of SIVmac239Δnef post-vaccination indicated that there is a shift following early , rapid virus escape due to immune pressure to variable regions targeted during the acute phase to a re-focussed immunological response to more conserved epitopes [40] . However , the level of sub-clinical antigenic drive required to deliver such an anentropic state requires clarification , perhaps also in the face of host responses to the vaccine virus , since it was also noted that macaques with undetectable plasma viraemia experienced ongoing sequence evolution of the vaccine virus . It is perhaps noteworthy that in our study we observed distinct sequence changes in the rtTA gene rescued from viral RNA in plasma , taken as a measure of recently replicating virus , in the two SIVrtTA protected macaques ( E65 , E70 ) during the early , post-acute phase of virus replication which further marked these macaques out as being virologically distinct from the other SIVrtTA vaccinates . Hence , viral evolution as a driver for improved virological fitness in vivo during the post-acute phase appears to have had a marked effect in terms of the overall protection status conferred on these two macaques . SIVrtTA replication in macaques will also probably induce immune responses not only against viral proteins but also against rtTA itself [41] . Therefore , it is plausible that the observed amino acid changes mediate a mechanism of immune-escape of the rtTA protein , which would likely improve persistent virus replication , but this was not formally investigated . Hence , SIVrtTA vaccination of Indian rhesus macaques appears on the cusp of delivering potent vaccine protection . If SIVmac239Δnef-induced protection correlates with an expanded T cell anentropy to highly conserved epitopes with an associated increased depth of response generated over time , this likely explains the relatively poor ability of a ‘one-hit’ vaccine response , such as single cycle SIV to ensure long-lived vaccine protection . Compensations in vaccine replication appear important in conferring protection mediated by SIVrtTA , although whether these are sufficient to explain features of early protection from heterologous challenge , for example , remains unclear . Highly attenuated viral vaccines such as SIVmacΔ4 [11] which have a reduced replication potential in vivo , but which fail to persist , exhibit an intermediate protection profile . Hence the ability of SIVrtTA to exhibit low , continuous replication provides a clear advantage compared to these approaches . Lack of an increased magnitude of SIV-specific CD8 T cell responses in lymph nodes correlating with proposed mechanisms of protection for cellular responses at key sites of virus replication in the body [40] , suggest that the role of CD8 T cells in this mode of vaccine protection is far from resolved , whereby a higher viral replication in turn leads to higher CD8 T cells responses in lymphatic tissue [10] . On the face of it our data appears to strongly support the view that CD8+ polyfunctional TEM are critical in protective immunity induced by live attenuated SIV as suggested by Fukazawa et al [10] for lymph node responses . However , technical limitations precluded the ability to assign ICS responsiveness specifically to memory phenotype in our study , and as in other studies , our observations remain correlative . Indeed , antibody responses to Gag p27 before and after vaccine challenge are also predictive of outcome but are unlikely of mechanistic significance . If the current paradigm of live attenuated vaccine protection is correct , it must also explain why superior responses in the host that prevent viral infection are established in the same host where host control of the vaccine is poorest . This counterintuitive observation requires a cogent answer irrespective of localisation of the vaccine virus e . g . in T-follicular helper cells which may be subject to immune privilege , or magnitude and breadth of measurable immune responses such as CD8 T cell responses which are potentially capable of targeting and controlling both vaccine and challenge viruses , yet the vaccine virus is able to persist at these key sites . Taken together , our data provide further insight into the highly dynamic process of live attenuated SIV vaccine outcomes where the replicative properties and persisting nature of the vaccine virus appear crucial to vaccine efficacy . SIVrtTA provides a novel tool in our armoury to understand more fully processes of occult and patent virus replication at niche anatomical sites where issues of viral latency and persistence are crucial in understanding retrovirus and immune interactions .
Non-human primates were used in strict accordance with UK Home Office guidelines , under a licence granted by the Secretary of State for the Home Office which approved the work described . Animal work at NIBSC is governed by the Animals ( Scientific Procedures ) Act 1986 that complies with the EC Directive 86/609 and performed under licence ( PPL 80/1952 ) granted only after review of all procedures in the licence by the NIBSC local Animal Welfare and Ethical Review Body . All study macaques were purpose bred and group-housed for the entire duration of the study , with daily feeding and access to water ad libitum . Given the limited availability of suitable macaques , age , sex and weight matching was not possible , nor central to the study outcome . Regular modifications to the housing area were made by husbandry staff including introduction of novel structures ( eg swings and perching stations ) and foodstuffs in novel manners to encourage foraging for food , to further enrich the study environment . The environmental temperature ( 15–24°C ) , was appropriate for macaques and rooms were subject to a 12 hour day/night cycle of lighting . Animals were acclimatised to their environment and deemed to be healthy by the named veterinary surgeon prior to inclusion on the study . All animals were sedated with ketamine prior to bleeding or virus inoculation by venepuncture . Frequent checks were made by staff and any unexpected change in behaviour by individuals on study followed up , including seeking of veterinary advice where necessary . Regular blood evidence of incipient disease and veterinary advice were sought when persisting abnormalities detected . The study was terminated and all animals killed humanely by administering an overdose of ketamine anaesthetic prior to development of overt symptomatic disease . All efforts were made to minimise animal suffering , including provision of a high standard of housing quarters and monitoring of animal health and well-being and the absence of procedures not essential to the study . 16 UK purpose-bred Indian rhesus macaques ( Macaca mulatta ) were used in the study , in accordance with UK Home Office guidelines ( Code of Practice 1988 ) and local ethical approval . The basic construction and mode of action of the SIV-rtTAΔnef ( SIVrtTA ) vaccine , based on a SIVmac239 genetic backbone , is depicted in Fig 1A . In a challenge study experiment , eight macaques were inoculated intravenously with 5 x 103 TCID50 SIVrtTA vaccine receiving 100mg daily oral dosing with dox . In four macaques ( Group A ) , dox was removed eight weeks prior to SIVmac239 wild-type challenge . In the remaining four SIVrtTA vaccinates ( Group B ) dox dosing was maintained at 100 mg daily oral dosing . Group C comprised four macaques inoculated with 104 TCID50 SIVmac239Δnef . All vaccinates were challenged with wild-type SIVmac239 in addition to four additional macaques which served as naïve challenge controls ( Group D ) . The study outline is summarised in Fig 1B . Veterinary procedures deployed the use of ketamine hydrochloride prior to sedate macaques . Plasma concentrations of dox were monitored ex vivo using a previously described assay [41] . Macaques were genetically characterised for host MHC profiles , by Dr David Watkins ( Univ . Wisconsin , S2 Table ) . Distribution of TRIM5α and TRIMcyp alleles was determined as previously described . Mamu7 represents macaques harbouring the TRIMcyp allele [42] . Initial quantitative measures were made in peripheral blood using quantitative gag-based real-time PCR assays as previously described [6] . Plasma vRNA levels were determined for EDTA-treated plasma samples with a limit of detection of 50 SIV RNA copies/ml and SIV DNA levels on PBMCs with limit of detection one SIV DNA copy/100 , 000 cell equivalents . SIVrtTA-specific levels were determined using primers designed to amplify a region of the rtTA gene using PCR conditions comparable to those described for the total gag estimations against an rtTA plasmid containing unique sequences to the rtTA gene . SIVrtTA-specific amplification sequences were CGCCGTGGGCCACTT ( forward ) , and CTTTCCTCTTTTGCTACTTGATGCT ( reverse ) ; internal rtTA probe sequence was FAM-CACTGGGCTGCGTATTGGAGGAACAG-BHQ1; primers and probes were used at 100nM concentrations . Wild-type SIVmac239-specific amplifications were made with CTCAGGACCAGGAATTAGATACC ( forward ) , AAGGGTCATCCCACTGGGAAGT ( reverse ) and internal probe sequence FAM-TCCCTGTAAATGTATCAGATGAGGCACAGGAGG-BHQ1 targeting the nef gene . Primers were used at 100nM and probe at 120nM concentrations . Detection limits of virus-specific amplification in plasma were determined to be 100 RNA copies/ml with an amplification efficiency of >98% . Cell-associated RNA determinations were made for SIVrtTA , SIVmac239Δnef and wild-type SIVmac239 respectively by adapting a previously reported method [7] . Total RNA was isolated from spleen , mesenteric and peripheral lymph nodes using an RNeasy kit ( Qiagen ) , subjected to on-column DNAase treatment in accordance with the manufacturers’ protocol . Virus-specific targets were amplified by one-step RT-PCR using 50ng total RNA input , adapting the SIVrtTA and SIVmac239 wild-type specific primers described above and employing those described previously in [43] for SIVmac239Δnef-specific amplification as follows: cttaggagaggtggaagatggatactc ( forward ) , CTTTTCTTTTATAAAGTGAGACCTGTTCC ( reverse ) and internal probe sequence FAM- CAATCCCCAGGAGGATTAGACAAGGGCTTG -BHQ1 . Primers were used at 300nM and probe at 75nM . All CA-RNA determinations were made using normalised values of GAPDH , in co-amplification reactions as described in [7] . All amplifications were performed with Invitrogen Ultrasense kits with a thermoprofile of RT step 52°C for 30 mins; 10 mins at 95°C then 40 cycles of 95°C for 30 seconds and 60°C for 60 seconds . Limits of detection for SIVrtTA , SIVmac239Δnef , wild-type SIVmac239 CA-RNA assays were determined as 50 , 34 and 80 SIV RNA copies per 50ng total RNA . All CA-RNA quantitative PCR assays had an efficiency of >95% , typically 98–99% efficiency of amplification . The SIVrtTA assays were validated using a plasmid construct denoted rtTAV16 diluted to an extinction end-point in quantitative assays . Plasma anti-SIV p27 IgG responses were quantified by ELISA . Briefly , medium binding 96-well plates ( Greiner , UK ) were coated with 1μg/ml recombinant SIV p27 ( CFAR , UK , Cat no: EVA664 ) . Test plasma and standard positive and negative control samples were added to washed plates and bound IgG detected with goat anti monkey IgG-HRP ( Serotec ) followed by addition of substrate to induce a colour reaction in reactive samples . Memory phenotype and intracellular cytokine staining were performed separately in each sample per animal due to limitations of the flow cytometry capability available . Peripheral blood lymphocytes ( PBL ) were isolated using Percoll gradient centrifugation and mesenteric lymph node mononuclear cells ( MNC ) were isolated by mechanical disaggregation of tissue . To delineate memory T cell subsets , PBL were simultaneously surfaced stained with anti-CD3-V500 ( clone SP32 , BD Horizon ) , anti-CD4-V450 ( clone L200 , BD Horizon ) , anti-CD8-APCCy7 ( clone SK1 , BD Biosciences ) , anti-CD95-PECy7 ( DX2 , BioLegend ) , anti-CD28-PerCP-Cy5 . 5 ( eBiosciences ) , and anti-CCR7-FITC ( R&D systems ) . Gates on lymphocyte subpopulation were defined as central memory CD8+C95+CD28+CCR7- and CD8+CD95+CD28- CCR7- as effector memory . SIV-specific T cell responses were determined by cytokine production after incubation with 5 μg/ml of either SIV Gag , Tat , Rev or ( for MLN MNC additionally Nef peptides from the nef -unique coding region ) ( 15mers overlapping 11 residues , CFAR/NIBSC , Potters Bar , UK ) plus 10 μg/ml CD49d , 50μg/ml anti-CD28 , Golgi Stop ( 10ng/ml , BD ) , and incubated at 37°C in a 5% CO2 environment with RPMI 1640/10% FCS for 14h . Stimulated cells were surfaced stained for CD3 , CD4 and CD8 , permeabilised ( Fix and Perm kit , Caltag ) , and then stained for intracellular cytokine detection with anti-IFNγ-PErCPCy5 . 5 ( clone B27 ) , anti-IL-2-PE ( MQ1-17H21 , eBiosciences ) , anti-TNF-α-APC ( MAB11 , eBiosciences ) and anti-IL-17-Pacific Blue ( BioLegend ) . Polyfunctional T cells were determined by a gating strategy as shown in the representative plots ( S9 Fig ) . In detail , within CD4 and CD8 subsets , distribution of TNF-α and/or IL-2 producing cells were specified using contour FACs profile quadrants . Each quadrant within these cell populations were sequentially analysed for IFN-γ and/or IL-17 production in combinatory plots . For group comparisons ( partial versus complete ) , total frequencies of ICS-stained cells were derived by adding mono , bi , tri and quadruple functional frequencies for each animal . The relative distribution of the cytokine producing cells in each animal was summarised in pie charts using SPICE software . All peripheral and tissue derived mononuclear cells were acquired and analysed using a BD Canto II flow cytometer ( BD Immunocytometry ) with FACS DIVA software as described previously [26] . Graphing and associated statistical analyses , as specified , were performed using Sigma Plot 11 ( Systat Software , Inc . ) . Kruskal-Wallis analyses of variance with Dunn’s post-hoc test were determined using the Minitab version 17 software . In addition , analysis and graphical representation of cytokine production were conducted using the data analysis programme Simplified Presentation of Incredibly Complex Evaluations ( SPICE , version 5 . 3 ) provided by M . Roederer , National Institutes of Health , Bethesda , MD . Immunochemical staining for Nef was performed with the KK77 antibody ( CFAR; ARP3093 ) which is an IgG2a isotype raised to recombinant SIVmac251 Nef and which detects wild-type Nef only , using protocols as previously described [7] . | Development of an HIV vaccine remains a global health priority . In non-human primates live-attenuated SIV induces a potent vaccine effect . Following disappearance of vaccine virus from the peripheral circulation replication persists in lymphoid tissue . To address whether this occult replication is critical to the generation of protective immunity we used a novel construct ( SIVrtTA ) based on the prototypic live attenuated SIVmac239Δnef but which requires the presence of the antibiotic doxycycline to replicate . Protection appeared independent of doxycycline status at the time of virulent virus challenge suggesting that occult replication may not be absolutely necessary for persistence of immunity; however , stronger protection was observed in monkeys vaccinated with SIVrtTA where vaccine replication persisted for longer after peak viraemia . Moreover , some evidence of very low level breakthrough of vaccine virus replication was seen and protection was weaker than that obtained with SIVmac239Δnef . Both vaccination and challenge perturbed circulating T cell populations , but only the frequency of SIV-specific CD8+ polyfunctional T cells measured on the day of challenge was associated with protection . Replication-conditional mutants such as SIVrtTA have great potential in unlocking the complex interactions between the vaccine virus and host responses in the generation of potent anti-viral protection in vivo . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"blood",
"cells",
"viral",
"vaccines",
"medicine",
"and",
"health",
"sciences",
"innate",
"immune",
"system",
"immune",
"cells",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"cytokines",
"pathogens",
"immunology",
"microbiology",
"vertebrates",
... | 2016 | Role of Occult and Post-acute Phase Replication in Protective Immunity Induced with a Novel Live Attenuated SIV Vaccine |
Histone modifications are known to play an important role in the regulation of transcription . While individual modifications have received much attention in genome-wide analyses , little is known about their relationships . Some authors have built Bayesian networks of modifications , however most often they have used discretized data , and relied on unrealistic assumptions such as the absence of feedback mechanisms or hidden confounding factors . Here , we propose to infer undirected networks based on partial correlations between histone modifications . Within the partial correlation framework , correlations among two variables are controlled for associations induced by the other variables . Partial correlation networks thus focus on direct associations of histone modifications . We apply this methodology to data in CD4+ cells . The resulting network is well supported by common knowledge . When pairs of modifications show a large difference between their correlation and their partial correlation , a potential confounding factor is identified and provided as explanation . Data from different cell types ( IMR90 , H1 ) is also exploited in the analysis to assess the stability of the networks . The results are remarkably similar across cell types . Based on this observation , the networks from the three cell types are integrated into a consensus network to increase robustness . The data and the results discussed in the manuscript can be found , together with code , on http://spcn . molgen . mpg . de/index . html .
The study of gene regulation is traditionally based on DNA sequence analysis , gene interactions and transcription factor binding events . It has however over the past decade been revolutionized by genome-wide maps of epigenetic marks , specifically DNA methylation and histone modifications . Histone modifications are post-translational modifications of the histone proteins which form nucleosomes by wrapping about 147 base pairs of DNA . These modifications can have effects on biological processes including transcription , DNA repair , splicing , dosage compensation and more [1] , [2] , either by altering the chromatin structure or by recruiting key proteins [1] . The observation of different histone modifications co-occurring in different contexts has raised the possibility of combinatorial effects and has led to the histone code hypothesis [3] , whereby combinations of histone modifications have a biological meaning and lead to distinct downstream effects . In particular , there has been much evidence for a strong role of histone modifications in the regulation of gene expression [4] , [5] , not only at promoters and enhancers , but also along the gene body . Many authors have contributed genome-wide pattern analyses of modifications around regulatory regions [6]–[10] . For example , it has been found that acetylation marks generally co-occur with active genes , whereas methylation marks can be associated with active genes or repressed genes , depending on the modified residue . Histone modifications can be clustered according to their average level around promoters into two groups , one group containing active marks and the other repressive marks [7] . Ernst et al . [9] used hidden Markov models to extract genome-wide epigenetic states , many of which can be thought of as characterizing the transcriptional process at various positions along the gene body , or different kinds of enhancers , or splicing or heterochromatin , etc . Although it is still unclear whether they are causes or effects of transcription , these observations clearly demonstrate a connection between different combinations of histone marks and different transcription states . For instance , it is well established that promoters carry H3K4me3 and/or H3K27me3 and that actively transcribed genes carry H3K36me3 [11] , whereas enhancers are marked by H3K4me1 and H3K27ac [11] , [12] . Histone modifications have even been successfully used to determine the presence of regulatory elements such as promoters or enhancers [11] , [13]–[17] . Beyond these qualitative findings , a remarkable quantitative relationship with mRNA expression levels has been demonstrated in [18] . However , so far all of these studies deal with co-occurrence but do not provide insights about associations between histone modifications . In this article , we are interested in building networks of histone modifications . This is a problem that benefits from relatively few variables ( histone modifications ) and many samples ( genomic regions of interest ) , allowing the use of rigorous statistical methods . In such networks , nodes represent histone modifications , and edges connections between them . The nature of these connections depends on the construction method used to obtain the network . Other authors , again particularly in the context of promoters , could capture associations using Bayesian networks ( BNs ) of histone modifications [19]–[23] . They aimed at establishing causal links: which modifications are required for the presence of another one . However claims about causality in BNs are controversial [24]–[27] , especially in the presence of hidden confounding factors , which occur quite frequently in biological systems . Additionally , BNs do not allow cycles or feedback mechanisms , which seems unrealistic in biological systems . The ChIP-seq data currently available represents a summary of the epigenome , averaged over many cells . For each histone modification , the read counts represent the average frequency at which it is found in the population of cells . This has three main implications for the interpretation of the edges . Firstly , it is very hard to make any claims about causality , as temporal information is missing . Secondly , discretization of the read counts is less plausible . Even if a histone modification is either present or absent at a specific region in a specific cell , the read counts represent the average over many cells , and discretizing these averages is no longer meaningful . Thirdly , given that only an average picture is available , it can safely be assumed that various states will be represented in the data and will appear in the network . Being in one particular state will mean highlighting relevant associations and downplaying others , but all associations will be present in the same network . In a way , we expect to infer the wiring of the circuit as opposed to the flow in the circuit , i . e . the statics as opposed to the dynamics . Edges can reflect co-occurence , mutual exclusivity , or they can mean that two modifications occur sequentially as part of the same pathway . We cannot distinguish between these scenarios with the data at hand . An observed correlation between two variables may either reflect a direct association or an induced association that may be due to a mutual association with a third variable . For example , if the lack of sports generates both a drop in fitness and a bad mood , a correlation between the variables and will be observed when , actually , they are only connected through the variable and do not interact otherwise . The third variable ( here ) is often referred to as confounding factor . Confounding factors , which can be accountable for part of the associations between other variables , are often presented as a nuisance - experimental techniques for instance may lead to biases that are undesirable confounding factors - however they need not be . For example , expression level is a confounding factor of great interest . In any case , looking at how apparent associations may be explained away can be very insightful . Let us suppose we have two variables of interest and . The correlation coefficient is a powerful tool but it cannot distinguish direct associations from those due to confounding factors . The partial correlation coefficient was designed to remedy that very problem [28] . The idea is to subtract from and the information contained in a control group of variables by linearly regressing ( resp . ) against , and to keep the residuals ( resp . ) . We then compute the correlation between and . This correlation is called a partial correlation , written and is a measure of the correlation between and that remains after the explanatory power of is taken out . Let us assume we have a set of variables , and we compute the correlation matrix such that . Let denote the partial correlation matrix ( PCM ) that contains the pairwise partial correlations , each using as control the remaining variables , i . e . the matrix such that . Note that , in this framework , each variable in turn is treated as a confounding factor , regardless of its expected biological relevance . A property of partial correlations is that may be obtained by simply inverting , normalizing and negating the correlation matrix [29]–[32] . This procedure , that we will use throughout the study , is a very fast alternative to the linear regressions . It also shows the involvement of all variables in the computation of through the inversion step , as opposed to that is only computed on and . It is common practice to recover the undirected network connecting these variables by simply building a fully connected network and by removing all edges for which [29]–[32] . This rests on the theoretical grounds that the variables are normally distributed and are linearly related , therefore having is equivalent to having independence between and conditioned on the other variables [29]–[32] , which is exactly the requirement for the absence of edge in an undirected network . Such networks are therefore referred to as graphical Gaussian models ( GGMs ) [29]–[32] . In case the true network is Bayesian ( i . e . directed and acyclic ) then the GGM will contain the original edges and will connect the parents of a same child . GGMs provide a simple and efficient method , whereby networks can be built in just a few seconds . They have been successfully applied to infer gene regulatory networks , even in the presence of small sample size , and a short review of these applications can be found in [33] . In this study , we propose to focus on edges that represent direct dependencies . We want to draw edges between histone modifications that are directly linked in a pathway or that act together , i . e . whose association cannot solely be explained by confounding factors . We build on GGMs , and put forward a robust method to compute sparse partial correlation networks ( SPCNs ) . To the best of our knowledge , PCNs have not yet been applied to histone modifications . In contrast to gene regulatory networks , here the sample size is very large and the variables are few . Formally , partial correlations require normal distributions . In our work this need is overcome and outliers accounted for by rank-transforming the input data . Sparseness is achieved via a cross-validation scheme . Our SPCNs reveal edges that are symptomatic of direct associations , mutual exclusivities , direct edges in a pathway , indirect edges where the intermediate variable ( s ) are not available , or collaborative work to produce a third variable . Zhao's group was one of the first to produce genome-wide profiles for a large number of histone modifications , they did so in CD4+ cells [6] , [7] . In the meantime , several other groups have contributed to the Roadmap Epigenomics project [34] , a database that now contains data for varying numbers of histone modifications in different cell types . Based on this data , the cell types with the largest number of histone modifications were chosen: CD4+ , IMR90 and H1 . CD4+ cells are lymphocytes ( white blood cells ) , they are part of our immune system . IMR90 cells are fibroblasts ( cells involved in the synthesis of tissues' external structure ) in the lung , and H1 cells are embryonic stem cells . 21 histone modifications are available for all three cell types , we keep only those . Histone modification data is obtained via ChIP-seq experiments , so openness of the chromatin is a potential confounding factor to include in the analysis via DNaseIHS , which marks the hypersensitivity of the DNA to the enzyme DNaseI . The relationship of histone modifications to mRNA levels is of particular interest because of the role of histone modifications in transcription , so mRNA data is included . We look at the amounts of ChIP-seq reads for these 23 variables in the [−2000 , +2000] around the transcription start sites ( TSSs ) of known genes , and at the amounts of RNA-seq reads in the exons of those genes . Antibodies can also play a role as confounding factors ( because of their cross-reactivity ) , and may also vary from experiment to experiment . Antibodies are an interesting case because , although they are not semantically “hidden” ( we know which ones are used and we know they can cross-react and act as confounding factors ) , they are technically hidden since we do not know how they cross-react as no data is available . However , we can build a table of cross-reactions and look it up as a possible source of explanation for links between histone modifications . Details about data collection and antibody can be found in Materials and Methods .
We modify GGMs in two respects: first by rank-transforming the input data , and second by enforcing sparseness via a cross-validation scheme . A global view of the algorithm is shown in Figure 1 . Precision is favored over completeness: an edge is only found in a network if it is strongly supported by the data . Therefore interpreting edges is favored over interpreting the lack thereof . Details about the computation of the PCMs , the p-values and the q-values can be found in Materials and Methods . “Explaining away” in machine learning is “a common pattern of reasoning in which the confirmation of one cause of an observed event reduces the need to invoke alternative causes” [37] . We take over this concept and translate it into our own context . A connection between and is explained away by when is negligible compared with , because we assume that was the main cause of the apparent connection between and and that therefore the need to find further causes is alleviated . When controlling for confounding factors , the partial correlation coefficient is substituted to the correlation coefficients and the difference can be very large . is generally smaller ( in terms of absolute value ) as it is explained away by the control variables , but it can also be greater as control variables tie and together . For example , if and are independent co-parents of such that , they become dependent upon conditioning on , such that may be different from 0 . We would like to know which variables are responsible for most of the change from to . Running an exhaustive search on combinations of about 20 variables is neither possible nor desirable . Instead we condition on a single variable . We repeat the operation for every possible in the dataset and identify the that leads to the biggest discrepancy between and , i . e . the control variable that has the highest impact on the correlation . The impact of all variables is shown for some pairs in It needs to be established that networks remain stable upon using input data from different experiments or from different cell types . To this end , we define an index of overlap between PCMs , based on the ranking of the entries which represent the associations between pairs of variables . For each PCM ( ) , the pairs of variables are ranked by increasing q-values and the first pairs ( ) are stored in a list . The number of pairs that occur in all lists divided by is a measure of the similarity between all the when pairs are considered . Results are presented in plots where varies from 1 to . The overlap expected at random depends on the number of matrices being compared and on the number of pairs being examined . It is easily computed , as seen in Materials and Methods . For , it follows a hypergeometric distribution , and therefore p-values are directly available . We now turn to a detailed analysis of the CD4+ network . Note that , the data containing 23 variables , the SPCN has edges maximum . The resulting network is shown in Figure 2a , all the partial correlation coefficients , their q-values and the mask are given in Text S1 Section 7 . Looking at edges around mRNA , we find it is negatively connected to H3K27me3 ( a mark of repression ) and positively to H3K27ac ( a mark of activation ) , H3K79me2 and H4K20me1 ( marks of elongation ) , which have been , with the exception of H3K27me3 , found to be important in predicting expression in CD4+ cells [18] . Interestingly , H3K36me3 has no link to mRNA , in line with [18] . The scatter plots in Text S1 Section 9 . 1 confirm the lack of relationship . Note that there is no standard correlation either . The data for H3K36me3 is not abundant , very few reads map to the regions of interest . This could come from H3K36me3's preference for exons [39] . Indeed exons are only a small part of the studied region , as shown in Text S1 Section 3 , so the lack of connection to expression could be due to poor data , it is hard to tell . Expected connections are numerous , such as the negative link between H3K27ac and H3K27me3 . These two histone modifications are by nature mutually exclusive , and therefore need not be explained by any other histone modification . The strong connections between the various methylation states of H3K4 , with H3K4me2 in between , are explained by the fact that these different methylation states are coupled by bidirectional links from H3K4me1 to H3K4me2 and to H3K4me3 . Alternatively , it can be explained by antibody cross-reactivity , but it may not be explained by any other histone modification . Connections between DNaseIHS and H3K4me3 and H4K20me1 reflect the need for open chromatin to have transcription . Finding expected associations is a requirement , however it is more interesting to find unexpected connections . H3K27me3 and H3K9me3 are positively associated ( see scatter plots in Text S1 Section 9 . 2 ) . They have been thought to be mutually exclusive , H3K9me3 encoding constitutive heterochromatin , H3K27me3 facultative heterochromatin . Both would act as repressors but as part of two different processes ( involving the PRC1/2 complex for H3K27me3 and the HP1 proteins for H3K9me3 ) , that have been assumed mutually exclusive [40] . Clearly it is not the case here . It has been found that SUZ12 , which is part of PRC2 and involved in setting H3K27me3 , promotes H3K9 methylation [41] , giving a straightforward explanation for our finding . The negative edge between H3K79me2 and H3K4me1 is puzzling given that they are two marks associated with transcription , and that the trend is mostly tue in active genes ( see scatter plots in Text S1 Section 9 . 3 ) . However a possible explanation is that H2BK120ub1 , which is required both for the production of H3K4me2/3 and of H3K79me1/2 [42] , acts as hidden confounding factor . Some expected edges exist albeit with an unexpected sign . In particular , H3K4me3 and H3K36me3 , associated with initiation and elongation , are positively linked to the repressive mark H3K27me3 ( see scatter plots in Text S1 Section 9 . 4 ) . In fact , for high levels of H3K27me3 , this trend already exists in the raw data . This may indicate that some promoters cycle between the repressed H3K27me3 state and the active H3K4me3/H3K36me3 state . The cycling idea of epigenetic states is not without precedent . It has been shown that the estrogen receptor target TFF1 is cyclically methylated and demethylated [43] , [44] . In some cells promoters are active ( H3K4me3 ) , in some cells they are repressed ( H3K27me3 ) , and in some cells they may be bivalent ( H3K4me3 AND H3K27me3 ) . All we measure is the population average . If these fluctuations are stochastic , we expect no correlation . However if promoters can move from being active ( H3K4me3 ) to being inactive ( H3K27me3 ) in a regulated manner , then we expect a positive correlation . This could be due to the cell cycle , e . g . promoters get active during S-phase and are rendered inactive thereafter [45] . When looking at the scatter plots in Text S1 Section 9 . 4 , the correlation seems to come from repressed genes , and a little bit from bivalent genes , supporting this hypothesis . Another example is the negative link between H4K20me1 and H4K5ac ( see scatter plots in Text S1 Section 9 . 5 ) , which seems at first glance counter-intuitive because H4K20me1 is positively linked to expression and acetylations are generally thought to be associated with transcription . This apparent paradox can be resolved by the following reasoning: H4K20me1 is mainly associated with transcription elongation , while acetylations are heavily enriched around the promoter . It has been shown in Drosophila that H4K20me1 recruits the factor RPD3/HDAC1 , leading to the deacetylation of H4K [46] . Thus it seems that H4K20me1 helps to prevent cryptic initiation in the transcribed gene body . Since mechanisms are to a large degree cell-type-independent , the precision and robustness of the results can be increased by integrating information from all available cell types . A SPCN is created for each cell type . Figure 2bc shows the consensus network which contains only those edges that are found in at least two cell-type-specific SPCNs . Light blue edges show negative associations that are found in two cell types , blue edges negative associations found in all three cell types . Pink edges show positive associations that are found in two cell types , red edges positive associations found in all three cell types . It looks very similar to the CD4+ SPCN in Figure 2a . Important associations such as mRNA-H3K27me3 , mRNA-H3K79me2 , DNaseIHS-H3K4me3 , DNaseIHS-H4K20me1 and H3K27ac-H3K27me3 are conserved across cell types . Surprising connection such as H3K27me3-H3K9me3 and H4K20me1-H4K5ac are also stable . The strong connection between H3K4me1 and H4K20me1 is only found in CD4+ . Some of the edges that are common to all networks ( marked in bright red and blue ) are of particular interest . The antibody table in Text S1 Section 2 ( see Materials and Methods ) shows that there is antibody cross-reactivity for H3K4's various methylations and for H3K79me1/2 . The edges may reflect biologically meaningful associations but may ( also or instead ) be due to cross-reactions . H3K23ac's antibody reacts with H3K14ac , H3K18ac's with H4K5ac , and H3K27ac's with H3K9ac , which explains partially these three connections . The group H2BK12/20/120ac remains unexplained , however it is plausible that it may be the result of unreported antibody cross-reactions . Other edges that may be explained by antibody cross-reactivity are H4K5ac-H3K27ac and H4K5ac-H3K18ac as well as H3K14ac-H3K18ac . The explaining away procedure was applied . Text S1 Section 10 shows some of the plots that are obtained for all the edges of interest . Figure 4 summarizes the critical information into one matrix . The colors give the magnitude of the differences between and . If zooming in is available , the numbers on the lower part of the diagonal give the actual difference , and the text on the upper part of the diagonal gives the histone modification that has the most incidence on . Partial correlations work in such a way that , in order to explain the correlation between and , it is sufficient that a control variable explain . The variable with the most impact then says something about regardless of . Symptomatic of this scenario , the first explanatory variable is then often the same along the column of the matrix corresponding to . For example , in the column associated with H3K27me3 , H3K27ac is very often the most influential variable . It can be assumed that H3K27ac explains H3K27me3 and therefore leads to the loss of correlation between H3K27me3 and other variables . H4K5ac seems to explain H3K14ac . This may be due to antibody cross-reactivity , as H4K5ac is often seen in H3K23's column , and H3K14ac's and H3K23ac's antibodies are known to cross-react . An interesting example that shows how well this procedure works is the pair H3K4me1 and H3K4me3 . After glancing at Text S1 Section 8 . 1 or after zooming into Figure 4 , it can be seen that the variable most responsible for the correlation is H3K4me2 . This makes a lot of sense biologically , as H3K4me2 is an intermediate state of methylation . Another example is the correlation between mRNA and H3K4me3 , which seems to be largely explained by H3K27ac . This maybe due to the fact that H3K4me3 recruits the SAGA complex required for acetylation [47] which puts H3K27ac , which in turn is predictive of mRNA levels , as was seen in [18] . The relationship between H3K4me3 and H4K20me1 is fully explained by DNaseIHS . One possible reason for this is that chromatin openness favors transcription , thereby explaining H3K4me3 . The role of H4K20me1 in HDAC recruitment has been demonstrated in the context of chromatin reassembly [46] . Thus it seems that transcription may lead to higher histone turnover , which results in higher levels of H4K20me1 . Similarly to the networks , a consensus effect matrix is shown in Text S1 Section 8 . 4 . It is surprising to see how well the effect of partial correlation and the explanatory variables are conserved across cell types . Indeed , out of 21 possible variables that are all correlated , in most cases the same one comes out in at least two cell types .
We put forward SPCNs , a fast and robust tool , to construct undirected networks of histone modifications . By definition SPCNs can handle continuous data . Moreover they contain all relevant links , and allow for cycles and symmetric relationships . Edges in a SPCN may be seen as controlled associations , where the link between two variables is only established after controlling for potential confounding factors ( the other variables at hand ) . We believe they are the perfect tool for our purposes . The algorithm is designed to maintain a high precision level in the reconstruction of the networks . To be present , an edge must appear in 7 out of 10 sub SPCMs , i . e . be highly supported by the data . Some edges may be missed , and the lack of edges must be carefully interpreted , however given that only 10% of the maximal drop in performance is allowed , we believe that most contributing edges are recovered , and that the lack of edges mainly corresponds to the lack of relevant associations . We used the availability of data from different experiments and different cell types to our advantage and quantified the variability that could be expected . Firstly , it is interesting to note that the variability across experiments , for the same cell type , is not low . This tends to show that biological data is difficult to reproduce , that results should be interpreted with care , and that evidence may not be overwhelming even though a phenomenon is true . Here , the cell type is the same so it is true that the mechanisms should be the same , yet the evidence is not as high as one might have expected . Secondly , the variability across cell types is marginally higher than the one across experiments , showing that the networks are stable across cell types , and that the variability is mostly due to experimental noise . This last observation is a significant result . Histone-modifications-related mechanisms are often assumed to be the same in all cell types , but it is not systematically checked . Our simulations show that meachanisms are strikingly similar across cell types , almost as similar as two different experiments in the same cell type . Gathering information on antibody cross-reactivity was difficult but it proved insightful as it revealed important biases in the data . In particular , different methylation states , such as H3K4me1/H3K4me2/H3K4me3 or H3K79me1/H3K79me2 , are difficult to distinguish . The edges between such histone modifications may be biologically relevant or/and due to antibodies' lack of specificity , probably both , it is impossible to tell with the data at hand . A similar phenomenon was observed for acetylations . This ought to be a warning for the community . Antibodies are too trusted in many ChIP-seq studies . Instead cross-reactivities should be documented and biases reported when appropriate . In fact , cross-reaction studies are missing for many antibodies , and biases may be more important than we think . The SPCN gives a global view of the associations between histone modifications , however this view assumes a closed environment containing only the variables in the network . This is an intrinsic limitation of the method . If the set of variables is increased , the new network will not necessarily contain the previous one , all edges might be affected . How much they might be affected depends on the relevance of the variables that are introduced , and on the number of these variables . This makes the network very hard to test experimentally , as the presence of other variables in the cell will make the network by definition obsolete . However such assumptions are not new in biology , where subsets of variables are often chosen , and consequently studied as if they were isolated from the rest of the world . The effect matrix on the other hand gives a detailed view of what partial correlation does . It shows the difference between the correlation and the partial correlation conditioned on all other variables . In particular , it allows to see which variable causes the highest difference between and . This is of high biological interest , not only because it identifies potential hidden interactions , but also because such effects can be in principle verified experimentally . Associations of histone modifications are interesting as a first step to understanding their relations . However their connections are not physical and therefore remain abstract . Edges in a SPCN are as direct as possible given the variables at hand , but they can most probably be explained away by enzymes or proteins that float around and provide a physical interface for histone modifications , in particular chromatin modifiers . The next step is therefore to include data for such proteins . Ram et al . have now produced data for chromatin regulators [48] . Including them in the network and particularly in the effect matrix would allow to gain much deeper insight into the physical mechanisms . Further steps should also include transcription factors , and various genomic regions , such as proximal promoters and enhancers .
With two lists of selected pairs from a pool of pairs , the number of common pairs follows a hypergeometric distribution with equal number of white balls and drawn balls ( ) and with a total number of balls of , and a hypergeometric test is appropriate to compute p-values . The probability for pairs to appear in the two lists is obtained through the hypergeometric distribution with successes ( white balls ) in draws from a finite population of size containing successes ( white balls ) , so . The expected number of same pairs in the two lists is therefore , so the expected proportion is , i . e . a straight line . The p-value is then given by the hypergeometric test: . The appropriate call in R is . With three lists , things are more complicated . The probability for a pair to appear in the three lists is obtained through a Binomial distribution with number of trials 3 and probability , so . The expected number of pairs common to the three lists is , the expected proportion is therefore , i . e . a quadratic curve . For an observation , the p-value is computed by simulating intersections between three lists containing pairs sampled randomly from with replacement , and by counting the proportion of times the length of these intersections was at least as high as . If the result is 0 , is reported as upper bound . | Nucleosomes are protein complexes around which the DNA is wrapped for compactness . They are made of histone proteins that can be post-translationally modified and these histone modifications can affect the expression of surrounding genes . In the past decade , scientists have developed a strong interest in the part of gene regulation provided by epigenetics , i . e . those heritable characteristics that are not based on the DNA sequence and that can therefore be cell-type-specific , such as histone modifications . Striking patterns about the co-occurrence of modifications have been discovered , leading to the hypothesis that different combinations of modifications lead to different effects . Different histone modifications could act jointly to recruit certain proteins , or be required sequentially , which is reflected in statistical dependencies in measured data . The focus of this article is on building a network that represents the global dependencies by extracting direct associations of histone modifications . We find that , although histone modifications patterns are cell-type specific ( modifications may not necessarily appear at the same loci ) , the dependencies are to a large degree cell-type independent , which is supported by a large overlap of the inferred associations in the networks built for different cell types . We are able to find meaningful associations , both known and novel . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Finding Associations among Histone Modifications Using Sparse Partial Correlation Networks |
Recent research has produced a number of methods for forecasting seasonal influenza outbreaks . However , differences among the predicted outcomes of competing forecast methods can limit their use in decision-making . Here , we present a method for reconciling these differences using Bayesian model averaging . We generated retrospective forecasts of peak timing , peak incidence , and total incidence for seasonal influenza outbreaks in 48 states and 95 cities using 21 distinct forecast methods , and combined these individual forecasts to create weighted-average superensemble forecasts . We compared the relative performance of these individual and superensemble forecast methods by geographic location , timing of forecast , and influenza season . We find that , overall , the superensemble forecasts are more accurate than any individual forecast method and less prone to producing a poor forecast . Furthermore , we find that these advantages increase when the superensemble weights are stratified according to the characteristics of the forecast or geographic location . These findings indicate that different competing influenza prediction systems can be combined into a single more accurate forecast product for operational delivery in real time .
In the United States , seasonal influenza outbreaks occur every winter; however , the timing and severity of these seasonal outbreaks vary considerably from year to year . Accurate forecast of key characteristics of influenza outbreaks would allow public health agencies to better prepare for and respond to epidemics and pandemics . To this end , there has been significant work in recent years to develop forecasts of seasonal influenza outbreaks[1–5] . Proposed forecast methods include a range of statistical and mechanistic models , and incorporate data sources including syndromic and viral surveillance data , meteorological information , and internet search queries[3] . A comparison of coordinated influenza forecasts for the 2013–2014 season showed substantial disagreement in predicted outbreak characteristics among forecast methods[1] . This disagreement presents a barrier to the use of forecasts in decision-making . While some forecast methods may be consistently superior to others , the relative performance of individual forecast methods can also vary according to the specifics of the population , location , or outbreak of interest . For example , a forecast method developed for a densely populated urban area may be less accurate in a sparse rural setting . Alternatively , some forecast methods may be better suited for prediction of outbreaks that are smaller or larger than typically observed . The optimal forecast method may also depend on when in the season the forecast is being generated . Previous comparisons of subsets of the individual forecast methods used in this study have indeed revealed differences in forecast performance . Yang et al . [5] compared the performance of six filtering methods coupled with a mathematic model of disease transmission in retrospective forecasting of influenza epidemics . The relative performance of the filter methods varied by the timing of the forecast , the location being forecast , and the number of observed peaks in the outbreak . In retrospective forecasts of dengue fever , the relative accuracy of a model-filter forecasting system and two statistical methods varied by forecast timing , forecast target of interest ( i . e . timing of outbreak peak , maximum incidence and total incidence ) , and the similarity between the outbreak being forecast and previously observed outbreaks[6] . In weather and climate forecasting , discordant forecasts from competing models are combined into superensemble forecasts in order to offset the biases of each individual model . The resulting superensemble forecasts are more accurate than forecasts from any one model form [7–9] . Recent work has shown this superensemble approach to be effective in improving the accuracy of forecasts of dengue outbreaks[6] and influenza [10 , 11] . Here , we compare the accuracy of a suite of 21 competing forecast methods , as well as a superensemble forecast , in retrospective forecasts of influenza epidemics ( See Methods and S1 Appendix ) . We group the 21 individual forecasts into three categories: ensemble filter systems; particle filter systems; and a statistical model . The ensemble-based filter systems include three types of filters: Ensemble Kalman Filter ( EKF ) , Ensemble adjustment Kalman Filter , ( EAKF ) and Rank Histogram Filter ( RHF ) . Two particle filters are used: a basic particle filter ( PF ) and a particle Markov Chain Monte Carlo method ( pMCMC ) . The five filter methods are each coupled with four standard compartmental models of disease transmission: SIR , SEIR , SIRS and SEIRS ( see Methods ) . For the final individual forecast method , we use a statistical model called Bayesian Weighted Outbreaks ( BWO ) . We find that the superensemble forecasts are more accurate than any individual forecast system , and that this advantage increases when superensemble weights are stratified according to the characteristics of the forecast , or by geographic location .
For the ensemble and particle filter systems , observations of influenza incidence from the beginning of the season to the week of forecast initiation are used to optimize each of the 4 mathematical models . The optimized model is then run to the end of the season to generate a weekly forecast of influenza incidence , as measured by ILI+ , an estimate of influenza positive patients per 100 patient visits to outpatient health care providers ( see Methods ) . For the BWO , only the most recent 8 weeks are used for training . A forecast is then generated as a weighted average of historical ILI+ epidemic trajectories ( see Methods ) . Thus , each individual forecast method produces a weekly estimate of ILI+ from the time of forecast through the end of the influenza season . These trajectories were used to calculate key characteristics of each outbreak: the peak incidence , peak timing , and total incidence . Forecast skill was assessed based on the mean absolute error ( MAE ) in predictions of each of these metrics . Distributions of seasonal influenza outbreak peak timing , peak ILI+ and total ILI+ across 48 states and 97 cities are shown in Fig 1 . The overall error of each of the 21 individual forecast methods ( see Methods for names and descriptions of each forecast system ) used to predict these observed outbreak characteristics is shown in Fig 2 , relative to the MAE of the Baseline superensemble as a reference value ( discussed below in Superensemble results ) . Forecast MAE ranged from 1 . 8 weeks to 3 . 5 weeks for peak timing . Individual forecast MAEs for peak incidence ranged from 1 . 2 to 3 . 5 ILI+ , and from 0 . 5 to 1 . 0 ILI+ for total incidence . Among individual forecast methods , the lowest MAEs for peak timing and total incidence were generated by the dynamical models that included an exposed compartment ( i . e . the SEIR and SEIRS structures ) coupled with ensemble filters . However , the advantage of the exposed compartment was less clear in forecasts of peak incidence . Rather , the EKF model-filter systems produced the most accurate forecasts of peak incidence . Ensemble filter methods and BWO consistently outperformed particle filter methods for peak week and peak incidence; for total incidence , several particle-filter forecast systems performed comparably to ensemble methods . The pMCMC-SEIR and pMCMC-SEIRS forecast systems had exceptionally large errors for peak and total incidence . Fig 3 shows the MAE of each individual forecast method by forecast lead time and influenza season . Lead time refers to timing of the forecast relative to the outbreak peak , in weeks , with negative values indicating forecasts made prior to the peak and positive values for forecasts made after the peak . There were clear differences in relative forecast accuracy when discriminating by these factors . For example , the ensemble filters and BWO generally outperformed the particle filters in forecasts of peak ILI+ and total ILI+ , with the exception of 2007–2008 and 2012–2013 , two years with relatively large outbreaks . The particle filters produced the most accurate forecasts of peak week 1 to 5 weeks before the peak , but were among the worst performers following the peak . Inaccurate predictions of outbreak peak are possible even after the true peak has passed , as the models may predict a continued increase in incidence resulting in a later peak . The pMCMC-SEIR and pMCMC-SEIRS systems were especially prone to this type of error . Unlike the other filter methods , the pMCMC requires the same set of parameters be used to fit the entire observed time series [5] . As a result , the filter is less able to adapt to shifts in outbreak dynamics during an influenza season . This can result in poor forecasts , particularly for multi-peak outbreaks . For the initial baseline set of superensemble forecasts , a single set of weights for each target metric was applied to the competing individual forecasts each season ( Fig 4 ) . Each season , a new set of superensemble weights were computed using training data from all previous years . As such , we expected increased stability in the weights as more years were used to generate the weights . For all target metrics , the BWO forecast was typically assigned large weight , ranging from . 10 to . 53 for peak week , . 10 to . 32 for peak ILI+ , and . 03 to . 33 for total incidence . However , note that the BWO approach is the most dissimilar among the forecast approaches . If a subgroup of forecasts predicts similar outcomes , the superensemble weights , which sum to one , are expected to be split among that subgroup . The contribution of the individual model-filter forecasts , which share filter methods and model structures , may thus be diluted among similar forecasts . This circumstance may explain the heavy weighting of the BWO forecast , despite its larger than average MAEs among individual forecasts . The twelve ensemble filter systems contributed a total weight of . 29 to . 68 for peak week , . 49 to . 68 for peak incidence , and . 36 to . 69 for total incidence . The eight particle filter systems contributed a total weight of . 14 to . 31 for peak week , . 14 to . 21 for peak incidence . , and . 26 to . 43 for total incidence . The MAEs of superensemble forecasts are shown in Table 1 . On average , the baseline superensemble forecasts were more accurate at predicting the timing of the outbreak peak than any of the individual forecasts , with the exception of the EKF-SEIRS ( Fig 2 ) . Baseline superensemble forecasts of peak and total incidence had smaller MAE than all individual forecasts . We then stratified the weights by geographical region , forecast week , lead time relative to predicted peak , lead time relative to observed outbreak peak , population size and population density . These variables were pre-specified on the basis of previous work indicating that they may influence the accuracy of individual forecast methods ( for example [5 , 6] ) . The MAEs of the stratified superensemble forecasts are shown in Table 1 . For both peak timing and peak incidence , stratifying by lead time relative to observed peak led to the greatest improvements in forecast performance , decreasing MAE by 0 . 4 weeks for peak timing and 0 . 06 ILI+ for peak incidence compared to the baseline superensemble forecast . On average , this forecast outperformed individual forecasts of peak timing and peak incidence at all times during the course of the outbreak ( Fig 5 ) . In contrast , several individual forecasts were more accurate in forecasting peak timing than the baseline superensemble early in the season , while others outperformed the baseline superensemble late in the season ( Fig 5 ) . Averaged within influenza seasons , the superensemble forecasts generally outperformed individual forecasts ( S1 Fig ) The most notable single seasonal exception was the set of forecasts of total incidence in 2012–2013 , the year with unusually high influenza incidence . Stratifying superensemble weights by forecast week or lead time relative to predicted peak , which can serve as real-time proxies for actual lead time , led to decreases in MAE for peak timing , but had only a small effect on forecasts of peak and total incidence . Meanwhile , stratifying by geographical region and population density led to small decreases in MAE for peak incidence , but degraded forecast accuracy for peak timing ( Table 1 , Fig 5 ) . Stratifying weights by geographical region led to the lowest MAE for total incidence , but the improvement over the baseline forecast was small ( 0 . 02 ILI+ ) . We further assessed the accuracy of superensemble forecast credible intervals by determining the fraction of observations falling within the credible intervals specified by each forecast . In a well-calibrated forecast , we would expect this fraction to correspond to the value of the credible intervals; for example , 95% of observed outcomes should fall within the 95% credible intervals of a forecast method . Overall , the forecasts were well-calibrated . The calibration varied between the three target metrics , as well as between choices of stratification variable for superensemble weighting ( S2 and S3 Figs ) Forecasts of peak week and peak ILI+ were well calibrated at 90% and 95% credible intervals but somewhat overdispersed at lower confidence intervals . Forecast coverage for total ILI+ was well calibrated for 50% , but underdispersed at 95% and 99% credible intervals . In addition to comparing forecast errors averaged over many forecasts , we also compiled a ranking of individual forecast outcomes . For each forecast , we ranked the 21 point-estimates from the individual forecasts and the resulting baseline superensemble forecast from 1 ( lowest absolute error ) through 22 ( highest absolute error ) and summed the frequency of each ranking ( Fig 6 ) . For peak week and peak incidence , we restrain the analysis to forecasts made prior to the observed forecast peak , as most forecasts report the true peak values after the peak has been observed , resulting in equal ranking . The superensemble gave the most accurate forecast 8 . 6% of the time for peak timing , which was more frequently than 12 of the individual forecast methods , but less frequently than BWO and most models-filter combinations using SEIR and SEIRS models . While the BWO , pMCMC-SEIR and pMCMC-SEIRS forecasts of peak timing had the most frequent first place rankings , they also produced many predictions that received the lowest rankings . In contrast , the superensemble forecast , as well as the SEIR and SEIRS models coupled with EAKF , EKF , and RHF ensemble filter methods , had few low rankings . When superensemble weights were stratified by lead time of forecast relative to the observed outbreak peak , the resulting forecast rankings dramatically improved , with 54 . 3% of superensemble forecasts receiving a rank of 1 through 4 , surpassing all individual forecast methods ( S4 Fig ) . In forecasts of peak incidence , the baseline superensemble forecast gave the best prediction 10 . 8% of the time , which was more frequent than all individual forecasts except BWO , which was the highest ranked method for 17 . 1% of predictions . The superensemble was among the 4 worst forecasts less than 0 . 5% of the time , compared to 17 . 7% of BWO forecasts ( Fig 6 ) . The proportion of high to low ranking forecasts increases when superensemble weights are stratified by lead time of forecast ( S4 Fig ) . These results indicate that the superensemble provides a consistent advantage in forecasts of peak incidence , both in aggregate , as well as for any given prediction . For forecasts of total incidence , the baseline superensemble forecast had an average number of first place forecasts ( 5% ) , and was most often ranked between 7 and 13 . However , as with the other two metrics , the superensemble had far fewer low rankings than any individual method . Among individual ensemble models , those using the SEIRS structure were ranked highest . BWO , pMCMC-SEIR , pMCMC-SEIRS , PF-SEIR and PF-SEIRS had frequent rankings in both the top 4 and the bottom 4 .
Disagreement between competing forecasts of infectious disease outbreaks presents an obstacle to the interpretation and utilization of such forecasts . Here , we have presented a method for reconciling the disagreement among forecasts , while simultaneously improving overall forecast accuracy . We have shown that overall forecast accuracy for the timing and magnitude of peak influenza transmission is improved by combining individual forecasts into weighted-average superensemble forecasts . These superensemble forecasts were , on average , more accurate than any individual forecast method . In particular , the superensemble was less prone to producing a poor forecast . The advantage of the superensemble approach increases in circumstances where the relative accuracy of individual forecasts varies according to characteristics of the outbreak , or the location being forecast . The 21 individual forecast methods compared in this study varied in their performance , as well as their reliability . The SEIRS dynamical model coupled with the EKF , EAKF and RHF ensemble filter methods were consistently among the better individual forecasts . Other forecast methods , namely pMCMC-SEIRS and pMCMC-SEIR , and to a lesser extent , BWO , performed inconsistently in that predictions were either among the best or the worst of the competing forecasts . This type of inconsistent performance presents a challenge to the superensemble approach , as the good forecasts can cause the forecast to receive a relatively high weighting in the superensemble; however , by identifying the circumstances that lead to differences in relative forecast performance , adaptive weighting can then be employed to variably weight an individual forecast method highly when it is prone to perform well and discount it in other circumstances . Here we found that the performance of individual forecast methods varied according to geographic location , influenza season , and the timing of the forecast . We found that the timing of the forecast with respect to the outbreak peak was an important factor in determining relative forecast accuracy; consequently , stratifying superensemble weights by the actual lead time of the forecasts led to improvements in superensemble forecast accuracy . This improvement outweighed the gains made by simply eliminating the two most inconsistent forecast methods ( pMCMC-SEIR and pMCMC-SEIRS ) . While the idealized process of weighting individual forecasts by actual forecast lead is not possible in real time , stratifying weights by forecast-predicted lead or simply calendar week , which can serve as real-time proxies for actual forecast lead , proved beneficial in improving forecast accuracy . Stratifying superensemble weights by HHS region improved forecasts of peak and total incidence . This benefit may be related to regional differences in baseline and seasonal levels of influenza activity , or could be reflective of differences in the progression of influenza among regions . These findings provide a robust methodology for generating superensemble forecasts of influenza and other infectious diseases; however , the superensemble weights and the optimal stratification partitioning must be continually reevaluated and updated as new years of data become available , or as the geographical scale or resolution are altered . While this study focused on the forecast of point estimate outcomes , the superensemble approach can also be used to produce probability distribution functions of target metrics . For example , the superensemble forecasts presented here were associated with reasonable credible intervals , which were influenced by the choice of stratification variable ( S3 Fig ) . The calibration of individual and superensemble probabilistic forecasts remains an area of ongoing research . The algorithm used to create the superensemble is flexible , and can combine any number and any type of individual forecast method , provided retrospective forecasts are available for superensemble training . These findings can thus be applied operationally to competing forecasts of infectious disease in order to improve forecast accuracy , and to present a streamlined prediction to public health decision-makers .
Regional influenza activity is monitored by the U . S . Centers for Disease Control and Prevention ( CDC ) through the U . S . Outpatient Influenza- like Illness Surveillance Network ( ILINet ) . The CDC provides weekly near real-time estimates of regional influenza-like illness ( ILI ) , defined as the number of patients with flu-like symptoms ( fever with sore throat and/or with cough ) divided by the total number of patient visits at ILINet outpatient healthcare facilities in order to account for temporal and spatial variability in patient volume and reporting rates of health care providers [1] . At the city and state level , ILI was estimated by Google Flu Trends ( GFT ) , which used a statistical model relating weekly CDC ILI data to Google internet search queries[12] . GFT estimates are available for up to 115 cities and 50 states , from 2003 until the program was discontinued in 2015 . ILI is not specific to influenza , as it encompasses a range of respiratory infections . The World Health Organization and National Respiratory and Enteric Virus Surveillance System ( NREVSS ) provide weekly reports of the proportion of laboratory-confirmed positive tests for influenza virus . A more specific estimate of influenza activity can be obtained by multiplying ILI with corresponding weekly regional viral isolation information , resulting in a measure we refer to as ILI+ , defined as the number of influenza positive patients per 100 patient visits . As in our previous studies , we use city and state level GFT ILI estimates multiplied by regional NREVSS viral isolation rates to obtain ILI+ , as the metric of observed influenza incidence [5] . We produced weekly forecasts of three target metrics for each influenza outbreak: the highest observed weekly ILI+ ( peak incidence ) ; the week during which peak ILI+ occurred ( peak week ) ; and the total ILI+ over the influenza season , which we define as a 20-week period beginning on the 45th calendar week of the year ( total incidence ) . Weekly forecasts of influenza outbreak trajectories , outbreak peak ILI+ , total ILI+ and outbreak peak timing were produced using 21 different forecast methods . Of the 21 methods , 20 are variations of a mathematical model of disease transmission coupled with a data assimilation , or filtering , method , and 1 is a statistical model based on historically observed outbreaks . We generated retrospective forecasts for 95 cities and the 48 contiguous United States with available records during the 2005–2006 through 2014–2015 influenza seasons , excluding the pandemic seasons of 2008–2009 and 2009–2010 . Pandemic seasons were excluded because the individual forecast systems used in this study were designed specifically for seasonal outbreaks . While the model-filter forecasts could be adapted to forecast pandemics or irregularly timed outbreaks ( for example [13–15] ) , the Bayesian Weighted Outbreaks method in particular is not appropriate for forecasting pandemics as it relies on the assumption that the outbreak in progress will be similar in timing and magnitude to previously observed outbreaks . Superensemble forecasts were created for the 2005–2006 through 2014–2015 influenza seasons ( excluding the pandemic seasons 2008–2009 and 2009–2010 ) by taking the weighted average of the 21 individual weekly forecasts for each location . The superensemble weights , which dictate the contribution of each individual forecast to the superensemble , are determined using maximum likelihood estimation of the conditional probability distribution function ( PDF ) over a selected number of training forecasts: p ( y'|f'1 , m , … , f'21 , m ) =∑k=121wk , mgk ( y'|f'k , m ) ( 4 ) where the left hand side of the equation is the probability distribution of the superensemble forecast , and the right hand side is the weighted sum of the 21 individual forecast distributions , gk ( y’|f’k , m ) . More formally , y’ is the true value of the training outbreak metric m ( peak timing , peak incidence or total incidence ) , wk , m is the probability that individual forecast method k is the most accurate method , and gk ( y’|f’k , m ) is the PDF of y’ , conditional on training forecast f’k , m , given that f’k , m is the most accurate forecast of y’m . This conditional PDF is assumed to be normal with mean f’k , m and standard deviation σ . For simplicity , σ is assumed equal for all individual forecasts , and is determined through the maximum likelihood estimation of Eq 4 to obtain wk , m , which serve as the superensemble weights ( see Raftery et al . [23] for full details ) . Superensemble weights for season N are trained using individual forecasts from 2003–2004 through season N-1 . The set of BWO training forecasts , f’BWO , m , are produced using a leave-one-out approach for training seasons 1 through N-1 . That is , training forecasts for each season from 1 through N-1 were constructed using trajectories using all other seasons between 1 and N-1 . The number and diversity of candidate trajectories increases over time , as each subsequent year adds 145 additional ILI+ trajectories to the pool . The model-filter forecasts do not use historical observations , and thus do not require a leave-one-out approach for training forecasts . The superensemble weights wk , m are then applied to the point estimates of the target metric from the 21 individual forecasts , fk , for the time t , location l , and metric m of interest to obtain the superensemble forecast , SE: SEm ( t , l ) =∑k=121wk , mfk , m ( t , l ) ( 5 ) The probability distribution function of the superensemble forecast obtained from Eq 4 is used to determine credible intervals around each forecast ( for example , S2 Fig ) . The width of the credible intervals is a function of both the spread of the point estimates from the 21 individual forecast methods , as well as the estimated variance of each individual forecast over the training period ( σ2 ) [23] . The baseline superensemble forecasts were made by applying a single set of superensemble weights across all locations and times within an influenza season . However , based on previous analyses of individual forecast system performance ( e . g . [2 , 5 , 6] ) , we hypothesized that superensemble performance would improve if superensemble weights were stratified by the following variables: calendar week of forecast , lead time relative to forecast peak ( weeks between the week of forecast initiation and the predicted peak ) , geographic region , population size , and population density . A final set of forecasts was produced by stratifying superensemble weights by lead time relative to the actual peak ( weeks between the week of forecast initiation and the week of the true peak ) . While this weighting scheme could not be implemented in an operational real-time forecast , it is useful to know how the superensemble would perform under this idealized condition , as this may represent an upper bound to improvements that can be achieved using a weighting scheme based on forecast timing . The method for stratifying superensemble weights consisted of dividing forecasts into bins according to the variable of interest . We then obtained weights for each bin by including only training forecasts falling into that bin in the algorithm described in Eq 4 . In setting the bin sizes for each variable , our objective was to resolve potential differences in forecast performance . This objective was balanced by the need to include sufficient training forecasts in each bin to avoid over-fitting . Forecasts stratified by lead time ( actual or predicted ) were grouped using the following bin edges , where negative values indicate weeks prior to the peak and positive numbers are weeks after the peak: [<-8 , -8 , -7 , -6 , -5 , -4 , -3 , -2 , -1 , 0 , 1 , 2 , 3 , 4 , 5–8 , 9–12 , >12] . We selected a fine resolution of 1 week around peak , with wider bins at either end , as fewer outbreaks have lead times in these categories . Actual lead time is simply the week of the forecast minus the week that peak is eventually observed . Forecast predicted lead time is calculated by taking the mean prediction of peak week from the 21 individual forecasts , and subtracting this mean value from the week of forecast . Geographic regions were grouped according to the ten US Health and Human Services Regions , as these are the standard geographical groupings used by the CDC to describe influenza activity . Calendar week groupings were delineated by individual weeks . Population density and populations size were arbitrarily binned into quintiles for cities and terciles for states . | Timely forecasts of infectious disease transmission can help public health officials , health care providers , and individuals better prepare for and respond to disease outbreaks . Work in recent years has led to the development of a number of forecast systems . These systems provide important information on future disease incidence; however , all forecasting systems contain inaccuracies , or error . This error can be reduced by combining information from multiple forecasting systems into a superensemble using Bayesian averaging methods . Here we compare 21 forecasting systems for seasonal influenza outbreaks and use them together to create superensemble forecasts . The superensemble produces more accurate forecasts than the individual systems , improving our ability to predict the timing and severity of seasonal influenza outbreaks . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"influenza",
"applied",
"mathematics",
"health",
"care",
"simulation",
"and",
"modeling",
"algorithms",
"seasons",
"health",
"care",
"providers",
"probability",
"distribution",
"mathematics",
"forecasting",
"statistics",
"(mathemat... | 2017 | Individual versus superensemble forecasts of seasonal influenza outbreaks in the United States |
The proteins involved in smooth muscle's molecular contractile mechanism – the anti-parallel motion of actin and myosin filaments driven by myosin heads interacting with actin – are found as different isoforms . While their expression levels are altered in disease states , their relevance to the mechanical interaction of myosin with actin is not sufficiently understood . Here , we analyzed in vitro actin filament propulsion by smooth muscle myosin for -actin ( A ) , -actin-tropomyosin- ( A-Tm ) , -actin-tropomyosin- ( A-Tm ) , -actin ( A ) , -actin-tropomyosin- ( A-Tm ) , and -actin-tropomoysin- ( A-Tm ) . Actin sliding analysis with our specifically developed video analysis software followed by statistical assessment ( Bootstrapped Principal Component Analysis ) indicated that the in vitro motility of A , A , and A-Tm is not distinguishable . Compared to these three ‘baseline conditions’ , statistically significant differences ( ) were: A-Tm – actin sliding velocity increased 1 . 12-fold , A-Tm – motile fraction decreased to 0 . 96-fold , stop time elevated 1 . 6-fold , A-Tm – run time elevated 1 . 7-fold . We constructed a mathematical model , simulated actin sliding data , and adjusted the kinetic parameters so as to mimic the experimentally observed differences: A-Tm – myosin binding to actin , the main , and the secondary myosin power stroke are accelerated , A-Tm – mechanical coupling between myosins is stronger , A-Tm – the secondary power stroke is decelerated and mechanical coupling between myosins is weaker . In summary , our results explain the different regulatory effects that specific combinations of actin and smooth muscle tropomyosin have on smooth muscle actin-myosin interaction kinetics .
Differential expression of smooth muscle contractile proteins has been associated with organismal development [1] , contractile phenotypes [2]–[4] , and pathologies , e . g . preterm labour , hypertrophic bladder , or airway hyper-responsiveness [5]–[7] . While the role of the smooth muscle myosin isoforms has been extensively investigated [7]–[9] , the functional implications of the differential expression of specific actin and actin-regulatory protein isoforms remain elusive [4] . In smooth muscle , actin isoforms are expressed from four different genes , yielding “vascular muscle” - and “enteric muscle” -actin , as well as non-muscle ( cytoplasmic ) - and -actin . The muscle isoforms are associated with the contractile apparatus , the non-muscle isoforms with cytoskeletal structures [5] . Muscle -actin is generally associated with tonic , -actin with phasic smooth muscles [5] , [10] , [11] . An anti-proportional relationship between the absolute levels of - and -actin has been established [2] . Disease-related expression differences in - vs . -actin have been found [6] . Functional differences between - and -isoforms were searched for in molecular mechanics experiments , but , to our knowledge , no differences were detected [12]–[15] . Insight from tissue level mechanics seems lacking , too [4] . Smooth muscle tropomyosin affects the weak to strong binding of ATP-activated myosin to actin: tropomyosin can be in an ON state supporting myosin strong binding , or an OFF state hindering myosin strong binding [10] , [16] . When regulated by caldesmon-calmodulin , dependent on the caldesmon-calmodulin activation state , smooth muscle tropomyosin is stabilized in the open or the closed state , increasing or decreasing the rate of myosin cycling compared to the rate without any tropomyosin being present [10] , [17] . Tropomyosin forms chains along actin filaments by a head-to-tail overlap of consecutive tropomyosin molecules . This overlap leads to an increased cooperativity in the switching between the ON and the OFF state . Compared to striated muscle tropomyosin isoforms , a stronger cooperativity between tropomyosin displacement due to stronger end-to-end binding between tropomyosin molecules is observed , as well as a greater bias for the ON conformation [16] , [18] , [19] . Similar to striated muscle tropomyosin , smooth muscle tropomyosin facilitates cooperative binding of myosin to actin: above a critical ratio of myosin heads per actin monomer , myosin heads cooperatively displace tropomyosin into the ON state so that further myosin binding is facilitated; below a critical density or activation by phosphorylation , tropomyosin remains mostly in the OFF state [20] , [21] and inhibits myosin cycling [10] , [19] . Tropomyosin is expressed from the same two genes in non-muscle , striated muscle , and smooth muscle cells . In smooth muscle , alternative splicing yields two smooth muscle specific isoforms ( tropomyosin- and tropomyosin- ) , one from each gene [22] . In vivo , tropomyosin- and tropomyosin- mostly occur as heterodimers , making functional differentiation between the isoforms difficult [10] , [22] . In disease states , however , expression differences between both isoforms can be observed [6] , raising the question of functional differences between these two isoforms , especially in interaction with other differentially expressed contractile protein isoforms . Crystallized N-terminal fragments of tropomyosin- and tropomyosin- displayed differences in the heterodimerization properties of tropomyosin- vs . tropomyosin- and a greater head-to-tail overlap of tropomyosin- than that of tropomyosin- [23] . These structural results were interpreted as indication of negligible differences in tropomyosin's interface for actin binding and more important differences in the surfaces available for mediation of actin-myosin interactions as well as the binding of other proteins [23] . However , actin affinity ( in terms of binding constants ) of smooth muscle tropomyosin- was found to be times greater than that of tropomyosin- [24] , [25] . In this study , we use an in vitro motility assay to investigate differences in the propulsion of “vascular” -actin vs . “enteric” -actin by smooth muscle myosin in the presence of smooth muscle tropomyosin- , tropomyosin- , or in the absence of tropomyosin , see Fig . 1 A and Tab . 1 . We develop and simulate a mathematical model to establish the differences in actin-myosin interaction kinetics that underlie the experimentally observed differences .
Using our specifically developed analysis software , we extracted the following features of actin sliding: mean sliding velocity ( ) , the motile fraction ( ) , the average run time ( ) , and the average stop time ( ) ( Fig . 1 B , C ) . These features were extracted for the different experimental conditions ( Tab . 1 ) and resolved by actin filament length ( ) ( Fig . 2 ) . For A-Tm a consistent increase is apparent ( Fig . 2 A ) . , , and do not immediately suggest consistent differences , ( Fig . 2 B–D ) . In spite of high filament counts ( Fig . 2 D , inset ) , the width of the confidence intervals compared to potential differences makes a direct , conclusive inference difficult , especially for and at . The resolved features represent a simultaneous measurement of values , whose interdependence cannot be judged a priori . We applied a Principal Component Analysis ( PCA ) to reduce the dimensionality of our data and remove correlations between values , which would otherwise inflate statistical significance . Transformation into the three Principal Components ( PCs ) explaining most of the variance indicates that consistent differences between the experimental conditions exist ( Fig . 3 A , B ) . Our statistical analysis detected no differences between A , A , and A-Tm , which will therefore be referred to as baseline conditions that show no effect; A-Tm , A-Tm , and A-Tm are all different from the baseline conditions , as well as from each other ( Fig . 3 C , D ) . To support the conclusions from our statistical analysis , we executed a hierarchical cluster analysis . Based on the relatively large reduction of linkage when going from four to five clusters , a number of four clusters was chosen ( Fig . 3 E ) . In the PC space , the four clusters appear similar to the above separation into one baseline and three regulated conditions ( Fig . 3 F , G ) . Indeed , the four clusters form a clear representation of the A , A , A-Tm baseline conditions , and the three distinctly regulated conditions A-Tm , A-Tm , and A-Tm , ( Fig . 3 H ) . Thus , two independent methods of statistical assessment indicate that only A-Tm , A-Tm , and A-Tm are significantly regulated , while for each of them the regulation affects actin sliding in the in vitro motility assay in a distinctly different manner ( Fig . 3 I ) . Next , we wanted to attribute the differences that had been detected using PCA to molecular mechanical features . Thus , we evaluated the motility features' fold changes relative to A , averaged over . For A-Tm , is statistically significantly increased to 1 . 12 times the baseline value ( Fig . 4 A ) . For A-Tm , is decreased to 0 . 96-fold , is increased by a factor of 1 . 6 relative to the baseline value , though both changes show up only as strong trends ( Fig . 4 B , C ) . For A-Tm , is elevated 1 . 3-fold , which also shows up as a strong trend only ( Fig . 4 D ) . When and are analyzed together , the joint fold changes for A-Tm become statistically significant ( Fig . 4 E ) . When only short actin is considered , is statistically significantly elevated to 1 . 7 times the baseline value ( Fig . 4 F ) . Note that each condition's differences are found in different features , which is coherent with the PCA finding that the regulated conditions are each affected by tropomyosin in a distinct manner . To theoretically understand the regulatory effect that tropomyosin has on actin-myosin interactions , we constructed a mathematical model of the kinetics of a myosin-coated surface interacting with actin filaments of different length . Stochastic simulations of our model produce time courses ( Fig . 1 C ) . Averaging these time courses gives , all other features of actin sliding can be extracted in exactly the same way as from experimental data . Our model is an extension of our earlier model of the group action of myosins propelling actin filaments in the in vitro motility assay [26] . Briefly , the model assumes that myosin moves actin by two mechanical steps , the main power stroke and a secondary mechanical step preceding myosin detachment [27] , [28] . When several myosins are simultaneously bound to the same actin filament , they are mechanically coupled via the filament . Thus , the individual myosins' steps cause a change in the mechanical configuration of the overall system of bound myosins and the actin filament . Consequently , mechanical work might have to be exerted on or might be released from the actin-myosin system during the execution of an individual myosin's mechanical step . This mechanical work affects the strain-dependent rates of both mechanical transitions , the main power stroke and the secondary pre-detachment step . The overall number of myosin binding sites that are accessible on a given actin filament ( ) is assumed to be proportional to . Using the helix repeat of actin ( 0 . 0355 μm ) as an approximate binding site distance [27] , [29] , the ranges were adjusted to correspond to the ranges used in the different analysis steps . For details regarding our mathematical model , see Text S1 . A set of model parameters was determined to mimic the baseline condition ( Fig . 5 ) . These baseline parameters were altered so as to mimic the changes in resolved features that were observed experimentally for the A-Tm , A-Tm , and A-Tm conditions ( Fig . 5 ) . The scalar fold changes in motility features were determined in the same way as from the experimental data ( Fig . 6 ) . The resolved motility features as well as the fold changes capture the experimentally observed differences between the baseline conditions and the conditions that exhibited statistically significant effects . The changes in model parameters that were necessary to mimic the experimentally observed differences point towards the aspects of actin-myosin interaction kinetics that are changed in the different conditions ( Fig . 7 ) . For A-Tm , all kinetic rates ( , , ) are increased 1 . 15-fold . For A-Tm , the impact of mechanical coupling between myosins on the rate of the mechanical transitions ( ) is increased by a factor of 1 . 2 . For A-Tm , is reduced to of the baseline value , and is reduced to of the baseline value .
We investigated in vitro the relevance of actin and smooth muscle tropomyosin isoforms to the mechanical action of smooth muscle myosins on actin . In accordance with prior studies [4] , [12]–[15] , no differences between actin isoforms could be detected . However , the sequence differences between actin isoforms are confined to regions of interaction with regulatory proteins [30] , suggesting potential mechano-chemical differences in the presence of such regulatory proteins . In vitro studies in solution ( i . e . not on a motility surface ) showed a different binding affinity between actin and smooth muscle tropomyosin [24] , [25] . Here , we establish that , in the presence of both tropomyosin- and tropomyosin- , the molecular mechanics differ between - vs . -actin . Thus , the sequence differences between actin isoforms not only affect actin-tropomyosin interactions , but also actin-myosin mechano-chemistry . Importantly , we found that -actin is significantly regulated only by tropomyosin- , while -actin is regulated by both tropomyosin- and tropomyosin- . More specifically , the regulation by tropomyosin has distinct effects on in vitro molecular mechanics in three regulated actin-tropomyosin combinations ( experimentally determined ) , suggesting three different modes by which tropomyosin regulation affects actin-myosin mechano-chemistry ( determined by model parameter adjustment ) : ( 1 ) A-Tm – is increased 1 . 2-fold . This is caused by a 1 . 15-fold increase in the myosin attachment rate to actin , the unstrained myosin main power stroke rate , and the unstrained rate of detachment of unloaded myosin from actin . ( 2 ) A-Tm – is reduced to 0 . 96-fold and is increased 1 . 6-fold . This is caused by an increase in the impact that myosin-to-myosin mechanical coupling has on rates of mechanical steps of myosin by a factor of 1 . 2 . ( 3 ) A-Tm – is increased 1 . 7-fold for short actin . This is caused by a decrease in the unstrained rate of detachment of myosin from actin to 0 . 75 times the baseline value and a decrease to 0 . 8-fold in the impact that myosin-to-myosin mechanical coupling has on rates of mechanical steps of myosin . Note that no quantitative adjustment , e . g . minimization of sum of squared errors , was used to determine the model parameter changes stated above . In consequence , the numeric parameter changes stated above should be understood as qualitative indicators of the general nature of changes in actin-myosin interaction kinetics . The changes in kinetic parameters determined for A-Tm using our model-based assessment are in line with what is known for this condition from ATPase assays with skeletal muscle myosin and actin . Sobieszek determined that gizzard smooth muscle tropomyosin increases the ATPase , while the affinity of myosin for the actin-tropomyosin complex was not affected at myosin∶actin ratios of less than one myosin head per 4 to 6 actin monomers – which is the relevant regime for our experiment [31] . These observations were attributed to increases in the rates of the kinetic steps after myosin binding to the actin-tropomyosin complex , which is concurrent with the general increase in the unstrained kinetic rates we observed for A-Tm . Williams et al . found results that are similar to Sobieszek's and were measured at low myosin concentrations and low ionic strengths corresponding to those used in our motility assays [18] . Sufficient evidence exists to state that smooth muscle tropomyosin does regulate smooth muscle myosin interactions with actin , and thus , the resulting molecular mechanics [10] , [20] , [21] , [32] . Regarding the functional relevance of the smooth muscle tropomyosin isoforms , however , several not mutually exclusive mechanisms by which the isoforms could affect molecular mechanics have been put forward [22]: With regards to smooth muscle contraction , smooth muscle myosin is the most central interaction partner of actin . We investigated its mechanical action on actin in the background of different actin and tropomyosin isoforms' interaction . Because we found that tropomyosin isoforms are indeed relevant to the regulation of actin-myosin interactions , all three mechanisms are possible for actin-tropomyosin-myosin interactions . However , the observed difference between the tropomyosin isoforms depends on the actin isoform . This suggests direct interactions between the actin filament and tropomyosin , highlighting the second mechanism . Our mathematical model does not include tropomyosin-mediated myosin binding cooperativity . Binding cooperativity is often assessed by changing the myosin-actin ratio or the myosin activation level [10] , [19]–[21] . Within the scope of this study , one detectable effect of binding cooperativity differences would be a shift in the actin length at which bifurcations between non-motile and motile behavior occur [26] . These bifurcation lengths depend on the number of myosins effectively bound to actin and would be affected by cooperativity-mediated changes in the effective rate of myosin binding to actin . We found no significant shifts in these lengths between the conditions , and therefore no indication of differences in binding cooperativity . Like any automation of a manual analysis procedure , our video analysis software makes the analysis of large data sets feasible and prevents differences occurring between different days or operators . A specific advancement is the automated machine learning-based approach to quality control of the filament traces . Further , a result management framework was devised , which allows keyword-based queries into annotated data sets and the application of custom analysis functions . Utility functions allow the creation of customized MatLab scripts to interact with results . This supports customized analyses of existent data sets also by computational scientists without their own motility assays , as well as the “high throughput” necessary for determining statistical distributions and resolved curves of motility features . The MatLab scripts with instructions are released as open source ( In Vitro Motility Assay Automated Analysis – ivma3 , http://code . google . com/p/ivma3/ ) . FIESTA is another openly accessible analysis software that can be used for in vitro motility assays [35] . It reaches nanometer precision and allows interactive assessment of filament motility in a graphical user interface . Differently , our software provides less precise image analysis and tracking at the benefit of fast processing of a high number of experiments and the possibility to execute specific analyses on large data sets in an automated fashion . The statistical assessment uses bootstrapping to maintain the high filament count that is necessary for a high resolution while still giving account of the variation present in the experiment . To explore the results and counteract inflation of statistical significance resulting from resolved analysis , PCA was used on the bootstrapped data sets . We could not find existent examples of this combination of PCA and bootstrapping – other studies estimate the variation of PCA itself [36] , [37] , or assess the variation of bootstrap scores ( loadings ) [38] , [39] . More detailed assessment of in vitro motility and the observed specificity of regulation require more specific theoretical explanations of the molecular mechano-chemistry underlying these observations . Our relatively simple stochastic model generated data sets that were analyzed in the same way as actual experimental data , indicating how the different actin and tropomyosin isoform combinations affect actin-myosin interaction kinetics . While providing a perspective beyond mere presentation of our experimental findings , the simplicity of our model as well as the procedure by which model parameters were adjusted to mimic the experimental observations call for future work . From an experimental perspective , molecular mechanical assays using expression and site-directed mutagenesis of actin and tropomyosin seem promising .
We developed an automated video analysis software which executes the following steps . Raw video data are preprocessed ( image enhancement and frame merging to a time resolution of s ) and turned into binary images . Filament objects and their properties are extracted from individual frames using connected components methods . Filaments are tracked throughout consecutive frames based on their centroid position and area . Frame-to-frame velocities ( ) are calculated from centroid displacements between two consecutive frames . Filament length ( ) and travelled path lengths are determined based on a transformation of image objects into rectangles of same area and perimeter , the longer edge representing lengths . A filament's mean trace velocity ( ) is determined by dividing the total distance that the filament's tip has travelled by the time the filament was present for ( ) . Filament traces with filament crossing events or signs of irregular motion were removed by a machine-learning algorithm , which was trained on subset of our data that we scored by hand . The automated video analysis was assessed using computer-generated mock motility videos , the automated quality control was evaluated against hand-scored data sets . For details see Text S1 . Statistical significance was assumed for . Statistical comparisons were executed by bootstrapping of the compared statistic; statistical significance was assumed where no overlap exists between the confidence intervals of the compared conditions . For details see Text S1 . | Dependent on the required physiological function , smooth muscle executes relatively fast contraction-relaxation cycles or maintains long-term contraction . The proteins driving contraction – amongst them actin , tropomyosin , and the contraction-driving myosin motor – can show small changes in the way they are constructed , they can be expressed as different “isoforms” . The isoforms are supposedly tailored to support the specific contraction patterns , but for tropomyosin and actin it is unclear exactly how the isoforms' differences affect the interaction of actin and myosin that generates the muscle contraction . We measured actin movement outside the cellular environment , focusing on the effects of different isoform combinations of only actin , myosin , and tropomyosin . We found that the actin isoforms cause differences in the mechanical interaction only when tropomyosin is present , not without it . Also , all different actin-tropomyosin combinations affected the mechanical interactions in a different way . In our experiments we could not directly observe the mechanical interactions of actin , tropomyosin , and myosin , so we reconstructed them in a mathematical model . With this model , we could determine in detail how the different actin-tropomyosin combinations caused the differences that we observed in our experiments . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Molecular Mechanical Differences between Isoforms of Contractile Actin in the Presence of Isoforms of Smooth Muscle Tropomyosin |
Epidemiological data from Zimbabwe suggests that genital infection with Schistosoma haematobium may increase the risk of HIV infection in young women . Therefore , the treatment of Schistosoma haematobium with praziquantel could be a potential strategy for reducing HIV infection . Here we assess the potential cost-effectiveness of praziquantel as a novel intervention strategy against HIV infection . We developed a mathematical model of female genital schistosomiasis ( FGS ) and HIV infections in Zimbabwe that we fitted to cross-sectional data of FGS and HIV prevalence of 1999 . We validated our epidemic projections using antenatal clinic data on HIV prevalence . We simulated annual praziquantel administration to school-age children . We then used these model predictions to perform a cost-effectiveness analysis of annual administration of praziquantel as a potential measure to reduce the burden of HIV in sub-Saharan Africa . We showed that for a variation of efficacy between 30–70% of mass praziquantel administration for reducing the enhanced risk of HIV transmission per sexual act due to FGS , annual administration of praziquantel to school-age children in Zimbabwe could result in net savings of US$16–101 million compared with no mass treatment of schistosomiasis over a ten-year period . For a variation in efficacy between 30–70% of mass praziquantel administration for reducing the acquisition of FGS , annual administration of praziquantel to school-age children could result in net savings of US$36−92 million over a ten-year period . In addition to reducing schistosomiasis burden , mass praziquantel administration may be a highly cost-effective way of reducing HIV infections in sub-Saharan Africa . Program costs per case of HIV averted are similar to , and under some conditions much better than , other interventions that are currently implemented in Africa to reduce HIV transmission . As a cost-saving strategy , mass praziquantel administration should be prioritized over other less cost-effective public health interventions .
Sub-Saharan Africa continues to bear a disproportionate share of the global HIV burden [1] . Parallel to HIV , schistosomiasis is highly prevalent in sub-Saharan Africa , where approximately two-thirds of schistosomiasis cases result from urinary and genital tract infections caused by Schistosoma haematobium [2] , [3] . Epidemiological studies have observed that female genital schistosomiasis ( FGS ) is associated with increased odds of having HIV [4]–[7] . Thus , mass preventive chemotherapy against schistosomiasis may not only reduce schistosomiasis morbidity and mortality in sub-Saharan Africa , but could simultaneously offer an innovative approach to HIV/AIDS prevention [8] , [9] . There are three lines of evidence that indicate to an association between FGS and an elevated risk of HIV infection . Firstly , there is a strong statistical association between FGS and HIV transmission . Several cross-sectional epidemiological studies have reported that in sub-Saharan Africa , the region most heavily affected by the HIV/AIDS pandemic , women with FGS have a three- to four-fold increased odds of having HIV compared to women without FGS [5] , [7] . Secondly , it is physiologically plausible that FGS elevates susceptibility to HIV infection through its lesions and chronic inflammation of genital tract , as well as chronic immunomodulatory effects [4] , [10] . Thirdly , the presence of schistosomal lesions is common in the vulva and the lower vagina of FGS infected women before puberty [11] , [12] , making it likely that the schistosomal infection typically precedes HIV infection . Thus , collectively , the strong statistical association between FGS and HIV , the biologic plausibility of the association , and the temporal association suggest that FGS infection exacerbates HIV transmission . Praziquantel is a highly effective anti-schistosomal chemotherapy agent against schistosomal morbidity that may be able to prevent FGS and the clinical manifestations associated with enhanced HIV susceptibility [12]–[14] . We conducted a cost-effectiveness analysis of mass praziquantel administration for HIV/AIDS prevention from the perspective of health payers , such as national government or international donors , which are the major providers of both mass schistosomiasis treatment and HIV antiretroviral therapy in sub-Saharan Africa [8] , [15] , [16] . For this , we constructed a model of the joint dynamics of HIV and FGS among sexually active individuals that we parameterized with epidemiological data from a cross-sectional study of rural Zimbabwean women [3] , [7] . We calculated cost-effectiveness ratios of a potential large-scale intervention based on praziquantel as a preventive anthelminthic chemotherapy in terms of reducing HIV incidence . We found that mass preventive chemotherapy of schistosomiasis might prove not only very cost-effective , but even cost-saving in preventing HIV infection in S . haematobium-endemic areas .
To estimate the potential cost-effectiveness of preventing HIV infection in sub-Sahara Africa through mass treatment of schistosomiasis , we constructed a mathematical model for genital schistosomiasis and HIV infections in the adult population ( aged 15–49 , corresponding to the standard age-range for WHO reporting of HIV prevalence [17] ) . We parameterized the model by applying a Bayesian inference approach to cross-sectional epidemiological data on FGS and HIV among rural Zimbabwean women [3] , [7] . We calculated the cost-effectiveness from a health care system perspective , because the Zimbabwean health care system and international donors are the primary providers of treatment costs for HIV [15] , [18]–[20] . Only direct medical costs to the health provider were considered , including the costs of mass administration of praziquantel and lifetime treatment costs of an HIV infection . We quantified the cost-effectiveness of praziquantel in terms of HIV cases averted and averted medical care costs over the duration of the intervention . Costs and benefits were discounted at an annual rate of 3% , according to WHO recommendations [21] . Initially , we modeled a Zimbabwean population with no praziquantel treatment . We divided the population into males and females , and high and low sexual activity risk groups defined according to rate of sexual partner change . The state variable of the model are given by : is gender ( 1 = female , 2 = male ) , is the sexual activity group ( 1 = high risk , 2 = low risk ) , is the HIV infection status ( 1 = susceptible , 2 = infected ) , and is the FGS infection status for women ( 1 = non-infected , 2 = infected ) . For men was always equal to 1 . Individuals enter the model at the onset of sexual activity ( assumed to be age 15 ) , with a proportion of women infected with FGS . From age 15 to 49 , women acquire FGS and/or HIV at rates dependent on their infection status of each disease . As re-infection rates of schistosomiasis are high in S . haematobium-endemic areas [12] , [22] , we assumed no natural recovery from FGS . The acquisition of HIV was modeled as a function of the rate of partner change , the mixing between individuals of the different sexual risk groups , the number of sex acts in partnerships per year , and the rate of HIV transmission per sex act ( ) . The rate of forming sexual partnerships is allowed to decline with HIV mortality to model potential behavior change as a response to the HIV epidemic [23] . FGS infection is assumed to elevate the risk of HIV transmission per sex act by a factor , which was parameterized from epidemiological data on HIV-FGS co-infection . Therefore , the rate of HIV transmission from men to FGS-infected women is . We assumed that FGS is primarily acquired in childhood [13] , [22] . To determine S . haematobium prevalence , we developed a model for S . haematobium dynamics ( see Appendix for details ) from which we quantify FGS prevalence in the HIV model . The compartmental diagram in Figure 1 illustrates the flow of individuals as they face the possibility of acquiring each infection . We used Bayesian Markov Chain Monte Carlo ( MCMC ) [24] to fit the HIV-FGS model to epidemiological data on FGS and HIV prevalence and co-infection among rural Zimbabwean women [3] , [7] . The MCMC approach allowed us to estimate uncertain epidemiological parameters by combining prior information about these parameters from epidemiological studies ( Appendix ) , empirical HIV-schistosomiasis data , and dynamic model prevalence predictions . We determined the likelihood function of our MCMC approach from empirical data [3] , [7] , assuming normal distributions for HIV and FGS prevalence and lognormal distribution for the odds ratio . Using this approach , we derived a posterior distribution for each epidemiological parameter ( Table 1 ) for which the model gives the best-estimated trajectory for HIV and FGS prevalence and odds ratio of the association between HIV and FGS . Our model predicted the annual prevalence of HIV/AIDS in a Zimbabwean setting over a baseline period of 10 years starting in 2000 . To validate the model , we compared these predictions of HIV prevalence to observed HIV prevalence from Zimbabwean antenatal clinics [17] . Praziquantel is a highly effective anti-schistosomal therapy agent against schistosomal morbidity [8] , with no serious or long-lasting side effects [25] , [26] . The WHO recommends that mass administration of praziquantel be undertaken annually in schistosomiasis-endemic areas , targeting school-age children [27] . We investigated a strategy for praziquantel administration , where praziquantel is annually administered to school-age children ( ages 5–14 years ) ( school-age strategy ) . We assumed that all praziquantel-treated girls reaching age 15 uninfected with FGS were less likely to develop FGS [14] , [22] than those who had been infected in childhood . We considered two scenarios for the potential effect of mass praziquantel administration for reducing HIV incidence either in terms of reducing the risk of HIV transmission or reducing FGS prevalence . In the first scenario , we assumed that women who have received praziquantel treatment during childhood have a reduced risk of HIV transmission ( 30–70% ) relative to FGS infected women who did not receive treatment during childhood , thus reducing by 30–70% . In the second scenario , we assumed that women who have received praziquantel treatment during childhood have a reduced FGS prevalence relative to those who did not receive treatment ( 30–70% ) . We simulated each mode of action to predict the potential effect of mass treatment of schistosomiasis on reducing HIV incidence at the population level . In these analyses we assumed that the Zimbabwean population aged 15–49 years old was 4 million in 2000 [28] , [29] . To estimate the funding required to implement the alternative strategies of praziquantel treatment , we used the WHO dose for praziquantel ( 2 . 5 tablets of 600 mg per child per year ) [27] which costs US$0 . 08 [30] , while the total delivery cost per child treated , including delivery , training , social mobilization , capital equipment , and administrative costs , was US$0 . 21 ( US$0 . 06–2 . 23 ) [31] , [32] . We used US$0 . 29 ( 0 . 008+0 . 21 ) as a base value for the total cost of treatment per individual . Medical costs for HIV treatment and care included provider-initiated testing ( diagnostic and routine offer of testing ) , treatment and prophylaxis for opportunistic infections , antiretroviral therapy , laboratory monitoring of antiretroviral therapy , and palliative care . The lifetime treatment costs of an HIV infection were based on recent estimates of the costs in sub-Saharan Africa of US$3469 in 2004US$ at a 5% discount rate [33] . For a 3% discount rate , this cost becomes US$3695 . The antiretroviral therapy coverage in Zimbabwe was assumed to be 34% ( 28–40% ) [34] . The Zimbabwean government expenditure on health , other than HIV related spending , was estimated to be US$26 ( US$12–US$41 ) per capita annually [16] , [35] . To compute the non-HIV/AIDS health expenditure per HIV case averted , we assumed that average age of HIV/AIDS acquisition among Zimbabwean is 25 years old [36] , and the life expectancy at age 25 is 28 years [37] . All costs were discounted at 3% and given in US$2004 [21] . We calculated the potential cost-effectiveness of mass praziquantel administration to school-age children for reducing HIV transmission . The status quo was to be no mass treatment of schistosomiasis , as is currently the case in Zimbabwe . We measured the effectiveness of the intervention in terms of the number of HIV cases averted during a baseline intervention period of 10 years . We measured cost-effectiveness in terms of program costs per HIV case averted and averted medical care costs over the ten-year intervention period as the base line duration , which was also varied to assess the impact of the duration of intervention on its cost-effectiveness . To identify the contribution of each model parameter to the variability of the number of HIV cases averted , we calculated the partial rank correlation coefficients ( PRCCs ) [38] . PRCC quantifies the degree of monotonicity between a specific input parameter and an outcome measure . For this purpose , we used a Latin Hypercube Sampling [38] to sample 10 , 000 estimates of input parameters from the posterior distributions of the epidemiological parameters ( Table 1 ) and distributions of the cost and efficacy of mass administration of praziquantel as well as antiretroviral therapy coverage and non-HIV/AIDS health expenditure ( Table S1 ) .
To fit our dynamic model to the epidemiological HIV/FGS data from rural Zimbabwe [3] , [7] , we used a Markov Chain Monte Carlo ( MCMC ) method to draw values of epidemiological parameters from prior distributions based on estimates available in the literature . We derived a posterior distribution for each epidemiological parameter ( Table 1 ) for which the model gives the best-estimated trajectory for HIV and FGS prevalence ( Table 2 ) . We estimated the value for , the coefficient by which FGS increases the risk of HIV transmission per sexual act , as 5 . 9 ( 95% CI: 3 . 8–9 . 1 ) The lower bound of the 95% credible interval ( CI ) , for the estimate of , lies above one , consistent with the hypothesis that FGS enhances the risk of HIV transmission . Moreover , these estimates are in agreement with those of epidemiological and clinical studies that have observed that the presence or history of genital ulcers is associated with greater susceptibility to HIV transmission per sex act of 1 . 4 to 19 . 5 relative to the absence of genital ulcerative disease [39] . We validated the projected HIV prevalence of our model against the Zimbabwean antenatal clinic data from 1990–2006 ( Figure 2 ) , which had not been used to fit the model . We used the estimated mean values of the epidemiological parameters of the dynamic model input to calculate the incidence of HIV and FGS from 2000 to 2009 . We considered two scenarios for the mechanism through which annual praziquantel administration for may reduce HIV transmission . In the first scenario , we assumed that for women who received praziquantel during their childhood , treatment reduces their enhanced risk of HIV transmission per sex act . In this scenario , the model predicted that , for an efficacy of 30% , annual administration of a single dose of praziquantel to all school-age children could avert 21 , 120 ( 95% CI: 11 , 000–55 , 395 ) cases of HIV at a program cost of US$259 . 31 per HIV case averted over a ten-year period ( Figure 3 ) . When adjusted for averted medical care costs , the net saving for school-age strategy compared to the status quo was estimated to be US$15 . 8 ( 95% CI: -US$13 . 0–50 . 7 ) million ( Figure 3 ) . For an efficacy of 70% , annual praziquantel administration to school-age children could avert 106 , 200 ( 95% CI: 74 , 025–207 , 230 ) cases of HIV at a program cost of US$51 . 68 per HIV case averted over a ten-year period ( Figure 3 ) . When adjusted for averted medical care costs , the net saving for school-age strategy compared to the status quo was estimated to be US$101 . 4 ( 95% CI: US$39 . 0–233 . 6 ) million ( Figure 3 ) . In the second scenario , we assumed that women who received praziquantel during their childhood have a reduced FGS prevalence . In this scenario , the model predicted that , for an efficacy of 30% , annual praziquantel administration to school-age children could avert 41 , 500 ( 95% CI: 34 , 000–79 , 190 ) cases of HIV at a program cost of US$131 . 94 per HIV case averted over a ten-year period ( Figure 3 ) . When adjusted for averted medical care costs , the net saving for school-age strategy compared to the status quo was estimated to be US$36 . 4 ( 95% CI: US$3 . 4–77 . 5 ) million ( Figure 3 ) . For an efficacy of 70% , annual praziquantel administration to school-age children could avert 96 , 945 ( 95% CI: 79 , 370–185 , 260 ) cases of HIV at a program cost of US56 . 50 per HIV case averted aver a ten-year period ( Figure 3 ) . When adjusted for averted medical care costs , the net saving for the school-age strategy compared to the status quo was estimated to be US$92 . 3 ( 95% CI: US$38 . 0–200 . 0 ) million ( Figure 3 ) . For the first scenario of the mechanistic basis of praziquantel in which the elevated risk of HIV acquisition is mitigated , variation in the number of HIV cases averted was primarily driven by , the probability of acquiring FGS from adult infection , and , the annual number of sex acts in low risk partnerships ( Figure 4 ) . For the second scenario , when mass praziquantel reduces FGS prevalence , variation in the number of HIV cases averted was primarily driven by , the coefficient by which FGS enhances HIV transmission rate per sex act , and , the annual number of sex acts in low-risk partnerships ( Figure 4 ) , highlighting the importance of this parameter to our analysis .
Our analysis demonstrated that mass administration of praziquantel could possibly be a cost-saving intervention for preventing HIV infection in Zimbabwe . For a wide range of efficacies , annual treatment of school-age children between 5 and 14 years could result in saving between US$15 . 8 and US$101 . 4 million medical costs over a ten-year period for a program cost of US$51 . 68–259 . 31 per HIV case averted . We found that even the indirect benefits of reducing HIV transmission alone may be sufficient to make mass praziquantel administration a cost-saving intervention . As a cost-saving intervention , mass praziquantel administration should be prioritized over other less cost-effectiveness public health interventions . Given the additional health benefits of reducing schistosomiasis morbidity that were not considered , because our model does not explicitly capture the complex age-structure dynamics of S . haematobium [40] , [41] , our cost-effectiveness results are conservative . Moreover , our model does not take into account the WHO recommendation of reducing the frequency of mass administration of praziquantel as schistosomiasis prevalence falls below 10% [27] . Accounting for this potential reduction in the frequency of mass treatment would reduce costs and might further enhance the cost-effectiveness of mass praziquantel administration . The results of our cost-effectiveness model may be applicable to other regions of sub-Saharan Africa with socio-cultural and epidemiological settings similar to those of Zimbabwe . The donation of praziquantel by pharmaceutical companies and international donors would substantially contribute to reducing the cost of mass schistosomiasis treatment in sub-Saharan Africa , making it even more cost-effective for HIV prevention . Cost-savings arise at the interface of the low cost of praziquantel , its high efficacy , and elevated risk of HIV among FGS-infected women . Mass treatment of schistosomiasis also compares favorably to other biomedical interventions against HIV transmission . Treatment of sexually transmitted diseases is estimated to cost between $304 and $514 per HIV case averted [42] . Male circumcision is estimated to cost between $174 and $2808 per HIV case averted [43] . Antiretroviral therapy , while cost-effective in preventing HIV-associated morbidity and mortality , is less efficient in preventing new HIV infections ( cost per HIV case averted>$20 , 000 ) [42] . Our analyses indicate the potential for mass treatment of schistosomiasis as an innovative and cost-saving public health tool for preventing HIV infections in sub-Saharan Africa . Given that epidemiological and clinical data on FGS dynamics are scarce , our model does not account for the potential natural acquisition of partial immunity to FGS , with age , among adult women [44] . Furthermore , the cost-effectiveness of mass administration of praziquantel on reducing HIV transmission in sub-Saharan Africa should also be explored in the context of increasing antiretroviral therapy coverage and other HIV prevention measure such as male circumcision . We anticipate that an increase in antiretroviral therapy coverage would result in increasing the initial annual national spending on HIV treatment [23] , thus making mass treatment of schistosomiasis more cost-effective for HIV prevention . In the long term , the national spending on HIV treatment may decrease with HIV prevalence [23] , thereby reducing the cost-effectiveness of mass treatment of schistosomiasis . Allying schistosomiasis control with HIV/AIDS control programs might offer synergetic opportunities for administration to reduce costs of delivery and increase the coverage of implementation . Praziquantel prevents long-term schistosomiasis induced morbidity by killing egg-laying schistosomes in the host [27] . In schistosomiasis-endemic areas , individual treatment is of minimal benefit to the recipient as the risk of post-treatment reinfection remains very high [29] , [45] . To significantly impact schistosomiasis morbidity in endemic areas , mass praziquantel administration should be administered periodically to entire communities or targeted to school-age children , which is the age-group at highest risk of infection [27] . However , intervention strategies that are based exclusively on mass praziquantel administration are likely to be unsustainable and to establish an indefinite chain of dependence on a pharmacological intervention . To complement the effectiveness of schistosomiasis control program and ensure sustainability of control efforts , mass praziquantel administration should be coupled with improvement of sanitation facilities , clean water supply , and health and hygiene education [27] , [46] . Epidemiological studies have shown that women infected with genital schistosomiasis have a three- to four-fold increased odds of having HIV compared to women without genital schistosomiasis [5] , [7] . However , each of these studies is cross-sectional and as such they can only determine an association , rather than a cause-effect relationship , between genital schistosomiasis and HIV infection . Definitive proof of a cause-effect relationship can only be established through longitudinal studies . A prospective randomized controlled study to assess the effect of praziquantel treatment on HIV incidence has been proposed as a necessary step toward developing a new protocol to treat schistosomiasis for HIV prevention [47] . There is pressing need for future epidemiological studies and control trials will fill the gaps in our knowledge on the association between schistosomiasis and HIV . Our analyses indicate that future control trials should not only aim to provide information on the relative risk of HIV acquisition given schistosomiasis infection and the efficacy of praziquantel treatment to reduce the increased susceptibility to HIV infection , but should also document the rates of FGS acquisition among children and adults as well as confounding sexual behavior , which we found to be fundamental in determining the effectiveness and cost-effectiveness of mass praziquantel administration . Genital schistosomiasis may interact with other risk factors , such as other genital ulcerative diseases and sexual behavior , exacerbating the risk of HIV infection . Our results suggest that mass treatment of schistosomiasis in sub-Saharan Africa would not only have direct health benefits of reducing schistosomiasis infections , it may also avert cases of HIV and reduce the cost burden on the Zimbabwean medical system . By considering the impact of this neglected disease on HIV transmission , public health efforts can be expanded to include a broader set of cost-effective control strategies . | Evidence from epidemiological and clinical studies supports the hypothesis that genital infection with Schistosoma haematobium increases the risk of becoming infected with HIV among women in sub-Saharan Africa . Praziquantel is an oral , nontoxic , inexpensive medication recommended for treatment of schistosomiasis , which might be able to prevent the development of genital schistosomiasis . We constructed a mathematical model of female genital schistosomiasis and HIV infections , which we calibrated using epidemiological data from Zimbabwe . We used this model to investigate the potential cost-effectiveness of mass drug administration with praziquantel as an intervention strategy for reducing HIV transmission in sub-Saharan Africa . We showed that mass drug administration with praziquantel may be a timely , innovative , and cost-saving intervention strategy for HIV prevention in sub-Saharan Africa . As a cost-saving strategy , mass drug administration with praziquantel should be prioritized over other less cost-effective public health interventions . Our findings indicate the possible benefit of scaling up schistosomiasis control efforts in sub-Saharan Africa , and especially in areas were Schistosoma haematobium and HIV are highly prevalent . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"schistosomiasis",
"sexually",
"transmitted",
"diseases",
"disease",
"dynamics",
"aids",
"population",
"dynamics",
"neglected",
"tropical",
"diseases",
"population",
"biology",
"biology",
"infectious",
"disease",
"control",
"infectious"... | 2013 | Potential Cost-Effectiveness of Schistosomiasis Treatment for Reducing HIV Transmission in Africa – The Case of Zimbabwean Women |
The host protein viperin is an interferon stimulated gene ( ISG ) that is up-regulated during a number of viral infections . In this study we have shown that dengue virus type-2 ( DENV-2 ) infection significantly induced viperin , co-incident with production of viral RNA and via a mechanism requiring retinoic acid-inducible gene I ( RIG-I ) . Viperin did not inhibit DENV-2 entry but DENV-2 RNA and infectious virus release was inhibited in viperin expressing cells . Conversely , DENV-2 replicated to higher tires earlier in viperin shRNA expressing cells . The anti-DENV effect of viperin was mediated by residues within the C-terminal 17 amino acids of viperin and did not require the N-terminal residues , including the helix domain , leucine zipper and S-adenosylmethionine ( SAM ) motifs known to be involved in viperin intracellular membrane association . Viperin showed co-localisation with lipid droplet markers , and was co-localised and interacted with DENV-2 capsid ( CA ) , NS3 and viral RNA . The ability of viperin to interact with DENV-2 NS3 was associated with its anti-viral activity , while co-localisation of viperin with lipid droplets was not . Thus , DENV-2 infection induces viperin which has anti-viral properties residing in the C-terminal region of the protein that act to restrict early DENV-2 RNA production/accumulation , potentially via interaction of viperin with DENV-2 NS3 and replication complexes . These anti-DENV-2 actions of viperin show both contrasts and similarities with other described anti-viral mechanisms of viperin action and highlight the diverse nature of this unique anti-viral host protein .
The interferon ( IFN ) response is triggered in cells infected by RNA viruses , including members of the Flaviviridae family via a number of RNA recognition pathways that ultimately act to limit viral replication [1] , [2] . Production of type I interferons ( IFNs; IFN-α and IFN-β ) by virus infected cells results in up-regulation of anti-viral IFN-stimulated genes ( ISGs ) and cytokines [3] . Infection of cells with the flavivirus , dengue virus ( DENV ) is recognised by the toll-like receptor-3 ( TLR3 ) , retinoic acid inducible gene–I ( RIG-I ) and melanoma differentiation associated gene-5 ( MDA5 ) pathways to induce the IFN response [4] , [5] . Microarray studies have shown up-regulation of ISGs , including viperin during DENV infection in cell lines and patient peripheral blood mononuclear cells ( PBMC ) [6] , as well as during DENV infection of macaques in both macrophages and B-cells [7] . We and others have demonstrated that viperin is induced by infection with a number of diverse viruses as well as able to limit viral infection in most instances , including the first reported up-regulation of viperin in human cytomegalovirus ( HCMV ) infected cells [8] , [9] , [10] . Subsequently , viperin has been shown to have anti-viral actions in other viral infections such as hepatitis C virus ( HCV ) , influenza virus , human immunodeficiency virus ( HIV ) , sindbis virus ( SINV ) , the flaviviruses Japanese encephalitis virus ( JEV ) and West Nile virus ( WNV ) [9] , [10] , [11] , [12] , [13] , [14] , [15] and more recently , Bunyamwera virus [16] and Chikungunya virus [17] . The roles and actions of viperin in these different viral infections appear diverse and multifaceted with anti-viral activity in some cases dependent on alterations to lipid rafts ( influenza , [14] ) , membrane localisation ( HCV , [9] ) , the radical S-adenosylmethionine ( SAM ) enzymatic activity of viperin ( HIV , [10] ) , negated by viral proteins ( JEV , [11] ) and even an enhancing role under some conditions for HCMV [18] . Viperin also has anti-viral activity against DENV infection [6] , [13] however the interaction of DENV and viperin has not been thoroughly investigated . In this study we have further defined the induction of viperin and its mechanisms of anti-viral actions in DENV-2-infected cells , using an infectious DENV-2 in vitro replication model and including primary monocyte-derived macrophages ( MDM ) which represent a target cell type for DENV in vivo . Results show that DENV-2 infection induces viperin mRNA and protein , that expression of viperin is anti-viral , requiring the C-terminal but not N-terminal regions of viperin protein , and restricts DENV-2 infection by reducing viral RNA production . Viperin co-localised and interacted with DENV-2 CA , viral RNA and NS3 proteins . The interaction of viperin and NS3 but not membrane association , however , is necessary for viperins anti-viral actions . These results show both similarities and differences to our recent data suggesting that the anti-viral actions of viperin relate to its interaction with HCV NS5A and VAP-A in HCV replication complexes [9] and supports the growing evidence for both conserved and unique mechanisms of action of viperin against viral infections , even within the closely related Flaviviridae family of viruses .
Vero African green monkey kidney cells , A549 , a human lung carcinoma cell line , Huh-7 and Huh-7 . 5 human hepatoma cells and primary monocyte-derived macrophages ( MDM ) were used for DENV-2 infection studies and maintained as previously described . Primary MDM were generated by adherence from PBMC that were isolated from voluntary blood donation at the Australian Red Cross Blood Service . Blood was provided anonymously and used with approval from the Southern Adelaide Clinical Human Research Ethics Committee . Infections utilised DENV-2 , Mon601 , a derivative of the New Guinea C strain [19] that was produced from in vitro transcribed RNA , transfected into BHK-21 , baby hamster kidney cells , amplified in C6/36 insect cells and titred in Vero cells . Viperin shRNA and control cells were generated in Huh-7 cells , as previously described [9] . Cells were infected at a multiplicity of infection ( MOI ) of 0 . 1 or 1 for cell lines and an MOI of 3 for MDM for 90 min at 37°C , as described previously [20] , [21] , [22] . At the indicated time points post infection ( pi ) cell culture supernatants were collected , clarified by centrifugation and stored at −80°C prior to performing a plaque assay in Vero cells as previously described [21] . MDM were generated and DENV-2 infected , as above and at 48 h pi cells lysed and lysates subjected to SDS-PAGE . Proteins were transferred to nitrocellulose membranes and probed for viperin ( in house rabbit anti-viperin antibody , 1/1000 [12] ) with detection of complexes with goat-anti-rabbit-HRP conjugate and chemiluminesence . Protein loading was normalised by re-probing filters for β-actin ( anti-rabbit β-actin , 1/500 , BioVision ) . Images were captured with a LAS-4000 imaging system ( Fuji Corp ) and quantitated using Carestream Molecular Imaging Software 5 . 02 ( Carestream Health Inc ) . Wild type ( WT ) viperin and viperin mutant constructs were as described previously [9] . The DENV-2 NS3-GFP and pEPI–GFP CA constructs were a kind gift from Professor David Jans ( Monash University , Australia ) . Viperin mCherry fusion proteins were created utilising pLenti6-mCherry; WT and viperin mutants were cloned in frame ( XhoI/SacII ) into the construct using previously described primers [9] . Cell lines were transfected using FuGene6 ( Roche , IN ) as per manufacturer's instructions . The viperin coding region was cloned into the lentiviral vector pLenti6/V5-D-TOPO ( Invitrogen , CA ) and the control lentiviral plasmid pLenti6/V5-D-TOPO-tdTomato was obtained from Dr Yuka Harata-Lee ( University of Adelaide , Adelaide ) . Infectious lentivirus was generated as previously described [23] . Primary MDM were transduced with tdTomato control or viperin expressing lentivirus for 90 min at 37°C and were DENV-2 challenged at 24 h post transduction . Total cellular RNA was isolated from cells using Trizol ( Invitrogen ) , DNase treated and quantitated by spectrophotometry . For DENV-2 strand specific RT-PCR , 100 ng of denatured RNA was reverse transcribed at 37°C for 1 h with 20 pmol of DENV-2 specific primer ( DENV5 . 1 or DENV3 . 2 [21] ) attached to a 19mer long sequence ( Tag ) [24] and 10 U MMuLV RT ( Promega , WI ) . The tagged DENV-2 cDNA was then subjected to real time PCR using SYBER Green PCR mix ( Applied Biosystems , CA ) and 20 pmole of each primer , Tag , DENV3 . 2 for negative ( −ve strand ) and DENV5 . 1 for positive ( +ve strand ) , as previously described [21] . Real time PCR for viperin and the control gene RPLPO was performed as previously described on the ABI 7000 prism [12] . Primer sequences for IFIT1 were 5′ AACTTAATGCAGGAAGAACATGACAA and 5′ CTGCCAGTCTGCCCATGTG . Cells were cultured on gelatin coated glass coverslips , and fixed in either 1% ( v/v ) formaldehyde for MDM and HeLa , or acetone∶methanol ( 1∶1 ) for Huh-7 cells and stored at −20°C . Slides were washed in PBS , and the formaldehyde fixed cells permeabilised with 0 . 05% ( v/v ) IGEPAL before blocking in 4% ( v/v ) goat serum , 2% ( v/v ) human serum , 0 . 4% ( w/v ) bovine serum albumin ( BSA ) in Hanks buffered salts solution ( Gibco BRL , NY ) . Cells were immunolabelled using mouse anti-DENV-2 ( serotypes 1–4 , Santa Cruz Biotechnology Inc , 1/100 dilution ) , a rabbit anti-viperin , a mouse anti-FLAG ( Sigma , MO ) or a mouse anti-DENV CA antibodies ( a kind gift from Prof David Jans , Monash University , Australia ) . Immunoreactivity was detected with goat anti-mouse IgG-Alexa 488 , a goat anti-rabbit IgG-Alexa 647 or a goat anti-mouse IgG-Alexa 555 secondary antibodies ( Molecular probes , CA ) . Nuclei were labelled with Hoechst 33342 ( Molecular Probes , CA ) . BODIPY 493/503 ( Invitrogen ) was prepared as a stock solution of 1 mg/ml in ethanol . Fluorescence was visualised by confocal laser scanning microscopy ( Biorad Radiance 2100 or Leica SP5 Spectral Confocal Microscope ) . For some experiments , quantification of intensity of immunofluorescence labelling was performed using ImageJ software ( National Institutes of Health ) . The mean grayscale value was obtained for each channel for all cells where the image plane passed through the nucleus and excluding any cells at edge of the image and clusters of overlapping cells . Thresholds for detection of DENV-2 immunoreactivity were set at a grayscale value of three standard errors of the mean above the mean grayscale value measured in mock infected cells . 293T cells were transfected with pLenti6/V5-D-TOPO-viperin for 3 h then allowed to recover for 2 h prior to infection with DENV-2 at an MOI = 3 . At 24 h pi cells were lysed ( 10 mM Tris , pH 7 . 5 , 100 mM NaCl , 0 . 5% ( v/v ) Triton X-100+complete mini protease inhibitors [Roche] ) , lysates clarified and incubated for 1 hr with rabbit anti-FLAG antibody . Complexes were recovered with protein A-sepharose , washed six times ( 10 mM Tris , pH 7 . 5 , 100 mM NaCl , 0 . 05% [v/v] Triton X-100+protease inhibitors ) and resuspended in water . Precipitates were analysed for proteins by western blot and total DENV-2 RNA by RT-PCR with DENV5 . 1 and DENV3 . 2 primers , as described above with the exception that the reverse transcription step was non-primer directed . Acceptor photobleaching was carried out as previously described in [9] with the use of GFP and mCherry tagged protein constructs . Pre and post-bleaching images were aligned using ImageJ and the difference in fluorescence ( DIF ) analysed in 5–10 regions of each cell where lipid droplets and/or cytoplasmic stained structures were positive for both proteins . At least 10 different cells in each of at least two independent experiments were analysed to ensure reproducibility . Negative slides were prepared by imaging cells with only the donor molecule present and treated in parallel photobleaching experiments . Student t-tests were utilised to analyse the distributions of 2 normally distributed data sets and experiments were performed a minimum of three times , in triplicate or duplicate . Statistical analysis was performed using SPSS 10 .
As mentioned previously , a number of viruses are able to induce viperin expression , and to extend these observations we infected cell lines and primary MDM with DENV-2 . Cells were lysed at the indicated time points pi , RNA extracted and analysed by RT-PCR for viperin and DENV-2 negative strand ( −ve ) RNA , which is a marker of productive DENV-2 replication . Viperin mRNA was significantly induced in DENV-2 infected human cell lines with approximately 25 fold induction in A549 lung carcinoma cells ( Figure 1A ) and a lesser 4 fold increase in Huh-7 hepatoma cells co-incident with high level DENV-2 −ve strand RNA production ( Figure 1B ) . In contrast viperin mRNA was not increased following DENV-2 infection of Huh-7 . 5 cells , a cell line which is defective in dsRNA signalling via a mutation in the pathogen-recognition receptor RIG-I [25] ( Figure 1C ) . Up-regulation of viperin by DENV-2 infection was also demonstrated in primary MDM , with a much greater , approximately 1000 fold induction of viperin mRNA at 24 h pi ( Figure 1D ) . We next assessed up-regulation of viperin protein in MDM since these showed the most significant change in viperin mRNA . Cells were DENV-2 infected , lysed and analysed for viperin by western blot with IFN-α treated cells used as a positive control . Results show increased levels of viperin protein in DENV-2 infected primary MDM , at levels greater than that induced by IFN alone ( Figure 2A ) . We further characterised viperin protein in DENV-2-MDM by confocal microscopy . As can be seen in Figure 2B , at 24 h pi viperin was elevated in DENV-2 infected compared with mock infected MDM . Interestingly MDM positive for DENV antigen displayed reduced amounts of viperin protein , whereas DENV-antigen negative bystander cells ( indicated via arrows , Figure 2B , upper panel ) were shown to express significantly increased levels of viperin . The intensity of viperin staining in these different populations was quantitated . Results support the visual up-regulation of viperin in DENV-2 compared to mock infected cells , but most predominantly in the antigen negative , bystander cells of the DENV-2 infected population ( Figure 2C ) . Ectopic expression of viperin has been previously shown to inhibit DENV infection using virus and reporter virus-like particles in vitro [6] , [13] , however neither the anti-viral mechanism ( s ) or the interaction with viperin has been fully explored . Here we first transfected HeLa cells with plasmid to express WT viperin and at 24 h post transfection infected with DENV-2 . Cells were fixed 24 h pi and immunolabelled for dsRNA and viperin . Results are suggestive of anti-DENV activity of viperin , with individual cells expressing viperin harbouring either no or very little DENV-2 RNA ( Figure 3A ) . We next quantitated this potential anti-DENV-2 activity of viperin using a panel of viperin mutants that have previously been used to investigate viperins anti-HCV activity [9] . Although little is known about the structure/function relationship of viperins anti-viral activity , recent work by us and others has demonstrated that both the localisation of viperin to the ER membrane through its N-terminal amphipathic helix , as well as its C-terminal residues are essential for its ability to limit the replication of HCV [9] , [26] . Transient expression of WT viperin in Huh-7 cells significantly inhibited DENV-2 −ve strand RNA levels by up to 61% ( Figure 3B ) . A significant reduction of DENV-2 −ve strand RNA was also observed for cells transfected with viperin mutants in the SAM ( SAM 1-3 ) domain , leucine zipper ( LZ ) and N-terminal deletions from 17 and up to 50 amino acid residues , suggesting that these regions play no role in the anti-viral activity of viperin ( Figure 3B ) . In contrast , C-terminal deletions , as small as 17 amino acids , completely abolished the anti-DENV-2 activity of viperin ( Figure 3B ) . These C-terminal deletion mutants have previously been shown in our laboratory to retain the same expression level and localisation as WT viperin [9] . Mutation of the single C-terminal residue of viperin ( CTM ) partly abrogated the anti-DENV-2 activity of WT viperin , although compared with the no viperin control the CTM still produced a significant reduction in DENV-2 −ve strand RNA levels ( Figure 3B ) . These results highlight the importance of the C-terminal end of viperin for anti-viral activity and the C-terminal , 3′Δ17 mutant is used as a control in subsequent experiments . Huh-7 or A549 cells were transfected to transiently express WT or the C-terminal 3′Δ17 mutant viperin lacking anti-viral activity , infected with DENV-2 and were analysed at 24 h pi ( Figure 3C and D respectively ) . Results confirmed a significant reduction in both infectious virus release as determined by plaque assay of media from infected cells and production of −ve strand RNA induced by WT but not 3′Δ17 viperin expression in these two different cell types . An important cell type for DENV infection in vivo are cells of the monocyte-macrophage lineage . Additionally , these cells are major contributors to the IFN response . As such we have analysed the anti-viral actions of viperin in primary MDM . Given the difficulty in transfecting MDM , we expressed viperin via lentivirus-mediated transduction . MDM were transduced with a td-Tomato-red fluorescent protein control or viperin encoding lentivirus expression vector , infected with DENV-2 and infection analysed . Results show a significant reduction in infectious virus release from lentivirus-viperin transduced MDM compared with lentivirus td-Tomato transduced control MDM , with a significant 30 and 4 fold decrease seen at 24 and 48 h pi respectively ( the 8 h time point is considered a measure of input virus and is not significantly different between control and viperin transduced cells ) ( Figure 4A ) . At 48 h pi cells were fixed and immunostained for viperin and DENV-2 antigens , followed by confocal microscopy . Enumeration of >600 cells from 10 different fields and two different infections showed a significant reduction in the number of DENV-2 antigen +ve cells in viperin compared with tdTomato transduced cells ( images not shown , 5 . 9%±0 . 8 vs 9 . 5%±0 . 6 , p<0 . 05 , Students unpaired t-test ) . Additionally , we observed dramatically higher levels of viperin protein in the DENV-2 infected cells compared with mock-infected viperin-lentivirus transduced cells ( Figure 4B ) . Further , this up-regulation of viperin was again only observed in the DENV-2 antigen −ve bystander cells of this population ( Figure 4B ) . This likely represents up-regulation of endogenous viperin , as demonstrated previously in DENV-2 infected MDM ( Figure 2B ) . We next assessed the requirement for induction of viperin to restrict DENV-2 replication using a well characterised viperin shRNA Huh-7 cell line . Cells were DENV-2-infected at a lower MOI ( 0 . 1 ) to avoid DENV-induction of viperin mRNA , as in Figure 1 , potentially overwhelming the capacity of the viperin shRNA . DENV-2 infection of viperin shRNA cells resulted in a significant , approximately 2 fold enhancement of infectious DENV-2 release at 24 h pi compared with control shRNA expressing cells ( Figure 5A ) . By 48 h pi , infectious virus release was comparable between viperin shRNA and control shRNA expressing cells , possibly due to enhanced cytopathic effects in viperin shRNA cells associated with the earlier and higher level of DENV-2 replication , although this was not specifically quantitated . DENV-2-infection did not induce viperin mRNA in viperin shRNA expressing cells at 24 h pi ( Figure 5B ) , although expression was detected at 48 h pi in some instances . In all cases , the unrelated ISG , IFIT1 mRNA , utilised as a control , was induced to comparable levels in both DENV-2-infected viperin shRNA and control shRNA expressing cells , demonstrating effective induction of other anti-viral responses in the absence of viperin ( Figure 5C ) . The lower levels of DENV-2 −ve strand RNA and viral release observed in HeLa , Huh-7 cells and primary MDM following infection of viperin expressing cells could be consistent with restriction of DENV-2 entry . To investigate this possibility Huh-7 and A549 cells were transfected to express viperin and following DENV-2 infection cells were immediately lysed and +ve strand DENV-2 RNA , indicative of intracellular genomic input RNA , quantitated by RT-PCR . Results showed no difference in intracellular levels of +ve strand DENV-2 RNA between cells transfected to express viperin and the inactive viperin C-terminal mutant , 3′Δ17 ( Figure 6A ) . Additionally , DENV-2-infections were performed as above and cells lysed at 6 h pi and −ve strand DENV-2 RNA quantitated . Results demonstrated a significant reduction in the intracellular level of DENV-2 −ve strand RNA at this early time point in both A549 and Huh-7 cells transfected with viperin ( Figure 6B ) , demonstrating a post-entry restriction in early DENV-2 RNA replication . Our prior studies with HCV and viperin have demonstrated a requirement for the anti-viral actions of viperin mediated through lipid droplet and replication complex localisation and association with NS5A [9] . We next assessed the ability of viperin to associate with DENV-2 replication complexes by immunoprecipitation ( IP ) . 293-T cells were transfected to express FLAG-tagged viperin , infected with DENV-2 then at 24 h pi cells lysed and IPed with anti-FLAG antibody . Precipitates were analysed for the presence of total DENV-2 RNA by RT-PCR . Results demonstrate successful co-precipitation of DENV-2 RNA with FLAG-antibody ( Figure 6C ) . Concurrent analysis of precipitates by western blot , however failed to detect co-precipitated DENV-2 NS3 protein ( data not shown ) . The above data suggests the association of viperin with complexes containing DENV RNA ( ie . replication complexes ) . Such complexes reportedly are also associated with cellular membrane structures and DENV-2 NS3 protein [27] , [28] . The maintenance , however of the anti-DENV-2 activity of viperin containing N-terminal deletion mutants suggests that the ability of viperin to associate with membranes is not required for its restriction of DENV-2 infection ( Figure 3A ) . We thus assessed the cellular localisation of viperin in DENV-2 infected cells . Viperin primarily localises to lipid droplets in Huh-7 cells as we have shown previously [9] , and as can be seen in Figure 7A; this distribution remains unaltered in DENV-2 infected Huh-7 cells . The DENV capsid ( CA ) protein has also been demonstrated to localise to lipid droplets [29] , [30] and consistent with these previous reports we observed partial co-localisation between DENV-2 CA and viperin at the interface of lipid droplet-like structures ( white arrows , Figure 7B ) . Interestingly , viperin and the CA protein appear to coat the surface of the droplet at distinct loci , with small overlapping areas of co-localisation . Huh-7 cells were also co-transfected with a DENV-2 NS3-GFP expression plasmid and viperin . A clear co-localisation between DENV-2 NS3 and viperin is observed at the surface of lipid droplet-like structures as well as in distinct cytoplasmic loci ( Figure 7C ) . The N-terminal viperin deletion mutant ( Vip5′Δ33 ) which loses its ability to localise to lipid droplets [9] , [31] , but remains anti-viral against DENV-2 ( Figure 3A ) , remained co-localised with DENV-2 NS3 , although the pattern of localisation was solely cytoplasmic ( Figure 7C ) . This observation demonstrates that viperin's anti-viral activities may be exerted through a possible interaction with NS3 but not necessarily at the lipid droplet interface . As described above , although viperin co-precipitated DENV-2 RNA , IP experiments could not demonstrate co-precipitation of NS3 with viperin from either DENV-2 infected or NS3/viperin co-transfected cells ( data not shown ) . These co-precipitation experiments are likely confounded by our observation of low levels of viperin protein in DENV-infected cells ( Figure 2B ) and low levels of DENV-2 antigens in viperin transfected cells ( Figure 3A ) . Further , viperin is a lipid associated protein , which are notoriously difficult to extract and retain physiological protein-protein interactions . We therefore investigated the physical interaction of viperin with DENV-2 NS3 and CA by fluorescence energy resonance transfer ( FRET ) . Huh-7 cells were transfected with expression plasmids for DENV-2 CA-GFP or DENV-2 NS3-GFP in conjunction with either mCherry N-terminally tagged viperin-WT , viperin 5′Δ33 or viperin 3′Δ17 and FRET acceptor photobleaching performed . Results show positive FRET for viperin-WT and the DENV-2 CA protein at the surface of the lipid droplet ( Figure 8A ) demonstrating an interaction of WT-viperin and CA proteins at this site . FRET analysis also demonstrated an interaction of DENV-2 NS3 and WT-viperin in distinct cytoplasmic foci ( Figure 8B ) , similar to that seen in the confocal co-localisation studies of these two proteins ( Figure 7C ) . No positive FRET was detected between DENV-2 NS3 and viperin surrounding lipid droplet like structures , despite our prior observation of co-localisation at these sites ( Figure 7C ) . The ampipathic helix mutant ( 5′Δ33 ) of viperin , which retains its anti-DENV-2 activity , but has lost its membrane localisation ability , demonstrated a positive interaction by FRET analysis with DENV-2 NS3 , once again at distinct cytoplasmic foci within the cells ( Figure 8C ) . In contrast , the C-terminal viperin mutant , 3′Δ17 , which has no anti-DENV-2 activity but maintains WT viperin localisation [9] , showed no positive FRET with DENV-2 NS3 suggesting this protein is unable to interact with DENV-2 NS3 . These results indicate that the C terminus of viperin mediates its anti-DENV-2 activity through an interaction with DENV-2 NS3 but does not require lipid droplet or membrane association .
Viperin is emerging as an important virus-induced ISG that can be up-regulated by both IFN-dependent and independent pathways and has a diverse array of anti-viral actions . Viperin can be induced in an IFN independent manner via IFN regulatory factor-1 ( IRF-1 ) , following infection with the RNA virus , vesicular stomatis virus ( VSV ) [32] . In contrast , SINV induction of viperin requires IFN but JEV induction of viperin occurs in an IFN-independent manner that requires IRF-3 and AP-1 [11] . In this study we show that viperin is induced early in DENV-2 infection and similar to our observation in HCV infected cells , does not occur in Huh-7 . 5 cells that are deficient in RIG-I [9] . IRF-3 and AP-1 are transcription factors downstream of RIG-I activation suggesting that the RIG-I pathway has an important role in induction of viperin , at least for the Flaviviridae members JEV , HCV and DENV . Furthermore , our studies in DENV-2-infected viperin shRNA cells suggest that the viperin already present or induced intracellularly in the DENV-2 infected cell acts to restrict or control DENV-2 infection in this initial target cell . Additionally our results from DENV-2-infected MDM show strong induction of viperin protein in DENV antigen negative bystander cells . This indicates that induction of viperin in these bystander cells , probably secondary to the release of IFN from the DENV-2 infected cell , is likely to also be important for restricting DENV-2 spread . Our observation of a far greater level of induction of viperin following DENV-2 infection compared with IFN stimulation of primary MDM suggests that induction of viperin by DENV-2 , either in the initial DENV-2 infected cell or the uninfected bystander cell , occurs via factors other than IFN that are yet to be defined . Viperin protein contains N-terminal ampipathic helical domains that direct viperin cellular localisation to the endoplasmic reticulum ( ER ) and lipid droplets [9] , [31] , [33] . The C-terminal portion of viperin is relatively unstructured and highly conserved amongst species; however its current function remains unknown . Previously we have demonstrated that the anti-viral actions of viperin are dependent on a number of functional domains of the viperin protein ( i ) the N-terminus for intracellular ER and lipid droplet localisation of viperin , ( ii ) the extreme C-terminus in the context of HCV replication [9] , and ( iii ) the radical SAM domain in the context of HIV egress [10] . Using DENV-1 virus-like particles ( VLP ) and a luciferase reporter replicon system , a previous study has shown that viperin is induced by DENV-1 infection , inhibits DENV-1 RNA production and requires the N-terminal SAM1 domain of viperin [13] . This same study showed a similar induction of viperin , inhibition of RNA production and requirement for the viperin SAM1 domain and in part , residues within the first 50 amino acids of viperin during WNV infection . In contrast , our study observed anti-DENV-2 activity of viperin SAM1-4 mutants at levels comparable to WT viperin . The differences in the requirement for the viperin SAM1 domain seen in our current study compared with previous results with DENV-1 and WNV may be due to ( i ) the level of expression of viperin through use of a tet-induction system compared with the transient viperin transfection system in the current study; ( ii ) the analysis of different markers of infection with viperin SAM1 reducing infectious DENV-1 release [13] but in our study not DENV-2 −ve strand RNA; and/or ( iii ) the use of DENV-1 compared to DENV-2 herein . Regardless , studies clearly suggest that the anti-viral actions of viperin can be mediated by residues outside of the SAM1 domain , including the N-terminal 50 amino acid residues for WNV [13] and in our previous work with HCV , also the C-terminal regions of viperin [9] . Consistent with this requirement of the C-terminal region of viperin for anti-HCV activity , in our current study we have similarly defined anti-DENV activity to reside in the C-terminal 17 amino acids of viperin . The specific regions of viperin necessary for anti-viral activity between various viruses differs , as does the biological effect of viperin during different virus infections . Viperin is reported to inhibit release of influenza virus by disruption of lipid rafts [14] , to inhibit HIV egress [10] and to diminish viral protein production in HCMV infection [8] . In this study we show that viperin inhibits early post-entry DENV-2 RNA replication consistent with prior reported effects of viperin in inhibiting RNA replication in other Flaviviridae family members , HCV [9] and WNV [13] . In contrast , viperin is induced but is not anti-viral against the related flavivirus , JEV due to mechanisms of JEV that proteolyse and degrade viperin in infected cells [11] . However , contrary to earlier reports of anti-viral activity of viperin against HCMV , a recent study has shown that the vMIA protein of HCMV induces re-localisation of viperin from the ER to mitochondria , resulting in an increase in HCMV infection [18] . These contrasting effects of viperin suggest that its effects on viral infection are multifaceted , virus specific and involve multiple mechanisms of action including alterations in the subcellular localisation of viperin . Viperin has been shown to localise to lipid droplets and the ER . Similarly we have shown here that viperin co-localises with the lipid droplet marker , BODIPY , in DENV-2 infected cells and thus the cellular localisation of viperin is unchanged during DENV-2 infection . The DENV CA localises to lipid droplets and preventing this CA-lipid droplet association reduces DENV RNA replication and infectious virus particle production [30] . While we confirm that viperin is able to co-localise and interact with the DENV-2 CA at lipid droplet-like structures ( Figure 7B , 8B ) , our observation that viperin N-terminal mutants , which lose the ability to localise to lipid droplets , still retain substantial anti-DENV-2 activity suggests that viperin has significant anti-DENV-2 activities independent of its lipid droplet/CA association . We also demonstrate that viperin co-localises and interacts with the DENV-2-NS3 protein and co-precipitates with DENV-2 RNA , both of which are components of DENV replication complexes [27] , [28] . The interaction of viperin and DENV-2 NS3 was independent of the N-terminal ampipathic helix , but reliant on the C-terminus of viperin . Further , the ability of viperin to co-localise and interact with DENV-2 NS3 correlated with anti-viral activity . We propose that viperin has anti-viral activity mediated by a C-terminus interaction with DENV-2 NS3 that reduces early RNA production by interfering with DENV-2 replication complexes . It remains to be determined whether this occurs solely through a direct interaction with the DENV-2 NS3 protein , or also through an intermediate pro-viral host cell factor , such as is the case for HCV , whereby viperin interacts with both NS5A and the pro-viral factor VAP-A [9] , [26] . Currently , the only known pro-viral host factor for DENV that interacts with NS3 in the context of functional replication complexes is fatty acid synthetase ( FASN ) [34] . We feel FASN is an unlikely candidate as a target of viperin's actions since viperin and FASN exist in alternate cellular compartments ( Figure S1 ) . In conclusion , this study has revealed further critical functions of viperin during DENV-2 replication and highlighted similarities and differences in the mechanisms of induction and in the anti-viral actions of viperin between DENV-2 and other medically important Flaviviridae . This data highlights the incredibly diverse anti-viral nature of viperin and the complexity of the viperin-virus interaction . | Viperin is a virally induced host protein that has been previously shown to have antiviral activity against a variety of viruses . Here we have demonstrated that viperin is also anti-viral against the medically significant arbovirus , dengue virus . Viperin was able to inhibit dengue virus at the level of viral replication , and cell lines unable to produce normal levels of viperin grew the virus to higher titres . These anti-dengue effects of viperin were mediated by amino acid residues in its C-terminus , and did not require structural domains of the N-terminal region as has been previously shown by us and others for the related virus , hepatitis C virus . Viperin was also demonstrated to co-localise and interact with the dengue capsid protein on the surface of lipid droplets , as well as with the NS3 protein and viral RNA . Viperin's association with NS3 was further demonstrated to be involved in its anti-dengue activities . The anti-viral activities of viperin presented in this manuscript show both similarities and contrasts with other described anti-viral mechanisms for the protein and highlight the diverse nature of this unique anti-viral host protein . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"immunity",
"dengue",
"fever",
"innate",
"immunity",
"neglected",
"tropical",
"diseases",
"dengue",
"biology",
"microbiology",
"viral",
"diseases"
] | 2013 | Viperin Is Induced following Dengue Virus Type-2 (DENV-2) Infection and Has Anti-viral Actions Requiring the C-terminal End of Viperin |
Zinc is an essential trace element involved in a wide range of biological processes and human diseases . Zinc excess is deleterious , and animals require mechanisms to protect against zinc toxicity . To identify genes that modulate zinc tolerance , we performed a forward genetic screen for Caenorhabditis elegans mutants that were resistant to zinc toxicity . Here we demonstrate that mutations of the C . elegans histidine ammonia lyase ( haly-1 ) gene promote zinc tolerance . C . elegans haly-1 encodes a protein that is homologous to vertebrate HAL , an enzyme that converts histidine to urocanic acid . haly-1 mutant animals displayed elevated levels of histidine , indicating that C . elegans HALY-1 protein is an enzyme involved in histidine catabolism . These results suggest the model that elevated histidine chelates zinc and thereby reduces zinc toxicity . Supporting this hypothesis , we demonstrated that dietary histidine promotes zinc tolerance . Nickel is another metal that binds histidine with high affinity . We demonstrated that haly-1 mutant animals are resistant to nickel toxicity and dietary histidine promotes nickel tolerance in wild-type animals . These studies identify a novel role for haly-1 and histidine in zinc metabolism and may be relevant for other animals .
Zinc is a trace nutrient that plays critical roles in all biological systems . Zinc contributes to protein structure and enzymatic activity and functions in signal transduction processes [1] , [2] , [3] . The important role of zinc in biological systems is demonstrated by its impact on human health , since both zinc deficiency and excess can be deleterious . Zinc deficiency in humans causes a wide spectrum of symptoms that result from functional defects in the epidermal , gastrointestinal , central nervous , immune , skeletal , and reproductive systems , and inadequate dietary intake of zinc is a major worldwide problem [4] , [5] . Zinc deficiency is also associated with mutations in genes encoding human zinc transporters such as Zip4 , the causative gene in acrodermatitis enteropathica [6] . Excess zinc is also deleterious . The mechanisms underlying toxicity caused by excess zinc are not well defined . However , excess zinc may displace other trace metals or bind low-affinity sites , leading to protein dysfunction [7] . In humans , zinc toxicity associated with excess dietary intake has been reported , but occurs rarely [8] , [9] . By contrast , pathological conditions that lead to focal disruptions of zinc metabolism may be more common . For example , zinc release from dying cells during ischemic brain injury is postulated to be a major contributor to cell death and functional deficits [10] . Zinc metabolism appears to modulate the pathology of Alzheimer's disease , since precipitation and toxicity of the Aβ peptide that causes the disease are influenced by interactions with metal ions such as zinc [11] , [12] , [13] . Zinc-mediated toxicity is also associated with pancreatic islet cell destruction during diabetes [14] . Because zinc plays critical roles in human health and disease , it is important to understand the biological processes that mediate zinc metabolism and protect against zinc toxicity . Organisms have evolved several strategies to promote zinc homeostasis and protect against zinc toxicity . One strategy is to regulate zinc uptake and excretion such that zinc uptake is downregulated and zinc excretion is upregulated in the presence of high levels of dietary zinc . For example , vertebrate cells downregulate zinc importers in the ZIP family in response to high levels of zinc [15] . A second strategy is to sequester zinc in an intracellular organelle , as illustrated by the import of zinc into the vacuole of S . cerevisiae [16] . A third strategy is chelation of zinc by small molecules such as glutathione or proteins . Dietary zinc causes vertebrate cells to upregulate expression of metallothionein , a small protein that can bind multiple zinc atoms [17] . The nematode C . elegans is a powerful model system that has been used to characterize fundamental and highly conserved biological processes such as RNA interference ( RNAi ) and apoptosis [18] , [19] . It has also been used to develop innovative experimental techniques such as in vivo expression of green fluorescent protein [20] . C . elegans is a relevant model system for the study of metal biology , since it has been used to analyze zinc signaling , metal toxicity , and iron and heme metabolism [21] , [22] , [23] , [24] , [25] , [26] , [27] . We are using C . elegans to study zinc metabolism , since these animals have a simple and well-characterized anatomy , culture methods that permit precise control of dietary zinc are established , and powerful genetic approaches such as forward genetic screens are available [28] , [29] . To identify genes involved in C . elegans zinc metabolism , we conducted a forward genetic screen for chemically induced mutations that caused resistance to high levels of dietary zinc [28] . Nineteen mutations that confer significant resistance to dietary zinc were identified , and these strains represent the first report of mutant animals with increased resistance to zinc toxicity . Here we describe the use of whole genome sequencing to identify the gene affected by two of these mutations as histidine ammonia lyase ( haly-1 ) . C . elegans haly-1 encodes a protein that is conserved in vertebrates and predicted to cause the reductive deamination of histidine to urocanic acid [30] . We demonstrated that mutations in haly-1 cause elevated levels of histidine , leading to the hypothesis that histidine protects against zinc toxicity . Supporting that model , we demonstrated that dietary histidine protected against zinc toxicity in wild-type worms . Mutations in haly-1 and dietary histidine also caused resistance to nickel toxicity , suggesting that the mechanism of histidine protection is likely to be chelation of zinc and nickel . These results provide novel insights into the role of the haly-1 gene and histidine in modulating zinc toxicity .
To identify genes that mediate zinc metabolism , we conducted a forward genetic screen for C . elegans mutants that were resistant to toxic levels of dietary zinc [28] . After screening approximately 300 , 000 mutagenized haploid genomes , nineteen mutations were identified that caused significant resistance to zinc toxicity . These mutations were positioned in the genome by linkage to single nucleotide polymorphism ( SNP ) markers . Here we focus on two mutations , am130 and am132 , that caused strong resistance to dietary zinc ( Figure 1 ) . These mutations displayed tightest linkage to the same SNP , pkP6160 , positioned at +2 . 50 map units on chromosome X . Three factor mapping experiments using visible markers indicated that am132 is positioned between unc-115 and egl-15 , an interval that contains 863 kb ( Figure 2A ) [28] . To identify the lesions in these alleles , we refined the interval that contains the am132 mutation using high resolution mapping relative to SNP markers [31] . These experiments established that the am132 mutation is positioned between +2 . 40 and +2 . 84 map units , a 270 kb interval that contains 48 predicted open reading frames ( Figure 2A ) ( see Material and Methods ) . To identify the gene affected by the am132 mutation , we performed whole genome sequencing using DNA from the am132 mutant strain . Candidate mutations were identified by comparing the am132 DNA sequence to the wild-type reference sequence [32] . One candidate mutation caused a predicted nonsense change in the F47B10 . 2 gene . The presence of this mutation was confirmed using standard DNA sequencing . The mutation is a C to T transition in the wobble position of codon 296 that changes a tryptophan to a stop codon ( Figure 3B ) . F47B10 . 2 is predicted to encode a histidine ammonia lyase , an evolutionarily conserved enzyme that converts histidine to urocanic acid ( Figure 3C ) [30] . Therefore , we named the gene haly-1 . The mutant allele is predicted to encode a truncated protein lacking a significant portion of the conserved regions . Thus , haly-1 ( am132 ) is likely to be a strong loss-of-function mutation . Similar to am132 , the am130 mutation displayed linkage to SNP pkP6160 that is positioned close to haly-1 [28] . To determine if the am130 mutation affects the haly-1 gene , we analyzed the sequence of the haly-1 locus using DNA from am130 mutant animals . A single G to A transition was detected , and the change is a missense mutation predicted to change amino acid 536 from the negatively charged aspartic acid to the polar uncharged asparagine ( Figure 3B ) . The aspartic acid at position 536 is conserved in histidine ammonia lyase found in humans and other vertebrates . The identification of two , independently derived mutations that both affect the haly-1 gene , suggests that mutations in haly-1 cause resistance to dietary zinc toxicity . If the changes in the haly-1 locus detected in the am130 and am132 strains cause zinc resistance , then the introduction of wild-type copies of haly-1 might restore zinc sensitivity . To test this hypothesis , we generated transgenic haly-1 ( am132 ) animals containing fosmid WRM0624bE06 that contains the entire haly-1 locus and seven other predicted open reading frames ( Figure 2B ) . Two independently derived transgenic strains were analyzed for zinc sensitivity using noble agar minimal media ( NAMM ) with a range of supplementary zinc concentrations . Both transgenic strains displayed zinc sensitivity similar to wild-type animals ( Figure 2B ) . To determine if the haly-1 locus is sufficient for the rescue activity , we subcloned a 4318 bp fragment that includes 1567 bp upstream of the haly-1 START codon and 114 bp downstream of the haly-1 STOP codon . Transgenic haly-1 ( am132 ) animals containing the haly-1 locus displayed zinc sensitivity similar to wild-type animals ( Figure 2B ) . To determine if an intact haly-1 open reading frame is necessary for the rescue activity , we generated a haly-1 locus with a deletion mutation that removes exons 4-10 . haly-1 ( am132 ) animals containing or lacking this mutant haly-1 locus displayed similar zinc tolerance , indicating that the rescue activity requires an intact open reading frame that has the capacity to produce HALY-1 protein . To investigate the effect of the am132 mutation , we used site-directed mutagenesis to change the tryptophan located at codon 296 into a stop codon . haly-1 ( am132 ) animals containing or lacking this mutant haly-1 locus displayed similar zinc tolerance , indicating that the am132 mutation causes a loss of haly-1 activity ( Figure 2B ) . These results demonstrate that the haly-1 locus is sufficient to rescue the mutant phenotype and the rescue activity requires an intact haly-1 open reading frame , indicating haly-1 is the gene affected by the am132 mutation . To determine the products generated from the haly-1 locus , we analyzed haly-1 mRNA . The C . elegans EST project isolated multiple cDNAs corresponding to haly-1 , and we determined the complete DNA sequence of six cDNAs . Two cDNAs included the predicted start codon , and five cDNAs included the intact 3′ end including the polyA tail . Thus , these data document the complete predicted open reading frame . The analysis indicated that there was only a single spliced form composed of 11 exons ( Figure 3A ) . To analyze the regulation of haly-1 mRNA , we cultured wild-type animals in C . elegans minimal maintenance medium ( CeMM ) , a fully defined , axenic liquid medium [33] . CeMM is formulated from purified vitamins , growth factors , amino acids , nucleic acids , heme , β-sitosterol , sugar , salts , and trace metals . CeMM with no added zinc can be supplemented with a wide range of zinc to analyze both zinc restriction and zinc excess [29] . We cultured wild-type animals in CeMM containing a low concentration of 10 µM zinc or a high concentration of 500 µM zinc , isolated RNA from adult stage animals and measured the level of haly-1 mRNA using quantitative real time PCR ( qRT-PCR ) . The level of haly-1 mRNA varied less than 1 . 5 fold in animals cultured at 10 µM and 500 µM zinc compared to control genes , indicating that the level of haly-1 mRNA is not regulated by dietary zinc ( see Materials and Methods ) . Based on the haly-1 mRNA structure , the predicted HALY-1 protein contains 677 amino acids . A BLAST search was used to identify related proteins , and Figure 3B displays an alignment of C . elegans HALY-1 with human , zebra fish and bacterial proteins . The amino acid sequence of C . elegans HALY-1 is 54% identical to HAL from humans and 37% identical to HAL from Cupriavidus metallidurans CH34 , a bacteria that has been used to study metal tolerance [34] . Histidine ammonia lyase catalyzes the reductive deamination of histidine to create urocanic acid ( Figure 3C ) [30] . The high degree of sequence conservation strongly supports the model that C . elegans haly-1 is descended from a common ancestral gene that was conserved in humans . Three members of the cation diffusion facilitator family of zinc transporters have been characterized in C . elegans , cdf-1 , cdf-2 and sur-7 [22] , [27] , [29] . Loss-of-function mutations in these genes cause sensitivity to dietary zinc . To investigate the interactions between haly-1 and these cdf genes , we constructed and analyzed double mutants . Because loss-of-function mutations in haly-1 and cdf genes cause the opposite phenotype , the analysis of double mutant animals can elucidate relationships between these genes . Figure 4A shows that compared to wild-type animals , haly-1 ( am132 ) mutant animals displayed resistance to dietary zinc , whereas cdf-1 loss-of-function mutant animals displayed sensitivity . cdf-1 ( lf ) haly-1 ( lf ) double mutant animals displayed an intermediate phenotype . In particular , in the presence of 0 . 02 – 0 . 06 mM supplemental zinc the double mutant animals were significantly more resistant than cdf-1 ( lf ) single mutant animals ( Figure 4B ) . At 0 . 08 mM and higher concentrations of supplemental zinc , the double mutant animals displayed sensitivity similar to the cdf-1 ( lf ) single mutant animals . We performed a similar analysis of haly-1 ( lf ) sur-7 ( lf ) and haly-1 ( lf ) cdf-2 ( lf ) double mutant animals . In both cases , the double mutant animals displayed zinc sensitivity that was intermediate compared to the single mutant animals ( Figure 4C , 4D ) . These results indicate that the haly-1 ( lf ) mutations promote zinc resistance in genetic backgrounds characterized by zinc sensitivity as well as in wild-type animals . Furthermore , these findings suggest that haly-1 functions in parallel to cdf genes to modulate zinc sensitivity . The zinc resistance of haly-1 mutant animals can be explained by two general models . One possibility is that haly-1 mutant animals have lower levels of zinc , perhaps as a result of reduced uptake or increased excretion . A second possibility is that haly-1 mutant animals have the same or higher levels of zinc compared to wild-type animals , but the mutant animals have improved tolerance . To distinguish between these possibilities , we used the method of inductively coupled plasma mass spectrometry ( ICP-MS ) to measure total zinc content [29] . A mixed-stage population was cultured in CeMM , harvested and analyzed for zinc content . The total zinc content of haly-1 ( am130 ) mutant animals was not significantly different from wild-type animals when cultured with optimal or high levels of dietary zinc ( Figure 5 ) . ICP-MS was also used to analyze the levels of magnesium , manganese , iron and copper . haly-1 ( am130 ) and haly-1 ( am132 ) mutants had levels of Mg , Mn , Fe , Cu that were similar to wild-type animals ( Table S1 ) . These results suggest that mutations in haly-1 cause zinc resistance by promoting tolerance to zinc rather than reducing the levels of zinc or other metals . HAL is a key enzyme in histidine metabolism , and mutations that diminish HAL activity cause elevated histidine levels in vertebrates [35] , [36] . To test the hypothesis that haly-1 ( lf ) mutations cause elevated histidine , we developed methods to measure total histidine levels in C . elegans . We cultured animals in CeMM containing 2 mM histidine , harvested a population consisting of mixed developmental stages , and measured the levels of amino acids ( Figure 6D ) . Wild-type animals contained 0 . 011 ± 0 . 002 nmoles histidine/µg protein . haly-1 ( am130 ) and haly-1 ( am132 ) animals displayed significantly higher levels , 0 . 039 ± 0 . 006 and 0 . 066 ± 0 . 02 nmoles histidine/µg protein , respectively ( Figure 6D ) . The haly-1 mutant animals displayed elevated levels of histidine compared to wild type when cultured at 0 mM , 0 . 075 mM or 1 . 5 mM zinc , indicating that the level of dietary zinc has little effect on the level of histidine ( data not shown ) . By contrast , the levels of the other amino acids were not consistently different between haly-1 mutant animals and wild-type animals ( data not shown ) . The levels of urocanic acid , the product of HAL enzymatic activity , have not been determined . These results indicate that haly-1 mutant animals have a specific defect in histidine metabolism that results in elevated levels of histidine and that the am130 and am132 mutations cause a reduction of haly-1 activity . The observation that haly-1 ( lf ) mutations cause elevated levels of L-histidine led us to hypothesize that the elevated histidine causes zinc resistance . To investigate this hypothesis , we analyzed the effects of feeding animals histidine . If elevated histidine levels promote zinc tolerance , then animals cultured with high levels of dietary histidine are predicted to display zinc tolerance . Wild-type hermaphrodites cultured on NAMM supplemented with 0 . 1 mM histidine displayed significantly increased tolerance to dietary zinc ( Figure 6A ) . The effects of dietary histidine were dose dependent: weak protection was observed at 0 . 03 mM , optimal protection was observed from 0 . 1 to 20 mM , and concentrations greater than 25 mM caused toxicity ( data not shown ) . Dietary histidine further increased the tolerance of haly-1 ( am130 ) and haly-1 ( am132 ) mutant animals to high levels of dietary zinc ( Figure 6B , 6C ) . To evaluate the specificity of the protection provided by feeding histidine , we analyzed the remaining amino acids by culturing wild-type hermaphrodites with 0 . 1 mM amino acid and 0 . 3 mM zinc . Histidine provided the most dramatic protection; 42% of animals grew to adulthood over 7 days when cultured with histidine compared to 8% when cultured with no amino acid ( Table 1 ) . Cysteine ( 21% ) also provided significant , but lower levels of protection , whereas the other 18 amino acids did not provide significant levels of protection ( Table 1 ) . These results demonstrate that only a small number of amino acids provide protection against zinc toxicity and histidine is the most effective . In the feeding experiments described above , histidine and zinc were both placed in the culture media . Therefore , these molecules have the opportunity to interact outside the animal and/or inside the animal after ingestion . To determine if histidine acts externally to the worms to provide zinc protection , we used a modified feeding procedure where worms were first exposed to histidine in the absence of zinc , and then exposed to zinc in the absence of histidine . If histidine acts externally to promote zinc tolerance , then animals subjected to this procedure are predicted to be zinc sensitive . By contrast , if histidine acts internally , then animals that are pre-treated with histidine are predicted to be resistant to a subsequent zinc challenge . Figure 6E shows that pre-treatment with histidine provided significant protection to wild-type animals that were challenged with 0 . 4 mM zinc . These findings indicate that ingested histidine is sufficient to protect against zinc toxicity . haly-1 mutant animals and wild-type animals fed a high histidine diet both display enhanced zinc tolerance , indicating that elevated histidine causes zinc tolerance . One possible mechanism is that histidine directly binds zinc and reduces its toxicity to the animal . This possibility is consistent with the fact that free histidine and histidine in proteins display high affinity interactions with zinc [37] , [38] . A second possible mechanism is that elevated levels of histidine trigger a biological response that promotes zinc tolerance; for example , a transcriptional response . To investigate these possibilities , we analyzed the effects of D-histidine . L- and D-histidine have identical chemical properties , such as pKa and binding affinity for zinc . However , L-histidine is utilized by biological systems for protein synthesis and other enzymatic reactions , whereas the enantiomer D-histidine is not recognized by enzymes or incorporated into proteins . If L-histidine protects against zinc toxicity by directly binding to zinc , then D-histidine is predicted to provide similar protection . By contrast , if L-histidine protects against zinc toxicity by initiating a biological response , then D-histidine is predicted to be inactive in promoting zinc tolerance . We compared wild-type animals cultured on NAMM plates with L- and D- histidine; both enantiomers caused similar levels of tolerance to zinc toxicity ( Table 1 ) . These results indicate that L-histidine promotes zinc tolerance by directly binding zinc . Our results indicate that haly-1 ( lf ) mutant animals have elevated levels of histidine that protect against zinc toxicity . If histidine binding to zinc is the mechanism of protection , then haly-1 mutant animals might be resistant to additional metals that can bind to histidine . Nickel binds to histidine [39] , as demonstrated by the use of nickel affinity chromatography to purify proteins containing a multi-histidine epitope tag [40] . To determine whether haly-1 mutant animals are resistant to nickel , we established the dose response of wild-type animals cultured in NAMM to supplemental nickel . Nickel caused dose-dependent toxicity , and no wild-type animals matured to adulthood in seven days at concentrations of 0 . 04 mM nickel or higher ( Figure 7A ) . By contrast , haly-1 ( am132 ) and haly-1 ( am130 ) mutant animals displayed striking resistance to nickel toxicity compared to wild-type animals ( Figure 7A ) . To investigate the specificity of haly-1 resistance to transition metals , we cultured wild-type and haly-1 ( am132 ) mutant animals on NAMM supplemented with iron , copper , cobalt , selenium , manganese , or cadmium . haly-1 mutant animals and wild-type animals displayed similar dose responses to selenium and cadmium ( Figure S1 ) . haly-1 mutant animals were slightly resistant to copper , slightly sensitive to cobalt and iron , and substantially sensitive to manganese compared to wild-type animals ( Figure S1 ) . These results demonstrate that haly-1 mutant animals are specifically resistant to a subset of transition metals including zinc and nickel , supporting the model that elevated histidine binds these metals to promote tolerance . To test the hypothesis that elevated levels of histidine cause the nickel resistance displayed by haly-1 mutant animals , we analyzed the effect of dietary histidine on nickel toxicity . Wild-type animals cultured with 0 . 1 mM histidine displayed striking resistance to nickel toxicity ( Figure 7B ) . These findings indicate that elevated levels of histidine promote tolerance to dietary nickel .
Histidine levels in animals are regulated by dietary intake and excretion and the activity of catabolic enzymes . Histidine is an essential amino acid in animals that is obtained from the diet [41] . For example , the fully-defined CeMM used to culture C . elegans includes L-histidine [33] . Although animals cannot synthesize histidine , they have catabolic enzymes . Histidine ammonia lyase was identified as an enzyme that converts L-histidine to urocanic acid , and HAL is the first enzyme in the catabolism of L-histidine . Other enzymes that modify L-histidine include histidine decarboxylase , histidyl-tRNA synthetase , and 1-methyl transferarase [42] . These findings demonstrate the central role of HAL in histidine metabolism . We used genetic analysis to characterize the function of the C . elegans haly-1 gene , which has not been previously characterized . C . elegans HALY-1 protein displays a high level of identity with vertebrate and bacterial enzymes that have been demonstrated to display histidine ammonia lyase catalytic activity , suggesting that C . elegans HALY-1 has a similar catalytic activity , although this has not been tested biochemically . Furthermore , loss-of-function mutations in C . elegans haly-1 caused elevated levels of histidine . These results support the model that C . elegans HALY-1 converts L-histidine to urocanic acid . haly-1 mutant animals display greater resistance to zinc toxicity than wild-type animals . However , haly-1 mutant animals and wild-type animals displayed similar total zinc levels , indicating that the resistance to zinc toxicity is not caused by reduced levels of zinc . These results suggest that haly-1 mutant animals accumulate zinc in a form that has reduced toxicity . haly-1 mutant animals also displayed striking resistance to nickel toxicity , but they were not highly resistant to other metals . Thus , reducing the activity of haly-1 caused specific resistance to zinc and nickel toxicity . Mutations in HAL have been characterized in mice and humans . In both animals , mutations in HAL cause elevated levels of histidine , consistent with a critical role for the enzyme in histidine catabolism . In humans , mutations in HAL cause a syndrome of histidinemia [35] . Several different missense mutations in HAL have been identified in affected families [43] . Histidinemia is a prevalent genetic disorder in certain ethnic groups such as Japanese where it affects 1 in 8000 live births . Patients with this disorder display alterations in zinc biology , including elevated excretion of histidine and zinc in the urine and mild zinc deficiency in some children as determined by hair analysis [44] . These results suggest that in humans with elevated levels of histidine , histidine can bind to zinc , and the complex can be excreted in the urine . The syndrome may predispose patients to disorders of the central nervous system [35] , [36] . In a mouse model of histidinemia , the disease is autosomal recessive , and the histidine ammonia lyase gene located on chromosome 10 is predicted to encode a protein with a single amino acid change [45] , [46] , [47] , [48] . The histidinemic mouse lacks a visible mutant phenotype , however , offspring of mutant mothers have increased risk of nervous system defects such as circling and head tilting . A low-histidine diet given to the his/his mother prevents the nervous system effects in offspring [46] , [49] . Thus , in C . elegans and humans , mutations in HAL cause an elevation in histidine levels and affect zinc metabolism indicating that haly-1 mutant animals may be a relevant model for the human disease . The findings reported here regarding the role of haly-1 mutations in metal toxicity suggest that metal chelation due to elevated levels of histidine may contribute to the pathophysiology of human histidinemia . The analysis of haly-1 suggests a model for the role of histidine in zinc biology in animals . We propose that elevated levels of histidine in haly-1 mutant animals chelate zinc and protect against zinc toxicity . This model predicts that dietary administration of histidine to wild-type animals can phenocopy the haly-1 ( lf ) mutant and protect against zinc toxicity . Our results confirmed this prediction and demonstrated that dietary histidine acts inside the animals to promote zinc tolerance . Histidine has been demonstrated to bind zinc , suggesting that elevated histidine acts by direct chelation [50] . An alternative possibility is that elevated histidine triggers a biological response that promotes zinc tolerance . We used two approaches to test these possibilities . First , we demonstrated that dietary supplementation with D- and L-histidine protected against zinc toxicity . Since D-histidine has the same chemical properties as L-histidine , but lacks biological activity , these findings suggest that L- and D-histidine act by directly chelating zinc . Second , we analyzed the specificity of the protective effects and demonstrated that haly-1 mutant animals are strongly resistant to zinc and nickel , but not other metals . Like zinc , nickel binds histidine with high affinity [39] , and these results suggest that elevated histidine in haly-1 mutant animals directly chelates nickel to protect against nickel toxicity . Furthermore , dietary histidine protected against nickel toxicity , consistent with the chelation model . haly-1 mutant animals were sensitive to excess dietary manganese , indicating that abnormalities in histidine metabolism can be deleterious and result in susceptibility to some stresses . The response of haly-1 mutant animals to other metals such as copper might be a combination of protection mediated by histidine chelation of the metal and susceptibility caused by abnormal histidine metabolism . The effects of dietary histidine supplementation have been analyzed in vertebrates . In humans and rats , dietary supplementation with histidine increases urinary excretion of histidine and zinc , and in some cases is associated with symptoms of zinc deficiency [51] , [52] , [53] . These results are consistent with the model that elevated levels of histidine promote chelation of zinc and document the relevance of the studies of C . elegans to vertebrate biology . Supplementation with histidine has been shown to affect zinc uptake in a variety of physiological assays , including absorption by intestinal preparations from fish , crustaceans and mammals and zinc uptake by cells such as erythrocytes [54] , [55] , [56] , [57] , [58] , [59] . These studies indicate that histidine may increase zinc solubility and/or availability for transporters , or that zinc and histidine may be cotransported across membranes . Ralph et al . recently analyzed the ability of amino acids in the medium to protect cultured astrocytes from the toxicity of zinc and demonstrated that histidine was the most effective , and cysteine , glutamine and threonine showed smaller protective effects [60] . The analysis of C . elegans are consistent with these findings , since dietary histidine was the most effective , and cysteine showed a smaller but significant effect protecting worms from zinc toxicity . The results presented here contribute to this field by demonstrating that elevated histidine levels modulate zinc metabolism in an intact animal and can provide protection against zinc toxicity . These findings document a physiological role for histidine binding to zinc in vivo . C . elegans have been demonstrated to respond to dietary metals , and an interesting issue raised by these studies is the possibility that histidine levels are regulated as a protective mechanism in response to high dietary zinc . In response to dietary cadmium , worms display a range of transcriptional changes including induction of metallothionein genes [24] , [61] , [62] . Davis et al . showed that the zinc transporter cdf-2 was induced by high dietary zinc [29] . We found that wild-type animals cultured in fully defined medium with low , optimal , or high concentrations of zinc displayed similar levels of histidine . Furthermore , the level of haly-1 mRNA was not significantly affected by dietary zinc . These results indicate that haly-1 activity and levels of histidine may not be regulated in response to dietary zinc . Several important human diseases have been associated with tissue-specific zinc toxicity , such as ischemic brain injury [10] , Alzheimer's disease [11] , [12] , [13] , and some forms of diabetes [14] . Our findings suggest the possibility that modifying the activity of HAL could provide protection against zinc toxicity in these cases . For example , chemical inhibitors of HAL have been described [63] , [64] , [65] , and such chemicals might elevate histidine levels and reduce zinc toxicity . Further research is necessary to evaluate the feasibility and potential benefits of manipulating HAL activity .
C . elegans strains were cultured at 20°C on nematode growth medium ( NGM ) seeded with E . coli OP50 unless otherwise noted [66] . The wild-type C . elegans and parent of all mutant strains was Bristol N2 . The following mutations were used: haly-1 ( am130 ) and haly-1 ( am132 ) [28] , cdf-1 ( n2527 ) [22] , cdf-2 ( tm788 ) [29] , sur-7 ( ku119 ) [27] , dpy-6 ( e14 ) [67] and egl-15 ( n484 ) [67] . Double mutant animals were generated by standard methods , and genotypes were confirmed by PCR or DNA sequencing . To make NAMM , we prepared a solution with 1 . 7% Noble agar ( U . S . Biological , Swampscott , MA ) and a final concentration of 5 mg/liter cholesterol using a stock solution of 5 mg/ml cholesterol in 100% ethanol using water from a Milli-Q synthesis A10 machine ( Millipore , Billerica , MA ) . The solution was autoclaved for 30 minutes – autoclave times greater than 45 minutes impaired solidification . Metals such as zinc chloride , nickel chloride , sodium selenite , cadmium chloride , cobalt ( II ) sulfa hepa hydrate , copper chloride , ammonium iron ( II ) sulfate hexahydrate or manganese chloride tetrahydrate ( Sigma-Aldrich , St . Louis , MO ) were added to yield the desired final concentrations , and 7 ml of molten agar was immediately dispensed to 6 cm Petri dishes . NAMM was allowed to harden overnight at room temperature . E . coli OP50 was grown overnight in LB , concentrated ten-fold in Milli-Q water , and 100 µL was dispensed to each dish . To make NAMM supplemented with L- and D- amino acids ( Sigma-Aldrich , St . Louis , MO ) , amino acids were added to the molten agar and dispensed to Petri dishes . To analyze the response of worms to dietary metals and/or amino acids , we generated a population of hermaphrodites cultured on NGM plates , treated the animals with alkaline hypochlorite to obtain eggs , and cultured the eggs in M9 media overnight to obtain arrested first larval stage ( L1 ) animals . L1 animals were pipetted onto each NAMM dish , counted and cultured at 20°C . The number of worms that had matured to the adult stage as judged by body size and vulval development over a period of seven days was determined using a dissecting microscope . The percent adult was calculated by dividing the number of adult stage animals by the number of L1 animals originally dispensed . For experiments shown in Figure 6E , L1 animals were placed on NAMM with or without supplemental amino acids for 24 hours , washed , then transferred to NAMM plates supplemented with zinc . To eliminate amino acids in the intestinal lumen , we washed the worms three times in M9 , incubated the worms for thirty minutes in M9 with 1 mM seratonin to stimulate pharyngeal pumping and defecation , then washed the worms two additional times in M9 . dpy-6 ( e14 ) haly-1 ( am132 ) hermaphrodites were crossed to males of the wild isolate CB4856 that contains multiple polymorphisms compared to N2 , F1 cross progeny were selected , and 18 F2 self-progeny were selected as non-Dpy hermaphrodites that displayed zinc resistance . Similarly , haly-1 ( am132 ) egl-15 ( n484 ) hermaphrodites were crossed to CB4856 males , F1 cross progeny were selected , and one F2 self-progeny was selected as a non-Egl hermaphrodite that displayed zinc resistance . Hermaphrodites homozygous for the recombinant chromosomes were selected and used to prepare genomic DNA . Nine SNP markers positioned on chromosome X between dpy-6 at position 0 . 0 and egl-15 at position at +2 . 86 were analyzed . The most informative non-Dpy zinc resistant recombinant contained the CB4856 SNP marker CE6-177 at position +2 . 40 , indicating that am132 is positioned to the right of this marker . The non-Egl zinc resistant recombinant contained the CB4856 SNP marker CE6-1202 at position +2 . 84 , indicating that am132 is positioned to the left of this marker . DNA was isolated from mixed stage animals grown on NGM plates using the Purification of Total DNA from Animal Tissues Spin Column Protocol from the DNeasy Blood and Tissue Kit ( Qiagen ) with minor modifications . The gDNA was fragmented by sonication and used to generate Illumina random whole genome sequencing libraries consisting of two size fractions , 250-300 bp and 350–400 bp . The libraries were amplified in situ on Illumina flow cells according to the manufacturer's protocol , and sequence data consisting of 50 bp reads were obtained using the Solexa/Illumina platform [32] . The number of short DNA sequences that were determined corresponds to approximately 30-fold coverage of the C . elegans genome . Sequences that corresponded to the 270 kb mapping interval for the am132 mutation were aligned to the reference wild-type C . elegans sequence using the maq utility [http:maq . sourceforge . net/]; read-depth > = 3; mapping quality >40 , consensus quality > = 15 . 251 candidate base changes were identified in the 270 kb interval . The majority of these candidate base changes were identified with low confidence , indicating they were likely to be sequencing errors . All candidate base changes were analyzed by determining the effect on predicted open reading frames . Only one candidate base change was predicted to cause a nonsense mutation , and this candidate base change was identified with high confidence . The presence of this candidate base change in the am132 strain was confirmed by conventional DNA sequencing , and other candidate base changes were not further analyzed . We prepared DNA using standard methods , determined the sequence using Applied Biosystems 3730 and/or 3130xl DNA sequencers , and analyzed data using Sequencher ( Gene Codes Corporation , Ann Arbor , MI ) . To determine the DNA sequence of the haly-1 locus , we analyzed from 75 bp upstream of the predicted START codon to 100 bp downstream of the predicted STOP codon . Plasmid pJM1 is pBlueScript SK+ ( Stratagene ) containing a 4 , 318 bp fragment of C . elegans genomic DNA from fosmid WRM0624bE06 . The fragment extends from 1 , 567 bp upstream of the predicted haly-1 START codon to 114 bp downstream of the predicted STOP codon . To generate pJM2 , we modified pJM1 by digestion with NcoI ( New England Biolabs ) and religation , resulting in the deletion of 1542 bp that removes haly-1 exons four through ten . To generate pJM3 , we performed site-directed mutagenesis using PCR-mediated overlap extension [68] to generate a mutation in exon six that changes codon 296 from tryptophan to STOP . Transgenic animals were generated by co-injecting fosmids with the transformation marker pEL125 that has homology to the fosmid backbone and expresses GFP or co-injecting plasmids with the dominant transformation marker pRF4 [69] . haly-1 ( am132 ) hermaphrodites were injected , and we selected Rol or GFP self progeny and then selected strains that transmitted the marker . For each of these strains , the Rol or GFP phenotype was transmitted to only a sub-set of self-progeny , indicating that these transgenes were extrachromosomal . We defined rescue as a significant difference in zinc sensitivity between F1 progeny that displayed the marker phenotype compared to F1 progeny that did not display the marker phenotype and were presumed to lack the extrachromosomal array ( P<0 . 05 Student T-Test ) . We analyzed six haly-1 ESTs obtained from the National Institutes of Genetics , Japan ( yk1370d8 , yk1228g10 , yk1325f01 , yk1035h03 , yk1086a03 and yk1286h11 ) . We determined the complete DNA sequence of the inserts using standard techniques . A poly A tail began 104 bp from the stop codon in three cDNAs and 125 bp from the stop codon in two cDNAs . None of the cDNAs contained a spliced leader sequence , indicating that they did not include the 5′ end . haly-1 mRNA levels were analyzed as described by Davis et al . with minor modifications [29] . Briefly , wild-type worms were cultured in CeMM containing 10 µM or 500 µM zinc chloride for six days . The COPAS Biosort was used to collect 1000 adult animals for RNA preparation . Quantitative real-time PCR was performed using an Applied Biosystems 7900HT Fast Real-Time PCR System and the Applied Biosystems SYBR Green Master Mix . Forward and reverse amplification primers for haly-1 were ctattcacgctgtggccaag and caacgcttgcagcgacaatgatg , respectively . We obtained a large population of animals by culturing worms in CeMM with 0 . 075 mM zinc . The worms were placed in 75 cm2 flasks containing 7 . 5 mL CeMM at a concentration of 10 , 000 worms per mL . Then , 7 . 5 mL of CeMM containing zinc was added to make a final volume of 15 mL . The worms were cultured at 20°C for 18 days . Worms were washed three times in magnesium-free M9 solution , incubated in 1 mM serotonin in Mg-free M9 solution for 30 minutes , washed twice in Mg-free M9 solution , transferred to pre-weighed tubes ( Stockwell Scientific , part #3220N ) and immediately frozen at −80°C . Serotonin stimulates pharyngeal pumping and defecation , and the incubation with serotonin improves the accuracy of the measurement of zinc content by promoting exchange of zinc-containing culture medium in the intestinal lumen with Mg-free M9 solution [29] . The metal content was determined using ICP-MS [29] . Samples were freeze-dried , reweighed to obtain the dry pellet weight , and digested by heating in a hot block digester at 90°C for 1 . 5 h with concentrated nitric acid ( HNO3 ) and hydrogen peroxide ( H2O2 ) solution . The solution was diluted to a volume of approximately 10 mL with deionized water , and internal standards were added to correct for matrix effects . Instrument calibration standards were prepared from multi-element stock solutions ( High-Purity Standards , Charleston , SC ) to generate a linear calibration curve , and samples were analyzed using a VG Axiom high-resolution ICP-MS ( Thermo Fisher Scientific ) . Blank tubes were included in all processes as a control . The content of zinc , iron , copper , magnesium and manganese was determined as a value in parts-per-million ( ppm or µg/g ) by dividing measured metal content by the dry pellet weight . We obtained a large population of animals as described above by culturing worms in CeMM with 0 . 075 mM zinc and then transferring worms to CeMM containing 0 mM , 0 . 075 mM or 1 . 5 mM zinc . The worms were cultured at 20°C for 7 days , then prepared for analysis . A whole worm extract was made by sonicating worms in a 20 mM HEPES buffer pH 7 . 55 using a Branson Digital Sonifier with a Model 102C CE Converter ( Danbury , CT ) . The extract was placed on ice and frozen at −80°C . The protein concentration was determined using Bradford reagent ( Bio-Rad , Hercules , CA ) according to the manufacturers instructions and calibrated to a standard curve using bovine serum albumin . Total amino acid content was measured using a Hewlett Packard Amino Quant II System by the Protein Chemistry Laboratory at Texas A&M University . To compare two sets of values , we used the Student's T Test when the number of samples was greater than two and the Fisher's Exact T Test when there were two samples ( Microsoft Excel , Seattle WA ) . | Zinc is an essential nutrient that is critical for human health . However , excess zinc can cause toxicity , indicating that regulatory mechanisms are necessary to maintain homeostasis . The analysis of mechanisms that promote zinc homeostasis can elucidate fundamental regulatory processes and suggest new approaches for treating disorders of zinc metabolism . To discover genes that modulate zinc tolerance , we screened for C . elegans mutants that were resistant to zinc toxicity . Here we demonstrate that mutations of the histidine ammonia lyase ( haly-1 ) gene promote zinc tolerance . haly-1 encodes a protein that is similar to vertebrate HAL , an enzyme that converts histidine to urocanic acid . Mutations in the human HAL gene cause elevated levels of serum histidine and abnormal zinc metabolism . Mutations in C . elegans haly-1 cause elevated levels of histidine , suggesting that histidine causes resistance to excess zinc . Consistent with this hypothesis , we demonstrated that dietary histidine promoted tolerance to excess zinc in wild-type worms . Mutations in haly-1 and supplemental dietary histidine also caused resistance to nickel , another metal that can bind histidine . A likely mechanism of protection is chelation of zinc and nickel by histidine . These studies suggest that histidine plays a physiological role in zinc metabolism . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome",
"sequencing",
"genetic",
"mutation",
"genetic",
"screens",
"cloning",
"genetics",
"molecular",
"genetics",
"biology",
"genomics",
"genetics",
"and",
"genomics"
] | 2011 | Histidine Protects Against Zinc and Nickel Toxicity in
Caenorhabditis elegans |
In Brazil , tungiasis is endemic in some resource-poor communities where various domestic and sylvatic animals act as reservoirs for this zoonosis . To determine the effect of control measures on the prevalence and intensity of infestation of human and animal tungiasis , a repeated cross-sectional survey with intervention was carried out . In a traditional fishing community in Northeast Brazil , humans and reservoir animals were treated , and premise-spraying using an insecticide was done , while a second fishing community served as a control . Both communities were followed up 10 times during a 12-month period . At baseline , prevalence of tungiasis was 43% ( 95% confidence interval [CI]: 35%–51% ) and 37% ( 95% CI: 31%–43% ) in control and intervention villages , respectively . During the study , prevalence of tungiasis dropped to 10% ( 95% CI: 8%–13%; p<0 . 001 ) in the intervention village , while the prevalence remained at a high level in the control village . However , after one year , at the end of the study , in both communities the prevalence of the infestation had reached pre-intervention levels . Whereas the intensity of infestation was significantly reduced in the intervention community ( p<0 . 001 ) , and remained low at the end of the study ( p<0 . 001 ) , it did not change in the control village . Our study shows that a reduction of prevalence and intensity of infestation is possible , but in impoverished communities a long-lasting reduction of disease occurrence can only be achieved by the regular treatment of infested humans , the elimination of animal reservoirs , and , likely , through environmental changes . Controlled-Trials . com ISRCTN27670575
Tungiasis is a parasitic disease caused by the sand flea Tunga penetrans . Female fleas penetrate into the epidermis where they undergo a process of so-called neosomy and expel several hundred eggs into the environment . After a period of six weeks , the parasite dies in situ and is sloughed off the epidermis by tissue repair mechanisms [1] . During the last decades growing urbanization and improved housing has resulted in a reduction of prevalence . Today , the occurrence of tungiasis is confined to resource-poor communities located at the coast or in the rural hinterland , and to slums of rapidly growing urban agglomerations in Latin America , the Caribbean and Sub-Saharan Africa [2] , [3] . In these settings prevalences range between 15% and 51% [2] , [4]–[6] . Since prevalence , intensity of infestation , and morbidity are positively related [7] , debilitating and disfiguring sequels are common in resource-poor rural and urban communities . Tungiasis is clearly a neglected disease of marginalized populations [8] , [9] . Although by its nature a self-limiting disease , tungiasis causes considerable morbidity [10] , [11] . Fissures , ulcers , gangrene , lymphedema , deformation and loss of nails and auto-amputation of digits are known sequels [10] . In non-immune individuals tungiasis is a risk factor for tetanus [12] , [13] . Superinfection of the lesions is virtually constant [14] , [15] , and a variety of aerobic and anaerobic bacteria have been isolated from embedded sand fleas [16] , [17] . Beside humans , T . penetrans parasitizes a range of domestic animals , such as dogs , cats , pigs and rodents [18] , [19] . In Brazil , dogs and cats act as important reservoirs for the intra- and peridomiciliary transmission of sand fleas [18] , [20] . When humans live in close contact with infested animals , the risk of infestation is high and the intensity of infestation is high [20] . The control of tungiasis in a resource-poor population with interventions targeted at the human and the animal population has never been described . Here we present the results of an intervention performed in collaboration with public health services in an endemic community in northeast Brazil . The results show that combining treatment of humans with treatment of animals and focal spaying of an insecticide reduced prevalence and intensity of infestation .
To evaluate the effect of multiple interventions on the prevalence of tungiasis , two endemic communities were selected in Ceará State , northeast Brazil ( Balbino and Pedro de Souza ) . Both communities are situated in Cascavél Municipality about 60 km south of Fortaleza , the state capital . The fishing communities are separated about 6 km and located within sand dunes near the Atlantic Ocean , showing little fluctuation of their population . The two communities are very similar with regard to demographic , crowding , social and economic characteristics . Houses are located on rather large compounds surrounded by fences and built on sandy soil . The quality of housing is poor and streets are not paved . Kitchens are indoors or consist of open stalls on the compound . Both communities are integrated in the national Family Health Program ( “Programa da Saúde da Família” ) and served by community health care workers ( “agentes comunitários de saúde” ) . During the study , the closest primary health care center was located in another community , some 10 km away . In October 2002 , Balbino was inhabited by 148 families with a total population of 630 . The community of Pedro de Souza comprised 251 individuals in 60 families . Inhabitants of all age groups were eligible for the study , provided they had spent at least four days per week in the village during the last three months . Dogs and cats were included if they were born in the villages or had lived there for at least two months . Pigs , goats , sheep , cows and horses , other species of domestic animals occurring in the villages were previously excluded to be animal reservoirs of T . penetrans in this setting [20] . The study was conceived as a repeated cross-sectional survey with intervention . To assess the impact of interventions on the prevalence and intensity of infestation , two complete villages were compared . In the community Balbino various interventions were implemented while the Pedro de Souza community served as a control . Interventions were coordinated and implemented by the Mandacaru Foundation ( Fortaleza , Brazil ) in collaboration with the Health Secretariat of Cascavel Municipality . Between November 2002 and November 2003 , a total of 10 surveys were planned and carried out in each community . The villages were visited during identical periods according to pre-defined dates with a maximum delay of 10 days between intervention and control community . The study started in November 2002 , in the middle of the dry season when the prevalence of tungiasis peaks [21] . During the preparatory phase , contact was made with community leaders , and the objectives of the study were explained . Censuses of the human and the animal population were performed and all houses mapped using a global positioning system ( GPS ) . All data were collected by door-to-door surveys . The primary outcome was the prevalence of tungiasis ( dichotomous ) , and the secondary outcome intensity of infestation ( continuous ) . All surveys were carried out by the same investigators ( S . S . , L . W . ) , accompanied by community health agents . During the surveys the human and the animal population were carefully examined for the presence of embedded sand fleas , according to previously established guidelines . In humans the entire body was examined ( except the genitals ) to identify any ectopic lesions [22] . In dogs and cats examination focused on paws , abdomen and muzzle , the topographic areas most commonly affected [18] . The following findings were considered to be diagnostic for human as well as animal tungiasis [1]: a red-brownish spot with a diameter of 1–3 mm with visible posterior segments of the penetrated flea ( early stage ) ; a circular whitish lesions with a diameter of 4–10 mm with a central black dot ( mature stage ) , round black crust surrounded by necrotic tissue ( late stage with dead parasite ) . Typical residuals in the epidermis , lesions altered through manipulation ( such as partially or totally removed fleas leaving a characteristic crater-like sore in the skin ) , and suppurative lesions ( caused by the use of nonsterile instruments ) , were recorded as well . Lesions were differentiated into viable , dead and manipulated lesions . At each assessment , participants were also asked about any adverse events that may have occurred in consequence of the intervention . Clinical experience shows that infestation with more than ten sand fleas oftentimes result in considerable morbidity . We therefore considered infestation with up to five embedded fleas as low , between six and as 10 moderate and >10 lesions as a high intensity of infestation , in analogy to a previously used classification [23] . In animals the stratification was <10 , 11–20 and >20 , respectively . After baseline examination in November 2002 , the control measures were carried out in the intervention village ( Balbino ) . From November 2002 through January 2003 , from all infested individuals embedded sand fleas were extracted every two to three weeks by experienced health care professionals under sterile conditions . The remaining sore was treated with an antibiotic ointment . During the same period all cats and dogs were treated with trichlorphone 97% in oily solution ( Neguvon , Bayer do Brasil , São Paulo , Brazil ) or neck collars impregnated with propoxur and flumethrin ( Kiltix , Bayer Bayer do Brasil , São Paulo , Brazil ) . In case of loss of neck collars these were substituted at the next survey . In February 2003 , deltamethrin was used for focal premise treatment . Focal spraying was performed by trained personnel of the Health Secretariat of Cascavel Municipality . The insecticide was sprayed on the ground next to the houses targeting areas in which off-host development of T . penetrans was suspected to occur , such as preferred whereabouts of dogs and cats , and shady places under trees [24] , or inside houses in the case of a sandy floor . Focal premise treatment using insecticides was repeated twice during a period of six weeks . Table 1 summarizes the type and the period of interventions . The study protocol was approved by the Ethical Review Board of the Federal University of Ceará Fortaleza , Brazil ( Protocol no . 195/02 ) . In addition , an ad hoc ethical committee , consisting of physicians , community members and professionals of the Health Secretariat of Cascavel Municipality approved the study protocol . Informed written consent was obtained from all study participants . In the case of minors , written consent was obtained from the minors and their carers . Written consent was also obtained from pet owners . At the end of the study surgical treatment of tungiasis was offered to affected individuals in both communities . Data were entered in Epi Info version 6 . 04d ( CDC , Atlanta , USA ) , checked for entry errors and transferred to SPSS 11 . 04 for Macintosh ( SPSS Inc . , Chicago , IL , USA ) for analysis . The χ2 test was employed to determine the significance of difference of proportions between population groups , and between the intervention and control communities . For the comparison of point prevalences within one community , the McNemar test was used . Infestation intensities were compared by the Mann-Whitney test . The prevalence ratio [PR] for tungiasis and the significance of differences in the relative distribution of tungiasis were calculated in contingency tables . To calculate the relative prevalence reduction Balbino and Pedro de Souza were regarded as one study population , and belonging to either community was considered an exposure variable . The relative prevalence reduction was calculated as follows: prevalence of tungiasis in Pedro de Souza–prevalence of tungiasis in Balbino/prevalence of tungiasis in Pedro de Souza .
The intervention community was considerable bigger , but socio-demographic characteristics were similar in both communities . Table 2 summarizes baseline characteristics of the study populations . At baseline ( November 2002 ) the prevalence of tungiasis in the human population was slightly higher in Pedro de Souza , the control village ( 43% , 95% confidence interval [CI]: 35–51% ) , than in Balbino , the intervention village ( 37% , 95% CI: 31–43%; ) . However , the difference was not significant ( Figure 1A; p = 0 . 11 ) . At the first follow-up in December 2002 –after completion of the first cycle of intervention measures 1 and 2 ( see Table 1 ) – the prevalence of tungiasis dropped to 25% ( 95% CI: 35–51% ) in Balbino , while it remained unchanged in Pedro de Souza ( pre-intervention versus one month post-intervention p = 0 . 05 and p = 0 . 47 , respectively; Figure 1A ) . In the intervention village prevalence continued to decrease to 18% ( 95% CI: 15–22% ) in January 2003 ( p = 0 . 001 ) , 15% ( 95% CI: 12–18% ) in February ( p = 0 . 17 ) , and 10% ( 95% CI: 8–13% ) in March ( p = 0 . 001; all p compared to the preceding survey ) . During the same period no significant reduction in prevalence of tungiasis was noted in Pedro de Souza ( Figure 1A ) . Here , prevalence in March 2003 remained as high as 36% ( 95% CI: 29–44% ) . During the rainy season a strong reduction in prevalence was observed in the control community , too . Prevalence decreased significantly from 36% ( 95% CI: 29–44% ) in March to 9% ( 95% CI: 4–13% ) in May ( p<0 . 001 ) . With the beginning of the dry season ( June–July 2003 ) prevalence started to rise in both communities , and prevalence curves were almost parallel . By November 2003 one year after beginning of the study , prevalences in both communities had reached the pre-intervention level . During the 12-month study period , the prevalence curves of animal tungiasis showed a similar pattern as compared to human tungiasis , with higher baseline values in Pedro de Souza ( 86% , 95% CI: 71–100% ) than in Balbino ( 64% , 95% CI: 52–75%; p = 0 . 02 ) and an impressive decrease in prevalence during the rainy season ( Figure 1B ) . However , the variation of measurements was considerably higher than in the human population . The PR of tungiasis showed no significant association at the beginning of the intervention in November 2002 ( PR = 0 . 81 , 95% CI: 0 . 65–1 . 03; p = 0 . 11 ) . After cessation of the intervention ( March 2003 ) the PR had decreased significantly in Balbino , as compared to Pedro de Souza ( PR = 0 . 28 , 95% CI: 0 . 2–0 . 39; p<0 . 001 ) . At the end of the study in November 2003 the PR remained slightly but significantly lower in the intervention village ( PR = 0 . 69 , 95% CI: 0 . 52–0 . 93; p = 0 . 017 ) . The prevalence reduction in Balbino was 55% and 19% in March and November 2003 , respectively . In the intervention population , the prevalence of individuals with a high and a moderate intensity of infestation decreased significantly from pre-intervention ( November 2002 ) to four months after intervention ( March 2003 ) : prevalence of individuals with high intensity 8 . 4% ( 95% CI: 6–11% ) versus 0 . 8% ( 95% CI: 0 . 02–1 . 5%; p<0 . 001 ) and with moderate intensity 3 . 3% ( 95% CI: 1 . 6–5% ) versus 0 . 6% ( 95% CI: 0 . 001–1 . 2% ) , p = 0 . 002 ( Figure 2A ) . In contrast , no significant reduction was observed in the control village: 3% ( 95% CI: 0 . 4–6% ) versus 2% ( 95% CI: 0–4%; p = 0 . 99 ) and 6 . 8% ( 95% CI: 2–11% ) versus 3 . 4% ( 95% CI: 0 . 4–6% ) , respectively ( p = 0 . 25; Figure 2A ) . At the end of the study ( November 2003 ) the prevalence of heavy and moderate infested individuals remained significantly lower in Balbino as compared to the pre-intervention level: 8 . 4% ( 95% CI: 6–11% ) versus 3% ( 95% CI: 1–5% ) after intervention ( p = 0 . 001 ) , and 3 . 3% ( 95% CI: 1 . 6–5% ) versus 1 . 6% ( 95% CI: 0 . 3–2 . 8%; p = 0 . 013 ) , respectively . In contrast , at the end of the study in Pedro de Souza the proportion of heavily and moderate infested individuals was even higher than at baseline: 3% ( 95% CI: 0 . 4–6% ) versus 4% ( 95% CI: 0 . 5–7%; p = 0 . 25 ) and 6 . 8% ( 95% CI: 2–11% ) versus 6 . 7% ( 95% CI: 2–11; p = 0 . 99 ) , respectively ( Figure 2A ) . Similar , the total number of embedded sand fleas per person was significantly reduced in the intervention village , from a median of 18 in November 2002 to a median number of one in March 2003 ( p<0 . 001 ) and remained significantly lower in November 2003 ( median of five lesions; p<0 . 001 compared to baseline ) . There were no significant differences in the reduction of intensity of infestation in the animal population ( Figure 2B ) . No adverse events related to the intervention were recorded . However , as our therapy did not differ from the treatment done by most community members themselves , we assume that people did not report any minor adverse events associated with surgical extraction ( such as pain and superficial bleeding ) .
The high prevalence of tungiasis in endemic areas and the important morbidity associated with this parasitic skin disease call for the implementation of control measures . As a first step to prove that successful intervention is possible , we determined the impact of repeated rounds of surgical extraction of embedded sand fleas in humans in combination with on-host treatment of dogs and cats , and focal spraying of an insecticide on the premises . Our study shows that the interventions were effective to control tungiasis in the short-term but failed to show an impact on the long run . This is reflected by a minuscule reduction of the prevalence ratio at the end of the study , i . e . 12 months after start of interventions . Several factors seem to be responsible for the re-increase of prevalence to baseline level in the intervention village at the end of the study . Firstly , the backbone of the intervention , the surgical extraction of embedded sand fleas , obviously has several shortcomings . As inhabitants of endemic areas rarely remove embedded sand fleas in a systematic manner–it is time-consuming , painful and often results in superinfection [25]– individuals could have abstained from the treatment . As a consequence , the human reservoir of T . penetrans may not have diminished as intended [23] . Presumably the inflammatory reaction of the skin at sites of embedded sand fleas facilitates penetration [25] . Since the barrier function of the epidermis is not immediately reconstituted after sand fleas have been taken out ( the sore produced by the surgical manipulation even might temporarily increase the surface of the skin particular susceptible to penetration ) , the idea that i ) the reduction of the parasite burden would prevent re-infestation , and that ii ) by consequence , the number of eggs expelled into the environment would reduce transmission , might be an invalid assumption . Actually , people at risk for immediate re-infestation could be more susceptible to the infestation with T . penetrans after extraction and may only profit from the surgical extraction after complete healing of the skin . We suggest that the prevention of infestation , rather than the surgical extraction of already embedded sand fleas , may interrupt transmission more effectively . Zanzarin , a plant-based repellent , has been shown to effectively prevent the infestation with T . penetrans in areas with high attack rates [26] . This compound would be an ideal candidate for prophylaxis but was not available in Brazil when the study was designed . Secondly , many cats and dogs remained infested despite the on-host treatment ( Figure 1B ) . By consequence these animals continued spreading T . penetrans and contributed to ongoing transmission in the community [27] . After on-host intervention had been stopped , more and more animals became re-infested , with an even higher prevalence at the end of the study as compared to baseline data ( Figure 1B ) . Assumably , these animals carried sand fleas to the compounds of their owners where they fuelled peri- and intradomiciliary transmission [20] . Actually , in both communities the strongest increase in prevalence of tungiasis in humans followed a strong increase in prevalence in animals ( Figures 1 A , B ) . Recently developed on-host products , such as a combination of imidacloprid and permethrin ( Advantix ) , effectively prevented infestation with T . penetrans in animals and lowered parasite burden [27] . Again , this product was not available when the study was conceived . Finally , focused premise treatment with deltamethrin aimed to interrupt the off-host cycle of T . penetrans [18] , [28] . For an optimal efficacy , focal spraying has to be applied at all sites where off-host development takes part in the soil , which means , that those sites have to be identified first . Due to delivery problems and constraints in qualified personnel , focal premise treatment only started in January , i . e . , very late in the seasonal cycle of T . penetrans . In addition , there was no expertise to examine soil samples for developmental stages of sand fleas . Obviously , spraying of breeding sites is better done before the parasite population has expanded , i . e . , at the beginning of the dry season [21] . However , the significant reduction in prevalence after the implementation of this intervention seems to have booster the protective effect of the preceding interventions . When transmission of T . penetrans is altered–e . g . through an intervention–the intensity of infestation reflects the infestation rates individuals have experienced during the last months [7] . Thus , intensity of infestation is a better outcome measure to evaluate the impact of an intervention than assessment of the prevalence . In addition , it is a better proxy of morbidity reduction , as morbidity is significantly correlated to intensity of infestation [7] . In contrast , prevalence merely measures absence or presence of infestation at a given point of time and will therefore rapidly increase if the attack rate is high , such as at the beginning of the dry season . Prevalence does not allow inferring on decreasing attack rates prior to the assessment since embedded sand fleas and their remains are visible for up to three months [1] . Our study shows that in the intervention community a prolonged effect on the intensity of infestation of the human population occurred . This means , that despite a persistent high prevalence of tungiasis , attack rates were reduced . However , this effect was not observed in the animal population ( Figure 2B ) . This indicates that the intervention measures applied were not very effective in reducing the parasite burden in dogs and cats . Any assessment of intervention methods is hampered by the characteristic seasonal variation of tungiasis . Disease occurrence decreases as soon as the rainy season starts and re-increases with the beginning of the dry season [21] . We therefore opted to include a control community where no intervention was done . Although both fishing communities were selected by their similarity with respect to demographic , physical and socio-cultural characteristics , one cannot exclude that they have differed in an unknown factor of epidemiological relevance , an additional , hitherto neglected animal reservoir . Such a factor could be responsible for different dynamics in prevalence between Balbino and Pedro de Souza during the year . However , different dynamics in the same epidemiological setting have never been reported and the similar development of prevalence towards the end of the study renders such an explanations unlikely . One major problem in the interpretation of data arises from the differences in population size between the two communities . This difference bares the possibility that significances in the changes of prevalence and/or intensity of infestation in the control village during the study period were not detected due to the smaller sample size . Non-participation may have biased the assessment of point-prevalences during the observation period . For surveys with a low participation , it is conceivable that over- or under-representation of infested individuals has skewed the results towards higher or lower prevalence . Fluctuation of participation was particularly high in animals , indicating that point prevalences in the animal population were likely to be biased . Another shortcoming of the study is that the study design does not allow identifying the relative effective of the three interventions applied in Balbino village . To address such an issue , various communities would have to be included , in each of which a particular intervention has to be performed . For operational reasons and financial constraints this could not be done . Due to the operational nature of the study , we did not include any multivariate analysis . The study was not designed as a typical trial and thus did not include detailed collection of data on possible confounders . We aimed to describe whether an intervention has any effect rather than identifying independent factors for success or failure . Off-host stages of T . penetrans develop best in dry soil or in dusty soil containing organic material [9] , [21] , [29] . Measures aiming to interrupt the off-host development should therefore focus on physically changing the environment in which eggs , pupae , and larva develop . This can be done through paving streets , cementing floors , and eliminating uncontrolled disposal of waste in public areas and private compounds [30] , [31] . However , theses interventions require substantial funds and are beyond the economic capabilities of most communities where tungiasis is endemic . Although the present study was conducted in a rural fishing community , it is conceivable that interventions are also applicable to urban settings where the animal reservoirs are similar [18] . Our study is a first step in the exploration of possible control measures against T . penetrans . Further studies are needed to assess the full potential of putative interventions . To do so , a cluster-randomized phased implementation study with various communities phasing-in specific interventions would be an ideal approach [32] . The Brazilian Unified Health System ( SUS ) with a network of primary health care clinics , many located in endemic communities , would offer a convenient way to coordinate such a study . Randomized clinics could implement intervention measures , such as prevention of infestation by application of Zanzarin , surgical extraction of embedded fleas , distribution of animal treatment , and coordination of premise spraying with insecticides . This would allow identifying the most effect measure over a prolonged period of time and controlling for seasonal variation [32] . | Tungiasis is a disease caused by the sand flea Tunga penetrans , a parasite prevalent in many impoverished communities in developing countries . The female sand flea penetrates into the skin of animals and humans where it grows rapidly in size , feeds on the host's blood , produces eggs which are expelled into the environment , and eventually dies in situ . The lesions become frequently superinfected and the infestation is associated with considerable morbidity . Clearly , tungiasis is a neglected disease of neglected populations . We investigated the impact of a package of intervention measures targeted against on-host and off-host stages of T . penetrans in a fishing community in Northeast Brazil . These measures decreased disease occurrence only temporarily , but had a sustained effect on the intensity of the infestation . Since infestation intensity and morbidity are correlated , presumably the intervention also lowered tungiasis-associated morbidity . Control measures similar to the ones used in this study may help to effectively control tungiasis in impoverished communities . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"microbiology/parasitology",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"infectious",
"diseases/skin",
"infections",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] | 2008 | Controlling Tungiasis in an Impoverished Community: An Intervention Study |
The intracellular pathogen Brucella abortus survives and replicates inside host cells within an endoplasmic reticulum ( ER ) -derived replicative organelle named the “Brucella-containing vacuole” ( BCV ) . Here , we developed a subcellular fractionation method to isolate BCVs and characterize for the first time the protein composition of its replicative niche . After identification of BCV membrane proteins by 2 dimensional ( 2D ) gel electrophoresis and mass spectrometry , we focused on two eukaryotic proteins: the glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) and the small GTPase Rab 2 recruited to the vacuolar membrane of Brucella . These proteins were previously described to localize on vesicular and tubular clusters ( VTC ) and to regulate the VTC membrane traffic between the endoplasmic reticulum ( ER ) and the Golgi . Inhibition of either GAPDH or Rab 2 expression by small interfering RNA strongly inhibited B . abortus replication . Consistent with this result , inhibition of other partners of GAPDH and Rab 2 , such as COPI and PKC ι , reduced B . abortus replication . Furthermore , blockage of Rab 2 GTPase in a GDP-locked form also inhibited B . abortus replication . Bacteria did not fuse with the ER and instead remained in lysosomal-associated membrane vacuoles . These results reveal an essential role for GAPDH and the small GTPase Rab 2 in B . abortus virulence within host cells .
Brucella abortus invades both phagocytic and non-phagocytic cells [1]–[6] residing inside a membrane-bound compartment called the Brucella-containing vacuole ( BCV ) . Bacteria ensure their survival and replication within host cells by avoiding fusion with lysosomes and by controlling interactions with the endoplasmic reticulum ( ER ) [1] , [5] . The membrane of the BCV is converted into an ER-derived organelle that is permissive for replication [1] . Interactions between BCV and ER occur at dynamic membrane complexes named ERES for ER exit sites , where membrane fusion and fission events take place . These events are regulated by the small GTPase Sar 1 . Sar 1 controls the assembly of COPII complexes on the ER mediating vesiculation and tubulation of the ER membrane towards the Golgi apparatus [7]–[9] . These events were shown to be essential for B . abortus intracellular replication at early stages of infection [10] . The Brucella replicative organelle has been , until now , characterized by the presence of ER chaperones such as calnexin , calreticulin , the translocator sec61β , and the ER resident enzyme protein disulfide-isomerase PDI [1] , [5] , [11] . Aside from these ER resident proteins , no other eukaryotic or prokaryotic proteins have been associated with the BCV membrane as yet , identification of these proteins is essential to understand how Brucella maintains interactions with the ER and keeps replicating within this compartment . In this work , we investigated by proteomic approaches , the composition of the BCV membrane and characterized 2 eukaryotic proteins that are essential for B . abortus survival . We modified a fractionation method , initially used to analyse latex bead-containing phagosomes [12] , [13] , to isolate BCVs obtained from cells infected with B . abortus . Mass spectrometry analysis of BCV proteins separated by two-dimensional ( 2D ) gel electrophoresis revealed the presence of the eukaryotic protein GAPDH ( glyceraldehyde-3-phosphate dehydrogenase ) . Further work on GAPDH revealed a role for GAPDH and the small GTPase Rab 2 in the intracellular replication of B . abortus .
To analyse the protein composition of the BCV membrane , a large number of purified BCVs is required . We first tried to determine which cell type was more susceptible for B . abortus infection by monitoring the infection of primary phagocytic cells ( bone marrow-derived macrophages: BMDM ) and phagocytic and non phagocytic cell lines ( Raw 264 . 7 , HeLa , baby hamster kidney: BHK-21 ) . Although no difference was observed in the percentage of infected cells at 48 h post-infection ( p . i . ) between the different cell types ( nearly 35% infected cells ) , the intracellular replication of B . abortus was 10 times higher within BHK-21 cells than in the other cell types ( Figure 1A ) . As in macrophages [1] and in dendritic cells [14] , B . abortus GFP ( in green ) in BHK-21 cells were located inside a membrane-bound vacuole labelled with the ER marker calnexin ( in red ) , suggesting that B . abortus replicates in ER-derived compartments within BHK-21 ( Figure 1B ) . This result was confirmed by electron microscopy analysis of infected cells ( Figures 1C , 1D and 1E ) . B . abortus was located inside a membrane-bound compartment resembling the ER with ribosomes lining the vacuolar membrane . However , unlike what has previously been described in BMDM and HeLa cells [1] , [5] , several bacteria resided inside a unique vacuole ( Figures 1C and 1D ) . This may explain the increase of B . abortus intracellular replication within BHK-21 cells . Taken together , these results show that B . abortus extensively replicates in an ER-derived compartment in BHK-21 cells , validating BHK-21 cells as a good cell model for studying the proteic composition of BCV membranes . In order to obtain high concentration of membrane proteins , we optimized a fractionation method to isolate and purify BCVs . Within the post-nuclear supernatant ( PNS ) of BHK-21 cells infected with B . abortus for 48 h , approximately 1 . 5% of vesicles were GFP positive as detected by flow cytometry and we found that 67% of BCVs remained positive for the ER marker calnexin ( Figure S1 ) . After PNS preparation , BCVs were first purified on a 50%–12% sucrose gradient . BCVs were present at the interface , which corresponded to 37% sucrose as indicated by the densitometer measurement ( data not shown ) . Bacteria were only detected in the interface fraction ( Figure 2A , lane F ) by the presence of the Brucella transmembrane outer membrane protein Omp 25 [15] . On the contrary , ER was detected in each fraction of the sucrose gradient with an anti-calnexin antibody ( Figure 2A ) . This first purification step allowed the elimination of Rab 7-positive late endosomes and cathepsin D-positive lysosomes from the BCV membrane fraction ( Figure 2A , lane F ) . The interface fraction was then loaded onto a second sucrose step gradient . This gradient allowed the removal of early endosomes ( Figure 2A , lane H ) and ER structures non-associated with the BCV as detected by immunofluorescence and electron microscopy ( data not shown ) . Although the fraction of BCVs was now devoid of endocytic organelles ( Figure 2A , lane H ) , a few mitochondria detected by the anti-mitochondrial VDAC 1 protein antibody were still present . These could be eliminated by incubating the interface fraction with dynabeads ( M-500 subcellular ) coated with anti-VDAC 1 antibody , but most of BCVs were lost ( data not shown ) . As a consequence , for proteomic analysis we chose not add the dynabeads step . We analysed BCVs by immunofluorescence and electron microscopy to determine if BCVs were still intact after fractionation . 77% of B . abortus GFP were surrounded by ER-positive vacuoles ( Figure 2B ) . Electron microscopy analysis showed that BCVs were still intact after subcellular fractionation ( Figure 2B ) . Bacteria were surrounded by an ER membrane-bound compartment and residual ribosomes were still located on the ER vacuolar membrane . Taken together , these data indicate that we successfully isolated BCVs and preserved their membrane integrity . To determine the protein composition of the BCV membrane , vacuolar proteins were solubilized by a mild Triton X-100 treatment , precipitated by a trichloroacetic acid/acetone and analysed by 2D gel electrophoresis . Approximately one hundred spots were detected by silver gel staining ( Figure 2C ) and further analysed by mass spectrometry . Proteins of the BCV membrane identified by mass spectrometry are listed in Table S1 . Spot numbers on 2D gel ( Figure 2C ) correspond to the numbers of identified proteins listed in Table S1 . As expected , 18% of proteins identified were ER proteins ( i . e . calreticulin , ERP 57 and PDI ) and 13% ribosomal proteins confirming that the replicative niche of B . abortus is derived from the ER . Interestingly , 29% of proteins were not ER-related proteins ( 19% bacterial and 10% eukaryotic proteins ) . The remaining 40% proteins were mitochondrial proteins . Among the identified proteins , we focused on one peculiar eukaryotic protein: the Glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) annotated by spot number 49 on the 2D gel ( Figure 2C ) and in Table S1 . GAPDH has multiple functions within host cells such as glycolysis and apoptosis [16] , [17] More interestingly , GAPDH interacts with the small GTPase Rab 2 to control vesicular retrograde transport between the ER and the Golgi [18] . The presence of GAPDH on the BCV membrane leads us to hypothesise that vesicular retrograde transport may be involved in Brucella replication within the ER . Therefore , we investigated the role of GAPDH and Rab 2 in Brucella survival within host cells . First , the presence of GAPDH on the BCV membrane was confirmed by immunoblotting ( Figure 3A , lane C ) . Its partner , the small GTPase Rab 2 , was also detected in the enriched BCV fraction ( Figure 3A , lane C ) . The small GTPase Rab 1 , which is known to be a resident of the ER as well as vesicular and tubular clusters ( VTCs ) and Golgi apparatus [19] , [20] was also detected in BCVs ( Figure 3A , lane C ) . As no commercial antibody was suitable for immunofluorescence detection of GAPDH on BCVs obtained from BHK-21 cells , we analysed the presence of its partner the small GTPase Rab 2 . We found that 35% of isolated BCVs were surrounded by Rab 2 staining ( Figure 3B ) . Together these results confirm the presence of GAPDH and Rab 2 on BCVs . We further studied the role of Rab 2 and GAPDH in HeLa cells , a well-established cell culture model for Brucella infection . We showed that the GTPase Rab 2 is present on the vacuolar membrane of purified BCVs at 48 h p . i . ( Figure 3A and 3B ) . In order to determine the kinetics of Rab 2 acquisition on BCV membranes , we first examined the presence of endogenous Rab 2 on BCVs within infected HeLa cells . Although we could detect Rab 2 on isolated BCVs , labelling of infected cells was extremely weak . As a consequence , we overexpressed Rab 2 within HeLa cells by using a dominant positive form of Rab 2: Rab 2 Q65L , which corresponds to Rab 2 locked in its GTP-bound form . HeLa cells were transfected with Myc Rab 2 dominant positive Q65L for 24 h and then infected with B . abortus GFP . The intracellular replication of Brucella within cells transfected or not with Q65L Rab 2 was similar ( data not shown ) . Figure 4A represents the quantification of BCVs positive for Rab 2 Q65L at different times p . i . At 6 h p . i . few BCVs were surrounded by the active form of Rab 2 ( 20±4 . 2% ) . Then , at 10 h p . i . 68 . 4±5 . 7% of BCV had acquired Rab 2 Q65L and remained positive for Rab 2 Q65L until 48 h p . i . Figure 4B illustrates the recruitment of Rab 2 on BCVs at 10 h p . i . as indicated by the arrow . These results indicate that the recruitment of the active form of Rab 2 takes place just before the interaction of BCVs with the ER , known to occur around 12 h p . i [5] . Indeed , most of BCVs surrounded by Rab 2 Q65L were still LAMP-1-positive at 10 h p . i . ( data not shown ) . Retrograde vesicle formation from the VTC is mediated by the exchange of GDP to GTP on the GTPase Rab 2 [21] . In order to inhibit the retrograde transport between the ER and the Golgi , we used a dominant negative form of Rab 2: Rab 2 I119 , which corresponds to Rab 2 locked in its GDP-bound form . HeLa cells infected with B . abortus Ds Red were transfected with either GFP plasmid as a transfection control , myc Rab 2 , myc Rab 2 dominant negative I119 , GFP Rab 1 or GFP Rab 1 dominant negative S25N . Figures 5A , 5B and S2A illustrate the level of intracellular replication at 48 h p . i . in the different transfected cells . We observed extensive Brucella replication in control cells or cells transfected either with GFP or myc Rab 2 ( Figure 5A , 5B and S2A ) . Interestingly , transfections with either the GFP Rab 1 or its dominant negative did not affect Brucella replication , contrasting with results obtained from cells infected with Legionella pneumophila and transfected with the dominant negative Rab 1 S25N ( Figure S2 ) [22] . On the contrary , in cells transfected with myc Rab 2 dominant negative I119 ( Figure 5A and 5B ) , Brucella replication was strongly decreased ( six times ) and 89 . 8±1 . 62% of the bacteria were located in a lysosomal LAMP-1-positive compartment at 48 h p . i ( Figure 5C and S2B ) , a time point were most of the Brucella ( 88% ) are within an ER-positive , LAMP-1-negative compartments ( Figure 5C and 5D ) . This shows that in cells transfected with the Rab 2 dominant negative form , Brucella was not able to reach the ER . Indeed , only a small percentage of BCVs were positive for cathepsin D at 48 h p . i . ( 22% of BCVs within cells overexpressing Rab 2 dominant negative ) ( Figure 5E ) . In contrast 92% of heat-killed Brucella BCVs were already cathepsin D-positive at 2 h p . i . ( data not shown ) . Therefore , inhibition of retrograde vesicle formation from the VTCs mediated by Rab 2 affects Brucella replication . Taken together these results indicate that the trafficking between ER and Golgi controlled by Rab 2 is important for entry of Brucella in the ER and subsequent intracellular replication . To investigate the role of GAPDH on Brucella pathogenesis , we down-regulated the expression of GAPDH in HeLa cells infected with B . abortus by using small interfering RNA . HeLa cells transfected with GAPDH siRNA for 72 h efficiently and specifically reduced the expression level of GAPDH ( Figures 5F and 5H ) , whereas , siRNA control did not affect the GAPDH expression ( Figures 5F and 5H ) . Inhibition of GAPDH expression induced a 10 fold reduction in Brucella replication as compared to non-transfected cells or cells transfected with the siRNA control ( Figure 5G and 5H ) . This result indicates that the presence of GAPDH on the BCV membrane is required for Brucella replication . We showed above that in cells transfected with the Rab 2 dominant negative form , Brucella was not able to reach the ER and remained in a LAMP-1 compartment . Similarly , in siRNA GAPDH-treated cells , Brucella was found in a LAMP-1-positive compartment ( Figure S3 ) . In addition , inhibition of GAPDH expression prevented Rab2 recruitment on BCVs , as shown after BCV purification ( Figure S3 ) . Quantification showed that 77% of BCVs were positive for Rab 2 in control cells whereas only 12% of siRNA GAPDH-treated cells were able to recruit Rab 2 . These results show that GAPDH is an important host factor for BCV biogenesis . GAPDH is known to play several functions within host cells [16]–[18] , [23] . To confirm the involvement of the retrograde transport of the early secretory pathway in Brucella pathogenicity , we investigated the role of other key components known to control the vesicular trafficking between ER and Golgi , such as the kinase PKC ι/λ and the coat COPI complex . Inhibition of these components was performed by infecting HeLa cells transfected with small interfering RNAs . We used a PKC ι siRNA to silence the expression of the kinase PKC ι/λ , a COP B siRNA to silence the subunit β of the COPI complex , a Rab 2 A siRNA to silence the GTPase Rab 2 and a α-Enolase siRNA to silence the Enolase , an enzyme involved in glycolysis . Cellular extracts prepared from HeLa cells transfected with the appropriate siRNA for 72 h efficiently and specifically reduced the expression of PKC ι , COPI , Rab 2 and Enolase ( Figure 6A ) , whereas , siRNA control did not ( Figure 6A ) . Figure 6B shows the intracellular replication of B . abortus at 48 h p . i . within infected-HeLa cells transfected with different siRNAs . Brucella replication was reduced 2 . 2 , 2 . 3 and 5 fold in cells transfected with PKC ι siRNA , Rab 2 A siRNA and COP B siRNA , respectively as compared to cells transfected with the siRNA-A control . This result indicates that each member of the complex GAPDH/COPI/Rab2/PKCι/λ is required for Brucella replication . Surprisingly , we noticed that the intracellular replication of Brucella was reduced 4 fold under the inhibition of Enolase expression . This result suggests that host glycolysis is necessary for Brucella survival within host cells . Taken together , these results demonstrate the role played by the early secretory pathway , in particular the GAPDH/COPI/Rab2/PKCι/λ retrograde vesicles , to ensure replication within host cells at late stages of infection .
Many intracellular bacteria , with the aim of generating a suitable niche of replication , have been shown to alter the phagosomal membrane composition to avoid fusogenic interactions with lysosomes [24] . For example , Salmonella typhimurium secretes multiple effector molecules onto the vacuolar membrane ( via its type III secretion systems ) , which interact with host proteins to modulate vesicle transport and vacuolar membrane dynamics [25] . This enables Samonella to replicate in a vacuole that interacts with late endosomes whilst avoiding fusion with lysosomes . In contrast , Legionella and Brucella replicate in ER-derived compartments [1] , [26] . In the case of Legionella , several recent studies have highlighted the role of type IV secreted proteins in recruiting eukaryotic proteins to the vacuolar membrane and by mechanisms that are still unclear sustain intracellular replication [27] . Much less in known for Brucella , particularly regarding the membrane composition of BCVs . Previous work has demonstrated that the small GTPase Sar1 is implicated in Brucella intracellular survival [10] . However , apart from ER-resident proteins no other eukaryotic molecules have been associated with the BCV membrane . Phagosomal proteomic studies using latex bead-containing phagosomes have significantly helped to decipher phagosome biology [28] . Using a modified procedure for phagosome purification , we analysed in detail the protein composition of the BCV membrane in order to identify eukaryotic proteins recruited to BCVs during intracellular replication . We developed a fractionation method to isolate intact BCVs from infected BHK-21 cells . This method allowed us to establish for the first time the BCV membrane protein map of the replicative niche of Brucella . As expected , a proportion of proteins on the BCV membrane were ER and ribosomal proteins . Interestingly , one of the proteins identified was GAPDH , a non-ER eukaryotic protein which is normally located on VTCs between ER and the Golgi apparatus . This protein has been extensively studied by Tisdale et al [29]–[35] . GAPDH forms an active complex with the small GTPase Rab 2 and the protein kinase C ( PKCι/λ ) , which is necessary for secretory vesicular transport . First studies showed that an inactive form of Rab 2 had a negative effect on anterograde transport of vesicles from the ER to the Golgi [21] . Recently , the group of Tisdale has shown that Rab 2 modulates protein retrograde transport from the Golgi to the ER by recruiting GAPDH to VTCs which allows the release of retrograde-directed vesicles [31] , [35] . This retrograde transport requires a functional GAPDH/COPI/Rab2/PKCι/λ complex . Presence of GAPDH and Rab 2 on the BCV membrane suggests that Brucella is somehow interacting with VTCs or intercepting vesicle trafficking of the retrograde transport . Consistent with this hypothesis we found that inhibition of GAPDH resulted in reduced intracellular replication of Brucella . However , we cannot exclude that its role in the host cell glycolysis also contributes to the intracellular survival of Brucella . Indeed , silencing of enolase , another enzyme involved in glycolysis , also resulted in inhibition of Brucella replication . Further work is necessary to determine if Brucella is directly using the host glycolysis to its advantage , for example as a source of energy . Nevertheless , the implication of retrograde transport in Brucella virulence is clearly demonstrated by the inhibition of the bacterial replication upon silencing of each member of the complex GAPDH/COPI/Rab2/PKCι/λ . In addition , we found that Rab 2 is recruited on the BCV membrane before fusion with the ER suggesting that the retrograde transport might have an important role in the establishment of the bacterial replicative niche . Previous work has demonstrated that BCV-ER fusion events occur specifically at ER exit sites and are mediated by the small GTPase Sar1 . This work also demonstrated that the anterograde pathway mediated by COPI/ARFI-coated vesicles are not involved in the BCV-ER fusion and that COPII complex formed on ER membranes was found in close apposition to BCV [10] . Our results also implicate the small GTPase Rab 2 , in the early events of BCV biogenesis since inactivation of Rab 2 using the dominant negative prior to infection prevented fusion of BCVs with the ER . When inactivation is performed after infection , Brucella is able to start to replicate at 24 h ( data not shown ) whereas at 48 h p . i . there is a strong replication effect . Overall , these results suggest that Rab 2 is also necessary for Brucella survival after it has established its ER-derived replication niche . Consistent with this hypothesis , we found that Brucella replication was also affected with prolonged brefeldin A treatment , which causes the Golgi apparatus to redistribute to the ER ( data not shown ) . Regeneration of the Golgi apparatus after brefeldin A treatment allowed Brucella to recover and replicate ( data not shown ) . Therefore , it is possible that trafficking of secretory vesicles from the Golgi apparatus to the ER is beneficial for Brucella replication . Perhaps , extensive bacterial replication requires an additional membrane input that may come from vesicles from the secretory pathway . Brucella seems to manipulate host cells differently than Legionella pneumophila , another bacterial pathogen which replicates in the ER . Indeed , Legionella was shown to manipulate host cells by secreting specific type IV secretion system bacterial effectors , for example DrrA ( or LidA ) [36] , [37] . DrrA is able to mimic the guanosine exchange factor ( GEF ) of the small GTPase Rab 1 to catalyse the exchange of GDP to GTP which lead to the activation and the recruitment of Rab 1 on the Legionella-containing vacuole [36] , [37] . Interestingly , another secreted effector LepB acts like a GTPase-activating protein ( GAP ) which inactivates Rab 1 [38] . Inhibition of Rab1 impairs Legionella replication [22] . In this study , we demonstrate that Rab 1 is not involved in intracellular replication of Brucella . Rab 1 is implicated in the anterograde pathway whereas Rab 2 is also essential for the retrograde pathway . After 24 h of replication within its ER-derived vacuole , Legionella lyses its vacuole to reach the cytosol and to infect the neighbouring host cells [39] . This suggests that Legionella does not need an influx of membrane since the bacteria lyse the host cell relatively quickly compared to Brucella , which continues to replicate within vacuoles for a long period of time ( at least another 48 h ) [1] . Therefore it is possible that Brucella might require an extensive influx of membrane for its own survival . Targeting of the retrograde pathway could ensure continued fusion with incoming vesicles and an influx of membrane . At this stage , it is unclear how Brucella is controlling biogenesis of ER-derived BCVs . It is possible that some secreted effector proteins not yet identified are directly recruiting GAPDH and Rab 2 to the BCV membrane . The type IV secretion system VirB , essential for sustained interaction and fusion with ER membranes [1] , could secrete these effectors to recruit Rab 2 onto the BCV membrane . We are currently undertaking further studies to identify the bacterial effector ( s ) which target ( s ) Rab 2 and GAPDH .
The bacterial strains used in this study were the smooth virulent B . abortus strain 2308 GFP [1] or 2308 Ds Red kindly provided by Jean-Jacques Letesson ( URBM , Immunology Laboratory , FUNDP , Namur , Belgium ) . Bacteria were inoculated in tryptic soy broth ( TSB . Sigma-Aldrich ) with kanamycin and grown at 37°C overnight ( 16 h ) as previously described for infections [1] . GFP vector used was pEGFP-C1 ( Clontech ) . Plasmids pGEM Rab 2 , pGEM Rab 2 I119 and pGEM Rab 2 Q65L were kindly given by Craig Roy ( Section of Microbial Pathogenesis , Yale University School of Medicine , New Haven , USA ) . The rab2 sequence from the vectors pGEM Rab 2 , pGEM Rab 2 I119 and pGEM Rab 2 Q65L was amplified by PCR with two specific primers: O-403 ( 5′GGGGACAAGTTTGTACAAAAAAGCAGGCTTCGCGTACGCCTATCTCTTCAAGTACAT3′ ) and O-404 ( 5′GGGGACCACTTTGTACAAGAAAGCTGGGTCCTAACAGCAGCCGCCCCCAGCCTGCTG3′ ) . The myc tag ( pCMV-myc ) was added by Gateway procedure according to the manufacturer's instructions ( Invitrogen ) to obtain myc Rab 2 , myc Rab 2 I119 and myc Rab 2 Q65L . Plasmids GFP Rab 1 and GFP Rab 1 S25N were also kindly given by Craig Roy . The silencing of endogenous GAPDH was performed by small interfering RNA with the commercial siRNA GAPDH ( Applied Biosystems Ambion ) . A negative control siRNA was also used ( Applied Biosystems Ambion ) as a transfection control . Silencing of endogenous PKC ι , COP B , Rab 2 A and α-Enolase were performed by small interfering RNA with the corresponding commercial siRNA ( Santa Cruz Biotechnology ) . A negative control siRNA-A was also used ( Santa Cruz Biotechnology ) as a transfection control . The primary antibodies used were: a mouse monoclonal anti-actin given by Lena Alexopoulou ( CIML , Marseille , France ) ; a rabbit polyclonal anti-calnexin ( Stressgen ) ; a rabbit polyclonal anti-cathepsin D [40]; a goat polyclonal T-14 COP B ( Santa Cruz Biotechnology ) ; a mouse monoclonal anti-GAPDH ( Sigma Aldrich ) ; a rabbit polyclonal anti-giantin ( Covance ) ; a mouse monoclonal anti-GM130 ( Transduction Laboratories ) ; a rabbit polyclonal anti-LAMP-1 ( Abcam ) ; a mouse monoclonal 4A1 anti-LAMP-1 kindly given by Jean Gruenberg ( University of Geneva , Switzerland ) ; a mouse monoclonal 9E10 anti-c-myc tag ( Santa Cruz Biotechnology ) ; a mouse monoclonal anti-Omp 25 ( A59/05F01/C09 ) kindly given by Michel Zygmund ( INRA , Tours , France ) ; a mouse monoclonal H-12 PKC ι ( Santa Cruz Biotechnology ) ; a rabbit polyclonal anti-Rab 1 ( Santa Cruz Biotechnology ) ; a polyclonal rabbit anti-Rab 2 and a mouse monoclonal anti-Rab 5 kindly given by Marino Zerial ( Max Planck Institute of Molecular Cell Biology and Genetics , Dresden , Germany ) ; a rabbit polyclonal anti-Rab 7 [40] and a rabbit polyclonal anti-VDAC 1 ( Abcam ) . The secondary antibodies used were: goat anti-mouse HRP ( Sigma Aldrich ) ; goat anti-rabbit HRP ( Sigma Aldrich ) for western blotting; and donkey anti-rabbit Texas red ( Jackson ImmunoResearch ) ; donkey anti-mouse Texas red ( Jackson ImmunoResearch ) ; donkey anti-mouse Cy3 ( Jackson ImmunoResearch ) ; donkey anti-mouse FITC ( Jackson ImmunoResearch ) ; donkey anti-mouse Cy5 ( Jackson ImmunoResearch ) ; phalloidin-tetramethylrhodamine isothiocyanate ( TRITC . Sigma Aldrich ) for immunofluorescence and a IgG anti-rabbit coupled to phycoerytrin ( PE . Serotec ) for flow cytometry . Baby hamster kidney ( BHK-21 ) cells were cultured at 37°C with 5% CO2 atmosphere in GMEM ( Glasgow's modified Eagle's medium . Gibco ) supplemented with 10% Tryptose Phosphate Broth ( Sigma Aldrich ) , 5% FCS ( Perbio ) and 1% L-glutamine ( Gibco ) and seeded 24 h before infection on 55 cm2 culture dishes ( 1 . 6×106 cells per dish ) for fractionation or at a surface ratio of 1/10 in 24-well plates containing 12-mm glass coverslips for immunofluorescence and CFUs . HeLa cells and Raw 264 . 7 macrophages were cultured at 37°C in a 5% CO2 atmosphere in DMEM supplemented with 10% FCS , 1% non-essential amino acids and 1% L-glutamine and seeded 24 h before infection at a surface ratio of 1/10 in 24-well plates containing 12-mm glass coverslips . Bone marrow-derived macrophages ( BMDM ) were isolated from femurs of 6 to 10-week-old C57Bl/6 female mice and differentiated into macrophages as previously described [41] . Infections were performed at a multiplicity of infection of 200∶1 by centrifuging bacteria onto BHK-21 , HeLa cells , BMDM or Raw 264 . 7 macrophages at 400 g for 10 min at 4°C , and then by incubating the cells for 1 h ( for BHK-21 and HeLa cells ) or 15 min ( for BMDM and Raw 264 . 7 macrophages ) at 37°C under a 5% CO2 atmosphere . Cells were extensively washed with their respective medium to remove extracellular bacteria and were incubated for an additional hour in their respective medium supplemented with 50 µg/ml gentamycin to kill extracellular bacteria . Thereafter , the antibiotic concentration was decreased to 10 µg/ml . To monitor Brucella intracellular survival , infected cells were washed three times with PBS and lysed with 0 . 1% ( vol/vol ) Triton X-100 in PBS . Serial dilutions in PBS of lysates were plated onto TSB agar plates to enumerate CFUs . The expression of GFP , myc Rab 2 , myc Rab 2 I119 , myc Rab 2 Q65L , GFP Rab 1 , GFP Rab 1 S25N were performed by transfecting HeLa cells using the FuGENE transfection reagent ( Roche ) , according to the manufacturer's instructions . Depending of experiment , transfections were performed either 24 h before or 2 h after Brucella infection and were left to proceed until the time of analysis . The transfection of small interfering RNA si GAPDH and siRNA control were performed on HeLa cells using the siPORT™ Amine transfection agent ( Applied Biosystems Ambion ) according to the manufacturer's instructions . The transfection of PKC ι , COP B , Rab 2 A , α-Enolase siRNAs and control siRNA-A were performed on HeLa cells using the siRNA Transfection Reagent ( Santa Cruz Biotechnology ) according to the manufacturer's instructions . 24 h later , transfected HeLa cells with the specific siRNA or siRNA-A control were infected with B . abortus as described before . To maintain the silencing of these specific proteins , HeLa cells were again transfected 2 h p . i . Cellular extracts were prepared after lysis of HeLa cells transfected 72 h with either siGAPDH or siRNA control with 0 . 1% ( vol/vol ) Triton X-100 in PBS . At 48 h p . i . , 6 infected BHK-21 55 cm2 dishes were washed once with GMEM . 1 ml of GMEM was added and infected cells were recovered by scraping with a rubber policeman . Then , several steps of washes were done with PBS , PBS/1 mM EDTA , pH 7 . 4 , homogenization buffer ( 3 mM imidazole/250 mM sucrose/0 . 5 mM EDTA/0 . 5 mM EGTA , pH 7 . 4 ) . A centrifugation at 80 g for 5 min at 4°C was performed between each wash . Pellets were resuspended very gently in the homogenization buffer and cells were mechanically broken through 5 passages into a 22G needle . PNS were recovered after a centrifugation at 80 g for 10 min at 4°C . Then , a first step of purification was performed by loading the PNS on the top of a 50%–12% sucrose gradient . After centrifugation at 800 g for 45 min at 4°C , a cloudy layer containing BCVs appeared at the 50%–12% interface . This cloudy layer ( called interface fraction ) was carefully recovered and loaded at the bottom of a SW 60 centrifugation tube . Three layers of sucrose were sequentially added on top of the interface fraction: 30% , 20% and 5% sucrose respectively . The enriched BCV fraction was obtained after an ultracentrifugation at 35000 rpm for 1 h at 4°C and was localized in the pellet . The pellet was then resuspended in the homogenization buffer . Depending of experiments , a supplementary step was added between the two steps of sucrose gradient in order to eliminate mitochondria from the enriched BCV fraction . This step consisted of incubating the interface fraction with dynabeads ( M-500 subcellular , Invitrogen ) coated with a rabbit anti-VDAC 1 antibody ( according to the manufacturer's instructions ) overnight recovering the supernatant from dynabeads retained with a magnet ( according to the manufacturer's instructions ) . BCV staining within PNS was adapted from what was previously described [40] . PNS was incubated with a rabbit anti-calnexin antibody and then incubated with a rabbit-PE antibody 30 min on ice . The preparation was fixed for 20 min in 3% final PFA and then diluted to 1% final PFA before analysis on a FACScalibur cytometer ( Becton Dickinson ) . Data were analysed using FlowJo software ( Tree Star ) . 250 µg of the enriched fraction of BCVs was treated for 30 min at room temperature with 0 . 1% Triton X-100 . The BCV membranes were separated from bacteria by centrifugation at 10000 rpm for 5 min at 4°C . The corresponding supernatants enriched in BCV membrane proteins were precipitated with trichloroacetic acid ( Sigma Aldrich ) 10% final for 5 min on ice and then with trichloroacetic acid 5% final for 5 min on ice . Trichloroacetic acid was then removed by three washes with 90% acetone . Each step was followed by a centrifugation at 10000 rpm for 3 min at 4°C . The pellet was finally resuspended in 400 µl of destreack rehydration buffer containing 2% carrier ampholytes pH 3–10 ( IPG Amersham Biosciences ) . BCV membrane proteins were separated by isoelectric focusing ( IEF ) . The IEF was performed using 18 cm gels with an immobilized linear pH gradient of 3–10 ( Immobiline DryStrips , Amersham Biosciences ) in IPGphor strip holders ( Amersham Biosciences ) on a MultiphorII machine ( Amersham Biosciences ) . The IEF protocol was as follows: 300 V for 1 min; 500 V gradient for 30 min; 3500 V gradient for 1 . 5 h; 3500 V for 6 h . Temperature was set at 20°C . Prior to SDS PAGE , IPG strips were equilibrated during 20 min in an equilibration buffer ( 6 M urea , Tris , pH 8 . 8 , 50 mM , 2% SDS , 65 mM DTT , 38 . 5% glycerol ) . The second dimension was performed using a Protean II xl Multicell separation unit ( Biorad ) and home-made 10% SDS PAGE gels . Temperature was set at 20°C . Gels , made of Tris-HCl , 0 . 1% SDS and 10% acrylamide were run at 20°C using the following running buffer ( 25 mM Tris , 192 mM glycine and 0 . 1% SDS ) for the cathode part and 2×running buffer for the anode . Electrophoresis was conducted at 10 mA per gel overnight and stopped when the bromophenol blue front dye reached the bottom of the gel . Proteins were stained by a PlusOne Silver Staining Kit , Protein ( GE Healthcare ) according to the manufacturer's instructions ( without glutaraldehyde to allow the mass spectrometry analysis ) . Protein spots immediately excised from silver-stained gels were destained and subjected to in-gel digestion with trypsin ( Sequencing grade modified porcine trypsine; Promega , Madison , WI , USA ) according to a modified protocol from Shevchenko et al . [42] . Tryptic peptides were then extracted from the gel by successive treatment with 5% formic acid and 60% acetonitrile/5% formic acid . Extracts were pooled and dried in a Speedvac evaporator . Peptides resuspended in an α cyano-4-hydroxycinnamic acid matrix solution ( prepared by diluting 6 times a saturated solution in 50% acetonitrile/0 . 3% trifluoroacetic acid ) , were then spotted on the metal target . Mass analyses were performed on a MALDI-TOF Bruker Ultraflex spectrometer ( Bruker Daltonics , Wissembourg , France ) . Mass spectra were internally calibrated using autolytic peptides from trypsin . The peptide mass lists were used to identify the protein using Mascot software available on site . Criteria used for protein identification are given by Mascot as a Probability Based Mowse Score . Ions score is −10*Log ( P ) , where P is the probability that the observed match is a random event . Protein scores greater than X ( X is a number between 60 and 74 , from one search to the other ) are significant ( p<0 . 05 ) . The protein concentration was determined with the BCA™ Protein Assay Kit ( Pierce ) . Volumes corresponding to 60 µg of proteins from each main step of the fractionation were resuspended in 1×laemmeli-buffer and loaded on a 12% SDS polyacrylamide gel . Then proteins were transferred onto a PVDF membrane using a semi-dry transfer . The PVDF membrane was then blocked for 1 h with 4% milk/PBS/0 . 1% Tween 20 and incubated with different antibodies . The PVDF membrane was washed 3 times with PBS/0 . 1% Tween 20 before incubation with the secondary antibody and the detection . The detection was carried out using the ECL™ western blotting detection kit ( Amersham ) . To analyse the PNS and the sucrose step gradient enriched in BCVs , 10 µl of sample were put in 24-well plates containing 12-mm glass coverslips pre-treated with poly-L-lysine and incubated for 15 min at 37°C to allow the adherence onto glass coverslips . Then BCVs or infected cells were fixed with 3% paraformaldehyde , pH 7 . 4 at room temperature for 15 min , and then processed for immunofluorescence staining as described previously [5] . Specimens were observed on a Zeiss LSM 510 laser scanning confocal microscope for image acquisition . Images of 1024×1024 pixels were acquired and assembled using Adobe Photoshop CS2 . B . abortus-infected BHK cells were fixed for 1 h at room temperature with 2 . 5% glutaraldehyde ( Sigma , St Louis , MO , USA ) in 0 . 1 M cacodylate buffer , pH 7 . 2 , containing 0 . 1 M sucrose , 5 mM CaCl2 and 5 mM MgCl2 . After two successive 15 min washes with the same buffer , cells were postfixed for 1 h at room temperature with 1% osmium tetroxide ( Electron Microscopy Sciences , Hatfield , PA , USA ) in the same buffer devoid of sucrose . The cells were scraped off the culture dishes with a rubber policeman and concentrated in 2% agarose in the same buffer . After 1 h incubation at room temperature with 1% uranyl acetate in veronal buffer , the samples were dehydrated in a graded series of acetone and embedded in Epon resin . Thin sections were stained with uranyl acetate and lead citrate . The PNS and enriched BCVs obtained from the BHK-infected cells were first prefixed for 20 min at room temperature with 5% glutaraldehyde in cacodylate buffer diluted at a 1∶1 volume ratio with the PNS or BCV fractions . These fractions were then processed as described above for BHK-infected cell . | A key determinant for intracellular pathogenic bacteria to ensure their virulence within host cells is their ability to bypass the endocytic pathway and to reach a safe replication niche . Brucella bacteria reach the endoplasmic reticulum ( ER ) to create their replicating niche called the Brucella-containing vacuole ( BCV ) . The ER is a suitable strategic place for pathogenic Brucella . Bacteria can be hidden from host cell defences to persist within the host , and can take advantage of the membrane reservoir delivered by the ER to replicate . Interactions between BCV and the ER lead to the presence of ER proteins on the BCV membrane . Currently , no other proteins ( eukaryotic or prokaryotic ) have yet been associated with the BCV membrane . Here we show that non-ER related proteins are also present on the BCV membrane , in particular , the glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) and the small GTPase Rab 2 known to be located on secretory vesicles that traffic between the ER and the Golgi apparatus . GAPDH and the small GTPase Rab 2 are involved in Brucella replication at late post-infection . Similarly , integrity of secretory vesicle trafficking is also necessary for Brucella replication . Here , we show that recruitment of the two eukaryotic proteins GAPDH and Rab 2 on BCV membranes is necessary for the establishment of the replicative niche by sustaining interactions between the ER and secretory membrane vesicles . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/membranes",
"and",
"sorting",
"biochemistry",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis"
] | 2009 | The Glyceraldehyde-3-Phosphate Dehydrogenase and the Small GTPase Rab 2 Are Crucial for Brucella Replication |
Stem and differentiated cells frequently differ in their response to DNA damage , which can determine tissue sensitivity . By exploiting insight into the spatial arrangement of subdomains within the adult neural subventricular zone ( SVZ ) in vivo , we show distinct responses to ionising radiation ( IR ) between neural stem and progenitor cells . Further , we reveal different DNA damage responses between neonatal and adult neural stem cells ( NSCs ) . Neural progenitors ( transit amplifying cells and neuroblasts ) but not NSCs ( quiescent and activated ) undergo apoptosis after 2 Gy IR . This response is cell type- rather than proliferation-dependent and does not appear to be driven by distinctions in DNA damage induction or repair capacity . Moreover , exposure to 2 Gy IR promotes proliferation arrest and differentiation in the adult SVZ . These 3 responses are ataxia telangiectasia mutated ( ATM ) -dependent and promote quiescent NSC ( qNSC ) activation , which does not occur in the subdomains that lack progenitors . Neuroblasts arising post-IR derive from activated qNSCs rather than irradiated progenitors , minimising damage compounded by replication or mitosis . We propose that rather than conferring sensitive cell death , apoptosis is a form of rapid cell death that serves to remove damaged progenitors and promote qNSC activation . Significantly , analysis of the neonatal ( P5 ) SVZ reveals that although progenitors remain sensitive to apoptosis , they fail to efficiently arrest proliferation . Consequently , their repopulation occurs rapidly from irradiated progenitors rather than via qNSC activation .
The response of stem cells to DNA damage plays a major role in determining the tissue response . The nature of the stem cell response following exposure to ionising radiation ( IR ) can underlie tissue sensitivity or resistance and may also determine sensitivity to radiation-induced carcinogenesis . Since DNA damage is encountered during normal growth and development , the response of stem cells and their immediate progenitors to exogenous damage can also provide insight into the regulation of tissue homeostasis . Here , we focus on the DNA damage response ( DDR ) to IR , which is of significance in evaluating the impact of radiotherapy and can influence the ageing characteristics of tissues and risk of accumulating mutations [1] . Remarkably , our knowledge of the response of stem cells to IR is fragmentary , although increasing evidence indicates that stem cells display distinct responses to somatic cells [2] . Further , recent studies have exposed , at least in certain tissues , the distinct responses of quiescent stem cells versus replicating progenitor cells [2] . The haematopoietic system is one of the best studied stem cell models , in part because haematopoietic stem cells ( HSCs ) and their immediate progenitors can be efficiently distinguished by fluorescence-activated cell sorting ( FACS ) analysis using cell-surface–specific markers . Interestingly , quiescent HSCs ( qHSCs ) are endowed with protective mechanisms to restrict DNA damage ( e . g . , a low metabolic rate and expression of ATP-binding cassette [ABC] transporter activities ) [3] . Nevertheless , after IR , they inevitably incur DNA damage . Strikingly , qHSCs are resistant to apoptosis after IR , in contrast to their multipotent progenitors ( MPPs ) and common myeloid progenitors ( CMPs ) , and have been reported to reenter the cell cycle to replenish the depleted progenitors [4–6] . The resistance of qHSCs to apoptosis and cell cycle reentry is p21-dependent but p53-independent , whilst apoptosis of MPPs and CMPs is p53-dependent and p21-independent [7] . Importantly , the repeated activation of HSCs out of dormancy has been reported to cause HSC attrition during ageing [6] . Quiescent mammary stem cells ( MaSCs ) , crypt stem cells , and bulge stem cells ( BSCs ) of the hair follicle similarly display resistance to apoptosis in contrast to their respective transit amplifying progenitors ( TAPs ) , which sensitively activate apoptosis [7 , 8] . Quiescent cells rely on DNA nonhomologous end-joining ( NHEJ ) , which is perceived to be error-prone [5] . However , hair follicle BSCs have been proposed to up-regulate NHEJ as a mechanism to enhance recovery to IR [8] . Additionally , to help maintain genomic stability , it has been proposed that HSCs can differentiate if they harbour excess DNA damage [9 , 10] . Here , we focus on neural stem cells ( NSCs ) located in the subventricular zone ( SVZ ) , the largest germinal region in the adult forebrain , which provides neurogenesis into adulthood [11] . Adult NSCs , designated type B cells , have an astrocytic phenotype and express Glial Fibrillary Acidic Protein ( GFAP ) [11] . They are relatively quiescent , but around 10% are activated to proliferate . They undergo symmetric division to generate further NSCs as well as asymmetric division to generate type C TAPs , which express mammalian Achaete-scute homologue 1 ( Mash1 ) . TAPs also undergo symmetric division to expand in number and asymmetric division to generate doublecortin+ ( Dcx+ ) neuroblasts ( NBs ) ( type A cells ) ( Fig 1A ) . NBs migrate through the rostral migratory stream to the olfactory bulb and terminally differentiate into subpopulations of mature interneurons ( S1A Fig ) . TAPs divide 3 times , and NBs divide once or twice [12] . Several studies have examined the response of these distinct cell types to DNA damage . Exposure to the antimitotic agent , cytosine-β-d-arabinofuranoside , resulted in the elimination of proliferating ( type A and C ) progenitors , but type B cells remained and promoted repopulation [13] . In the rat brain , 5 Gy IR induced apoptosis in the subependymal layer , representing cells adjacent to the lateral ventricle , with subsequent repopulation indicating a radiation-resistant stem cell subset [14] . Using a FACS approach , a more recent study identified and quantified quiescent NSCs ( qNSCs ) and proliferating progenitors and observed loss of proliferating progenitors with subsequent exit of qNSCs from dormancy after IR exposure [15] . Collectively , these studies provide evidence that the SVZ may respond to DNA damage in a similar manner to that observed in the HSC compartment , although details of which cells die by apoptosis , the underlying mechanism , and the coordination with proliferation in vivo is lacking . The adult SVZ has been shown to be spatially heterogeneous and can be divided along the dorso-ventral axis into subdomains designated as dorsal , medial , ventral , and dorsolateral [16–18] ( Fig 1B ) . Each subdomain has a distinct embryonic origin and gives rise to different olfactory bulb neuronal subtypes ( S1A Fig ) . Additionally , the subdomains differ in their proliferative/quiescent status . Most studies to date examining the response to IR in the SVZ have not considered the topographical division into different subdomains . The subdomain organization , however , could be informative if they differ in the degree of proliferation . A further aspect affecting NSC proliferative status is postnatal age . Multiple NBs are generated in newborn mice , but this rapidly diminishes with progression into adulthood , which is accompanied by an increased level of qNSCs [19] . Significantly , young mice and people are sensitive to IR-induced neuronal carcinogenesis in contrast to the resistance of adults [20–22] . Whether neonatal and adult mice differ in their response to IR has not been examined yet might be significant in evaluating the specific IR-induced cancer predisposition of the juvenile brain . In the first part of our study , we examine the response to IR within the adult SVZ , taking advantage of knowledge of the subdomain structure . Importantly , our analysis is undertaken in vivo allowing us to assess the response in each subdomain under the in vivo proliferation state . We then exploit this insight to evaluate distinctions in the response between adult and neonatal mice . This approach proved to be insightful in assessing the distinct responses of NSCs , TAPs , and NBs . We demonstrate that progenitors ( TAPs and NBs ) but not NSCs ( quiescent and activated ) undergo apoptosis and characterize 2 novel responses , namely proliferation arrest and progenitor marker loss . We show that these responses are ataxia telangiectasia mutated ( ATM ) -dependent and drive qNSC activation , which replenishes the pool of NBs . The ability to activate apoptosis is cell type- rather than proliferation-dependent and does not appear to be driven by distinctions in DNA damage induction or repair capacity . We propose that the benefit of apoptosis is not to confer sensitive cell death but rather to remove progenitor cells to aid repopulation , a beneficial feature since proliferating cells are likely to be the ones more sensitive to cancer-driving genetic changes . Importantly , our analysis of these responses in neonatal mice reveals distinctions to the response in adults , demonstrating that progenitor repopulation occurs more rapidly from irradiated progenitors that fail to arrest proliferation rather than via qNSC activation .
Delineation of the subdomains within the lateral ventricle is shown in Fig 1B . First , we assessed the percentage of NSCs ( type B cells; GFAP+ ) , TAPs ( type C cells; Mash1+ ) , and NBs ( type A cells; Dcx+ ) in the adult SVZ ( defined as 3-month-old mice ) in each subdomain ( Fig 1C ) . We observed high ( >90% ) costaining of GFAP with Sox2 and Nestin in all 4 subdomains . Thus , to aid clarity , we have designated all GFAP+ cells as NSCs , although they may encompass a small number of astrocytes . Hereafter the term neural progenitors will be used to represent TAPs and NBs . Cell numbers are expressed as the percentage of 4′ , 6-diamidino-2-phenylindole ( DAPI ) + cells within each subdomain . GFAP+ cells are located in all 4 subdomains , with higher levels in the medial and dorsal regions ( Fig 1C ) . Consistent with the more quiescent nature of the medial and dorsal domains , there were very low ( or undetectable ) numbers of neural progenitors ( Mash1+ TAPs and Dcx+ NBs ) in these 2 subdomains; indeed , the progenitors were predominantly restricted to the ventral and dorsolateral domains ( Fig 1C and 1D ) . To examine proliferation within these subdomains , we assessed the distribution of the proliferation marker , Ki67 . Consistent with the proliferative nature of the progenitors ( TAPs and NBs ) , Ki67+ cells were also predominantly localized to the ventral and dorsolateral domains , with 20%–30% of the cells in these 2 subdomains being Ki67+ ( Fig 1C and 1D ) . At 6 h following exposure to 2 Gy , apoptotic cells ( assessed by terminal deoxynucleotidyl transferase-mediated dUTP nick end-labelling [TUNEL] staining ) were observed in the lateral ventricle ( Fig 1E ) with little or no apoptosis in the surrounding differentiated tissue ( S1B Fig ) . Strikingly , the distribution of apoptotic cells within the subdomains revealed that apoptosis occurred almost uniquely in the ventral/dorsolateral regions where the proliferating progenitors were located ( Fig 1F ) . Collectively , these findings based on spatial analysis of the subdomains demonstrate that the GFAP+ cells , which are abundant in the medial/dorsal regions , are resistant to apoptosis . Additionally , the findings provide strong evidence that it is the neural progenitors that undergo apoptosis . These findings are consistent with studies on the haematopoietic system and provide initial evidence that it is specifically the neural progenitor cells ( TAPs and/or NBs ) that succumb to apoptosis [7] . To further evaluate the nature of cells undergoing apoptosis , we assessed the percentage of each cell type expressing Ki67 in the ventral or dorsolateral regions . Consistent with the quiescent nature of NSCs , only approximately 5%–10% of GFAP+ NSCs were Ki67+ ( Fig 2A ) . Additionally , consistent with cell lineage analysis , approximately 80%–90% of the Mash1+ TAPs and 50% of Dcx+ NBs expressed Ki67 , respectively ( Fig 2A ) . From this quantification and that in Fig 1C , the distribution of Ki67+ cells in the ventral and dorsolateral regions showed that Dcx+ NBs represent the predominant progenitor cell type in the ventral and dorsolateral domains ( S1C Fig ) . Assessment of the percentage of each cell type undergoing apoptosis revealed that approximately 50%–60% of Mash1+ or Dcx+ cells were TUNEL+ at 6 h post IR ( Fig 2B ) . There were very few GFAP+TUNEL+ cells , supporting the notion that qNSCs are resistant to apoptosis ( Fig 2B ) . As expected from the cell distribution , the predominant cell type undergoing apoptosis were Dcx+ NBs , which represented 60%–70% of the apoptotic cells at 6 h post IR ( S1D Fig ) . These findings demonstrate that both progenitor cell types ( TAPs and NBs ) , which are located predominantly in the dorsolateral and ventral subdomains , undergo apoptosis after 2 Gy IR . In these domains , we observed that approximately 50% of Dcx+ NBs are Ki67+ , and approximately 50% Dcx+ NBs undergo apoptosis after 2 Gy . Thus , after 2 Gy , either 100% of the Dcx+Ki67+ NBs ( i . e . , actively proliferating Dcx+ NBs ) succumb to apoptosis , or 50% of all Dcx+ NBs undergo apoptosis irrespective of proliferative status . Thus , we sought to address the relationship between apoptotic sensitivity and proliferative status . We did not wish to use the Ki67 marker to assess this relationship since , as confirmed below , we predicted that this marker might be affected by IR exposure at 6 h post IR ( when apoptosis is quantified ) . To address whether all progenitors can become apoptotic or only those that are actively proliferating , we prelabelled proliferating cells with BrdU at 4 h and 1 h prior to exposure to 2 Gy IR and at 6 h post IR assessed the level of apoptotic BrdU+ cells out of the total BrdU+ cells ( Fig 2C ) . For technical reasons , apoptosis was assessed using cleaved Caspase-3 ( Casp3 ) staining for this analysis , but control experiments verified that a similar level of apoptosis was detected by TUNEL or Casp3 staining ( S2A Fig ) . Significantly , at 6 h post IR , all BrdU+ cells were localized to the ventral/dorsolateral domains but only 50%–60% ( depending on the precise subdomain ) were Casp3+ ( Fig 2D and 2E ) . This demonstrates that 50% of the progenitors , despite being BrdU labelled immediately prior to IR ( i . e . , actively proliferating ) , do not undergo apoptosis . This strongly suggests that 50% of progenitors undergo apoptosis after 2 Gy regardless of their proliferative status . ( Note also that we observed that approximately 30% of the Casp3+ cells were BrdU+ , which was consistent with the conclusion above but is less informative because the BrdU labelling may not detect all proliferating cells [S2B Fig] ) . This conclusion is also consistent with the finding that , although approximately 80% of Mash1+ cells are Ki67+ , only 50% are TUNEL+ ( Fig 2A and 2B ) . This analysis provides strong evidence that the sensitivity to apoptosis is a feature of the progenitor cell type ( Mash1+ or Dcx+ ) rather than being a direct consequence of proliferation per se . Nonetheless it is the neural progenitors , the proliferative cells in the SVZ , which are apoptotic-sensitive . To gain insight into the downstream consequences of IR exposure , we examined the number of Dcx+ and Ki67+ cells at longer times post IR in the ventral and dorsolateral domains . For this analysis , we focused on Dcx+ cells since Mash1+ TAPs represent a small subset ( approximately 6% ) of progenitors . First , we will discuss the events taking place up to 48 h post IR . Analysis of TUNEL+ cells showed that apoptosis peaked at 6 h post 2 Gy X-rays with no detectable apoptosis at 48 h ( Fig 3A ) . Strikingly , the number of Ki67+ cells in the ventral and dorsolateral regions was dramatically diminished at 6 h and 48 h post IR ( Fig 3B and 3C ) . Since apoptotic cells are detectable at 6 h post IR ( and 50% of the Dcx+ cells at 6 h are TUNEL+ ) , this strongly suggests that the Ki67 marker is lost following IR exposure in a manner that cannot be attributed to cell loss by apoptosis . Somewhat distinct to these findings , we found that the percentage of Dcx+ cells was less dramatically decreased at 6 h but by 48 h was very low ( Fig 3B and 3C ) . Notably , in the dorsolateral domain , although >50% of the Dcx+ cells retain the NB marker at 6 h post IR , very few display the Ki67 marker demonstrating the distinct kinetics between these 2 responses . Additionally , although only 50% of Dcx+ cells appear to undergo apoptosis , by 48 h , very few express the Dcx+ marker . This strongly suggests that there is a striking loss of progenitor cells expressing the Dcx marker between 6 h and 48 h post IR . Significantly , we observed only a 20%–25% decrease in overall DAPI-staining cells and a similar decrease in subdomain thickness at 48 h post IR , which is consistent with the fraction of total cells that undergo apoptosis; furthermore , these changes are only observed in the subdomains that undergo apoptosis ( S2C and S2D Fig ) . Since NBs represent approximately 50% of the total cells in the ventral and dorsolateral regions ( Fig 1C ) , this supports the notion that they do not all undergo apoptosis but , as a distinct response , cease proliferation and lose their defining marker . Ki67 is an indirect cell surface marker of proliferation . We also examined proliferation using ethynyl deoxyuridine ( EdU ) incorporation as a more direct monitor of replicating cells . Using the approach described in Fig 3D , we labelled cells replicating prior to IR exposure using BrdU ( added at 4 h and 1 h prior to IR ) and monitored ongoing replication from 5 h to 6 h post IR by EdU incorporation . We assessed the fraction of BrdU+ or EdU+ cells with or without IR exposure . These findings confirmed a dramatic arrest in replication ( EdU+ cells ) between 5 h to 6 h post IR exposure ( Fig 3E and 3F ) . We also observed an approximately 2-fold loss of BrdU+ cells by 6 h post IR , which may be a consequence of the onset of DNA degradation during apoptosis and/or cell division and dilution of the BrdU+ signal in the unirradiated control cells . Additionally , to provide further evidence for the distinct nature of apoptosis and proliferation arrest , we carried out a dose-response analysis monitoring the level of TUNEL+ and Ki67+ cells reasoning that they may show a distinct dose-response relationship . Indeed , while we observed a linear dose response for apoptosis up to 3 Gy , we observed a loss of Ki67+ cells even after exposure to 1 Gy ( Fig 3G–3I ) . Thus , we conclude that IR exposure to ≥1 Gy causes arrest of proliferation in a nonapoptotic-dependent manner . Collectively , these findings identify 2 novel responses to IR exposure in the SVZ , namely the rapid inhibition of proliferation followed by a loss of cells bearing the Dcx marker . We next aimed to gain insight into the basis underlying the loss of the NB ( Dcx ) marker . The number of DAPI cells in the SVZ only diminished by 20%–25% at 48 h , a level entirely attributable to apoptosis . If loss of the Dcx marker represents elimination of immature neurons ( e . g . , by movement out of the SVZ ) , it must be compensated by movement of cells into the SVZ . We therefore examined whether the number of microglia , the most likely candidate cell type moving into the irradiated SVZ , increases after IR ( S3 Fig ) . Following staining for the microglial marker , Iba1 , we did not observe any significant increase in either the ventral or dorsolateral subdomains , although there was a small but nonsignificant increase when the entire lateral ventricle was analysed ( S3A–S3C Fig ) . There was no difference in the percentage of Iba1+ cells in the SVZ or the differentiated , nonapoptotic isocortex ( S3D and S3E Fig ) . Next , we examined whether the Dcx marker loss represents differentiation of NBs into more mature neurons within the SVZ and looked at the expression of early ( neuron-specific class III β-tubulin [Tuj1] ) and late ( microtubule-associated protein 2 [MAP2] ) neuronal differentiation markers in the ventral and dorsolateral regions at 48 h post IR ( Fig 4A and 4B ) . As expected , the percentage of Tuj1+ cells decreased after 2 Gy compared with control ( Fig 4A–4C ) due to its partial expression in NBs , which undergo apoptosis after IR ( Fig 2B ) . The decrease in Tuj1+ cells was , however , less than that observed for the Dcx marker ( Fig 3B and 3C ) , suggesting the possibility that early committed neurons are not lost . The overall percentage of mature MAP2+ neuronal cells after IR was slightly greater than control , although the difference was not statistically significant ( Fig 4B ) . However , this may be a result of the relatively low number of MAP2+ cells arising from differentiation compared to the number already present in the SVZ . To further examine whether NBs can more rapidly commit to a neuronal fate following IR , we injected BrdU at 4 h and 1 h before IR with 2 Gy and waited 48 h before collection ( Fig 4D ) . The BrdU-labelling analysis , when compared to the data obtained at 6 h ( Fig 3E ) , revealed a decrease in the percentage of BrdU+ cells in the ventral subdomain but not the dorsolateral subdomain , consistent with migration of the BrdU+ cells towards the rostral migratory stream ( Fig 4E ) . In fact , the percentage decrease in BrdU+ cells after IR could predominantly be attributed to loss by apoptosis ( approximately 50% ) . Since the majority of BrdU+ cells are Dcx+ in unirradiated mice , this raises the possibility that BrdU+ cells are retained in the SVZ and undergo differentiation . To examine this possibility , we co-stained the BrdU labelled cells with Dcx , Tuj1 , and MAP2 ( Fig 4F ) . In control mice , the majority of BrdU+ cells had a NB phenotype , expressing Dcx and Tuj1 , and virtually none of the BrdU+ cells co-expressed the more mature MAP2 neuronal marker . Importantly , after IR , a higher percentage of BrdU+ cells expressed MAP2 in both the ventral and dorsolateral subdomains , indicating cell cycle exit and differentiation into more mature neurons ( Fig 4F and 4G ) . Collectively , these findings suggest that loss of the Dcx marker can be attributed to the differentiation of Dcx+ cells into mature ( MAP2+ ) neurons within the SVZ . To assess the longer-term consequences of apoptosis and proliferative arrest , we extended our analysis to 7 and 14 days post IR . First , in the ventral and dorsolateral subdomains , Ki67+ cells , which , as described above , were almost undetectable at 6 h and 48 h post IR , increased to levels slightly greater than or similar to that of unirradiated control animals by 7 days ( Fig 5A ) . Thus , either there is a transient ( approximately 48 h ) arrest of proliferation or a defined mechanism to repopulate the pool of proliferating progenitors . The percentage of GFAP+ cells did not change significantly over this time period , although we observed a modest increase at 48 h and 7 days ( Fig 5B ) . Strikingly , GFAP+Ki67+ ( i . e . , activated stem cells ) , which are low in number in control cells , showed a slight increase at 48 h and a >4-fold increase above the control level at 7 days . This was followed by a return to the level seen in unirradiated animals by 14 days ( Fig 5C ) . This demonstrates that exposure to 2 Gy causes marked qNSC activation . Analysis of Dcx+ NBs revealed a slower rate of recovery with only half the Dcx+ cell numbers being detectable at 7 days and a further increase by 14 days ( Fig 5D ) . We focused our detailed analysis above on the ventral and dorsolateral subdomains since only these domains contain neural progenitors . Examination of the dorsal and medial regions showed no evidence for NSC activation at 7 days post IR ( Fig 5E and 5F ) , demonstrating that there is no cross talk between the subdomains and that exposure to 2 Gy cannot cause qNSC activation in the medial or dorsal domains . Detailed temporal analysis of the dorsal and medial subdomains for additional markers is shown in S4A–S4D Fig . These findings reveal that the return of GFAP+Ki67+ cells precedes the replenishment of Dcx+ cells , suggesting that NBs are replenished by activation of qNSCs . This suggestion is further supported by intervening temporal analysis of mice at 3 and 5 days post IR , showing that the increase in GFAP+Ki67+ cells clearly precedes the return of Dcx+ cells ( S4C and S4D Fig ) . Additionally , at 7 days , a larger fraction of Dcx+ cells are Ki67+ , which is consistent with their recovery occurring via stem cell activation ( Fig 5G and 5H ) . Thus , we propose that the major route of Dcx recovery after 2 Gy IR is via qNSC activation . IR causes double-strand break ( DSB ) formation and activates ATM-dependent DNA damage-response signalling [23] . However , in cycling cells , ataxia-telangiectasia-related ( ATR ) can also be activated during replication . In previous studies , we have observed that after 500 mGy , apoptosis in the embryonic ventricular zone ( VZ ) /SVZ is approximately 70% ATM-dependent , suggesting that ATM is the major activating kinase , but another kinase , most likely ATR , can also contribute [24] . To assess the ATM dependency of the DDR responses in the adult SVZ , we examined apoptosis , proliferation arrest , and Dcx marker loss following exposure to 2 Gy in Atm-/- mice . First , we examined the distribution of Ki67+ cells in the SVZ subdomains and found that , as for control mice , Ki67+ cells were predominantly located in the ventral and dorsolateral domains ( Fig 6A ) . The domains also showed a similar distribution of NSCs ( GFAP+ ) , TAPs ( Mash1+ ) , and NBs ( Dcx+ ) to that of the wild-type ( WT ) SVZ ( Fig 6B ) . Significantly , we failed to observe either apoptosis or proliferation arrest at 6 h in Atm-/- mice ( Fig 6C–6E ) . There was a small reduction of Dcx+ cells at 6 h ( similar to that observed in WT mice ) , but , by 48 h , more Dcx+ cells remained than observed in WT mice ( although slightly lower than at 6 h ) ( Fig 6F and 6G ) . Collectively , we conclude that the 3 described DDR responses ( apoptosis , Ki67 loss , and differentiation ) to 2 Gy IR are ATM-dependent . Further , we did not observe any increase in GFAP+Ki67+ cells in Atm-/- mice at 48 h or 7 days post 2 Gy ( S5 Fig ) . These findings are consistent with the notion that a failure to activate the DDR responses precludes qNSC activation . However , despite failing to activate cell death by apoptosis , Atm-/- cells lack many responses to DSBs and are markedly radiosensitive , undergoing cell death by other processes , such as necrosis , senescence , or ATR-dependent permanent checkpoint arrest . These alternative forms of cell death may also contribute to a failure to activate Atm-/- qNSCs . The analysis of Mash1+ and Dcx+ cells demonstrates that sensitivity to apoptosis after 2 Gy is a feature of these neural progenitors rather than being a direct consequence of their proliferative status ( Fig 2C and 2D ) . We , therefore , sought to investigate whether , conversely , the resistance of GFAP+ cells to apoptosis is a consequence of their quiescent status . To address this , we examined whether activated GFAP+Ki67+ cells undergo apoptosis . It was difficult to address this after a single dose of 2 Gy because the number of GFAP+Ki67+ cells is very low ( approximately 10% of GFAP+ cells ) , and the Ki67 marker is not expressed at 6 h after 2 Gy , the optimal time for assessing apoptosis . We , therefore , took an alternative approach to address this question by exploiting the marked increase in activated NSCs ( i . e . , GFAP+Ki67+ ) at 7 days post 2 Gy IR . Three-month-old mice were exposed to 2 Gy IR , then 7 days later exposed to a second dose of 2 Gy , and GFAP+TUNEL+ cells quantified 6 h later ( Fig 7A ) . The magnitude of apoptosis was compared to that observed in NSCs at 6 h post a single dose of 2 Gy ( Fig 7B ) . Strikingly , despite a 4-fold increase in activated GFAP+ cells at 7 days post 2 Gy ( Fig 5C ) , the second exposure to IR did not promote any increase in apoptosis amongst GFAP+ cells in the ventral or dorsolateral domain . This experiment provides strong evidence that activated NSCs , like their quiescent counterpart , are resistant to IR-induced apoptosis , consolidating the notion that sensitivity ( or resistance ) to apoptosis is a feature of the cell type rather than an indirect consequence of proliferation . It has also been proposed that hair follicle BSCs have an upregulated capacity to repair DSBs , which is causally related to their resistance to apoptosis [8] . To gain insight into the basis underlying the resistance of NSCs and sensitivity of the progenitors to apoptosis , we examined whether NSCs in the SVZ incur the same DSB induction level as Dcx+ NBs and whether they display any changes in DSB repair capacity . To assess this , adult mice were exposed to 100 mGy or 2 Gy and p53-binding protein 1 ( 53BP1 ) foci , a marker of DSB formation , quantified 0 . 5 h and 6 h later in GFAP+ and Dcx+ cells . A dose of 100 mGy was used since this induces <1% apoptosis in progenitors , allowing foci numbers to be enumerated in predominantly nonapoptotic cells . Although 100 mGy only induces 1 to 2 foci/cell , foci numbers could be accurately assessed by scoring a large area and multiple tissue sections . Repair was not assessed in Dcx+ cells after 2 Gy due to presence of apoptotic cells . We observed similar levels of induced DSBs in Dcx+ and GFAP+ cells ( Fig 7C and 7D ) . The slightly greater levels in the Dcx+ cells after 100 mGy is not significant ( P = 0 . 075 ) and insufficient to account for the marked distinction in apoptotic sensitivity and was not observed after 2 Gy . Assessment of foci numbers at 6 h post 100 mGy showed a similar level of residual DSBs in Dcx+ and GFAP+ cells , suggesting a similar DSB repair capacity , at least over this time scale ( Fig 7C and 7D ) . Finally , we also examined whether NSCs undergo senescence rather than apoptosis using β-galactosidase ( SA-β-gal ) as a marker of senescent cells . Whilst we were able to detect the senescence marker , SA-β-gal , in the kidney of aged 21-month-old mice , we did not detect any SA-β-gal+ cells in any of the SVZ subdomains at 7 days post 2 Gy ( Fig 7E ) . Thus , although limited by the restrictions of in vivo analysis , we have not found any evidence that the resistance of NSCs to apoptosis is a direct consequence of a protective environment , quiescence , or an enhanced DSB repair capacity . Additionally , we could not detect any senescence at 7 days post 2 Gy in the SVZ subdomains . The embryonic VZ/SVZ represents the predominant forebrain region in the embryo and undergoes rapid proliferation from E11 . 5 to E16 . 5 . Although the size of the neonatal SVZ and the extent of proliferation diminishes post birth , both are greater at P5 than observed in the adult SVZ [17] . Slowly dividing qNSCs increase in level and adopt an astrocytic morphology during the early neonatal stage with the extent of proliferation and the number of progenitors decreasing [17] . Whether the distinct subdomains differ in the rate at which these changes arise has not been examined . First , we assessed the percentage of Ki67+ and GFAP+ cells in the distinct subdomains of P5 mice ( Fig 8A and 8C ) . Since the SVZ is larger in neonatal ( P5 ) mice compared to 3-month-old mice , we expressed the results as a percentage of cells within each domain . These findings reveal , as expected , that there is an increased level of proliferative cells in all subdomains of P5 mice compared to adult mice; the increase is approximately 2-fold in the ventral and dorsolateral regions , in which there are detectable Ki67+ cells in the adult; a more dramatic increase is observed in the medial and dorsal regions , in which few proliferative cells are observed in the adult ( Fig 8A ) . The number of GFAP+ cells in each subdomain showed the converse picture , that is , a lower number of GFAP+ cells ( Fig 8B ) . Quantification of apoptosis revealed approximately 5-fold increased apoptosis in the entire SVZ , most likely a consequence of the increased size of the SVZ as well as the increased level of apoptotic-sensitive progenitor cells ( Fig 8D ) . However , when assessed as the percentage of cells in each subdomain , a similar level of apoptosis to that found in the adult was observed in the ventral and dorsolateral regions . Since the number of Ki67+ cells was slightly enhanced in these 2 subdomains , this suggests that there is a similar or possibly slightly reduced sensitivity for each progenitor cell to undergo apoptosis at P5 ( Fig 8E and 8F ) . Additionally , apoptosis was enhanced in the dorsal region and slightly in the medial region due to the presence of proliferating cells at this stage ( Fig 8E ) . Collectively , these findings suggest that although the overall level of apoptosis in the neonatal SVZ is greater than in the adult SVZ , this is largely attributable to the increased size of the neonatal SVZ . When evaluated on a per cell basis , the sensitivity of each progenitor cell to activate apoptosis appears similar to that in the adult SVZ . In adult mice , we observed a marked decrease ( 5- to 6-fold ) in Ki67+ cells in the ventral and dorsolateral subdomains at 6 h post 2 Gy IR , which occurs in most progenitor cells independently of whether they succumb to apoptosis ( Fig 3B and 3C ) . We , therefore , examined whether the Ki67 marker is similarly lost following exposure of P5 mice to 2 Gy . Strikingly , at 6 h post 2 Gy , there was a much smaller reduction ( 2-fold ) in the ventral , dorsolateral , and dorsal subdomains of P5 mice ( Fig 8G ) . The reduction in Ki67+ cells in the medial region was more dramatic , but nonetheless , cells expressing the Ki67 marker were readily detectable . This difference between adult and neonatal mice appears to represent a distinction in the percentage of cells arresting proliferation rather than simply being a consequence of the increased number of Ki67+ cells in the neonatal SVZ . In adult mice , the marked reduction in Ki67+ cells remained at 48 h post IR ( Fig 3B and 3C ) . To assess the response at 48 h in the neonates , it was necessary to estimate the percentage of Ki67+ cells in unirradiated P7 mice ( i . e . , the equivalent age of mice exposed at P5 and assessed 48 h post IR ) since there are dynamic changes in proliferative capacity during this early neonatal period . This analysis revealed that the number of Ki67+ cells in the medial region in unirradiated P7 neonates is substantially lower than in P5 neonates , suggesting that the proliferative capacity of this subdomain diminishes ( i . e . , “ages” ) more rapidly than the other subdomains ( Fig 8G ) . The number of Ki67+ cells was similar in the unirradiated control or irradiated medial subdomain at P7 , suggesting that the decrease with age is not additive with any radiation-induced decrease . In the other subdomains , the percentage of Ki67+ cells was similar between P5 and P7 unirradiated control neonates ( Fig 8G ) . Following IR at P5 , there was a small increase in Ki67+ cells at 48 h compared to the level of 6 h , suggesting that replenishment was already commencing . At P9 , the distribution of Ki67+ cells within the subdomains of unirradiated mice resembled that of adult mice , although the percentage of proliferating cells was nearly 2-fold greater than in the adult mice . At 4 days following exposure of P5 mice ( i . e . , P9 ) , the number of Ki67+ cells in each subdomain was similar to unirradiated P9 mice and greater than at 6 h post IR . Thirty percent of the neonatal SVZ cells undergo apoptosis by 6 h ( Fig 8E ) , suggesting that all the remaining cells ( active NSCs and/or progenitors ) express the Ki67 marker by 48 h ( Fig 8G ) . We consider it unlikely that these cells were recovered by qNSC activation since this takes longer than 48 h in the adult SVZ ( Fig 5C ) . To substantiate this , we quantified the number of GFAP+Ki67+ cells following IR exposure of P5 , P7 , and P9 mice exposed or not to IR at P5 ( Fig 8H ) . At 6 h , the number of GFAP+Ki67+ cells was reduced 2- to 4-fold relative to unirradiated neonatal mice similar to that observed for all Ki67+ cells ( which predominantly reflect progenitor cells and not NSCs ) . By 48 h and 4 days post IR , the level of GFAP+Ki67+ NSCs had increased to that observed in unirradiated neonates , but the rate of increase paralleled the response of all Ki67+ cells . We conclude that the kinetics of proliferating progenitor cell recovery strongly suggests that it occurs from surviving irradiated progenitors rather than via the activation of qNSCs as in adult mice . Collectively , these findings reveal important distinctions between the DDR to 2 Gy IR in the neonatal compared to the adult SVZ . Although the ability of progenitor cells to activate apoptosis in neonates is similar to adults with substantially more apoptosis taking place due to the enhanced size of the SVZ and increased number of progenitors , the neonatal SVZ fails to efficiently arrest proliferation , and the recovery of proliferating progenitors occurs more rapidly . The kinetics of recovery and analysis of activated NSCs , moreover , strongly suggests that in neonatal mice the progenitors repopulate from inefficiently arrested irradiated progenitors rather than via qNSC activation .
In this study , we examine the response to DNA damage in the SVZ exploiting recent understanding of its spatial organization . This revealed distinct responses of the neural stem and progenitor cells and insight into the process leading to qNSC activation . Additionally , a temporal analysis has shown distinctions in the responses of the adult versus the juvenile SVZ to DNA damage , which is of significance given their differing sensitivities to carcinogenesis . First , we show that NSCs are resistant to apoptosis , whilst TAPs and NBs activate apoptosis in a dose-dependent manner . This parallels the response of HSCs and mammary stem cells and their respective progenitors [7] . HSCs and NSCs have distinct turnover rates , suggesting that the apoptotic sensitivity is not unique to more proliferative stem cell compartments [25] . The apoptotic sensitivity of progenitors is not a direct consequence of replication but rather represents a programmed cell type–specific response . We report 2 additional novel responses to exposure to 2 Gy IR: ( 1 ) rapid ( by 6 h ) proliferation arrest , measured by Ki67 and BrdU/EdU incorporation , and ( 2 ) loss ( by 48 h ) of cells bearing the progenitor marker , Dcx . Cell cycle checkpoint arrest is a well-characterized response of cultured cells to IR and has also been observed in tissues [24 , 26] . However , it is unclear if the Ki67 marker loss and lack of EdU incorporation observed in the adult SVZ after 2 Gy represents a consequence of checkpoint arrest or a distinct response . In the SVZ , after 2 Gy , the Ki67 marker loss appears to be sustained for at least 48 h , with recovery arising from activated NSCs rather than from preexisting progenitor cells . In contrast , after 2 Gy G1/S or G2/M checkpoint arrest in cultured primary cells is more transient [27] . We provide evidence that the second response observed , namely loss of Dcx+ progenitors , can be attributed to enhanced differentiation of immature into mature neurons . One study using cultured cells observed that DNA damage can induce astrocytic differentiation in NSCs derived from embryonic stem cells [28] . Additionally , DNA damage has been shown to induce terminal differentiation in HSCs [9 , 10] . Further studies are required to understand the basis underlying this important observation . Importantly , apoptosis , proliferative arrest , and differentiation are ATM-dependent , suggesting that they represent a coordinated response to DNA damage in the adult neural stem cell compartment . These responses appear linked to the activation of qNSCs since NSC activation is only observed in the subdomains where these responses are seen . Further , qNSC activation is not observed in Atm-/- mice . However , we resist causally linking these endpoints since ATM loss , despite overriding cell death by apoptosis , can lead to dramatic radiation sensitivity ( via cell death by other processes ) , rendering Atm-/- qNSCs unable to become activated . Other studies describing the resistance of quiescent stem cells to apoptosis frequently describe them as being radioresistant [29] . However , apoptosis represents a form of cell death that can be sensitively monitored , and the absence of apoptosis should not be taken to assume radioresistance . Cultured primary fibroblasts , for example , do not undergo IR-induced apoptosis , yet they lose proliferative capacity due to lethal chromosome breaks or aberrations [30] . Although there are differences in the precise shape of the survival curves , assessing survival of NBs by the percentage of cells without apoptosis shows a similar survival response to analysis of radiosensitivity of cultured fibroblasts via clonogenic survival assays ( S6A Fig ) . Furthermore , the level of DNA damage incurred and NHEJ repair appears similar between qNSCs and progenitors . Thus , as a working hypothesis , we propose that progenitors undergo apoptosis , a form of cell death that causes rapid cell loss , to promote recovery via quiescent stem cell activation . Two additional responses , namely proliferation arrest and progenitor cell differentiation potentially serve to enhance quiescent stem cell activation . These ATM-regulated responses have the important consequence of changing the balance of signals that confer homeostasis between stem cell quiescence/activation causing activation of qNSCs . Enhancing differentiation may function to diminish progenitor cells , thereby increasing the signal for qNSC activation . Additionally , progression through replication or mitosis in the presence of DNA damage can cause lethality or enhance misrepair . Although cells activate DDR checkpoints , they have limitations and , as evident from the studies on potentially lethal damage repair ( PLDR ) , cycling cells have reduced survival and chromosome rearrangements relative to cells that have exited the cell cycle [27] . Thus , activating responses to restrict the recovery of irradiated progenitors , the cells most vulnerable to IR-induced rearrangements via their proliferative capacity , will have the additional advantage of restricting carcinogenesis . Instead , promoting recovery via the activation of a quiescent population will , as demonstrated by many classic radiobiology studies , enhance survival and , importantly , reduce translocations , despite the employment of NHEJ [30] . In summary , we propose that apoptosis is not so much a form of sensitive cell death but rather a form of cell death that causes rapid cell loss in the germinal centre , thereby promoting quiescent stem cell activation . The processes of proliferation arrest and differentiation represent complementary components of this DDR . The ability of DNA damage to promote exit from dormancy in HSCs has also recently been demonstrated , and , here , we link this to cell loss by apoptosis [6] . Significantly , we do not observe qNSC activation in the medial or dorsal regions , which do not undergo apoptosis . Radiation exposure of the neonatal brain has been shown to diminish cognitive function and to enhance carcinogenesis [20–22 , 31] . Our subdomain analysis substantiates other studies showing that the juvenile SVZ has a larger number of proliferating progenitors than the adult brain and , as a corollary , fewer qNSCs [32] . Although this results in an increased number of apoptotic-sensitive progenitors , their sensitivity appears similar to that of adult progenitors . Consequently , overall there is a greater level of apoptosis in the neonatal compared to the adult SVZ , which likely reflects the enhanced sensitivity of the neonatal brain to IR . Strikingly , however , the neonatal progenitor cells have a diminished ability to undergo proliferative arrest at P5 compared to adult progenitors and , moreover , recover proliferative capacity more readily . Indeed , the marked difference in the kinetics of recovery of proliferation strongly suggests that the irradiated progenitors regain proliferative capacity after a brief arrest rather than repopulation arising as a consequence of quiescent stem cell activation . Whilst exposure of the adult brain to 2 Gy appears to promote the repopulation of NBs from qNSCs , the progenitors are rapidly replenished in neonates via a mechanism that does not require extensive NSC activation ( shown in S6B Fig ) . Thus , the NBs are derived from irradiated proliferating cells in neonates but from activated NSCs in the adult mice , a distinction which may influence the level of rearrangements they contain , and hence their ability to become carcinogenic . These findings also suggest that after 2 Gy IR , the activation of apoptosis is not the sole factor promoting qNSC activation but involves , additionally , proliferative arrest and differentiation . We note also that in this study we have focused on the response to 2 Gy IR . However , such a dose is rarely encountered outside a radiotherapy setting , and it will be important to assess these responses using much lower , more physiological doses . More studies are needed to determine whether similar responses arise in other stem cell compartments . Central to our analysis is the exploitation of recent insight into the subdomain structure of the SVZ around the lateral ventricle [17 , 18 , 33] . Indeed , it is our combined spatial and temporal analysis which has provided novel insight into the DDR in the SVZ . Importantly , in the context of adult neurogenesis and the DDR , our findings show that the distinct subdomains differ in their degree of “ageing” , which is detectable both from the adult and neonatal analysis . The medial region appears to “age” and lose proliferative capacity most rapidly , followed by the dorsal region . The ventral and dorsolateral regions maintain some neurogenic capacity at least to 3 months of age . This feature lies at the basis of the difference in the DDR between the subdomains . A consequence is that the neuronal subtypes derived from each subdomain will differ in their renewal capacity with age and , of relevance here , differ in their response to DNA damage . In summary , by exploiting spatial analysis of subdomains within the SVZ , we show that NSCs are resistant to radiation-induced apoptosis , whilst the progenitors , TAPs and NBs , succumb to apoptosis . This response is not a direct consequence of progenitor cell proliferation but reflects a programmed cell type response . Additionally , in the adult SVZ , the proliferation and progenitor markers are lost within 6 h to 48 h post IR in an ATM-dependent manner . This drives activation of qNSCs , which repopulate the progenitor cell pool by 14 days post exposure ( see S6B Fig for a model ) . Temporal analysis of this response reveals that the progenitor cells in the neonatal SVZ similarly undergo apoptosis but display a milder decrease in the proliferation marker and more rapidly repopulate the progenitor pool . Collectively , the findings suggest that after 2 Gy adult mice replenish progenitor cells by qNSC activation , whilst the progenitor cells in neonatal mice that do not succumb to apoptosis can rapidly resume proliferation . We propose that apoptosis is not so much a sensitive form of cell death but rather represents a form of cell death that encourages quiescent stem cell activation , and that this response is additionally promoted by arrest of progenitor proliferation and marker loss , likely reflecting differentiation .
C57BL/6 and Atm-/- ( 129/SV x C57BL/6 ) mice were used in this study . Atm-/- mice were generated and genotyped as described previously [34 , 35] . The day of birth was designated as postnatal day 0 ( P0 ) . Neonatal mice were used at P5 , P7 , P9 , and P12 . Adult mice were 3-month-old . Untreated age-matched mice were used as control . All animal experiments were performed in accordance with accepted standards of animal welfare approved by the United Kingdom Home Office and complied with the Animals ( Scientific Procedures ) Act 1986 . Irradiation was performed using an HS X-ray system ( A . G . O . installation , UK ) . Mice were exposed to a total-body dose of either 100 mGy or 2 Gy ( dose rate of 0 . 5 Gy/min at 250 kV potential ) . Coronal sections of the adult and neonatal brain were taken to correspond to where the lateral ventricle ( LV ) is maximally extended along the dorsal-ventral axis at plane positions relative to bregma between +1 . 2 and +0 mm ( brain atlas: Paxinos and Franklin , 2008 ) . The SVZ was topographically divided into 4 subdomains: ventral , the area delineating the lower ventral one-third of the LV in correspondence of the nucleus accumbens; medial , the area corresponding to the middle-upper medial wall of the LV in correspondence of the caudal part of the lateral septal nucleus; dorsal , the area adjacent to the corpus callosum encompassing the dorsal roof of the LV; and dorsolateral , the upper dorsal one-third of the caudoputamen [16 , 33] . Brains were first fixed with 4% paraformaldehyde ( PFA ) ( Santa Cruz Biotecnology , cat# SC-281692 ) , cryoprotected in 25% sucrose/PBS solution , and frozen after OCT embedment ( Thermo Fisher Scientific , cat# 12678646 ) . Samples were cryosectioned ( coronal view , 7 μm ) on a CM1900 cryostat ( Leica ) and mounted on Superfrost slides ( Thermo Fisher Scientific , cat# 10149870 ) . Antigen retrieval was performed according to the manufacturer’s instructions ( Histo VT One , Fine Chemicals Products Ltd , cat# 06380–05 ) . Slides were then washed 3 times with PBS and blocked in 5% serum ( donkey or goat ) , 1% bovine serum albumin ( BSA ) , and 0 . 4% Triton X-100 in PBS for 1 h at room temperature ( RT ) . Primary antibody incubation was done in a humidified chamber overnight at RT . The following antibodies were used in this study: 53BP1 ( rabbit , 1:500 , Bethyl Laboratories , cat# A300-272A ) , Cleaved Caspase-3 ( rabbit , 1:200 , Cell Signaling , cat# 9579S ) , Doublecortin ( goat , 1:400 , Santa Cruz Biotechnology , cat# sc-8066 ) , GFAP ( rabbit , 1:500 , Abcam , cat# ab7260; mouse , 1:400 , Millipore , cat# MAB360 ) , Ki67 ( rat , 1:500 , eBioscience , cat# 14–5698 ) , Mash1 ( rat , 1:100 , R&D Systems , cat# MAB2567; mouse , 1:100 , BD Biosciences , cat# 556604 ) , Tuj1 ( mouse , 1:400 , Sigma-Aldrich , cat# T8578; rabbit , 1:100 , Sigma-Aldrich , cat# T2200 ) , MAP2 ( rabbit , 1:400 , Millipore , cat# AB5622 ) , Iba1 ( rabbit , 1:1000 , Wako Chemicals USA , cat# 019–19741 ) , Sox2 ( mouse , 1:100 , Cell Signaling , cat# 4900 ) , and Nestin ( mouse , 1:200 , Millipore , cat# MAB353 ) . Slides were washed 3 times with PBS and then incubated with the relative Alexa Fluor 488 , 555 , 594 , and 647 conjugated secondary antibodies for 1 h at RT ( host either donkey or goat , 1:500 , Thermo Fisher Scientific ) . Sections were counterstained with DAPI and mounted with Vectashield ( Vector Laboratories Ltd , cat# H-1000 ) . BrdU ( Sigma-Aldrich , cat# B5002 ) was injected intraperitoneally at 50 mg/kg body weight at 4 h and 1 h before irradiation . EdU ( Thermo Fisher Scientific , cat# A10044 ) was injected intraperitoneally at 50 mg/kg body weight at 1 h before the mice were killed . For BrdU detection , sections were denatured in 2 . 5 M HCl for 25 min at RT , washed with PBS , and subjected to standard immunofluorescence staining using antibodies against BrdU ( MoBU-1 ) ( 1:100 , mouse , Thermo Fisher Scientific , cat# B35128 ) and BrdU ( BU1/75 ) ( 1:100 , rat , Biorad , cat# OBT0030CX ) . EdU incorporation was measured using the Click-iT EdU Alexa Fluor 647 imaging kit ( Thermo Fisher Scientific , cat# C10340 ) . TUNEL staining was performed after immunofluorescence staining according to the manufacturer’s instructions ( Roche , cat# 11684795910 ) . Brains were embedded in OCT and flesh frozen in liquid nitrogen after resection . Frozen tissues were immediately sectioned ( coronal , 4 μm ) and mounted onto Superfrost slides . Cellular senescence was assessed using the Senescence β-Galactosidase Staining Kit ( Cell Signaling , cat# 9860 ) according to the manufacturer’s instructions and counterstained with Nuclear Fast Red ( Vector Laboratories , cat# H-3403 ) . Olympus IX73 and Carl Zeiss microscopes were used for immunofluorescence imaging with 40x and 100x objectives using Micro-Manager software . Bright-field imaging was performed using a Zeiss Axiovert microscope . Image processing and analysis was performed using ImageJ ( NIH ) . Between 3 to 6 tissue sections were analysed for each individual mouse . Quantification of marker-positive cells by immunofluorescence was carried out in at least 200 cells for each SVZ subdomain per section ( with a minimum of 600 cells/subdomain scored per mouse ) . Quantification of 53BP1 foci was done by scrolling through the entire section depth with a 100x objective in at least 100 cells per section . Values represent the mean and standard error of at least 3 independent experiments . Unpaired Student t tests were used for the determination of P values . We used between 3 to 6 mice per each experimental group to allow basic statistical inference given a significant P value of 0 . 05 . | The response of stem cells to DNA damage is poorly understood , although there is increasing evidence that they respond distinctly to differentiated cells . We have monitored the different responses of adult neural stem and progenitor cells to exposure to X-ray irradiation . We see that progenitor cells activate apoptosis , undergo rapid proliferation arrest , and premature differentiation . However , quiescent stem cells do not activate radiation-induced apoptosis . Instead the responses of the progenitor cells promote the activation of these quiescent neural stem cells , thereby replenishing the neuroblasts . These responses and quiescent stem cell activation are dependent on the ataxia telangiectasia mutated ( ATM ) kinase . We propose that this coordinated response functions to remove damaged progenitor cells and to aid repopulation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"cell",
"death",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"neuronal",
"differentiation",
"neuroscience",
"cell",
"differentiation",
"dna",
"damage",
"developmental",
"biology",
"stem",
"cells",
"dna",
"developmental",
"neuroscience",
"animal",
... | 2017 | A coordinated DNA damage response promotes adult quiescent neural stem cell activation |
Enterovirus 71 ( EV71 ) and coxsackieviruses ( CV ) are the major causative agents of hand , foot and mouth disease ( HFMD ) . There is not currently a vaccine available against HFMD , even though a newly developed formalin-inactivated EV71 ( FI-EV71 ) vaccine has been tested in clinical trial and has shown efficacy against EV71 . We have designed and genetically engineered a recombinant adenovirus Ad-EVVLP with the EV71 P1 and 3CD genes inserted into the E1/E3-deleted adenoviral genome . Ad-EVVLP were produced in HEK-293A cells . In addition to Ad-EVVLP particles , virus-like particles ( VLPs ) formed from the physical association of EV71 capsid proteins , VP0 , VP1 , and VP3 expressed from P1 gene products . They were digested by 3CD protease and confirmed to be produced by Ad-EVVLP-producing cells , as determined using transmission electron microscopy and western blotting . Mouse immunogenicity studies showed that Ad-EVVLP-immunized antisera neutralized the EV71 B4 and C2 genotypes . Activation of VLP-specific CD4+ and CD8+/IFN-γ T cells associated with Th1/Th2-balanced IFN-ɣ , IL-17 , IL-4 , and IL-13 was induced; in contrast , FI-EV71 induced only Th2-mediated neutralizing antibody against EV71 and low VLP-specific CD4+ and CD8+ T cell responses . The antiviral immunity against EV71 was clearly demonstrated in mice vaccinated with Ad-EVVLP in a hSCARB2 transgenic ( hSCARB2-Tg ) mouse challenge model . Ad-EVVLP-vaccinated mice were 100% protected and demonstrated reduced viral load in both the CNS and muscle tissues . Ad-EVVLP successfully induced anti-CVA16 immunities . Although antisera had no neutralizing activity against CVA16 , the 3C-specific CD4+ and CD8+/IFN-γ T cells were identified , which could mediate protection against CVA16 challenge . FI-EV71 did not induce 3C-mediated immunity and had no efficacy against the CVA16 challenge . These results suggest that Ad-EVVLP can enhance neutralizing antibody and protective cellular immune responses to prevent EV71 infection and cellular immune responses against CV infection .
Enterovirus 71 ( EV71 ) and coxsackievirus ( CVA ) infections are the most common causative factors of hand , foot , and mouth disease ( HFMD ) and other neurological disorders . Severe neurological disorders , including encephalitis , acute flaccid paralysis , pulmonary edema ( PE ) , and hemorrhaging , culminating in fatality , particularly in EV71-infected children under 5 years old , have been reported [1–5] . Because HFMD and EV71 infections can potentially become a new threat to global public health [1 , 6–11] , effective antiviral drugs and prophylactic vaccines are urgently needed . Enterovirus is a nonenveloped , single-stranded RNA virus of the Picornaviridae family . Its genome consists of a single open reading frame that encodes the P1 , P2 , and P3 poly-proteins . The P2 and P3 regions encode nonstructural proteins ( e . g . , 3CD ) responsible for virus replication and virulence . During viral RNA translation , the 2A protein catalyzes its N-terminal cleavage in cis , thereby releasing the capsid proteins in the P1 region from the nascent nonstructural proteins in the P2 and P3 regions . 3CD is released from the P3 precursor by autocatalytic cleavage . A 3C’ cleavage site in the polyprotein resides between the 3C and 3D portion of 3CD to generate 2 products , 3C’ and 3D’ . When the P1 precursor is encoded by the P1 region , it can be cleaved by the 3C’ protease into VP0 , VP1 , and VP3 . These 3 proteins spontaneously assemble into an icosahedral procapsid and pack the RNA genome into the provirion that could be a non-infectious empty ( E ) -particle or infectious full ( F ) -particle [12 , 13] . Human scavenger receptor class B , member 2 ( hSCARB2 ) and human P-selectin glycoprotein ligand 1 ( PSGL-1 ) have been identified as the important cell receptors for EV71 infection [14 , 15] . Our group [16] and Fujii et al . [17] have successfully developed transgenic mice expressing the human hSCARB2 receptor . In this promising model , transgenic animals infected with clinical EV71 isolates of the B4 and B5 subgenotypes developed HFMD-like skin rashes , whereas those inoculated with EV71 C2 and C4 subgenotypes or CVA16 suffered severe limb paralysis and death . In addition , passive administration of the monoclonal anti-EV71 VP1 neutralizing antibody N3 [18] reduced EV71 B5 infection-induced symptoms and protected the transgenic mice against EV71 C2-induced severe limb paralysis and death[19] . In a previous study[20] , we produced a formalin-inactivated EV71 strain E59 ( FI-EV71 ) vaccine candidate formulated with alum adjuvant , and found that FI-EV71 displayed high efficacy in the hSCARB2-Tg mouse challenge model [16] . In a human phase I clinical trial[21] , FI-EV71 was safe and could elicit strong neutralizing antibody responses against current circulating EV71 isolates , but failed to protect against CVA16 infections . A replication-incompetent adenovirus ( Ad ) containing a fusion protein ( F ) gene of the respiratory syncytia virus ( RSV ) could induce viral antigen specific neutralizing antibody and cellular immunity against RSV infection in a mouse model [22] . In addition , Ad-expressing human immunodeficiency virus gp120 was tested as a vaccine in human trials and was confirmed to be safe [23] . However , some reports have argued that preexisting anti-adenovirus antibodies can influence the efficacy of the intramuscular injection of Ad vaccine intramuscular in clinical use [23] . Previous studies have shown that preexisting antibodies do not affect the subsequent generation of humoral responses to the antigen expressed by a recombinant Ad when administered mucosally [24 , 25] . In this study , we designed a recombinant Ad-5 carrying both P1 and 3CD genes of EV71 ( Ad-EVVLP ) . Upon Ad-EVVLP replication in HEK-293A cells , EV71 VLPs could self-assemble through the cleavage of P1 polyprotein into individual proteins ( VP1 , VP3 , and VP0 ) by the 3C’ protease . We investigated the immunogenicity of Ad-EVVLP in mice to determine whether neutralizing antibodies and Th1/Th2 cellular responses were induced to cross-protect against EV71 and CVA16 in animal models .
All animal experiments were conducted in accordance with the guidelines of the Laboratory Animal Center of the National Health Research Institutes ( NHRI ) , Taiwan . Animal use protocols were reviewed and approved by the NHRI Institutional Animal Care and Use Committee ( Approval Protocol No . NHRI-IACUC-100125-A ) . In EV71 challenge experiments , survival rate was used as an endpoint to assess the protective efficacy of the anti-EV71 treatment . Survival rate used as an index of pathogenesis of EV71 infection has been reported by numerous studies in experimental animal models [16 , 26–28] . After investigation , tested animals were euthanized by 100% CO2 inhalation for 5 min followed by cervical dislocation to minimize suffering . To perform virus challenge , mice were placed in an anesthetic inhalator chamber containing isoflurane ( initial phase: 5%; maintenance phase: 1 . 5%–2 . 5% ) for 1 min before s . c . or i . p . EV71 immunization . African green monkey kidney ( Vero ) ( ATCC No . CCL-81 ) and human rhabdomyosarcoma ( RD ) ( ATCC No . CCL-136 ) cells were provided by the Taiwan Centers of Disease Control ( Taiwan CDC ) ; the original cell lines were obtained from the American Type Culture Collection ( ATCC ) , United States . Vero cells were cultured in a VP-SFM medium ( Gibco-Invitrogen , CA , USA ) supplemented with 4 mM L-glutamine ( Gibco-Invitrogen , CA , USA ) . The RD cell line was cultured in DMEM medium containing 10% fetal bovine serum ( Gibco-Invitrogen , CA , USA ) . Cells were maintained in a 37 °C incubator equilibrated with 5% CO2 . Clinically isolated strains of EV71 , E59 ( B4 ) ( GenBank: GQ150746 . 1 ) , Neu ( pinf7-54A ) strain ( C2 ) ( GeneBank DQ060149 ) , Tainan/5746/98 ( C2 ) ( GenBank: AF304457 . 1 ) , and one strain of CVA16 , 5079 ( GenBank: AF177911 . 1 ) were obtained from Dr . Jen-Ren Wang , National Cheng-Kung University , Tainan , Taiwan , and were propagated in Vero cells based on the microcarrier cell culture bioprocess [29 , 30] . Human adenivirus 5 ( Ad5; ATCC No . VR-1516 ) was purchased from ATCC and propagated in 293A cells . Virus stocks were stored at—80 °C . Virus stock titers were tested in a standard plaque-forming assay [31] , and the number of plaque-forming units ( pfu ) was calculated . The monoclonal antibody and Mab979 against the VP0/VP2 capsid protein of EV71 [32] was purchased from Millipore Inc . ( MA , USA ) . A VP1-specific monoclonal antibody E1 was produced in-house , as described previously [32] . The anti-3C antibody was purchased from GeneTex ( Cat . No . GTX630191 ) . PE-Cy5-labeled mouse-specific CD4 and CD8 antibodies ( Cat . No . 15-0042-82 and 15-0081-82 , respectively ) were purchased from eBioscience . PE-labeled rat anti-IFN-ɣ ( Cat . No . 554412 ) was obtained from BD Bioscience . The anti-adenovirus type 5 antibody was obtained from Abcam ( Cat . No . ab6982 ) . Horseradish peroxidase ( HRP ) -conjugated donkey anti-mouse antibody ( Cat . No . 715-036-150 ) and HRP-conjugated rabbit anti-goat antibody ( Cat . No . 305-035-003 ) were purchased from Jackson Immunoresearch Inc . ( PA , USA ) . The P1 and 3CD genes of the EV71 Neu ( pinf7-54A ) strain were amplified by PCR and individually inserted into the shuttle vector pENTR4 ( Invitrogen ) . The nucleotide element of the elongation factor-1α ( EF-1α ) promoter was inserted into the 3’ end of the P1 gene and the 5’ end of the 3CD insert to generate the pENTR4-P1/EF-1α/3CD construct . The 3CD gene alone was inserted into pENTR4 to generate the pENTR4-3CD construct . The pENTR4-P1/EF-1α/3CD and pENTR4-3CD constructs were enzymatically recombined into the ΔE1/ΔE3 ( replication-incompetent ) Ad5 vector pAd/CMV/V5-DEST [33] to form recombinant pAd-EVVLP and pAd-3CD , respectively . pAd DNA was transfected into the 293A packaging cell line to produce the recombinant adenoviruses designated Ad-EVVLP and Ad-3CD . Ad-LacZ carrying a luciferase reporter gene as a vector control was obtained from Invitrogen . The recombinant viruses were purified and concentrated using Vivapure adenoPACK 100RT ( Satorius Stedin Biotech ) . The purified virus titers were determined using a modified standard plaque assay . Various Ad virus dilutions were added to each well of 293A cells plated in a 6-well tissue culture plate . After overlaying the cultures with DMEM containing 0 . 75% methylcellulose , the cultures were incubated at 37 °C for 10 to 12 days and plaques were counted . The typical yield of adenoviruses was approximately 1 × 109 pfu/mL . Western blotting was performed as described previously [31] . Total cell lysates were prepared by treating 1 to 2 × 106 cells with 100 μL ice cold lysis buffer ( 0 . 5% sodium deoxycholate , 0 . 1% sodium dodecyl sulfate ( SDS ) , 0 . 5% NP-40 , 50 mM TRIS , 150 mM NaCl ) plus a protease inhibitor cocktail ( Roche , French ) and 1 mM PMSF ( Sigma-Aldrich , CA , USA ) . Lysates were centrifuged for 20 min at 10 , 000 rpm at 4 °C to sediment the cell debris . The protein concentration of the cell lysates or fractions was measured using the Bradford method [34] . Cell lysates containing 10 μg protein were mixed with loading dye and loaded into each well of a 10% SDS-polyacrylamide gel ( SDS-PAGE , Amersham Biosciences-GE Healthcare , USA ) and subjected to electrophoresis in 1x Tris-glycine SDS-running buffer . The resolved proteins were transferred onto nitrocellulose membrane ( Hybond-ECL , Amersham Biosciences-GE Healthcare , USA ) . Membranes were soaked in 5% skim milk in 1x PBS for 30 min at room temperature , then washed 3 times with 1x PBS plus 0 . 05% Tween 20 ( PBS-T ) . The membrane was incubated with rat anti-3C ( 1:1000 ) , MAB979 ( 1:5000 ) , or anti-VP1 antibody ( 1:1000 ) for 14 to 16 h at 4 °C and subsequently washed with PBS-T followed by incubation with HRP conjugated anti-rat or donkey anti-mouse ( for MAB979 ) antibodies . After 1 h incubation , the membrane was washed 5 times with PBS-T , and then Super Signal West Pico chemiluminescent substrate ( Pierce , IL , USA ) was layered onto the membrane , and it was exposed to X-ray film ( Kodak , NY , USA ) . When necessary , the membranes were stripped using Restore buffer ( Pierce , IL , USA ) and blotted with another antibody . Splenocytes were harvested from BALB/c mice and labeled with 5- ( 6 ) -carboxyfluorescein diacetate succinimidyl ester ( CFSE ) ( Cat . No . C34554 , Molecular Probes ) . They were restimulated in vitro with 107 pfu/mL UV-inactivated EV71 5746 or 1 . 4 μg/mL purified recombinant E59 3C proteins expressed by E . Coli . ( provided by Dr . Pele Chong , a coauthor of this study ) for 5 days . Proliferation of splenocytic CD4+ T cells was analyzed by flow cytometry ( BD FACSCalibur ) using PE-Cy5-labeled anti-CD4 antibodies . The population of no fluorescence signal-shifting in CFSE-prestained CD4+ T cells without antigen stimulation was set to 0% , and the population of negatively shifted CD4+ T cells ( proliferating cells ) after antigen stimulation was quantified . The mean percentage corresponding to the individually proliferating CD4+ T cells in each group was calculated . To detect the population of CD8+IFN-γ+ T cells , splenocytes were cocultured with the EV71 antigen for 2 days and then with brefeldin A ( Cat . No . 00-4506-51 , eBioscience ) for 3 h before harvesting . Stimulated splenocytes were stained with PE-Cy5-labeled anti-CD8 antibody for 30 min , followed by subsequent fixation and permeabilization . A portion of these cells was further stained with PE-conjugated anti-IFN-γ+ antibody ( BD Bioscience ) for 30 min to detect intracellular IFN-γ . After washing , the samples were analyzed by flow cytometry . pAd-EVVLP plasmid DNA was used as a template to detect the P1 , 3CD , and EF-1α promoter regions within pAd-EVVLP by PCR using the respective primer pairs . The PCR conditions were as follows: 95 °C for 3 min; 35 cycles at 95 °C for 1 min , 60 °C for 1 min , and 72 °C for 3 min; and a final incubation at 72 °C for 2 min . Total RNA was purified from tissues using TRIZOL reagent ( Invitrogen , CA , USA ) following the manufacturer’s instructions and was subjected to real time RT-PCR . Total RNA was converted into cDNA using random primers ( Genomics BioSci&Tech , Taiwan ) and reverse transcriptase ( Bionovas , Toronto , Canada ) . The synthesized cDNA was subjected to quantitative PCR analysis ( LightCycler 480 SYBR Green Real-Time PCR system ) using primer pairs specific to the VP1 region of EV71 P1 RNA . Human β-actin gene expression was used as an internal control . The PCR conditions were as follows: 95 °C for 3 min; 40 cycles at 95 °C for 10 s , 65 °C for 20 s , and 72 °C for 2 s; and a final incubation at 72 °C for 2 min . The number of cycles required for amplification of transcripts was obtained . The relative expression of EV71 P1 RNA was calculated as follows: the individual Ct obtained from the experimental group or control group was subtracted by its respective Ct ( β-actin ) to gain normalized Ct . Then , 2Normalized Ct ( VP1 of P1 RNA from the sample without viral infection ) was divided by 2Normalized Ct ( VP1 of P1 RNA from the sample with viral infection ) . The forward and reverse primers , [5_-ACGCGCAAATGCGTAGAAAGGT-3_—forward and 5_-TTAGTGGCAGTTTGCCATGCGA-3_—reverse] , were used to amplify and detect VP1 RNA . human β-actin mRNA was amplified using the primer pairs 5_-ACCAACTGGGACGACATGGAGAAA-3_—forward and 5_-TAGCACAGCCTGGATAGCAACGTA-3_—reverse . Primer pairs targeting the P1 , 3CD , and EF-1α promoter regions of Ad-EVVLP are as follows: P1: 5_-ATCG GAATTCATGGGCTCACAGGTGTCCAC-3_—forward and 5_-CTTGTCGACTTAGAGAG TGGTAATTGCTG-3_—reverse , 3CD: 5_-ATCGGAATTCATGGGGCCGAGCTTGGAC-3_—forward and 5_-ATCGCTCGAGAAACAATTCGAGCC-3_—reverse , EF-1α promoter: 5_-ATCGACGCGTGTGAGGCTCCGGTGCCC-3_—forward and 5_-ATCGCCCGGGGTTTTCACGACACCTG-3_—reverse . All primer sets were commercially synthesized by Genomics BioSci&Tech , Taiwan . HEK-293A cells ( 1 x 107 ) were seeded in a 10-well plate 1 day prior to Ad-EVVLP infection with MOI = 1 . After 24 h of infection , the cells were harvested and lysed in 1% NP-40 lysis buffer ( 50 mM Tris/HCl , pH 7 . 5 , 150 mM NaCl , 2 mM EDTA , and 1% NP-40 ) on ice for 30 min and centrifuged at 1000 xg for 10 min to remove the cell debris . The supernatants were harvested and concentrated by ultracentrifugation at 100 , 000 xg for 1 h at 4 °C and then dissolved in 30 μL PBS . The samples were loaded into self-generated iodixanol gradients , which were prepared by mixing 0 . 6 mL solution S ( 0 . 25 M sucrose , 15 mL EDTA , 30 mM Tris/HCl , pH 8 . 0 ) and 0 . 42 mL 60% ( w/v ) iodixanol ( Cat . No . 1114542 , Optiprep; Axis Shield , UK ) to form a homogenous solution . Gradients were generated through centrifugation at 162 , 000 xg for 24 h at 4 °C . The various fractions were manually harvested from the top ( named Fraction No . 1 ) , 0 . 1 mL per fraction , and 10 fractions were serially collected for each sample . These fractions were subjected to Western blot using Mab979 antibodies or transmission electron microscopy . HEK-293A cells were harvested 24 h after Ad-EVVLP infection , and cell pellets were frozen and thawed twice at—80 °C for 30 min and 37 °C for 15 min . Lysates were centrifuged at 3000 rpm for 15 min at room temperature , and supernatants were harvested and subjected to examine adenovirus using a JEOL JEM-1400 transmission electron microscope ( TEM ) . The lysate was fractionated through density gradient centrifugation , and fractions were concentrated through ultracentrifugation at 100 , 000 xg for 1 h and resuspended in 200 μL PBS . The fractions were then cleaned by centrifugation in a 100 KDa-cut-off spin-XR UF 20 column ( Corning ) . Samples were treated with uranyl acetate and inspected by TEM . To detect anti-EV71 , anti-Ad , or anti-3C antibodies in sera , 96-well plates were coated with 100 μL per well of heat-inactivated ( 56 °C for 1 h ) 103 pfu EV71 5746 ( C2 genotype ) or E59 ( B4 genotype ) strains , 200 pfu purified Ad5 , or 700 ng recombinant 3C protein in carbonate coating buffer . Serum samples collected from immunized mice were inactivated at 56 °C for 30 min . Two-fold serial dilutions of the sera were performed beginning from an 8-fold initial dilution . The diluted sera were added to the wells and incubated at room temperature for 2 h . After washing with PBS-T , HRP-conjugated donkey anti-mouse IgG antibodies were added to the wells for 45 min . The reaction was developed by incubation with 100 μL TMB substrate ( 3 , 3’ , 5 , 5’-etramethyllbenzidine ) for 20 min in the dark and terminated by adding 50 μL 2 N H2SO4 . The optical density at 450 nm was determined using a microplate absorbance reader ( SPECTRA , MAX2 , M2 ) . To detect cytokines secreted by splenocytes , the supernatants from 2-day cultures of splenocytes restimulated with 107 pfu/mL UV-inactivated EV71 5746 were analyzed using a calorimetric sandwich IFN-γ , IL-4 , IL-13 , and IL17A ELISA kit ( Cat . No . 887314 , 88–7044 , 88–7137 , and 88–7371 , respectively , eBioscience ) . The assays were conducted according to the manufacturer’s instructions , and the optical densities at 450 nm were determined using a microplate absorbance reader . To detect the neutralizing activity as described in our previous study [32] , each sample was serially diluted 2-fold in fresh cell culture medium . A total of 100 μL 100 TCID50 virus suspension , E59 , 5746 , or CVA16 strain was added to each tube containing 100 μL serially diluted serum . After incubation at 4 °C for 18 to 24 h , 100 μL virus serum mixture was added to 96-well plates seeded with rhadomyosarcoma ( RD ) cells and incubated for 7 days at 37 °C; TCID50 values were measured by counting cytopathic effects ( CPE ) . The 50% neutralization inhibition dose ( ID50 ) was calculated as the reciprocal of the serum dilution compared to normal serum using the Reed–Muench method [35] . A mouse anti-EV71 Mab979 antibody ( Chemicon International ) was used as an internal positive control . Suspensions containing 5 × 106 RBC-free splenocytes were prepared from individual mice and seeded in individual wells of 96-well filtration plates ( Millipore ) pre-coated with capturing monoclonal antibodies for murine IL-4 or IFN-γ ( 0 . 5 μg/well ) ( Cat . No . 16-7041-68 or 16-7313-68 , respectively , eBioscience ) and blocked with conditioned medium ( CM ) for 1 h at room temperature . The splenocytes were added to 106 pfu/well UV-inactivated EV71 5746 dissolved in CM ( 100 μL ) . Splenocytes incubated with Con A ( 10 g/mL ) were used as a positive control . Unstimulated splenocytes were used as a negative control . Plates were maintained in a 37 °C incubator equilibrated with 5% CO2 for 48 h . The individual wells of the ELISPOT plates were washed 3 times with PBS-T , and 0 . 2 g of the corresponding biotinylated detection monoclonal IL-4- or IFN-γ-specific antibody was added to detect the respective cytokines . The plates were washed after 2 h incubation at room temperature , and 100 L streptavidin-alkaline phosphatase ( 1:250 dilution ) was added to the individual wells . The plates were incubated at room temperature for 45 min . Finally , the plates were washed 4 times with wash buffer , and 100 L AEC ( 3-amine-9-ethylcarbazole , Sigma-Aldrich ) substrate was added to each well and allowed to react for 30 min at room temperature in the dark . The plates were washed with water , air-dried overnight , and the spots on each well were scored using an immunospot counting reader ( Immunospot , Cellular Technology Ltd . ) . The results were expressed as the number of cytokine-secreting cells per 5 × 105 splenocytes seeded in the initial culture . hSCARB2-Tg mice in a C57BL/6 background generated were previously generated by our group and were maintained by cross-mating hSCARB2-Tg subjects to obtain inbred mice [16] . One-day-old hSCARB2-Tg mice were inoculated s . c . with PBS , 3 × 107 pfu Ad-LacZ , or 3 × 106 or 3 × 107 pfu Ad-EVVLP , or 1 μg FI-EV71 vaccine on Days 1 and 7 , and then challenged s . c . with 3 × 106 pfu EV71 5746 or 5 × 105 pfu CVA16 on Day 14 . The mice were monitored daily for survival for 15 days after the challenge . The list of accession number/ID numbers for cells and viruses was shown below: ATCC No . CCL-81: African green monkey kidney ( Vero ) cell . ATCC No . CCL-136: Human rhabdomyosarcoma ( RD ) cell . GenBank , GQ150746 . 1: EV71 , E59 ( B4 ) . GeneBank , DQ060149: EV71 Neu ( pinf7-54A ) ( C2 ) . GenBank , AF304457 . 1: EV71 Tainan/5746/98 ( C2 ) . GenBank , AF177911 . 1: CVA16 , 5079 . ATCC No . VR-1516: Human adenovirus type 5
We used the E1- and E3-deleted adenovirus-5 genome to construct the Ad-EVVLP expression vector , which carried full-length P1 and 3CD genes of EV71 ( Fig 1A ) . We performed polymerase chain reaction ( PCR ) to confirm the inserts: P1 , 3CD , and elongation factor-1α promoters ( EF-1p ) using the respective primers in the Ad-EVVLP construct . The PCR products corresponded to 2585 bps ( P1 ) , 2020 bps ( 3CD ) , and 1186 bps ( EF-1p ) ( Fig 1B ) . Upon transfection in competent HEK-293A cells constitutively expressing the E1 protein , the recombinant Ad was generated . We examined Ad-EVVLP and the control Ad-LacZ by western blot using the polyclonal anti-Ad5 antibody . They each expressed Ad structural proteins , including hexon , penton , and protein V , VI , and VII ( Fig 1C ) . Previous studies have shown that multiple capsid proteins of VP0 ( 38 KDa , a precursor product of VP2+VP4 ) , VP1 ( 36 KDa ) , VP2 ( 28 KDa ) , VP3 ( 25 KDa ) , and VP4 ( 8 KDa ) can be detected in EV71-infected cells [32 , 36] . The VLP expression by Ad-EVVLP was characterized; translated products of EV71 VLPs , including VP0 ( a precursor of VP4-VP2 ) but not VP2 , were detected in 293A cells using the VP2-specific monoclonal antibody Mab979 . EV71 antigens VP0 and VP2 were detected by Western blotting ( Fig 1C ) . We confirmed VP1 expression using a VP1-specific antibody , which corresponded to the 34 to 36 kDa bands in Ad-EVVLP-infected lysates and in the sample of EV71 antigens ( Fig 1C ) . The antigenic profile of VLPs expressed by Ad-EVVLP was similar to E-particles ( composed by VP0 , VP1 , and VP3 ) from EV71 , which does not contain viral RNA , compared to F-particles ( composed by VP2 , VP4 , VP1 , and VP3 ) . In addition , 3C’ processed the P1 polyprotein to form VP0 , VP1 , and VP3 in the absence of EV71 genetic RNA [36] . We detected an 18 kDa band using an anti-3C antibody . In contrast , we could not detect these bands in uninfected or Ad-LacZ-transfected 293A cells ( Fig 1C ) . However , no VP3-specific antibody is available to detect VP3 . We could not detect a VP3 signal by blotting with sera from EV71-infected mouse ( Fig A in S1 Text ) . To demonstrate that the EV71 VLP particles were cogenerated in Ad-producing cells , we purified the virions from the cytosol of Ad-EVVLP transfectant using fractionation through density gradient centrifugation . We fractionated EV71 particles as a control . We characterized each fractionated sample by Western blot analysis using Mab979 . The major band was 38 kDa , corresponding to VP0 , but there was minor expression of a 28 kDa corresponding to VP2 ( intensity ratio of VP0/VP2 = 4 and 6 , respectively ) , which together make the E-particle of EV71 in Fractions 6 and 7 . In contrast , the opposite pattern of VP0/VP2 expression ( 0 . 8 , 0 . 9 , and 0 . 8 , respectively ) corresponding to F-particles was observed in Fractions 8 , 9 , and 10 . However , only VP0 signals were detected in Fractions 6 to 9 of Ad-EVVLP-infected lysate ( Fig 2A ) . A similar antigenic profile of EV71 has been previously reported [37] . We pooled Fractions 7 and 8 of Ad-EVVLP samples and examined them by TEM . TEM analysis revealed some fractured VLPs ( f ) and cellular impurities in the samples due to sample preparation . Two sizes of complete particles were also present; particles over 100 nm in diameter corresponded to Ad particles ( Fig 2D ) , and particles approximately 30 nm in diameter corresponded to VLPs expressed by Ad-EVVLP ( Fig 2B ) . EV71 particles in the pool of Fractions 8 and 9 of EV71 sampleswere also examined ( Fig 2C ) . To examine the immunogenicity of Ad-EVVLP compared to the FI-EV71 vaccine , we intraperitoneally ( i . p . ) , subcutaneously ( s . c . ) , or orally administered adult BALB/c mice with 1 × 108 pfu of Ad-EVVLP or Ad-LacZ on Days 1 and 14 . Animals in separate groups were s . c . administered 0 . 1 μg or 1 μg FI-EV71 twice to evaluate the virus-specific immune responses compared with those of recombinant adenoviruses . The results of ELISA assays showed ( Fig 3A ) that the mean anti-EV71 titer against EV71 5746 ( C2 subgenotype ) in Ad-EVVLP-immunized serum samples collected on Day-21 were 2240 , 7040 , and 130 for s . c . , i . p . , and orally , respectively . We did not detect a titer in serum from s . c . Ad-LacZ-immunized mice ( Fig 3A ) . The mean titer of serum antibodies reacting with the EV71 E59 strain ( B4 subgenotype ) from the Ad-EVVLP-immunized animals was to 2240 , 8960 , and 180 , for s . c . , i . p . , or orally , respectively . Again , no E59 reactivity was detected in serum of the mice immunized with Ad-LacZ ( Fig 3B ) . Sera from Ad-EVVLP-immunized mice possessed EV71 neutralizing activity ( Table 1 ) . Higher virus neutralization titers ( 1/128 ) were found in i . p . and s . c . Ad-EVVLP-immunized mice compared to a considerably low neutralizing titer in orally administered animals . Neutralizing antibodies produced in Ad-EVVLP-immunized mice exhibited potent neutralizing activity against EV71 B and C strains . Comparable titers ( 1/256 and 1/512 ) of neutralizing antibody in the mice s . c . administered 0 . 1 μg FI-EV71 vaccine . No anti-CVA16 neutralizing activity was found in the serum from mice immunized with Ad-EVVLP , FI-EV71 vaccine , or PBS ( < 1:8; Table 1 ) . These results are consistent with previous reports [38] that FI-EV71 vaccine could not elicit cross-neutralizing antibody against CVA16 . Recent studies on host immune responses against EV71 have suggested that T cell immunity plays a critical role in the protection against EV71 infection and control of the disease [39 , 40] . Therefore , we investigated whether the VLP-specific CD4+ and CD8+ T cell responses could be elicited in Ad-EVVLP-immunized mice . Seven days post-immunization , we isolated lymphocytes from the spleen , followed by in vitro restimulation with UV-inactivated EV71 ( UV-EV71 ) . Lymphocytes from Ad-LacZ-immunized mice produced background IFN-γ levels . In contrast , substantially higher IFN-γ levels were measured in lymphocyte cultures from mice administered Ad-EVVLP ( Fig 4A ) . Lymphocytes from FI-EV71 vaccine-immunized mice secreted background IFN-γ levels ( Fig 4A ) . Within the panel of Th2 cytokines assayed , IL-4 ( Fig 4B ) and IL-13 ( Fig 4C ) were moderately secreted by lymphocytes from Ad-EVVLP-immunized mice , indicating that balanced Th1/Th2 responses were activated . Interestingly , immunization of the FI-EV71 vaccine led to the production of the highest IL-4 and IL-13 levels , indicating that a Th2 biased response was induced ( Fig 4B and 4C ) . This result supports our findings that FI-EV71 vaccination in hSCARB2-Tg mice induced splenocytic IL-4 but not IFN-γ secretion , as shown previously [16] . The results obtained from IFN-γ and IL-4 ELISPOT assays confirmed that i . p . Ad-EVVLP immunization induced significant splenocytic IFN-γ production and low levels of IL-4 secretion in Ad-EVVLP-vaccinated mice ( Fig B in S1 Text ) . A considerable amount of IL-17A was produced by splenocytes from Ad-EVVLP-immunized mice in response to EV71 antigens . This was in sharp contrast to the barely detectable amount of IL-17 secreted by splenocytes of animals immunized with Ad-LacZ or the FI-EV71 vaccine ( Fig 4D ) . These results indicate that Ad-EVVLP drives T cell activation , leading to the differentiation of a subpopulation of T cells that bear the Th1 , Th2 , and IL-17 producing phenotypes . We measured VLP-specific CD4+ T cell proliferation in vaccine-immunized splenocytes followed by restimulation with UV-EV71 by examining the negative shift of fluorescent signal in 5- ( 6 ) -carboxyfluorescein diacetate succinimidyl ester ( CFSE ) -prestained CD4+ T cells using flow cytometry . Compared to little or no shift of signals in the PBS- and Ad-LacZ-immunized groups ( 3% and 8 . 4% , respectively ) , a substantial shift was detected in Ad-EVVLP-immunized group ( 42%; Fig 5A ) . The proliferation of CD4+ T cells corresponding to UV-EV71 was barely detectable in the FI-EV71-immunized mice ( 8%; Fig 5A ) , indicating that the antigenicity of FI-EV71 reacting to VLP was altered , and therefore the immunized CD4+ T cells could not be fully reactivated by exposure to EV71 particles . We further examined the response of VLP-specific CD8+ T cell activation in Ad- and FI-EV71-vaccinated animals . After UV-EV71 restimulation , we stained splenocytes with fluorescence dye-conjugated antibodies reacting to surface CD8 molecules and intracellular IFN-γ and analyzed the cells by flow cytometry . We found that the number of CD8+IFN-γ+ T cells in Ad-EVVLP-immunized mice ( 6 . 5% ) was higher than in Ad-LacZ- or FI-EV71-immunized mice ( 0 . 9% or 1 . 5% , respectively; Fig 5B ) . These results suggest that Ad-EVVLP activates EV71 VLP-specific cellular immunity . We further assessed the efficacy of Ad-EVVLP in protecting against EV71 infection using the hSCARB2-Tg mice model . One-day-old hSCARB2-Tg mice were primed and s . c . boosted with Ad or FI-EV71 vaccine on Days 1 and 7 , followed by s . c . challenge of 3 × 106 pfu EV71 5746 strain 14 days after birth . Mice were monitored daily for survival . As shown in Fig 6A , mice immunized with 3 × 107 pfu Ad-LacZ or PBS died 8 to 9 days after challenge . In contrast , 75% of the mice survived after receiving as little as 3 × 106 pfu of Ad-EVVLP , and 100% of the mice survived when injected with a 10-fold higher dose of Ad-EVVLP . In comparison , EV71-challenged mice received 0 . 1 μg FI-EV71 vaccine and 100% of the mice survived ( Fig 6A ) , indicating that the protective efficacy of Ad-EVVLP against EV71 infection was comparable to the FI-EV71 vaccine . We further examined the viral loads in different tissues of vaccine-immunized animals followed by viral challenge . We extracted RNA from various organs of EV71-challenged Tg mice on Day 4 post-infection to quantify EV71 transcripts using real time RT-PCR with VP1 region-specific primers . Ad-EVVLP immunization substantially reduced VP1 expression in the brainstem , spinal cord , and muscle , compared to considerably high expression in Ad-LacZ-vaccinated mice ( Fig 6B ) , confirming that Ad-EVVLP can suppress EV71 infection and replication . 3C and 3D are proteins conserved between EV71 and CVs ( A16 , A6 , A10 , and A4 ) that share at least 90% homology in their amino acid sequences ( Table A in S1 Text ) . Technical limitations restricted our ability to obtain recombinant 3D protein , but we were able to generate recombinant 3C protein . We therefore examined whether 3C-specific immunities were induced by Ad-EVVLP vaccination . We collected and assayed serum from mice on Day 7 post-prime-boost s . c . with Ad-EVVLP , Ad-LacZ , or FI-EV71 . Serum from Ad-EVVLP-immunized mice elicited activity against 3C protein in a recombinant 3C-protein-coated ELISA capturing assay . Anti-3C binding activity was not detected in serum from Ad-LacZ- or FI-EV71-immunized mice ( Fig 7A ) . Like antisera obtained from FI-EV71 , antisera from Ad-EVVLP-immunized mice showed no virus neutralizing activity against CVA16 ( Table 1 ) . Moreover , serum from mice primed with 10 μg recombinant 3C protein formulated with complete Freund’s adjuvant ( CFA ) and boosted with the same dose of 3C protein adjuvanted with incomplete FA ( IFA ) at an interval of 14 days elicited 3C-binding activity , but did not neutralize EV71 or CVA16 infection ( Fig C in S1 Text ) . Taken together , these results suggest that the induction of 3C-specific antibody does not contribute to the protection against EV71 infections . We further examined 3C-specific cellular immunity in mice immunized with Ad-EVVLP . We isolated splenocytes on Day 7 after vaccine boost and restimulated them with recombinant 3C protein in vitro and observed of CD4+ and CD8+ T cell activation by flow cytometry ( Fig 7 ) . CD4+ T cells from the Ad-EVVLP-immunized group responding to 3C were activated ( mean = 32 . 6% ) , but there were no or minimally activated splenocytes in the PBS- , Ad-LacZ- , and FI-EV71-immunized mice ( mean = 2 . 3% , 2 . 2% , and 1 . 9% , respectively; Fig 7B ) . Activated CD8+ ( CD8+IFN-γ+ ) T cells corresponding to 3C protein in the Ad-EVVLP-immunized splenocytes were markedly activated ( mean = 3 . 8% ) , in contrast to the minimal CD8+IFN-γ+ T cells in Ad-LacZ ( mean = 0 . 5% ) or FI-EV71 ( mean = 0 . 5% ) and background levels of CD8+IFN-γ+ T cells obtained from mice immunized with PBS buffer alone ( mean = 0 . 6%; Fig 7C ) . These results confirm that Ad-EVVLP can induce CD4+/CD8+ T cell responses against VLP and 3C protein . In addition to the protection against EV71 infection , we investigated whether Ad-EVVLP or FI-EV71 can facilitate hSCARB2-Tg mice in resisting lethal CVA16 challenge . After Ad-EVVLP immunization , 100% of hSCARB2-Tg mice survived , in contrast to 0% survival of hSCARB2-Tg mice that received PBS or Ad-LacZ after CVA16 challenge ( Fig 8A and Table 2 ) . Ad-EVVLP fully protected animals challenged with a 6-fold higher CVA16 dose ( 3 × 106 pfu; Table 2 ) . Immunization with 1 μg FI-EV71 vaccine did not protect hSCARB2-Tg mice against 5x105 pfu CVA16 challenge , leading to 0% survival ( Fig 8B and Table 2 ) . Taken together , these results suggest that the Ad-EVVLP vaccine elicits potent CD4+/CD8+ T cell immune responses to control EV71 and CVA16 , whereas the FI-EV71 vaccine protects against only EV71 challenge . This demonstrated a correlation with the results shown in Table 1 , and the results of phase I clinical trials in which sera from subjects immunized with FI-EV71 vaccine neutralized distinct EV71 genotypes , but could not cross-neutralize CV [21 , 38] .
In previous studies , EV71 subunit vaccines including DNA vaccine and recombinant VP1 protein induced an incomplete immune response and showed lower efficacy [27 , 41] . Oral vaccines , such as those against attenuated Salmonella enterica expressing EV71 VP1 , have demonstrated limited efficacy against EV71 , elevating the survival rates to only 50% after viral challenge [42] . Transgenic tomatoes [43] and peptide vaccines [44] expressing VP1 have also been developed , but the vaccine efficacy has not been assessed in vivo . A denatured virus particle containing formalin as a vaccine ( FI-EV71 ) was tested in a hSCARB2-Tg mice model [16] and in human clinical trials [38] , in which its safety and protective efficacy was demonstrated . A previous study on the development of influenza VLP as a vaccine showed that disrupting the influenza VLP structure abolished humoral immune responses and protective immunity [45] . In addition , the denatured EV71 particle possesses linear epitopes to elicit anti-EV71 antibodies; however , most of them are likely to be nonneutralizing , similar to the case of poliovirus [46] . Loss of the induction of effective neutralizing antibodies may be associated with the loss of antigenic determinants during inactivation , such as denatured EV71 particles by formalin . VLPs expressed in insect cells elicited even lower levels of neutralizing antibody titer , proliferation , and cytokine production in monkeys [47] . This may be due to differential post-translational modification of VLP proteins in nonhuman cells to induce differential immune responses . In contrast , intact VLPs produced from host cells preserve conformation-dependent epitopes , which might enable direct interaction of VLPs with B-cell receptors , activating B cells and antigen internalization through antigen-presenting cells [48] . This triggers potent antibody responses [49] and cross switching through cooperation with stimulated CD4+ T cells [50] . Furthermore , recent studies have shown that neutralizing antibodies , specifically those against the EV71 capsid proteins , cannot cross-protect against CV infection [19 , 21] , indicating that the vaccines currently being developed protect against only EV71-induced HFMD . In this study , we evaluated the potential of adenovirus-expressing EV71 VLP as a vaccine candidate against EV71 and CVA16 infections through comparison with the efficacies and immune responses elicited by Ad-EVVLP and the classical preparation of formalin-inactivated EV71 vaccine . Immunization with Ad-LacZ elicited no EV71-specific antibody titers and low levels of T cell responses , compared to Ad-EVVLP and FI-EV71 vaccines , which strongly induced the anti-EV71 antibody titer ( Fig 3 , Table 1 ) . Antibodies induced by Ad-EVVLP exhibited cross reactivity against the clinically isolated EV71 C2 and B4 genotypes ( Fig 3 ) . In addition to anti-VLP antibody , the Ad-EVVLP vaccination induced anti-3C antibody ( Fig 7 ) . However , we did not observe the neutralizing activity against CVA16 in the serum of Ad-EVVLP- and FI-EV71-immunized mice ( Table 1 ) . This may explain why the anti-3C antibody could not bind to the 3C protein , which was either not expressed or was in the EV71 or CV inner capsid . Previous studies have shown that preexisting anti-adenovirus antibodies do not affect subsequent generations of humoral responses to an antigen expressed through a mucosally administered recombinant adenovirus vector [24 , 25 , 51] . However , Ad-EVVLP oral immunization induced a decreased immune response compared to the mice receiving systemic Ad-EVVLP immunization ( s . c . or i . p . ; Fig 3 and Table 1 ) . Our results showed that the existence of low anti-Ad antibody in sera of vaccine-primed animals ( Fig D in S1 Text ) did not influence the secondary VLP-specific antibody in the sera of mice administered a second dose of Ad-EVVLP orally , s . c . , or i . p . The actual immuno-efficacy of Ad-EVVLP still needs to be assessed in clinical trials . Ad is a strong DC activator , which enzymatically processes and presents antigenic peptides associated with MHC class I and II molecules on the surface , and subsequently coordinates and stimulates T helper and cytotoxic T-cell responses [52] . Ad-EVVLP immunization induced capsid protein-specific cellular immune responses , which was confirmed by the EV71 VLP induction of CD4+ and CD8+ T cell activation ( Fig 5 ) and cytokine production ( Fig 4 ) . Compared to FI-EV71 vaccine immunization that activated Th2-mediated responses [16] associated with IL-4 and IL-13 secretion ( Fig 4 ) , the high IFN-ɣ , IL-4 and IL-13 levels produced by Ad-EVVLP-immunized splenocytes ( Fig 4 ) suggested a mixed Th1/Th2 immune response , which potentiates both the activation of effector cellular responses and antibody production . These results are consistent with the induction of Th1/Th2 immune responses from the VLP of the influenza virus [45] and human papillomavirus [53] . Conversely , the CD4+ and CD8+ T cell activation corresponding to VLP was not observed in the FI-EV71-immunized mice ( Fig 5 ) , indicating that the epitopic antigenicity of VLP in the FI-EV71 vaccine after formalin inactivation was changed from its native form of EV71 VLP . However , structural analysis has shown that FI-EV71 is not different from infectious EV71 virions [13] , and immunogenicity studies have revealed that the formalin-inactivated F- and E-particles of EV71 can induce the neutralizing antibody , even though the F-particle was more potent than E-particles in mice [37] . Thus , the antigenicity of the Ad-EVVLP-expressed VLP compared to the FI-EV71 vaccine VLP in the activation of cellular responses will be investigated in the future . CD4+ and CD8+ T cell-mediated cellular responses corresponding to the recombinant 3C protein in Ad-EVVLP- but not FI-EV71 vaccine-immunized mice was also observed ( Fig 7 ) . We demonstrated that Ad-EVVLP immunization fully protected hSCARB2-Tg mice against EV71 ( Fig 6 ) and CVA16 challenge ( Fig 8 and Table 2 ) . These results suggest that protection against EV71 infection through Ad-EVVLP is mediated by the induction of EV71-VLP-specific neutralizing antibodies , as well as VLP- and 3C-specific cellular immunities . The lower titer of neutralizing antibodies accompanied by higher transmission rates in children and infants indicates that neutralizing antibodies are crucial for the prevention of EV71 infection [54 , 55] . Our study also demonstrated that challenge of hSCARB2-Tg mice with EV71 followed by VP1 specific monoclonal antibody treatment might prevent EV71-induced pathology [19] . However , serum in 80% of EV71-infected patients contain neutralizing antibodies 1 day after illness onset; the level of antibody titer does not correlate with disease severity [56] . In contrast , cellular immune responses correlate with disease progression and clinical outcome [39 , 57] . Decreased cellular immunity is associated with increased disease severity in EV71 patients , whereas neutralizing antibodies display no difference between mild and severe cases [40] . These studies suggest that cellular immunity might be crucial in the protection against enterovirus infection . Our results showed that the 3C-specific cellular immunity induced by Ad-EVVLP might be sufficient to protect against CVA16 infection ( Fig 8 and Table 2 ) even though Ad-EVVLP did not induce a CVA16-VLP-specific neutralizing antibody ( Table 1 ) . Therefore , we constructed Ad-3CD only expressing the 3CD gene and immunized hSCARB2-Tg mice followed by EV71 or CVA16 challenge . The results showed that Ad-3CD fully protected animals from EV71 and CVA16 challenges ( Fig E in S1 Text ) . They indicate that 3CD-specific cellular immunities are sufficient to provide protection against EV71 and CVA16 infections . Therefore , we will further examine 3C and/or 3D-specific immune responses induced by the Ad-3CD vaccine and the identification of CD4 and CD8 epitopes in 3C and 3D proteins . In conclusion , VLP expression in host cells through the replication of defective adenovirus mimicking the natural structure of EV71 particles induced antibodies against VLP and 3C proteins and cellular immunities specific to VLP and 3C proteins . Because the 3C protein is highly conserved between EV71 and CVA ( Table A in S1 Text ) , we demonstrated that Ad-EVVLP acts as a multivalent vaccine to suppress EV71 and CVA16-induced disease . We achieved several breakthroughs in the development of a medically necessary enterovirus vaccine . First , instead of the subunit EV71 vaccine , inactivated EV71 vaccine , or protein-typed VLPs that protect against only EV71-induced HFMD , Ad-EVVLP prevents EV71- and CVA-induced HFMD . Second , induction of 3C-specific cellular immunity might sufficiently protect against CVA infection . | The spread of enterovirus-induced HFMD could be controlled through a robust vaccination program . Formalin-inactivated EV71 ( FI-EV71 ) vaccines have been evaluated in human clinical trials in China , Taiwan and Singapore and were found to be safe and to elicit strong neutralizing antibody responses against EV71 , which is currently circulating in Asia . However , the results from recent three phase III clinical trials performed in young children indicate that the lack of efficacy against CVA16 infections is a major challenge to the current EV71 vaccine and future HFMD vaccine development . In this study , we developed an adenovirus-based vaccine , Ad-EVVLP with the EV71 P1 and 3CD genes to express VLPs . Ad-EVVLP immunization induced EV71-specific neutralizing antibodies and Th1/Th2-balanced cellular responses in mice . Ad-EVVLP provided protection against both EV71 and CVA16 challenges in the hSCARB2-Tg mice model , whereas FI-EV71 vaccine activated only Th2-mediated EV71 neutralizing antibody responses to protect against EV71 challenge . Because Ad-EVVLP vaccination-induced antibodies had no virus neutralizing activities against CVA16 , the cross-protective immunity against CVA16 was mediated by conserved 3CD-specific cellular immunity activation . These results indicate that Ad-EVVLP meets a medical need as a universal HFMD vaccine against both EV71 and CV infections . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Recombinant Adeno-Vaccine Expressing Enterovirus 71-Like Particles against Hand, Foot, and Mouth Disease |
Microbiome studies suggest the presence of an interaction between the human gut microbiome and soil-transmitted helminth . Upon deworming , a complex interaction between the anthelminthic drug , helminths and microbiome composition might occur . To dissect this , we analyse the changes that take place in the gut bacteria profiles in samples from a double blind placebo controlled trial conducted in an area endemic for soil transmitted helminths in Indonesia . Either placebo or albendazole were given every three months for a period of one and a half years . Helminth infection was assessed before and at 3 months after the last treatment round . In 150 subjects , the bacteria were profiled using the 454 pyrosequencing . Statistical analysis was performed cross-sectionally at pre-treatment to assess the effect of infection , and at post-treatment to determine the effect of infection and treatment on microbiome composition using the Dirichlet-multinomial regression model . At a phylum level , at pre-treatment , no difference was seen in microbiome composition in terms of relative abundance between helminth-infected and uninfected subjects and at post-treatment , no differences were found in microbiome composition between albendazole and placebo group . However , in subjects who remained infected , there was a significant difference in the microbiome composition of those who had received albendazole and placebo . This difference was largely attributed to alteration of Bacteroidetes . Albendazole was more effective against Ascaris lumbricoides and hookworms but not against Trichuris trichiura , thus in those who remained infected after receiving albendazole , the helminth composition was dominated by T . trichiura . We found that overall , albendazole does not affect the microbiome composition . However , there is an interaction between treatment and helminths as in subjects who received albendazole and remained infected there was a significant alteration in Bacteroidetes . This helminth-albendazole interaction needs to be studied further to fully grasp the complexity of the effect of deworming on the microbiome . ISRCTN Registy , ISRCTN83830814 .
Shortly after birth , the human body is colonized by a community of bacteria [1 , 2] with relatively simple composition which increase in number and complexity with age [3] . The densest colonization with commensal microbes of the human body is found in the intestine [4] which has a beneficial impact on gastro-intestinal function and host health by providing support for host metabolism , protection against pathogenic microbes , integrity of intestinal mucosa , and modulation of the immune system [2 , 3 , 5] . Furthermore , it has been shown that intestinal microbiota is associated with dietary habits [6 , 7] , physiological factors such as age , gender and BMI [8 , 9] as well as diseases , such as inflammatory bowel disease and obesity [1 , 5 , 10] . Apart from intestinal microbiota , certain pathogens such as soil-transmitted helminths ( STH ) may coexist in the human intestine . It is estimated that STH , largely represented by Ascaris lumbricoides , hookworm such as Necator americanus and Ancylostoma duodenale , and whipworm Trichuris trichiura , infect 2 billion people in the majority of developing countries and mostly children [1 , 11] . These infections have been reported to cause impairments in physical , intellectual , and cognitive development [12] . At the same time , these parasitic worms have a long co-evolutionary interaction with their host . The result of this co-evolutionary trajectory , seems to be that helminths lead to immune regulatory responses that allow their long term survival within their host [13 , 14] . Since intestinal microbiota and helminths share the same niche in their host , it is hypothesized that the presence or absence of intestinal helminths may affect their interaction with each other within the host . In an interesting study , evidence was provided for the beneficial effects of the microbiome on successful completion of whipworm life cycle [15] . Currently , there is also much interest to determine whether helminth infections affect the gut microbiome and whether the effects of worms on human health is mediated via alteration in the microbiome composition . It is becoming increasingly clear that the gut microbiota has important link to the immune system and several disease outcomes . With the mass drug administration programs underway to eliminate intestinal helminths in many endemic regions , it is essential to fully understand the consequences of deworming on community health by characterizing the effect on the gut microbial composition . Recently , several studies investigated the relationship between the intestinal microbiome and intestinal helminth infections . In swines , a statistically significant association between Trichuris infection and the gut microbiome composition was shown [16 , 17] , evident from the altered abundance of the genus Paraprevotella and phylum Deferribacteres in the infected pigs . The chronic infection of Trichuris muis in C57BL/6 wild-type mice increased the relative abundance of Lactobacilli [18] , while giving T . trichiura ova to macaques with chronic diarrhea increased the phylum Tenericutes and resulted in clinical improvement [19] . Therefore , in animal models , Trichuris infection seems to be associated with alternation in the gut microbiome . However , in humans , findings are not consistent . In an observational study in Ecuador , comparing the gut microbiome of infected and uninfected school children , no significant differences at various taxonomical levels were found [20] . On the contrary , two other observational studies in rural villages of Malaysia [21] and Zimbabwe [22] found a significant increase in diversity and abundance of certain bacteria taxa in infected compared to uninfected subjects . An increase in Paraprevotellaceae was seen in the Malaysian study , which seemed to be associated with Trichuris infection while an increase in Prevotella was reported in the study in Zimbabwe that was attributed to S . haematobium infection . Furthermore , in an interventional study carried out in another rural village in Malaysia [23] , a significant change in order Bacteroidales and Clostridiales was observed after deworming while deworming of S . haematobium in an interventional study in Zimbabwe [22] did not seem to alter the microbiome . The study designs which were used to investigate the human-gut microbiome in relation to helminth infections were either observational [20–22] or interventional without a control group [20–23] hampering the estimation of the true relationship between helminth infection and the microbiome composition . Motivated by the findings from previous studies of helminths on microbiome , we used samples from a larger randomized placebo-controlled trial of albendazole treatment in a population living in an area endemic for soil transmitted helminth infections [24] to further characterize the effect of helminth infection and treatment at before and 21 months after treatment . The study design allowed the investigation of the effect of helminths on the fecal microbial community through comparing helminth infected and uninfected at baseline and subsequently assessing the effect of treatment with albendazole . We also explored the effect of the interaction between treatment and infection status on the faecal microbiome . In addition , we used the opportunity to assess whether albendazole has a direct effect on the microbiome by analyzing those who received albendazole and were uninfected throughout the study . The placebo group enables the estimation of the effect of deworming on the microbiome composition in the absence of anthelminthic treatment which itself could affect the microbiome . The analyses carried out in this study aim to characterize the joint effects of several predictors , such as helminth infection and treatment on each bacterial category . For comparing the gut microbiome of premature infants with different severities of necrotizing enterocolitis , a Dirichlet—multinomial model was used [25] . Here , we consider the same approach for modelling and hypothesis testing for the association between treatment and helminth infection on microbial composition at the phylum level . Our approach addresses the possible correlation between bacteria categories , the compositional feature of the microbiome data [26] , and the multiple testing issue .
This study was nested within the ImmunoSPIN study , a double blind placebo-controlled trial conducted in Flores Island , Indonesia [24] . The ImmunoSPIN study has been approved by the Ethical Committee of Faculty of Medicine , Universitas Indonesia , ref:194/PT02 . FK/Etik/2006 and has been filed by ethics committee of the Leiden University Medical Center . The clinical trial was registered with number: ISRCTN83830814 in which the protocol for the trial and supporting CONSORT checklist are available elsewhere [27] . The subjects gave their informed consent either by written signature or thumb print . Parental consent was obtained for children below 15 years old . Households were randomized to receive either a single dose of 400 mg albendazole or placebo once every 3 months for 2 years . To assess the effect of treatment on the prevalence of soil transmitted helminth infection , yearly stool samples were collected on a voluntary basis . T . trichiura infection was detected by microscopy and a multiplex real time PCR was used for detection of hookworm ( A . duodenale , N . americanus ) , A . lumbricoides and Strongyloides stercoralis DNA . For the current study , paired DNA samples before and at 21 months after treatment from 150 inhabitants in Nangapanda were selected based on the treatment allocation and infection status as well as the availability of complete stool data at pre and post-treatment ( Fig 1 ) . The procedure for sample collection and processing is already described elsewhere [24] . Briefly , prior to DNA isolations , approximately 100 mg unpreserved faeces ( kept at -20°C ) were suspended in 200μl PBS containing 2% polyvinylpolypyrolidone ( PVPP;Sigma , Steinheim , Germany ) . Suspensions were heated at 100°C for 10 min and were treated subsequently with sodium dodecylsulphate-proteinase K at 55°C for 2 h . DNA was isolated using QIAamp DNeasy Tissue Kit spin columns ( QIAgen , Venlo , The Netherlands ) . The whole procedure of DNA isolations and setup of PCR plates were performed using a custom-made automatic liquid handling station ( Hamilton , Bonaduz , Switzerland ) . As published already , sequences of the A . lumbricoides and N . americanus-specific primers and probes as well as the A . duodenale specific XS-probes were used to accommodate the specific fluorophor combinations of the CFX real-time PCR system ( S1 Table ) [24 , 28] . The real-time PCRs were optimized first as monoplex assays with 10-fold dilution series of A . duodenale , N . americanus and A . lumbricoides DNA , respectively . The monoplex realtime PCRs were thereafter compared with the multiplex PCR with the PhHV internal control . The cycle threshold ( Ct ) values obtained from testing the dilution series of each pathogen in both the individual assay and the multiplex assay were similar , and the same analytical sensitivity was achieved . Amplification reactions were performed in white PCR plates in a volume of 25μl with PCR buffer . Amplification consisted of 15 min at 95°C followed by 50 cycles of 15 s at 95°C , 30 s at 60°C , and 30 s at 72°C . Amplification , detection , and analysis were performed with the CFX real-time detection system ( Bio-Rad laboratories ) . The PCR output from this system consists of a cyclethreshold ( Ct ) value , representing the amplification cycle in which the level of fluorescent signal exceeds the background fluorescence and reflecting the parasite-specific DNA load in the sample tested . In this manuscript , we set the ct value 30 as a threshold for the infection status i . e . subjects with PCR lower than 30 was identified as clearly infected and PCR above 30 as uninfected or very low infection . The analyses were carried with regard to the infection status and we do not consider the analysis in the level of infection . Genomic DNA samples were isolated from 100 mg of fresh stool , which were also used for detection of helminth infection by real time PCR . The DNA amplification and pyrosequencing followed the protocols developed by the Human Microbiome Project ( HMP ) [29] at the McDonnell Genome Institute , Washington University School of Medicine in St . Louis . Briefly , The V1-V3 hypervariable region of the 16S rRNA gene was amplified by PCR and the PCR products were purified and sequenced on the Genome Sequencer Titanium FLX ( Roche Diagnostics , Indianapolis , Indiana ) , generating on average 6 , 000 reads per sample . The filtering and analytical processing of 16S rRNA data for this cohort has been previously described in details [30] . The assembled contigs count data as a result of RDP classification was organized in matrix format with taxa in columns and subjects in row . The entries in the table represent the number of reads for each phyla for each subject . Rarefaction to 2000 reads was performed using an R package ( vegan ) [31] . We obtained the count data of 609 bacterial genera and 18 bacteria phyla . In the analysis at phylum level , we retained the 5 most prevalent phyla ( Actinobacteria , Bacteroidetes , Firmicutes , Proteobacteria , and Unclassified Bacteria ) and pooled the remaining phyla into a pooled category such that there are only 6 phyla categories . The Unclassified bacteria represents the category where all the sequences cannot be assigned into a phylum . We conducted further analyses by decomposing the statistically significant phylum ( Bacteroidetes ) into the two most prevalent genera ( Bacteroides and Prevotella ) and the remaining genera into a pooled Bacteroidetes category and combining the Proteobacteria , Unclassified Bacteria and Pooled in a Pooled Phyla category . In total we have six categories since we also selected Actinobacteria and Firmicutes at phylum level . The within sample diversity ( Shannon and richness diversity ) indices as well as the between sample diversity ( Bray-Curtis distance ) were computed at baseline and follow-up using the dataset at genera level . Clustering of samples and bacteria was studied by plotting a heat map of bacteria genera which were present in at least one sample and which had an average relative abundance of more than 1% . This cutoff was chosen to exclude rare genera . Unless stated otherwise , the rest of the analyses were done at the phylum level . A Pearson’s chi-squared test statistic was used to test for differences of infection prevalence between the two treatment groups at pre and at post-treatment . Although the study design allows for the pairwise analysis , unfortunately no method is available for multivariate categorical count data . For this reason , we used the Dirichlet-multinomial regression where the characterization of infection and treatment are similar to the interpretation in loglinear model . Each count outcome within a category was assumed to follow the negative binomial distribution . This distribution is the result of a Poisson distribution for counts with the additional assumption that the underlying parameter is a random variable which follows the conjugate distribution ( Gamma ) . By assuming that the underlying parameter was random , the presence of overdispersion due to multiple counts observed within a sample was modelled . To incorporate the fact that the total count is fixed per sample , we conditioned the probability of the multivariate count outcome on the total count per sample . This model is equivalent to the approach of Guimaraes and Lindrooth [32] , i . e . the Dirichlet-multinomial regression model . The model parameters are log of odds ratios which compare the prevalence rate of each bacteria phyla associated with the covariates with the reference category . In all analyses , Firmicutes was used as reference since it has the highest abundance among the phyla . The covariates were infection status and treatment allocation which are both binary variables . The likelihood ratio statistic was used to test the null hypothesis of no effect of the covariate on the microbiome composition . The test statistic follows asymptotically a χ2 distribution with J degrees of freedom , representing the J − 1 bacterial comparison with the reference and one overdispersion parameter . As the Dirichlet—multinomial regression is available for cross-sectional setting , we modelled the association between microbiome composition and covariates including treatment at 21 months after treatment . First , we modelled the association between treatment and microbiome composition by including all study participants . Next , we selected subjects who were infected with at least one single helminth at baseline and included a categorical variable representing the four combinations of treatment allocation and infection status at post-treatment in the model . The R package MGLM [33] was used for analyses . The results were reported in terms of odds ratios , 95% confidence intervals and p-values . To confirm our finding with this method , we used the univariate pairwise analysis for single bacterial categories of interest in albendazole arm . For this purpose , the inverted beta binomial test was applied to test the null hypothesis that the relative abundance of certain bacteria category at pre-treatment is similar to the relative abundance at post-treatment . Note that the inverted beta-binomial regression model is only defined for two categories and is equivalent to the Dirichlet-multinomial . The R package ibb [34 , 35] was used for this test . All computations were conducted in R version 3 . 1 . 0 [36] .
At baseline , 94 out of 150 ( 62 . 7% ) individuals were infected with one or more helminth species , and hookworm was the most dominant species ( 52 . 1% ) followed by T . trichiura ( 44 . 7% ) and A . lumbricoides ( 37 . 2% ) . The baseline characteristics such as age , gender , and helminth prevalence were similar between the two treatment arms although the prevalence of N . americanus was slightly higher in albendazole group , but not statistically significant ( Table 1 ) . The additional relevant characteristics of the participants are listed in S2 Table . With regard to the microbiome composition , the proportions of each bacterial phyla were also similar between two treatment arms with the highest abundance at the phylum level being Firmicutes followed by Actinobacteria , Proteobacteria and Bacteroidetes . At 21 months after treatment , the prevalence of STH infection was 21 . 7% in the albendazole arm and 54 . 3% in placebo arm ( p-value < 0 . 001 ) . Albendazole had the greatest effect on hookworm ( 24 . 7% ( placebo ) vs 4 . 3% ( albendazole ) ) followed by A . lumbricoides ( 28 . 4% ( placebo ) vs 4 . 3% ( albendazole ) ) and lastly T . trichiura ( 28 . 4% ( placebo ) vs 15 . 9% ( albendazole ) ) . These percentages are similar to what was seen in the whole ImmunoSPIN trial [24] . These data show that while infections with A . lumbricoides and with hookworms decrease at post-treatment , the infections with T . trichiura was not affected much by albendazole and therefore the proportion of individuals infected with T . trichiura increased when considering those that remained infected at post-treatment ( Fig 2 ) . In the placebo group , there was no such difference in the composition of helminth species at post-treatment . It was noted that 12 ( 2 from albendazole and 10 from placebo ) out of 56 uninfected subjects at baseline ( 21 . 4% ) gained helminth infection over the study time period . Using bacterial data at the genus level ( a total of 609 genera ) , we calculated the within sample diversity ( richness and Shannon index ) and between sample diversity ( Bray-Curtis dissimilarity ) . We observed a similar within-sample diversity at pre and post-treatment as evident from the Shannon diversity index ( 2 . 99 vs 2 . 96 ) and the richness index ( 66 . 17 vs 62 . 16 ) . The Bray-Curtis dissimilarity measures the percentage of similarities between two samples in a community and the values range from 0 ( completely similar ) to 1 ( completely dissimilar ) . As reported earlier [30] , the Bray-Curtis dissimilarities calculated from 150 subjects at pre-treatment was 0 . 61 and the same average was obtained when calculating the Bray-Curtis dissimilarities at post-treatment , indicating that in average there was 61% dissimilatory percentages between each pairs of samples . When stratifying all samples based on infection status at pre-treatment and on randomization arm at post-treatment , again we observed similar beta-diversities , indicating that neither infection nor treatment induced a shift in diversity . When analyzing the genera in relation to infection status rather than treatment , the average Shannon diversity index as well as the average richness was similar between the infected and the uninfected group at pre-treatment and post-treatment ( S1 Fig ) . The average relative abundances of all bacterial genera at both time-points were below 10% , with the highest being in the phylum Firmicutes , specifically the genus Catenibacterium ( 6 . 7% at pre-treatment ) and the unclassified genus belonging to the family Ruminococcaceae ( 5 . 6% at post-treatment ) . The relative abundance at the genus level as well as the dominant genera vary between populations as observed in studies where samples in rural Ecuador [20] or Malaysia were compared with the US [21] or in studies where samples of healthy European and American adults were analysed [37] . To illustrate the bacterial genera profile in relation to infection and treatment status , we selected the 29 genera ( at pre and post-treatment ) with an average of relative abundance across all samples larger than 1% . Genera from phylum Firmicutes are the most dominant ( 21 of 29 genera belongs to Firmicutes ) . As shown in heatmaps based on composition of the most prevalent genera , no significant clustering could be seen , neither at the level of bacteria nor at the level of individuals ( Fig 3A and 3B ) in relation to helminth infection or treatment , which indicates that neither helminths nor treatment affected the predominant genera in the gut . Using the Dirichlet-multinomial regression model , we observed that there was no difference on the microbiome composition at the phylum level when subjects with any helminth infection were compared with uninfected ones either at pre ( Fig 4A ) or at post treatment ( Fig 4B ) time points . The same was the case when infection with a specific helminth species was considered ( Fig 4A and 4B ) . The Dirichlet-multiomial regression model was also used to discern the effect of helminths and treatment on the microbiome data at post treatment . Six bacterial categories were considered in the analyses with Firmicutes used as a reference . The effect of treatment on microbiome composition in all individuals irrespective of whether they were infected or not at post-treatment was not significant . No differences were observed between placebo and albendazole at post-treatment ( p-value = 0 . 305 , Table 2A , likelihood ratio test ) . We further selected subjects who were infected at baseline ( N = 94 ) and characterized their microbiome composition at post-treatment with regard to their infection status and treatment arm , namely: subjects who lost their infection either in the albendazole ( group 1 , N = 34 ) or placebo arm ( group 2 , N = 13 ) , and subjects who remained infected in either the albendazole ( group 3 , N = 13 ) or placebo arm ( group 4 , N = 34 ) . We compared the microbiome composition of the first three groups to the group of remained infected in the placebo arm ( group 4 ) as the latter group were neither influenced by treatment nor the changing of infection status . When subjects who were infected at pre-treatment and lost their infection in the albendazole arm were compared to subjects who remained-infected in placebo group , no differences were observed ( p-value of 0 . 371 , Table 2B ) , indicating that removing helminths with albendazole did not change the microbiome profile at a phylum level . Furthermore , in subjects who lost their infection in the placebo arm , there was a trend for decrease in Bacteroidetes and pooled category ( OR 0 . 49 , 95% CI: ( 0 . 27 , 0 . 91 ) and OR 0 . 47 , 95% CI: ( 0 . 23 , 0 . 96 ) , respectively , Table 2B ) , moreover , the whole composition in this group did not differ significantly from that in the group of remained infected in the placebo arm ( p-value of 0 . 069 ) . These two comparisons suggest that removing helminths regardless of treatment did not alter the microbiome composition when analysed at a phylum level . Interestingly , the comparison of microbiome composition between subjects who remained infected in the albendazole group was significantly different from the microbial composition in subjects who remained infected in the placebo group ( p-value of 0 . 004 , Table 2B ) . This difference was driven by the increasing odds of having Actinobacteria ( OR 1 . 57 , 95% CI of ( 1 . 05 , 2 . 35 ) ) and the decreasing odds of having Bacteroidetes ( OR 0 . 35 , 95% CI: ( 0 . 18 , 0 . 70 ) ) . To further analyse the direct treatment effect without the influence of helminth infection , we selected subjects who were uninfected at baseline and remained-uninfected at post-treatment ( N = 44 ) . For these subjects , we compared the microbial composition at post-treatment of subjects who received albendazole versus those who received placebo . No difference was observed ( the estimate odds ratios range from 0 . 88 , 95% CI: ( 0 . 56 , 1 . 39 ) to 1 . 42 , 95% CI: ( 0 . 88 , 2 . 29 ) , p-value = 0 . 666 , illustrated in Fig 5 ) , indicating that albendazole alone does not seem to affect the microbiome composition in uninfected subjects when compared at a phylum level . As neither treatment alone nor the infection affected the microbial composition , we further hypothesized that the significant difference in microbiome composition in subjects who remained infected and received albendazole compared to the group that remained infected in the placebo arm was caused by the alteration of the abundance of Actinobacteria and Bacteroidetes during the treatment period . To test this hypothesis , we used the inverted beta-binomial test to compare the relative abundance of Actinobacteria and Bacteroidetes in subjects who remained infected in albendazole group at pre-treatment to the relative abundances of these bacterial phyla at post-treatment . While the relative abundance of Actinobacteria did not change significantly between pre and post-treatment ( p-value of 0 . 155 , inverted beta binomial test ) , the relative abundance of Bacteroidetes was estimated to be 1 . 88 fold higher at pre-treatment compared to post-treatment ( p-value of 0 . 012 , inverted beta binomial test ) . This result indicates that there is a complex interaction between helminths and treatment , which induces a change in bacterial composition during the treatment period . Using the same analysis , the direct effect of albendazole was assessed by comparing subjects who were uninfected but received albendazole at pre treatment and remained uninfected at post treatment . Although some differences were seen in the microbiome composition between pre and post-treatment , specifically in the phyla Actinobacteria , Bacteroidetes and Proteobacteria , these differences were not statistically significant ( p-values of 0 . 149 , 0 . 267 and 0 . 064 , respectively ) . This is in line with the finding when we used the Dirichlet-multinomial regression model where no direct effect of albendazole on the microbiome composition was found . In addition , similar microbiome composition was seen in subjects free of helminth infection at baseline who received placebo and remained uninfected at post-treatment , which suggests that the microbiome was stable over time . In the Dirichlet—multinomial regression analysis carried out at the phylum level , Bacteroidetes was the phyla that showed significant differences in subjects who remained infected in the albendazole arm compared to those who remained infected in the placebo arm . We dissected this further to assess which Bacteroidetes genera accounted for this difference using the Dirichlet-multinomial regression model on 6 bacterial categories which were obtained as follows . The phylum Bacteroidetes was divided into three categories , namely the Bacteroides , Prevotella and pooled Bacteroidetes . The first two genera were chosen as they were the two most abundant in the phylum Bacteroidetes . In the analyses , as 6 categories are needed , we included another three phyla , i . e . , Actinobacteria , Firmicutes and pooled remaining phyla ( pooled Phyla ) . As for the modelling at the phylum level , Firmicutes was used as a reference . Similar to the analyses at the phylum level , we characterized the association of infection and treatment on these 6 bacterial categories that comprised the genera belonging to Bacteroidetes . When considering the whole study subjects irrespective of infection status , there was no difference between albendazole and placebo ( Table 3A ) . When 94 infected subjects at pre-treatment were selected and 6 bacterial categories as above were analysed with regard to infection and treatment , we observed a decrease in odds of having Prevotella in subjects who lost their helminth infection in placebo group ( OR 0 . 44 , 95% CI: ( 0 . 21 , 0 . 90 ) ) compared to subjects who remained infected in placebo group although this fell short of statistically significant ( p-value of 0 . 086 , Table 3B ) . Furthermore , in line with the finding at the phylum level , we also observed a significant difference in microbial composition of subjects who remained infected with albendazole compared to the microbial composition of subjects who remained infected in the placebo group ( p-value of 0 . 016 ) . This alteration was mainly due to the increase in odds of having Actinobacteria ( OR 1 . 54 , 95% CI: ( 1 . 00 , 2 . 35 ) ) and a decrease in odds of having Prevotella ( OR 0 . 44 , 95% CI: ( 0 . 21 , 0 . 94 ) ) , suggesting that the decrease in Bacteroidetes at the phylum level observed in Table 2B was driven by Prevotella .
The 16S rDNA assembled sequences , annotation and abundances from all the Indonesia samples are available for download from Nematode . net ( nematode . net/Indonesia_Microbiome . html ) [48] . ” | Studying the relationship between soil-transmitted helminthiasis and gut microbiota is becoming more important as both have been implicated in modulating immune system in various inflammatory diseases . However , findings of previous studies of the effect of helminth on the microbiome are inconsistent . In this study , an optimal design , a placebo-controlled anthelminthic trial was conducted to dissect the effect of helminths and anthelminthic treatment on gut microbial profile . In addition , a novel statistical model was used to analyse the association by taking into account the correlation structures between bacterial categories by applying multivariate analysis whereby the multiple testing correction is not needed . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"microbiome",
"helminths",
"microbiology",
"hookworms",
"parasitic",
"diseases",
"animals",
"bacterial",
"diseases",
"ascaris",
"ascaris",
"lumbricoides",
"molecular",
"biology",
"techniques",
"bacteria",
"microbial"... | 2018 | Dynamic changes in human-gut microbiome in relation to a placebo-controlled anthelminthic trial in Indonesia |
To investigate factors contributing to drug side effects , we systematically examine relationships between 4 , 199 side effects associated with 996 drugs and their 647 human protein targets . We find that it is the number of essential targets , not the number of total targets , that determines the side effects of corresponding drugs . Furthermore , within the context of a three-dimensional interaction network with atomic-resolution interaction interfaces , we find that drugs causing more side effects are also characterized by high degree and betweenness of their targets and highly shared interaction interfaces on these targets . Our findings suggest that both essentiality and centrality of a drug target are key factors contributing to side effects and should be taken into consideration in rational drug design .
Regardless of their effectiveness , most drugs come with side effects of different types that affect patients' life quality and may even bring up additional health problems . It is estimated that around two million patients suffer from serious drug side effects each year and that the fourth leading cause of death in the United States is severe side effects of medication [1] , [2] . Of the total number of drug candidates failed during clinical trial phases II and III , 20% of these failures are because of safety issues [3] . Hence , evaluating potential side effects of drugs is important in rational drug design and development , as well as successful marketing . Binding of drugs to their on- and off-targets modifies the functions of these targets and therefore is believed to account for their efficacies as well as side effects [4] . Traditionally , properties of a drug such as binding fingerprint and chemical structure are evaluated to anticipate side effects [5] , [6] . Moreover , in vitro assays or phenotypic tests in model organisms may not be able to capture the same spectrum of side effects in human [7] , [8] . Recently , an increasingly accepted view is that integrating biological networks would provide unique insights into understanding disease mechanisms and identifying novel drug targets [9] , . Network-based methods have been explored and successfully applied in finding disease-associated genes and inferring underlying molecular mechanisms [11] , [12] . Similarly , phenotypic responses to drugs can be better rationalized by considering their overall effects in the context of molecular networks . Previous studies have shown that drugs with shared targets or those that are close in the interactome network often share similar side effects [13] , [14] . Also , similar side effect profiles have been used to predict drug-target interactions for potential drug repositioning [13] . Hase et al . examined network degree distribution of different categories of genes and suggested that connectivity is potentially important in inferring drug side effects [15] . However , no actual adverse effect data were used in their study . The relationships between drug target properties , especially in the context of biological networks , and its potential toxicity to human remains unexplored . Here , we systematically investigate major contributing factors of drug side effects , taking into consideration their direct targets and the local network structures of these targets .
We obtained a list of 996 drugs and the associated 4 , 199 side effects from SIDER 2 [16] and analyzed 645 FDA-approved drugs that have at least one known human protein target based on the DrugBank database [17] . Evaluation of severity of adverse effects varies among individuals and is often affected by an individual's underlying health conditions . In general , drugs that cause more side effects tend to have higher likelihood leading to severe outcomes , including death ( Figure 1 ) . Although tremendous efforts have been made on studying drug side effects in the pharmaceutical industry , the number of side effects for FDA-approved drugs significantly increases for those that were approved recently ( Figure S1 ) , indicating the necessity in further studying the contributing factors underlying drug adverse effects . By grouping drugs into the categories of “nutraceutical” , “approved” , and “withdrawn” drugs , we find that , unsurprisingly , the nutraceutical drugs have the least number of side effects ( P-value = 0 . 00023 , when compared to the approved therapeutical drugs; Figure 2A ) , while the withdrawn drugs cause significantly more side effects compared to the approved ones ( P-value = 0 . 04; Figure 2A ) . However , there is no significant difference between the average numbers of targets of the three drug groups ( Figure 2B ) . This indicates that the occurrence of side effects may not simply be explained by the number of targets a drug binds to . To investigate this further , we performed a generalized linear regression with negative binomial distribution for side effects over the number of targets . At first , we observed that the number of side effects significantly correlates with the number of targets ( β = 0 . 045; P-value = 0 . 0033; Figure 2C ) . However , further dissection of properties of drug targets reveals that the positive correlation is due to the presence of essential targets , those drug targets encoded by essential genes . We find that the positive correlation between the number of side effects and that of essential targets is much more significant ( β = 0 . 17; P-value = 1 . 8×10−5; Figure 2D ) . On the contrary , by analyzing drugs with no known essential targets , we find that the positive correlation between the number of side effects and targets no longer holds ( β = 0 . 004; P-value = 0 . 93; Figure 2D; see Figure S2 for the illustration separating the effects of essential and non-essential targets ) . This discovery suggests that it is the number of essential targets , rather than the number of total targets , that governs the occurrence of drug side effects . Moreover , the human interactome network has been demonstrated to be highly valuable in understanding pathogenic mechanisms of many disease genes [9] , since most proteins interact with other proteins to carry out their functions [18] . Therefore , it is also important to assess drug side effects by considering network properties of their targets within the human protein interactome . Here , we examined whether the degree ( number of proteins that directly interact with the targets ) and betweenness ( number of shortest paths going through the targets ) [19] of drug targets in the network contribute to side effects . These are two of the most important network parameters , measuring the centrality of the target proteins within the network . We constructed a high-quality human protein-protein interactome network that consists of 30 , 713 interactions between 8 , 357 proteins and then mapped all the drug targets onto the interactome ( Materials and Methods; the sub-network containing the drug targets is shown in Figure 3A ) . This high-quality human protein-protein interactome network can provide insights into potential toxicity of drugs based on the network properties of their targets . To systematically investigate the relationship between a drug's side effects and its target degree within the interactome network , we focused on drugs with only one non-essential target to separate potential confounding effects of the number of total and essential targets . The results show that the number of side effects correlates significantly with the degree of drug targets ( β = 0 . 31; P-value = 0 . 041; Figure 3B ) . Furthermore , we analyzed the occurrence of side effects with respect to the number of targets that are bottlenecks [19] ( network nodes with betweenness among top 20% ) and found significant positive correlation between them ( β = 0 . 21; P-value = 0 . 0057; Figure 3C ) . This positive correlation is consistent when we set the betweenness cutoff at top 5% , 10% , and 40% for identifying bottleneck proteins ( Figure S3 ) . This observation indicates that the centrality of drug targets in biological networks also plays a key role in producing various side effects . We further partitioned the drugs into cancer and non-cancer drugs and repeated the calculations for essentiality and centrality that we presented above . We found the same conclusions for both cancer ( Figure S4 ) and non-cancer drugs ( Figure S5 ) . Our recent study has shown that reconstructing the human protein interactome into a three-dimensional ( 3D ) structurally resolved network can provide insights into molecular mechanisms of disease genes and their mutations [12] . To understand distinct perturbations of the interactome network by various drugs , we then examined the properties of their targets within the framework of our 3D-interaction network . The structural details in this 3D-interaction network allow us to distinguish the effects of drug targets with distinct binding interfaces ( i . e . , multi-interface targets , which bind their different interaction partners at different interfaces ) and those with a common interface ( i . e . , single-interface targets , which bind their different partners at the same interfaces ) [20] . We hypothesize that more adverse effects are expected for a single-interface target due to a higher likelihood of altering all of its interactions by a drug disrupting its only interaction interface . By analyzing side effects of a drug with the proportion of shared interaction interfaces of each drug target with its interaction partners , we observe that the number of side effects increases significantly with the proportion of shared interaction interfaces on a target ( β = 1 . 5; P-value = 0 . 00014; Figure 3D ) . This observation confirms our hypothesis that single-interface targets are likely to cause more side effects than multi-interface ones . We show that this finding is not due to potential biases contributed by hubs or bottlenecks since these nodes tend to have smaller proportions of shared interaction interfaces ( Figure S6 ) . We further identified genes associated with human genetic disease and mapped them onto our human protein interactome network [12] . We calculated the average shortest distances between drug targets and disease-associated genes to represent potential molecular steps needed for a drug to affect the corresponding disease module/pathway . We find that although there is an enrichment of shorter distance between drug targets and their “indicated disease” genes , the distribution largely overlaps with that of distance between targets and unrelated disease genes ( Figure 4A ) . Furthermore , the drugs that fail to specifically interfere with the disease-associated module/pathway result in many more side effects ( Figure 4B ) . This result further demonstrates the importance of incorporating network properties of drug targets and corresponding disease genes in rational drug design and development . In summary , for the first time , we show that the number of essential targets , not the number of total targets , is a determinant of drug side effects . Furthermore , high incidence of drug side effects is also characterized by high degree and betweenness of their targets in the interactome network , as well as highly shared interaction interfaces on these targets . Our findings reveal that both essentiality and centrality of a drug target are important factors to be considered in the drug development pipeline in order to improve the efficiency of this lengthy and costly process . Incorporation of these factors will be useful in the selection of drug candidates at the early stages of the drug development pipeline . When choosing from several drug candidates with similar chemical properties , the one binding to proteins that are not essential and not central in the network would have a higher chance of passing clinical trials later . Moreover , in the efforts of computationally predicting drug side effects [21] , the inclusion of target essentiality and centrality as additional features would also improve the prediction performance . Furthermore , our results can serve as guidance for minimizing side effects in clinical applications , especially when prescribing multi-drug cocktails , which have been proven to be much more effective than single drug approaches [22] . With the increasing coverage of the protein-protein interaction network in human and the accessibility of interactions of high confidence levels [23] , more interesting analyses can be performed to further dissect the properties of drug targets and the associated side effects . This study of adverse effects of drugs within the framework of the protein-protein interactome network demonstrates that network-based pharmacology is of great importance in the field of drug development and application .
We downloaded 4 , 199 side effects associated with 996 drugs from the SIDER database release 2 [16] . For the drugs in SIDER 2 , we mapped them based on the generic drug names or PubChem IDs [24] to the DrugBank database [17] downloaded on November 6 , 2011 , and extracted all of their direct binding human protein targets ( 647 in total ) with available uniprot IDs . We did not differentiate on- and off-targets in all of our analyses with the rationality that they could all potentially produce side effects when bound by the corresponding drugs . Furthermore , we downloaded the database containing the approval dates for each drug from the Drugs@FDA database ( http://www . accessdata . fda . gov/scripts/cder/drugsatfda/ ) and the Orange Book ( http://www . accessdata . fda . gov/scripts/cder/ob/eclink . cfm ) . The earliest approval date was used when a drug had a history of multiple approval events . We then cross-checked the list with the ones reported by Rask-Andersen et al . [25] and removed the drugs with conflicting dates . A list of essential genes was obtained by taking the union of the human orthologs of mouse genes that result in embryonic or postnatal lethality when disrupted [26] and the genes reported as essential from a large-scale RNAi screen in human mammary cells [27] . A drug target that belongs to the essential gene list is abbreviated as an “essential target” . To find key factors contributing to the incidence of side effects , we performed a series of generalized linear regressions based on negative binomial distribution for side effects with the following probability density function:with mean μ and shape parameter θ . The expected value and variance for the number of side effects are: This model is used because we observed over-dispersion with Poisson distribution , which is normally modeled for count data . The generalized linear regressions were built using the log-link function:where X is the independent variable ( such as the number of targets ) , β is the unknown parameter , and is the linear predictor . To minimize the effects of extreme observations , we used median numbers of side effects as response variables for regression analysis . For each regression , we obtained a P-value for the effect of a tested factor based on the hypothesis testing: H0: β = 0 ( there is no effect of the tested factor ) vs . HA: β≠0 ( the incidence of side effects is contributed by the factor ) . Due to the lack of data points , a few observations at the margin were binned together . We first fitted regression for the number of side effects over that of total targets and that of essential targets . To distinguish the effect of total targets and essential targets on the incidence of side effects , we repeated the regression analysis on the drugs that do not have any essential targets . We compiled a list of human protein-protein interactions combining high-throughput high-quality yeast two-hybrid interaction datasets [28]–[31] with six major protein-protein interaction databases [32]–[37] . Since literature-curated interactions could contain low-quality interactions [38] , [39] , we filtered the dataset by applying the criteria that each interaction has to be either from a high-throughput high-quality experiment or supported by at least two independent publications . The interactome network contains 30 , 713 binary and co-complex interactions between 8 , 357 proteins . To evaluate network properties of drug targets , we mapped them to the high-quality protein-protein interactome network and calculated their network properties . To reconstruct the three-dimensional ( 3D ) structurally resolved network , we further filtered the interactions with binary evidence codes , since the concept of interaction interface does not apply when two proteins do not bind each other directly [12] . We then constructed the 3D-interaction network based on known co-crystal structures in the Protein Data Bank ( PDB ) [40] using a homology modeling approach as described earlier [12] . This approach has been demonstrated to be very effective and accurate in inferring protein-protein interaction interfaces [12] . The resulting structurally resolved protein interactome is composed of 6 , 594 interactions between 3 , 630 proteins . We compiled a list of diseases for each drug based on the “indication” field from the DrugBank database . For each drug , we then obtained the disease-associated genes for these diseases from the disease-gene association map we compiled earlier based on OMIM and HGMD databases [12] , [41] , [42] . We then calculated the average shortest distance on the binary interactome network for 1 ) pairs of target proteins and the genes associated with the “indicated” diseases and 2 ) pairs of target proteins and all other disease-associated genes ( Figure 4 ) . For each drug target protein T that can be mapped to the structurally resolved network with at least two interaction partners , we measured the proportion of shared interaction interfaces by calculating the Jaccard similarity coefficient [43]:where is the number of interacting domains on drug target protein T involved in both T-A and T-B interactions , and is the number of interacting domains involved in either T-A or T-B interaction . The mean of the Jaccard similarity coefficient was taken when a target protein has more than two interaction partners . To minimize potential confounding effects of essentiality , we analyzed the drugs with only one non-essential target to evaluate the effects of shared interaction interfaces of a drug target on the number of side effects . While the vast majority of drugs have average distances between their targets and corresponding disease genes comparable to network mean distance ( mean distance = 4 . 4 ) , there are some drugs enriched with much smaller distances ( distance<3; Figure 4A ) . We categorized the drugs into two classes using an average distance of 3 as cutoff to compare the median number of side effects . We carried out the bootstrapping approach to evaluate the difference of median number of side effects due to the observation of extremely unequal sample sizes ( 12 drugs with distance less than 3 and 319 drugs with distance equal to or bigger than 3 ) and variances between the two classes . For each drug class , we randomly sampled 10 observations with replacement and generated the median of these observations . The procedure was repeated 1000 times to obtain distributions of median number of side effects for each of the two drug classes . Then the Wilcoxon rank-sum test was used to evaluate the differences of median drug side effects between the two drug classes ( Figure 4B ) . By randomizing the protein-protein interactions , the disease gene sets , and the drug target sets , we demonstrated that the observation is not due to potential biases in the data ( Figure S7 ) . | The ultimate goal of medical research is to develop effective treatments for disease with minimal side effects . Currently , about 20% of drug candidates failed at clinical trial phases II and III due to safety issues . Therefore , understanding the determining factors of drug side effects is of paramount importance to human health and the pharmaceutical industry . Here , we present the first systematic study to uncover key factors leading to drug side effects within the framework of the human protein interactome network . Our results show that it is the number of essential targets , not the number of total targets , of a drug that determines the occurrence of its side effects . Furthermore , we find that the centrality , both degree and betweenness , of the drug targets is also an important determining factor of drug side effects . Our findings will shed light on new factors to be incorporated into the drug development pipeline . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"biology",
"computational",
"biology"
] | 2013 | Target Essentiality and Centrality Characterize Drug Side Effects |
Two of the crucial aspects of human immunodeficiency virus ( HIV ) infection are ( i ) viral persistence in reservoirs ( precluding viral eradication ) and ( ii ) chronic inflammation ( directly associated with all-cause morbidities in antiretroviral therapy ( ART ) -controlled HIV-infected patients ) . The objective of the present study was to assess the potential involvement of adipose tissue in these two aspects . Adipose tissue is composed of adipocytes and the stromal vascular fraction ( SVF ) ; the latter comprises immune cells such as CD4+ T cells and macrophages ( both of which are important target cells for HIV ) . The inflammatory potential of adipose tissue has been extensively described in the context of obesity . During HIV infection , the inflammatory profile of adipose tissue has been revealed by the occurrence of lipodystrophies ( primarily related to ART ) . Data on the impact of HIV on the SVF ( especially in individuals not receiving ART ) are scarce . We first analyzed the impact of simian immunodeficiency virus ( SIV ) infection on abdominal subcutaneous and visceral adipose tissues in SIVmac251 infected macaques and found that both adipocytes and adipose tissue immune cells were affected . The adipocyte density was elevated , and adipose tissue immune cells presented enhanced immune activation and/or inflammatory profiles . We detected cell-associated SIV DNA and RNA in the SVF and in sorted CD4+ T cells and macrophages from adipose tissue . We demonstrated that SVF cells ( including CD4+ T cells ) are infected in ART-controlled HIV-infected patients . Importantly , the production of HIV RNA was detected by in situ hybridization , and after the in vitro reactivation of sorted CD4+ T cells from adipose tissue . We thus identified adipose tissue as a crucial cofactor in both viral persistence and chronic immune activation/inflammation during HIV infection . These observations open up new therapeutic strategies for limiting the size of the viral reservoir and decreasing low-grade chronic inflammation via the modulation of adipose tissue-related pathways .
Human immunodeficiency virus ( HIV ) infection is characterized by massive CD4+ T cell depletion in the intestinal mucosa ( progressively affecting blood and lymphoid CD4+ T cells ) and sustained systemic immune activation and inflammation . The advent of antiretroviral therapy ( ART ) has drastically changed the outcomes of HIV infection by enabling a reduction in the viral load and the restoration ( at least in part ) of CD4+ T cell counts . In people receiving ART , chronic HIV infection is characterized by the persistence of viral reservoirs ( preventing the eradication of HIV ) and chronic immune activation and inflammation ( which are associated with all-cause , non-AIDS-related morbidity , such as cardiovascular disease and non-insulin dependent diabetes , and mortality [1–3] . Similar observations ( i . e . viral persistence and low level immune activation and inflammation ) apply–albeit to a lesser extent—to “HIV-controllers” , i . e . patients who are able to spontaneously control viral load [4 , 5] . The eradication or reduction of viral reservoirs remains a crucial therapeutic objective in the fight against HIV [6] , and both cellular and anatomical reservoirs require further investigation [7 , 8] . A second therapeutic objective is to circumscribe the sustained immune activation . It has been suggested that microbial translocation is a potent factor in the maintenance of chronic immune activation/inflammation [9] , along with viral persistence , CD4+ T cell lymphopenia , Th17 loss , a change in the regulatory T cell balance , disruption of the lymph node architecture , viral co-infection , accelerated ageing , the side effects of some antiretroviral drugs , and individual susceptibility [2 , 10 , 11] ) . HIV-infected patients on ART are not always able to reestablish gut mucosa integrity and/or normal CD4+ T cell counts , and chronic , low-levels immune activation appears to persist [12] . Taken as a whole , these data suggest that ( i ) immune activation and chronic inflammation are driven by multiple factors and ( ii ) targeting several inflammatory mechanisms may achieve better immune restoration . We hypothesized that adipose tissue has an important role in both chronic immune activation/inflammation and viral persistence . In fact , adipose tissue is not merely a metabolic and endocrine organ for lipid storage; it also exhibits strong immune activity: ( a ) adipose tissue is an important site of production of both pro-inflammatory molecules ( such as leptin , IL-6 , MCP-1 ( CCL2 ) , RANTES ( CCL5 ) and TNF-α ) and anti-inflammatory adipokines ( such as adiponectin ) [13–21]; ( b ) the stromal vascular fraction ( SVF ) contains immune cells ( CD4+ T cells and macrophages ) that are potentially important target cells for HIV; and ( c ) a growing body of evidence demonstrates the close relationship between the immune response and metabolic alterations [22 , 23] , such as the recruitment of activated CD8+ T cells and inflammatory macrophages and their participation in inflammation processes ( and then the modifications of the adipose tissue ) in obesity and non-insulin dependent diabetes [24–31] . In the context of obesity , the interplay between adipocytes and immune cells is being actively investigated . The two cell types are clearly “partners in inflammation” as their coordinated action leads to adipose inflammation [32] . Studies of the adipose tissue during HIV infection have essentially addressed the toxicity of certain antiretroviral drugs and their induction of metabolic alterations–even though a direct impact of infection per se was clearly documented by early studies [33 , 34] . Metabolic alterations [34–38] and elevated levels of pro-inflammatory cytokines have been described in both plasma and adipose tissue [39–43] . Previous analyses of adipocytes failed to demonstrate that adipocytes could be infected by HIV in vivo [44]—in contrast to the results of in vitro studies [45–47] . However , it has been shown that the HIV viral proteins Vpr and Nef are present in adipose tissue and have a negative impact on adipose homeostasis [48–50] . These observations provide a strong rationale for reconsidering the impact of HIV infection on adipose tissue by focusing on immune cells rather than adipocytes . We hypothesize that adipose tissue may constitute a neglected partner that drives viral persistence and chronic immune activation via two nonexclusive mechanisms: local infection and the abnormal local activation of immune cells . To assess the putative infection of adipose immune cells more precisely , we first analyzed tissues from chronically SIV-infected macaques . We chose to analyze both subcutaneous adipose tissue ( SCAT ) and visceral adipose tissue ( VAT ) because they differ in terms of metabolic activity and immune cell content [14 , 15 , 17 , 51] . Secondly , we extended these analyses to ART-treated HIV-infected patients . In the present report , we demonstrate that SIV infection is associated with changes in the composition of adipose tissue , such as elevated densities of both adipocytes and stromal vascular cells . Importantly , adipose tissue macrophages and CD4+ and CD8+ T cells exhibited a more intense activation profile ( relative to non-infected animals ) . Furthermore , SIV DNA and RNA was detected in total SVF and in sorted adipose tissue macrophages and CD4+ T cells . We observed similar results in ART-controlled , HIV-infected patients having undergone elective visceral surgery: their SVF samples were positive for HIV DNA . The presence of infected/virus-producing cells within adipose tissue was confirmed by the detection of HIV RNA in tissue sections via in situ hybridization . Lastly , we performed an in vitro reactivation assay on samples from six patients and found that adipose tissue CD4+ T cells were capable of producing replication-competent virus . Taken as a whole , our data show that adipose tissue as a viral reservoir with inflammatory potential .
We first determined the impact of SIV infection on adipocyte density ( as evaluated by microscopy ) ( Fig 1A ) . Adipocyte density in both SCAT and VAT was markedly higher in SIV-infected macaques than in non-infected animals ( median [interquartile range] number of adipocytes per field in SCAT: 51 [31–90] in infected animals and 17 [13–26] in controls , p = 0 . 032; in VAT: 64 [60–95] in infected animals and 31 [21–32] in controls , p = 0 . 0059 ) ( Fig 1B ) . We also counted SVF cells harvested from SCAT and VAT in infected and non-infected animals ( Fig 1C ) . To enable a valid comparison , SVF cell counts were expressed per gram of adipose tissue . In line with previous publications [47 , 52] , we found that non-infected animals had significantly higher numbers of SVF cells in VAT than in SCAT ( p = 0 . 009 ) . Significantly higher SVF cell counts were detected in SCAT and VAT from infected animals than in non-infected animals ( p<0 . 05 for both tissues ) . We next determined the proportion of CD45-expressing cells in SVF . CD45+ cells accounted for a small proportion of SVF cells in both SCAT and VAT recovered from non-infected animals . In animals with chronic SIV infection , we observed significantly lower percentages of CD45+ cells in SCAT SVF ( relative to non-infected animals ) ( Fig 1D ) . A similar trend was detected in VAT . To determine whether these lower percentages of CD45+ cells reflected a fall in CD45+ cell numbers or the recruitment/expansion of CD45- cells , we analyzed the absolute CD45+ cell count . As shown in Fig 1E , groups of infected and non-infected animals did not differ significantly in terms of the number of CD45+ cells recovered from SCAT or VAT; this was suggestive of changes in the numbers of CD45- cells . We thus demonstrated that chronic SIV infection modulated both adipocyte and SVF cells . In the latter cell population , the quantitative alteration essentially concerned CD45- cells . We next looked at whether or not SIV infection was associated with changes in the percentages of CD4+ and CD8+ T lymphocytes within adipose tissue . To this end , we determined the percentage of total T lymphocytes among CD45-expressing cells and the percentages of CD4+ and CD8+ T lymphocytes among CD3-expressing cells ( Fig 2 ) . Adipose tissue T lymphocytes accounted for approximately half of the CD45+ cells within the SVF; SIV-infected and non-infected animals did not differ significantly in terms of the proportion and number of CD3+ cells ( Fig 2A ) . However , SIV infection was associated with differences in CD4+ and CD8+ T cell relative percentages ( Fig 2B and 2C ) . The proportion of CD4+ T cells was significantly lower in SIV-infected macaques ( 17 . 6% [8 . 9–25] in SCAT , and 20 . 4% [11 . 0–28 . 0] in VAT ) than in non-infected animals ( 40 . 0% [26 . 4–49 . 0] in SCAT , and 39 . 8% [23 . 6–48 . 7] in VAT; p = 0 . 0043 for SCAT , p = 0 . 0317 for VAT ) . Conversely , CD8+ T cell percentages were significantly higher in infected animals . The low CD4+ T cell percentages may reflect the CD4+ T cell depletion associated with SIV infection , whereas elevated CD8+ T cell percentages may reflect either a passive increase ( due to CD4+ T cell decay ) , an active increase due to antiviral CD8 recruitment ( as described in lungs after SIV infection [53 , 54] ) , or viral-independent local recruitment ( as described in inflammatory adiposity [27] ) . To evaluate the direct impact of SIV infection on T cell subsets , we analyzed the numbers of CD4+ and CD8+ T cells recovered from adipose tissue of infected and non-infected groups of animals . Surprisingly , the CD4+ T cell number was not lower in infected animals , and there was even a non-significant trend towards CD4+ T cell accumulation in VAT ( Fig 2C ) . In contrast , CD8+ T cell numbers in VAT were significantly higher ( 0 . 48 106 [0 . 29–0 . 67] in SIV-infected animals than in non-infected animals ( 0 . 14 106 cells/g [0 . 11–0 . 24]; p = 0 . 0205 ) . A similar trend was observed for SCAT . Thus , the change in the CD4/CD8 ratio was mainly driven by an increase in CD8+ T cell numbers—a phenomenon that is often associated with inflammatory adiposity [27] . Lastly , we used immunochemical techniques to formally confirm the presence of T lymphocytes in adipose tissue ( Fig 3A ) . Given that the recruitment of peripheral blood cells into adipose tissue during local inflammation has been described , we evaluated the T cell distribution both in the vicinity of the capillaries and far from the capillaries in SIV-infected animals . The CD4+ T cells were mainly located far from the capillaries , whereas CD8+ T cells were essentially located in the capillary area ( Fig 3B ) . These observations are in accordance with massive influx of CD8+ T cells previously described in the context of adipose inflammation [27] . In contrast , CD4+ T cell counts in adipose tissue were only slightly affected by SIV infection , and most CD4+ T cells recovered in the SVF had not recently migrated into the adipose tissue from peripheral blood . We next evaluated the differentiation of adipose CD4+ and CD8+ T cells obtained from infected and non-infected animals ( Figs 4 and S1 ) . Testing for CD95 , CD28 and CCR5 expression enabled us to identify naïve ( Tn ) , central memory ( Tcm ) , transitional memory ( Ttm ) and effector memory ( Tem ) subsets ( Figs 4A and S1A ) [55 , 56] . In both infected and non-infected animals , naïve CD4+ T cells ( CD95- CD28int ) were virtually absent , whereas the Tcm ( CD95+ CD28+ CCR5- ) fraction was predominant in both SCAT and VAT . Interestingly , this profile was specific to CD4+ T cells since CD8+ T cells from adipose tissue were essentially Tem ( CD95+ CD28- ) ( S1 Fig ) . Among CD4+ T cells , the CD95+ CD28+ CD4+ T cell fraction which includes the two potential cellular CD4+ T cell reservoir , i . e . Tcm ( CCR5- ) and Ttm ( CCR5+ ) , accounted for 82 . 3% [69 . 0–89 . 0] of the total in SCAT and 77 . 6% [69 . 9–77 . 6] in VAT , vs . 45 . 3% [28 . 8–62 . 0] in peripheral blood mononuclear cells ( PBMCs ) ( p = 0 . 001 and 0 . 014 respectively ) ( S2A and S2B Fig ) . Interestingly , we did not detect significant differences in any of the CD4+ T cell fractions when comparing adipose tissue from SIV-infected animals ( n = 7 ) and non-infected animals ( n = 5 ) . To evaluate the proportion of resident memory T cells [57 , 58] , we next determined CD69 expression on CD4+ T cells recovered from adipose tissue and ( as a control ) in PBMCs ( Fig 4B ) . The fraction of CD4+ T cells expressing CD69 was significantly higher in adipose tissue than in PBMCs ( p = 0 . 0003 for SCAT and 0 . 0034 for VAT ) . However , SIV infection was not associated with a significant difference in the proportion of CD69-expressing cells . Thus , the maintenance of normal CD4+ T cell numbers was associated with preservation of memory CD4+ T cell subset distribution in general as well as the resident CD4+ T cell memory distribution . Lastly , we evaluated the activation profile of adipose tissue T cells during chronic SIV infection ( Fig 4C ) . We determined the expression of HLA-DR ( a standard marker of T cell activation ) on adipose tissue CD4+ and CD8+ T cells and ( as a control ) in PBMCs . In non-infected animals , there were no significant differences between SCAT and VAT in terms of CD4+ and CD8+ T cell activation ( CD4+ T cells: 7 . 2% [4 . 5–9] in SCAT and 5 . 5% [4 . 3–8 . 2] in VAT; CD8+ T cells: 8 . 0% [6 . 0–10 . 0] in SCAT and 7 . 8% [5 . 0–10 . 1] in VAT ) . The percentages of HLA-DR-expressing T cells were higher in both SCAT and VAT than for PBMCs ( 1 . 6% [1 . 3–3 . 5] for CD4+ T cells in PBMCs and 3 . 8 [1 . 9–5 . 0] for CD8+ T cells in PBMCs ) . In SIV-infected animals , the percentage of HLA-DR-expressing cells was significantly higher in CD4+ and CD8+ T cells recovered in SCAT , VAT and PBMCs , relative to non-infected controls ( Fig 4C ) . For example , the percentage of HLA-DR-expressing CD4+ T cells was 12 . 5% [5 . 5–15] in SCAT , 15 . 0% [10 . 5–17 . 3] in VAT and 9 . 5% [5 . 8–20 . 0] in PBMCs from SIV-infected animals . Interestingly , we could not detect any significant difference in the proportion of Ki-67-expressing CD4+ or CD8+ T cells ( S3 Fig ) , suggesting that change in activation profile was not associated with massive in situ proliferation . Overall , SIV infection slightly affected CD4+ T cell numbers and their differentiation profile , and was associated with increased T cell activation in adipose tissue . A constant concern when studying immunity in tissues is the potential bias induced by blood contamination . Here , special efforts were made to avoid this; adipose tissue was devascularized and washed in medium prior to digestion . Importantly , the low observed proportion of CD45-expressing cells constitutes important evidence of low blood contamination . It is noteworthy that B cells ( identified as CD20-expressing cells ) were virtually absent from SCAT ( 0 . 10% of SVF cells [0 . 01–0 . 40] ) and VAT ( 0 . 11% [0 . 01–1 . 10] ) but were detected in PBMCs ( 9 . 5% [3 . 0–10 . 0] ) ( S2C Fig ) . The absence of B cells is thus a reliable indicator of the absence of blood contamination of adipose tissue samples . Moreover , adipose tissue CD4+ T cells differed significantly from PBMCs , with significantly lower percentages of naïve cells ( the CD4+ Tn fraction represented 2 . 0% [0 . 3–3 . 6] of the cells in SCAT , 5 . 6% [1 . 3–22 . 0] in VAT and 22 . 0% [8 . 9–41 . 7] in PBMCs; p<0 . 0001 and p = 0 . 03 , respectively ) and higher percentages of the two memory T cell subsets that are preferentially infected ( i . e . Tcm and Ttm ) ( S2A and S2B Fig ) . Lastly , the immunochemical demonstration of different localizations of CD4+ and CD8+ T cells ( Fig 3 ) also suggests that blood contamination was absent or barely present . We next evaluated the changes among adipose macrophages ( Fig 5 ) , which are involved both in innate immunity and adipose homeostasis . At present , there is no clear phenotypic strategy for identifying tissue-resident macrophages and defining their activation profile . The pro-inflammatory ( M1 ) versus anti-inflammatory ( M2 ) distinction ( commonly used in murine models ) may not reflect the great heterogeneity of macrophage phenotypes in tissues . Indeed , macrophages probably develop across a continuum , with anti-inflammatory to pro-inflammatory profiles . In the present study , we considered macrophages to be CD45+CD3-CD14+ cells . The proportion of macrophages among CD45+ cells in SCAT was greater in SIV-infected animals than in non-infected animals ( Fig 5A ) . A similar trend was observed in VAT . These findings are in line with the macrophage accumulation previously described in the context of adipose inflammation [26] . We confirmed the presence of macrophages in adipose tissue by performing immunohistochemical analyses ( CD68 staining ) of tissue sections ( Fig 5B ) . We next evaluated the activation profile of adipose tissue macrophages by screening for so-called M2 markers ( CD206 and CD163 ) associated with anti-inflammatory activity . Due to the tissue macrophages’ high level of auto-fluorescence , analyses were performed using isotype controls for both CD206 and CD163 staining , and “fluorescence minus one” ( FMO ) strategies were applied to accurately identify the different subsets ( S4A Fig ) . In non-infected animals , most adipose tissue macrophages were found to express both CD206 and CD163 in SCAT and VAT—suggesting that anti-inflammatory M2 macrophages were predominant in non-infected adipose tissue , as previously reported for humans and mice [24 , 59] ( Fig 5C ) . Importantly , CD206 expression on CD14-expressing cells was restricted to adipose-resident cells and was not detected in PBMCs . In contrast , the CD206-CD163+ fraction was predominant in PBMCs ( 73 . 9% [62 . 3%-92 . 6%] ) but extremely rare in SCAT and VAT SVF ( S4B Fig ) . When considering SIV-infected animals , the CD206+CD163+ populations were highly predominant in adipose tissue . However , the frequency of the anti-inflammatory CD206+CD163- fraction was significantly lower in SIV-infected animals than in non-infected animals for both SCAT ( p = 0 . 0258 ) and VAT ( p = 0 . 0187 ) ( Fig 5C and S5 Fig ) . Conversely , the CD206-CD163- fraction of adipose-resident macrophages ( which presumably corresponds to pro-inflammatory macrophages ) was elevated in SIV-infected animals ( p = 0 . 0269 ) . It is noteworthy that opposite changes were observed for CD14-expressing cells in PBMCs ( Fig 5D ) : SIV infection was associated with a lower CD206-CD163- fraction and a higher CD206-CD163+ fraction . Taken as a whole , our results demonstrated that chronic SIV infection skews both adipose tissue T cell and macrophage populations towards a more activated profile . To date , assays for HIV DNA in adipose tissue have been performed on both whole tissue samples and adipocytes . No consensus has emerged from these studies , although it is currently assumed that adipocytes are not a major target for HIV in vivo . In the present work , we focused on the SVF and potential HIV/SIV target cells in particular ( CD4+ T cells and macrophages ) . We looked for cell-associated SIV DNA and RNA in total SVF ( n = 8 ) and sorted CD4+ T cell and macrophage fractions ( n = 5 each ) from SIV-infected animals . As shown in Fig 6A , SIV DNA was detected in SVF samples from SCAT and VAT in all animals tested and no significant difference between the two sites was detected . Sorted CD4+ T cell fractions from all SCAT and VAT samples ( n = 10 ) were positive for SIV DNA . The observations for adipose CD14+ cells were more heterogeneous: 3 of the 5 animals were positive for SIV DNA in both SCAT and VAT . The median SIV DNA levels were 3 . 7 log DNA copies/106 cells in total SVF and 3 . 2 log SIV DNA copies/106 cells in PBMCs . Additionally , 4 . 3 and 2 . 3 log viral DNA copies/106 cells were detected respectively in the sorted CD4+ fraction and CD14+ cells recovered from adipose tissue . We analyzed viral RNA in a total of 11 animals by performing an RT-PCR assay on cell suspensions ( n = 8 ) ( Fig 6B ) and in situ SIV RNA hybridization ( n = 3 ) ( Fig 6C ) . SIV RNA was detected in the SVF from all animals tested ( except for one SCAT sample ) . SIV RNA levels in SVF collected from SCAT and VAT ( 4 . 5 log SIV RNA copies/106 cells ) were similar to those observed in PBMCs ( 5 . 3 log SIV RNA copies/106 cells ) . Quantitative assays of SIV RNA were also performed in CD4+ and CD14+ cell subsets from 5 animals . The mean level of SIV RNA ( in log copies per million cells ) was 4 . 1 in a sorted adipose CD4+ fraction and 4 . 5 in adipose CD14+ cells . In situ hybridization was performed on SCAT and VAT samples and confirmed the presence of SIV RNA in both tissues as shown in one of the three animals tested ( Fig 6C ) . We next compared SIV DNA and RNA contents in CD4+ T cells and CD14+ cells recovered from various tissues/cell sources: adipose tissue , PBMCs and mesenteric lymph nodes ( LNs ) ( S6 Fig ) . Levels of SIV DNA and RNA within the CD4+ T cell fractions were similar in all tissues tested—suggesting that the proportion of CD4+ T cells in infected adipose tissue was equivalent to that in other tissues where viral replication is known to occur extensively . The same observation applied to CD14+ cells . In conclusion , we demonstrated that the SVF fraction from chronically SIV-infected animals included SIV-infected CD4+ T lymphocytes and macrophages . SIV RNA was detected in both the total SVF fraction and adipose CD4+ and CD14+ fraction—suggesting that viral production occurs in adipose immune cells . To ascertain whether adipose tissue may constitute a viral reservoir , we collected adipose tissue samples from 13 ART-treated HIV-infected patients who had undergone elective visceral surgery for non-HIV related causes ( Table 1 ) . In 11 patients , we searched for viral DNA in SVF cells and PBMCs . Even after clinically effective ART treatment , HIV DNA was detected in the SVF of all samples tested ( 11 from VAT and 4 from SCAT ) ( Fig 7A ) . The median level of HIV DNA in SVF from VAT ( 2 . 15 [2 . 01–2 . 47] log DNA copies/million cells ) was significantly lower than in PBMCs ( 2 . 94 [2 . 25–3 . 28] ) . However , this low level of infection in adipose tissue might primarily reflect the low proportion of hematopoietic cells in adipose tissue . We therefore quantified HIV DNA in sorted CD4+ T cells and sorted CD206+ CD14+ cells from three patients . Contamination by circulating monocytes was controlled for by measuring CD206 expression on CD14+ cells . In one sample , HIV DNA could not be detected in adipose CD4+ T cells—presumably because of low DNA input . In the samples with detectable virus , the median level of HIV DNA in CD4+ T cell fractions recovered from adipose tissue was 3 . 83 log DNA copies/million cells ( compared with 3 . 56 in sorted peripheral blood CD4+ T cells ) . We failed to detect HIV DNA in the three sorted CD14+ CD206+ samples tested . Therefore , in patients on long-term ART with no detectable viremia , we confirmed that HIV-infected cells are present in the SVF ( mainly tissue-resident CD4+ T cells ) . Given that the detected HIV DNA might have been replication-defective , we also screened for viral RNA in adipose tissue sections using in situ hybridization . Two SCAT samples and one VAT sample were obtained from ART-treated HIV-infected patients . As shown in Fig 7C , cells expressing HIV RNA were detected ( albeit in low numbers ) in the three adipose tissue samples studied . We therefore demonstrated that productive , HIV-infected cells are present in the adipose tissue of ART-treated patients . The ex vivo induction of viral replication in sorted cells is a further means of demonstrating that HIV is replication-competent . In six long-term ART-treated patients , we performed an in vitro viral reactivation assay on total SVF cells , sorted adipose CD4+ and CD206+CD14+ cells and sorted peripheral blood CD4+ or CD14+ cells in the presence of allogeneic pre-activated CD4+ T cells and phytohemagglutinin ( Fig 8 ) . In order to obtain sufficient numbers of cells , SCAT and VAT samples were pooled when both were available . Although small numbers of cells ( 5 104 to 5 105 ) were used for this reactivation assay , we detected the induction of viral replication in the sorted CD4+ T cell fraction from the SVF of all six patients . HIV RNA in supernatants from the ex vivo-activated , sorted SVF CD4+ T cells was detected from day 7 to day 21 in nearly all patients . In two patients ( #5 , 7 ) , HIV RNA was detected solely after the reactivation of sorted SVF CD4+ T cells but not in total SVF or in PBMCs—probably because of the small number of infected cells in the culture . In the four other patients ( #6 , 8 , 9 , 10 ) , HIV RNA was detected in similar numbers of ex vivo-activated CD4+ T cells recovered from both adipose tissue and PBMCs . Induction of HIV replication from SVF was detected in these four patients ( albeit at low levels in three—probably due to the small number of infected cells ) . No HIV RNA was detected in cultures of activated adipose CD14+CD206+ fractions . Importantly , HIV RNA was not detected in non-activated SVF cell cultures ( except in one patient ) , meaning that we could rule out the amplification of non-latent , pre-existing HIV . Our results show that adipose CD4+ T cells were infected by replication-competent HIV . In conclusion , we performed three levels of detection: ( i ) HIV DNA using PCRs , ( ii ) HIV RNA using in situ hybridization and ( iii ) HIV RNA in supernatants in an in vitro reactivation assay based on ultrasensitive RT-PCR . We demonstrated that latently HIV-infected cells were present in the SVF from both SCAT and VAT in 6 ART-treated patients . HIV DNA was consistently detected in adipose CD4+ T cells—strongly suggesting that adipose tissue is an HIV reservoir in aviremic patients on long-term ART .
Residual chronic inflammation and viral persistence are two key features of ART-treated , HIV-infected patients . Although several mechanisms have been described , the establishment and persistence of low-grade inflammation are not fully understood and remain to be characterized . Given that adipose tissue is a major site of inflammation in a context of obesity , we hypothesized that the adipose tissue changes associated with HIV infection may drive low-grade inflammation . Indeed , the literature data reveal a strongly pro-inflammatory profile of adipose tissue during HIV infection—although most of these studies were performed on ART-treated patients [36 , 41 , 60] . In order to rule out the potential confounding role of ART and to collect sufficient amounts of tissue , we studied the SIV/macaque infection model . Chronic SIV infection in vivo was associated with a markedly higher adipocyte density ( i . e . cell numbers per field ) and SVF cell count ( expressed per gram of adipose tissue ) . The elevated adipocyte density has been linked to an abnormally low adipocyte size , which in turn may reflect drastic alterations of adipogenesis [61 , 62] and/or lipolysis [48 , 60] . The massive increase in SVF cells in adipose tissue in chronically infected animals essentially reflected the accumulation of CD45- cells . The CD45- cells recovered in the SVF of infected animals need to be identified more accurately . Those in the SVF included pre-adipocytes , the accumulation of which may be related to changes in adipocyte differentiation . These striking changes in CD45- proportion need to be further investigated and could relate to multiple defects affecting adipose tissue during SIV infection such as lipodystrophy , inflammation or viral persistence . The SVF also contains mesenchymal stem cells [63–65] , and it remains to be established whether these cells can also be infected . When focusing on immune cells , there was no major change in total leukocyte counts . However , the percentage of CD4+ T cells was severely lower in adipose tissue from SIV-infected animals than in non-infected animals , whereas the CD8+ T cell percentage was markedly higher . We demonstrated that the change in the CD4/CD8 ratio reflected the accumulation of CD8+ T cells , in accordance with previous reports of massive CD8 accumulation during adipose tissue inflammation [27] . Intense recruitment of CD8+ T cells was confirmed by the perivascular location of CD8+ T cells . However , a direct impact of SIV in adipose tissue cannot be ruled out [53 , 54] . The fact that CD4+ T cell counts were generally unaffected was not expected in the context of SIV infection—especially given the high percentage of central memory CD4+ T cells ( a major HIV/SIV target ) in adipose tissue . Organ-dependent differences in the direction and amplitude of cell depletion have been described [66 , 67]: CD4+ T cell depletion is rapid and massive in the intestinal mucosa , whereas the follicular helper subset ( also a major HIV target ) even expands in secondary lymphoid organs [68] . Importantly , CD4+ T cells were mainly found recovered in the vicinity of adipocytes and far from capillaries . This remote site may favor viral persistence , as was recently described in follicular CD4+ T cells [69] . It thus appears that adipose tissue may constitute an additional site for the accumulation of CD4+ T cells . Phenotypic analysis revealed that the Tcm CD4+ T cells fraction was predominant in both SCAT and VAT , thus confirming the high potential for viral persistence of adipose tissue . The predominance of central memory CD4+ T cells , which usually identifies inductive lymphoid site , also raises the question of the lymphoid definition of the adipose tissue: Does adipose tissue constitute a lymphoid site and if so is it an inductive or effector lymphoid site ? We also found that phenotypic changes can be detected in both lymphocytes and macrophages during chronic SIV infection of adipose tissue . Expression levels of the activation marker HLA-DR on adipose tissue T lymphocytes were higher in SIV-infected animals . Macrophages displayed a moderate change in phenotype . SIV infection was associated with an increase in the proportion of CD206-CD163- adipose tissue macrophages and a concomitant decrease in the CD206+CD163- fraction—suggesting a shift from an anti-inflammatory profile towards a more activated profile or the recruitment of migrating blood monocytes . Overall , we observed features commonly associated with obesity-related inflammation: elevated SVF numbers , the specific recruitment of CD8+ T cells , a higher proportion of macrophages and greater expression of activation markers . These observations strongly suggest that adipose tissue is involved in inflammation during SIV infection . We next investigated the mechanisms that might underlie the increase in adipose tissue inflammation in chronically SIV-infected animals . There is increasing evidence to show that HIV infection per se interferes with adipose tissue homeostasis , although the mechanisms remains to be defined [34] . Adipose inflammation may be a consequence of systemic inflammation , with in turn is worsened by adipose inflammation in a vicious circle . Adipose inflammation may also be related to the spread of viral proteins [48–50] , CD4+ T cell lymphopenia [70] and/or microbial translocation [71 , 72] , all of which are known to alter adipocyte homeostasis . Furthermore , SIV may directly infect adipose-resident immune cells . Importantly , previous reports essentially failed to demonstrate consistent infection of adipocytes and did not provide any information on adipose immune cells . In the present study , SVF fractions were positive for SIV DNA and SIV RNA in all animals tested . SIV DNA was also detected in all sorted adipose CD4+ T lymphocytes . SIV DNA was detected less consistently in sorted CD14-expressing cells ( 3 out of 5 in both SCAT and VAT ) . The results for cell-associated SIV RNA essentially corroborated those for SIV DNA . We demonstrated that in SVF fractions , SIV was consistently present in CD4+ T lymphocytes and less frequently present in macrophages . Our detection of viral DNA and RNA in stromal vascular cells collected from adipose tissue of chronically viremic animals thus confirms that adipose tissue is a site of infection . We next sought to confirm the presence of viral infection in human adipose tissue and , more importantly , to characterize viral persistence in ART-suppressed HIV-infected patients . Three different assays were performed: ( i ) detection of viral DNA in SVF ( in 11 patients ) and in sorted CD4+ T cell fractions ( from 3 patients ) , ( ii ) in situ RNA hybridization on fixed sections of adipose tissue ( in 3 patients ) , and ( iii ) in vitro viral reactivation ( in 6 patients ) . HIV DNA was detected in all SVF samples tested; this finding is in line with Couturier et al . ’s report of HIV DNA in the SVF of 5 ART-treated , HIV-infected patients [73] . Importantly , we were able to detect HIV DNA in sorted adipose CD4+ T cell fractions but not in CD206+ CD14-expressing cells . It is noteworthy that sorted CD4+ T cells recovered from adipose tissue had much the same levels of HIV DNA as PBMCs . These results suggest that the proportion of infected cells is similar in adipose tissue CD4+ T cells and in peripheral blood CD4+ T cells . In contrast to the situation in chronically viremic macaques ( in which SIV DNA was detected in a CD14+ fraction recovered from the SVF ) , we did not detect HIV DNA in CD14+CD206+ cells recovered from the SVF in aviremic patients . We checked whether this discrepancy was related to the difference in macrophage selection ( i . e . CD14+ cells vs . CD14+CD206+ cells ) . CD14+CD206+ cells sorted from viremic macaques were still positive for SIV DNA , suggesting that viral DNA may be present in macrophage subsets in viremic stages but might not persist during long-term ART . One can reasonably presume that the half-life of macrophages is shorter than that of Tcm CD4+ T cells . Alternatively , the small number of macrophages collected may have prevented us from detecting a small proportion of infected cells . This aspect will be investigated further . Lastly , six samples were reserved for a viral replication assay . Although only small numbers of sorted adipose cells were available , we were able to monitor in vitro viral replication by detecting HIV RNA in supernatants after the incubation of adipose tissue CD4+ T cells with allogeneic pre-activated CD4+ T cells . HIV RNA was consistently detected in cultures of CD4+ T cells sorted from the SVF , suggesting that adipose CD4+ T cells are infected by replication-competent HIV . In four patients , ex vivo HIV replication was induced to the same extent in sorted CD4+ T cells from the SVF and from PBMCs . These results suggest that the proportion of infected cells is much the same among adipose tissue CD4+ T cells and peripheral blood CD4+ T cells; this hypothesis is supported by the fact that sorted CD4+ T cells from adipose tissue and PBMCs had similar HIV DNA contents . In contrast , replication-competent HIV was detected in the SVF only in two patients . The high frequency of memory CD4+ T cells in adipose tissue ( relative to PBMCs ) may explain the high proportion of latently infected cells . Otherwise the number of latently infected CD4+ T cells in adipose tissue might be higher than in PBMCs in some patients on ART . These findings emphasize the need to sample several tissues when studying HIV reservoirs in patients on ART . Further investigation ( using limiting dilution assays ) is necessary but would be technically challenging ( given the low numbers of resident tissue CD4+ T cells in patients ) . Overall , we were able to demonstrate the persistence of HIV DNA and RNA within stromal vascular cells in adipose tissue recovered from ART-treated HIV-infected patients . These observations were supported by the results of in vitro reactivation experiments , showing that adipose CD4+ T cells contained replication-competent HIV . This dataset strongly supports the hypothesis whereby adipose tissue constitutes an important viral reservoir . Adipose tissue may thus constitute a favorable environment for viral persistence for several reasons: ( a ) constant inflammation favors viral replication , ( b ) the presence of elevated fractions of activated and central memory CD4+ T cells , which are HIV’s natural targets , ( c ) the potentially insufficient distribution of some antiretroviral drugs into adipose tissue [44] , which may favor viral persistence , and ( d ) the specific metabolic and immune activity of adipose tissue , which may affect the effectiveness of immune responses [74] . In the present work , our analyses of SCAT and VAT from SIV-infected macaques and ART-suppressed patients , showed that ( i ) SIV infection induced immune activation and a pro-inflammatory profile in adipose tissue immune cells and ( ii ) these immune cells were indeed infected by SIV/HIV . These results indicate that adipose tissue constitutes a new , relatively large reservoir for the virus that could be involved in chronic immune activation and low-grade inflammation . Our observations have major implications in the context of HIV disease . Firstly , they emphasize the crucial requirement for the broad diffusion of antiretroviral drugs within tissues; combination therapy must include drugs that diffuse not only into adipose tissues but also into tissue CD4+ T cells and macrophages . Two main mechanisms may prevent efficient activity of ART on adipose infected cells: ( i ) low accessibility of ART to fat tissue [44] , ( ii ) sequestration of drugs inside the lipid droplets at the expense of adipose infected immune cells [75] . Secondly , they provide an interesting rationale for the use of drugs with metabolic activity . It might be interesting to reconsider the anti-inflammatory impact of statins [76 , 77] by focusing on adipose sites and addressing the drugs’ potential impact on viral reservoirs . Thirdly , gender differences in adipose tissue distribution , the inflammatory profile and immune cell content [78] may underlie differential susceptibility to the establishment of viral reservoirs . Fourthly , our results open up new therapeutic strategies for limiting the size of viral reservoirs , chronic inflammation and associated comorbidities ( via the modulation of adipose tissue related pathways rather than strictly immune pathways ) .
Twenty-three adult cynomolgus macaques ( Macaca fascicularis ) were infected via the intravenous , intravaginal or intrarectal route with 0 . 5 to 5 , 000 50% animal infectious doses ( AID50 ) of SIVmac251 biological isolate and monitored for 15 months ( median value , interquartile range: [11–18] ) . SIV infection in cynomolgus macaques closely recapitulates the major features of HIV infection , including progression towards AIDS . At sacrifice , the plasma viral load in SIV-infected animals was 4 . 4x104 [0 . 7x104-6 . 7x104] RNA copies/mL . Twenty one non-SIV-infected animals were used as controls . At sacrifice , blood samples and 10 to 35g samples of abdominal subcutaneous adipose tissue ( SCAT ) and visceral adipose tissue ( VAT ) were collected . Adipose tissues were devascularized to prevent blood contamination . PBMCs were isolated from EDTA-anticoagulated blood by Ficoll density gradient centrifugation . Adult cynomolgus macaques were imported from Mauritius and housed in the animal facility at the Commissariat à l’Energie Atomique et aux Energies Alternatives ( CEA , Fontenay-aux-Roses , France ) . Non-human primates ( NHPs , which include cynomolgus macaques ) are housed and handled in accordance with French national regulations and subject to inspection by the veterinary authorities ( CEA Permit Number A 92-032-02 ) . The CEA facility complies with the Standards for Human Care and Use of Laboratory of the Office for Laboratory Animal Welfare ( OLAW , USA ) under OLAW assurance number #A5826-01 . The use of NHPs at CEA is also in line with the European Directive ( 2010/63 , recommendation Nu9 ) . The animals were used under the supervision of the veterinarians in charge of the animal facility . The study protocols were reviewed by the CEA’s Animal Care and Handling Committee ( Comité d’Ethique en Expérimentation Animale , registered with the French Ministry of Research ) . Animals were housed in adjoining , individual cages ( allowing social interactions ) and under controlled humidity , temperature and light conditions ( 12-hour light/12-hour dark cycles ) . Water was available ad libitum . Animals were monitored and fed with commercial monkey chow and fruits once or twice daily by trained personnel . Macaques were provided with environmental enrichment , including toys , novel foodstuffs and music under the supervision of the CEA’s Animal Care and Handling Committee . After sedation with ketamine chlorhydrate , animals were sacrificed by intracardiac injection of sodium pentobarbital ( Vetoquinol , Paris , France; 180 mg/kg ) . All 13 patients provided their written , informed consent to participation . The study protocol was approved by the regional investigational review board ( Comité de Protection des Personnes Ile-de-France VII ( Paris , France ) ) . Adipose tissue samples from 13 ART-treated HIV-1-infected patients were recovered during elective abdominal surgery for non-AIDS-related indications . SVF was isolated from fresh samples . The 13 patients were on long-term ART and had displayed an undetectable viral load for over four years . In eleven patients , SVF samples were screened for HIV DNA . Five SVF samples were reserved for an HIV DNA assay in sorted fractions and 6 were reserved for the in vitro reactivation assay . In situ hybridization was performed on fixed adipose sections from three patients on ART and on a prostate sample ( from a viremic patient ) as a positive control . SCAT and VAT were weighed , washed twice in PBS 1X 5% fetal bovine serum ( FBS ) , cut into pieces of 2 to 3 mm and then digested in a bath of collagenase ( C2139 , Sigma ) at a concentration of 0 . 33 mg /mL in DMEM supplemented with 5% FBS for 30 min at 37°C with constant shaking . Mechanical dissociation by suction/discharge with a 10 mL syringe was then performed . Next , the adipose suspension was filtered through a 100 micron mesh . Following an initial low-speed centrifugation ( 300g , 10 min ) , adipocytes ( the upper phase ) were separated and the lower phase ( comprising the SVF cells ) was centrifuged further . After two additional washes , the pellet containing the SVF was resuspended in PBS with 5% FBS . SVF cell suspensions were then counted in Malassez counting chambers ( C-chip , NanoEntek , Seoul , Korea ) under the microscope , using Trypan blue to exclude dead cells . SVF was either directly frozen for molecular analyses , frozen in FBS 10% DMSO for cell preservation , and/or sorted or analyzed immediately by flow cytometry . Staining was performed after the saturation of Fc receptors by incubation with FC block ( BD ) , mouse serum ( eBiosciences ) and healthy macaque serum ( an in-house preparation ) for 30 min at 4°C . Amine-reactive blue dye ( Live/dead Fixable , Life Technologies ) was used to assess cell viability and exclude dead cells from the analysis . Cells were stained with monoclonal antibodies ( incubation for 15 min at 4°C ) , washed in PBS 1X/10% FBS and fixed in commercial fixative solution ( CellFIX , BD Biosciences ) . A variety of antibody panels were used to study adipose tissue T lymphocytes and macrophages [55] . Macrophages were identified using the following panel: CD16 V450 ( Clone 3G8 ) / CD206 FITC ( Clone 19 . 2 ) / CD14 PerCPCy5 . 5 ( M5E2 ) / CD163 APC ( GHI/61 ) / CD11b PECY7 ( BEAR1 ) / HLA-DR PECF574 ( G46-6 ) / CD3 AF700 ( SP34-2 ) . Corresponding isotype controls for CD163 and CD206 were used at the same concentrations as the reference antibody , in accordance with the manufacturer’s instructions . Lymphocyte subsets were analyzed as follows: CD45 V500 ( D058-1283 for macaques , HI-30 for humans ) / CD4 PerCPCy5 . 5 ( L200 ) / CD8 V450 ( leu2a ) / CD95 APC ( DX-2 ) / CD28 PE ( CD28 . 2 ) / HLA-DR PECY7 ( G46-6 ) / CD20 AF700 ( 2H7 ) / Ki-67 FITC ( B56 ) . Additional combinations with CCR5 APC ( 3A9 ) , CD69 PECY7 ( FN50 ) antibodies were also applied . All antibodies were purchased from BD Biosciences , with the exception of anti-CD11b ( Beckman-Coulter ) . Data were acquired with an LSR Fortessa cell analyzer ( BD Biosciences ) and analyzed with FlowJo software ( Treestar ) . Samples of SCAT and VAT were fixed in 4% buffered formalin and embedded in paraffin . Sections ( thickness: 3 microns ) were stained with hematoxylin/eosin/saffron reagent . Adipocytes were counted in an average of 10 high-power fields ( HPFs ) . Parallel sections were immunostained for CD3 ( clone F . 2 . 38 , Dako ) , CD4 ( clone 4B12 , Leica Biosystem ) , TIA1 ( TIA1 cytotoxic granule-associated RNA binding protein , clone 2G9 , Immunotech ) and CD68 ( clone PG-M1 , Dako ) , in order to identify CD4+ T lymphocytes , cytotoxic T cells and macrophages . Unless otherwise stated , intravascular leucocytes were excluded from counting , although T cells located in the vicinity of capillaries or far from capillaries were also studied . SVF cells from SCAT and VAT were stained as described above , using amine-reactive blue dye to identify the dead cells . CD4+CD3+CD45+L/D- cells ( CD4+ T cells ) in macaques and humans , CD14+CD3-CD20-CD45+L/D- cells in macaques and CD14+CD206+CD3-CD20-CD45+L/D- cells ( macrophages ) in humans were sorted on a FACS ARIA cell sorter ( BD Biosciences ) . The gating strategy for macrophage isolation used CD14 fluorescence minus one ( FMO ) control staining to eliminate contamination due to potential auto-fluorescence of SVF cells . Sorting purity profile are presented in S7 Fig . In both NHPs and humans , sorting purity was consistently greater than 95% ( CD14-expressing cells: 97 . 2% [95 . 1–98 . 7] , CD4+ T cells: 96 . 9% [96 . 1–99 . 1] ) . RNA and DNA were extracted from cells as follows: 1x106 PBMCs or SVF cells and various numbers of sorted CD4+ and CD14+ cell numbers ( ranging from 50000 to 200000 ) were resuspended in 350 μL of RLT buffer ( Qiagen , France ) . 200 μL of TE buffer were added . Next , NaCl ( 5M ) was added . The mixture was incubated at 4°C for between 15 and 60 min , followed by centrifugation for 15 min at 3000 rpm and 4°C . RNA was precipitated from the supernatant using pH 4 . 5 phenol/chloroform ( Sigma-Aldrich ) at a v/v ratio of 5:1 . 200 μl of TE buffer was added to the pellet . DNA was then precipitated from the mixture in phenol/chloroform/isoamyl alcohol 25:24:1 ( pH = 8 , Sigma Aldrich ) . Cell-associated viral DNA and RNA were measured by qPCR as described previously , using primers and probes designed specifically for SIVmac251 isolates [79] . PCRs were performed in duplicate; the detection limits were 35 SIV DNA copies/106 cells and 50 SIV RNA copies/106 CCR5 copies . Only reactions for which duplicates gave consistent results were included . Positive and negative controls were used to rule out sample contamination . SIVmac plasmids were used as standards to calculate SIV DNA copy numbers . Primers and plasmids for CCR5 were used to normalize the viral levels against the number of cells [80] . Plasma SIV RNA was quantified as previously described [81] . Total HIV DNA was quantified in an ultrasensitive real-time PCR assay of frozen SVF cells , sorted adipose CD4+ and CD14+ cells in SVF , and PBMCs using the GENERIC HIV-DNA assay from Biocentric ( Bandol , France ) , as previously described [82 , 83] . Total DNA was extracted with a QIAamp All prep DNA/RNA microkit or minikit ( Qiagen ) , depending on the number of cells available ( less than and more than 1 million cells , respectively ) . The entire HIV-DNA extract was tested in two to four replicates , with a threshold of 2 HIV-DNA copies per PCR . The thresholds varied according to the available cell numbers and were calculated for each sample . 1x106 cells were analyzed for total SVF and PBMC samples . Sorted SVF CD4+ and CD14+CD206+ cell numbers were consistently lower ( from 14000 to 320000 ) , and so the detection limit was set to <3 . 23 log copies/106 cells . Following in vitro reactivation , HIV-RNA was quantified in culture supernatants using an ultrasensitive real-time PCR assay ( GENERIC HIV , Biocentric , Bandol , France ) . The extracts were tested in two to five replicates [84] . Paraformaldehyde-fixed , paraffin-embedded tissues were assayed for SIVmac and HIV-1 RNA expression using a digoxigenin-antidigoxigenin technique , as previously described [67 , 85] . The digoxigenin-UTP-labeled riboprobe spanned the whole SIVmac or HIV-1 genome ( Lofstrand Labs Ltd , Gaithersburg , MD , USA ) . Nitroblue tetrazolium-5-bromo-4-chloro-3-indollylphosphate toluidinium revelation was used to detect infected cells in the tissues . The specificity of the hybridization signal was always checked by hybridizing sense probes on parallel sections and anti-sense probes on non-infected adipose tissues . Prostate tissues from a viremic SIV-infected macaque and from HIV-infected patients were used as positive controls , as previously described [86–88] . Cell fractions recovered from ART-treated , HIV-infected patients were stimulated with phytohemagglutinin ( 0 . 3 μg/mL , Sigma ) , IL-2 ( 25 μg/mL , Immunotools ) ( for SVF and T cell fractions ) and LPS ( 1 μg/mL , Sigma ) ( for CD14-expressing cells ) and co-cultured with allogeneic pre-activated CD4+ T cells at a ratio of 1:2 . 5 . Allogeneic CD4+ T cells were purified from PBMCs from healthy donors by positive selection using magnetic beads ( Miltenyi ) . At day ( D ) 7 and D14 , reactivation was sustained by novel addition of pre-activated CD4+ T cells . Supernatants were collected at D3 , 7 , 11 , 14 , 17 and 21 and cell pellets were collected at D21 . The number of cells used in the reactivation assay depending on the cell fraction isolated from adipose tissue SVF: 1x106 for the total SVF fraction , 1x105 to 5x105 for sorted CD4+ T cells , and 5x104 to 2x105 for sorted , CD14-expressing cells . As negative controls , 1x106 SVF cells were cultured in the absence of allogeneic , pre-activated CD4+ T cells . Data are quoted as the median [interquartile range] . Statistical analyses were carried out with GraphPad Prism 5 . 03 software ( GraphPad Software Inc . ) . A Mann-Whitney non-parametric test was used to compare data from SIV-infected and non-infected macaques . A Wilcoxon matched-pair signed rank test was used to compare different tissues from the same animal . In graphs , the thresholds for statistical significance are indicated as follows: * p<0 . 05; ** p<0 . 01 , *** p<0 . 001 . | Chronic immune activation/inflammation and viral persistence in reservoirs are important features of chronic HIV infection—even in patients receiving ART . We sought to evaluate the involvement of adipose tissue in chronic HIV/SIV infections . Adipose tissue accounts for 15 to 20% of the body weight , contains both adipocytes and ( within the stromal vascular fraction ) immune cells , and exerts crucial metabolic and immune activities . We postulated that adipose tissue might provide an ideal environment for HIV persistence and immune inflammation . We first showed that viremic SIV-infected macaques had elevated levels of immune activation and inflammation in adipose tissue , and that both resident CD4+ T cells and macrophages were infected . In similar experiments in ART-controlled HIV-infected patients , HIV DNA was detected in the stromal vascular fraction of adipose tissue ( more specifically , in adipose tissue CD4+ T cells ) . Replication-competent HIV was detected in ex vivo- activated , sorted adipose tissue CD4+ T cells from six aviremic , ART-treated patients . Thus , adipose tissue may constitute a viral reservoir and be involved in long-term immune activation and inflammation—even in ART-suppressed patients . Given that adipose tissue is strongly regulated by both metabolic and immune pathways , modulating adipose tissue may constitute a valuable means of limiting both viral persistence and chronic inflammation in HIV-infected patients even ART-controlled . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Adipose Tissue Is a Neglected Viral Reservoir and an Inflammatory Site during Chronic HIV and SIV Infection |
Spinocerebellar ataxia type 11 ( SCA11 ) is a rare , dominantly inherited human ataxia characterized by atrophy of Purkinje neurons in the cerebellum . SCA11 is caused by mutations in the gene encoding the Serine/Threonine kinase Tau tubulin kinase 2 ( TTBK2 ) that result in premature truncations of the protein . We previously showed that TTBK2 is a key regulator of the assembly of primary cilia in vivo . However , the mechanisms by which the SCA11-associated mutations disrupt TTBK2 function , and whether they interfere with ciliogenesis were unknown . In this work , we present evidence that SCA11-associated mutations are dominant negative alleles and that the resulting truncated protein ( TTBK2SCA11 ) interferes with the function of full length TTBK2 in mediating ciliogenesis . A Ttbk2 allelic series revealed that upon partial reduction of full length TTBK2 function , TTBK2SCA11 can interfere with the activity of the residual wild-type protein to decrease cilia number and interrupt cilia-dependent Sonic hedgehog ( SHH ) signaling . Our studies have also revealed new functions for TTBK2 after cilia initiation in the control of cilia length , trafficking of a subset of SHH pathway components , including Smoothened ( SMO ) , and cilia stability . These studies provide a molecular foundation to understand the cellular and molecular pathogenesis of human SCA11 , and help account for the link between ciliary dysfunction and neurodegenerative diseases .
Primary cilia play a critical role in many aspects of embryonic development . Cilia are important for the development of the brain and central nervous system , which accounts for the structural brain defects , cognitive impairments , and other neurological disorders that are characteristic of many human ciliopathies[1–3] . Cilia are present on a wide variety of neurons and astroglia within the adult brain , although the specific requirements for these organelles in the function of the adult brain are not well understood . In prior work , we identified a Serine/Threonine kinase , Tau tubulin kinase 2 ( TTBK2 ) , that is essential for initiating the assembly of primary cilia in the embryo[4] . TTBK2 is a 1244 amino acid protein ( 1243 amino acids in mouse ) that was initially purified from bovine brain tissue as a microtubule-associated protein[5–7] . The protein is comprised of a kinase domain ( AA 21–284 ) , and a long C-terminus that is important for targeting TTBK2 to the mother centriole[4] , as well as mediating its interactions with end-binding proteins at the microtubule +tips[8] , and likely for additional regulation of TTBK2 function . TTBK2 was initially shown to phosphorylate the microtubule-associated proteins TAU and MAP2 in addition to β-Tubulin in vitro[6] , and more recent evidence suggests that TTBK2 can phosphorylate the centriolar distal appendage protein CEP164[9] , as well as the atypical kinesin KIF2A at the microtubule +tips[10] . In addition to the critical requirement for TTBK2 in ciliogenesis , particular dominant mutations that disrupt TTBK2 cause a hereditary ataxia , spinocerebellar ataxia type 11 ( SCA11 ) [11] . Like other subtypes of SCA , SCA11 is a progressive neurodegenerative condition predominantly affecting the cerebellum . At the cellular level , SCA11 is characterized by cerebellar atrophy resulting from a degeneration of Purkinje cells ( PCs ) of the cerebellum . However , the molecular basis underlying this pathology as well as for the dominant mode of SCA11 inheritance remain unknown . Three different heterozygous , familial , SCA11-associated mutations in TTBK2 cause late-onset ataxia . These mutations are insertions or deletions of one or two bases that result in frame shifts , and produce similar truncations of TTBK2 protein C-terminal to the kinase domain , at approximately AA 450[11 , 12] . A fourth mutation in TTBK2 that causes an earlier-onset disease truncates the protein at AA 402[13] . Because TTBK2 is essential for the biogenesis of primary cilia , which are in turn critical for the development of the nervous system , we hypothesized that the SCA11-associated mutations disrupt the function of TTBK2 in cilia formation . In previous structure-function experiments , we tested the ability of truncations of TTBK2 to restore cilia in Ttbk2 null mutant cells and found that those corresponding with the SCA11-associated mutations were unable to rescue cilia formation . We also found that , when over-expressed in Wild-Type ( WT ) cells by retroviral transduction , these truncations also partially suppress cilia formation[4] , suggesting the SCA11-associated truncations may act by interfering with the activity of the wild-type gene product . In the present study , we examined phenotypes of mice with SCA11-like truncating mutations knocked into the endogenous Ttbk2 locus . We found that , with respect to cilia , Ttbk2sca11 homozygotes are indistinguishable from the null allele . Specifically , in these mutants , cilia initiation fails and cilia-dependent SHH signaling is blocked . Using a series of Ttbk2 alleles , we showed that SCA11-associated truncated proteins dominantly interfere with the function of full-length ( wild-type ) TTBK2 in cilium assembly . In addition , these allelic combinations have uncovered a previously unappreciated function for TTBK2 in the regulation of cilia stability . TTBK2 localizes to the mother centriole prior to cilia formation and remains at the transition zone of the cilium following completion of assembly . In this study , we present evidence from hypomorphic allelic combinations that TTBK2 also acts after cilium initiation to regulate cilium stability , in part by countering a cilium disassembly pathway .
To examine the effects of SCA11-associated TTBK2 truncations on the function of the protein in mediating cilia formation and function , we used an allele of Ttbk2 in which a mutation precisely recapitulating one of the human SCA11-causing mutations was knocked into the mouse Ttbk2 genomic locus[14] . Ttbk2sca11/sca11 homozygous embryos were previously reported to die by E11 , but their developmental and cellular phenotypes were not described[15] . We found that E10 . 5 Ttbk2sca11/sca11 embryos exhibit morphological phenotypes that are strikingly similar to those that we previously described in embryos homozygous for an ENU-induced null allele of Ttbk2 , Ttbk2bby/bby [4] ( referred to from this point as Ttbk2null/null ) , including holoprosencephaly , a pointed midbrain flexure , and randomized heart laterality ( Fig 1A ) . The SHH pathway is essential for the patterning of many tissues within the developing embryo , including the neural tube , where a gradient of SHH from the notochord specifies and patterns ventral neural progenitors , including floorplate , V3 interneuron progenitors , and motor neuron progenitors[16] . Similar to Ttbk2null/null embryos , the Ttbk2sca11/sca11 embryos exhibited neural patterning defects consistent with a failure to respond to SHH , including the absence of the NKX2 . 2+ V3 interneuron progenitors that require high levels of SHH activity , and ISL1+ motor neurons that are shifted ventrally to span the midline ( Fig 1B vs . 1E and 1C vs . 1F ) . We have also examined the phenotype of Ttbk2sca11/null , confirming that these embryos have the same phenotype ( S1 Fig ) . Ttbk2sca11/sca11 embryos lacked cilia in mesenchymal cells surrounding the neural tube ( Fig 1D vs . 1G ) , as assayed by immunostaining for the ciliary membrane protein ARL13B . Mouse embryonic fibroblasts ( MEFs ) derived from Ttbk2sca11/sca11 embryos failed to recruit IFT proteins to the basal body and retained the cilium-suppressing centrosomal protein CP110 at the distal mother centriole ( S2 Fig ) , cellular defects identical to those originally reported for the Ttbk2null/null allele[4] . Truncated TTBK2 protein produced by the SCA11-associated mutations was reportedly detected in tissues from heterozygous knockin animals[14] , however our data indicate that these truncations are unable to function in cilia formation , despite the inclusion of the kinase domain . To better understand how the SCA11-associated mutations might lead to a dominant neurological condition in humans , we next undertook a series of studies to investigate the effect of these truncations on ciliogenesis in the presence of varied amounts of full length TTBK2 . Neither Ttbk2null/null [4] , Ttbk2sca11/null , nor Ttbk2sca11/sca11 embryos form any cilia ( Fig 1 , S1 and S2 Figs ) . However , in cells derived from Ttbk2null/null embryos , cilia can be fully rescued by expression of WT TTBK2-GFP via retroviral transduction . We previously found that overexpression of TTBK2SCA11 in WT fibroblasts using the same method modestly suppresses cilia formation[4] . This led us to propose that Ttbk2sca11 may be a dominant-negative ( antimorphic ) allele of Ttbk2 . To test this hypothesis , we expressed WT TTBK2-GFP in MEFs derived from both Ttbk2null/null and Ttbk2sca11/sca11 embryos using the same retroviral transduction system we previously employed for rescue experiments , and compared the ability of WT TTBK2 to rescue cilia formation in cells of these two genotypes . The frequency of cilia rescue was approximately 2-fold lower in Ttbk2sca11/sca11 MEFs compared to Ttbk2null/null MEFs ( 34 . 1 +/- 4 . 6% vs 66 . 2 +/-3 . 3%; p = 0 . 0002; Fig 2A and 2B ) . The intensity of ARL13B was also reduced in cilia of Ttbk2sca11/sca11 MEFs expressing TTBK2-GFP compared to cilia in Ttbk2null/null MEFs ( WT: 120 . 4 +/- 4 . 96 A . U . , Ttbk2bby/bby+ TTBK2-GFP: 103 . 2 +/- 3 . 37 A . U . , Ttbk2sca11/sca11+ TTBK2-GFP: 58 . 93 +/- 6 . 14 A . U . ; Fig 2C and 2D ) , although the cilia did not differ significantly in length ( WT: 3 . 468 +/- 0 . 154 , Ttbk2bby/bby+ TTBK2-GFP: 2 . 874 +/- 0 . 074 , Ttbk2sca11/sca11+ TTBK2-GFP: 3 . 438 +/- 0 . 194 , Fig 2C and 2E ) . Together , these results suggest that the ability of exogenous TTBK2-GFP to restore cilia in mutant fibroblasts is inhibited by the presence of the truncated TTBK2SCA11 protein in these cells . To begin to investigate how the truncated SCA11-associated protein might interfere with the function of WT TTBK2 , we tested whether TTBK2SCA11 could physically interact with full-length TTBK2 . Previous studies have found that TTBK2 molecules physically associate and that TTBK2 can phosphorylate its own C terminus[8] , suggesting that like other members of the CK1 family[17] , TTBK2 may form a homodimer and this association could have regulatory significance . We therefore tested the ability of different fragments of TTBK2 ( Fig 3A ) to interact with full-length TTBK2 by co-immunoprecipitation ( in all cases GFP or V5 tags were placed at the N terminus of the protein or protein fragment ) . Consistent with previous reports , V5-tagged full-length TTBK2 co-precipitates with full-length TTBK2-GFP when both constructs are expressed in HEK293T cells ( Fig 3B ) . Full-length TTBK2-V5 also co-precipitates with the C-terminus of TTBK2 ( TTBK2306-1243-GFP ) , but not with a SCA11-associated TTBK2 truncation ( TTBK21-443 ) ( Fig 3C ) . Thus , the C-terminus of TTBK2 ( amino acids 450–1243 ) is essential for this self-interaction . Consistent with this finding , TTBK2SCA11-V5 is also unable to co-immunoprecipitate with TTBK2SCA11-GFP . The loss of these interactions has implications for the regulation of TTBK2SCA11 as well as its interactions with substrates . As the SCA11 truncation does not bind the full-length protein , it is likely that the SCA11 protein acts as a dominant negative by competing with the full-length protein for binding to another protein or proteins . To test genetically whether the SCA11-associated mutations to Ttbk2 are antimorphic alleles , we again turned to the Ttbk2sca11 knockin mice . Since the human disease SCA11 is an adult-onset phenotype seen in individuals heterozygous for TTBK2 mutations , we examined the phenotype of adult Ttbk2scall/+ mice . At 3 months of age , the cerebellar architecture of the Ttbk2scall/+ animals was indistinguishable from that of wild type ( S3 Fig ) . SCA11 in humans reportedly results in mild ataxia , with patients having a normal lifespan and predominantly remaining ambulatory even years after the onset of symptoms[11 , 18–20] . We cannot yet rule out that the mice may have subtle or late-onset phenotypes . Another possibility is that the relative dosage of WT or full length TTBK2 required to maintain sufficient ciliary function and/or neuronal function and survival is lower in mice than it is in humans . To further test whether TTBK2SCA11 dominantly interferes with WT TTBK2 function in vivo , we next generated an allelic series of Ttbk2 . We generated mice that carry a gene trap allele of Ttbk2 ( Ttbk2tm1a ( EUCOMM ) Hmgu ) from ES cells obtained from the European Mutant Mouse Cell Repository ( EuMMCR ) . Although the targeting strategy was designed to trap splicing of an early Ttbk2 exon ( S4A Fig ) , the homozygous gene trap mice ( Ttbk2gt/gt ) were viable past weaning , developing variably penetrant hydrocephalus and polycystic kidneys by 6 months of age ( S4C and S4D Fig ) . Transcript analysis showed that this allele produced mRNAs with the predicted gene trap transcript and a wild-type RNA formed by splicing around the gene trap insertion ( S3B Fig ) . Consistent with this , by Western blot we detected a small amount of TTBK2 protein , running at the same molecular weight as WT TTBK2 ( S3E Fig ) . We conclude that Ttbk2gt is a partial loss-of-function ( hypomorphic ) allele that produces a reduced amount of wild-type , full-length TTBK2 protein . Consistent with the hypomorphic character of the gene trap allele , Ttbk2null/gt embryos had a phenotype intermediate between that of Ttbk2null/null and the Ttbk2gt/gt homozygotes: Ttbk2null/gt embryos and neonates were recovered at nearly Mendelian frequencies up to birth ( P0 ) but died by P1 ( S1 Table ) . At E15 . 5 , in contrast to Ttbk2gt/gt embryos , which showed wild-type morphology , Ttbk2null/gt embryos had fully penetrant polydactyly on all 4 limbs , consistent with a disruption in Hh-dependent limb patterning ( Fig 4A–4C , S2 Table ) . We reasoned that the Ttbk2gt allele , with lowered levels of TTBK2 protein , might provide a sensitized genetic background to better compare the effects of the Ttbk2null and Ttbk2sca11 alleles . Ttbk2sca11/gt embryos showed similar overall morphology to Ttbk2null/gt embryos at E15 . 5 , with fully penetrant polydactyly on all 4 limbs . While some Ttbk2sca11/gt neonates were recovered at P0 , they were present at a sub-Mendelian frequency: only 9 . 5% of pups ( 6/63 compared with 16/63 expected; p = 0 . 0189 ) recovered at birth from Ttbk2gt/+ x Ttbk2sca11/+ crosses genotyped as Ttbk2sca11/gt ( summarized in S1 and S2 Tables ) , suggesting some prenatal lethality . We compared ventral neural patterning in Ttbk2gt/gt , Ttbk2null/gt , and Ttbk2sca11/gt embryos to determine whether Ttbk2sca11/gt embryos had more severe disruption of SHH signaling than seen in Ttbk2null/gt embryos . We found that neural patterning in E10 . 5 Ttbk2gt/gt embryos was similar to that in Ttbk2gt/+ embryos ( Fig 4E , 4F , 4I and 4J ) , and Ttbk2null/gt embryos exhibited only mild defects in neural patterning . In these mutants , the distribution of motor neurons , labeled with Islet1 ( ISL1 ) was very similar to that of Ttbk2gt/+ though a small number of motor neurons were found at the ventral midline ( Fig 4G and 4M ) . We also observed an increase in the number of cells positive for both NKX2 . 2 and OLIG2 relative to Ttbk2gt/+ ( Fig 4K and 4N ) . In contrast , the ISL1+ motor neuron domain was shifted ventrally in Ttbk2sca11/gt embryos and ISL1+ cells were found at the ventral midline in all sections examined ( Fig 4H and 4M ) . We also observed extensive intermixing of OLIG2+ and NKX2 . 2+ progenitor populations , with a larger number of cells positive for both NKX2 . 2 and OLIG2 compared to other genotypes ( Fig 4L and 4N ) . In addition , OLIG2+ cells were often found at the ventral midline in the Ttbk2sca11/gt embryos , in contrast with the other genotypes in which this dramatic ventral shift was not observed ( Fig 4L and 4O ) . Together , these data are consistent with a more severe disruption in SHH-dependent patterning in the Ttbk2sca11/gt embryos . This enhanced SHH patterning phenotype , combined with the increase in embryonic lethality of the Ttbk2sca11/gt animals , provides genetic support for Ttbk2sca11 as a dominant negative allele . To assess whether the more severe developmental defects in Ttbk2sca11/gt embryos were due to greater defects in ciliary trafficking , structure , or stability , we analyzed cilia in MEFs derived from embryos of each genotype of the Ttbk2 allelic series . Following serum starvation , we found that a mean of 69 . 1 +/- 3 . 64% of Ttbk2gt/+ cells were ciliated ( Fig 5A and 5J ) , whereas in Ttbk2gt/gt and Ttbk2null/gt an average of 45 . 9 +/- 3 . 66% and 43 . 8 +/- 3 . 35% of cells were ciliated , respectively ( Ttbk2gt/+ vs Ttbk2gt/gt p = 0 . 0003; Ttbk2gt/+ vs Ttbk2null/gt p<0 . 0001; Ttbk2gt/gt vs Ttbk2null/gt p = 0 . 9772; Fig 5B , 5C and 5J ) . There were clearly fewer cilia in Ttbk2sca11/gt cells , with an average of 18 . 9 +/- 3 . 65% of cells having a cilium ( Ttbk2null/gt vs Ttbk2sca11/gt p<0 . 0001; Fig 5D and 5J ) . These findings suggest that the increased severity of the embryonic phenotypes correlates with a decrease in cilia number . While the mean cilia length was reduced in all of the Ttbk2 mutants relative to Ttbk2gt/+ cells , cilia length did not differ significantly between the different mutant allelic combinations ( Fig 5K ) . We observe similar overall trends examining the cilia of the embryo ( mesenchymal tissue surrounding the neural tube ) where the frequency of cilia is reduced in the hypomorphic mutants , and more dramatically reduced in Ttbk2sca11/gt ( S5 Fig ) . We also examined the percentage of cells with endogenous TTBK2 at the mother centriole or basal body in MEFs derived from each genotype . Relative to Ttbk2gt/+ , the percentage of cells with TTBK2 localized at the mother centriole/basal body in Ttbk2gt/gt trended towards being slightly reduced , though this was not statistically significant ( Fig 5E , 5F and 5L; 38 . 1 +/- 9 . 11% for Ttbk2gt/+ cells vs 25 . 5 +/- 2 . 76% for Ttbk2gt/gt; p = 0 . 052 ) . As expected , centriolar TTBK2 was further reduced in Ttbk2null/gt and Ttbk2sca11/gt cells ( Fig 5G , 5H and 5L; 13 . 0 +/- 1 . 86% and 11 . 1 +/- 1 . 12% , respectively; Ttbk2gt/+ vs Ttbk2null/gt p = 0 . 001; Ttbk2gt/+ vs Ttbk2sca11/gt p = 0 . 0006 ) , but there was no significant difference between these two genotypes ( Fig 5L; p = 0 . 9608 ) , implying that the presence of TTBK2SCA11 does not interfere with full length TTBK2 function by impairing its localization to the presumptive basal body . Since our data indicate that hypomorphic Ttbk2 mutants have shorter cilia in addition to forming cilia at a reduced frequency , we hypothesized that TTBK2 may be required for ciliary trafficking and stability as well as for the initiation of ciliogenesis . To further investigate the role of TTBK2 in cilia structure and/or trafficking following initial assembly of the axoneme , we examined trafficking of HH pathway components in the cilia of MEFS of each genotype . The transmembrane protein SMO is critical for HH pathway activation , and becomes enriched within the cilium upon stimulation of the pathway with SHH or various agonists[21 , 22] . We found that the amount of SMO in the cilium upon stimulation of cells with SMO agonist ( SAG ) was comparable between Ttbk2gt/+ cells and either Ttbk2gt/gt or Ttbk2null/gt cells , as measured by average intensity of SMO within the Acetylated α-Tubulin+ cilium ( mean intensity of 82 . 2 +/- 3 . 43 , 89 . 2 +/- 5 . 3 , and 78 . 9 +/- 4 . 65 A . U . , respectively ) . In contrast , SMO intensity was clearly reduced in the axonemes of Ttbk2sca11/gt cells ( mean intensity of 51 . 0 +/- 3 . 78 A . U . ) relative to Ttbk2null/gt ( p = 0 . 0003 ) , as well as each of the other genotypes ( vs Ttbk2gt/+ and Ttbk2gt/gt p<0 . 0001 ) , consistent with the exacerbated SHH signaling-related phenotypes observed in these mutant embryos ( Fig 6A and 6B ) . We also examined the trafficking of other HH pathway components within cilia in response to pathway activation . GLI2 is a transcription factor that mediates activation of target genes in response to HH ligands . GLI2 localizes to the tips of cilia , and becomes strongly enriched at the cilium tip in response to HH pathway activation[23] by SHH or SAG . There was no difference in GLI2 ciliary tip localization or intensity in response to SAG between Ttbk2gt/+ and any of the mutant alleles ( Fig 6C and 6D ) . KIF7 is the vertebrate homolog of the Drosophila protein COS2 and essential for the establishment and maintenance of the microtubule structure of the cilium in mammals , and for the stability of the axoneme[24 , 25] . Like GLI2 , KIF7 normally becomes enriched at the tips of cilia in response to SAG , however in contrast to the results with GLI2 , the percentage of cells with KIF7 localized to the tip of the cilium in the presence of SAG was significantly reduced in Ttbk2sca11/gt mutants relative to other genotypes ( Ttbk2gt/+: 82 . 2 +/-2 . 16% , Ttbk2gt/gt: 75 . 6 +/- 1 . 73% , Ttbk2null/gt: 62 . 6 +/- 7 . 93% , Ttbk2sca11/gt: 37 . 8 +/- 1 . 1%; Fig 6E and 6F ) , and was clearly less than in Ttbk2null/gt ( p = 0 . 012 ) . Thus , consistent with the more severe SHH-related patterning phenotypes observed in the Ttbk2sca11/gt embryos , trafficking of a subset of signaling molecules is impaired in cells of this genotype , consistent with possible disruptions in ciliary trafficking . To further investigate whether ciliary trafficking is disrupted Ttbk2 hypomorphic mutant cells , and whether this is exacerbated in Ttbk2sca11/gt cells in particular , we assessed other factors that control cilia trafficking and stability[25] . Since the shorter cilia observed for each of the Ttbk2 hypomorphic allele combinations relative to Ttbk2gt/+ cells could be due to defects in the protein machinery that mediates assembly of the ciliary axoneme , the intraflagellar transport ( IFT ) machinery , we examined the localization of IFT components in MEFs of each genotype . We measured the average intensity of IFT81 , IFT88 , and IFT140 within the ciliary axoneme of MEFs of each genotype . For IFT81 , the average intensity was not significantly changed within the axoneme in Ttbk2bby/gt and Ttbk2sca11/gt cells relative to the other genotypes ( Fig 7A and 7B; Ttbk2gt/+: 11 . 49 +/- 0 . 68 AU , Ttbk2gt/gt: 13 . 15 +/- 0 . 88 AU , Ttbk2null/gt: 16 . 07 +/- 1 . 93 AU , Ttbk2sca11/gt: 15 . 49 +/- 1 . 56 AU ) . For IFT88 , average intensity within the axoneme also varied only modestly by genotype , with modest increases in the average intensity within the axoneme in Ttbk2null/gt and Ttbk2sca11/gt relative to the other genotypes ( Fig 7D and 7E; Ttbk2gt/+: 40 . 54 +/- 1 . 73 AU , Ttbk2gt/gt: 45 . 06 +/- 2 . 55 AU , Ttbk2null/gt: 65 . 61 +/- 3 . 73 AU , Ttbk2sca11/gt: 51 . 89 +/- 3 . 00 AU ) . For IFT140 , the average intensity was reduced in the axoneme of Ttbk2sca11/gt cells relative to other genotypes ( Fig 7G and 7H; Ttbk2gt/+: 54 . 4 +/-3 . 58 AU , Ttbk2gt/gt: 65 . 55 +/- 6 . 11 AU , Ttbk2null/gt: 74 . 62 +/- 7 . 74 AU , Ttbk2sca11/gt: 37 . 13 +/- 3 . 26 AU ) . Thus , we identified a specific reduction in IFT-A machinery to the axonemes in Ttbk2sca11/gt cells . We also separately examined the average intensity of IFT81 , IFT88 and IFT140 at the basal body in cells of each genotype . We found that for each of the IFT proteins we examined , this basal body pool of IFT proteins was significantly reduced in average intensity in the Ttbk2null/gt and Ttbk2sca11/gt cells relative to Ttbk2gt/+ or Ttbk2gt/gt . For both IFT88 and IFT140 , average intensity at the basal body was significantly lower in Ttbk2sca11/gt cells than in Ttbk2null/gt ( Fig 7C , 7F and 7I ) . Taken together , this data suggests that reduced function of TTBK2 affects the localization of IFT components , particularly in the pools of IFT that form at the basal body , with the amount and distribution of IFT proteins within the axoneme affected to a lesser degree . These findings are aligned with our prior work showing that the basal body pools of IFT proteins are lost in Ttbk2null/null cells[4] , a defect that is thus far specific to TTBK2 and to proteins such as CEP164 that act upstream of TTBK2 to mediate its localization to the distal appendages . The defects we observed in the Ttbk2 hypomorphic mutants with respect to changes in IFT proteins as well as impaired trafficking of SMO and KIF7 led us to hypothesize that the cilia of these mutants may have defects in their structure and/or stability . Post-translational modifications of axonemal microtubules are often impaired in mutants , such as Kif7 , in which the structure and stability of the axoneme is disrupted[25] . We have therefore examined both acetylation and glutamylation of tubulin , two modifications associated with stability of the ciliary axoneme[26] . We have not observed any alterations in the acetylation of microtubules within the cilia of the Ttbk2 hypomorphic mutants , however we observe a marked effect on tubulin polyglutamylation , which is important for establishing ciliary structure and length[27 , 28] . Intensity of polyglutamylated tubulin within the cilium was comparable between Ttbk2gt/+ and Ttbk2gt/gt cells ( mean intensity of 111 . 8 +/- 5 . 42 and 93 . 64 +/- 5 . 66 A . U . respectively ) but was significantly reduced in Ttbk2null/gt cells ( mean intensity of 75 . 8 +/- 4 . 85 A . U . ) and further reduced in Ttbk2sca11/gt cells ( mean intensity of 55 . 1 +/- 4 . 32 A . U . ; Fig 8A and 8B ) . Reduction of tubulin polyglutamylation is associated with defects in cilium assembly and stability in a variety of organisms[27–29] , and recent evidence also suggests that hypo-glutamylation of ciliary microtubules promotes disassembly of primary cilia and impairs the trafficking of signaling molecules within the axoneme[30] . To further examine cilium stability across the Ttbk2 allelic series , we treated MEFs derived from embryos of each genotype with nocodazole . Because the microtubule doublets of the ciliary axoneme are more stable than cytoplasmic microtubules , treatment of WT cells with nocodazole for a short period has a limited effect on cilia length or frequency[25] . After treatment of MEFs with nocodazole for 10 or 30 minutes , the percentage of ciliated cells in WT or Ttbk2null/gt cells decreased modestly ( Fig 8C; for WT , 77 . 7 +/- 1 . 73% of cells were ciliated at T0 , 72 . 2 +/- 6 . 33% at 10 minutes , and 69 . 3 +/- 3 . 5% at 30 minutes; for Ttbk2null/gt , 54 . 6 +/- 2 . 31% of cells were ciliated at T0 , 45 . 5 +/- 0 . 64% at 10 minutes , and 46 . 3 +/- 1 . 96% at 30 minutes ) . In contrast , in Ttbk2sca11/gt cells treatment with nocodazole caused a rapid reduction in ciliated cells ( from 26 . 6 +/- 3 . 38% at T0 to 13 . 6 +/- 0 . 62% after 10 minutes of treatment , and 8 . 1 +/- 1 . 24% after 30 minutes of treatment ) . The length of the remaining Ttbk2sca11/gt cilia reduced over time in a manner that was proportional to the other genotypes: for Ttbk2sca11/gt cells , cilia length at 30 post nocodazole was 63 . 2% of the starting length , compared with 65 . 4% for Ttbk2null/gt and 55 . 9% for WT ( Fig 8D ) . These data suggest that cilium stability is more compromised in Ttbk2sca11/gt cells than in Ttbk2null/gt , with cilia in Ttbk2sca11/gt cells rapidly lost in the presence of nocodazole , consistent with the dominant negative nature of the sca11 allele . To further investigate the role of TTBK2 in cilium stability and ciliary structure , we performed transmission electron microscopy ( TEM ) on neural tube sections from E10 . 5 embryos of each genotype to assess the cilia ( Fig 8E–8H ) . We did not observe dramatic differences in the overall structure of cilia between Ttbk2gt/gt or Ttbk2null/gt and Ttbk2gt/+ . By contrast , the structure of the cilia in Ttbk2sca11/gt embryos differs noticeably from the other genotypes in a number of ways . Consistent with our previous findings from Ttbk2null/null cells which have normal distal and subdistal appendages , we observe these structures in Ttbk2sca11/gt cilia , as well as the extension of axonemes that contain microtubules . However the microtubules appear less distinct than in the cilia of the other genotypes , with the proximal cilium/transition zone in particular having a less organized appearance . In addition , we frequently observed what appear to be vesicles within the ciliary axonemes of Ttbk2sca11/gt embryos , but not in cilia in embryos of the other genotypes . These TEM images , together with our data showing that polyglutamylated tubulin is reduced in Ttbk2 hypomorphic mutants , and in particular in Ttbk2sca11/gt mutants , suggest that TTBK2 is important for the structure and stability of the microtubule axoneme . To examine the possible molecular mechanisms by which reduced TTBK2 could affect ciliary structure and stability , we tested whether a pathway important in cilium suppression and disassembly was altered upon reduced TTBK2 function . KIF2A is an atypical kinesin of the Kinesin 13 family that mediates microtubule depolymerization in a number of cellular contexts[31] . KIF2A was recently identified as a substrate of TTBK2 at the plus ends of cytoplasmic microtubules; in this context , phosphorylation of KIF2A by TTBK2 at S135 reduced the ability of KIF2A to bind microtubules , thereby impairing its depolymerase activity and stabilizing microtubules[10] . Given this association , we tested whether the localization of KIF2A was altered in Ttbk2 mutant cells . In WT MEFs , KIF2A was localized to the centrosome and was also occasionally seen within the proximal ciliary axoneme ( Fig 9A ) . Centrosomal localization was maintained in the Ttbk2 mutant alleles . However , quantification of KIF2A intensity at the base of ciliated cells revealed that the level of KIF2A at the centrosome was increased in Ttbk2null/gt ( mean pixel intensity of 43 . 1 +/- 1 . 96 A . U . ) cells relative to Ttbkgt/+ ( mean pixel intensity of 19 . 98 +/- 0 . 92 ) or Ttbk2gt/gt ( mean pixel intensity of 22 . 8 +/- 0 . 90 A . U ) . The intensity of KIF2A was further increased at the ciliary base of Ttbk2sca11/gt relative to all other genotypes ( mean pixel intensity of 53 . 0 +/- 1 . 87 A . U . , Fig 9B; p<0 . 0001 ) . This suggests that KIF2A accumulates at the ciliary base when TTBK2 levels are reduced , where it could contribute either to structural defects in cilia , or to the observed reduction in ciliated cells by promoting cilium disassembly , or both . To assess whether and how loss or reduction in S135 phosphorylation of KIF2A may contribute to the ciliary phenotypes seen in the TTBK2 hypomorphic mutants , we tested the effects of expressing a non-phosphorylatable variant of KIF2A ( S135A ) in WT MEFs . We found that in MEFs over-expressing KIF2AS135A , cilia were significantly shortened compared to those overexpressing WT KIF2A ( Fig 9C and 9D ) , although we did not observe a significant change in the percentage of ciliated cells between these conditions . Thus , we propose that increased activity of KIF2A in Ttbk2 hypomorphic mutants contributes to defects in ciliary stability and structure , and to the exacerbated phenotypes observed in the Ttbk2sca11/gt embryos and neonates .
In this study , we show that the human SCA11- associated mutations to Ttbk2 produce truncated proteins that interfere with the function of full-length TTBK2 in cilia formation . Consistent with our previous data showing that familial SCA11-associated mutations are unable to restore primary cilia in null mutant cells , our analysis of Ttbk2sca11/sca11 mutants revealed a phenotype that is essentially indistinguishable from that of our previously described ENU-induced null allele . Like Ttbk2null/null , homozygous SCA11 mutants lack cilia in all tissues examined at E10 . 5 , and the cells of these mutants exhibit an identical set of cellular defects to those of embryos lacking Ttbk2 . These results indicate that TTBK2SCA11 truncations are completely unable to function in mediating ciliogenesis , despite having an intact kinase domain and producing a protein product[14] . This inability to function in ciliogenesis is likely the result of the SCA11-associated truncations lack of the C-terminus , which we and others have shown is required to target TTBK2 to the basal body and for its interaction with the distal appendage protein CEP164[4 , 9 , 32] . In our prior studies we also found that expression of TTBK2SCA11-GFP in WT fibroblasts led to a modest but significant reduction in ciliogenesis[4] , consistent with the classical definition of a dominant negative[33] . We hypothesized based on this that the SCA11-associated mutations to Ttbk2 function as antimorphic alleles . In the current work , we present two major lines of evidence in support this hypothesis . First , we found that expression of WT TTBK2-GFP only partially rescues cilia formation in Ttbk2sca11/sca11 mutant cells whereas full rescue is achieved by stable expression of the same construct in Ttbk2null/null cells . This is seen both at the level of ciliogenesis , where many fewer ciliated cells are found in rescued Ttbk2sca11/sca11 cells and also with respect to the structure of the cilium: ARL13B localization is significantly impaired in the rescued Ttbk2sca11/sca11 relative to rescued null mutant cells . Second , we present genetic evidence for the dominant negative function of Ttbk2sca11 . The combination of Ttbk2sca11 with a hypomorphic allele that produces a reduced amount of TTBK2 protein ( Ttbk2gt ) results in more severe phenotypes than the null allele in combination with Ttbk2gt on the same genetic background . We propose a model wherein TTBK2’s functions in cilium assembly are highly dosage sensitive , with alterations in the amount of functional TTBK2 protein below a certain threshold causing a range of phenotypes related to defects in ciliary trafficking and signaling . In human SCA11 patients , the presence of SCA11 truncated protein is sufficient to cause a phenotype limited to a specific tissue- the cerebellum . In mice , we did not identify any changes in the architecture of the cerebellum between Ttbk2sca11/+ animals and their WT siblings by 3 months of age . While we can’t yet exclude the emergence of more subtle defects occurring at advanced age , it does not appear one allele of Ttbk2sca11 is sufficient to cause phenotypes recapitulating human SCA11 in the presence of a second WT allele of Ttbk2 , ( ie Ttbk2sca11/+ ) in mice . However , on a sensitized background with a reduced amount of full-length TTBK2 , the dominant negative effects of TTBK2SCA11 become apparent , such as in the allelic series . Our studies of the ciliary defects occurring in the Ttbk2 allelic series have also yielded valuable insights about the role of TTBK2 in cilia formation and trafficking . Our prior work based on a null allele of Ttbk2 demonstrated the essential role played by this kinase in initiating cilium assembly upstream of IFT . However , examination of hypomorphic alleles in this study points to additional requirements for TTBK2 following initial cilium assembly . For example , cilia are shorter in cells derived from all of the hypomorphic Ttbk2 alleles compared with WT or Ttbk2gt/+ cells , pointing to a role for TTBK2 in cilia structure and trafficking . Identifying the molecular targets of TTBK2 in both cilium initiation and in ciliary trafficking and/or stability will be critically important to our understanding of the pathways that regulate ciliogenesis . We identified modest disruptions in the concentration of IFTB and IFTA components in our Ttbk2 hypomorphic allelic combinations , consistent with a role for TTBK2 particularly in maintaining the basal body pools of IFT proteins , in addition to the requirement for TTBK2 in the recruitment of IFT components to the basal body that we identified previously in our analysis of Ttbk2 null mutant cells[4] . Identifying the mechanisms by which TTBK2 mediates the localization of IFT proteins to the basal body , as well as testing whether TTBK2 contributes to IFT mediated ciliary trafficking will be a focus of our future studies . We have also uncovered a role for TTBK2 in maintaining the stability of the ciliary axoneme , with these defects becoming particularly evident in Ttbk2sca11/gt mutant cells . Consistent with a requirement for TTBK2 in cilia structure , KIF7 is reduced in Ttbk2sca11/gt cells compared with Ttbk2null/gt cells with respect to the percentage of cilia that are positive for KIF7 . The Ttbk2sca11/gt cells also exhibit a subset of the defects found in Kif7-/- cells , including a reduction in polyglutamylated tubulin[25] . Unlike Kif7-/- cells however , we did not observe any reduction in tubulin acetylation in any of the Ttbk2 hypomorphic cells . The Ttbk2sca11/gt cells do exhibit increased instability in the presence of nocodazole . The highly modified microtubules of the cilium are typically relatively resistant to this microtubule-depolymerizing drug [25 , 34] , and in WT or Ttbk2null/gt cells neither the proportion of ciliated cells nor the length of the cilium changes dramatically when the cells are treated with nocodazole for up to 30 minutes . In contrast , in the Ttbk2sca11/gt cells the percentage of ciliated cells drops dramatically upon treatment with nocodazole , consistent with a requirement for TTBK2 in the stability of the axonemal microtubules . In addition , increased levels of the microtubule depolymerizing kinesin KIF2A are present at the centrosome of ciliated cells in the Ttbk2 hypomorphic mutants , with the highest amounts seen in Ttbk2sca11/gt cells . This suggests that TTBK2 may oppose the activity the PLK1-KIF2A cilium disassembly pathway , and that an increase in the activity of this pathway in the Ttbk2 hypomorphic mutants contributes to the reduction in ciliated cells , in addition to defects in cilium stability . Loss of tubulin glutamylation in cilia has also recently been shown to perturb ciliary trafficking and the enrichment of HH pathway components within cilia upon stimulation of cells with SAG[35] . Thus , the reduction in SMO enrichment that we observed in the Ttbk2sca11/gt cells could result from the additional reduction in tubulin glutamylation we see within the cilia of these cells relative to other genotypes , with disrupted trafficking in HH pathway components in turn leading to exacerbated embryonic phenotypes related to HH-dependent patterning . The additional impairment of TTBK2 function in the Ttbk2sca11/gt animals results in a greater perturbation of cilia than the defects seen in Ttbk2null/gt cells . These include reduced numbers of cilia , disrupted cilium stability , and impaired trafficking of signaling molecules such as SMO to the axoneme , although we have not yet precisely defined the biochemical mechanisms by which the human disease-associated truncations interfere with TTBK2 function . Our data argue against a model where TTBK2SCA11 directly binds to full length TTBK2 and inhibits its function through a direct association . Rather , it seems more likely that TTBK2SCA11 , having lost critical regulatory motifs as well as the ability to efficiently translocate to the centrosome , may sequester some important TTBK2 substrate or substrates , resulting in the further impairment of cilia structure and signaling that in turn causes the modest exacerbations in SHH-dependent developmental patterning . While our data indicate that the SCA11-associated Ttbk2 mutations interfere with cilia formation and stability , pointing to a strong possibility that SCA11 pathology is related to disrupted ciliary signaling , we cannot exclude the possibility that TTBK2 has non-ciliary roles within the brain that could also contribute to neural degeneration . For example , TTBK2 phosphorylates Synaptic Vesicle Protein 2A , and this event is important for the formation and release of synaptic vesicles[36] . The mechanisms of TTBK2 regulation and the specific substrates of this kinase in cilium assembly , as well as possible non-ciliary roles for TTBK2 within the brain are key topics in our ongoing research . Having shown that TTBK2SCA11 is both unable to mediate cilium assembly and also impairs the function of TTBK2WT in ciliogenesis , another important area of investigation is the relationship between cilia and ciliary signaling pathways and the maintenance of neural connectivity and function .
The use and care of mice as described in this study was approved by the Institutional Animal Care and Use Committees of Memorial Sloan Kettering Cancer Center ( approval number 02-06-013 ) and Duke University ( approval numbers A246-14-10 and A218-17-09 ) . Euthanasia for the purpose of harvesting embryos was performed by cervical dislocation , and all animal studies were performed in compliance with internationally accepted standards . We used two previously described alleles of Ttbk2: Ttbk2null is an ENU-induced allele ( also called Ttbk2bby ) [4] , and Ttbk2sca11 is a knockin recapitulating one of the familial SCA11-associated mutations[14] . Genotyping for both of these alleles was performed as previously described . Ttbk2 “knockout first” genetrap ( Ttbk2tm1a ( EUCOMM ) Hmgu , here referred to Ttbk2gt ) targeted ES cells were purchased from the European Mutant Mouse Consortium . One clone ( HEPD0767_5_E08 , parental ESC line JM8A3 . N1 , agouti ) was injected into host blastocysts by the Mouse Genetics Core Facility at Sloan Kettering Institute . Resulting chimeric male mice were bred to C57BL/6 females to test germline transmission and obtain heterozygous mice . PCR genotyping ( F: ATACGGTTGAGATTCTTCTCCA , R1: TCTAGAGAATAGGAACTTCGG , R2: TGCAATTGCATGACCACGTAGT ) yields a band corresponding to the mutant allele at 407bp and to the WT allele at 762bp . To obtain embryos at the identified stages , timed matings were performed with the date of the vaginal plug considered embryonic day ( E ) 0 . 5 . Pregnant dams were sacrificed by cervical dislocation and embryos were fixed in either 2% ( E11 . 5 or earlier ) or 4% ( later than E11 . 5 ) paraformaldehyde ( PFA ) overnight at 4C . For cryosectioning , tissue was cryoprotected in 30% Sucrose overnight and embedded in Tissue Freezing Medium ( General Data TFM-5 ) . Tissue was sectioned at 16μm thickness . To harvest tissues from adult mice , animals were anesthetized with 12 . 5mg/mL Avertin , and a transcardially perfused with of Phosphate Buffered Saline ( PBS ) followed by 4% PFA . Kidneys and brains were dissected and incubated in 4% PFA for an additional 2 hours at 4°C . Tissue was then prepared for cryosectioning as described above . MEFs were isolated from embryos at either E10 . 5 or E12 . 5 , and maintained as previously described[37] . To induce cilia formation , cells were shifted from 10% to 0 . 5% fetal bovine serum ( FBS ) and maintained in low serum conditions for 48 hours . Cells were grown on coverslips and fixed in 4% Paraformaldehyde ( PFA ) in Phosphate Buffered Saline ( PBS ) for 5 minutes at room temperature followed by methanol for 5 minutes at -20C . Cells were then washed in PBS + 0 . 2% Triton X-100 ( PBT ) and blocked in PBT + 5% FBS + 1% bovine serum albumin for 30 minutes . Cells were then incubated with primary antibodies diluted in blocking solution overnight at 4 °C , and finally incubated with Alexa-coupled secondary antibodies and DAPI in blocking solution for 30 minutes at room temperature and affixed to slides for microscopy . Embryonic and adult tissue sections were collected onto slides , dried , washed in PBT + 1% serum , and incubated with primary antibodies as described above . The human Kif2a Gateway-ready clone was obtained from the Human ORFeome Collection ( Dharmacon clone BC031929 ) . The KIF2AS135A mutation was introduced via site directed mutagenesis using the Quick Change II Mutagenesis Kit ( Agilent ) . Using Gateway LR Clonase ( Invitrogen ) , both KIF2AWT and KIF2AS135A clones were transferred into Gateway Destination vectors compatible for retroviral vector expression modified to contain eGFP and FLAG tags . Retroviral transduction was carried out as previously reported[4] . The SMO antibody was raised in rabbits ( Pocono Rabbit Farm and Laboratory Inc . ) using antigens and procedures described[20]; diluted 1:500 . Antibodies against KIF7[23] ( 1:1000 ) , ARL13B [38] ( 1:2000 ) , GLI2 [39] ( 1:2000 ) and TTBK2 [14] have been previously described . Commercially available antibodies used in these studies were: mouse anti- NKX2 . 2 , ISL1 ( Developmental Studies Hybridoma Bank , each 1:10 ) ; mouse anti- Pericentrin , ( BD Biosciences #611814 , 1:500 ) γ-Tubulin ( Sigma SAB4600239 , 1:1000 ) , Acetylated α-Tubulin ( Sigma T6793 , 1:1000 ) , polyglutamylated Tubulin ( Adipogen AG-20B-0020 , 1:2000 ) ; rabbit anti- IFT88 ( Proteintech 13967-1-AP , 1:500 ) , rabbit anti- IFT81 ( Proteintech 1174-1-AP , 1:1000 ) , rabbit anti- IFT140 ( Proteintech 17460-1-AP 1:500 ) , mouse anti-FLAG ( Sigma F1804 , 1:1000 ) , rabbit anti- KIF2A ( Abcam ab37005 , 1:500 ) , TTBK2 ( Proteintech 15072-1-AP , 1:1000 ) , Calbindin ( Cell Signaling Technology 13176-S , 1:250 ) , VGLUT2 ( EMD Millipore AB2251 , 1:2500 ) . Immuno-fluorescence images were obtained using a Zeiss AxioObserver wide field microscope equipped with an Axiocam 506mono camera and Apotome . 2 optical sectioning with structured illumination . Z-stacks were taken at 0 . 24μm intervals . Whole mount images of embryos and tissues were captured with a Zeiss Discovery V12 SteREO microscope equipped with an Axiocam ICc5 camera . Image processing and quantifications were performed using ImageJ . To quantify the signal intensities of ciliary proteins , Z stack images were captured using the 63X objective . A maximum intensity projection was then created for each image using ImageJ , background was subtracted . Cilia were identified by staining with Acetylated α-Tubulin and γ-Tubulin . Each cilium or portion of the cilium was highlighted using either the polygon tool or the line tool ( for line-scan analysis ) , and the mean intensity was recorded for the desired channel ( measured on an 8 bit scale ) , as described[25] . To measure the mean intensity , ImageJ software was used to calculate total intensity divided by the area selected . Measurements taken within the cilium therefore take into account the length per measurement recorded . Statistical analysis was done with the Prism7 statistical package ( GraphPad ) . E10 . 5 embryos were fixed in a 0 . 1M cacodylic acid buffer ( pH7 . 4 ) containing 2% PFA and 2 . 5% glutaraldehyde . The samples were washed three times with 0 . 1M cacodylic acid buffer and post stained with 1% osmium tetroxide in cacodylic buffer for 1h . The samples were then prestained with 1% uranyl acetate ( Polaron Instruments Inc . , Hatfield , PA ) overnight at 4°C . The samples were washed and carried through acetone dehydration steps . Infiltration was done using the Epon embedding kit ( EMS ) . Samples were ultrathin sectioned ( 60-70nm ) on a Reichert Ultracut E ultramicrotome and sections were stained with 2% uranyl acetate in 50% ethanol for 30min and SATO’s lead stain for 1min . Samples were imaged on a Philips CM12 electron microscope . HEK-293T cells were transfected with constructs for tagged proteins of interest using Lipofectamine 3000 ( Thermo Fisher ) according to the manufacturer’s instructions . Constructs used were TTBK2FL-GFP , TTBK2FL-V5 , TTBK2SCA11-V5 ( 1-443aa , Ttbk2Cterm-GFP ( 306-1243aa ) . For western blots , cells or tissues were lysed in buffer containing 10mM Tris/Cl pH7 . 5 , 150mM NaCl , 0 . 5mM EDTA , 1% Triton , 1mM protease inhibitors ( Sigma #11836170001 ) and 25mM β-glycerol phosphate ( Sigma 50020 ) , and total protein concentration was determined using a BSA Protein Assay Kit ( Thermo Fisher #23227 ) . For co-IP experiments , cells were lysed in buffer containing 20mM Tris-HCl pH7 . 9 , 150mM NaCl , 5mM EDTA , 1% NP-40 , 5% glycerol , 1mM protease inhibitors and 25mM β-glycerol phosphate . Immunoprecipitation of lysates was performed using analysis was done using GFP-Trap beads ( Chromotek GTA-20 ) blocked with 3% BSA in Co-IP lysis buffer overnight prior to pull-down . rabbit α-GFP ( Invitrogen A11122 , 1:10 , 000 ) , mouse α-V5 ( Invitrogen R96025 , 1:7 , 000 ) , HRP-conjugated secondaries ( Jackson ImmunoResearch ) . Quantification of the cerebellar tissue was done using ImageJ software . Images for the molecular layer analysis were taken at 20x . For measuring the molecular layer , a line was drawn from the top of the PC cell soma to the pial surface and the distance was recorded . That same line was then brought down from the pial surface to the top of the nearest VGLUT2 puncta along that line , the distance was recorded , and a ratio was calculated . Measurements were pooled equally from both sides of the primary folia of the cerebellum , and from four slices per animal . Images for the VGLUT2 analysis were 10μm thick z-stacks taken at 63x . VGLUT2 puncta analysis was performed using the ImageJ “Analyze Particles” plug-in with the following stipulations: Size exclusion: 0 . 5-infinity , Circularity: 0–1 . Measurements were pooled from 5 areas in the cerebellum , and from four slices per animal . RNA was extracted from brains dissected from p30 animals using the Qiagen RNeasy Mini Kit ( Qiagen , 74104 ) . cDNA was then made from 1μg of RNA using the BioRad iScript cDNA Synthesis Kit ( BioRad , 1708891 ) . PCR primers were designed to span the exon 4–5 boundary of Ttbk2 ( F: ATGCTCACCAGGGAGAATGT , R: TGCATGACCACGTAGTTGAAA ) , lacZ ( F: AGCAGCAGTTTTTCCAGTTC , R: CGTACTGTGAGCCAGAGTTG ) , and GAPDH ( F: ACCACAGTCCATGCCATCAC , R: TCCACCACCCTGTTGCTGTA ) .
Indicated statistical comparisons were performed using Graphpad Prism7 . For multiple comparisons , a Tukey-Kramer post-hoc test was performed . | Defects in primary cilia structure and function are linked to a number of recessive genetic disorders , now collectively referred to as ciliopathies . Most of the characteristics of these disorders arise from disruptions to embryonic development , with the requirements for primary cilia in adult tissues being less well-defined . We previously showed that a kinase associated with an adult-onset neurodegenerative condition is required for cilium assembly and ciliary signaling during development . Here , we show that the human disease-associated mutations act as mild dominant negatives , interfering with the function of the full-length protein in cilia formation and ciliary signaling . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods",
"Statistics"
] | [
"centrosomes",
"microtubules",
"neuroscience",
"alleles",
"motor",
"neurons",
"developmental",
"biology",
"embryos",
"tubulins",
"cellular",
"structures",
"and",
"organelles",
"cytoskeleton",
"embryology",
"animal",
"cells",
"proteins",
"genetic",
"loci",
"hedgehog",
"sig... | 2018 | Spinocerebellar ataxia type 11-associated alleles of Ttbk2 dominantly interfere with ciliogenesis and cilium stability |
Tuberculosis ( TB ) is one of the most devastating infectious diseases worldwide . Whilst global burden estimates for M . tuberculosis infection ( MtTB ) are well established , accurate data on the contribution of zoonotic TB ( zTB ) caused by M . bovis or M . caprae to human TB are scarce . The association of M . bovis infection with extrapulmonary tuberculosis has been suggested repeatedly , though there is little scientific evidence available to support this relationship . The present study aimed to determine globally the occurrence of extrapulmonary TB and the primary site ( i . e . primary body location affected ) of zTB in comparison with MtTB , based on previously published reports . A systematic literature review was conducted in 32 different bibliographic databases , selecting reports on zTB written in English , French , German , Spanish or Portuguese . Data from 27 reports from Africa , America , Europe and the Western Pacific Region were extracted for analyses . Low income countries , in Africa and South-East Asia , were highly underrepresented in the dataset . The median proportion of extrapulmonary TB cases was significantly increased among zTB in comparison with data from registries of Europe and USA , reporting mainly MtTB cases ( 47% versus 22% in Europe , 73% versus 30% in the USA ) . These findings were confirmed by analyses of eight studies reporting on the proportions of extrapulmonary TB in comparable populations of zTB and MtTB cases ( median 63% versus 22% ) . Also , disparities of primary sites of extrapulmonary TB between zTB and MtTB were detected . Our findings , based on global data , confirm the widely suggested association between zTB and extrapulmonary disease . Different disability weights for zTB and MtTB should be considered and we recommend separate burden estimates for the two diseases .
Tuberculosis ( TB ) is a disease distributed worldwide , accounting for an estimated 1 . 5 million deaths annually [1] . The primary causative agent of human TB is Mycobacterium tuberculosis . Tuberculosis in cattle and goats is typically caused by M . bovis and M . caprae , respectively . Zoonotic transmission of these bacteria from livestock to humans is well documented [2] . Whilst M . tuberculosis-related human TB ( MtTB ) is one of the most devastating infectious diseases worldwide , the public health relevance of zoonotic TB ( zTB ) appears to be minor in industrialized countries , where milk pasteurization is regularly performed [3] . However , in developing countries where control programmes for TB in livestock are absent and milk pasteurization is not done routinely , zTB may be of considerable importance [4]–[6] . A recent systematic literature review estimated incidence rates of zTB of approximately 7/100'000 human population/year in Africa [7] . This relatively low incidence rate notwithstanding , the study revealed that considerable rates of zTB exist in certain high-risk populations . Moreover , cases of zTB could be considerably under-diagnosed and consequently under-reported , particularly in developing countries [4]–[6] . Disease burden estimates are commonly calculated in terms of disability-adjusted life years ( DALYs ) , taking into account disease frequency but also its severity and long-term effects ( sequelae ) [8] . The disability weights of zTB caused by M . bovis and M . caprae may differ from those of MtTB . A main reason may be associated with the different primary sites ( i . e . primary body location affected ) of disease for zTB and MtTB and potentially distinct secondary consequences resulting from this . An association of M . bovis infection with extrapulmonary TB has been suggested repeatedly in the literature , and current dogma suggests zTB to be transmitted to humans primarily by ingestion of contaminated products of animal origin rather than by aerosols , a hypothesis that would be in line with the more frequent extrapulmonary presentation of zTB in humans [3] , [6] , [9] . This view is mainly based on analysis of historical data from high-income countries and few other studies in geographically restricted regions or including small study populations . More comprehensive analyses investigating on a global scale the relative proportion of pulmonary versus extrapulmonary TB cases in humans and the primary sites of disease for zTB infection are missing . Other factors potentially affecting the clinical picture of zTB appear to be age , sex and co-morbidities ( especially co-infection with the Human Immunodeficiency Virus; HIV ) . For example , an early study conducted in the UK revealed that TB caused by M . bovis in children was nearly always extrapulmonary [10] . Males may have a higher risk of infection with zTB due to their more frequent intensive contact to cattle , either at farms or in abattoirs [11] . The present study was mandated by the World Health Organization ( WHO ) Foodborne Disease Burden Epidemiology Reference Group . It aimed , based on available scientific evidence , to determine the proportion of extrapulmonary TB and the primary site of zTB and its association with patient demographic parameters and HIV co-infection status . It combines for the first time observational and intervention studies on this topic from all geographical regions and without restriction to specific study populations and used systematically selected high-quality data to explore the association between zTB and extrapulmonary TB .
A systematic multi-lingual literature search was performed according to Cochrane guidelines with certain modifications ( http://cochrane-handbook . org/ ) . Thirty-two different bibliographic databases were searched for potentially relevant reports on putative zTB cases ( M . bovis or M . caprae infections ) in humans published until March 2010 using a highly sensitive search syntax ( Tables S1 and S2 ) . We included all types of observational and interventional studies on zTB or M . bovis and M . caprae infections in humans , unless the study reported exclusively on cases with evident human-human transmission . Before and after removal of duplicated reports 18'485 and 12'176 records , respectively , were identified ( Figure 1 ) . Titles and abstracts were screened to exclude reports which were unlikely to contain information on zTB cases; 1'203 potentially relevant reports remained of which 447 ( 37% ) were available online and assessed for eligibility . We focused on reports available online for convenience while this was an important factor to improve the efficiency of the work . Moreover , it can be assumed that reports available online are of higher quality and that most of the more recent reports were available . Eligible records were written either in English , French , German , Spanish or Portuguese . Additionally , studies had to report on either the occurrence of extrapulmonary TB , primary site or sequelae of the disease . They had to include at least 10 individuals with zTB . No restrictions were made on the year when the study was undertaken . Eligibility of the relevant reports was assessed independently by three operators on 100 randomly selected reports . Ambiguities and diverging judgements were examined in order to harmonize the selection procedure . The remaining records were randomly assigned to one operator only . Thirty-seven reports were considered eligible for data extraction . Data was extracted stratified by WHO region , sex , age group and HIV co-infection of the patients , where possible . Data were sought for 20 variables ( Table S3 ) . After harmonizing the procedure of data extraction done by three operators based on 15 reports , the remaining reports were randomly allocated to one operator only . If any of the included reports were referring to relevant accessible external data which were not included in the database during earlier steps , these data were included as well in the analysis . Data on more than one study setting ( different geographical regions and study periods ) were available from two reports which enlarged the database by five records . A total of 15 reports had to be excluded for different reasons ( Figure 1 ) . The final database included 27 records from 26 different reports . Within these reports , differentiation of M . bovis , M . caprae and M . tuberculosis was done by molecular ( e . g . PCR , spoligotyping ) , biochemical or both methods , or was not further specified for six reports . Anonymized human medical data was used . Human TB registries from the USA and Europe were screened for information on the proportions of extrapulmonary TB cases and the primary sites of the disease . The reports did not distinguish between different species of the M . tuberculosis complex and consequently included M . bovis cases . Data from registries of the USA for the years 2005 to 2010 was used ( Reported Tuberculosis in the United States , Department of Health and Human Services , [12] ) . For Europe , data were sourced from the EuroTB reports ( InVS/KNCV , funded by the European Commission , [13] ) and from the Tuberculosis surveillance in Europe reports ( ECDC and WHO , [14] ) for the years 1998 to 2006 and 2007 to 2010 , respectively . The former included data from the European Economic Area ( EEA ) states , Israel and Switzerland; the latter included data of the EEA states only . The primary sites of TB infection were categorized according to the ICD10 ( International Statistical Classification of Diseases and Related Health Problems 10th Revision ) online tool ( http://apps . who . int/classifications/apps/icd/icd10online/ ) . As an exception , we classified TB of the pleura as extrapulmonary , in agreement with most reports and registries that classify pleural TB as extrapulmonary TB , unless co-existing in a patient with pulmonary TB . Affected lymph nodes ( also including non-defined lymph nodes ) were counted as extrapulmonary TB whereas affected mediastinal lymph nodes were classified as pulmonary TB . Tuberculosis of the nervous system included meningeal TB , TB of the brain and spinal cord and of undefined nervous tissue . Tuberculosis of the intestines , peritoneum and mesenteric glands were categorized as intestinal TB . Patients with both pulmonary and extrapulmonary TB were defined as extrapulmonary cases throughout the study for both zTB and MtTB . We compared the frequency of extrapulmonary disease and of primary sites of zTB with MtTB in two ways . First , zTB cases from the selected reports were compared with cases reported in TB registries ( mainly M . tuberculosis ) , using the one sample Wilcoxon signed rank test . Second , information on comparable populations of M . bovis and M . tuberculosis cases within the same study were reported in eight reports regarding the occurrence of extrapulmonary TB and in four reports regarding the primary sites of TB infection ( Table S4 , footnote 3 and 4 ) . The Wilcoxon matched pairs signed rank test was used for comparing the reported proportions . No weighting by study size was applied . The bibliography database was stored in Reference Manager v11 . 0 . 1 bibliographic software . The data extracted was collated and stored in an access file . Data analysis was done using the statistical software R version 2 . 13 . 1 ( http://cran . r-project . org/ ) .
Twenty-seven reports from four different WHO regions ( Africa , 1 report; America , 8; Europe , 15; Western Pacific region , 3 ) were included in the present study ( Tables S4 and S5 , Figure 2 ) . The single study from Africa was done in Madagascar [15] . Half of the reports from Europe were carried out in the UK and Ireland . Six reports originated from the USA , 50% of which reported data from San Diego County , California , an area with a high proportion of Hispanic residents [16]–[18] . Reports included data from 1927–2007 whereas only 7 reports included data from earlier than 1980 ( Table S4 ) . Two studies from Europe reported on TB caused by M . caprae [19] , [20] while the remaining of the zTB cases reported was caused by M . bovis . Sample sizes of zTB patients ranged from 10–296 cases with a median of 69 cases . Information on the proportion of extrapulmonary TB cases among all zTB patients was presented in 26 of the studies included ( Table S4 , footnote 1 ) and varied substantially between and within WHO region ( Table 1 , Figure 3 ) . The region with the highest median value of extrapulmonary TB cases ( 73% , range 46–95% ) was the Americas . Two studies from USA , which reported high proportions of extrapulmonary TB ( 95% and 74% , respectively ) , were conducted in specific populations . The first report included children only ( 0–15 years ) [18] and the second was related to a food-borne outbreak caused by unpasteurized cheese , where a high proportions of extrapulmonary TB is expected [21] . Nevertheless , two other studies conducted in representative populations from the State of Michigan and the whole of the USA covering a period of 10 years , also reported proportions of 73% and 74% [22] , [23] . According to the estimates in the studies conducted in Europe , the median proportion of extrapulmonary TB was 47% ( 21–99% ) . The only report from Africa presented a low proportion of extrapulmonary TB ( 11% ) [15] . The three studies included from the Western Pacific region reported proportions of extrapulmonary TB from 13–43% with a median of 28% [11] , [24] , [25] . The proportion of extrapulmonary zTB based on our dataset was significantly higher than within TB cases ( mostly consisting of MtTB ) reported in TB registries . For Europe , the median proportion of extrapulmonary TB among cases of zTB was more than two-fold of that of MtTB ( 47% for zTB vs . 22% for MtTB , n = 15 , V = 119 , p<0 . 001 ) . For the USA , this difference was even more pronounced ( 73% for zTB vs . 30% for MtTB , n = 6 , V = 21 , p = 0 . 031 ) . These findings were confirmed by analyses of the proportions of extrapulmonary TB within the eight studies reporting on comparable populations of both , zTB cases and MtTB cases ( Table S4 , footnote 3 ) . The proportion of extrapulmonary TB was significantly higher among zTB cases than among MtTB cases ( median 63% vs . 22% , n = 16 , V = 36 , p = 0 . 008; Figure 4 ) . Four of these reports originated from the USA [16]–[18] , [22] with three from the San Diego County , California , two from England [26] , [27] and one each from Ireland [28] and New Zealand [25] . Only one eligible study reported data stratified by sex and found a more than two-fold increased proportion of extrapulmonary cases in women ( 21/43 patients , i . e . 49% ) than in men ( 20/103 patients , i . e . 19% ) [11] . The primary sites of extrapulmonary zTB were reported and analysed in 19 reports ( Table S4 , footnote 2 ) . Overall , lymph nodes ( median of the proportions of affected lymph nodes among all extrapulmonary cases within each study = 30% ) and the genitourinary system ( median = 25% ) were most often affected , followed by bones and joints ( median = 13% ) , intestines and peritoneum ( median = 6% ) and the nervous system ( median = 5% ) ( Table 2 ) . This frequency distribution did not differ greatly between studies from Europe and the Americas with the exception of genitourinary TB having been detected in only two of five studies conducted in the Americas [29] , [30] whereas in 8 of 10 studies conducted in Europe . Comparing the primary sites of extrapulmonary zTB of our dataset with the data from European TB registries reporting mainly MtTB cases , extrapulmonary zTB was negatively associated with infections of the pleura ( median = 0% vs . 27% in MtTB , n = 10 , V = 1 , p = 0 . 004 ) and by trend with miliary TB ( median = 0% vs . 6% in MtTB , n = 10 , V = 10 , p = 0 . 07 ) but positively associated by trend with infections of the genitourinary system ( median = 25% vs . 10% in MtTB , n = 10 , V = 46 , p = 0 . 07 ) and bones and joints ( median = 17% vs . 8% in MtTB , n = 10 , V = 46 , p = 0 . 07 ) . For the USA , no different lesion location was detected when comparing data from TB registries with those from zTB reports . Information on the primary sites of extrapulmonary TB cases within comparable populations of zTB and MtTB cases were reported in two reports from the USA [18] , [22] , and one each from Mexico [29] and England [26] of our dataset ( Table S4 , footnote 4 ) . Lymph nodes ( in all four reports ) , the genitourinary tract , the intestine tract and the skin tended to a positive association for extrapulmonary zTB ( Figure S1 ) . In contrast , bones and joints , the nervous system and the pleura were more often affected by trend in extrapulmonary MtTB cases . No significant difference in primary sites between zTB and MtTB was found using the Wilcoxon matched pairs signed rank test .
Global data on zTB infection and its relation to extrapulmonary TB are scarce . Studies investigating these aspects are needed in order to accurately estimate the global burden of this disease and to explore how it differs from the burden of MtTB . Moreover , until now , the assumption of zTB being associated with extrapulmonary TB , has been based mostly on historical data from high-income countries and on few individual studies from geographically restricted regions or including small study populations . The present study analysed the frequency distribution of extrapulmonary zTB based on previously published reports . By combining the results from multiple reports without restriction to geographical regions or population groups , it aimed to explore the best possible evidence of the association between zTB and extrapulmonary TB . Additionally , the findings were compared to MtTB using registries on human TB infections and comparing comparable populations of M . bovis and M . tuberculosis infected patients within the same studies . To our knowledge , this is the first study globally investigating the association between zTB and extrapulmonary TB as well as the primary sites of infection of extrapulmonary zTB . Our study showed an increased proportion of extrapulmonary TB among zTB compared to MtTB . While the comparison between zTB of the individual records and MtTB of the registries may be questionable because of diverse study settings , the comparison between zTB and MtTB within the same study is more robust and revealed an association of zTB and extrapulmonary TB as well . Similar results were reported previously , but mostly performed in geographically restricted regions ( e . g . [9] , [31] ) . The reason for this association may be explained by the different transmission routes , as zoonotic infection is primarily caused by consumption of unpasteurized dairy products . However , since the introduction of milk pasteurization in the 1950/60s in Europe and Northern America , patients with M . bovis infection in these regions suffered more often from reactivation of old infections , which mainly leads to pulmonary TB [3] , [11] , [30] , [32] . This supports the decreasing trend of association of zoonotic and extrapulmonary TB over time that we observed in our analysis ( Figure S2 ) . The occurrence of extrapulmonary zTB was very heterogeneous . This heterogeneity was found within and across different WHO regions and countries , likely due to different populations at risk in the different areas . However , there may be other factors underlying the different affected organs for zTB . The data extracted from our review does not allow a detailed investigation of these potential factors . While we did not perform a formal risk of bias analysis , data were extracted from the reports to explore possible factors influencing our findings ( Table S5 ) . For example , the study population never corresponded to a representative population of a region . When the sampling strategy and the case finding strategy were reported , it was mostly and always , respectively , done by convenience and passively , whose effectiveness might differ between geographical regions and individual studies . Also , differentiation methods of M . bovis from M . tuberculosis were not homogenous between the different studies . However , we did not find a significant difference between the proportion of extrapulmonary TB in various reports which used molecular and biochemical identification methods ( Wilkoxon rank sum test , n = 21 , W = 64 , p = 0 . 238 ) . Furthermore , our analyses were influenced by the technical constraints of the studies included . For example , the paucibacillary nature of extrapulmonary TB complicates culture and speciation of bacteria , in particular , in low-income areas with limited laboratory infrastructure and still common use of the classical Löwenstein-Jensen medium which is less sensitive than the newer BACTEC systems . However , since this may affect all ecotypes of the M . tuberculosis complex , it is unlikely to affect our conclusions regarding the frequency and distinct primary sites of extrapulmonary infection for M . bovis and M . tuberculosis . The proportion of extrapulmonary TB for zTB and MtTB may be overestimated by the categorization of patients with both pulmonary and extrapulmonary TB as extrapulmonary cases . This categorization would be incorrect in cases where extrapulmonary disease is an advanced clinical picture of a primary pulmonary infection . However , as this categorization was applied for zTB and MtTB , the conclusion for the comparison of the two agents may not be affected . Comparison of the occurrence of primary extrapulmonary sites between zTB and MtTB also revealed differences . This observation was made for studies conducted in Europe but not in the USA , which may be caused by different transmission routes in the two geographical regions . Further investigation would be needed to assess the transmission ways for different regions and populations at risk . For Europe and North America , available data cover large regions over a period of more than 50 years ( 1954–2007 ) providing a good overview of the situation in these regions . Low income countries , as Africa and South-East Asia , were highly underrepresented in the dataset . As milk pasteurization in most of these countries is not yet routinely done [6] , it is very likely that the consumption of dairy products is the main transmission route of zoonotic TB and therefore it is expected that the differences between the occurrence of extrapulmonary TB owing to zTB and MtTB are more pronounced than in Europe and America . However , the almost complete lack of information from these areas makes it almost impossible to draw conclusion which are also valid for Africa and Asia . Additionally , our findings might be subject to publication bias because small studies without spectacular findings are less likely to get published . Data on demographic parameters was rarely reported in the publications included in the review . Only one study , reporting on TB sites stratified by sex , had a sufficiently high sample size to draw any conclusion [11] . Also , data stratified by age category were limited with unevenly reported age classes and contradictory report results . Three studies reported higher proportions of extrapulmonary TB in young children [21] , [33] , [34] . Two of them were surveys which retrospectively analysed M . bovis infections of 10 and 11 hospitalized patients , respectively , over a period of 5 years from 1986–1990 in Spain and 6 years from 2000–2005 in France , respectively [33] , [34] . It was hypothesized in both reports that the higher proportion of pulmonary TB was a result from reactivated older infection caused by dairy products , which had not been routinely pasteurized when these patients were born . Three patients from France , all aged between 35 and 40 years including two with extrapulmonary TB , were born in Africa where the risk of infection by unpasteurized milk products , and therefore extrapulmonary disease , is higher [33] . The third report was on a foodborne outbreak associated with unpasteurized cheese in New York City , where all of the five children below five years of age suffered from extrapulmonary TB while 9 of the 30 patients aged from 5–76 years showed pulmonary TB [21] . One study , a survey from 1995 in France , reported higher proportions of extrapulmonary TB in adults [35] . However , only two patients , both with pulmonary TB , were younger than 15 years . Interestingly , both of them were born in Africa from where extrapulmonary TB would be expected to be more likely . Finally one larger study including patients from 1986–1990 in England and Wales reported similar proportions of pulmonary and extrapulmonary TB for patients above and below 30 years of age [36] . Only two reports were available which reported on deaths among extrapulmonary and pulmonary zTB and they revealed contradictory results [16] , [37] . To draw conclusions on the fatal outcome of extrapulmonary versus pulmonary zTB is therefore impossible . Only one report from Mexico included information on primary sites of extrapulmonary zTB stratified by the HIV co-infection of the patients [29] and it was therefore not possible to analyse the influence of HIV/AIDS on the TB sites . For a proper DALY calculation , demographic and mortality data on the respective disease are required . The limitations of the availability of these data in our dataset impede therefore DALY calculations for zTB . However , the significant differences in the occurrence of extrapulmonary TB and primary sites of the disease between zTB and MtTB detected in the present study demonstrate evidence for different disease sequelae . We may conclude that , regarding the differences in disease sequelae for the two causative agents , separate parameters for the DALY calculation should be used for zTB . Our findings , based on previously published global data , corroborated the widely stated , albeit rarely demonstrated association between TB caused by M . bovis or M . caprae and extrapulmonary disease . Disparities between zTB and MtTB were also suspected for specific sites of extrapulmonary TB . We do not know how well this conclusion fits the African and Asian situation , owing the almost complete lack of data . Therefore , investigations in these regions , where zoonotic TB is still relevant , are urgently required to assess the global universality of the conclusions . Nevertheless , there is evidence that the degree of disability weights for zTB and MtTB differ and therefore , to measure most accurately the socio-economic impact of TB , burden estimation should be conducted separately for zTB and MtTB . | Tuberculosis ( TB ) is one of the most devastating infectious diseases worldwide . The impact estimation of worldwide human TB is well established; however , that of TB transmitted by cattle , goat or sheep ( i . e . zoonotic TB ) is not . The affected body sites of human and zoonotic TB are repeatedly suggested to be different , which would influence the severity and impact of the diseases . The present study aimed to determine globally the association of affected body site and zoonotic TB by a systematic literature review . Data from 27 reports from Africa , America , Europe and the Western Pacific Region were included in the analyses . We found that the proportion of extrapulmonary TB was significantly higher in zoonotic TB than in human TB . Also , disparities of the specific body sites of extrapulmonary TB between zoonotic TB and human TB were detected . Our findings , which are based on global data , confirm the widely suggested association between zoonotic TB and extrapulmonary disease . Therefore , different measurements for estimating the impact of the two diseases should be considered . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"bacterial",
"diseases",
"infectious",
"diseases",
"zoonoses",
"bovine",
"tuberculosis",
"tuberculosis",
"bovine",
"tuberculosis",
"in",
"humans",
"epidemiology",
"neglected",
"tropical",
"diseases",
"miliary",
"tuberculosis",
"disease",
"informatics"
] | 2013 | Differences in Primary Sites of Infection between Zoonotic and Human Tuberculosis: Results from a Worldwide Systematic Review |
The degradation of cytosol-invading pathogens by autophagy , a process known as xenophagy , is an important mechanism of the innate immune system . Inside the host , Salmonella Typhimurium invades epithelial cells and resides within a specialized intracellular compartment , the Salmonella-containing vacuole . A fraction of these bacteria does not persist inside the vacuole and enters the host cytosol . Salmonella Typhimurium that invades the host cytosol becomes a target of the autophagy machinery for degradation . The xenophagy pathway has recently been discovered , and the exact molecular processes are not entirely characterized . Complete kinetic data for each molecular process is not available , so far . We developed a mathematical model of the xenophagy pathway to investigate this key defense mechanism . In this paper , we present a Petri net model of Salmonella xenophagy in epithelial cells . The model is based on functional information derived from literature data . It comprises the molecular mechanism of galectin-8-dependent and ubiquitin-dependent autophagy , including regulatory processes , like nutrient-dependent regulation of autophagy and TBK1-dependent activation of the autophagy receptor , OPTN . To model the activation of TBK1 , we proposed a new mechanism of TBK1 activation , suggesting a spatial and temporal regulation of this process . Using standard Petri net analysis techniques , we found basic functional modules , which describe different pathways of the autophagic capture of Salmonella and reflect the basic dynamics of the system . To verify the model , we performed in silico knockout experiments . We introduced a new concept of knockout analysis to systematically compute and visualize the results , using an in silico knockout matrix . The results of the in silico knockout analyses were consistent with published experimental results and provide a basis for future investigations of the Salmonella xenophagy pathway .
Inside the host , Salmonella enterica serovar Typhimurium , in the following referred to as Salmonella , can invade phagocytic and non-phagocytic cells , such as intestinal epithelial cells . After invasion of epithelial cells , the bacterium resides and starts replication inside the Salmonella-containing vacuole ( SCV ) . A fraction of the bacteria does not stay inside the SCV . The SCV membrane becomes damaged , and Salmonella gets access to the host cytosol and starts to replicate with high rates [1–3] . To protect the host cell , Salmonella rapidly becomes a target of xenophagy . Xenophagy , also known as antibacterial autophagy , is a process of capturing and eliminating cytosolic pathogens , like Salmonella . Xenophagy is an early defense mechanism against pathogens that enter the cytosol [4] . Cytosolic bacteria are targeted with ubiquitin or galectin-8 [4–6] , leading to the recruitment of the autophagy receptors , p62 [7] , nuclear dot protein 52 kDa ( NDP52 ) [8] , and optineurin ( OPTN ) [9] . Autophagy receptors simultaneously bind ubiquitin- or galectin-8-labeled Salmonella as well as the autophagic protein of the microtubule-associated protein 1 light chain 3 ( LC3 ) /γ-aminobutyric acid receptor-associated protein ( GABARAP ) family to recruit LC3/GABARAP-positive autophagosomal membranes [10] . Then , the bacteria are engulfed by an isolated membrane termed phagophore . The phagophore elongates and forms a double-membrane compartment , the autophagosome . The autophagosome can fuse with lysosomes , resulting in the destruction of the bacteria in the autolysosome . Defects in this defense pathway of xenophagy are linked to chronic inflammation , like Crohn‘s disease [11–16] . The xenophagy pathway has recently been discovered , and the exact molecular details are still not completely understood . Many bacteria have been shown to be targeted for the xenophagy pathway . Until now , Salmonella is the best-studied model organism for xenophagy . Neither quantitative nor qualitative mathematical models , describing the process of Salmonella xenophagy , have been presented so far . Existing literature provides a rich repertoire of molecular interaction information . To compile the known information and to analyze the biological system of xenophagy , we developed a mathematical model of Salmonella xenophagy in HeLa cells . We applied the Petri net ( PN ) formalism [17] to analyze the structure of the model and the basic dynamics of the system , even without explicit knowledge of the kinetic data . PN have been successfully employed to study signaling pathways and regulatory networks [18–23]; for an overview see [17 , 24] . The model recapitulates in detail the molecular mechanisms of Salmonella xenophagy , including the galectin-8 and ubiquitin-dependent targeting of cytosolic Salmonella and of Salmonella inside the damaged SCV . Additionally , the model comprises regulations like nutrient-dependent regulation of autophagy and TANK-binding kinase 1 ( TBK1 ) -dependent activation of the autophagy receptor , OPTN . The mechanism of TBK1 activation in HeLa cells upon Salmonella infection is not known , but it has been shown that a knockdown of TBK1 increases Salmonella replication in HeLa cells [8 , 9 , 25] . To model the activation of TBK1 , we proposed a new mechanism of TBK1 activation , suggesting an autoactivation by TBK1 oligomerization . We analyzed invariant properties to check the model for consistency and biologically meaningful behavior [17] . To observe the model behavior and the corresponding biological effect , we performed in silico knockouts of specific proteins , e . g . , NDP52 , OPTN , and TBK1 . We compared the model behavior with published knockout and knockdown experiments to verify the biological credibility of the model . Knockouts that has not been experimentally investigated yet provide new hypotheses for further studies . This can help to plan further experiments more precisely .
Due to the well-established analysis techniques , the expandability to a quantitative model , and the intuitive graphical representation of PN , the PN formalism is widely used for modeling and simulating biological systems . In the following , we give a short explanation of PN terms and methods that are necessary to understand the study . For a more comprehensive introduction to the formalism and analysis techniques of PN , we refer to Koch et al . 2011 and Reisig 2012 [17 , 27] . PN are directed , bipartite , labeled graphs , consisting of two types of nodes , places and transitions , connected by directed , weighted arcs . Places , modeling the passive part , represent biological substances , such as proteins or protein complexes . Transitions , representing the active part , stand for chemical reactions , such as binding reactions , dissociations , or degradations . A transition is called input transition if the transition has only outgoing and no incoming arcs . A transition is called output transition if the transition has only incoming and no outgoing arcs . Places and transitions are graphically represented as circles and rectangles , respectively . Arcs are labeled by integer numbers , which correspond to the stoichiometric factors of the biochemical reactions . Places can carry movable objects called tokens , which can move along the arcs by firing of transitions and reflect the system’s dynamics . The distribution of tokens over all places is called marking . The marking of the model defines the state of the system . The system’s state before any transition has fired is called initial state with an initial marking . A token can have different meanings . Tokens may represent a discrete amount of chemical substances , e . g . , one molecule or number of organisms as Salmonella . A token on a place may alternatively represent a fulfilled precondition of a reaction . Such a precondition could be , e . g . , the availability of a substance above a given threshold level . To simplify the graphical layout of the PN , we applied logical places for substances participating in many reactions . Logical places are identical graphical copies of a biological substance involved in various reactions of the model . To design , analyze , and simulate the PN , we used the open-source tool MonaLisa [28] . We structurally verified the model by the analysis of invariant properties called transition invariants ( T-invariants ) and place invariants ( P-invariants ) [17] . Mathematically , a T-invariant represents a set of transitions , whose firing reproduces an arbitrary state of the system . Biologically , a T-invariant describes a functional module at steady-state . Each T-invariant has to represent a biological meaningful subpathway of the system . We verified the biological interpretation of each T-invariant . If each transition is a member of at least one T-invariant , the PN is covered by T-invariants . Mathematically , a P-invariant represents a set of places , for which the weighted number of tokens remains constant . Each P-invariant corresponds to the conservation of a biological substance , e . g . , of an enzyme that is neither synthesized nor degraded in the network . We checked the biological interpretation of each P-invariant . For a more strict verification , we deleted all input and output transitions of the PN and checked the biological interpretation of the P-invariants of the resulting subnetwork . The subnetwork without input and output transitions has much more P-invariants than the original and should be covered by P-invariants , i . e . , each place is a member of at least one P-invariant . A biological PN should be connected , covered by T-invariants , and each of the invariants has to have a biological meaning [20 , 24] . Once we had constructed and verified a model , we performed in silico knockout analyses to examine the model behavior for perturbations of specific proteins . If a transition is knocked out , all T-invariants that contain the corresponding transition are affected . We performed single as well as double knockout experiments . We visualized the results of the knockout experiments using a sensitivity matrix . The rows and columns indicate the proteins of the model , whose synthesis , i . e . , input transitions , has been knocked out . The entries represent the influence of the knockout experiments on the system . This influence is measured by the percentage of the affected T-invariants—a measure for the model’s sensitivity to perturbations . The diagonal of the matrix shows the influence of the single knockouts on the system . To investigate the influence of a knockout on proteins or protein complexes , we introduce a new concept of in silico knockout analyses based on T-Invariants and visualize the results in the in silico knockout matrix . The rows of this matrix represent a certain knockout , e . g . , the deletion of an input transition . The columns represent the proteins or protein complexes , which may be affected by the knockout . The entries of the matrix are binary values encoded by either a red or a green circle . An entry is green , if the substance can be produced under steady-state conditions , i . e . , the substance is the outgoing place of a transition that is part of a T-invariant . An entry is red , if the substance can not be produced under steady-state conditions , i . e . , the substance is not the outgoing place of a transition that is part of a still functional T-invariant . Biologically , a green entry indicates that the knockout has no effect on , e . g . , the production of a substance or a protein complex , and a red entry represents a negative effect , e . g . , the substance or the protein complex can not be produced anymore . Fig 1 exemplifies the concept of in silico knockout matrices . The example PN in Fig 1A describes the synthesis of two proteins , A and B , the binding of these proteins to form a complex , AB , and the outflow of the complex to the environment . The T-invariant consists of the reactions SynA , SynB , Bin , and Out . In this example , the syntheses of A ( input transition SynA ) and B ( input transition SynB ) are knocked out . The resulting in silico knockout matrix ( see Fig 1B ) visualizes the influence of these knockouts on A , B , and AB . The rows of the matrix represent the input transitions , SynA and SynB , and the columns the substances , A , B , and AB . The knockout of either the syntheses of A ( input transition SynA ) or B ( input transition SynB ) is sufficient to have a negative effect on all substances in the model , A , B , and AB . Consequently , all entries in the matrix are red . However , we would expect that the synthesis of protein B has no effect on protein A and vice versa . We correct this effect by the addition of output transitions for A and B ( see Fig 1C ) . Biologically , these output transitions , DegA and DegB , represent the degradation of the proteins . The additional output transitions generate two new T-invariants , which separately include the synthesis and the degradation of A and B , respectively . The first T-invariant consists of the transitions SynA and DegA , and the second T-invariant of SynB and DegB . Now , B is not affected by a knockout of A and vice versa , see Fig 1D . The concept of sensitivity and in silico knockout matrices is directly related to Mauritius maps , which compute and visualize the impact of knockouts as a binary tree structure [19] .
We proposed a model of Salmonella xenophagy illustrated in Fig 2 . To get a comprehensive overview of the current state of research , we included all known processes that target Salmonella to xenophagy in epithelial cells . Inside the lumen of the mammalian small intestine , Salmonella infiltrates epithelial cells , see Fig 2 upper part . Once inside the cell , most of the bacteria stay in the SCV . Inside the SCV , Salmonella has non-ideal life conditions , such as an acidic environment [29] with limited space and likely nutrient-poor [30] . A small fraction of Salmonella disrupts their vacuolar membrane of the SCV [4 , 31] , and enters the cytosol . Inside the cytosol , Salmonella has ideal life conditions , such as access to many nutrients , sufficient space , and a neutral pH . Under these conditions , Salmonella replicates inside the cytosol at a higher rate than inside the vacuole [1–3] . The PN describes the biological model shown in Fig 2 . The model includes all known processes of Salmonella xenophagy in epithelial cells , which were biochemically proven and published , see S2 Table . We included a hypothetical model of TBK1 activation , suggesting a spatial and temporal regulation of this process , see detailed view in Fig 2 . The resulting PN model of Salmonella xenophagy is available in S1 File . It comprises 61 places , including nine logical places , and 69 transitions connected by 184 arcs . A graphical representation of the PN model is depicted in Fig 3 . All places and transitions , including their biological description , are listed in S1 and S2 Tables . Places can be proteins ( e . g . , p62 , NDP52 , OPTN , TBK1 ) , macromolecular complexes ( e . g . , S:Gal8 , S:Ub:NDP52:OPTN:p62 ) , organisms ( S-cyt ) , or even signals ( e . g . , SignalSCVdamage , NDP52’ ) . The place E3 ligase represents an unknown E3 ubiquitin ligase . Ubiquitin and phosphate are not directly modeled as a place in the PN . Transitions can be binding processes ( e . g . , T1 , T2 , T3 ) , escape from the SCV ( T17 ) , degradations ( e . g . , Deg1 , Deg2 , Deg2i ) , damage of the SCV ( SCVdamage ) , or syntheses ( Syn1-9 ) . The PN has ten input and seven output transitions . Input transitions stand for the syntheses of proteins or protein complexes ( Syn1-9 ) or the invasion of Salmonella to the cytosol ( SCVdamage ) . Output transitions are degradations of captured Salmonella inside the autophagosome ( e . g . , Deg1 , Deg2 , Deg2i ) or degradation of protein complexes ( Output ) . The temporal order of binding events in Salmonella xenophagy is still unknown . For instance , it is not known whether Nap1/Sintbad binds to NDP52 followed by a binding of the NDP52-Nap1/Sintbad complex to ubiquitinated Salmonella or whether NDP52 binds to ubiquitinated Salmonella followed by a binding of Nap1/Sintbad to NDP52-positive Salmonella . Likely , both temporal orders of binding events are possible . We decided to model one of these temporal orders as a typical system’s behavior . To model all possible orders is beyond the scope of this work and would unnecessarily increase the complexity of the model . The problem of combinatorial complexity emerges because the proteins can be modified and combined in various ways to form multicomponent complexes , reviewed by Hlavacek et al . [55] . To observe the effect of specific proteins on the network behavior , we performed in silico knockout experiments and compared the results with published experimental data . The aim was to check the biological credibility of the model . The model describes 16 different ways of capturing Salmonella for the xenophagy pathway . If we knockout each of the autophagy components , unknown E3 ubiquitin ligase , LRSAM1 , galectin-8 , p62 , NDP52 , OPTN , Nap1/Sintbad , or TBK1 , all of these 16 ways or even a few could be affected . S5 and S6 Tables list the knocked out proteins and their effect on Salmonella xenophagy measured in the percentage of affected T-invariants . In Fig 4 , the results are visualized in the sensitivity matrix . The knockout analysis showed that only the knockout of NDP52 affects 100% of the T-invariants , resulting in a complete loss of function of both galectin-8 and ubiquitin-dependent xenophagy . None of the 16 T-invariants remained functional , which demonstrates the central role of NDP52 in Salmonella xenophagy . The knockout of the E3 ubiquitin ligase LRSAM1 , other E3 ubiquitin ligases , and the autophagy receptors , p62 and OPTN , affected 94% of the T-invariants . Only the galectin-8-dependent xenophagy pathway remained functional . A double knockout of p62 and NDP52 showed a complete loss of function of the galectin-8 and ubiquitin-dependent xenophagy . This result is consistent with the double knockdown experiment of NDP52 and p62 in HeLa cells [38] . The knockdown of both autophagy receptors caused a reduction of LC3/GABARAP association to Salmonella to a similar extent than the knockdown of one receptor . A knockout of the kinase TBK1 would affect only 38% of the T-invariants . In the model , TBK1 is not absolutely required in xenophagy , but may have an enhancing effect on Salmonella xenophagy . Nevertheless , TBK1 siRNA-treated HeLa cells show similar Salmonella replication than NDP52 siRNA-treated cells [8 , 9] . This may be due to a higher amount of fast-replicating , cytosolic bacteria caused by a reduced vacuolar integrity in TBK1-depleted cells [25] . To avoid effects as described in Fig 1 , we included additional output transitions to generate new T-invariants for the construction of the in silico knockout matrix . The modified PN model is available in S2 File . Fig 5 shows the resulting in silico knockout matrix . The rows of the matrix represent protein knockouts and the columns the effects of the perturbation on macromolecular Salmonella complexes . The matrix has 200 entries , 41 of those had been experimentally investigated in eight studies [6–9 , 25 , 33 , 34 , 38] . 12 entries are biologically obvious and need no further experimental investigation , e . g . , the knockout of galectin-8 has a negative effect on the binding of galectin-8 to Salmonella inside the damaged SCV . The other 147 entries can be interpreted as hypotheses for future investigations . In the following , we describe the effect of some knocked out proteins ( see Fig 5 ) . We constructed a mathematical model , which provides a compilation of currently available experimental findings of Salmonella xenophagy in a verified and consistent network topology of the elementary molecular processes . The PN model combines the molecular processes of ubiquitination , binding of the autophagy receptors , regulatory processes , like nutrient-dependent regulation of xenophagy , TBK1-dependent activation of the autophagy receptor , OPTN , and galectin-8-dependent xenophagy . Modeling the unknown process of TBK1 activation in HeLa cells , we proposed a new hypothetical model of TBK1 activation . We assume a spatial and temporal regulation of this process . We checked the model structure for consistencies and correctness . We found 16 basic functional modules , which describe different pathways of the autophagic capturing of Salmonella and reflect the basic dynamics of the system . We introduced a new concept of in silico knockout analyses based on T-Invariants and visualized the results as an in silico knockout matrix . The model correctly reproduces the experimental results published so far , ensuring the biological credibility of the model . Based on the in silico knockout analyses , we postulate hypotheses for future investigations . Due to the fact that xenophagy was discovered recently and the exact molecular details are still unknown , the proposed model is probably not complete and will be extended by new insights . If quantitative data become available , the PN model can be easily extended to model the quantitative system’s behavior . | Salmonellae are Gram-negative bacteria , which cause the majority of foodborne diseases worldwide . Serovars of Salmonella cause a broad range of diseases , ranging from diarrhea to typhoid fever in a variety of hosts . In the year 2010 , Salmonella Typhi caused 7 . 6 million foodborne diseases and 52 000 deaths , and Salmonella enterica was responsible for 78 . 7 million diseases and 59 000 deaths . After invasion of Salmonella into host epithelial cells , a small fraction of Salmonella escapes from a specialized intracellular compartment and replicates inside the host cytosol . Xenophagy is a host defense mechanism to protect the host cell from cytosolic pathogens . Understanding how Salmonella is recognized and targeted for xenophagy is an important subject of current research . To the best of our knowledge , no mathematical model has been presented so far , describing the process of Salmonella Typhimurium xenophagy . Here , we present a manually curated and mathematically verified theoretical model of Salmonella Typhimurium xenophagy in epithelial cells , which is consistent with the current state of knowledge . Our model reproduces literature data and postulates new hypotheses for future investigations . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"Discussion"
] | [
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"autophagic",
"cell",
"death",
"pathology",
"and",
"laboratory",
"medicine",
"hela",
"cells",
"enzymes",
"pathogens",
"cell",
"processes",
"biological",
"cultures",
"microbiology",
"enzymology",
"ubiquitin",
"l... | 2016 | In Silico Knockout Studies of Xenophagic Capturing of Salmonella |
Leptospirosis , caused by pathogenic Leptospira , is a worldwide zoonotic infection . The genus Leptospira includes at least 21 species clustered into three groups—pathogens , non-pathogens , and intermediates—based on 16S rRNA phylogeny . Research on Leptospira is difficult due to slow growth and poor transformability of the pathogens . Recent identification of extrachromosomal elements besides the two chromosomes in L . interrogans has provided new insight into genome complexity of the genus Leptospira . The large size , low copy number , and high similarity of the sequence of these extrachromosomal elements with the chromosomes present challenges in isolating and detecting them without careful genome assembly . In this study , two extrachromosomal elements were identified in L . borgpetersenii serovar Ballum strain 56604 through whole genome assembly combined with S1 nuclease digestion following pulsed-field gel electrophoresis ( S1-PFGE ) analysis . Further , extrachromosomal elements in additional 15 Chinese epidemic strains of Leptospira , comprising L . borgpetersenii , L . weilii , and L . interrogans , were successfully separated and identified , independent of genome sequence data . Southern blot hybridization with extrachromosomal element-specific probes , designated as lcp1 , lcp2 and lcp3-rep , further confirmed their occurrences as extrachromosomal elements . In total , 24 plasmids were detected in 13 out of 15 tested strains , among which 11 can hybridize with the lcp1-rep probe and 11 with the lcp2-rep probe , whereas two can hybridize with the lcp3-rep probe . None of them are likely to be species-specific . Blastp search of the lcp1 , lcp2 , and lcp3-rep genes with a nonredundant protein database of Leptospira species genomes showed that their homologous sequences are widely distributed among clades of pathogens but not non-pathogens or intermediates . These results suggest that the plasmids are widely distributed in Leptospira species , and further elucidation of their biological significance might contribute to our understanding of biology and infectivity of pathogenic spirochetes .
Leptospires are thin , spiral , highly motile bacteria that belong to the order Spirochaetales , an early branch of eubacteria . The genus Leptospira includes at least 21 species based on 16S rRNA phylogeny , further distinguished into three clades: pathogens , non-pathogens and intermediates[1] . Pathogenic Leptospira are comprised of at least 14 species , which share a common branch in evolution , genetically distinct from non-pathogens . Leptospires are also serologically classified into serovars , including more than two hundred that are pathogenic in human and animals[1] . Pathogenic Leptospira is known to cause the widespread water-related zoonosis , called leptospirosis . Hosts usually become infected through direct contact with soil or water contaminated by the urine of infected animals [2] . Infection produces a wide spectrum of clinical manifestations , ranging in severity from a mild influenza-like disease to an acute , potentially lethal infection . Pathogenic Leptospira species , such as L . interrogans , L . borgpetersenii , L . kirschneri , L . noguchii , and L . weillii [3–5] , are the causative pathogens of leptospirosis , among which L . interrogans and L . borgpetersenii are most prevalent globally . Although infection with L . interrogans and L . borgpetersenii cause similar clinical symptoms , the transmission modes are different: for example , L . interrogans is commonly waterborne , whereas L . borgpetersenii is transmitted via direct host-to-host contact [6] . Currently , the completed genome sequences include five pathogenic strains of L . interrogans , two pathogenic strains of L . borgpetersennii , and two strains of saprophytic L . biflexa [6–12] . Comparison of the L . interrogans and L . borgpetersenii genome sequences revealed genome reduction in the L . borgpetersenii genome that is supposedly IS-mediated [6] . Plasmids are regarded as one of the most effective vehicles for bacterial communication of genetic information [13] , promoting the rapid evolution and adaptation abilities of bacteria [14] . Because of the diverse genetic information they carry , plasmids often play specific biological roles in the host bacterium . Also , they can potentially be engineered as efficient genetic tools for microbial genetic manipulation and analysis through the introduction , modification , or removal of target genes [15] . In fact , many plasmids were found in spirochaetes , with a majority identified in the genus Borrelia [13] . In addition to its linear chromosome , Borrelia contains multiple circular and linear plasmids within a single cell [16] . As for the genus Leptospira , plasmid P74 and bacteriophage LE1 were first reported within the saprophytic L . biflexa serovar Patoc strain Patoc I genome , and then a LE1-like prophage was found in the intermediate species L . licerasiae [9 , 17] . It was long believed that pathogenic Leptospira species contain only two chromosomes [5] . Recently , a 54-kb genomic island , LaiGI , was confirmed as an extrachromosomal replicon stable within the first sequenced pathogenic strain of L . interrogans strain Lai [18 , 19] . Two plasmids , pGui1 and pGui2 , in L . interrogans pathogenic strain Gui44 and three plasmids , lcp1 , lcp2 , and lcp3 , in L . interrogans pathogenic strain 56609 were reported in succession , which significantly contribute to revealing the diversity of the pathogenic Leptospira genome[10 , 12] . Based on the plasmids identified , L . interrogans–Escherichia coli shuttle vectors with the predicted replication rep gene or rep combined with parAB loci from the three plasmids of L . interrogans pathogenic strain 56609 were reported to be successfully transformed into both saprophytic and pathogenic Leptospira species , which is considered as a new milestone in research efforts involving pathogenic Leptospira [12 , 16] . Although the plasmids have originally been identified in L . interrogans , the plasmid sequence in other infectious strains , including L . borgpetersenii is still unknown . In this study , two extrachromosomal circular elements of L . borgpetersenii serovar Ballum strain 56604 were detected and estimated by S1 nuclease digestion following pulsed-field gel electrophoresis ( S1-PFGE ) , which allowed detection of low-copy replicons [20] . These two plasmids were further characterized in details through whole genome sequencing . We subsequently used S1-PFGE to identify the plasmids in the remaining 14 reference strains of Leptospira in China belonging to species L . borgpetersenii , L . interrogans , and L . weilii . These efforts will contribute to a better understanding the genetic complexity of this bacterium , delivering the most effective genomic information , and accelerating the process of complete genomic sequencing of the genus Leptospira .
Leptospira borgpetersenii serogroup Ballum serovar Ballum strain 56604 was obtained from the Institute for Infectious Disease Control and Prevention ( CDC ) , Beijing , China . The remaining 14 domestic reference strains of Leptospira were also obtained from the Chinese CDC and used in this study: Leptospira interrogans serogroup Icterohaemorrhagiae serovar Lai strain 56601 , serogroup Canicola serovar Canicola strain 56603 , serogroup Pyrogenes serovar Pyrogenes strain 56605 , serogroup Autumnalis serovar Autumnalis strain 56606a and 56606v , serogroup Australis serovar Australis strain 56607 , serogroup Pomona serovar Pomona strain 56608 , serogroup Grippotyphosa serovar Linhai strain 56609 , serogroup Hebdomadis serovar Hebdomadis strain 56610 , serogroup Bataviae serovar Paidjan strain 56612 , serogroup Sejroe serovar Wolffi strain 56635; Leptospira borgpetersenii serogroup Javanica serovar Javanica strain 56602 , serogroup Tarassovi serovar Tarassovi strain 56613 , serogroup Mini serovar Mini strain 56655; Leptospira weilii serogroup Manhao serovar Qingshui strain 56615 . Of note , 56606a is avirulent strain derived from strain 56606v and has lost its virulence after long time in vitro passages in our laboratory . The strains were grown in liquid Ellinghausen–McCullough–Johnson–Harris ( EMJH ) medium under aerobic conditions at 28°C to mid-log phase and then collected at an optical density of 1 . 3–2 . 0 at 600 nm . The genome of L . borgpetersenii serovar Ballum strain 56604 was sequenced using a 454 GS20 system ( 454 Life Sciences ) . Six micrograms of 56604 genomic DNA was prepared to create sequencing libraries according to the manufacturer’s protocol ( 454 Life Sciences ) . A total of 421 , 952 reads ( average read length 282bp ) were generated and 421 , 410 reads of high quality ( 99 . 8% ) were selected for genome assembly , providing 29 . 5 fold coverage . 179 contigs ( 144 contigs >500bp ) were yielded and the N50 size of the contigs was 28 , 039bp . In the finishing process , the reference genome sequences of L . borgpetersenii serovar Hardjo strain L550 and JB197 were used to determine the suppositional contig order of 56604 . In the following physical gap closing , PCR was performed thousands of times based on the following conditions: 5 min at 95°C , followed by 35 cycles of 30 sec at 95°C , 30 sec at 58°C , and 30 sec at 72°C . After that , sequence assembly was accomplished using Phred , Phrap , and Consed programs[21–23] . The sequence alignments were performed using BLAST ( http://blast . st-va . ncbi . nlm . nih . gov/Blast . cgi ) . The open reading frames ( ORFs ) were predicted and manually checked by the combined use of GLIMMER , GeneMark , and Z-curve programs[24 , 25] . Clusters of orthologous groups ( COG ) functional annotation for each gene was performed through RPS-BLAST in the NCBI Conserved Domain Database ( CDD ) , and conserved domains were analyzed to further verify and supplement the annotation by searching the Pfam database[26 , 27] . Transfer RNA genes were identified with tRNAscan-SE ( http://selab . janelia . org/tRNAscan-SE/ ) . Insertion sequence ( IS ) elements were determined using the IS-finder online tool ( https://www-is . biotoul . fr/ ) . Orthologous proteins were identified by performing BLAST searches against the NCBI non-redundant protein database of L . borgpetersenii serovar Hardjo strain L550 and strain JB197 genomes and L . borgpetersenii serogroup Ballum serovar Ballum strain 56604 and subsequently checked manually . Whole-genome sequence comparison was performed at the nucleotide level using the program BLASTn ( with a cutoff E value of 1e-10 ) and visualized with EasyFig[28] . Bacteria were grown in EMJH medium under aerobic conditions at 28°C to mid-log phase . Each culture was suspended in 100 μl cell suspension buffer ( 50mM Tris-HCl pH7 . 2 , 25M NaCl , 50mM EDTA ) to a turbidity of 1 . 3–2 . 0 OD600 , as specified in S1 Table . An equal volume of plasmid-harboring bacteria was embedded in agarose plugs ( 1% ) ( SeaKemGold Agarose gels ) , immediately dispensed into wells of plug molds , then incubated at 55°C with proteinase K solution ( 2% ( w/v ) Na-deoxycholate , 10% ( w/v ) Na-lauroyl sarcosine , 0 . 5M EDTA pH 8 . 0 , 0 . 5 M NaCl , 0 . 5% proteinase K ) overnight . The plugs were washed twice in tubes shaking in a 54°C water bath for 15 min each time with both pre-warmed double distilled water and TE buffer to inactivate the proteinase K , then used immediately or stored in TE buffer at 4°C . S1 nuclease , a specific enzyme that can convert supercoiled plasmids into full-length linear molecules , was used to digest the 2-mm gel plugs . For different Leptospira species , the digestion reaction was similar in time but different in the enzyme added ( S1 nuclease , Thermo Scientific; NotI and PstI , New England Biolabs ) , as shown in S1 Table . The standard strain Salmonella enterica serotype Braenderup ( H9812 ) , kindly provided by the Department of Clinical Microbiology , Ruijin Hospital ( Shanghai , China ) , was digested with XbaI and used as a molecular weight marker . Gel plugs were subjected to PFGE immediately after completion of the digestion reaction using a contour-clamped homogeneous electric field machine ( CHEF-DR III; Bio-Rad ) . Electrophoresis with linear ramp time from 5 to 65 s at a gradient of 6 V/cm and an included angle of 120° was performed for 20 h to separate the DNA fragments , and gels were cooled continuously at 14°C during the running process . Gels were then stained in 1μg/mL ethidium bromide for 40 min and visualized in a gel image acquisition and analysis system . The plasmid DNA was transferred and cross-linked to positively charged nylon membranes ( Roche Diagnostics ) and hybridized against digoxigenin-labeled probes generated through polymerase chain reaction amplification ( PCR DIG Probe Synthesis Kit , Roche ) according to the manufacturer’s instructions with some modifications . Briefly , the membrane was washed in 2X SSC twice for 5 min each time at 25°C , then washed twice in 0 . 1X SSC for 15 min at 68°C . After blocking in 1% blocking reagent for 30 min , the membrane with DIG-labeled probe was detected with anti-Digoxigenin-AP , Fab fragments ( Roche ) , and CDP-Star ( Roche ) . All primers used are listed in S2 Table . The complete genomic sequences of L . borgpetersenii strain 56604 have been deposited in GenBank under the following accession numbers CP012029 , CP012030 , CP012031 and CP012032 .
The genome of strain L . borgpetersenii strain 56604 consists of two circular chromosomes and two circular extrachromosomal replicons . Chromosome CI and CII are 3 , 550 , 837 bp and 361 , 762 bp , respectively , with GC content of 40 . 2% . Table 1 summarizes the genome features of strains 56604 , L550 and JB197 . The genome of strain 56604 is smaller than L . interrogans , similar to that of L . borgpetersenii strains L550 and JB197[6] . The genome sequence of strain 56604 , L550 and JB197 shared extensive collinear disrupted by few rearrangements , but not L . interrogans strain Lai ( Fig 1A ) . The numbers of conserved genes amongst the three strains are 2 , 536 as shown in Fig 1B . Two additional circular extrachromosomal replicons , designated lbp1 and lbp2 , which are 65 , 435 bp with GC content of 41% and 59 , 545 bp with GC content 39 . 7% , respectively ( detailed in Table 1 and S1 Fig ) were present in strain 56604 . Several insertion sequence ( IS ) elements have been described in pathogenic Leptospira , including IS1500 , IS1501 , IS1533 , and ISLin1[5 , 29] . We identified a total of approximately 54 ISs scattered in chromosomes of strain 56604 , including 31 copies of IS1533 , 15 copies of ISLin1 , 4 copies of IS1502 , 2 copies of IS1500 , 1 copy of IS1501 , and 1 copy of ISLin2 ( S3 Table ) . The number of IS copies is apparently less than those in L . borgpetersenii strain L550 and JB197 [6 , 30 , 31] . Two circular extrachromosomal replicons , designated lbp1 and lbp2 , were identified in strain 56604 through whole genome sequencing . According to sequence-reads coverage of each assembly level , the two plasmids were estimated to share equal copy numbers with the chromosomes . The GC content is higher than that of plasmids lcp1 , lcp2 , and lcp3 ( 35% , 35% , 39% ) in L . interrogans strain 56609 as well as pGui1 ( 34 . 63% ) and pGui2 ( 33 . 33% ) in L . interrogans strain Gui44 ( Table 2 ) [10 , 12] . Sequence comparisons showed low similarity amongst them ( less than 20% coverage ) . BLASTp analysis showed that plasmid lbp1 has 1 , 3 , 0 , 10 and 1 orthologs in plasmid lcp1 , lcp2 , lcp3 , pGui1and pGui2 respectively; that plasmid lbp2 has 1 , 1 , 0 , 1 and 3 orthologs in plasmid lcp1 , lcp2 , lcp3 , pGui1 and pGui2 respectively . 99% identity with 100% coverage with lcp2-rep . parAB genes were also found to be located immediately upstream of rep genes in lbp1 and lbp2 respectively . The parAB genes are less conserved and they showed low similarities between each other . S1-PFGE separation followed by Southern blotting with lcp1-rep , lcp2-rep , and lcp3-rep specific probes confirmed that lbp1 could hybridize with lcp1-rep sequence and lbp2 with lcp-2-rep sequence ( Fig 2 ) . Genomic DNA digested with selected restriction enzyme followed by in situ PFGE-based Southern blot analysis also confirmed that lbp1 and lbp2 were not integrated into chromosomes . lbp1 , cutting with a single restriction enzyme NotI in lbp1was confirmed by Southern blotting using lcp1-rep . lbp2 , cutting with a single restriction enzyme PstI in lbp2 was confirmed by Southern blotting using lcp2-rep . The obtained DNA bands were same as above ( Fig 2 ) . These data showed that lbp1 contained same rep sequence with lcp1 and lbp2 contained same rep sequence with lcp2 , both of which are extrachromosomal . Further bioinformatics analysis showed that the rep of lbp1 shared 87% identity with 100% coverage with lcp1-rep , and the rep of lbp2 shared The large size and low copy of plasmids within Leptospira cells presents a challenge for their isolation[12] . Moreover , the high proportion of homologous sequence also makes it difficult to differentiate them from the chromosome and affects the assembly process , even during whole genome sequencing[12] . S1 nuclease treatment can convert the supercoiled plasmids into full-length linear molecules . When the bacteria harboring plasmid embedded in agarose are digested with S1 nuclease followed by pulsed-field gel electrophoresis ( PFGE ) , plasmids can be detected and their sizes can be estimated with appropriate linear DNA markers [20] . In this study , 15 Chinese epidemic Leptospira strains were subjected to S1-PFGE , a distinct approach to detect the presence and the size of plasmids within Leptospira cells . According to the results of S1-PFGE , plasmids can be directly detected in 13 out of 15 tested Leptospira strains , with the exception of L . borgpetersenii strain 56602 and L . interrogans strain 56608 ( Fig 3 and Table 3 ) . The sizes ranged from 50 kb to 150 kb ( Fig 3 ) , and one Leptospira cell can contain up to three plasmids . It showed that the plasmids have been detected in all tested species , including L . borgpetersenii , L . weilii , and L . interrogans . In our previous study , the whole genome of L . interrogans serovar Linhai strain 56609 was sequenced , and three extrachromosomal replicons , designated lcp1 , lcp2 , and lcp3 , were in the cell[12] . The homology of rep genes in the plasmids detected by S1-PFGE was tested by Southern blot hybridization using lcp1 , lcp2 , and lcp3-rep specific probes . As shown in Fig 4 and Table 3 , each of the plasmids detected by S1-PFGE in 13 Chinese epidemic strains can hybridize with one of three known rep specific probes . Eleven plasmids distributed in 10 strains can hybridize with lcp1-specific probes ( Fig 4A ) . Of note , two plasmids in strain 56612 can hybridize with lcp1-specific probes . Eleven plasmids distributed in 11 strains can hybridize with lcp2-specific probes ( Fig 4B ) . Two plasmids distributed in two strains can hybridize with lcp3-specific probes ( Fig 4C ) . Plasmid partitioning system encoded two partitioning proteins ParAB protein and a replication protein rep protein . Plasmids with the same replication control are “incompatible” , whereas the plasmids with different replication control are “compatible”[32] . rep has been frequently used to classify plasmids[33] . It seems that the plasmids in Chinese epidemic strains can be divided into three types: lcp1 , lcp2 , and lcp3 . When compared with the previous results to test for presence of the rep gene in the same 15 Chinese epidemic strains , more plasmids were detected in this study[12] . The only difference is that in the previous study , the embedded bacterial cells were digested with the selected restriction enzyme , whereas in this study , the cells were digested with S1 nuclease . In the previous study , some plasmids without cleavage sites for the selected restriction enzyme may have gone undetected because of being supercoiled . S1-PFGE is a general method independent of knowing restriction enzyme cleavage sites and thus can detect all the plasmids in a cell[20] . The genus Leptospira is comprised of 21 species clustered into following groups , pathogen ( Group I ) , intermediate pathogen ( Group II ) , and non-pathogen based on 16S rRNA phylogeny[1] . The infectious group includes at least 14 species , nine species of pathogens , and five species of intermediate pathogens . The genomes of 319 Leptospira strains belonging to 20 species have been sequenced and deposited in the GenBank database . Blastp search of the lcp1 , lcp2 , and lcp3-rep genes against the sequenced Leptospira genomes showed that 6 species ( 299 strains sequenced ) including 77 strains harbored homologous sequence with lcp1-rep , 6 species ( 299 strains sequenced ) including 80 strains harbored homologous sequence with lcp2-rep , and 7 species ( 301 strains sequenced ) including 71 strains harbored homologous sequence with lcp3-rep ( Table 4 ) . All of the 7 species belonged to pathogens ( Group I ) . Of the 319 sequenced strains , 304 strains belonged to pathogens ( Group I ) ; only 8 strains belonged to intermediate pathogens , and 7 strains belonged to the non-pathogens group . Obviously , more genome data of intermediate pathogens and non-pathogens are needed to understand the relationship between extrachromosomal replicons and their host range .
Heterogeneity within the genus Leptospira is well confirmed based on recent studies adopting DNA-DNA hybridization , comparative genomics , and whole genome sequencing efforts [2 , 5 , 12] . There are currently 21 species within the genus Leptospira , 14 of which are supposedly pathogenic . However , based on prevalence and pathogenicity , L . interrogans and L . borgpetersenii are the two largest species , containing about half of the known 230 pathogenic serovars , although the latter strain is less well characterized [6] . L . borgpetersenii genomes are 700 kb smaller than that of L . interrogans [6 , 30] and its genome reduction may be IS-mediated , because there are more copy numbers of IS elements in sequenced L . borgpetersenii serovar Hardjo compared with in L . interrogans serovars Lai and Copenhageni [6] . IS sequence , as a mobile genetic element , varies between serovars . Several IS elements , including IS1500 , IS1501 , IS1502 , IS1533 , and ISLin1 , were identified in Leptospira [2 , 5 , 6 , 29 , 30] . Although the genome size of L . borgpetersenii serovar Ballum strain 56604 showed a similar trend of genome reduction as L . borgpetersenii serovar Hardjo , the type and copy number of IS elements in serovar Ballum were apparently less than those in serovar Hardjo , with the exception of two single copy extrachromosomal elements . Plasmids are important genetic vehicles playing an important role in the processes of genomic organization , adaptation , evolution , and virulence for bacterial pathogens [13] . Also , plasmids are very important genetic tools to manipulate and analyze microorganisms through introduction , modification , or removal of target genes [13] . However , in some cases due to their occurrence in low copy numbers and containment of repeat sequences ( homologous to chromosomes ) , the method to detect plasmids in Leptospira is primarily dependent on whole genome sequencing and assembly[10 , 12] . Even using whole genome sequencing , plasmid could be miss annotated and assembled into chromosomes [19] . Based on the plasmids identified , shuttle vectors constructed with the rep genes can be successfully transformed into some pathogenic Leptospira[12 , 16] . Although genetic incompatibility is supposed to hinder successful transformation , the precise mechanism is still unknown[12] . Identification of more replicons in different strains will help to understand the mechanism and facilitate the manipulation of genetic tools . In this study , S1-PFGE was applied as an effective method to detect the plasmids within Leptospira cells , which yielded accurate bands and sizes of plasmids in all tested Leptospira strains , although despite our best efforts bands were smeared to some degree . We speculate that despite S1-PFGE method is sensitive , the low copy number of plasmids in Leptospira probably render them more difficult for detection . We tested the plasmid distribution among 15 Chinese epidemic Leptospira strains , including L . borgpetersenii , L . weilii , and L . interrogans using S1-PFGE , which enable us to detect potentially all plasmids in sequenced strains 56601 , 56603 ( Gui44 ) , and 56609 . The band numbers ( corresponding to one , two , and three bands , respectively ) and size values were consistent with previously identified plasmids . Furthermore , Southern blotting using known lcp1 , lcp2 , and lcp3 specific probes further confirmed them as extrachromosomal elements . Interesting , it seems that plasmids detected in Chinese epidemic strains can be divided into three types: lcp1 , lcp2 , and lcp3 . They can exist together or separately within Leptospira cells . Moreover , same type plasmids ( lcp1 in strain 56612 ) can also coexist in one cell . Further analysis of the plasmid-encoded genes did not reveal any unique gene cluster but confirmed the distribution and diversity of plasmids in the genus Leptospira . Blastp search of lcp1 , lcp2 , and lcp3-rep homologous sequence with 319 sequenced Leptospira genomes further showed they are widely distributed in pathogens ( Group I ) of Leptospira . Based on the information provided in this study , we speculate that additional assessment of plasmid contents in additional globally prevalent Leptospira strains , should shed new light on significance of these intriguing extrachromosomal DNA elements in leptospiral biology and evolution . | Leptospirosis , caused by a diversity of pathogenic species within the genus Leptospira , is a worldwide public health problem affecting both developed and developing countries . In 2003 , the whole genome sequencing of pathogenic Leptospira interrogans strain 56601 opened the genomic research of this specific pathogen and accelerated genomic sequencing speed afterwards . The availability of whole genome sequences of the pathogenic species L . interrogans and L . borgpetersenii , saprophyte L . biflexa and the intermediate L . licerasiae , which represented three phylogenetic groups of Leptospira spp . , has undoubtedly facilitated the understanding of the genetic complexity of this organism . Genome drafts of more than 300 strains have been released recently . The analysis of these sequences will provide invaluable information for understanding the evolution and adaption to various environment of the Leptospiraceae . Extrachromosomal replicons are important for communication of genetic information between bacteria . Despite first leptospiral genome was sequenced in 2003 , these small auto-replicons , however , were not identified in the whole genome sequence of pathogenic L . interrogans until 2014 . How to effectively identify the presence of small auto-replicons in Leptospira cell is a technical challenge which hinders researchers in fully understanding the complexity and diversity of Leptospira genomes . In this study , we report identification of small auto-replicons in 15 Leptospira strains endemic in China by S1-PFGE analysis , independent of genome sequence assembly . Further analysis suggested that these plasmids are widely presented in Leptospira species . The study , combined with the recently available genomic sequencing data , will help us to elucidate the diversity of Leptospira genomes and a deeper insight into evolutionary perspective of this unique clade of organism . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Distribution of Plasmids in Distinct Leptospira Pathogenic Species |
The cystic fibrosis transmembrane conductance regulator ( CFTR ) is an epithelial chloride channel mutated in patients with cystic fibrosis ( CF ) . The most prevalent CFTR mutation , ΔF508 , blocks folding in the endoplasmic reticulum . Recent work has shown that some ΔF508-CFTR channel activity can be recovered by pharmaceutical modulators ( “potentiators” and “correctors” ) , but ΔF508-CFTR can still be rapidly degraded via a lysosomal pathway involving the CFTR-associated ligand ( CAL ) , which binds CFTR via a PDZ interaction domain . We present a study that goes from theory , to new structure-based computational design algorithms , to computational predictions , to biochemical testing and ultimately to epithelial-cell validation of novel , effective CAL PDZ inhibitors ( called “stabilizers” ) that rescue ΔF508-CFTR activity . To design the “stabilizers” , we extended our structural ensemble-based computational protein redesign algorithm to encompass protein-protein and protein-peptide interactions . The computational predictions achieved high accuracy: all of the top-predicted peptide inhibitors bound well to CAL . Furthermore , when compared to state-of-the-art CAL inhibitors , our design methodology achieved higher affinity and increased binding efficiency . The designed inhibitor with the highest affinity for CAL ( kCAL01 ) binds six-fold more tightly than the previous best hexamer ( iCAL35 ) , and 170-fold more tightly than the CFTR C-terminus . We show that kCAL01 has physiological activity and can rescue chloride efflux in CF patient-derived airway epithelial cells . Since stabilizers address a different cellular CF defect from potentiators and correctors , our inhibitors provide an additional therapeutic pathway that can be used in conjunction with current methods .
Protein-peptide interactions ( PPIs ) are vital for cell signaling , protein trafficking and localization , gene expression , and many other biological functions . The PDZ ( PSD-95 , discs large , zonula occludens-1 ) family of proteins forms PPIs that play crucial physiological roles , including synapse formation [1] and epithelial cell polarity and proliferation [2] . The common PDZ structural core generally binds a specific sequence motif at the extreme C-terminus of its binding partner through -sheet interactions ( Fig . 1A ) . Recently , key PPIs have been discovered linking the trafficking of the cystic fibrosis transmembrane conductance regulator ( CFTR ) to PDZ domain containing proteins [3] ( Fig . 1B ) . Specifically , the PDZ domain of the CFTR-associated ligand ( CAL ) binds CFTR , targeting it for lysosomal degradation and reducing its half-life at the plasma membrane [4] , [5] . CFTR is an epithelial chloride channel that is mutated in cystic fibrosis ( CF ) patients . The most common disease-associated mutation , ΔF508-CFTR , is a single amino acid deletion that causes CFTR misfolding and endoplasmic reticulum-associated ( ER ) degradation . There is now evidence that the ΔF508-CFTR loss of function can be pharmacologically improved through the use of “correctors” [6] and “potentiators” [7] . Correctors , such as corr-4a [6] , [8] , work by correcting the folding defect of CFTR and preventing ER retention of CFTR . Potentiators combat mutant CFTR gating defects and increase the flow of ions through CFTR channels present at the cellular membrane . Despite these interventions , the half-life of ΔF508-CFTR in the membrane is still reduced compared to that of the wild-type protein [9] . However , the CAL-mediated degradation of ΔF508-CFTR can be reduced by RNA interference or by mutagenesis of the CAL PDZ domain , suggesting that a competitive inhibitor of the CAL binding site could act as a CFTR “stabilizer” and thus ameliorate CF symptoms [3] , [10] . Since stabilizers address a different underlying CF defect than correctors and potentiators , combined application can achieve additive rescue of ΔF508-CFTR activity [11] . Since PDZ domains have an inherent affinity for peptides , here we focus on the use of protein design methods to rationally design a competitive peptide inhibitor that could serve as a ΔF508-CFTR stabilizer . Indeed , the development of successful peptide inhibitor design tools would provide a means to target a wide variety of PPIs for both mechanistic and therapeutic applications . Several aspects of our new design algorithm ( described below ) are well suited to the requirements of this class of problems . In general , structure-based computational protein design seeks amino-acid sequences that are compatible with a specific protein fold . Often , additional functional constraints are applied to the problem in order to design a protein with a given binding or catalytic activity . Because protein conformational space is large , design algorithms often assume a fixed backbone conformation and reduce side-chain configuration space by using discrete conformations called rotamers [12]–[15] . Thus , most current design methods try to solve the traditional design problem , which can be defined as: for a given input model ( protein structure , rotamer library , and energy function ) , find the side chain rotamers that yield a single , global minimum energy conformation ( GMEC ) for the entire protein [16]–[34] . However , in reality , a protein in solution exists as a thermodynamic ensemble and not just a single low-energy structure [35] . Accounting for such ensembles can help find true native protein structures [36]–[39] . The design algorithm we present here , , takes this into account by computing Boltzmann-weighted partition functions over structural molecular ensembles to find provably-accurate approximations to the binding constant for a protein complex [40] , [41] . The value of this approach is reflected in previous applications of the algorithm to design a switch in enzyme specificity for an enzyme in the non-ribosomal peptide synthetase pathway [40] and to predict resistance mutations for antibiotic targets [42] . As with the established algorithm , most successful protein design studies have focused on protein/small molecule systems , since predicting PPI binding is more challenging than small molecule binding , due to PPIs' much larger , flexible , and energetically shallow binding surfaces . The methodologies that have been developed to study protein-protein interactions and , more specifically , PDZ domain interactions , can be divided into sequence- [43] , [44] and structure-based [38] , [45]–[49] methods . Sequence-based methods require a large amount of sequence and binding information for the protein family and do not provide direct structural information on the modeled interaction . Among the previous structure-based alternatives , most focus on finding the single GMEC conformation , although one study suggests that designing to a set of different backbone conformations can improve recovery of PDZ domain binding motifs [45] . In addition , only the work of Altman et al . [46] utilizes provable techniques , and none use both provable techniques and protein ensembles . In comparison , the algorithm is more general , requiring only a starting template structure and preserving structural information on the modeled interaction . It also evaluates energy-weighted ensembles , employs provable guarantees for finding the optimal sequence , and uses the minimization aware dead-end elimination ( minDEE ) pruning criteria [16] , [41] to permit continuous minimization of rotamers during the search . As a result , complements existing approaches while addressing some of their methodological limitations . Here we report the development of new extensions to the algorithm , enabling the software to design novel PPIs . Using this new tool we designed high-affinity CAL PDZ inhibitors and validated them in both biochemical and cell-culture experiments . We present peptide array data which shows that CAL binds a specific sequence motif , but does not bind all sequences within that motif . Therefore , it is important that the algorithm is able to differentiate the affinities of peptides that share the motif , rather than just separating motif from non-motif sequences . Overall , searched 2166 peptide inhibitor sequences within the CAL binding motif ( approximately possible conformations ) and generated top-ranked peptides that had up to a 170-fold improvement in binding to CAL compared to the wild-type CFTR sequence . The best binder was able to rescue ΔF508-CFTR function in human cells .
computationally searches over peptide amino acid substitutions ( mutations ) for a given protein-peptide complex and assigns each candidate sequence a score , called a score [40] , [41] . To compute the score for a given protein-peptide complex candidate sequence , evaluates the low-energy conformations for the sequence and uses them to compute a Boltzmann-weighted partition function . Partition functions are computed for each protein binding partner using rotamer-based ensembles defined as , , where is the partition function for protein bound to protein , and and are the partition functions for the unbound proteins , and . The score is defined as the ratio of partition functions: , which is an approximation of the protein complex association constant , [41] . Candidate sequences are ranked based on their score , where sequences with a higher score are considered to have a higher affinity for the target protein . The algorithm has been described previously [16] , [40] , [41] . Briefly , to calculate a partition function for a given sequence , finds low energy conformations by performing a rotamer search as follows . First , uses an enhanced version of dead-end elimination ( DEE ) , minDEE [16] , [41] , [50] , to prune side-chain rotamers that provably cannot be part of low-energy structures . Since rigid-rotamer DEE [34] , [51] often eliminates rotamers and sequences that are involved in bona fide low-energy conformations [50] , prunes rotamers using minDEE , which allows local side-chain rotamer minimization to relieve clashes that are incorrectly pruned by rigid rotamer design methods . In order for minDEE to account for minimization during the rotamer search , it computes energy lower bounds for each rotamer pair . The branch-and-bound algorithm [30] is used to enumerate conformations in gap-free order of their minimum energy bounds . These conformations are minimized and their Boltzmann-weighted energy is incorporated into the partition function . The partition function is computed with respect to the input model ( protein structure , energy function , and rotamer library ) , so the accuracy of the partition function is bounded by the accuracy of the input model . Refer to Fig . 2 to see the general framework for the algorithm . The energy minimization scheme that is used for both the energy lower bounds computation and the minimization of a full conformation is similar to previous descriptions [41] . The algorithm's minimization protocol separates a protein's degrees of freedom ( DOF ) into three categories: ( 1 ) backbone dihedrals ( and angles ) ( 2 ) side-chain dihedrals ( up to four angles per side chain ) and ( 3 ) rigid body rotation and translation ( ) . The minimization process holds the backbone dihedrals fixed while allowing the side-chain dihedral and rigid body DOF to minimize . The minimization over these DOF is performed using gradient descent . To prevent rotamers from minimizing from one rotamer to another , each side-chain dihedral was only allowed to move a maximum of from its modal rotameric value . relies on the mathematically provable guarantees of each of its steps ( Fig . 2 ) to compute an accurate score . If we were to use heuristic steps to find the low energy conformations , it could not be guaranteed that all the low energy conformations are found and we would lose the ability to calculate a provably-good -approximation ( where is user-defined ) to each partition function for the design system . Because of the provable aspects of , if makes an errant prediction , we can be certain that it is due to an inaccuracy in the input model and not a problem ( such as inadequate optimization ) with our search algorithm . This makes it substantially easier to improve the model based on experimental feedback , as we show in Section S2 of Text S1 . Before applying to PPI designs , we first had to ensure that the mathematical framework of could be extended to cover larger systems . For large designs such as PPIs , the provable guarantees of no longer hold as they did for small design systems . Specifically , the previous proofs [41] for intermutation pruning and guaranteeing the accuracy of the score , relied on properties of small molecule design systems that are not true for PPIs . We now show that it is possible to improve the algorithm to maintain these critical provable guarantees . As a result , systems where both binding partners in the protein complex are flexible or mutable during the search can be accurately studied using . Intermutation pruning uses computed partition functions to truncate the conformation enumeration process for candidate sequences when they will provably fail to achieve a score close to the best score . To show that an intermutation pruning criterion [41] exists for PPI design we seek a halting condition for the conformation enumeration such that we know we have an -approximation to the bound partition function for a given protein complex . First we observe: , where is the score of the current sequence , is the best score observed so far , and is a user-selected parameter . In the following lemma , is the number of conformations in the search that remain to be computed , is the number of conformations that have been pruned from the search with DEE , is the lower energy bound on all pruned conformations , is the universal gas constant , and is the temperature . The full partition function for the protein-protein complex , and unbound proteins are , , and respectively , while , , and denote the current calculated value of the partition functions during the computational search . The previously-determined NMR structure of the CAL PDZ domain bound to the C-terminus of CFTR ( PDB ID: 2LOB ) was used to model the binding of CAL to CFTR . To prepare the protein complex for the computational design , the initial complex structure was obtained by molecular dynamics refinement of the NMR structure as described previously [52] . Hydrogens were added to the structure using Reduce [53] . The CFTR peptide in the NMR structure was truncated to the six most C-terminal amino acids . An acetyl group was modeled onto the N-terminus of the peptide using restrained molecular dynamics and minimization in which the N-terminus of the peptide was allowed to move , while the remainder of the protein complex was restrained using a harmonic potential [54] . The coordinates of this starting structure are provided as supporting information ( Text S2 ) . An 8 Å shell around the peptide hexamer was used as the input structure to . The CFTR C-terminal residues , VQDTRL , were mutated to the following residues during the design search: to W , stayed fixed to Q , to all amino acids except Pro , to T/S , to all amino acids except Pro , and to I/L/V . In addition , the Probe program [55] was used to determine the side-chains on CAL that interact with the CFTR peptide mimic . The nine residues that interact with the peptide , as well as the two most N-terminal residues on the peptide , were allowed to be flexible during the design search ( Fig . 1A ) . To explore the feasibility of our new algorithms , unless otherwise noted , full partition functions were not computed and a maximum of conformations were allowed to contribute to each partition function . Rotamer values were taken from the Penultimate Rotamer Library modal values [14] . The energy function used to evaluate protein conformations has been previously described [40] , [42] . The energy function , , consists of a van der Waals term , a Coulombic electrostatics term , and an EEF1 implicit solvation term [56] . The EEF1 solvation term implicitly models water solvent during all of the computational designs . All design runs used the Amber98 [57] forcefield terms except for one prospective design run which used the Charmm19 [58] forcefield parameters . Previously-determined experimental binding constants [59] for 16 of CAL's natural ligands were used to train the energy function weight parameters ( See Text S1 Section S2 ) . scores were computed for each of the natural ligands . For this training , the CAL-CFTR structure only included the four most C-terminal residues of the peptide inhibitor . A gradient descent method was used to optimize the correlation between the scores and the experimental values . The final parameters chosen for the design runs are as follows: a van der Waals scaling of 0 . 9 , a dielectric constant of 20 , and a solvation scaling of 0 . 76 . was used to predict binding between the CAL PDZ domain and the HumLib set of 6223 human protein C-termini . The binding of the C-termini peptides to CAL was experimentally assessed using a peptide SPOT array [59] , [60] . Due to experimental restrictions , all cysteines in the HumLib peptide set were replaced by serine in the peptide array . For consistency , all computational predictions compared to the array modeled serines in the place of cysteines . A summary of the peptide array data is presented in Fig . 3 while the complete binding results from the array are provided as Supporting Information ( Table S1 ) . The algorithm was used to evaluate 4-mer structural models of 6223 peptide-array sequences to verify the accuracy of the algorithm's predictions . To compare the array data with the predictions , the quantitative array data , measured in biochemical light units ( BLUs ) , was converted into a binary yes/no CAL binding event . In other words , by using a fixed cutoff value , each sequence from the array was classified as either a CAL binder or non-binder . The cutoff value was chosen as three standard deviations away from the average BLU value of the array . A receiver operating curve ( ROC ) , which uses a floating cutoff to compare array data to scores , was used to evaluate the ability of to predict the array binding data . After the predictions were calculated , the binding of C-termini peptides to CAL was also experimentally assessed using an additional SPOT array . The profile library array ( ProLib; Fig . S3 in Text S1 ) was designed based on the following motif: bbbb ( B = permutation of a defined set of amino acids , b = mixture of 17 amino acids , without C , M and W ) . The defined set of amino acids were selected based on the HumLib results combined with substitutional analyses [60] with = A/C/D/E/F/I/K/L/M/N/Q/R/S/T/V/W/Y , = S/T , = A/C/D/E/F/I/K/L/M/N/Q/R/S/T/V/W/Y , = I/L/V ( Total number of peptides = 1734+22 internal control sequences ) . Incubation condition: His-tagged CAL PDZ domain detected by anti-His ( Sigma; 1∶2600 ) /anti-mouse-HRP ( Calbiochem; 1∶2000 ) antibody sandwich . was used to search over all peptide sequences within the CAL PDZ domain sequence motif ( excluding prolines ) to find new CAL peptide inhibitors . For computational efficiency the number of conformations enumerated by A* for each partition function was limited to conformations . Two sets of peptides ( promising designs and poorly ranked designs ) were chosen to be experimentally validated . In order to choose the most promising peptide inhibitors , a second design was done where scores for the top 30 sequences were re-calculated with the number of enumerated conformations per partition function increased to . Several top-ranked sequences were chosen to be experimentally tested . First , the top 7 ranked sequences from the second run were chosen . In addition , two sequences that greatly increased in ranking from the first to second run ( rank 29 to 9 , and rank 28 to 11 ) were chosen as well . Finally , a run was conducted using Charmm forcefield parameters instead of Amber parameters . Two sequences that scored high on both the Amber and Charmm runs were chosen to be experimentally tested as well ( Table 1 ) . The poorly-ranked designs were chosen to minimize the sequence similarity among the set of poorly-ranked peptides ( Table 2 ) . First , the worst-ranked peptide was chosen and added to initialize the set of negative sequences . Next , sequences were successively chosen from the worst 200 ranked sequences and added to the set in order to maximize the amino acid sequence diversity with all the sequences already in the set . The similarity between two sequences was determined using the PAM-30 similarity matrix [61] . In total 23 ( eleven top-ranked and twelve poorly-ranked ) K*-computed peptide inhibitor sequences were experimentally tested . The inhibitor dissociation constants of top- and poorly-ranked peptide sequences from the CAL-CFTR design were experimentally determined . As a control , the best known peptide hexamer was also retested . The corresponding N-terminally acetylated peptides were purchased from NEO BioScience ( Cambridge , MA ) and the values for the peptides were detected using fluorescence polarization ( FP ) , using the method previously described in [59] . Briefly , the CAL PDZ domain was incubated in FP buffer ( 25 mM Tris-HCl pH 8 . 5 , 150 mM NaCl; supplemented to a final concentration of 0 . 1 mg/mL bovine IgG ( Sigma ) and 0 . 5 mM Thesit ( Fluka ) ) with a labeled peptide of known binding affinity . Each peptide inhibitor was serially diluted and the protein-peptide mixture was added to each dilution . Finally , the amount of competitive inhibition was tracked using residual fluorescence polarization at temperatures between . Each value is reported as an average of three FP experiments conducted on separate days along with the corresponding standard deviation . Ussing chamber experiments were performed as described previously [11] . Polarized monolayers of patient-derived bronchial epithelial cells , CFBE-F cells ( a generous gift of Dr . J . P . Clancy [62] , [63] ) , were maintained in MEM with 2 mM l-glutamine , 10% fetal bovine serum , 50 units/mL penicillin , streptomycin , puromycin , plasmocin , and amphotericin B . Cells were grown at in 5% . Twenty four hours before treatment the cells were moved to MEM with only penicillin and streptomycin . Peptides were dissolved in DMSO and diluted to in PBS . Peptide solutions were applied to cells following incubation with BioPORTER delivery reagent ( Sigma ) . The final DMSO concentration did not exceed 0 . 03% . Following a 3 . 5 hour incubation with peptide , short circuit currents ( ) were monitored in Ussing chambers . Following treatment with amiloride , forskolin , and genistein , ΔF508-CFTR chloride flux was measured as the change in when the CFTR-specific inhibitor , [64] , [65] , was applied to the cell monolayer . All measurements were performed at .
To validate the algorithm , we compared predictions for CAL peptide inhibitors against peptide array binding data . First , peptides from the 6223 peptide HumLib library were tested for CAL binding using a SPOT array [59] . The array was able to find over one hundred peptides that clearly bind the CAL PDZ domain ( Fig . 3 ) . Second , predictions were made for all of the peptide sequences in the HumLib library . Fig . 4A shows the resulting receiver operating curve ( ROC ) when comparing the scores to the binding measurements ( BLU values ) of the peptide array . The ROC has an area under the curve ( AUC ) of 0 . 84 which shows that greatly enriches for peptides that bind CAL . Specifically , according to the peptide array , out of the top 30 predicted sequences , 11 are expected to bind CAL . Notably , this is a 20-fold increase over the number of binders that would be expected to be found if the CAL binding peptides were distributed randomly within the predictions . To investigate the success of the algorithm in more detail , we evaluated the importance of the CAL binding motif in determining predictions . The amino acid frequencies from the top binding peptides of the HumLib library ( Fig . 3C ) and natural binding partners of CAL [59] reveal that the canonical sequence motif of CAL is X-S/T-X-L/V/I . As expected , among the full set of HumLib peptides , enriches for sequences that conform to this motif . Furthermore , if we allow to design peptides varying at the primary motif positions 0 and −2 , it achieves an AUC of 0 . 94 ( Text S1 Section S3 and Fig . S2 in Text S1 ) , confirming its ability to identify the motif de novo . While also identified a few non-motif sequences in each case , the HumLib suggests that CAL actually can bind to such sequences , albeit less frequently ( 10 of 5867 sequences ) . Of course , the identification of motif residues , while a necessary test of the algorithm , does not by itself represent a major advance in affinity prediction . The HumLib library shows that only 70 out of 261 sequences with the CAL binding motif bind to CAL . A much more stringent test of the design algorithm is thus to determine how well enriches for binders among sequences that match the known CAL binding motif . As a first test , we recalculated the ROC curve considering only peptides in the HumLib library that match the CAL sequence motif , and was still able to significantly enrich for CAL peptide binders ( AUC = 0 . 71; Fig . 4B ) . This search , together with the blind test of rankings described below , provides a true test that the success of in predicting HumLib binders is not merely due to its identification of peptides conforming to the known sequence motif , but also to its ability to distinguish high- and low-affinity binders among such peptides . While SPOT arrays have proven to be a powerful tool for the identification of CAL binding peptides , the highest affinity inhibitors identified to date are composed of at least 10 amino acids . For hexamers , the highest published affinity is for iCAL35 ( WQTSII; [60] ) . Since was able to successfully enrich for CAL binders found in the HumLib library , we then used to prospectively find novel , shorter CAL peptide inhibitors , searching over 2166 peptides containing motif-based combinations of the C-terminal four residues . To facilitate accurate experimental binding-constant measurements , each peptide was extended by a shared N-terminal addition of the most frequent and residues among HumLib binders ( WQ ) , yielding hexamer sequences that exhibit a higher baseline affinity [59] . Both top- and bottom-ranked sequences were chosen for experimental validation . The value for each peptide hexamer was determined using fluorescence polarization [59] ( Table 1 ) . We used the same FP protocol to confirm the affinity of the acetylated iCAL35 reference peptide for CAL ( ) . All of our top-ranked inhibitors are novel CAL ligands , for which neither predicted nor experimental affinities were previously available . Remarkably , all of the top predicted peptides bind CAL with high affinity ( Fig . 5A , Table 1 ) . The tightest binding predicted peptide ( kCAL01 , WQVTRV ) had a of . While this affinity is comparable to that of several other PDZ inhibitors [66] , [67] , solution-state measurements show that the CAL PDZ domain exhibits systematically weak interactions with target C-termini: note that the for the wild-type CFTR sequence ( TEEEVQDTRL ) is and the best known affinity natural ligand ( ANGLMQTSKL ) for CAL is [60] . Thus , our design algorithm successfully identifies high affinity peptide inhibitors of the CAL PDZ domain , with 170-fold higher affinity than the interaction we were trying to inhibit and 9-fold higher affinity than any comparable natural ligand . This peptide affinity advantage may be important in physiological applications , since the native CAL∶CFTR target interaction may involve additional sources of affinity outside the PDZ binding pocket [4] , [59] , not available to a peptide inhibitor . We also performed further analysis of the HumLib SPOT array used for validation . Selecting the most common amino acid at positions to among HumLib binders yields the sequence WQSTRL ( HumLib01 , Fig . 3C ) , which is ranked in the top 50 predictions ( out of 2166 ) . This sequence is also the strongest binder identified among the ProLib sequences ( see below , and Fig . S3 in Text S1 ) . However , when we measured the CAL binding for HumLib01 using fluorescence polarization ( FP ) it exhibited a value of , only a marginal improvement in affinity compared to iCAL35 ( ) . In comparison , five of the eleven top predicted sequences we measured with FP show an improvement in binding compared to both iCAL35 and HumLib01 , and kCAL01 shows a six-fold improvement over both iCAL35 and the HumLib01 sequence . The best inhibitor found through previous FP and array screens involves a fluorescein group modification to a peptide decamer ( F*-iCAL36 , F*-ANSRWPTSII , ) . kCAL01 rivals this binding affinity despite the computational search library restriction to only allow amino acids and hexamer sequences . Critically , at 830 Da , kCAL01 has approximately twice the binding efficiency ( ratio of inhibitor potency , G , to molecular mass ) of F*-iCAL36 and is much closer in size to typical drugs . This makes kCAL01 a very promising inhibitor compared to F*-iCAL36 and other discovered inhibitors . Furthermore , as suggested by our retrospective tests , the tight binding of our top-ranked sequences was not merely a consequence of the underlying CAL-binding motif used to select candidate sequences for evaluation . To establish this , we selected a set of poorly-ranked peptides to minimize sequence similarity and evaluated their CAL-binding affinity experimentally . Almost all of the poorly-ranked sequences bound CAL , consistent with their motifs ( Fig . 5A ) . Reflecting the enrichment of CAL binders in the pool , the two poorly-ranked peptides with the best affinities ( and , respectively ) were indeed close to the affinity of the weakest top-ranked sequence ( ) . However , all of the poorly ranked peptides bound CAL more weakly than any of the top-ranked sequences ( Table 1 ) , and none of them had improved affinity relative to prior biochemical efforts . This suggests that can efficiently distinguish among motif-bearing peptides , allowing it to predict sequences with CAL affinities unprecedented among hexamers . Detailed analysis of the predictions suggests that the use of both ensemble-weighting and minDEE approaches was important in the success of the algorithm . The ensembles generated by do not have a dominant conformation , i . e . , a conformation with significantly lower energy than the others , which would thus dominate in the partition function . For example , in the case of iCAL35 ( WQTSII ) , found 75 conformations that were within 0 . 5 kcal/mol and 454 conformations that were within 1 kcal/mol of the iCAL35 GMEC . In general , the ensemble conformations are consistent with canonical PDZ:peptide interactions and with the conformation of the CAL-bound CFTR peptide determined by NMR [52] . To determine the importance of the ensemble-based rankings we compared the predictions to two single-structure GMEC-based methods , minDEE [41] , and rigid-rotamer DEE ( rigidDEE ) [68] . Both minDEE and rigidDEE were run with the same energy parameters as the designs . However , since the single-structure designs only compute the energy of the bound state , reference energies [16] were included as in [69] to account for the energy of the unbound state . The inclusion of reference energies for single-structure designs have been deemed necessary by most protein designers to account for the unfolded/unbound state [24] , [69] , [70] . does not need reference energies since it calculates a partition function for both the bound and unbound states of the complex [16] , [40] . Therefore , reference energies are included to make the comparison between and the single-structure designs more fair . We compared the top 30 sequences from minDEE and rigidDEE and found they had no sequences in common . This supports previous work where we have shown that in over 69 protein design systems minDEE finds low energy sequences that rigidDEE discards by not allowing minimization [41] , [50] . In addition , when we compare the top 30 rigidDEE and minDEE results to the top designs we find that they have only three and four sequences in common , respectively . If we had used only GMEC-based approaches instead of , we would not have predicted most of the experimentally successful sequences that found , including the best inhibitor kCAL01 . In addition , the overall sequence rankings show a very poor correlation between the minDEE and predictions; the same is true of the rigidDEE and predictions ( = 0 . 1 and 0 . 09 respectively ) . The prospective peptide predictions demonstrate that can successfully find CAL peptide inhibitors . Our solution-state binding tests provide robust information for the best and worst K*-predicted peptides , but give little information about the CAL binding of the remaining peptides that match the CAL motif . To investigate this experimentally , we designed a peptide library SPOT array ( ProLib ) based on the HumLib motif combined with substitutional analyses [60] . The resulting sequences closely match our prospective prediction set and the binding of these sequences to CAL was assessed as described in the Materials and Methods section . Using a similar analysis to that performed on the HumLib peptide array we compared the predictions to the CAL binding observed with the ProLib array . We found an AUC = 0 . 88 ( Fig . 6 ) . Note that this AUC is much higher than the 0 . 71 found when only looking at CAL motif sequences within the HumLib array . One explanation for this improvement is that the experimental setup is closer to the design model used by . Specifically , the ProLib array uses a mixture of amino acids at to of the peptides , while the HumLib array is composed of decamer peptides . Thus , the ProLib data focuses on the identity of the last 4 C-terminal positions , which better matches the sequence and structure search space of the designs . A complete evaluation of the accuracy of affinity predictions would require the synthesis and FP binding analysis of all 2166 sequences within the CAL binding motif . However , taken together , the FP measurements for the designed peptides plus the ProLib blind test suggest that is a powerful filter , efficiently selecting tight binders from a pool of sequences with baseline affinity for the target . All of our top-predicted inhibitors successfully bound CAL , which suggests that they should disrupt the degradation pathway of CFTR . The ability of kCAL01 to restore ΔF508-CFTR function was assessed by measuring CFTR-mediated chloride efflux in CF-patient derived bronchial cells expressing ΔF508-CFTR ( CFBE-F ) using an Ussing chamber apparatus [11] . As a control peptide , we used kCAL31 ( WQDSGI ) , which was ranked as the weakest interactor by and for which no binding was detected experimentally ( Table 2 ) . Fig . 7 shows ΔF508-CFTR chloride secretion across polarized monolayers treated with either kCAL31 , the iCAL35 reference peptide , or kCAL01 . Previous studies with fluorescently labeled peptides have demonstrated delivery into CFBE-F cells using the BioPORTER reagent [11] . Significance of rescue was evaluated by comparing percentage improvement in chloride efflux to rescue from a well-established “corrector” under identical conditions , and by Student's -test ( -value ) . Compared to the non-binding control , the previously best hexamer , iCAL35 , yields only a slight ( non-significant ) improvement in chloride secretion ( 4% , ) . In contrast , chloride secretion following treatment with the designed inhibitor kCAL01 is significantly enhanced with respect to the control peptide ( 12% , ) and with respect to the reference ( 8% , ) peptide . Indeed , the biological activity of kCAL01 is very similar to that observed under similar conditions following treatment with either the best previously available CAL inhibitor ( F*-iCAL36 ) or the first-generation corrector corr-4a [6] , [11] .
The source code of our program is freely available , and is distributed open-source under the GNU Lesser General Public License ( Gnu , 2002 ) . The source code can be freely downloaded at http://www . cs . duke . edu/donaldlab/osprey . php . | Cystic fibrosis ( CF ) is an inherited disease that causes the body to produce thick mucus that clogs the lungs and obstructs the breakdown and absorption of food . The cystic fibrosis transmembrane conductance regulator ( CFTR ) is mutated in CF patients , and the most common mutation causes three defects in CFTR: misfolding , decreased function , and rapid degradation . Drugs are currently being studied to correct the first two CFTR defects , but the problem of rapid degradation remains . Recently , key protein-protein interactions have been discovered that implicate the protein CAL in CFTR degradation . Here we have developed new computational protein design algorithms and used them to successfully predict peptide inhibitors of the CAL-CFTR interface . Our algorithm uses a structural ensemble-based evaluation of protein sequences and conformations to calculate accurate predictions of protein-peptide binding affinities . The algorithm is general and can be applied to a wide variety of protein-protein interface designs . All of our designed inhibitors bound CAL with high affinity . We tested our top binding peptide and observed that the inhibitor could successfully rescue CFTR function in CF patient-derived epithelial cells . Our designed inhibitors provide a novel therapeutic path which could be used in combination with existing CF therapeutics for additive benefit . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biomacromolecule-ligand",
"interactions",
"algorithms",
"biochemistry",
"biochemical",
"simulations",
"computer",
"science",
"computational",
"chemistry",
"chemistry",
"biology",
"computational",
"biology",
"biophysics",
"drug",
"discovery"
] | 2012 | Computational Design of a PDZ Domain Peptide Inhibitor that Rescues CFTR Activity |
Burkholderia cenocepacia causes severe pulmonary infections in cystic fibrosis ( CF ) patients . Since the bacterium is virtually untreatable by antibiotics , chronic infections persist for years and might develop into fatal septic pneumonia ( cepacia syndrome , CS ) . To devise new strategies to combat chronic B . cenocepacia infections , it is essential to obtain comprehensive knowledge about their pathogenesis . We conducted a comparative genomic analysis of 32 Czech isolates of epidemic clone B . cenocepacia ST32 isolated from various stages of chronic infection in 8 CF patients . High numbers of large-scale deletions were found to occur during chronic infection , affecting preferentially genomic islands and nonessential replicons . Recombination between insertion sequences ( IS ) was inferred as the mechanism behind deletion formation; the most numerous IS group was specific for the ST32 clone and has undergone transposition burst since its divergence . Genes functionally related to transition metal metabolism were identified as hotspots for deletions and IS insertions . This functional category was also represented among genes where nonsynonymous point mutations and indels occurred parallelly among patients . Another category exhibiting parallel mutations was oxidative stress protection; mutations in catalase KatG resulted in impaired detoxification of hydrogen peroxide . Deep sequencing revealed substantial polymorphism in genes of both categories within the sputum B . cenocepacia ST32 populations , indicating extensive adaptive evolution . Neither oxidative stress response nor transition metal metabolism genes were previously reported to undergo parallel evolution during chronic CF infection . Mutations in katG and copper metabolism genes were overrepresented in patients where chronic infection developed into CS . Among professional phagocytes , macrophages use both hydrogen peroxide and copper for their bactericidal activity; our results thus tentatively point to macrophages as suspects in pathogenesis towards the fatal CS .
In cystic fibrosis ( CF ) patients , thick sputum obturates the airways as a result of CFTR chloride channel defect . This environment is populated by bacterial communities which often include pathogens such as Staphylococcus aureus , Pseudomonas aeruginosa , Haemophilus influenzae and Stenotrophomonas maltophilia [1] . Bacteria from Burkholderia cepacia complex ( Bcc ) , a monophyletic group within the genus which currently comprises 20 species [2] , have emerged during the 1980s as CF pulmonary pathogens [3] . Bcc are generally regarded the most harmful CF pathogens; the infections are associated with significant decline in lung functions and the lingering threat of development of cepacia syndrome [4] , a fulminant necrotizing pneumonia with high fatality rate [5] . B . cenocepacia ( representing the former genomovar III ) is one of the most prevalent Bcc species encountered in CF infections [6] . Among B . cenocepacia , two lineages were delineated based on recA sequence similarity: IIIA and IIIB [7] . The IIIA lineage , syn . clonal complex 31 ( multilocus sequence typing [MLST] CC31 ) , is by far the largest cluster in Bcc MLST database [8] , reflecting their frequent isolation from patients . Furthermore , the IIIA lineage was reported to show the most pronounced isolation bias; virtually all MLST sequence types were isolated from clinical sources [9] , suggesting tight association with humans or even an ongoing switch from environmental to host-associated lifestyle [10] . Studies clearly demonstrated the particularly destructive nature of IIIA infections in CF patients when compared with other Bcc bacteria [11 , 12] . B . cenocepacia IIIA are known to cause epidemic outbreaks [13]; recently , they were reported to dominate in Serbia ( ST856 [14] ) and Russia ( ST709 [15] ) . The most notorious epidemic IIIA bacterium is the ET12 clone , a hypervirulent transatlantic strain responsible for large infection outbreaks in Europe and North America [16] . Another globally distributed clone , ST32 , was detected in Italy , France , UK and Canada [17] . Czech CF patients were plagued by epidemic spread of this B . cenocepacia strain ( also called CZ1 [18] ) in the 1990s . Out of the 57 patients infected; only 15 were alive by 2015 [19] . In this work , we aimed to elucidate the evolution of B . cenocepacia ST32 during chronic pulmonary CF infections with fatal outcomes . A comprehensive comparative genomic analysis was conducted , covering multiple isolates from multiple patients . The genes and pathways exhibiting parallel evolution and the underlying mutational processes are reported .
First , we examined the genetic relationship of the Czech epidemic strain B . cenocepacia ST32 ( CZ1 ) with other sequenced strains from the B . cenocepacia group recA IIIA ( CC31 ) . Whole-genome phylogenetic analysis showed that ST32 was clearly distinct from the ET12 epidemic clone ( S1 Fig ) . To gain insight into ST32 evolution during chronic CF pulmonary infection , a total of 32 isolates were selected for whole-genome sequencing ( WGS ) . The isolates originated from 8 chronically infected patients who ultimately developed cepacia syndrome ( CS ) ( S1 Table ) . Each patient was represented by 4 chronological isolates . These covered most of the known period of their chronic infection ( average 7 . 9 years ) . The collection included the first archived sputum isolate , one mid-term sputum isolate and isolates from last sputum sample and blood culture; the two latter isolates corresponded to the time of CS . The only exception was patient 1 where no blood culture isolate at the time of CS was available; thus 1 additional sputum isolate was used to complete the set . Interestingly , patient 8 survived CS episode for one more year , so his last sputum and blood isolates were collected one year before the date of death . All other patients died soon after dates of collection of their CS isolates ( Fig 1A ) . In silico MLST analysis of WGS sequences confirmed that all 32 isolates belonged to ST32 . The presence of ST32-specific DNA sequence [20] was detected in genomic sequences of all isolates . From whole-genome sequences of ST32 isolates , phylogenetic tree was constructed ( Fig 1B ) and the results were assessed with respect to their spatiotemporal relationships . Virtually all isolates ( 30/32 ) clustered into patient-specific lineages . This implies that the populations have originated from a single colonization event per patient , followed by subsequent diversification . The intra-patient branching patterns did not always follow the chronology of isolation . For example , blood culture isolates 5D and 7D branched out at the earliest and so did the last sputum isolate 8C . This indicates long-standing co-existence of different subpopulations which evolved from the original colonization . The isolates accumulated single-nucleotide polymorphisms ( SNPs ) over the course of infection ( Fig 1C ) . SNP accumulation rate ( calculated as linear regression slope ) in patient 2 was over an order of magnitude higher than in remaining patients . This was in accordance with markedly increased genomic diversity of isolates retrieved from this patient ( Fig 1B ) . The mutation frequency values determined for isolates 2A-2D corresponded to hypermutator phenotype ( 4 . 9 x 10−6–8 . 5 x 10−6 ) , while all remaining isolates displayed nonmutator values ( 3 . 6 x 10−9–5 . 3 x 10−8 ) ( measured and interpreted according to Martina et al . [21] ) Hypermutability was associated with a 4-nt deletion in the mismatch-repair gene mutS . For nonmutator lineages , the weak correlation between SNP numbers and infection duration did not allow for reliable SNP rate calculation ( Fig 1C ) . The value calculated from best linear fit ( 1 . 66 SNPs/ year in patient 7 ) was slightly lower , yet comparable with SNP rates reported during chronic pulmonary infection for other CF pathogens such as B . dolosa ( 2 . 1 SNPs/year [22] ) , B . multivorans ( 2 . 4 SNPs/year [23] ) , P . aeruginosa ( 2 . 7 SNPs/year [24] ) and Burkholderia pseudomallei ( 3 . 6 SNPs/year [25] ) . Mapping of sequencing reads to reference genome detected many cases when large portions of reference genome showed missing coverage . Most of these large-scale deletions ( LSDs ) were not shared among patients and thus represented regions of ST32 genome lost during diversification in chronic infection ( Fig 2 , S3 Table ) . The putative genomic islands ( GIs ) specific for ST32 epidemic clone were detected by pairwise comparison of its genomic sequence with ET12 isolate B . cenocepacia J2315 [10] and termed GiST32-01 to GiST32-16 ( S4 Table ) . Analysis of possible association between LSDs and GIs revealed important differences among the four replicons ( Fig 2 ) . LSDs on essential replicons ( chromosomes 1 and 2 ) were localized almost exclusively in ST32 GIs . Different pattern was observed with non-essential replicons ( chromosome 3/megaplasmid [26] and plasmid [27] ) where LSDs removed significant portions of DNA regardless of GI positions , in accordance with the dispensable nature of these replicons . Although numerous LSDs were present in non-essential replicons , no complete loss of either chromosome 3 or plasmid was detected among our ST32 dataset . On the contrary to ST32-specific GIs , GIs that were shared with B . cenocepacia J2315 , including the previously characterized cenocepacia island cci [28] and the lxa locus [29] , were not affected by LSDs at all . This , in conjunction with their conservation between the two epidemic strains , corroborates their functional importance . Preliminary inspection revealed transposase genes at the borders of many LSD regions , which prompted us to look into ST32 transposable elements ( insertion sequences , ISs ) in more detail . 13 groups of ISs present at two or more copies in reference genome were detected , and their abundance in complete B . cenocepacia IIIA genomes was assessed . Strikingly , the most numerous IS group 01 ( 53 copies in ST32 reference genome ) was found to be absent from genomes of all closely related B . cenocepacia strains , indicating its very recent acquisition and transposition burst ( S5 Table ) . Insertions of all IS groups were analyzed in silico in the WGS dataset of ST32 isolates using ISMapper [30] ( Fig 2 , S6 Table ) . Numerous new insertions were detected in the dataset; the most abundant IS ( group 01 ) was also the most mobile . Many cases of intra-patient , lineage-specific IS insertions were observed , corroborating their relationships inferred from whole-genome phylogeny ( Fig 1B ) . IS insertions were substantially enriched in GIs in comparison with the rest of the genome ( Fig 2 , S6 Table ) . Detailed examination of ISs in the ST32 WGS dataset revealed that several genes experienced multiple , independent IS insertions during chronic infection . When analyzed for the presence of conserved domains [31] , metal-related functions were predicted . These genes ( TQ36_15160 , TQ36_15180 , TQ36_25385 and TQ36_35715/copD; S2 Fig ) were located within ST32-specific GIs and were surrounded by other metal-related genes , each on a different replicon . In addition to IS insertions , deletions of varying sizes were detected upon detailed analysis of mapped sequencing reads ( S2 Fig ) . The extent of parallelism was remarkable: each gene was inactivated in at least 6 out of total 8 patients . The plasmid-located copper resistance gene copD ( TQ36_35715 ) and neighboring copper-related genes were affected by the greatest number of deletions ( 8 independent events; S2 Fig ) . Together , parallel IS insertions and deletions suggest that inactivation events were positively selected during chronic infection in CF sputum . In the search of further evidence for convergent evolution among bacteria from different patients , we analyzed point mutations , i . e . single nucleotide polymorphisms ( SNPs ) and short insertions and deletions ( indels ) . We focused on intragenic , nonsynonymous mutations which have arisen during diversification of ST32 populations after initial colonization of patients’ lungs . Analysis of SNP distribution showed that within the ST32 WGS dataset , genes harboring 2 or more independent SNPs were overrepresented in comparison with neutral model ( S3 Fig ) ; ≥3 SNPs per gene were not predicted to occur in the neutral model . Genes were thus considered to have undergone parallel evolution if they received 3 or more independent nonsynonymous mutations . 16 out of 6 , 939 genes present in ST32 genome met this criterion ( Fig 3 ) . Multiple nonsynonymous mutations were typically present ( 4 . 75 unique mutations per gene on average ) , while synonymous mutations ( a measure of “background” mutation rate ) were absent in a great majority ( 13/16 ) of the investigated genes . These values deviate strongly from the theoretical frequencies calculated by Dillon et al . for B . cenocepacia ( 72% nonsynonymous vs . 28% synonymous mutations [32] ) , indicating positive selection acting upon the genes . Based on available literature , genes harboring parallel nonsynonymous mutations ( Fig 3 ) were grouped into 5 functional classes: antibiotic resistance ( 4 genes ) , global transcription regulation ( 4 genes ) , oxidative stress protection ( 3 genes ) , transition metal metabolism ( 2 genes ) and general stress protection ( 2 genes ) ; afcE , whose role in B . cenocepacia physiology is complex [33] , was the only uncategorized protein . Antibiotic resistance genes and global transcription regulators were previously described to undergo parallel evolution in B . dolosa [22]; mutations in gyrA affected the same amino acid positions and were associated with high-level levofloxacin resistance ( S1 Table ) . In contrast , genes predicted to function in oxidative stress protection and metal metabolism have not yet been reported to be subject to selection during chronic pulmonary infection in other CF pathogens ( see Discussion ) . KatG ( BCAL3299 , also called KatB ) , the major hydrogen peroxide-detoxifying enzyme with hybrid catalase/peroxidase activity [34 , 35] , was among the three proteins connected with oxidative stress protection . Mutations in KatG were restrained to three sites; two identical mutations ( 246M→L and 570A→E ) have arisen independently in two patients each , suggesting a tightly limited parallel evolution . Furthermore , the region containing the katG gene was multiplicated in isolates from patient 6; as deduced from sequencing read coverage , the copy numbers increased from 4 to 8 during the progress of infection ( Fig 3 ) . Together , these results indicate strong selection acting upon KatG . In contrast , other catalase genes present in ST32 genome ( homologs of BCAL3477 , BCAM0181 , BCAM0931 and BCAS0635 in B . cenocepacia J2315 ) did not harbor multiple mutations . Another oxidative resistance protein found to be under parallel evolution , YedY ( BCAL0269 ) , is a methionine-sulfoxide reductase which repairs periplasmic proteins [36] . In Escherichia coli , YedY repairs proteins damaged by hypochlorous acid , a molecule which also induces the expression of yedY [36] . YedY requires a molybdopterin cofactor for catalysis; it is the only molybdopterin-dependent enzyme in E . coli which uses this cofactor in its nucleotide-free form [37] . Since MoeA1 ( BCAL2891 ) catalyzes precisely the final step in the biosynthesis of nucleotide-free molybdopterin , it was regarded to belong to the same functional category as YedY . Among the most mutated genes , two were related to transition metal metabolism . BCAL0155 is a member of cation diffusion facilitator family of divalent metal efflux transporters [38] whose substrate specifity has not yet been determined . The sensory kinase CusS ( BCAM1417 ) senses periplasmic copper and through its cognate response regulator activates copper resistance mechanisms [39] . Importantly , the genes experiencing rapid inactivation by deletions and IS insertions were also predicted to perform metal-related functions ( see above ) . Given the rapid convergent evolution of genes involved in transition metal metabolism and the functional importance of oxidative stress defense , all genes under parallel evolution belonging to these two categories ( together with an essential gene rpoB ) were subjected to further analysis , which aimed to unravel the true extent of genetic parallelism in B . cenocepacia ST32 chronic CF infections . We investigated 12 sputum samples collected from 12 patients chronically infected with ST32 ( Table 1 ) . These patients were all clinically stable at the time of sputum collection and have not developed CS ( “non-CS patients” ) . The genes were PCR-amplified from total sputum DNA . Further processing differed according to the type of mutations observed in the WGS dataset: 6 genes affected by point mutations ( Fig 3 ) were subjected to deep population sequencing ( DPS ) , while PCR amplicons of 4 genes affected by structural variation ( S2 Fig ) were sized by agarose electrophoresis . The composition of detected mutations allowed us to predict the type of selection for all investigated genes ( Table 1 ) . In BCAL0155 and moeA1 , short deletions and frameshifts were present in addition to nonsynonymous substitutions , indicating selection for inactivation . In contrast , only nonsynonymous substitutions were detected in cusS , katG , yedY and the control essential gene rpoB ( BCAL0226 ) , suggesting selection for more subtle functional alterations of the encoded proteins . MoeA1 appeared as the first-line protein to mutate; typically , mutations were fixed or nearly-fixed in ST32 population , while other genes were mutated only in a subset of the same population . Indirectly , this implies substantial selective advantage of MoeA1 inactivation which had enabled the mutated bacterium to outcompete its direct ancestors and rise to fixation before other mutations appeared . For other proteins , different pattern was observed in some ST32 populations; up to 7 different mutations coexisted in a patient , typically covering the entire ST32 population ( as inferred from their summed frequencies ) ( Table 1 ) . Selective pressure for mutations in genes undergoing convergent evolution might therefore drive genetic diversification of pulmonary ST32 populations . Electrophoretic analysis of metal-related genes located on ST32 GIs ( S2 Fig ) confirmed their frequent inactivation during ST32 chronic infection ( Table 1 ) . Since PCR can neither detect deletions spanning amplified fragment nor reveal deleterious point mutations if present in presumably heterogeneous ST32 populations , the reported extent of inactivation is likely to be underestimated . To further characterize the wealth of mutations detected to have arisen during chronic ST32 infection , we mapped them to available 3D structures of homologous proteins ( RpoB , CusS , KatG and YedY ) ( Fig 4 ) . For reference , mutations extracted from genomic sequences of a diverse dataset of B . cenocepacia IIIA isolates representing various clonal lineages were included ( S9 Table [40] ) . Mutations in YedY appeared scattered through the structure . Most mutations in the catalytic subunit of RNA polymerase ( RpoB ) localized into a distinct cluster which overlapped or neighbored with the βi4 region ( also called dispensable region I ) [41] , whose function is yet to be established . Curiously , we found dataset-dependent mutation patterns . In KatG , all 3 residues mutated in isolates from CS patients ( WGS dataset ) were in direct contact with the catalytic MYW cofactor or arginine switch [42] , while mutations from both other datasets mapped to residues positioned further apart from the cofactor . Mutations in the sensory kinase CusS also showed different distribution of mutations . Residues mutated in CS patients resided exclusively in the DHp ( dimerization and histidine phosphotransfer ) domain [43] ( S10 Table ) and its immediate vicinity . On the other hand , mutations from non-CS patients localized randomly throughout the protein , without preference for DHp . Finally , we examined if the presumed adaptive mutations resulted in corresponding phenotypic changes . We compared in vitro susceptibilities of all ST32 isolates which were genotypically characterized by WGS ( Fig 1 ) towards following substances: copper ( II ) chloride ( CuCl2 ) , sodium hypochlorite ( NaClO ) and hydrogen peroxide ( H2O2 ) . All isolates exhibited uniform level of resistance to both CuCl2 and NaClO ( 8 mM and 0 . 0625% , respectively ) , despite the multitude of mutations affecting repair of NaClO-induced oxidative damage ( yedY , moeA1 ) and metabolism of copper ( cusS , copCD ) . In sharp contrast , H2O2 resistance varied significantly; 8-fold range of MIC values was observed ( Fig 5 ) . All point mutations in katG were associated with decreased resistance , implicating impaired detoxification of H2O2 by mutant KatG . Furthermore , a trend of MIC decrease during chronic infection was observed in 5 out of 8 patients ( Fig 5 ) .
A particular insertion sequence ( IS group 01 ) detected in the epidemic strain ST32 showed conspicuous characteristics . With over 50 copies in reference ST32 genome , this IS was more abundant than any other IS in Bcc genomes investigated [52] . Interestingly , IS group 01 was absent in all other clones within the IIIA lineage ( S5 Table ) . This implies that IS group 01 was acquired very recently during divergence of the ST32 clone and has since then undergone excessive transposition . Similar cases of IS proliferation were detected in flexible genomes of bacteria rapidly adjusting to new lifestyles [53] , for example Shigella spp . adapting to human intracellular environment [54] or Burkholderia mallei becoming a host-specific obligate pathogen [55] . Expansion of IS elements is recognized to be one of general mechanisms underlying evolution of bacterial species which recently diversified from a single clone into highly virulent human-restricted pathogens [56] . Interestingly , among B . cenocepacia genomes studied by Graindorge et al . [52] , the hypervirulent clone ET12 ( isolate J2315 ) was found to harbor the largest number of IS copies . Furthermore , in both ST32 and ET12 , genomic islands specific for each of the clones were substantially enriched in IS insertions ( Fig 2 , [52] ) . This suggests that IS proliferation is a general phenomenon in evolution of virulent lineages of B . cenocepacia . Our analysis of ST32 isolates also detected copious IS insertions which occurred during chronic CF infection ( S6 Table ) , directly demonstrating the extent of genetic plasticity conferred by these mobile elements . The second most numerous IS group 2 ( syn . ISBcen20 ) was previously noted to be highly mobile under conditions of oxidative stress [57] . Surprisingly , we observed many LSDs to arise in ST32 genomes during chronic CF infection , preferentially affecting GIs and nonessential replicons . Investigating the mechanism behind LSD formation , we often detected IS at the termini of deleted regions ( S2 Fig , S6 Table ) . Recombination between two congruent IS copies was reported as a major mechanism generating spontaneous deletions in E . coli experimental evolution [58 , 59]; apparently , LSDs in chronic ST32 infection arise by the same mechanism . A question is why LSDs have occurred in ST32 so frequently ( Fig 2 , S3 Table ) . Some genes located on GIs clearly experienced convergent inactivation , either by IS insertions or LSDs ( S2 Fig , Table 1 ) , raising the possibility that some LSD might be adaptive , i . e . subject to positive selection . In E . coli , deletions between two IS copies were found to occur at high frequency and became rapidly fixed in parallel evolving experimental populations if they provided a rather minor fitness gain [60] . LSD formation is an example of reductive genome evolution; like IS proliferation , reductive evolution is characteristic for bacteria evolving into host-dependent pathogens [56 , 61] . Several genes harbored point mutations in independent ( i . e . patient-specific ) ST32 lineages , a characteristic of parallel ( convergent ) evolution . Strikingly , most of these genes have not been reported to be subject to parallel evolution in B . dolosa , a distantly related Bcc bacterium whose genetic evolution in chronic CF infection was studied in detail [22 , 50] . The handful of shared genes ( 4/16 ) either received lowest numbers of parallel mutations in ST32 ( spoT , fixL , rpoD ) or were mutated due to antibiotic selective pressure ( gyrA ) ( Fig 3 ) . Altogether , this implies that evolution of both Bcc species in chronic CF infection is driven by different selective forces . This distinction may be genetically-grounded , predetermining both primarily environmental bacteria to establish persistent infection by own independent means . Genes linked to oxidative stress response and transition metal metabolism were markedly represented among the most mutated genes ( Fig 3 ) . These functional categories have not been reported to undergo adaptive within-patient evolution either in B . dolosa or in the well-characterized CF pulmonary pathogen Pseudomonas aeruginosa [62] . Proteins encoded by the three genes are involved in protection against two reactive oxygen species ( ROS ) : hydrogen peroxide ( KatG ) and hypochlorous acid ( YedY , MoeA1 ) . Both ROS are produced by leukocytes as bactericidal agents . Thus , our findings point to a fundamental role of host immune system in driving B . cenocepacia evolution during chronic CF infection . Chronic pulmonary infections are accompanied by persistent inflammation and neutrophil infiltration . Extracellular hypochlorous acid production by CF neutrophils is not compromised [63] and results in significant chlorination damage in sputum [64] . Although ST32 isolates from our collection which carried various presumably adaptive mutations in YedY and/or MoeA1 were uniformly sensitive to sodium hypochlorite in vitro , the unprecedented extent of parallelism suggests principal importance of these mutations in vivo . YedY and MoeA1 were frequently mutated not only in 8 patients who developed CS ( Fig 3 ) , but also in 12 control non-CS patients ( Table 1 ) and in other B . cenocepacia IIIA strains ( S9 Table [40] ) . Repair of oxidized periplasmic proteins is thus a previously unrecognized target of adaptive evolution during chronic infection progress in B . cenocepacia . Other genes carrying parallel mutations revealed transition metal ( copper ) metabolism as a target of adaptive evolution ( see below ) . Copper has recently been shown to concentrate in macrophage phagolysosomes , aiding in the clearance of ingested bacteria and fungi [65–67]; suggesting possible functional connection between these two categories . Pathogenesis of the fatal CS still remains largely unknown . All ST32 isolates initially characterized by WGS originated from CS patients . In addition , polymorphisms were analyzed in oxidative stress and transition metal metabolism genes in ST32 populations from non-CS patients . Upon the comparison of the two datasets , several genes were detected where mutations segregated among CS and non-CS patients; however , due to inevitably small numbers of patients included , the differences did not reach statistical significance . copCD operon was inactivated in every CS patient but one ( 7/8; 88% ) by independent deletions or IS insertions , these events were biased towards late isolates ( S2 Fig ) . In contrast , only in 5/12 ( 42% ) populations from non-CS patients were these mutations present in detectable frequencies ( Table 1 ) ( p = 0 . 07 , Fisher´s exact test ) . In the copper-sensing histidine kinase CusS , only mutations from CS patients localized exclusively to the DHp domain ( Fig 4 , S10 Table ) . Mutations in copper-related genes did not modulate the measured in vitro sensitivity of ST32 towards copper; however , their abundance and/or specific pattern in CS isolates strongly suggest a yet undisclosed role of this metal in CS pathogenesis . Catalase KatG was mutated in 2/12 non-CS patients ( 17% ) . In contrast , isolates from 4/8 CS patients ( 50% ) carried nonsynonymous mutations in katG ( p = 0 . 16 , Fisher´s exact test ) ; another CS patient was colonized with population whose katG region was multiplicated ( Fig 5 ) . Curiously , some epidemic B . cenocepacia isolates have been known to possess a paralog of KatG , which is 76% identical and performs different cellular functions than the canonical KatG [35] . Indeed , the hypervirulent ET12 lineage was the only clone among B . cenocepacia IIIA strains sequenced by Lee et al . [40] where katG paralog was present in genomic sequences . We speculate that the presence of katG mutations might indicate unfavorable outcome of chronic ST32 infection . This is further underlined by a recent fatality case: patient II ( Table 1 ) , upon completion of WGS and DPS analyses , underwent lung transplantation and developed CS within several months . A functional link between ROS protection and motility , another CS predictor we have reported previously to segregate between CS and non-CS patients [68] , is lacking . On a final note , we would like to emphasize that our results point to macrophages , the type of professional phagocytes which rely on both hydrogen peroxide and copper for their bactericidal activity , as putative key players behind the development of CS . B . cenocepacia has been known for its affinity to macrophages; several mechanisms were described which enable intracellular persistence [69] and the importance of macrophages in infection establishment has freshly been demonstrated [70] . Importantly , CF macrophages differ from normal macrophages by exhibiting both hyperinflammatory response to bacteria and their impaired phagocytosis and killing [71] . The observed attenuation of protective mechanisms against antimicrobial agents during within-patient evolution of ST32 is counterintuitive; in Mycobacterium tuberculosis , inactivation of katG or copper-resistance mechanisms lead to decreased virulence as a result of impaired survival of oxidative burst in macrophages [72 , 73] . We hypothesize that under increased stress encountered in CF macrophages , evolved bacteria might activate physiological processes ( e . g . global stress response , persistence ) which in turn can modulate the course of intracellular infection . The precise roles of macrophages ( at host side ) and defense mechanisms ( at pathogen side ) in chronic infection outcome warrant further investigation .
32 isolates of the B . cenocepacia epidemic clone ( CZ1 [18] , multilocus sequence type ST32 ) were collected during routine microbiological examinations of CF patients at the Centre for Cystic Fibrosis , Motol University Hospital , Prague , and kept deep-frozen . Frozen stocks were streaked and a single colony was selected and directly re-stocked; for all subsequent procedures , aliquots of these final stocks were plated and grown bacterial populations were used directly to minimize introduction of unwanted genetic variability . The complete annotated genome of ST32 isolate B . cenocepacia 1232 ( Genbank ID: GCA_001484665 . 1 ) was used as reference for all comparative analyses . Bacteria for genomic DNA preparation were harvested from an agar plate culture ( Mueller-Hinton , Oxoid ) inoculated directly from frozen stock and grown overnight at 37°C . DNA was isolated using ChargeSwitch gDNA Mini Bacteria Kit ( Invitrogen ) and quantified using Quant-iT PicoGreen dsDNA Assay Kit ( Invitrogen ) . Sequencing libraries were prepared using Nextera XT DNA Library Preparation Kit ( Illumina ) and sequenced on the MiSeq platform ( Illumina ) using MiSeq Reagent Kit v2 ( 300 cycle ) ( Illumina ) , resulting in 2 x 150 bp paired-end reads . Sequencing reads are available from Genbank ( Bioproject PRJNA397653 ) . Total DNA extracted from sputa of 12 CF patients with Amplicor Respiratory Specimen Preparation Kit ( Roche ) during periodic routine molecular microbiological examination in 2016 was used as template for PCR reactions . Q5 Hot Start High-Fidelity DNA Polymerase ( New England Biolabs ) was used to minimize amplification errors . Reaction mixtures with the Q5 High GC Enhancer were prepared according to manufacturer´s recommendations in a final volume of 50 μl , containing 1 μl template DNA and 0 . 5 μM of each primer ( S8 Table ) . PCR reactions were run for 35 cycles at annealing temperature 67°C . Aliquots of frozen stocks were plated on Mueller-Hinton agar plates ( Oxoid ) and incubated overnight . Grown bacteria were transferred into 1 ml Luria broth ( Sigma ) to obtain a suspension with OD600 approximately 0 . 05–0 . 1 . The cultures were incubated at 37°C with shaking for 2 hours to reach mid-exponential phase . Copper ( II ) chloride , hydrogen peroxide and sodium hypochlorite ( Sigma ) solutions in Luria broth were freshly prepared at 64 mM , 0 . 1% ( vol/vol ) and 0 . 5% ( w/vol ) concentrations , respectively , and sterile-filtered . The solutions were serially diluted 2-fold with sterile Luria broth , 100 μl were transferred to MIC microtiter plate wells and inoculated with 1 μl of bacterial cultures . The MIC plates were incubated aerobically for 24 hours at 37°C and MICs were recorded as the minimal concentration of antimicrobial compound which resulted in no visible growth . The experiments were repeated in three biological replicates . Sputum samples were taken with the written informed consent of the CF adult subjects . The study was approved by the Ethics Committee for Multicentric clinical trials of the University Hospital Motol , Prague on September 22 , 2010 , No . 4 . 2 . 6 . | The large Burkholderia cenocepacia populations which persist in cystic fibrosis lungs during many years of chronic infections have an inherent potential for adaptive evolution . The results provided by comparative genomics are key in understanding the processes involved . Mutational events which have taken place allow us to deductively reconstruct the history of chronic infection and to identify driving forces acting upon the bacteria . Beyond the conventional point mutation analysis of next generation sequencing data , we observed interesting phenomena such as large deletions and transposable element movement which represent another facet of adaptive evolution of B . cenocepacia during chronic infection . We also found , unexpectedly , that adaptive evolution in B . cenocepacia strain ST32 affects a set of genes conspicuously different from related species B . dolosa; these appear to be linked to host immune response . Our study provides clues to the complex puzzle of chronic B . cenocepacia infection establishment , persistence and outcome in cystic fibrosis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"genetic",
"diseases",
"microbiology",
"fibrosis",
"pulmonology",
"parallel",
"evolution",
"mutation",
"developmental",
"biology",
"cystic",
"fibrosis",
... | 2017 | What matters in chronic Burkholderia cenocepacia infection in cystic fibrosis: Insights from comparative genomics |
Mycobacterium ulcerans ( M . ulcerans ) , the causative agent of the devastating skin disease Buruli ulcer ( BU ) , is characterized by an extremely low level of genetic diversity . Recently , we have reported the first discrimination of closely related M . ulcerans variants in the BU endemic Densu River Valley of Ghana . In the study real-time PCR-based single nucleotide polymorphism ( SNP ) typing at 89 predefined loci revealed the presence of ten M . ulcerans haplotypes circulating in the BU endemic region . Here we describe the development of temperature-switch PCR ( TSP ) assays that allow distinguishing these haplotypes by conventional agarose gel-based analysis of the PCR products . After validation of the accuracy of typing results , the TSP assays were successfully established in a reference laboratory in Ghana . Development of the cost-effective and rapid TSP-based genetic fingerprinting method will thus allow investigating the spread of M . ulcerans clones by regular genetic monitoring in BU endemic countries .
Infection with M . ulcerans causes a chronic and necrotizing skin condition known as Buruli ulcer . This emerging disease occurs focally in more than 30 predominantly tropical countries worldwide , but mainly affects impoverished populations of West and Central Africa with limited access to health care services [1] . Recent findings suggest that M . ulcerans has diverged about a million years ago from the fish pathogen Mycobacterium marinum by the acquisition of a plasmid encoding the enzymes required for the production of mycolactone [2]–[4] . Mycolactone is a cytotoxic macrolide toxin that plays a key role in the unique pathology of BU [5] , characterized by the formation of progressive skin ulcers . While a potential transmission model implicating mammals as reservoirs and mosquitoes as vectors of M . ulcerans has been proposed for a local BU endemic region in south-eastern Australia [6] , [7] , epidemiologic information for BU endemic African settings is sparse . Remarkably little genetic diversity between M . ulcerans isolates from African BU patients has hindered molecular epidemiological studies tracing the spread of genetic variants of M . ulcerans . However , from a phylogenetic perspective the genetic monomorphism of this pathogen , which is associated with a clonal ancestry , holds a great potential to trace transmission pathways and evolutionary relationships . Several studies of other genetically highly homogeneous pathogens such as Mycobacterium leprae [8] , Bordetella pertussis [9] , Yersinia pestis [10] and Bacillus anthracis [11] have demonstrated the validity of genome-wide single nucleotide polymorphisms as markers for such phylogenetic analyses . In our previous work we compared genome sequences of three Ghanaian M . ulcerans isolates , selected on the basis of the three earlier identified variable number of tandem repeat ( VNTR ) types among 57 M . ulcerans strains from Ghana [12] , in order to detect a comprehensive set of SNP markers for genotyping studies . The subsequent development of 65 real-time PCR-based typing assays for the identified SNP loci allowed us to differentiate 75 M . ulcerans strains from a BU endemic area in the Densu River Valley of Ghana into six haplotypes . Genome re-sequencing of four haplotype representatives followed by further SNP detection and design of 24 additional real-time PCR assays led to the identification of ten haplotypes ( HT1–10 ) among the 75 M . ulcerans isolates ( Table 1 ) . Here we report the development of a simplified and cost-effective SNP typing method based on TSP and analysis of PCR products by conventional agarose gel electrophoresis . For this approach we have selected ten canonical SNP ( canSNP ) markers that facilitate a rapid differentiation of the ten described M . ulcerans haplotypes in the Densu River Valley of Ghana by the elimination of diagnostically redundant assays . This strategy is used to monitor the temporal as well as spatial distribution and spread of the M . ulcerans haplotypes in that region . Results of this ongoing study are expected to provide insights into the circulation of M . ulcerans variants in a BU endemic area .
M . ulcerans isolates analyzed in this study were cultivated for BU diagnosis . Ethical approval to use the isolates for immunological and microbiological research was obtained from the institutional review board of the Noguchi Memorial Institute for Medical Research , University of Ghana , Legon , Ghana ( Federal-wide Assurance number FWA00001824 ) . Written informed consent was provided by all patients involved in this study . We analyzed a total of 33 M . ulcerans isolates cultivated from wound specimen of BU patients living in the BU endemic Densu River Valley of Ghana . Ten of these isolates , used for the setup of SNP typing assays , had been typed previously by real-time PCR as SNP haplotypes 1–10 ( Agy99 ( HT1 ) , NM98/03 ( HT2 ) , NM83/03 ( HT3 ) , NM100/03 ( HT4 ) , NM27/02 ( HT5 ) , NM18/02 ( HT6 ) , NM74/03 ( HT7 ) , NM32/02 ( HT8 ) , NM28/02 ( HT9 ) , and NM78/03 ( HT10 ) ) [13] . Genomic M . ulcerans DNA was isolated by cell wall disruption and phenol-chloroform extraction as described earlier [14] . DNA was quantified by using Qubit Fluorometer ( dsDNA HS Assay Kit , Invitrogen ) . In our previous work we have detected ten M . ulcerans haplotypes ( HT1–10 ) in a relatively small BU endemic region within the Densu River Valley of Ghana by amplification refractory mutation system ( ARMS ) real-time PCR assays [13] . Here we constructed a phylogenetic tree of the ten identified haplotypes based on concatenated sequences of the 89 SNP loci ( Figure 1A ) . For the TSP assay development , we selected ten representative SNP loci ( TSP1 , 3 , 4 , 6 , 8 , 9 , 15 , 16 , 17 and 18 ) , providing a discrimination of all described M . ulcerans variants in the Densu River Valley by haplotype-specific allele combinations ( Figure 1B ) . TSP assay primers were designed on the basis of the strategy described by Hayden and Tabone [15] , [16] . For each SNP marker , locus specific ( LS ) primers amplifying the region surrounding the SNP of interest as well as a nested allele-specific ( NAS ) primer fully complementary to the sequence of reference strain Agy99 , but mismatched at the 3′end nucleotide for M . ulcerans strains harboring the SNP at this locus , were designed using Primer 3 Software [17] ( Figure 2 , Table S1 ) . LS primers were designed to have an optimum melting temperature ( Tm ) of 63°C ( range of 62–64°C ) and to amplify a PCR product greater than 400 bp . The NAS primer was designed to have a core region with an optimum Tm of 46°C ( range of 43–48°C ) and a non-complementary 5′tail region that increased the overall optimum primer Tm to 53°C ( range of 52–55°C ) . The forward ( TSP assays 1 , 3 , 4 , 6 , 8 , 9 , 15 , 16 ) or reverse ( TSP assays 17 , 18 ) NAS primer was positioned in at least 60 bp distance from the corresponding forward/reverse LS primer to ensure a clear distinction between the larger LS and the smaller NAS PCR product . Primer sequences and PCR product sizes are provided in Table S1 . The TSP method described by Hayden and Tabone [15] , [16] served as basis for the development of a TSP assay protocol for M . ulcerans . A number of technical details were modified , including standard PCR reagents , the addition of diluted DNA instead of DNA desiccation and the usage of a three-primer system with optimized primer concentrations . PCR assays were performed using 0 . 5 U FIREPol DNA Polymerase ( Solis BioDyne ) , Buffer BD , 3 mM MgCl2 , 0 . 25 mM dNTPs ( Sigma ) , 0 . 1 µM each of forward and reverse LS primer , 1 µM ( 0 . 25 µM , 0 . 5 µM , 0 . 75 µM ) of forward/reverse NAS primer and 1–50 ng genomic DNA in a total reaction volume of 5 µl . In order to avoid evaporation of the relatively small reaction volume , PCRs were carried out in 0 . 2 ml eppendorf PCR tube strips , which could individually be closed after addition of the reagents . PCRs were performed in a T-Professional thermocycler ( Biometra ) . Thermal conditions for PCR amplification of M . ulcerans genomic DNA after an initial denaturation step ( 95°C for 5 min ) were as follows: 15 cycles of 95°C for 30 s , 58°C for 30 s , 72°C for 1 min in order to enrich the LS product at a relatively high annealing temperature ( Figure 2A ) ; 5 cycles of 95°C for 10 s and 45°C for 30 s to enable a possible incorporation of the NAS primer into the enriched LS PCR product at a low annealing temperature ( Figure 2B ) ; 15 cycles of 95°C for 10 s , 53°C for 30 s and 72°C for 5 s to facilitate a competitive amplification of LS and NAS PCR products ( Figure 2C ) ; final extension step of 72°C for 10 min . 1 µl of the PCR products were analyzed on 2% agarose gels and stained for 1–2 hours in an ethidium bromide bath ( 1 µg/ml in 0 . 5xTBE ) . Whether the tested M . ulcerans strains harbored reference or SNP allele at the analyzed loci could then be determined by the presence of either the smaller or the larger PCR product , respectively . All TSP SNP typing results were validated by the recently described amplification refractory mutation system real-time PCR SNP typing technique [13] .
TSP parameters described by Tabone et al . included the desiccation of genomic DNA by evaporation prior to PCR amplification [16] . In order to reduce the risk of contamination of the laboratory with DNA template , we eliminated this step . The addition of M . ulcerans DNA dissolved in water necessitated a new setup for all assay parameters . While standard PCR reagents were used according to the manufacturer's ( Solis Biodyne ) recommendations , the most critical step for a specific detection of either the smaller NAS or the larger LS PCR product was to identify the optimal application of NAS and LS primers used for the PCR assays . Since initial PCR reactions using a four-primer system including both forward and reverse NAS and LS primers led to the amplification of additional non-specific PCR products ( Figure 3A ) , we applied a three-primer system with only one NAS primer . In order to ensure an accurate differentiation of reference and SNP alleles we determined optimal primer concentrations by performing TSP assays for the ten haplotype representatives at NAS∶LS primer ratios of 2 . 5∶1 , 5∶1 , 7 . 5∶1 and 10∶1 . Results of all TSP assays provided a correct differentiation of reference and SNP alleles for the four NAS primer concentrations ( Figure 3B–E ) . However , a clear-cut visualization of only the allele-specific PCR product was obtained by deploying a tenfold concentration of the NAS primer compared to the LS primers ( Figure 3E ) . The illustration of allele-specific PCR products could be further improved by loading only one-fifth of the PCR reaction on a 2% agarose gel . Gels initially overloaded by the whole PCR reaction volume showed traces of the non allele-specific PCR products . The optimal amount of DNA for the TSP assay reactions was assessed by performing TSP assays for the ten haplotype representatives with a NAS∶LS-primer ratio of 10∶1 and different genomic DNA concentrations ranging from 1–50 ng . TSP assays provided accurate results for all DNA concentrations tested . However , subsequent standardized TSP strain typing was performed using 5 ng DNA for each reaction . The performance of TSP assays was dependent on the purity of the extracted DNA . While samples with a high ratio of OD 260/280 ( >1 . 7 ) facilitated a clear visualization of allele-specific PCR products on the agarose gels , traces of non-allele-specific PCR products were detected when using DNA extracts with poorer quality . TSP assays were evaluated on ten haplotype representatives in order to confirm accurate performance of all assays for both reference and SNP alleles ( Figure S1 ) . We typed a total of 23 M . ulcerans strains isolated between 2009 and 2011 from BU patients living in the Densu River Valley of Ghana by the ten developed TSP assays using the optimized single standard condition ( Figure S2 ) . One additional strain served as a control for either the NAS or the LS PCR product amplification . Based on the resulting allele combinations at tested SNP loci , isolates could be differentiated into seven of the ten described M . ulcerans haplotypes ( HT1 ( 1 strain ) , HT2 ( 7 strains ) , HT3 ( 1 strain ) , HT5 ( 3 strains ) , HT6 ( 5 strains ) , HT7 ( 1 strain ) and HT9 ( 5 strains ) . ARMS real-time PCR SNP typing of the 23 isolates at the ten SNP loci confirmed the accuracy of TSP SNP typing results . In addition to clinical M . ulcerans strain typing , TSP assays were performed using DNAs directly extracted from BU lesion specimens . We selected four samples out of a panel of DNA extracts with the highest M . ulcerans DNA concentrations detected by IS2404 real-time PCR . However , all attempts to SNP-type these samples by TSP failed , since additional , unspecific PCR products were amplified . Robustness of the assay protocol and suitability of the technique for technology transfer to laboratories in BU endemic countries was verified by analyzing TSP assays in a laboratory of the Noguchi Memorial Institute in Accra , Ghana . For this purpose we tested the ten haplotype representatives again using the same PCR materials and optimized assay parameters . Each assay provided the expected TSP genotyping products and endpoint detection by 2% agarose gels and ethidium bromide staining was comparable to the TSP setup results ( Figure 4 ) .
A global overview of genetic diversity in M . ulcerans could be established by comparative genomic hybridization analysis detecting insertional-deletional variation in 35 clinical M . ulcerans isolates of world-wide origin [18] , [19] . However , efforts made to resolve the population structure of M . ulcerans strains from the same BU endemic countries were limited by the extensive genetic monomorphism of closely related isolates [12] , [20]–[25] . Genome sequencing of several clinical M . ulcerans isolates from a BU endemic region in the Densu River Valley of Ghana has facilitated for the first time the development of a SNP-based genotyping strategy with sufficient resolution for micro-epidemiological studies . This real-time PCR-based approach provided first insights into the actual diversity of such closely related M . ulcerans isolates as well as the distribution of M . ulcerans variants in the BU endemic area . Among 75 M . ulcerans strains isolated between 1999 and 2007 ten haplotypes showing different patterns of geographical distribution could be detected . The identification of geographically clustered emerging haplotypes provides a valuable opportunity to monitor the spatio-temporal spread of haplotypes in this region . Cost is an important issue for all genetic fingerprinting analyses , and this particularly applies to neglected pathogens like M . ulcerans . While the established real-time PCR technique required relatively expensive equipment , the TSP-based assays described here , provide a simple and cheap alternative system , thus increasing the access of genetic typing assays for M . ulcerans to laboratories in tropical BU endemic areas . Low cost for the PCR assays , direct endpoint detection through PCR product staining on agarose gels and simple implementation of the method provided a practical means for rapid and cost-effective SNP typing . The critical step for TSP assay development was to design primers with optimal melting temperatures for the envisaged reaction phase participations . Efficient primer design enabled the performance of all TSP assays under a single optimized standard condition . Genotyping accuracy and robustness of the TSP strategy was demonstrated by the successful application of the designed assays in two different research laboratories , yielding identical results . SNPs represent highly stable phylogenetic markers in genetically homogenous pathogens as they occur at very low rates and convergent or reverse mutations are highly unlikely [26] . Numerous genotyping studies of monomorphic pathogens have demonstrated that a limited set of SNPs can be used to define phylogenetic relationships [11] , [27]–[30] . The idea of a canonical SNP , a SNP that can be used to define species [31] , major genetic lineages of a species [27] , [28] or even specific strains [32] , [33] has recently been described in a review article focussing on the molecular epidemiology of Bacillus anthracis [26] . Here we present an extreme example of the canSNP concept where a small number of SNPs is used to distinguish between strains from a relatively small endemic region . For that purpose we replaced the 89 real-time PCR-based SNP typing assays by ten strategically placed canSNPs that enabled a differentiation of the ten described M . ulcerans haplotypes in our panel of strains . The detected canSNPs provide useful genetic markers for the geographical and temporal analysis of M . ulcerans variants in the Densu River Valley of Ghana and the established TSP assays are now routinely used in the laboratories of the Noguchi Memorial Institute for M . ulcerans haplotype identification . However , a comprehensive overview of the M . ulcerans population structure in Africa can only be achieved by genome sequencing of M . ulcerans strains from other African BU endemic regions and subsequent identification of local sets of informative SNPs for further phylogenetic analyses . | As the third most common mycobacterial disease after tuberculosis and leprosy , Buruli ulcer constitutes a considerable health problem in many parts of the world , but in particular in West and Central Africa . Although this emerging skin disease is commonly associated with proximity to aquatic habitats , the mode of transmission remains obscure . While the clonal population structure of Mycobacterium ulcerans provides a great potential to trace transmission pathways and evolutionary relationships , micro-epidemiological studies have long been hampered by a striking lack of genetic diversity among African M . ulcerans populations . Whole genome comparison of M . ulcerans strains isolated from patients living in the BU endemic Densu River Valley of Ghana led to the identification of single nucleotide polymorphisms between these closely related strains . The subsequent development of SNP typing assays enabled a differentiation of ten M . ulcerans haplotypes in the BU endemic region . Here we selected canonical SNP markers for a spatio-temporal analysis of M . ulcerans variants in the Densu River Valley of Ghana using a temperature-switch PCR-based approach . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"emerging",
"infectious",
"diseases",
"biology",
"genomics",
"microbiology",
"genetics",
"and",
"genomics"
] | 2012 | Development of a Temperature-Switch PCR-Based SNP Typing Method for Mycobacterium ulcerans |
The rapid evolution of Human Immunodeficiency Virus ( HIV-1 ) allows studies of ongoing host–pathogen interactions . One key selective host factor is APOBEC3G ( hA3G ) that can cause extensive and inactivating Guanosine-to-Adenosine ( G-to-A ) mutation on HIV plus-strand DNA ( termed hypermutation ) . HIV can inhibit this innate anti-viral defense through binding of the viral protein Vif to hA3G , but binding efficiency varies and hypermutation frequencies fluctuate in patients . A pivotal question is whether hA3G-induced G-to-A mutation is always lethal to the virus or if it may occur at sub-lethal frequencies that could increase viral diversification . We show in vitro that limiting-levels of hA3G-activity ( i . e . when only a single hA3G-unit is likely to act on HIV ) produce hypermutation frequencies similar to those in patients and demonstrate in silico that potentially non-lethal G-to-A mutation rates are ∼10-fold lower than the lowest observed hypermutation levels in vitro and in vivo . Our results suggest that even a single incorporated hA3G-unit is likely to cause extensive and inactivating levels of HIV hypermutation and that hypermutation therefore is typically a discrete “all or nothing” phenomenon . Thus , therapeutic measures that inhibit the interaction between Vif and hA3G will likely not increase virus diversification but expand the fraction of hypermutated proviruses within the infected host .
The HIV-1 population within an infected individual is characterized by extensive viral variation and continuous adaptation to its host . Such rapid evolution is the result of a combination of several factors: a large viral population , high replication and mutation rates , recombination , and various intra-host selective pressures [1] . The high mutation rate is associated with the inherent infidelity of HIV reverse transcriptase ( RT ) and RNA polymerase II ( RNA pol II ) [1] and has also been proposed to be partly caused by cellular cytidine deaminases such as hA3G , which can cause Guanosine-to-Adenosine ( G-to-A ) mutations on HIV plus-strand DNA [2]–[7] . Several observations appear to provide support for this hypothesis as lentiviral genomes are adenine rich [8] , [9] and G-to-A is the most frequent nucleotide mutation observed during HIV-1 replication both in vitro [10] , [11] and in vivo in both acute [12] and chronic infection [13] . In infected cells , hA3G can become incorporated into nascent virions as large , enzymatically inactive , ribonucleoprotein complexes termed ‘Intra-Virion A3G Complexes’ ( IVAC ) [14] . When a virion subsequently infects another cell , IVACs become active through the activity of viral RNaseH during reverse transcription [14] and hA3G restricts HIV replication through a combination of mutagenesis ( or editing ) [5] , [15] and possibly non-editing activities [16] . Editing is easily recognized because it results in extensive Cytidine-to-Uridine ( C-to-U ) deamination of single-stranded minus-strand DNA during reverse transcription [5] , [17] , [18] . The mutations appear as plus-strand G-to-A changes and hA3-induced mutations are usually reported as such and termed hypermutation [19] as G-to-A transitions far exceed all other mutations . As the preferred target is TGG ( encoding Tryptophan when in frame ) , many G-to-A mutations will produce stop-codons , TAG , resulting in viral inactivation [17] , [20] . The HIV accessory protein Vif can circumvent the protective role of hA3G , and other hA3 deaminases , by targeting them for proteasomal degradation and thereby preventing their incorporation into virions [21] . However , as various frequencies of hypermutated sequences are observed in HIV DNA from infected patients , the efficiency of these Vif-hA3 interactions must vary between them [4] , [22]–[24] . Two different scenarios could account for the in vivo variation in hypermutation frequency . First , editing could act to increase viral diversification , with possible advantages to the virus in a fluctuating fitness environment , but to do so , hA3G would have to induce mutations at a low , sub-lethal level . In such a situation , selection would act on Vif to moderate the number of hA3G molecules incorporated into virions . Alternatively , inefficient Vif-hA3G interactions could be the by-product of other hitherto undefined selective pressures and the resulting hypermutation considered a viral fitness cost , acting at the level of the viral population . Here , we investigate the fundamental question of whether hA3G-induced G-to-A mutation is always lethal to the virus or if it may occur at sub-lethal frequencies .
To examine whether limiting-levels of hA3G activity could result in sub-lethal mutation rates in HIV infections , we designed an in vitro hA3G titration and sequencing experiment ( Table 1 ) . Briefly , we made Vesicular Stomatitis Virus G protein ( VSV-G ) pseudotyped Δvif-HIV ( IIIB ) virions , which incorporated variable amounts of editing wild-type hA3G ( wt-hA3G ) . The total hA3G concentration was kept constant using the E259Q non-editing hA3G mutant ( E259Q-hA3G ) [25] . These viruses were used to infect TZM-bl cells ( a HeLa cell line expressing HIV coreceptors and a lacZ reporter gene under the control of an HIV LTR ) in a single-cycle infection assay from which DNA was extracted and provirus amplified using limiting-dilution nested-PCR . We examined total hA3G expression in both producer cell lysates ( Figure 1A ) and purified virions ( Figure 1B ) for each titration to test that transfections of both editing and non-editing hA3G were equally efficient . Viruses with hA3G ( wt- or E259Q-hA3G ) displayed large reductions in infectivity compared to virus generated without hA3G , and the presence of increasing concentrations of wt-hA3G conferred relatively greater losses of infectivity , in line with previous studies ( Figure 1C ) [26] . We amplified and sequenced 8–20 env-to-3′LTR fragments ( 2 . 1 kb ) from each hA3G titration . As the sequence of the parental HIV ( IIIB ) virus is known ( Figure S1 ) , and the infections in our experiments restricted to a single replication cycle , we could readily identify all mutations induced by hA3G using HYPERMUT ( www . hiv . lanl . gov ) . We found that 33/87 sequences had no plus-strand G-to-A mutations while 48/87 were hypermutated carrying greater than 4% GG-to-AG mutations ( Figure 1D , Table S1 ) . In the remaining six sequences , a single G-to-A change was found in either non-hA3G ( 5/6 ) or rare ( 1/6 ) hA3G contexts ( as defined in [20] ) , suggesting that RT/RNA pol II or PCR-related errors may have been responsible . Of the hypermutated sequences , all but one ( 47/48 ) carried stop codons , and as the sequenced region corresponds to only ∼20% of the protein-coding genome , stop-codons likely exist in the rest of the genome . Hypermutation levels in the lower three wt-hA3G titrations were significantly lower than those in the higher titrations ( p<0 . 0001 , unpaired t-test ) ( Figure 1D ) . To evaluate whether these in vitro hypermutation rates were representative of those occurring in vivo , we estimated the mutation levels of 39 near-full length hypermutated patient-derived proviruses ( www . hiv . lanl . gov ) . As the parental viral sequences were unknown , we made optimized reference sequences as in [20] . Briefly , reference sequences were estimated as the consensus of closely related sequences identified by NJ phylogenetic tree analysis of HIV subtype alignments in which potential hA3-type hypermutation sites were ‘repaired’ ( i . e . all AG and AA sites were changed to NG and NA , respectively , if a GG or a GA was also present at the same position in the alignment ) . We found that the hypermutation levels observed in vivo were similar to those observed in vitro ( Figure 1D ) . Due to the lack of original patient-derived non-hypermutated reference sequences , we were unable to distinguish whether GG-to-AG mutation levels at <5% of all GG targets in these sequences were caused by hA3G or RT/RNA pol II; however , an abundance of sequences with such low hypermutation levels would imply a bimodal distribution of mutation levels in natural infections , which would be inconsistent with the in vitro data . For hA3G editing to contribute to viral adaptation , the induced mutations would need to occur at low , sub-lethal levels . This is most likely to happen if just a single editing hA3G-unit is incorporated into the virion . As hA3G may undergo RNA-dependent oligomerization during virion assembly , the term hA3G-unit is used here to refer to the active hA3G deaminase [27] . We cannot know for certain whether the hypermutants we observed in vitro did result from the incorporation of a single editing unit , but conditional on assumptions about the incorporation process , we can estimate the probability that this was so . We examined the maximum number of hA3G units that could reside in a virion by considering the proportion of sequences carrying hypermutation at each titration to derive a maximum likelihood estimate ( MLE ) of the number of editing hA3G-units per virion ( Figure 1E , Figure S1 ) . As the estimate depends on hypermutation being observed , only the number of incorporated hA3G-units with editing activity is estimated . Our analysis assumed ( i ) that there is a finite number of positions in a virion that can be occupied by hA3G-editing units [28]; ( ii ) that the efficiency of transfection , protein expression , and virion incorporation is the same for editing and non-editing hA3G ( as supported by Figure 1A and [28]; ( iii ) that there was sufficient hA3G present in each titration for all positions to be occupied by either editing or non-editing hA3G ( as supported by the 100% detection rate when 100% wild-type hA3G was present ( Table S1 ) ( iv ) that hA3G editing , when it had occurred in a sampled sequence , was always successfully detected; and ( v ) that degradation of uracil-containing edited viral DNA by cellular uracil DNA glycosidases such as UNG2 and/or SMUG1 was insignificant [3] , [29]–[31] . Under these assumptions , the probability that an observed hypermutant resulted from a single wild-type hA3G unit is approximately 1− ( k−1 ) r/2 , where k is the maximum possible number of hA3G units that can be incorporated into a single virion , and r is the proportion of hA3G present that was wild-type when the hypermutant was generated ( see Materials and Methods for full details and Table 2 ) . The probability that we have observed the minimal level of hA3G-induced hypermutation therefore depends on the number of available positions , denoted k . Using assumptions ( i ) – ( v ) listed above , we were able to derive a maximum likelihood estimator of k that could be applied to the results of our titration experiments ( see Materials and Methods and Figure 1E ) . In this way , we estimated that a virion could accommodate k = 13 editing hA3G-units ( 95% CI: 6–26 units ) – an estimate that was robust to the removal of each titration condition in turn ( Figure 1E , Figure 2 ) . This estimate was similar to a previous biochemical estimate of 7+/−4 molecules [28] . This estimate implies that in our transfection condition 2 , in which 1% of the hA3G was wild-type ( r = 0 . 01 ) , an expected 1− ( 13−1 ) 0 . 01/2 = 94% of hypermutants are predicted to have resulted from the incorporation of a single virion . This figure rises to 97% if we take the previous biochemical estimates of k ( k = 7 molecules; [28] ) , and remains as high as 87 . 5% if we take our upper confidence interval ( k = 26 ) . Based on this analysis , it follows that our lower editing hA3G titrations ( with low r values ) are highly likely to have recorded hypermutation occurring at the lowest possible level . To further assess the effects of hypermutation occuring in this way , and to ensure that the hypermutation levels were not specific to the env-3′LTR region , we analyzed several near-full length proviral sequences from these lower editing hA3G titrations ( Figure 3 ) . In each case , 9–18% of all GG-motifs were mutated to AG , with mutation occurring either side of each polypyrine tract , suggesting that single editing hA3G-units can be active throughout the genome . It has been hypothesized that editing rates are highest in the regions most distal to the polypurine tracts , which are exposed as a single-stranded DNA substrate for the longest times forming a “twin gradient” of mutational burden across the genome [17] , [32] , [33] . Our previous study of in vitro and in vivo hypermutated sequences demonstrated that reduced levels of editing immediately downstream of the polypurine tracts were a common feature of hA3G editing although hypermutation gradients were not always evident [20] , in agreement with the single editing hA3G-unit data in Figure 3 . Transient transfections of hA3G in vitro have shown hA3G incorporation into IVACs , but have also demonstrated that overexpressed hA3G may become packaged external to the virion core [14] . However , as only IVAC-associated hA3G has been suggested to edit nascent viral DNA [14] , our estimate ( based on the proportion of edited sequences ) would be expected to just reflect the number of IVAC-incorporated hA3G molecules , regardless of potential hA3G overexpression . In the case non-IVAC associated hA3G contributed to editing in this experiment , even fewer hA3G-units would likely be incorporated in natural infection , underscoring that extensive hypermutation can be induced by a single or very few hA3G units . Together , these results suggest that even a single incorporated hA3G-unit is likely to cause extensive and inactivating levels of HIV hypermutation , and that therefore , hypermutation is typically a discrete “all or nothing” phenomenon . If hA3G-induced G-to-A mutations were to increase viral diversification [2]–[7] , they would have to be generated at a low , sub-lethal level ( Figure 4A ) . To determine how low this level should be to permit neutral or potentially beneficial mutations while avoiding lethal mutations ( i . e . stop codons ) , we determined hA3G tetranucleotide target preferences [20] and simulated editing in silico ( Figure 4B , 4C ) . A previous simulation study [34] assumed , in effect , that hA3G induced a single mutation per round of replication but this is in conflict with functional studies demonstrating that hA3G moves along its single stranded DNA template while inducing multiple mutations [35] . Accordingly , we simulated here the effects of increasing hA3G-mediated mutation rates on individual viruses . We assumed that all stop codons within HIV genes would result in non-functional virus and used the HIV ( IIIB ) open reading frames ( Figure S1 and Figure 4 ) to estimate the rate at which a lethal mutation was induced in 50% of viral offspring ( lethal mutation 50% - LM50 ) using three different nucleotide targets: G-to-A , GG-to-AG , and predefined hA3G-specific nGGn-to-nAGn tetranucleotide contexts [20] . The estimated LM50 rates depended strongly on nucleotide target specificity . Considering all G-to-A targets ( assuming that hA3G recognized all Gn dinucleotide targets equally ) an average of 9 targets would have to be mutated to give a 50% chance of at least one lethal mutation . However , if hA3G specificity was considered using its preferred dinucleotide GG , only 3 . 8 out of 667 GG targets would need to be mutated to yield a 50% chance of at least one lethal mutation . Furthermore , if specific hA3G tetranucleotide target preferences ( nGGn-to-nAGn ) [20] were used in the simulations , we estimate an LM50 of only 2 . 5 mutations , implying that the innate anti-viral hA3G protein generate stop codons very efficiently ( Figure 4B ) . At an nGGn-to-nAGn rate of 2 . 5% per context ( equivalent to only 11 mutations per genome ) , stop codons were induced in 99% of simulations ( LM99 ) . These estimates are highly conservative as they ignore the likely harmful effects of most non-synonymous ( NS , amino acid changing ) mutations and possible negative effects of synonymous ( S ) changes on RNA secondary structure [36]–[38] . Both NS and S mutations are more frequent than stop codons ( e . g . at the nGGn-to-nAGn LM50 rate , >80% of the simulations also had at least one NS mutation ( Figure 4C ) , and about 60% had multiple ) . At these rates , only a few hA3G-induced mutations are needed to inactivate progeny viruses and considering the hypermutation rates observed in vitro , we found that the lowest hypermutation frequency detected was ∼10 fold higher than the estimated LM50 rate and over double the estimated LM99 rate ( Figure 1D ) . Collectively , our results suggest that even a single virion-incorporated hA3G-unit rarely , if ever , generate G-to-A mutations at sub-lethal levels but is very likely to cause extensive and inactivating levels of HIV hypermutation .
Here we investigate the pivotal question of whether hA3G-induced G-to-A mutation is always lethal to the virus or if it may occur at sub-lethal frequencies . We examined whether limiting-levels of hA3G activity could result in sub-lethal mutation rates using an in vitro hA3G titration and sequencing experiment . The resulting in vitro mutation patterns and per replication cycle rates were similar to mutation levels found in in vivo hypermutated HIV DNA sequences implying that our experimental data reflected natural infection [20] . Second , based on the proportions of sequences carrying hypermutation in these datasets , we estimated that the maximum number of editing hA3G molecules packaged in a virion was 13 ( 95% CI , 6–26 ) , which was only slightly higher than a previous biochemical estimate of 7+/−4 molecules [28] . Using our estimate , we calculated that it was highly likely that the hypermutants we observed at the lowest wt-hA3G concentrations in the titration experiments ( Table 1 ) were caused by the incorporation of just a single hA3G-unit , and this becomes even more likely if the lower biochemical estimate is correct . As the editing observed was extensive and induced inactivating levels of G-to-A mutations , hypermutation typically seems to be an “all or nothing” phenomenon . It has been hypothesized that a proportion of hypermutated sequences might be degraded by the cellular uracil DNA glycosylases UNG2 and/or SMUG1 and that this may contribute to the antiviral effect of hA3G [39] , [40] . This hypothesis is however controversial [3] as few studies support it [41] while several have demonstrated that the absence or inhibition of UNG2 and/or SMUG1 activity neither abrogates hA3G inhibition of infection nor rescues viral cDNA accumulation in infected cells , suggesting that these enzymes are not involved in hA3G restriction of viral replication [3] , [29]–[31] . Without conclusive data demonstrating UNG-mediated degradation , it is impossible to model in a realistic manner . However , we estimate that UNG-mediated degradation , if it destroyed a large proportion of the hypermutated sequences , would increase our estimation of k ( the number of hA3G units in a virion ) . This would however not impact on our analyses of the role of hA3G in viral evolution in vivo as sequences that are degraded disappear and do not form part of the viral population . Third , we simulated editing in silico taking viral reading frames into account , to determine how low levels of hA3G-induced G-to-A mutations should be to increase viral diversification through neutral or potentially beneficial mutations while avoiding induction of lethal mutations ( i . e . stop codons ) . We found that due to hA3G tetranucleotide target preferences , which render it efficient at generating stop codons , only a few mutations were generally needed to inactivate progeny viruses . When we compared the estimated LM50 rate with in vitro hypermutation rates , we found that it was ∼10 fold less than the very lowest hypermutation frequency , suggesting that even a single hA3G-unit rarely , if ever , causes G-to-A mutations at potentially beneficial low levels . Examining the role of hA3G in HIV evolution is an area of active research . In vitro studies have used reporter-genes to extrapolate the effect hA3G editing on HIV diversification [42] and the nucleoside analog RT inhibitor 2′ , 3′-dideoxy-3′-thia-cytidine ( 3TC or Lamivudine ) to assess the effect of hA3G on the appearance of drug resistance mutations in lab-adapted HIV [43] . Population sequencing , which only detects polymorphisms present in >20–25% of the viral population [44]–[47] , was used to identify drug-resistance mutations and as Lamivudine accumulates to different degrees in different cell lines [48]–[51] and increases intracellular dATP levels [52] , which may affect RT misincorporation [13] , the relevance of these studies for HIV evolution in natural infection needs further examination . Studies of patient-derived HIV sequences either directly support our finding that hA3G is unlikely to contribute to viral diversification [13] , [53] or does not contrast it [54] , [55] . One report found that about 25% of rapidly diversifying sites in HIV were in sequence motifs that could be mutated by either hA3C , hA3F , hA3G or RT [54] . Another study indicated that RT misincorporation was affected by imbalances in dNTP pools , which could explain the observed bias of G-to-A mutations in HIV evolution , and found no sign of hA3F/G editing [13] . A third study of plasma virus sequences from HIV-1 infected patients that were either drug-naïve or had failed HAART demonstrated that Vif was highly polymorphic in both groups , but more so in pretreated patients [55] . One of the Vif substitutions ( K22H ) was further analyzed as another substitution ( K22E ) had previously been demonstrated to partially neutralize hA3F but not hA3G [4] . K22H was shown to partially neutralize hA3G whilst the effect on hA3F was not tested . In vitro culture of mutated virus in MT2 cells that express high levels of hA3F and hA3G [56] resulted in a minority of the sequences carrying sub-lethal mutations , which could be caused by either hA3F or hA3G . In contrast to hA3G , hA3C and hA3F are likely to sometimes induce sub-lethal G-to-A mutations as hA3F neutralization is dispensable for spread of HIV-1 in primary lymphocytes [56] and hA3C neutralization is not needed for viral spread in SupT1 cells , which does not express hA3F and hA3G [57] . A fourth cross-sectional study of patient-derived sequences found no evidence of an evolutionary footprint of hA3F/G [53] and studies of thousands of patient-derived sequences have found either no , or very few , hypermutated RNA sequences , suggesting that low-level hypermutation , or recombination between hypermutated and non-hypermutated viruses , very rarely occurs in vivo [12] , [23] . Such a recombination has been found only once in vitro after co-transfection of 32 hypermutated and non-hypermutated proviruses and 3TC drug selection [58] . As hA3G activity has such detrimental effects on HIV , strong viral selective pressures must act to optimize Vif's interaction with hA3G . However , as variable levels of hypermutation are observed in many HIV infected patients , other selective pressures may sometimes also affect vif evolution . Several studies have demonstrated that CD8+ cytotoxic T-cells ( CTL ) can target Vif [59]–[65] and we hypothesize that these CTL responses sometimes select for Vif variants that by chance interact less efficiently with hA3G . As hypermutation frequency has been found to correlate inversely with plasma viremia in three large patient cohorts [66]–[68] , but not in two smaller cohorts [69] , [70] , increasing hypermutation frequencies in patients through therapeutic measures is potentially beneficial . In conclusion , our study suggests that hA3G activity is unlikely to increase HIV evolution and that hA3G-activity is highly likely to inactivate HIV-1 .
pcDNA3 . 1 expression vectors with wild-type hA3G ( wt-A3G ) or non-editing E259Q mutant hA3G ( E259Q-hA3G ) , VSV-G and the vif-deficient HIV-1 ( IIIB ) ( pIIIB/Δvif ) proviral construct have been described previously [21] , [25] , [71]–[73] . Vif-deficiency was caused by the introduction of two nonsense mutations while all other accessory genes were functional . pIIIB/Δvif was furthermore modified with a G-to-A mutation at position 571 of the 5′LTR U5 region , which copies to the 3′LTR during reverse transcription , enabling discrimination of viral sequences that have passed through a replication cycle from those derived from the residual transfection cocktail . VSV-G pseudotyped Δvif-HIV-1 was produced by transfection of subconfluent monolayers of 293T cells using polyethylenimine ( PEI ) ( Polyscience ) . as in [25] . The transfection efficiency of PEI is reported to be over 98% [74] and the average number of transfected plasmids per cell using similar plasmid concentrations and cell numbers is about 105 plasmid molecules [75] . The pIIIB/Δvif construct , VSV-G , and varied ratios of wt-hA3G to E259Q-mutant hA3G were used ( summarized in Table 1 ) . Media were changed after 6 h and supernatants were harvested after 24 hr ( hA3G titration experiment ) or 48 hr ( patient-derived Vif experiment ) ; virus production was quantified by p24 Gag ELISA ( Perkin Elmer ) , prior to storage at −80°C and use in subsequent experiments . For preparation of purified HIV-1 virion associated proteins , virus supernatant equivalent to 30 ng of p24 Gag was diluted in media , and underlain with 20% sucrose solution . Samples were centrifuged for 2 hours at 14000 rpm at 4C° and supernatants removed . Purified virions or infected 293T cells were lysed , centrifuged to remove cell debris , and prepared for loading onto SDS-PAGE gels in a 1∶1∶1 mix of 3× SDS-PAGE sample buffer ( 180 mM Tris , pH 6 . 8; 9% ( w/v ) SDS; 30% glycerol; bromophenol blue ) , DTT ( in PBS , giving a final concentration of 100 mM ) and lysate , and were incubated for 10 minutes at 95°C . 5–10 µl of samples were loaded into a 4% stacking gel on a 12% separating gel and run for 1 hr at 25 mA/gel at maximum voltage . Proteins were transferred from gels onto PVDF membranes ( pre-soaked in methanol and running buffer ( WB: 0 . 1% Tween20 in PBS ) ) at 16 V overnight; membranes were blocked in 5% milk powder in WB for at least 30 minutes , prior to incubation with primary antibody ( either anti-hA3G ( recognizing both wt- and E259Q-hA3G ) or anti-p24CA ( loading control ) diluted in 5% milk powder/WB ) for 1 hr at room temperature . After rinsing 3 times and washing 4 times for 5′ with WB , membranes were incubated with horseradish-peroxidase conjugated secondary antibody ( diluted in 5% milk powder/WB ) for 40′ at room temperature , and the rinse/wash procedure was repeated . Membranes were then incubated for 1–5′ with ECL substrate before exposure to film as in [25] . TZM-bl cells ( a HeLa cell line expressing HIV-1 co-receptors and a lacZ reporter gene under control of an HIV-1 LTR promoter ) were infected with 293T cell produced VSV-G-pseudotyped Δvif-HIV-1 virions containing various ratios of wt-hA3G to E259Q-mutant hA3G ( hA3G titration experiment ) or Δvif-HIV-1 virions containing hA3G and patient derived Vif ( patient-derived Vif experiment ) . After 24 hrs , supernatants were removed and cells were washed with PBS , before lysing with 200 µl lysis solution . Following transfer to microfuge tubes , debris from cell lysates was pelleted by microcentrifugation at 14 , 000 rpm for 10 minutes and 20 µl cell extract was then added to 100 µl Galacton-Star ( reporter gene assay system for mammalian cells ) substrate ( Applied Biosystems Inc . , CA , USA ) diluted 1∶50 with reaction buffer diluent in white microplate wells . The light signal was measured every 10–15 minutes up to 2 hr after the start of the reaction on a luminometer , giving a read-out of β-galactosidase production , which is proportional to the infectivity of the infecting virus . For sequencing experiments , total DNA was extracted from infected cells using the DNeasy DNA extraction kit ( Qiagen Inc , CA , USA ) and digested with DpnI ( New England Biolabs ) , a restriction endonuclease that specifically targets methylated DNA , to remove carried-over transfection mixture . Near-full length proviral single genomes were amplified by limiting dilution nested PCR using Advantage 2 Polymerase mix ( TakaraBio/Clontech , Paris , France ) and HIV-1 specific oligonucleotide primers , as described previously [20] . The product of an 8 . 5 kb first-round PCR from gag-to-3′LTR was used as a template for a second-round PCR spanning env-to-3′LTR ( 2 . 1 kb , 8–20 fragments per hA3G transfection condition , 87 amplicons in total ) ( Figure S2 ) . For a subset of sequences , gag-to-pol , pol-to-vif , and vif-to-env fragments were amplified to derive near-full length sequences . Where possible , primers ( Table S2 ) were designed to exclude 5′GG or 5′GA ( plus-strand ) or 5′CC or 5′TC ( minus-strand ) motifs , the preferred contexts for hA3F and hA3G activity respectively , in order to reduce the potential for bias in amplification of hypermutated viruses . Amplicons were purified using the QIAquick PCR purification kit ( Qiagen Incorporated , CA , USA ) and both strands were sequenced directly using Dyedeoxy Terminator sequencing ( Applied Biosystems , CA , USA ) on an Applied Biosystems 3730xl DNA Analyzer as previously described [20] . DNA reads were assembled and proofread using the Pregap4 and Gap4 software within the Staden package [76] ( Figure S2 ) . Sequences lacking the engineered G-to-A mutation in the 3′LTR [72] were assumed to be carried-over transfection mixture and were discarded . Sequences were screened for evidence of hA3G-mediated editing/hypermutation ( defined as a mutational process in which G-to-A transitions far exceed all other mutations [19] ) using the HYPERMUT software ( www . hiv . lanl . gov ) [77] . The proportion of sequences carrying evidence of hypermutation at each titration was used to generate a MLE of the average number of deaminating hA3G units incorporated into a progeny virion . Our analysis assumes ( i ) that there are a limited number of positions in a virion that can be occupied by hA3G-editing units [28] . The number of such positions , denoted k , is unknown , but we can use our titrations to obtain a maximum likelihood estimate of its value . Let us denote as ri , the proportion of the hA3G in transfection condition i that is wild-type editing ( wt-hA3G ) , as opposed to non-editing ( E259Q-hA3G ) ; for example , from Table 1 and Table 2 , in condition 5 , r5 = 0 . 33 . How likely is a virion to incorporate an editing hA3G-unit under this condition ? To answer this question , we assume ( ii ) that the efficiency of transfection , protein expression , and virion incorporation is the same for editing and non-editing hA3G; and ( iii ) that there is sufficient hA3G present in each titration for all k slots to be occupied . Under these three assumptions , the probability that a virion incorporates one or more editing hA3G-units is simply qi = 1− ( 1−ri ) k . If we further assume ( iv ) that detectable hypermutation always ensues from the incorporation of one or more editing hA3G-units , then qi is also the probability that a sequence undergoes hypermutation . As such , given a sample of ni sequences , the probability that hi of them will be hypermutants is the binomial probability: . Because qi is a function of k we can now write the likelihood function of k as , and thereby obtain the value of k that is most likely to have given rise to our data , i . e . , the value that maximises L ( k ) . 95% confidence intervals on this estimate were obtained by assuming that twice the log likelihood ratio is χ2 -distributed with 1 degree of freedom , and the sensitivity of the analysis to each individual condition was assessed by jackknifing , i . e . , reestimating k after removing each condition in turn . Results of the analyses are shown in Figure 1E and Figure 2 , and the values of ri , ni , and hi are shown in Table 2 . The approximation for the probability that an observed hypermutated sequence has arisen from the incorporation of a single hA3G-editing unit is obtained from: The thirty-eight near-full length HIV genomes annotated as hypermutated in the Los Alamos HIV sequence database ( www . hiv . lanl . gov ) at the time of this analysis and one non-annotated hypermutated sequence ( EF036536 ) were used to estimate levels of hypermutation in HIV DNA ( Table S3 ) . EF036536 was identified by examining GenBank entries of 1725 near-full length HIV genomes . The sequences were tested by the search terms ‘stop’ , ‘truncated’ , ‘truncation’ , ‘terminated’ , ‘termination’ , ‘mutated’ , ‘mutation’ , ‘hypermutated’ , ‘hypermutation’ , ‘non-functional’ , and ‘nonfunctional’ , and those carrying more than 4 stop codons were tested for evidence of hA3G-induced mutations as previously described [20]; analyses of sequences with fewer mutations was not possible due to noise . GG-to-AG mutation rates were estimated for each of these 39 hypermutated sequences using reference sequences generated from closely related taxa identified by NJ phylogenetic tree analyses as described previously [20] . GG-to-AG mutation rates were corrected for probable non-hA3G-mediated mutation by subtracting the mean of the GC-to-AC and GT-to-AT mutation rates in each sample from the GG-to-AG mutation rate , after adjusting for the biased nucleotide composition of the HIV genome in each case ( GC and GT are seldom mutated by hA3G in single cycle in vitro infections [20] . The open reading frames of HIV-1 pIIIB , ( the virus used in the in vitro analyses ) were used in computer simulations of hA3G-induced mutation ( Figure S1 ) . Predefined nGGn-to-nAGn mutation rates and the array of defined hA3G nGGn-to-nAGn mutation preferences [20] were used to determine the probability of mutation of each nGGn context . The mutation rate required to induce at least one stop codon in open reading frames in 50% of the simulations ( the Lethal Mutation 50% or LM50 ) was determined from 100 , 000 simulations of 100 incremental mutation rates in simulations of G-to-A , GG-to-AG and nGGn-to-nGAn mutations . Other thresholds such as LM95 and LM99 were also determined . The proportion of simulations without non-synonomous substitutions and the LM50 was also determined using the defined nGGn-to-nAGn mutation preferences [20] . The simulations did not account for the proposed twin gradient hypothesis for hypermutation , whereby the hA3G-induced mutational burden across individual genomes is proposed to increase from minima at the polypurine tracts ( PPTs ) in a plus-strand 5′-3′ direction [17] , [33] as existing data are insufficient to model this effect [20] , [78] . Nevertheless , since the twin gradient hypothesis predicts higher levels of mutation in the structural pol and env genes ( most distal to the 3′ ends of the PPTs ) , we predict that at a given mutation rate , simulations modeling this effect would yield increased numbers of stop codons in these genes with respect to the simulations described; thus our estimates are conservative . | Human cells have conserved antiviral defense systems , which protect against a range of viruses . A key component of this innate , intra-cellular defense is APOBEC3G ( hA3G ) , which can cause extensive and inactivating G-to-A mutations ( termed hypermutation ) in viral DNA . To circumvent this , human immunodeficiency virus type-1 ( HIV-1 ) encodes a protein , Vif , which can bind hA3 and prevent its antiviral effects . Vif is however , not always fully efficient , and many HIV-1 infected patients harbor hypermutated sequences . A key question is whether hA3G also might generate sub-lethal levels of G-to-A mutations , which could increase viral evolution , possibly accelerating disease progression . If this were to occur , drugs and vaccine-induced CTL-responses targeting Vif might have counterproductive effects . We show through in vitro , in vivo , and in silico analyses that it is unlikely that hA3G-activity can enhance virus evolution . Thus , measures that inhibit the interaction between Vif and APOBEC3G are likely to only increase the fraction of hypermutated , inactivated HIV sequences in the infected host . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"computerized",
"simulations",
"medicine",
"infectious",
"diseases",
"computer",
"science",
"computer",
"modeling",
"immunology",
"biology",
"computational",
"biology",
"evolutionary",
"biology",
"microbiology"
] | 2012 | APOBEC3G-Induced Hypermutation of Human Immunodeficiency Virus Type-1 Is Typically a Discrete “All or Nothing” Phenomenon |
Transposon-mediated transformation was used to produce Anopheles stephensi that express single-chain antibodies ( scFvs ) designed to target the human malaria parasite , Plasmodium falciparum . The scFvs , m1C3 , m4B7 , and m2A10 , are derived from mouse monoclonal antibodies that inhibit either ookinete invasion of the midgut or sporozoite invasion of salivary glands . The scFvs that target the parasite surface , m4B7 and m2A10 , were fused to an Anopheles gambiae antimicrobial peptide , Cecropin A . Previously-characterized Anopheles cis-acting DNA regulatory elements were included in the transgenes to coordinate scFv production with parasite development . Gene amplification and immunoblot analyses showed promoter-specific increases in transgene expression in blood-fed females . Transgenic mosquito lines expressing each of the scFv genes had significantly lower infection levels than controls when challenged with P . falciparum .
Plasmodium falciparum , a causative agent of human malaria , is a vector-borne parasite that is responsible for more than 500 million clinical disease cases each year [1] . The selection of insecticide-resistant mosquitoes and drug-resistant parasites necessitates an ongoing search for new disease-control methods . A proposed strategy for interrupting transmission is to replace wild , malaria-susceptible mosquito populations with transgenic , Plasmodium-resistant mosquitoes [2]–[4] . Key components of this approach are effector molecules that inhibit parasite development when expressed from a transgene . The mechanisms by which effector molecules function can vary greatly , as the development of the malaria parasites within mosquitoes involves several transitions of environment , physiology and morphology [5] . When mosquitoes feed on infected humans , they ingest parasites in the form of gametocytes . These produce gametes that fuse to form diploid zygotes that develop into the motile ookinetes . The ookinetes invade and traverse the mosquito midgut epithelium and then rest beneath the basal lamina of the midgut , forming oocysts . Thousands of sporozoites develop within the oocysts before budding out into the circulatory fluid , the hemolymph , and invading the salivary glands . Several effector molecules have been tested for their ability to target the parasite during either early sporogony in the midgut , or late sporogony in the hemolymph or salivary glands [5]–[8] . An effector mechanism based on the mosquito signaling protein Akt is the only one to date shown to inhibit completely P . falciparum development in a transgenic Anopheles mosquito [7] . One effector molecule strategy exploits the finding that monoclonal antibodies ( mAbs ) that recognize surface-bound or secreted parasite molecules can inhibit pathogen development [9]–[14] . Two mAbs , 4B7 and 1C3 , target parasites early in their development within mosquitoes . 4B7 binds P . falciparum surface protein Pfs25 , a molecule expressed on the surface of ookinetes , and inhibits parasite development completely when fed to Anopheles mosquitoes in a gametocytemic bloodmeal [9] . In contrast , 1C3 binds a parasite-secreted enzyme , P . falciparum chitinase 1 , and inhibits oocyst formation of P . falciparum when incorporated into infectious bloodmeals [10] . A third mAb , 2A10 , binds P . falciparum circumsporozoite protein ( CSP ) , and when pre-incubated with sporozoites , greatly decreases their ability to infect cultured hepatocytes [11] , [12] . Although the size and complexity of mAbs exclude them from consideration as potential effector molecules , single-chain antibodies ( scFvs ) , which retain the binding specificity of a mAb , are much smaller and can be produced from a single transcription unit [15] . An scFv targeting the P . gallinaceum CSP inhibited sporozoite invasion of salivary glands in Aedes aegypti in both transient assays and transgenic mosquitoes [13] , [16] . Anopheles stephensi fed Escherichia coli expressing an anti-P . berghei scFv-immunotoxin were shown to have significantly-reduced oocyst densities when fed on parasite-infected mice [14] . Furthermore , an scFv derived from the 1C3 mAb reduced significantly P . falciparum parasite transmission to mosquitoes [17] . The experiments described in the work presented here test the scFv-based strategy on human malaria parasites in transgenic mosquitoes and support the further development and evaluation of these molecules as disease-control tools . scFvs based on the 1C3 , 4B7 and 2A10 mAbs were expressed in transgenic An . stephensi and their efficacy tested in parasite challenge assays with P . falciparum . Anopheles stephensi was chosen because it is a significant vector of urban malaria transmission in the Indian subcontinent and is an efficient model for transgenic research . To distinguish the novel scFvs developed in this study , we refer to them as “modified” 1C3 , 4B7 or 2A10 ( m1C3 , m4B7 , m2A10 ) . For the m4B7 and m2A10 transgenes , the An . gambiae Cecropin A gene ( AgCecA ) was joined to the scFv gene to form a single open reading frame ( ORF ) . Cecropin A is an antimicrobial peptide that has microbiocidal activity against both gram-negative and gram-positive bacteria , as well as multiple Plasmodium species [18] , [19] . This broad activity is due to its ability to form large pores in cell membranes [20] . With the addition of cecropin A , the m4B7 and m2A10 scFvs possess both parasite-binding and antimicrobial activity . The cecropin A peptide was not joined to m1C3 as the target of this scFv is a secreted molecule [17] . Anopheles gambiae carboxypeptidase A ( AgCPA [21] , [22] ) gene regulatory sequences were included in m4B7 and m1C3 transgenes to coordinate their expression with the development of ookinetes . Anopheles stephensi vitellogenin 1 ( AsVg1 [23] ) regulatory elements were joined to the m2A10 scFv to direct transgene expression in the female fat body . Thus , m2A10 secreted from the fat body into the hemolymph could encounter sporozoites migrating to the salivary gland . When challenged in multiple experiments with P . falciparum infectious gametocyte cultures , scFv-expressing transgenic lines displayed statistically-significant , reduced mean intensities of infection and in most trials lower parasite prevalence when compared to control mosquitoes .
The scFv genes were synthesized commercially to incorporate either the AgCPA signal sequence or the entire AgCecA ORF ( Figure 1 ) . Codons corresponding to the amino acids serine , proline , alanine , threonine , and arginine displayed the greatest frequency bias differences between Mus musculus and An . gambiae ( Table S1 ) [24] , and these were replaced in the mouse-derived scFv sequences by those favored by the mosquito . DNA sequence encoding a short polypeptide linker ( five amino acids ) was used to join the heavy- and light-chain variable fragments of m4B7 and m2A10 scFvs and a longer linker ( encoding 15 amino acids ) joined the two corresponding moieties of m1C3 . Long linkers permit intramolecular pairing of variable fragments , while short linkers favor the intermolecular joining of scFv molecules to form multimers containing multiple antigen recognition sites [25] . The m1C3 and m4B7 scFv genes were joined to AgCPA regulatory elements and inserted into a pBac [3xP3-EGFP] plasmid to construct the transformation vectors ( Figure 2 ) . Similarly , the m2A10 scFv gene was joined to AsVg1 regulatory elements and inserted into a pBac [3xP3-dsRed] plasmid . The three transformation plasmids pBac [3xP3-EGFP]-m1C3 , pBac [3xP3-EGFP]-m4B7 and pBac [3xP3-dsRed]-m2A10 were injected into 980 , 615 and 765 embryos , respectively . Three transgenic m1C3 mosquito lines ( 21 . 1 , 39 . 1 and P4 . 1 ) were established from EGFP-positive families derived from 78 surviving adults . Two transgenic m4B7 mosquito lines ( 25 . 1 and P6 . 1 ) were established from 89 adults , and seven transgenic m2A10 mosquito lines ( 18 . 1 , 20 , 34 . 1 , 39 . 1 , 44 . 1 , P5 . 1 and P7 . 1 ) were established from 105 adults . Southern blot analyses were used to verify transgene insertions and to determine the number of integrated constructs in each line ( Figure 2 ) . Hybridization of an m1C3 probe to genomic DNA digested with both SphI and XhoI restriction endonucleases produced a diagnostic fragment of ∼1 . 2 kilobase pairs ( kb ) in transgenic samples , confirming m1C3 integration . Genomic DNA digested with SpeI and hybridized to an EGFP probe produced multiple fragments in each transgenic sample , indicating that there were at least three , nine , and ten copies in lines 21 . 1 , 39 . 1 , and P4 . 1 , respectively . Genomic DNA digested with ApaI , AscI , and FseI , and hybridized to a probe complementary to the m4B7 gene and the AgCPA 3′UTR produced two diagnostic fragments of 940 and 810 base pairs ( bp ) , verifying transgene insertion . A second blot , comprising XhoI-digested genomic DNA recovered from transgenic mosquitoes and hybridized with a 3XP3 EGFP probe , revealed several fragments in each sample , indicating that at least four copies of the m4B7 transgene were present in each line . Lastly , genomic DNA digested with both XbaI and BamHI and hybridized to an m2A10 probe produced an ∼1 kb diagnostic fragment in each transgenic sample . The same probe hybridized to HindIII-digested genomic DNA bound multiple DNA fragments in each m2A10 sample , indicating the presence of six , three , three , four , six , seven and three copies in transgenic lines 18 . 1 , 20 , 34 . 1 , 39 . 1 , 44 . 1 , P5 . 1 , P7 . 1 , respectively . Transgenic lines were maintained by intercrossing at each generation . However , selection pressures on individual transgene insertions , small founding colony sizes and independent assortment likely result in loss over time of some of the insertions . Reverse-transcriptase-PCR ( RT-PCR ) and Real-time quantitative RT-PCR ( RT-qPCR ) were used to evaluate the presence and relative abundance of m1C3 , m4B7 or m2A10 transcription products in non-blood-fed and blood-fed mosquitoes in all of the established transgenic lines . No significant correlation was seen between transgene copy number and amount of transcription product detected ( data not shown ) . Therefore , the lines m1C3 P4 . 1 , m4B7 25 . 1 and m2A10 44 . 1 , each of which displayed the highest levels of transcript accumulation in their respective group , were selected for use in all further analyses . Southern blot analyses of the generations of m1C3 P4 . 1 , m4B7 25 . 1 and m2A10 44 . 1 used in the challenge assays indicated the presence of eight , four , and four copies of the respective transgenes . Transgene-specific transcript accumulation profiles detected by RT-PCR were similar in mRNA samples prepared from the dissected midguts of m1C3 P4 . 1 and m4B7 25 . 1 females ( Figure 3 ) . Both lines showed constitutive accumulation in midguts from non-blood fed mosquitoes . In addition , each line showed accumulation of their respective mRNAs at 4 hours post-bloodmeal ( hPBM ) , and signals were evident at 12 and 24 hPBM . m4B7 transcript also could be detected at low levels at 48 hPBM . No amplification products were produced from mRNA prepared from female carcasses or males of each line . As expected , control reactions using mRNA from midguts dissected at 4 hPBM from wild-type , non-transgenic females were negative . RT-qPCR analysis at multiple post-bloodmeal time points was used as an independent measure of m1C3 P4 . 1 expression . The highest measured level of m1C3 mRNA , 10 , 000-fold above the control level , was observed in the 16 hPBM midgut sample ( paired T-test , one tailed p-value = 0 . 005 ) , but similar elevated levels also were seen at 24 and 48 hPBM . At this time , we cannot account for the difference in the RT-PCR and RT-qPCR results at 48 hPBM , although this could result from individual females that responded differentially to the feeding regimen . This difference is not expected to have affected the outcome of the challenge experiments because this scFv targets parasites within the first 24 hPBM . Immunoblot analyses of m1C3 and m4B7 transgenic mosquitoes were unproductive despite repeated attempts . Although high levels of proteinase inhibitors were used during sample preparation , it is possible that the transgene products were degraded quickly in the strong digestive milieu of the post-feeding midgut lumen . Transgene transcripts detected in whole transgenic m2A10 44 . 1 females showed sex- and stage-specificity ( Figure 3 ) . No signals were seen in samples derived from mRNA prepared from males and non-blood fed transgenic or control wild-type , non-transgenic females . Specific transcript accumulation was evident in m2A10 44 . 1 females at 12 and 24 hPBM in an expression pattern similar to that of endogenous AsVg1 , but was not as abundant at 48 hPBM . While expression of the m1C3 and m4B7 scFvs is necessary only during the first 24 hours post-bloodmeal , expression of m2A10 must be sustained over several days , as oocysts can mature asynchronously [26] . Denaturing immunoblot analyses were performed on m2A10 44 . 1 females sampled over the course of four bloodmeals to evaluate whether protein expression could be induced repeatedly ( Figure 4; Figure S1 ) . Anti-E tag antibody specifically detected a polypeptide with an approximate Mr of 32 kiloDaltons ( kDa ) , consistent with the predicted size of m2A10 protein , in transgenic blood-fed females at 24 , 48 , 72 and 96 hPBM . The continuous presence of m2A10 was detected in females that were given bloodmeals once every five days . Expression of m2A10 also was observed at 12 hPBM in additional immunoblot analyses ( data not shown ) . Immunoblots of hemolymph samples indicated that m2A10 protein was present in the hemolymph of blood-fed transgenic females ( Figure 4 ) . Immunoblot analyses of hemolymph samples analyzed in non-denaturing conditions detected m2A10 protein in several multimeric conformations with estimated Mrs of 125 , 223 , 284 , and 485 kDa . Parasite challenge experiments were performed to test the efficacy of the anti-pathogen effector molecules . Transgenic and control mosquitoes ingested blood containing P . falciparum gametocytes through a membrane-feeding apparatus . Control mosquitoes for most experiments were non-transgenic ( wild-type ) mosquitoes . In addition , the oocyst prevalence and mean intensities of infection of a group of m2A10 44 . 1 females were examined for each challenge experiment to determine whether transgenesis alone had an impact on parasite development . The effect of transgene expression on parasite development for both the m4B7 25 . 1 and m1C3 P4 . 1 transgenic lines was measured by comparing the number of oocysts in transgenic and control mosquito midguts at nine days after the infectious bloodmeal ( Table 1; Figure 5 ) . The mean intensities of oocyst infection were reduced by 37–81% in three challenge experiments ( 1 , 2 and 3 , Table 1 ) of m4B7 25 . 1 . However , the mean intensities of infection were reduced by only 29–36% in two experiments ( 4 and 5 , Table 1 ) in which control mosquitoes had greater than 17 oocysts per midgut . Mean intensities of oocyst infection were reduced by 47–73% in mosquitoes expressing m1C3 when compared to controls . Furthermore , with the exception of the high infection-level experiments ( 4 and 5 , Table 1 ) , both m1C3 P4 . 1 and m4B7 25 . 1 transgenic mosquitoes had lower prevalence of infections than controls . Parasite challenge assays of m2A10 44 . 1 involved dissecting 7–11 mosquitoes of each group 10 days after the infectious bloodmeal to count midgut oocysts and to confirm that both transgenic and control mosquitoes were infected successfully ( Table 2 ) . No statistically significant difference in the number of oocysts between transgenic m2A10 44 . 1 and control mosquitoes was observed ( Mann-Whitney U test , one-tailed P value , 0 . 24<p<0 . 48 ) . The remaining mosquitoes ( n = 8–50 ) in each group were examined 17–19 days after infection for the presence of sporozoites in the salivary glands ( Table 2; Figure 5 ) . All mosquitoes were provided an uninfected bloodmeal every five days to maintain expression of m2A10 . Engorged and un-engorged females were not separated after the uninfected bloodmeals in experiments 1 , 2 , 3 and 4 , and a 52–84% reduction in mean intensity of sporozoite infection was observed in transgenic mosquitoes when compared to the controls . To obtain a more precise measurement of the effect of m2A10 expression upon P . falciparum development , an additional three experiments ( 5 , 6 and 7 ) were performed in which un-engorged females were discarded after each uninfected bloodmeal . A 96–97% reduction of mean intensity of infection was observed in m2A10 44 . 1 mosquitoes that fed every five days . Furthermore , m2A10 44 . 1 mosquitoes in experiment 7 had a 14% prevalence of infection compared to 78% observed in the corresponding control .
Previous evaluations of mosquitoes engineered genetically to express anti-Plasmodium effector genes featured analyses of transgene copy numbers , transgene transcription levels , detection of transgene effector proteins , binding of effector molecules to the target parasite stage and a phenotype of reduced parasite mean intensities of infection and prevalence [7] , [8] , [16] , [22] , [27]–[29] . Remarkably , no single study includes all of these data and the emphasis has been on the impact of transgene presence on parasite numbers . Expression of the two midgut-directed scFvs , m1C3 and m4B7 , was detected by RT-PCR , but not by immunoblots . The rapid degradation of these scFvs in the midgut environment may have inhibited immunoblot detection . However , the observation that m4B7 25 . 1 and m1C3 P4 . 1 transgenic mosquitoes have reduced parasite loads supports the conclusion that these scFvs are expressed in the midgut . Both transgene transcription and translation products were detected in m2A10 44 . 1 mosquitoes . The finding that the immunoblot analyses of non-denatured m2A10 44 . 1 samples detected the presence of scFv multimers is consistent with the expectation that the short polypeptide linker joining the VH and VL regions promotes intermolecular scFv interactions . The size of these multimers was similar to the predicted sizes of m2A10 multimers comprising four , seven , nine , and fifteen scFv molecules . Such scFv multimers are reported to have high affinity to target epitopes [25] . Both m1C3 and m4B7 expressed in transgenic lines P4 . 1 and 25 . 1 , respectively , inhibited parasite development during early sporogony , resulting in significantly reduced mean intensities of oocyst infection in eight of ten challenge experiments . The results of two of the m4B7 25 . 1 challenge experiments are consistent with the interpretation that there is a threshold level of initial parasite density above which this scFv , at the levels expressed in these transgenic lines , cannot efficiently inhibit ookinete development . The finding that m2A10 44 . 1 and wild-type control mosquitoes did not differ in midgut infection supports the conclusion that transgene integration alone does not necessarily impair parasite development . When expression of m2A10 in line 44 . 1 was induced repeatedly by blood feeding , a highly significant decrease in sporozoite load was observed in transgenic mosquito salivary glands . For this transgenic line , the greatest reduction in prevalence was found in an experiment in which the mean intensity of oocyst infection was low . It is likely that these scFvs would effectively impair P . falciparum transmission in field conditions , as infected wild-caught An . gambiae carry few oocysts . Studies of An . gambiae by Billingsley et al . [30] and Taylor [31] found mean numbers of oocysts per infected mosquito of 1 . 55 and 3 . 38 , respectively . Incorporation of multiple transgenes is typical for piggyBac-mediated insertions into An . stephensi [22] , [23] , [32] . Although it is reasonable to expect that higher transgene copy numbers should yield higher expression levels , no statistically-significant correlations have been reported . We hypothesize that many of the multiple copies have little or no expression as a result of position effects , and that the majority of transgene expression comes from single or small numbers of the transgenes . To mitigate copy-number issues , we have used piggy-Bac-mediated transposition to integrate target sites for φC31 site-specific recombination into multiple locations in the An . stephensi genome and are now testing individual lines for permissiveness for optimum transgene expression [33] . These lines have the added benefit of having been evaluated for the impact on fitness of the introduced exogenous DNA at the specific insertion site , and therefore the effects of anti-pathogen transgene product expression can be measured directly . Furthermore , we are eager to evaluate the phenotype of dual transgenes , for example , those combining m4B7 and 2A10 or m1C3 and m2A10 , on parasite mean intensities of infection and prevalence . Additional studies facilitated by this approach could include testing alternate gene regulatory sequences , such as those of the salivary gland-specific anopheline antiplatelet protein or the An . gambiae adult peritrophic matrix protein 1 , to measure the effect of different transgene expression patterns upon parasite development [34] , [35] . Although the scFvs in lines m1C3 P4 . 1 , m4B725 . 1 , and m2A10 44 . 1 inhibited parasite development significantly , no transgenic line displayed zero prevalence of infection . It has been demonstrated in an avian malaria model system comprising the vertebrate host , Gallus gallus , the mosquito host , Aedes aegypti , and the parasite , P . gallinaceum , that mosquitoes containing as few as 20 sporozoites in their salivary glands infected chickens during a blood meal [16] . This finding supports the conclusion that a target of zero prevalence is necessary for a transgenic mosquito to be incapable of disease transmission in this system . These results are in contrast to reports of experiments with transgenic mosquitoes and a rodent malaria parasite , P . berghei , in which the effector molecules SM1 , PLA2 and CEL-III show a significant inhibition of parasite development ( 81 . 6% , 87% and 84 . 8% , respectively ) [22] , [27] , [28] . Reductions of mean intensities of P . berghei sporozoite infection in salivary glands below ∼400 were sufficient to block transmission . In contrast , experimental infections of humans with P . vivax showed that 10 sporozoites were sufficient to cause malaria [36] . We have opted to take the conservative approach and are attempting to achieve zero prevalence of human parasites in mosquito salivary glands [16] . Two anti-Plasmodium effector molecule strategies have yielded transgenic mosquitoes with zero prevalence: expression of the signaling molecule Akt , and expression of a combination of Cecropin A and Defensin A [7] , [29] . The latter study was conducted with the P . gallinaceum/Ae . aegypti/G . gallus model system . The study of Akt demonstrated the feasibility of producing an Anopheles mosquito that is completely resistant to P . falciparum , however this effector molecule may not be an optimal component of a population replacement strategy as these mosquitoes have a significantly reduced lifespan [7] . A synthetic peptide designed to interact with P . yoelii reduced midgut infections of this parasite by 67–87% in An . gambiae but was considerably less efficacious against P . falciparum [8] . Quantitative comparisons of the efficacy of alternative effector molecules are hindered currently by differences in expression that result from variations in transgene location and copy number . Site-specific recombination approaches will allow such evaluations in well-characterized ‘docking-site’ mosquito strains [8] , [33] . The finding that m2A10 44 . 1 mosquitoes display up to 97% decreases in the mean intensity of P . falciparum infection , as well as decreased prevalence of infection , supports the argument that this scFv may be an effective component of a malaria resistance transgene . The m1C3 and m4B7 scFv genes conferred significant reductions in mean intensities of infection , and if expressed in higher quantities , also may be used in the design of a transgenic , parasite-resistant mosquito . Furthermore , expressing the scFv transgenes in additional malaria vectors , in particular , An . gambiae , and challenging these with a variety of P . falciparum isolates would help evaluate whether these effector molecules could be used in multiple transmission areas . The discovery and characterization of several effector molecules that completely inhibit P . falciparum development will support the engineering of mosquitoes that express multiple effector molecules . Such mosquitoes may have a reduced likelihood of selecting for resistant parasites . Along with vaccines , drugs , and insecticide-treated nets , parasite-resistant transgenic mosquitoes would be a useful component in a malaria-control strategy , especially in regions where existing interventions have been unable to eliminate disease transmission .
A colony of Anopheles stephensi ( gift of M . Jacobs-Lorena , Johns Hopkins University ) bred in our insectary for >5 years was used in the experiments . The mosquitoes were maintained in conditions that maximize larval nutrition , and adult size and fitness [37] . These conditions include maintenance of cultures at 27°C with 77% humidity and 12 hr day/night , 30 min dusk/dawn lighting cycle . Larvae were fed a diet of powdered fish food ( Tetramin ) mixed with yeast . Adults were provided water and raisins ad libitum . Anesthetized chickens , mice , or rabbits were used for blood feeding . Transgenic and wild-type control mosquitoes used in parasite challenge experiments were reared in parallel using standardized insectary procedures . This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Intuitional Animal Care and Use Committee of the University of California , Irvine ( NIH Animal Welfare Assurance number: A3416 . 01 ( approved February 20 , 2008 ) , Protocol Number: 1998- 1411 ( approved May 21 , 2010 ) . The vertebrates used as bloodmeal donors for mosquitoes were anesthetized on a regimen that avoids the build-up of drug tolerance , and all efforts were made to minimize suffering . The sequences of the 4B7 and 2A10 variable heavy- and light-chain regions ( VH and VL , respectively ) were derived from cDNA synthesized from 2A10 and 4B7 hybridoma cell lines ( obtained from E . Nardin [New York University] , and the Malaria Research and Reference Reagent Resource Center , respectively ) . 2A10 cDNA was synthesized from total RNA isolated from the hybridoma cell line using primers designed from the known VH and VL sequence [38] . VH and VL cDNA from the 4B7 hybridoma cell line was amplified from its total RNA using the heavy and light primer mixes respectively , provided in the Mouse ScFv Module/Recombinant Phage Antibody System ( Amersham Biosciences ) . The modified scFv genes , including either AgCPA signal sequence or the entire AgCecA ORF , were synthesized commercially ( Epoch Biolabs ) to allow for incorporation of novel features . The variable regions of 4B7 , 2A10 , and 1C3 [17] were optimized by replacing the codons corresponding to the amino acids serine , proline , alanine , threonine , and arginine in the mouse-derived sequences with those favored in An . gambiae ( Table S1 ) [24] . For the m4B7 and m2A10 scFvs , the variable regions were joined by sequence encoding a short polypeptide linker , G4S . The VH region of the m2A10 and m4B7 scFv sequences were joined to the AgCecA protein-coding sequence by a long polypeptide linker , ( G4S ) 3 [39] . The variable regions of m1C3 were joined by the same long polypeptide linker . The m2A10 VL was joined to sequence encoding a complete E tag , while the VL of m4B7 and m1C3 were joined to sequence encoding a partial E tag . For m4B7 and m1C3 , the remaining E tag coding sequence was joined to the partial E tag at a later cloning step . The pBacDsRed-AsVg5′-m2A10′-AsVg3′ plasmid was produced in two cloning steps . First , m2A10 sequence from the commercially-synthesized pBSKm2A10 plasmid replaced the CFP gene of pSLfa-AsVg5′-CFP-AsVg3′ [23] using XbaI and BamHI sites . Second , the AsVg5′-m2A10-AsVg3′ sequence was joined to pBacDsRed [37] using AscI sites . A pSLfa-AgCP5′-4B7-AgCP3′ plasmid supplied the AgCP regulatory sequences , as well as a partial E tag sequence , for both the pBacEGFP AgCP5′-m1C3-AgCP3′ and the pBacEGFP AgCP5′-m4B7-AgCP3′ plasmids . The pSLfa-AgCP5′-4B7-AgCP3′ plasmid was cloned in several steps . First , the Mouse ScFv Module/Recombinant Phage Antibody System ( Amersham Biosciences ) was used to produce a single chain antibody from VH and VL cDNA from the 4B7 cell line . The scFv was cloned into the pCANTAB 5E vector in frame with the E-Tag at the C terminus . BclI sites were added to both ends of 4B7 scFv by amplification using primers 4B7BclF [5′-CGTGATCAGTGAAGCTGGTGGAGTCT-3′] and 4B7BclR [5′-CGTGATCACTATGCGGCACGCGGTT-3′] from the 5′ and 3′ end and cloning into pCR4Blunt-TOPO . The pGEMT[AgCP-SM1] plasmid , containing AgCP ( AGAP009593 ) regulatory regions , was generously provided by Dr . Marcelo Jacobs-Lorena [22] . The BclI-cut 4B7 scFv fragment from TOPO [4B7BclI] was sub-cloned into the Bam HI sites of pGEMT[AgCP-SM1] thereby swapping the SM1 fragment with 4B7 . AgCP5′-4B7-AgCP3′ was cloned subsequently into pSLfa1180fa [40] using enzymes SacII and SalI . ApaI and SgrAI were used to replace the 4B7 region of pSLFA-AgCP5′-4B7-AgCP3′ with the commercially-synthesized m4B7 gene . AgCP5′-m4B7-AgCP3′ sequence was then joined to pBacEGFP [40] using a 5′ blunt ligation of AscI and KpnI sites , and a 3′ FseI ligation . To assemble the m1C3 transformation plasmid , the enzymes ApaI and BamHI were used to replace the 4B7 region of pSLfa-AgCP5′-4B7-AgCP3′ . AgCP5′-m1C3-AgCP3′ was then joined to pBacEGFP using AscI restriction sites . Microinjection of the pBac [3xP3-EGFP]-m1C3 , pBac [3xP3-EGFP]-m4B7 or pBac [3xP3-dsRed]-m2A10 plasmids with the piggyBac helper plasmid was performed as described previously , except that 0 . 1 mM p- nitrophenyl p′-guanidinobenzoate was omitted from isotonic buffer [41] . Each G0 male was mated with 15 virgin females and groups of 5–10 G0 females were mated with 5 males , and G1 progeny were screened as larvae with UV-fluorescence microscopy for the presence of the marker genes . Standard Southern blotting and hybridization techniques were used to detect transgene integration [42] . Genomic DNA was extracted from groups of six transgenic or wild-type control females as described previously , except that DNA pellets were re-suspended in 100 µl of dH2O [43] . The probe used to identify m1C3 integration was amplified from a plasmid thought to contain the EGFP ORF , but which in fact contained the ECFP ORF . These two ORFs share 99% nucleotide sequence identity , so it is likely that the ability of the probe to hybridize to the integrated gene was affected negligibly . The EGFP probe was generated from pMos[3xP3-EGFP] [37] using XbaI and SacI enzymes . The m4B7 probe was generated through a restriction digest of the pBac [3xP3-EGFP]-m4B7 plasmid with both NaeI and FseI . The m2A10 probe was generated through a restriction digest of the pBSK-m2A10 plasmid with both BamHI and BstBI . Probes were labeled with 32P using the Megaprime DNA labeling system ( Amersham ) . Total RNA was isolated from whole or dissected mosquitoes using Trizol ( Invitrogen ) . For m2A10 RT-PCR analyses , 10 males or 2–3 whole females were used for each RNA preparation . One microgram of RNA was treated with DNAseI ( Promega ) for each 50 µl RT-PCR reaction . For m4B7 and m1C3 RT-PCR analyses , 6 males , 4–15 female midguts , or 4 female carcasses were used for each RNA preparation . Two hundred fifty nanograms of RNA were treated with DNAseI for each 12 . 5 µl RT-PCR reaction . Gene-specific primers and a OneStep RT-PCR Kit ( Qiagen ) were used for amplification of diagnostic products from m2A10 , m4B7 , m1C3 , AsCPA [32] , AsVg1 [23] , or An . stephensi ribosomal protein S26 gene [23] transcripts ( Table S2 ) . For m2A10 RT-PCR analyses , amplification of diagnostic products from AsVg1 and ribosomal protein S26 gene-specific primers was performed in a single reaction . Diagnostic amplification reactions for the m4B7 25 . 1 and m1C3 P4 . 1 lines were initiated with one cycle at 50°C for 30 m , one cycle at 95°C for 15 m , 32 cycles denaturation at 94°C for 30 s , annealing at a reaction-specific temperature ( Table S2 ) for 30 s , and extension at 72°C for 1 m , followed by a final extension at 72°C for 10 m . Diagnostic amplification reactions for the m2A10 44 . 1 line were performed as described , except that 30 cycles of amplification were completed . For each sample , an additional control RT-PCR reaction tested for the presence of genomic DNA contamination using ribosomal protein S26 gene primers but omitting the reverse transcription step . Multiple biological replicates ( ≥2 ) were performed for selected time points for each of the RT-PCR series of experiments . Female mosquito midguts and carcasses and male midguts were dissected in phosphate-buffered saline ( PBS ) , homogenized in Trizol Reagent ( Invitrogen ) , and total RNA extracted . Midguts were dissected at different time points ( 4 h , 8 h , 16 h , 24 h , 48 h , 72 h , 7 d , and 15 d ) after a bloodmeal . RNA was treated with DNase I ( Invitrogen ) at 1 U/µg RNA to remove potential genomic DNA contamination . Further purification was performed using a DNA-free kit ( Ambion ) . A total of 0 . 4 µg of RNA was used for reverse transcription in a reaction volume of 20 µl using ThermoScript RT-PCR System ( Invitrogen ) . Real-time quantitative PCR was performed on an Opticon 2 Real-Time PCR Detection System using the Opticon Monitor software version 3 . 1 ( Bio-Rad laboratories ) . m1C3 expression was quantified with Platinum SYBR Green qPCR SuperMix UDG with ROX ( Invitrogen ) using gene-specific primers ( Table S2 ) to amplify a diagnostic fragment 211 bp in length . A series of quantitative standards were generated from serial 10-fold dilutions ( a range of 1010-1 molecules ) of TOPO-m1C3scFv , in which full-length m1C3 was cloned . Each assay was run in triplicate wells in a 25 µl final reaction volume containing 2 . 5 µl of Platinum SYBR Green , 400 nM each forward and reverse primer , and 2 . 5 µl cDNA sample . Each run included negative controls consisting of wild-type control cDNA and water instead of cDNA . The amplification protocol consisted of 2 min at 50°C , 2 min at 95°C , followed by 40 cycles of amplification ( 94°C for 15 s , 60°C for 45 s , plate read of SYBR Green I fluorescence ) , after which a melting-curve reaction was conducted from 42°C to 95°C with plate readings every 1°C . GraphPad Prism software was used to calculate statistical significance using paired T-tests . Mosquitoes were blood-fed on chickens and homogenized in a protease inhibitor solution made from complete mini ( Roche ) and Pefabloc SC ( Roche ) . An equal volume of Laemmli sample buffer ( Bio-Rad ) with 0 . 1 M dithiothreitol was added . Homogenates were separated on a 12% Tris-HCl polyacrylamide gel in 1×Tris/Glycine/SDS buffer ( Bio-Rad ) , transferred to Immun-Blot PVDF membrane ( Bio-Rad ) , and incubated with goat anti-E tag polyclonal antibody conjugated to horse radish peroxidase ( Abcam ) . ECL Plus Western Blotting Detection Reagents ( GE Healthcare ) were used to detect bound antibody . Ten females were used for each hemolymph sample preparation . Legs were removed with forceps and the proboscis was cut with a scissor . Individuals were inserted into a pipette tip plugged with glass wool and threaded through a 0 . 5 ml tube placed in a 1 . 5 ml collection tube . Centrifugation at 530 g for 10 min at 4°C extracted hemolymph . Each hemolymph sample was mixed with 25 µl 0 . 15 M NaCl and centrifuged at 2040 g for 5 min at 4°C . Fifteen microliters of the middle fraction of the sample was transferred to a new 1 . 5 ml tube , to which 10 µl of the protease inhibitor solution was added . For immunoblots with non-denatured samples , native sample buffer ( Bio-Rad ) , 4–15% Tris-HCl polyacrylamide gels ( Bio-Rad ) , 1×Tris/Glycine electrophoresis buffer ( Bio-Rad ) , and native transfer buffer ( 25 mM Tris , 25 mM Glycine , pH 9 . 2 ) were used . Four to six day-old transgenic and wild type female mosquitoes were fed with P . falciparum NF 54 gametocytes using a membrane feeding apparatus . After 15 min of feeding , un-engorged mosquitoes were removed and engorged mosquitoes were maintained in the insectary under standard conditions [37] with daily access to a 10% sucrose solution or water and raisins . Midguts were dissected 9 days after the infectious bloodmeal , stained with 0 . 1% mercurochrome and the number of oocysts in each preparation counted . Uninfected bloodmeals were provided to transgenic and wild-type control mosquitoes following the membrane feeding . For the m4B7 25 . 1 experiments , mosquitoes were allowed to feed on the first and second days post-infection . Mosquitoes in the m2A10 experiments were allowed to feed on the 4th , 8th , and 12th days post-infection . Engorged and un-engorged females were not separated after the uninfected bloodmeals in m2A10 experiments 1–4 , while un-engorged females were discarded in experiments 5–7 . Samples of wild-type control and m2A10 females were dissected for oocyst counts on the 10th day post-infection . The salivary glands of all remaining m2A10 and wild-type control females were dissected 17–19 days post-infection . A hemacytometer was used in m2A10 experiments 1–4 to count salivary gland sporozoites [13] . The samples in experiments 5–7 were dried on 6 mm well slides and stored at −20°C . Sporozoites were stained using SlowFade Gold antifade reagent with DAPI ( Invitrogen ) and counted with a Zeiss Axioskop using the Axiovision camera and software . Sporozoites were counted using one of three methods , depending on parasite density . Method 1: If the number of sporozoites in each of five fields was counted , and a total of 3 or more sporozoites was found , an average sporozoite/mm2 measurement was calculated . When a field contained greater than 50 parasites , Improvision Volocity software was used to count the number of sporozoite nuclei in the DAPI image ( Measurement protocol: 1 . Find 2D nuclei: separate touching nuclei with a separation guide of 0 . 4 µm , reject nuclei with an area of less than 0 . 2 µm2 . 2 . Exclude objects by size: exclude objects >10 µm2 ) . Method 2: If the 3 sporozoite requirement of method 1 was not met , fields were examined until 3 sporozoites were counted , and an average sporozoite/mm2 measurement was calculated . Sporozoite/mm2 values were used to calculate the total number of sporozoites present in the 6 mm2 slide well area . Method 3: If a total of 3 sporozoites was not found in up to 25 fields , the entire 6 mm2 slide well area was examined for an exact count . GraphPad Prism software was used to calculate statistical significance using Mann-Whitney U tests . The GenBank ( http://www . ncbi . nlm . nih . gov/Genbank/ ) accession numbers for the m1C3 , m4B7 , and m2A10 genes are HQ315886 , HQ315885 , and HQ315884 , respectively . | Malaria eradication will require vector-control strategies that are both self-sustaining and not affected by migration of infected humans and mosquitoes . Replacement of wild malaria-susceptible mosquito populations with transgenic strains refractory to parasite development could interrupt the cycle of disease transmission and support eradication efforts . Production of P . falciparum-resistant mosquitoes is a necessary first step towards investigating the population replacement strategy . Here we show that An . stephensi engineered to produce P . falciparum-targeting effector molecules are resistant to this important human malaria parasite . Two of the three effector molecules represent a novel combination of components derived from the immune systems of mosquitoes and mice . An important feature of these molecules is that they are unlikely to significantly harm the mosquito , as the mosquito component is an Anopheles antimicrobial peptide with activity against Plasmodium , while the other component is based on a murine antibody selected for its ability to bind specifically to a parasite protein . Transgenes with this design coupled with a gene-drive system could be used alongside vaccines and drugs to provide sustainable local elimination of malaria as part of a long-term strategy for eradication . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biotechnology",
"medicine",
"infectious",
"diseases",
"transgenics",
"gene",
"regulation",
"genetics",
"molecular",
"genetics",
"biology",
"malaria",
"parasitic",
"diseases",
"genetics",
"and",
"genomics",
"genetic",
"engineering"
] | 2011 | Engineered Resistance to Plasmodium falciparum
Development in Transgenic Anopheles stephensi |
Apicomplexan parasites are obligate intracellular parasites that infect a variety of hosts , causing significant diseases in livestock and humans . The invasive forms of the parasites invade their host cells by gliding motility , an active process driven by parasite adhesion proteins and molecular motors . A crucial point during host cell invasion is the formation of a ring-shaped area of intimate contact between the parasite and the host known as a tight junction . As the invasive zoite propels itself into the host-cell , the junction moves down the length of the parasite . This process must be tightly regulated and signalling is likely to play a role in this event . One crucial protein for tight-junction formation is the apical membrane antigen 1 ( AMA1 ) . Here we have investigated the phosphorylation status of this key player in the invasion process in the human malaria parasite Plasmodium falciparum . We show that the cytoplasmic tail of P . falciparum AMA1 is phosphorylated at serine 610 . We provide evidence that the enzyme responsible for serine 610 phosphorylation is the cAMP regulated protein kinase A ( PfPKA ) . Importantly , mutation of AMA1 serine 610 to alanine abrogates phosphorylation of AMA1 in vivo and dramatically impedes invasion . In addition to shedding unexpected new light on AMA1 function , this work represents the first time PKA has been implicated in merozoite invasion .
Malaria is one of the most devastating infectious diseases of mankind and is a leading cause of morbidity and mortality in tropical and sub-tropical regions where 40% of the world's population live . The most pathogenic species that infects humans is Plasmodium falciparum and in 2002 it was estimated that out a total of 515 million clinical cases , 2–3 million were fatal [1] . Central to malarial pathogenesis is the large-scale invasion of red blood cells ( RBCs ) by Plasmodium parasites . The invasive merozoite forms of the parasite infect RBCs via a complex multi-step process involving sequential receptor-ligand interactions and signal transduction events ( reviewed in [2] ) . Merozoite invasion is an intense area of investigation by many groups as it is a point in the parasite lifecycle that is particularly vulnerable to immune and drug intervention . While signalling within the parasite , particularly that triggered by calcium , is known to be involved in RBC invasion , the specific nature of this process including the identity of the key molecular players remains largely a mystery . To address this we have been studying an essential transmembrane protein present on the invasive merozoite surface , apical membrane antigen 1 ( AMA1 ) . AMA1 is one of the most promising blood-stage malaria vaccine candidates and is amongst the best studied of the ∼5000 Plasmodium proteins . In P . falciparum , AMA1 is synthesised during merozoite development towards the end of the blood stage cell cycle and is stored in apical secretory organelles called micronemes [3] , [4] . It is a type I integral membrane protein of 83 kDa with a large N-terminal ectodomain , a single transmembrane domain near the C-terminus and a small cytoplasmic tail of 56 amino acids ( PfAMA183 ) [5] , [6] . Before schizont rupture the N-terminal prosequence is cleaved resulting in a 66-kDa form ( PfAMA166 ) that translocates from the micronemes to the merozoites surface [7] . In a second proteolytic processing step during invasion the bulk of the ectodomain is shed quantitatively as 44-kDa and 48-kDa fragments by the membrane bound subtilisin-like protease PfSUB2 leaving a 22-kDa transmembrane fragment that is taken into the newly invaded RBC [8] , [9] . AMA1-specific antibodies and peptides block invasion at a step after the long-distant primary contacts between parasite and host cell have occurred but prior to the close interactions seen during tight junction formation [10]–[12] . Some functional insights of antibody-based inhibition were gained from using the monoclonal antibody 4G2 that targets an epitope adjacent to the conserved hydrophobic trough in the ectodomain [13]–[15] . This antibody binding led to the abrogation of AMA1 interaction with rhoptry neck proteins ( RONs , [14] ) that was previously established as important part of the tight junction [16] , [17] . Apart from the critical interactions of the ectodomain the cytoplasmic tail of AMA1 has also been shown to be essential for invasion [18] . Mutational analysis hinted an important role for tail phosphorylation in the invasion process . In this paper we show that the PfAMA1 tail is phosphorylated at a specific serine residue ( S610 ) and that the enzyme responsible for this event is the parasite-encoded protein kinase A ( PfPKA ) . Moreover , we show that S610 plays a crucial role for AMA1 function and parasite invasion . This phosphorylation event has implications for understanding the regulation of invasion , for the function of AMA1 and for the development of new therapeutic approaches .
To investigate if the cytoplasmic domain of AMA1 is phosphorylated by parasite kinases we generated a glutathione S-transferase ( GST ) fusion protein of the AMA1 cytoplasmic tail ( Figure 1A ) and performed in vitro phosphorylation assays with P . falciparum 3D7 lysates . Autoradiography of the AMA1 tail resolved by SDS-PAGE indicated it was specifically phosphorylated to comparable amounts by schizont and merozoite lysates ( Figure 1B ) . Control reactions with non-infected RBC lysate gave only a background signal , indicating that the AMA1 tail was phosphorylated by parasite kinases rather than RBC kinases ( Figure 1B ) . As loading controls the membrane was probed with an anti-AMA1 antibody that specifically recognised the AMA1 tail . No signal can be detected for the GST proteins because the antibody used was specific for the AMA1 tail only ( Figure 1 ) . Invasion of RBCs by Plasmodium is known to involve calcium ion ( Ca2+ ) fluxes [19] which might trigger AMA1 phosphorylation . To address this , calcium ions in parasite lysates were chelated by EGTA/EDTA before incubation with the AMA1 tail . Conversely , to increase the calcium concentration in the assay 2mM CaCl2 was added to the buffer . Whereas EGTA/EDTA had little effect on the intensity of the phosphorylation signal compared to the untreated sample ( Figure 1C , left panel and Figure 1D ) , the addition of Ca2+ to the buffer resulted in a 3 fold increase of the phosphorylation signal ( Figure 1C , left panel and Figure 1D ) . In addition to Ca2+ , cAMP is also a common second messenger and has been shown to be involved in parasite development during the asexual blood stages [20] , [21] . Thus , we tested if cAMP had an effect on AMA1 tail phosphorylation . As shown in Figure 1C ( right panel ) the addition of cAMP to the in vitro phosphorylation assay led to a dramatic 17-fold increase of AMA1 tail phosphorylation ( Figure 1D ) , which indicated an involvement of the protein kinase A ( PKA ) in the phosphorylation event . A multiple alignment of 13 AMA1 protein sequences using the software PRALINE ( www . ibi . vu . nl ) showed that the C-terminal cytoplasmic domain of apicomplexan AMA1 is highly conserved among different Plasmodium species as well as other apicomplexans ( Toxoplasma gondii , Babesia bovis , Theileria parva , Theileria annulata ) suggesting a common function during host cell invasion . Additionally , the AMA1 tails contain several potential phosphorylation sites with six amino acids being predicted as phosphorylation sites in PfAMA1 ( www . cbs . dtu . dk/services/NetPhos ) ( Y576 , Y585 , S590 , S610 , T612 , Y622 , Figure 2A ) . The dependence on cAMP suggested that parasite-encoded PKA ( PFI1685w ) , the only apparently recognisable cAMP dependent kinase expressed at this stage of the life cycle [22] , was responsible for the observed phosphorylation . Consistent with that , the NetPhosK program ( www . cbs . dtu . dk/services/NetPhosK ) predicted that residue S610 is phosphorylated by PKA , showing the highest score when compared to all the other serines , threonines and tyrosines in the AMA1 tail . To establish if the prediction of S610 phosphorylation by PfPKA was correct , site-directed mutagenesis was performed . Firstly , a GST-fusion protein mutant containing a stop codon at position S610 was generated ( Figure 2A ) that lacks the highly conserved C-terminus with its putative phosphorylation sites including the S610 . In vitro phosphorylation assays with 3D7 schizont lysates indicated this mutant was not phosphorylated in the presence of cAMP ( Figure 2B ) , suggesting that none of the less conserved proximal predicted phosphorylation sites Y576 , Y585 , S590 were involved in this phosphorylation event or that if these sites are phosphorylated their phosphorylation is dependent on the presence of serine 610 or residues downstream of S610 . Although tyrosine kinases are apparently absent in the parasite's genome [23] two predicted tyrosines were included in the mutagenic analyses as controls . The band for the stop mutant is missing in the loading control since the anti-AMA1 antibody detects a peptide C-terminally of S610stop ( Figure 2B ) . Secondly , to verify that S610A is responsible for protein phosphorylation , we exchanged the remaining phosphorylation sites ( including S601 ) and subjected these mutant AMA1s to in vitro phosphorylation . Only the stop mutant and S610 lacked a phosphorylation signal suggesting that S610 is either the only residue being phosphorylated under the given conditions or S610 phosphorylation enables the phosphorylation of other sites in the AMA1 tail ( Figure 2B ) . To confirm the involvement of PfPKA as the phosphorylating kinase , two PKA inhibitors , H89 and KT 5720 , were tested for their ability to block phosphorylation of the AMA1 cytoplasmic domain . These compounds are competitive antagonists of ATP's access to a binding pocket on the catalytic subunit of PKA and H89 has been used in Plasmodium parasites previously where it appears to affect the cell cycle and arrest proliferation during schizogany [20] , [21] . To determine biologically relevant concentrations of PKA inhibitors to use in the in vitro phosphorylation assays , the growth inhibitory effects of the drugs were measured in live cultures . These indicated that the IC90 for H89 and KT 5720 was about 50 µM and 10 µM , respectively ( data not shown ) . In the in vitro phosphorylation assays , both compounds completely blocked the stimulatory effect of cAMP upon the phosphorylation of the recombinant AMA1 tail indicating potent inhibition of PKA stimulation ( Figure 2C and 2D ) . In the absence of additional cAMP , H89 and KT 5720 decreased AMA1 tail phosphorylation by 70% and 55% respectively in in vitro phosphorylation assays using 3D7 schizont lysates ( Figure 2C and 2E ) . Although we can't completely rule out that the inhibition of cAMP stimulated kinase activity is due to methanol , the solvent of KT5720 , rather than KT5720 itself , it is very unlikely for two reasons: firstly , both inhibitors show similar inhibitions , even if only KT5720 is dissolved in methanol , whereas H89 is dissolved in H2O . Secondly , methanol per se had no effect on the outcome of the in vitro phosphorylation assay in the absence of additional cAMP as shown in Figure 2C and 2E . Although H89 and KT5720 have been used extensively in animal cell PKA studies for many years some doubts about the specificity of these compounds for PKA , especially at the used concentrations , remain and they may have effects on other kinases such as PKB [24] . We can therefore not be entirely sure that the background ( unstimulated ) level of AMA1 tail phosphorylation is completely due to PKA since other kinases may also be inhibited . To further validate PfPKA as an AMA1 phosphorylating kinase , we over-expressed HA-tagged PfPKA catalytic subunit in late blood stages . Transgenic expression was verified by either Western Blotting or IFA using anti-HA antibodies ( Figure S1 ) . The antibodies detected a double band at around 43 kDa corresponding well with the theoretical molecular weight of the fusion-protein of 42 . 5 kDa . The doublet might represent a phosphorylated form of PfPKA as has been described for other PKA proteins [25] . In vivo activation of PKA in Schizosaccharomyces pombe requires threonine phosphorylation at its activation loop and is dependent on PDK1 [26] . Subsequent immuno-precipitation of the HA-tagged PfPKA from 3D7PfPKA-HA parasites allowed the purification of the PfPKA-HA ( Figure 2F ) . Late stage specific over-expression of the catalytic subunit of PfPKA had no effect on parasite growth rate ( data not shown ) . This purified PfPKA readily phosphorylated the recombinant AMA1 tail ( Figure 2G ) . Taken together , the in vitro data suggest that cAMP triggers AMA1 tail phosphorylation on residue S610 by PfPKA . To investigate the phosphorylation status of native AMA1 we analysed mature schizont stage parasite extracts from 3D7 parental parasites by 2DE ( Figure 3A-C ) . When analysed by Western Blotting using antibodies that recognise the C-terminus of AMA1 , the 66 kDa AMA1 species was shown to be comprised of 5 distinct spots ( termed “a-e” respectively with spot “a” being the most negatively charged and “e” the least ) that separated on a basis of their isoelectric point . Three of these spots ( a , c and d ) incorporated radiolabelled phosphate indicating that these spots represent phosphorylated forms of AMA1 ( Figure 3B ) . To determine if S610 was phosphorylated in vivo the phosphorylation patterns of wild type AMA1 and the non-phosphorylatable S610A and PM ( all potential sites mutated ) mutants were compared by 2DE . Because previous studies had indicated these phosphorylation sites were essential for AMA1 function we used a complementation approach [18] . We generated 3D7 parasites episomally expressing the W2mef allelic form of AMA1 tagged at the C-terminus with the TY1 epitope to distinguish it from the endogenous 3D7 AMA1 protein . Three lines were created expressing either the W2mef wild type form ( AMA1-WT-TY1 ) , a form with the S610A mutation ( AMA1-S610A-TY1 ) or another form with each potential phosphorylation site mutated ( AMA1-PM-TY1 ) ( Figure 3D-I ) . Samples were analysed by 2DE and Western Blot which were probed with either TY1 antibodies or polyclonal AMA1-C-terminal antibodies where indicated . It was apparent that the 66 kDa species in the AMA1-WT-TY1 line separates into 8 visible spots ( Figure 3E ) consistent with the expression of two different forms of AMA1 in this parasite line . The W2mef TY1 tagged spots , termed a'-e' , were generally of slightly higher molecular weight than the untagged 3D7 species and their PIs were a little greater ( Figure 3D , E ) . We assume 10 spots are present in this line but that 2 spots ( b and b' ) are masked by other spots or are below detectable levels . Consistent with the masking of b' , anti-TY1 antibodies detect spots a' , c' , d' and e' in AMA1-WT-TY1 parasites ( Figure 3D ) . Strikingly , the banding pattern in AMA1-S610A-TY1 is much simpler with only 2 TY1-tagged spots observed , b' and e' ( Figure 3F and G ) . Mutation of all potential phosphorylation sites in the AMA1 tail to alanine ( including S610 ) gave an identical pattern to the S610A-only mutant ( Figure 3H and I ) . The mutation of S610 to an alanine residue appears to have eliminated the presence of all phosphorylated species of AMA1 . One possible scenario might be that spots a , c & d represent 3 separate phosphorylation sites on the protein with the modification of S610 as prerequisite for the phosphorylation of the other sites . Alternatively , all three spots could contain phosphorylated S610 , but two of the three contain additional charge-modifying posttranslational modifications depending on S610 phosphorylation . Indeed , our data suggest that there are at least two forms that differ in charge for reasons other than phosphorylation ( b & e ) . This data confirms a crucial role for S610 in the phosphorylation of AMA1 . Due to the inability to obtain perfectly synchronous parasites it is difficult to precisely determine when S610 phosphorylation is occurring . However , to address this issue to some extent we prepared a time course experiment examining AMA1 species present in schizonts , merozoites and ring-stages ( Figure 4 ) . By Western Blotting 2DE gels with an antibody that recognises the C-terminal tail of AMA1 we show that the 66 kDa AMA1 species ( a form that appears late in schizogony ) has a more complex banding pattern in merozoites than it does in schizonts , most notably the appearance of a more negatively charged , apparently phosphorylated , species ( arrowhead in Figure 4 ) . This is in contrast to the 83 kDa precursor form of AMA1 that appears to change little in pattern between schizonts and merozoites . While , all phosphorylated species of the 66 kDa fragment can be detected in some schizont preparations ( see Figure 3A-C for example ) , the time-course data shown in Figure 4 suggests that the bulk of secondary modifications of the AMA1 tail , including phosphorylation , occurs very late in blood stage development , perhaps even in the short-lived merozoites themselves . Interestingly , the cleaved C-terminal form of AMA1 in newly invaded ring-stage parasites reveals a much simpler pattern of secondary modifications , which could point towards - among other possibilities - a de-phosphorylation event of S610 during this stage ( Figure 4 ) . We caution that precise knowledge about the timing of AMA1 phosphorylation and its potential de-phosphorylation will require the generation of new reagents , most probably a functional S610 phospho-specific antibody . We next tested the functional consequences of S610A mutation in vivo . To do this we used a similar complementation approach as previously described [18] , [27] . Briefly , full length AMA1 bearing a C-terminal TY1 epitope tag for immuno-detection was ectopically expressed in the 3D7 parasite line under the AMA1 promoter ( Figure 5A ) . To test the invasion capability of phosphorylation defective mutants , a S610A mutation ( AMA1-S610A-TY1 ) or mutations of all six predicted phosphorylation sites in the AMA1 tail to alanines ( AMA1-PM-TY1 ) were introduced by site-directed mutagenesis ( Figure 5B ) . Importantly , all the AMA1-TY1 proteins , wild type and mutants , were derived from the W2mef parasite strain . In this strain the AMA1 protein bears crucial amino acid changes that makes it resistant to the invasion blocking effects of the R1 peptide that binds to the 3D7 AMA1 protein [28] . This strain specific binding blocks RBC invasion by 3D7 parasites though it does not prevent initial interaction steps between the merozoite and RBC [18] . All chimeric proteins were correctly expressed as TY1 fusion proteins ( Figure 5C , left panel ) . Proteolytic cleavage of the HA-fusion was indistinguishable from the endogenous protein as shown with the 3D7 specific monoclonal AMA1 antibody 1F9 [29] ( Figure 5C , right panel ) . As previously reported , over-expression of AMA1-WT-TY1 ( derived from W2mef ) functionally complements the endogenous AMA1 ( ∼78 . 5% ( +/− 5 . 1% s . d . ) invasion ) while both mutants AMA1-S610A-TY1 and AMA1-PM-TY1 revealed a drastically decreased invasion capability in the presence of R1 peptide ( ∼21 . 2% ( +/−5 . 9% s . d . ) and ∼20 . 3% ( +/− 5 . 2% s . d . ) invasion , respectively ) ( Figure 5D ) . In comparison , invasion is blocked up to ∼96% ( +/− 1 . 3% s . d . ) in the parental 3D7 parasite line with the R1 peptide . This comparable failure of the PM and the S610A mutant to complement AMA1 function demonstrates an important role for AMA1 residue S610 and , together with the above in vitro data , strongly implicates PKA-mediated phosphorylation of S610 as a vital step in promoting efficient invasion . Direct comparison of the temporal aspect of invasion blockade induced by either R1 peptide binding or S610A mutation using video-microscopy revealed no apparent differences . Both parasites are arrested after the re-orientation of the merozoite on the surface of the host cell ( Video S1 , Figure S2 ) . In summary , this data demonstrates that the cytoplasmic tail of AMA1 is phosphorylated by parasite-encoded PKA , and that this is crucial to AMA1 function during the invasion . The vital role of S610 was further investigated by targeting two other single amino acids that are predicted and could be putatively involved in phosphorylation of AMA1: Y576 and Y585 . The substitution of these amino acids with alanine and its subsequent ectopic expression ( AMA1-Y576A-TY1 and AMA1-Y585A-TY1 ) did not impair their function in complementation assays ( Figure 5D ) . Additionally , we have tried to mimic the phosphorylation state of AMA1 by expressing AMA1-S610E-TY1 and AMA1-S610D-TY1 mutants in the presence of R1 peptide , however they failed to complement the invasion efficiency of the AMA1-WT-TY1 allele and performed little better than the non-phosphorylatable S610A allele ( Figure 5D ) . This could be because the negatively charged amino acids fail to fully substitute the biological activity of a phosphate group or because they cannot lose their charge through the action of phosphatases . The latter seems a distinct possibility since the 2D pattern of AMA1 spots in ring-stage parasites is greatly simplified compared to merozoites suggesting the activity of phosphatases .
Merozoite invasion of RBCs is a rapid yet complex multi-step process . It can be broadly separated into a ∼11 sec pre-invasion stage which is characterised by long-distance interactions and extensive deformation of the host cell and a ∼17 second invasion step where the parasite has formed a moving tight junction [30] . The pre-invasion stage is particularly poorly understood both in terms of the receptor-ligand interactions involved and in the communication events that link the subsequent stages of invasion . However , it appears likely that during the pre-invasion stage a signalling cascade occurs to promote subsequent invasion steps , such as apical organelle secretion and activation of the actin-myosin motor . Most notably in this regard calcium dependent kinase 1 ( CDPK1 ) has been implicated to be involved in phosphorylating motor components [31] . In the present work , we have shown that the major transmembrane protein known to be present on the surface of free merozoites AMA1 , appears to be phosphorylated at a specific residue , S610 , in its cytoplasmic tail . We show that this occurs both in in vitro assays using recombinant AMA1 with parasite extracts and also in vivo in whole parasites where S610 appears to be a dominant site of phosphorylation . Given that we observed 3 phosphorylated isoforms of AMA1 it remains possible that another site ( s ) of the AMA1 tail is/are also phosphorylated . However , if this is the case , these sites must be subsequent to and dependent of serine 610 phosphorylation as all isoforms are absent when this residue is mutated . Several lines of evidence indicate that the enzyme responsible for this event is parasite-encoded PfPKA , an enzyme previously unknown to be involved in invasion . Firstly , in vitro S610 is phosphorylated strongly in a cAMP dependent manner and PfPKA is the only known cAMP dependent kinase expressed in the blood stages of P . falciparum . In fact , PfPKA transcription is co-regulated with that of AMA1 late in the blood-stage cycle and S610 is strongly predicted by publicly available software to constitute a PKA site . Secondly , two different PKA inhibitors H89 and KT5720 block this phosphorylation in vitro . Thirdly , we show that immuno-precipitated PfPKA phosphorylates the tail of AMA1 in vitro . Finally we use a complementation approach to show that S610 is important for efficient red blood cell invasion . In this experiment , mutant parasites that express an AMA1 with an alanine in this position are unable to efficiently invade host cells . Taken together , we conclude that PfPKA mediated phosphorylation of AMA1 at S610 is vital to its functioning and hence to the invasion process . So what is the role of this process in invasion and what can we learn from host cell entry at other parasite stages ? With respect to the latter , PfPKA is also expressed in the hepatocyte invasive sporozoite stages and indeed PKA has been implicated in invasion via genetic deletion of P . berghei adenylyl cyclase alpha ( ACα; [32] ) . ACα is responsible for rapidly generating cAMP from ATP in response to a stimulus , in this case uracil derivates [32] . The membrane associated ACα is not expressed in blood-stages however another adenylyl cyclase , ACβ , is expressed during the blood-stages [33] . In fact , the gene encoding ACβ is co-transcribed with genes encoding PKA and AMA1 late in the blood-stage cycle . Hence , ACβ is a strong candidate to produce the cAMP required to activate PKA during invasion . Experimental confirmation of this is required but if true the identification of the activation signals of ACβ should shed considerable light on the primary trigger for merozoite invasion of red blood cells . It will be interesting to discover how the phosphorylation of AMA1 renders the merozoites competent to invade . Phosphorylation does not appear to influence AMA1 trafficking since expression of a tail deletion mutant in transgenic parasites does not appear to effect localisation [18] . Previous experiments have indicated that the interaction of AMA1 with RON proteins is essential for tight junction formation [14] , [34] . Does phosphorylation of the AMA1 interfere with RON/AMA1 complex formation ? This appears unlikely as AMA1 without the cytoplasmic tail ( like wild type AMA1 ) remains capable of interacting with RON4 [18] . It remains to be determined whether the phosphorylation of the AMA1 tail by PKA has a direct effect on an AMA1 binding function or whether this event is a key step in signal transduction pathway . It is interesting to note that the AMA1-S610A-TY1 mutant invades at 20% of AMA1-WT-TY1 levels in the presence of R1 peptide whereas in the absence of a complementing W2mef protein the 3D7 parasites only invade with ∼4% efficiency . This indicates that there is not an absolute need for S610 phosphorylation for every successful invasion or that there is some cross-talk between the native 3D7 protein whose tail can presumably still be phosphorylated and the W2mef AMA1-S610A-TY1 mutant that cannot be phosphorylated but whose ectodomains probably remain functional . It is interesting to consider the likelihood of a link between cAMP signalling and calcium signalling during merozoite invasion . Both the Ono et al . study in sporozoites [32] and earlier work in P . falciparum blood-stages [20] have demonstrated an inter-relationship between these two pathways and indeed in higher order eukaryotes this relationship is now well established ( reviewed in [35] ) . How these are involved in invasion by merozoites and sporozoites is unclear although it appears likely that cAMP signalling operates upstream of intracellular calcium release [20] , [32] . In summary , this study opens up a new area of investigation for those interested in understanding P . falciparum merozoite invasion . Knowledge of the molecular events that trigger PKA signalling as well as those that follow as a consequence of AMA1 phosphorylation will be essential to a full understanding of merozoite invasion .
The culture of malaria parasites using donated blood and serum from the Australian Red Cross Society has been approved by The Walter and Eliza Hall Institute Human Research Ethics Committee and by the Bernhard Nocht Institute . The use of this material follows long-standing protocols and has not been associated with any adverse or other unforeseen events and no data of relevance to specific patients has been generated . A Supply Agreement has been executed between the Australian Red Cross Society and the Walter and Eliza Hall Institute and between the Australian Red Cross Society and the Burnet Institute covering the provision of blood and blood products for non-clinical use . Magnetic cell-sorted P . falciparum 3D7 parasites were cultured in the absence of RBCs until about half of the schizonts had ruptured to release their merozoites . A schizont enriched fraction was prepared by centrifuging the parasite culture at 1500×g for 5 min to pellet the schizonts . To harvest the merozoites , the supernatant was then spun at 3000×g for 15 mins and enrichment process was checked by microscopy of Giemsa-stained smears of the schizonts and merozoites . Schizonts were then released from host cells by saponin lysis . To make parasite lysates with kinase activity , the parasite pellets were lysed in 10 volumes of ice cold buffer B ( 50 mM β-glycerophosphate pH 7 . 3 , 1% Triton X-100 , 1 mM DTT , Complete Protease Inhibitor Cocktail [Roche] and phosphatase inhibitors , 50 mM Na3VO4 and 50 mM NaF , ) and incubated on a rotating wheel overnight at 4°C . Control extracts were made from similar amounts of whole uninfected red blood cells under identical conditions . To clear the lysates they were centrifuged at 16000×g and the protein concentration of the supernatant was determined by BCA protein assay ( Pierce ) . 3 µg total protein lysate ( or 1 µL of immuno-precipitated PfPKA-HA-sepharose ) was mixed with ∼10 µg of the GST or GST-AMA1-tail fusion proteins immobilized on glutathione-sepharose beads . To the beads buffer B containing different compounds was added to a final volume of 100 µL . The compounds used were 1 . 5 mM EGTA/1 mM EDTA , 2 mM CaCl2 , 1 µM cAMP , 10 µM KT 5720 , 50 µM H89 and 1% methanol for a control . The concentration of stock solutions , solvents and replicate numbers are indicated in the supplementary data ( Text S1 ) . The reaction was initiated by the addition of 3 µL ATP mixture composed of 100 µL 1 M MgCl2 and 150 µCi 32[P]ATP ( 10 µCi/mL ) . After 30 minutes at 30°C , the immobilised fusion proteins were pelleted and washed in buffer B . The AMA1-tail was removed from the beads as whole GST-AMA1-tail fusion protein with 10 mM glutathione in PBS or the tail only was cleaved from the GST with thrombin . Eluted proteins were immediately suspended in SDS gel sample buffer and were boiled for 5 min , resolved on a 4–12% gradient SDS-PAGE gel , blotted onto a PVDF membrane . Imaging of 32[P]-labelled protein bands was achieved by direct autoradiography ( 1–2 day exposure ) of dry blots using FUJIFILM BAS-TR2040 tritium imaging plates and a FLA-3000 luminometric detection system . PKA-HA was immunoprecipitated using anti-HA antibodies ( 5 µg/mL ) and protein-G-sepharose ( Invitrogen ) . 20 mL of a 10% synchronized schizont culture ( PfPKA-HA expressing and wild type parasite strain ) were saponin lysed . The resulting parasite pellet was washed in icecold PBS and hypotonically lysed in 10 volumes of ice-cold buffer B without detergent by several passages through a needle and 2 hours incubation on a rotating wheel at 4°C . The membrane fraction was separated from the soluble proteins by centrifugation at 13 . 000×g for 30 minutes . The resulting supernatant was precleared with protein-G-sepharose beads . The precleared lysate was incubated with anti-HA antibodies with a final concentration of 5 µg/mL for 3 hours . 20 µL of protein-G-sepharose was used to precipitate the PKA-HA antibody complexes . It was washed 5x with 10 volumes of cold PBS and subsequently stored in buffer B at −80°C until used . For each experiment , 3 µL of PKA-HA coupled beads were used . Parasite erythrocyte invasion assays were performed using 3D7 and transgenic parasites AMA1-WT-TY1 , AMA1-S610A-TY1 and AMA1-PM-TY1 ( 3D7 background ) . Parasitemia of sorbitol synchronized parasite culture was measured using the FACS . For the experiment a parasite culture with 0 . 5–1% parasitemia of late trophozoites ( 4% hematocrit ) was incubated in a 96-well Plate ( 100 µL per well ) under standard culturing conditions for 48 hours to allow re-infection in the presence or absence ( control ) of 100 mg/mL R1 . After reinvasion occurred , parasites were stained with 1 mg/mL ethidium bromide for 30 minutes at 37°C , washed three times with media and then counted using the FACS . Assays were performed in triplicates on three independent occasions . Metabolic labelling of phosphorylated parasite protein was achieved by incubating ∼1×1010 sorbitol-synchronized P . falciparum 3D7 parasites with 100 µCi/mL of 32[P]-monosodium phosphate ( Perkin-Elmer ) in 50 mL phosphate-free RPMI medium ( Gibco ) supplemented with 25 mM HEPES ( pH 7 . 2 ) , 0 . 5% ( w/v ) Albumax , 0 . 4% ( w/v ) glucose , 0 . 2% ( w/v ) Na2HCO3 at 37°C overnight . Late blood-stage parasites were released from host cells by saponin lysis , extensively washed in TBS , and parasite pellets extracted in 10 volumes of HNET ( 25 mM HEPES , pH 7 . 4 , 150 mM NaCl , 5 mM EDTA , 1% Triton X-100 ) for one hour with vortexing . Detergent-soluble schizont extract was clarified by ultracentrifugation at 75 , 000×g for 30 mins at 4°C . All procedures were carried out in the presence of protease and phosphatase inhibitors on ice , unless otherwise stated . 32[P]-labelled parasite extracts were snap-frozen in liquid nitrogen and stored at -80°C until required . Two-dimensional gel electrophoresis was performed using conditions required for optimal extraction and separation of P . falciparum-infected erythrocyte proteins [36] . Frozen parasite extracts were processed using 2-D Clean-Up Kit ( GE Healthcare ) . The resulting precipitates ( ∼100 mg protein ) were redissolved in 300 µL 2-DE sample buffer ( 7 M urea , 2 M thiourea , 2% ASB-14 , 1% DTT , 1% ampholytes ) , loaded onto 13 cm pI 4–7 IPG strips by passive re-hydration for 12 hours , and focussed using a fast voltage gradient ( 8000V max , 24 , 000 Vh ) at 15°C , using an Ettan IPGphor 3 system ( GE Healthcare ) . The second dimension was carried out on 7 . 5% polyacrylamide gels using a Hoefer SE 600 system ( GE Healthcare ) at 75V overnight . 2-DE gels were electrophoretically transferred onto Immobilon-PSQ PVDF membranes ( Millipore ) in Towbin's transfer buffer containing 20% methanol and 0 . 01% SDS . Complete transfer of total protein was confirmed using Deep Purple protein stain ( GE Healthcare ) . Imaging of 32[P]-labelled parasite phospho-protein spots was achieved by direct autoradiography of dry blots using FUJIFILM BAS-TR2040 tritium imaging plates and a FLA-3000 luminometric detection system ( 14 day exposure ) . 2DE Western blot analyses of protein extracts of 3D7 parasites expressing wild type and/or transgenic AMA1-WT-TY1 or AMA1-S610A-TY1 was carried out as detailed in the figure legend . The ama1 gene was either amplified from 3D7 or W2mef P . falciparum gDNA ( Table S1 ) . For the generation of GST fusion proteins the DNA sequence of the PfAMA1 C-terminal tail was cloned into BamHI and EcoRI restriction sites of the bacterial expression vector pGEX-4T-3 ( Pharmacia Biotech ) . This construct produces fusion proteins of PfAMA1 tail C-terminally bound to glutathione S-transferase ( GST ) . Different mutants of the GST-PfAMA1 tail fusion protein were achieved by using a site-directed mutagenesis kit ( Stratagene ) and sequences were confirmed by sequencing ( Table S1 ) . For transfecting 3D7 parasites ama1 was cloned into the KpnI and AvrII restriction sites of the pARL-AMA1-GFP Vector and sequences were confirmed by sequencing . To ensure correct timing of transcription , expression of the AMA1 transgenes were controlled by the AMA1 promoter . In vitro mutagenesis of ama1 was achieved by using a two-step primer directed PCR mutagenesis method ( Table S1 ) with proof reading Vent polymerase ( NEB ) . P . falciparum asexual stages were cultured in human 0+ erythrocytes according to standard procedures . W2mef is derived from the Indochina III/CDC strain . 3D7 parasites were transfected with 100 µg of purified plasmid DNA . Positive selection for transfectants was achieved using 10 nM WR99210 , an antifolate that selects for the presence of the human dhfr gene . For further and more detailed information see Text S1 . | The invasion of host cells by zoites of the phylum apicomplexa is an active event that is powered by the parasite invasion machinery . It can be divided in several distinct steps that involve binding to the host cell , reorientation and tight junction formation that are accompanied by sequential secretion of specialised organelles that store proteins involved in these events . A great number of proteins are now known to be involved in invasion but how the invasion process is regulated remains obscure . Recently , phosphorylation of some proteins with a defined function in invasion like GAP45 , MTIP and AMA1 were reported and provided the first insight into putative regulation mechanism of invasion . Using mutational analysis we now demonstrate that AMA1 is phosphorylated in the cytoplasmic domain at serine 610 in a cAMP dependent manner and that mutation of S610 dramatically reduces the efficiency of invasion into erythrocytes . We identified protein kinase A ( PfPKA ) as a late transcribed kinase that is responsible for phosphorylating AMA1 at this specific residue . This work describes for the first time PKA signalling being implicated in merozoite invasion , providing a new avenue for understanding the initiation and regulation of invasion . Significantly also , the PKA-AMA1 pathway defines a promising new and validated drug target for therapeutic intervention . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/protozoal",
"infections",
"biochemistry/cell",
"signaling",
"and",
"trafficking",
"structures",
"microbiology/parasitology"
] | 2010 | Protein Kinase A Dependent Phosphorylation of Apical Membrane Antigen 1 Plays an Important Role in Erythrocyte Invasion by the Malaria Parasite |
Several germline single nucleotide polymorphisms ( SNPs ) have been identified in the POLB gene , but little is known about their cellular and biochemical impact . DNA Polymerase β ( Pol β ) , encoded by the POLB gene , is the main gap-filling polymerase involved in base excision repair ( BER ) , a pathway that protects the genome from the consequences of oxidative DNA damage . In this study we tested the hypothesis that expression of the POLB germline coding SNP ( rs3136797 ) in mammalian cells could induce a cancerous phenotype . Expression of this SNP in both human and mouse cells induced double-strand breaks , chromosomal aberrations , and cellular transformation . Following treatment with an alkylating agent , cells expressing this coding SNP accumulated BER intermediate substrates , including single-strand and double-strand breaks . The rs3136797 SNP encodes the P242R variant Pol β protein and biochemical analysis showed that P242R protein had a slower catalytic rate than WT , although P242R binds DNA similarly to WT . Our results suggest that people who carry the rs3136797 germline SNP may be at an increased risk for cancer susceptibility .
DNA Polymerase β ( Pol β ) is the main polymerase involved in the base excision repair pathway ( BER ) , the pathway responsible for repairing up to 20 , 000 endogenous lesions per cell per day [1] , [2] . Pol β is a bifunctional polymerase , containing both deoxyribose phosphate ( dRP ) lyase and nucleotidyl transferase activities ( reviewed in [3] ) . One or both of these activities are essential , as Pol β knockout mice die shortly after birth [4] . Two germline SNPs of the POLB gene ( rs12678588and rs3136797 ) have been previously identified , and the variant alleles have been shown to be present in specific populations [5] , [6] . The rs12678588 SNP results in a nonsynonymous amino acid substitution of glutamine for arginine at residue 137 ( R137Q ) . In the wild-type ( WT ) protein , Arg137 is methylated by the protein arginine N-methyltransferase 1 ( PRMT1 ) , leading to a reduction in proliferating cell nuclear antigen ( PCNA ) binding [7] . R137Q is a slow polymerase with decreased BER activity in cell extracts , and cells expressing this variant have increased formation of AP sites following methyl methanesulfonate ( MMS ) exposure [8] . Little is known about the biochemical and cellular characteristics of the rs3136707 SNP , in which the proline at residue 242 is altered to arginine ( P242R ) or its role in human health . Carriers of this allele include populations from Eastern Europe [6] . Interestingly , patients heterozygous for this allele exhibited decreased survival when treated for either lung cancer or lymphoma [9] , [10] . Additionally , this residue is located at the base of Loop II , a region that has been shown by us to be critical for polymerase activity and fidelity [11]–[13] . In this study , we tested the hypothesis that the P242R germline POLB variant has a functional phenotype that could drive carcinogenesis . We found that expression of a cDNA encoding the P242R protein in both human and mouse cells induce chromosomal aberrations and cellular transformation . We also show that purified P242R protein is a slow polymerase that binds DNA tightly . In combination , our results suggest that cells expressing P242R accumulate BER intermediates that result in the induction of DSBs and chromosomal aberrations that lead to cellular transformation . Our results also indicate that the P242R germline variant of Pol β could result in aberrant BER in carriers of the allele , potentially leading to increased cancer predisposition .
Previous work on Pol β has shown that expression of certain tumor-specific single amino acid variants can induce genomic instability in the form of chromosomal aberrations [14] , [15] . Therefore , we characterized chromosomal aberrations in MCF10A normal human epithelial cells expressing the germline variant P242R . We generated stable MCF10A cell lines expressing C-terminally HA-tagged Pol β-WT or P242R at equal levels to the endogenous WT protein ( Figure S1A ) and scored metaphase spreads from each of these lines . Figure 1A–1B shows examples of metaphase spreads from MCF10A cells expressing WT or P242R Pol β , respectively . MCF10A cells expressing P242R had increased amounts of chromosomal fusions and fragments compared to cells expressing the WT ( Figure 1C ) . To determine if P242R was a mutator in cells , we performed the λcII forward mutation assay using a C127λb clonal cell line as we describe [16] . Briefly , these cells carry the λ genome and express either P242R or WT Pol β . The phage λ genome is packaged from isolated genomic DNA and the mutant frequency is obtained by infection of E . coli using differential plating conditions as we describe [16] . The spectra of mutations induced by P242R and WT Pol β are generated by sequencing purified plaques . We found that expression of P242R does not induce an increased frequency of point mutations nor a mutation spectrum different from that of WT Pol β , suggesting that it is not a mutator polymerase ( Table S1 and Figure S2 ) . Therefore , our results suggest that expression of P242R induces genomic instability in the form of chromosomal aberrations . A key role of Pol β is to fill gaps that arise from excision of DNA damage during BER . Chromosomal aberrations are known to arise from breaks in the DNA and aberrant BER can lead to the accumulation of DNA breaks [17] . Therefore , we wished to determine whether treatment with the alkylating agent methylmethane sulfonate ( MMS ) induces the formation of SSB and DSB BER intermediates in MCF10A cells expressing P242R . MMS induces DNA base damage that is repaired by the BER pathway and therefore treatment of cells with MMS leads to an increase in substrates recognized by Pol β . We used the alkaline comet assay to quantify SSBs induced by MMS . Following treatment with MMS for 30 min , cells expressing P242R have a significantly higher level of SSBs than cells expressing WT Pol β ( Figure 2A ) ( p<0 . 001 ) . Even after cells were allowed to recover from the MMS treatment for 30 or 60 min , greater levels of SSBs were still observed in P242R cells compared to WT cells ( p<0 . 001 ) . In fact , the percentage of tail DNA was increased in P242R cells allowed to recover compared to P242R cells treated with MMS for 30 min without any recovery ( p<0 . 001 ) suggesting that cells expressing P242R were unable to efficiently repair the damage or were continuing to accumulate damage . Although the alkaline comet assay measures SSBs , DSBs could also be present . To determine if this was the case , we treated cells with MMS for 30 min and allowed them to recover as in the comet assay , except we monitored γH2AX staining as an indicator of DSB formation . Two to about five percent of cells expressing either WT or P242R exhibited DSBs when treated with MMS for 30 followed by a recovery period for 0 , 30 or 60 min . ( Figure 2B ) . Because the levels of DSBs are similar in cells expressing WT or P242R using this protocol , our results suggest that the comet assay is not predominantly detecting DSBs . Importantly , DSBs are not observed to accumulate in P242R-expressing cells unless they are treated for at least 2 hrs with MMS . We then treated the MCF10A pools expressing WT or P242R with MMS for two hours and analyzed γH2AX staining as an indicator of DSB formation . We also stained with propidium iodide for cell cycle analysis . We found that exposure to MMS for 2 hours increased the levels of DSBs in both WT and P242R-expressing cells although MMS induced significantly more DSBs in P242R cells compared to WT Pol β ( Figure 2C ) ( p<0 . 001 ) . Moreover , cells expressing WT Pol β began to repair the damage after 2 hours following treatment ( p<0 . 05 ) whereas cells expressing P242R continued to exhibit γH2AX staining , suggesting that these cells continued to accumulate DSBs longer and/or had delayed repair compared to cells expressing WT Pol β . In fact , WT Pol β expressing cells are able to repair the damage after 2–4 hours and return to background levels of γH2AX staining whereas even after a 4 hour recovery , γH2AX positive foci were observed in cells expressing P242R Pol β ( Figure S3 ) . Strikingly , we also observe a significant increase in γH2AX staining in untreated cells expressing P242R Pol β compared to WT ( p<0 . 05 ) ( Figure 2C ) . The differences observed were not due to varying levels of expression as western analysis shows that WT and P242R were expressed at similar levels in these cells ( Figure S1A ) . Furthermore , although γH2AX positive cells were observed in all three phases of the cell cycle , cells expressing P242R had a significantly higher percentage of γH2AX cells in both S and G2/M phases ( Figure 2D ) ( p<0 . 001 ) suggesting that the DSBs may be formed , in part , during DNA replication . We tested the hypothesis that the genomic instability resulting from expression of the germline Pol β P242R protein induces cellular transformation . We generated clonal C127λb cell lines expressing exogenous HA-tagged human Pol β ( WT and P242R ) at approximately equal levels to endogenous Pol β in a tetracycline-repressible manner ( Figure S1B ) . In the focus formation assay , untransformed cells will grow to confluence forming a monolayer ( Figure 3A ) , while transformed cells will continue to grow after reaching confluence , forming foci ( Figure 3B ) . Expression of P242R Pol β induced cellular transformation whereas expression of WT Pol β did not ( Figure 3C–3D ) . To confirm transformation in these lines , we used a soft agar growth assay in which transformed cells that are capable of anchorage-independent growth will grow when plated on soft agar , while non-transformed cells will not . This assay confirms the results of the focus formation assay , showing that expression of P242R induces anchorage independent growth ( Figure 3E ) ( p<0 . 01 and 0 . 05 for P242R clones 5 and 15 , respectively ) . Next , we assessed anchorage independent growth in the human MCF10A cells . Similar to the C127λb clonal lines , MCF10A pools expressing P242R also displayed increased numbers of cells able to grow in soft agar compared to WT ( Figure 3F ) ( p<0 . 05 ) . Additionally , MCF10A cells expressing P242R had an increased rate of proliferation , another hallmark of cancer cells ( Figure 3G ) ( p<0 . 05 ) . Together , these data suggest that expression of Pol β P242R induces cellular transformation in both mouse and human cells . The accumulation of BER intermediates suggests that P242R may not function in the gap-filling step as well as WT Pol β . Expression of P242R in the Pol β−/− mouse embryonic fibroblasts ( MEFs ) partially rescued cellular survival in response to treatment with MMS , albeit not as well as expressing WT Pol β ( Figure 4A ) . The reduced ability of P242R to rescue cells from the effects of MMS compared to WT implies that P242R has a partially impaired BER function . However , the rs3136797 SNP , encoding P242R , is present as a heterozygous allele and rarely as a homozygous allele [6] . In addition , the cell lines we employed for our studies , namely , C127λb and MCF10A , both express WT Pol β . Therefore , we investigated whether expression of the P242R variant in the presence of WT Pol β sensitizes cells to MMS , as has been shown for the polymerase-dead E295K variant [15] . We conducted clonogenic survival assays using pools of Pol β+/+ MEFs and MCF10A cells expressing either WT or P242R Pol β or empty vector . Expression of P242R only slightly sensitized cells to high concentrations of MMS in a Pol β proficient background ( Figure 4B–4C ) . In combination with our chromosomal aberration studies , this suggests that some of the cells harboring genomic instability are likely to survive and could become transformed . The recombinant WT and P242R proteins were purified from E . coli and studied in a presteady-state burst assay . This assay used a radiolabeled 1 bp gapped DNA substrate , the preferred substrate for Pol β . Both proteins fit a biphasic burst of product formation , typical of Pol β activity ( Figure 5A ) . However , P242R had decreased initial burst ( kobs = 14±2 and 7 . 5±0 . 2 sec−1 , WT and P242R , respectively ) and steady-state rate compared to WT ( kss = 3 . 3±0 . 5 and 1 . 7±0 . 2 sec−1 , WT and P242R , respectively ) . This decreased activity was not due to decreased DNA binding , as the gel electrophoretic mobility shift assay showed that the P242R variant binds 1-bp gapped DNA with similar affinity to WT ( Kd = 5±1 and 5±1 nM , WT and P242R , respectively ) ( Figure 5B ) . Together , our data suggest that the slow rate of DNA synthesis catalyzed by P242R could result in unfilled gaps that lead to genomic instability and cellular transformation .
Pro242 of Pol β is located at the base of Loop II , a structure that we previously showed to be important for Pol β activity and fidelity [11] , [12] . Loop II is highly flexible , solvent exposed , and quite far away from the active site of Pol β . Mutations in amino acids within this loop result in decreased nucleotide discrimination and fidelity and induce mutations [11] , [12] , [18] , [19] . In addition , alterations in the amino acid sequence of the loop results in a reduced catalytic rate [12] . Pro242 is conserved between all members of the Pol X polymerase family [20] , [21] . Prolines can cause rigidity in the protein structure and this characteristic may help to anchor this flexible loop . We previously proposed that the specific geometry of this loop , and particularly of residue 242 , is important for maintaining the β-sheet structure of the Pol β active site . A change from Pro to Arg could disrupt the overall structure of the loop . Thus , it is not surprising to find that alteration of residue 242 from Pro to Arg results in an enzyme with low DNA polymerase activity . Cancer is a disease of aging . We suggest that slow accumulation of genomic instability over 50–60 years may occur in people carrying the P242R germline variant and that genomic instability could lead to cancer . P242R catalyzes single nucleotide gap filling at a rate half that of WT Pol β , although the protein binds to DNA with similar affinity as WT Pol β ( Figure 5 ) . Because P242R is a slow polymerase , gaps and SSBs undergoing repair with P242R may accumulate at levels somewhat higher than those repaired by WT Pol β . Indeed , treatment of cells with MMS for 30 min leads to significantly increased SSB levels in cells expressing P242R versus WT ( Figure 2A ) , even after the cells have had a chance to recover . At pH>12 . 6 , the alkaline comet assay detects SSBs , BER intermediates from incomplete repair , and alkali-labile sites [20] , [21] , but others report that it also detect DSBs [22] . Monitoring of γH2AX under our conditions suggests the presence of few DSBs ( Figure 2B ) . Our results suggest that SSBs and single nucleotide gaps accumulate in P242R-expressing cells , likely as a result of the slow gap filling activity of P242R . Treatment of cells for 2 hours with MMS results in the appearance of greater levels of DSBs in cells expressing P242R versus WT Pol β , and we show that they form predominantly during S-phase . These results suggest after two hours of MMS treatment , the BER system becomes overwhelmed with DNA damage , resulting in fewer gaps being filled by Pol β and leading to DSB formation upon encounter of the replication fork by a gap . However , cells expressing WT Pol β recover from this treatment more quickly than cells expressing P242R . This is likely due to deficient gap filling by P242R . Our results are also consistent with the possibility that treatment with MMS for 2 hours induces frank DSBs that are more rapidly repaired by WT versus P242R Pol β . We do not favor this explanation because to date there is no evidence for Pol β having a direct role in the repair of DSBs . In addition , our cell cycle data suggests that the DSBs accumulate during S-phase , which is consistent with the idea that they originate as a result of replication of a break or gap in the DNA . This mechanism has been suggested to explain similar findings that have been reported in Pol β−/− cells , where MMS treatment in G1 leads to unrepaired gaps that can form DSBs when cells enter G2/M [23] . Additionally , treatment of cells with a low dose of MMS and a PARP inhibitor , which effectively inhibits BER , resulted in DSB formation primarily during S-phase [24] . We have also shown that expression of P242R induces an elevated frequency of chromosomal aberrations compared to cells expressing WT Pol β ( Figure 1 ) . We suggest that these aberrations arise as the result of the persistent accumulation of BER intermediates in cultured cells ( Figure 2 ) , and that the presence of this type of genomic instability leads to cellular transformation ( Figure 3 ) . We show that cells expressing P242R undergo cellular transformation as demonstrated by formation of foci , anchorage-independent growth , and an increased rate of proliferation . The increase in proliferation suggests that critical cell cycle check points may be disrupted and cells are growing without effectively repairing the DNA . This uncontrolled cellular growth and replication of damaged DNA can induce the genomic instability we observe in these cells . For many of the experiments we performed we increased the levels of DNA damage over endogenous levels by treating with MMS , which would be expected to increase the amounts of Pol β substrates , namely , single nucleotide gaps in cells , in order to be able to detect breaks and aberrations . However , we find that the levels of endogenous DSBs are increased in cells expressing P242R compared to cells expressing only WT Pol β ( Figure 2B ) . Since Pol β is responsible for repairing at least 20 , 000 lesions/cell/day of endogenous damage , it is probable that humans expressing the P242R SNP may have more unresolved lesions compared to those with two WT alleles . Our finding of endogenous DSBs in cells expressing P242R suggests that over time , cells harboring this variant would incur more DSBs than cells without P242R even in the absence of exogenous DNA damage . We envision that the majority of DSBs are repaired accurately but that some of them are not , leading to deletions and insertions if repaired by non-homologous end joining , gene fusions , or other types of genomic instability , which could lead to cancer . Environmental exposures and/or diagnostic procedures over a P242R-carrying individual's lifetime could serve to enhance the rate or levels of genomic instability and perhaps decrease the latency of cancer . Given that the P242R variant is rare ( maximum allele frequency 2 . 4% ) [6] , epidemiologic studies have had limited success determining the role of this variant in human health [25] . It has been suggested that the 242Arg allele is associated with poor prognosis in lung cancer and lymphoma patients [9] , [10] , but the mechanism is unknown . One possibility is that the slow rate of DNA synthesis of P242R might be expected to enhance cancer cell death in the presence of DNA damage induced by radio- and chemotherapies , because many DNA gaps would remain unfilled . Indeed , we have shown that expression of P242R in Pol β-deficient MEFs does not rescue cells as completely as WT Pol β . This suggests that in the presence of P242R alone , for example in a homozygote who carries two alleles of P242R , would lead to cell death as a result of treatment with alkylating agents . However , expression of P242R in the presence of WT Pol β , as would be the case with a heterozygotic individual , only slightly sensitizes cells to high concentrations of MMS ( Figure 4 ) . This suggests that the majority of cells expressing both P242R and WT survive treatment with alkylating agents . We suggest that these survivors could have increased levels of genomic instability , based upon our results showing that when treated with MMS , cells expressing both P242R and WT Pol β accumulate BER intermediates and have an increased frequency of chromosomal aberrations . Thus , treatment of cells expressing both of these proteins could lead to cellular transformation or more aggressive disease . In conclusion , our results show that the presence of the P242R germline variant contributes to the increase in chromosomal aberrations and cellular transformation . Therefore , individuals carrying this germline variant may have increased cancer susceptibility , suggesting that aberrant BER at the level of the germline could be a driver of carcinogenesis .
All ultrapure deoxynucleoside triphosphates ( dNTPs ) were purchased from New England Biolabs . [γ-32P] ATP ( 5 mCi ) and ATP were purchased from Amersham Biosciences and Sigma-Aldrich , respectively . All oligonucleotides used for the in vitro biochemical assays were purchased from Keck Biotechnology Research Center at Yale University and purified as described [26] . All oligonucleotides used for cloning and PCR were purchased from Invitrogen and are shown in Table S2 . Human Pol β cDNA ( Genbank accession NM_002690 ) was cloned into the pET28a expression plasmid ( Novagen ) for expression with an N-terminal 6×His tag . WT Pol β cDNA sequence was verified by sequencing at the Keck DNA Sequencing Facility at Yale University . For cell culture experiments , human Pol β cDNA with a C-terminal hemagluttinin ( HA ) tag was cloned into the pRVYTet-Sis retroviral vector as described [14][15] . The 242Arg variant was introduced into the human WT ( 242Pro ) Pol β cDNA sequence using site-directed mutagenesis ( Stratagene ) following the manufacturer's protocols . For cloning of Pol β , E . coli DH5αMCR with the genotype mcrA ( mrr-hsdRMS-mcrBC ) φ80ΔlacZ ( M15 ) ( lacZYA-argF ) U169 deoR recA1 endA1 phoA supE44 thi-1 gyrA96 relA1 was used . Human Pol β was expressed in Rosetta ( DE3 ) cells ( Novagen ) . For the λcII forward mutagenesis assay , lysogen strain NM759 [27] was used for the preparation of sonication extracts and BHB2688 [28] , [29] was used for the freeze-thaw extracts . E . coli strain G1250 hflA::Tn5 hflB29 was used for selection and harvesting of packaged phage harboring mutations in the cII gene . Mouse embryonic fibroblast ( MEF ) cell lines 92TAg ( Pol β+/+ ) and 88TAg ( Pol β−/− ) were gifts from Leona Samson ( Massachusetts Institute of Technology ) [30] , [31] . These cells were maintained in high-glucose Dulbecco modified Eagle's medium ( Invitrogen ) supplemented with 10% fetal bovine serum ( Invitrogen ) , 1% penicillin-streptomycin ( Invitrogen ) and 1% L-glutamine and grown at 37°C in a 5% CO2 humidified incubator . MCF10A cells are immortalized , non-transformed epithelial cells derived from human mammary tissue ( ATCC ) . These cells were maintained in DMEM/F12 medium ( Invitrogen ) supplemented with 5% horse serum ( Invitrogen ) , 1% penicillin-streptomycin , epidermal growth factor ( 20 ng/ml ) , hydrocortisone ( 0 . 5 µg/ml ) , cholera toxin ( 100 ng/ml ) , insulin ( 10 µg/ml ) ( Sigma-Aldrich ) and grown at 37°C in a 5% CO2 humidified incubator . C127 cells have been described [32] . The C127λ cells were made by a procedure similar to the one described in [28] with the following exceptions . The λsup-Fneo vector ( kind gift from Dr . Peter Glazer , Yale University School of Medicine ) was transfected into C127 cells using FuGene 6 ( Roche ) . Single clones were selected in 900 mg/mL of G418 and expanded . A dot blot was used as described [28] to identify clones carrying the λsup-Fneo DNA . The C127λb clone was used in the experiments described here . C127λb cells were grown in Dulbecco modified Eagle's medium ( Invitrogen ) supplemented with 10% fetal bovine serum ( Invitrogen ) , 1% penicillin-streptomycin ( Invitrogen ) , and 600 µg/ml G418 at 37°C in a 5% CO2 humidified incubator . The GP2-293 virus packaging cell line ( Clontech ) was used for retrovirus preparation . These cells were maintained in Dulbecco modified Eagle's medium ( Invitrogen ) supplemented with 10% fetal bovine serum ( Invitrogen ) , 1% L-glutamine ( Invitrogen ) , 1% penicillin-streptomycin ( Invitrogen ) and 1 mM HEPES ( Invitrogen ) . Human Pol β WT and P242R constructs were packaged into retrovirus using the GP2-293 packaging line . pRVYTet and pVSV-G plasmids were co-transfected into GP2-293 cells using standard calcium phosphate transfection , cells were grown for 72 hours , and retrovirus was harvested . To infect C127λb , 92TAg ( Pol β+/+ ) , 88TAg cells ( Pol β−/− ) , and MCF10A , cells were grown to approximately 30% confluence and infected with retrovirus in the presence of 4 µg/ml polybrene . Cells were incubated overnight in fresh media with 4 µg/ml polybrene . For selection of pools , cells were split 1∶3 the day after infection and cells with the integrated construct were selected with 220 µg/ml hygromycin B for the C127λb and MEFs and 15 µg/ml hygromycin B for the MCF10A cells . For generation of stable clones , C127λb cells were split at several dilutions following infection and selected with 250 µg/ml hygromycin B . Single cell clones were grown and selected using cloning rings . Clonal cell lines were propagated in the presence of 160 µg/ml hygromycin B . Expression of exogenous HA-tagged Pol β was verified by Western blot . Cells were passed in parallel in the presence or absence of tetracycline . Approximately 80–90% confluent cells were harvested by scraping with hot SDS Loading Buffer ( 50 mM Tris pH 6 . 8 , 100 mM DTT , 2% SDS 10% glycerol ) . Lysates were boiled for 10 minutes and run on a 10% acrylamide SDS-PAGE gel . Proteins were transferred to nitrocellulose membrane using a semi-dry transfer apparatus and probed using monoclonal mouse anti-Pol β antibody ( Abcam #1831 ) . Four 10 cm dishes were seeded with 106 cells per dish each per cell line and grown at 37°C 5% CO2 overnight . The cells were fed fresh medium and colcemid ( Invitrogen ) was added to a final concentration of 100 ng/ml . MCF10A pools were incubated in colcemid for three hours before harvesting by mitotic shakeoff . Cells were harvested via centrifugation , washed twice with 1× PBS , and resuspended dropwise in 0 . 8% sodium citrate . Following lysis , cells were incubated at 37°C for 30 minutes before fixing in Carnoy's Fixative ( 75% methanol , 25% acetic acid ) . Finally , cells were dropped onto microscope slides , dried and stained with 5% KaryoMax Giemsa stain ( Invitrogen ) . Well-spread metaphases were identified under 100× objective ( Zeiss ) . Images were taken using Spot Camera software ( Diagnostic Instruments ) . Metaphase spreads were de-identified and scored by eye for chromosomal fusions , breaks , acentromeric chromosomes and fragments . Equal numbers of cells ( 4×105 ) were plated in 60 mm dishes . The following day , the cells were treated with 2 mM MMS for 30 min . After treatment , the cells were prepared and analyzed immediately according to published procedures [33] using Cometslides ( Trevigen Cat # 4250-200-03 ) . Image analysis of 100–125 cells was performed using CometScore software ( TriTek , Sumerduck , VA ) . Data are represented as mean ± SEM ( n = ) . MCF10A cells expressing WT or P242R Pol β were untreated or treated with 2 mM MMS for 2 h . Cells were rinsed then with PBS and replaced with fresh media . Cells were allowed to recover for 0 h and 2 h post treatment . Cells were harvested by trypsinization , washed once with PBS , and pelleted . The pellet was resuspended by adding 70% ice cold ethanol dropwise while vortexing . Cells were fixed overnight at −20°C . The cells were incubated with primary phospho-γH2AX antibody ( Millipore 05-636 ) 1∶500 overnight at 4°C . Following the incubation , cells were washed twice with PBS and incubated with anti-mouse secondary antibody conjugated to FITC 1∶500 for 1 h at room temperature . Cells were washed twiced with PBS and resuspended in 500 µl PI/RNase staining buffer ( BD Pharmingen ) . Fluorescence was analyzed by flow cytometry using the BD FACSCalibur and analyzed using FlowJo 8 . 8 . 6 software . The focus formation assay was conducted as in [15] . Briefly , cells were passaged every 3–4 days in the presence of 160 µg/ml hygromycin B and the presence ( non-induced ) or absence ( induced ) of 3 µg/ml tetracycline . Every four passages , 1×10 5 cells were seeded into each of four T25 flasks . These cells were fed every 3–4 days with media containing or lacking 3 µg/ml tetracycline . After 21 days , the cells were stained with 0 . 25% crystal violet . The presence of foci was also monitored by microscopic examination as described previously [14] , [15] . Anchorage independent growth was assessed as previously described [14] . Approximately 1×104 MCF10A cells were mixed with media containing 0 . 7% noble agar ( USB ) . This mixture was poured onto a layer of media containing 1 . 0% noble agar in a well of a 6-well dish . Cells were fed twice weekly for 4 weeks . The number of colonies present in each of five microscope fields per well from a total of 6 wells per experiment were counted after 4 weeks of growth . Cells ( 2 . 2×105 ) from MEF pools were seeded into 60 mm dishes and incubated for 48 hours . Cells were treated with varying concentrations of MMS for two hours followed by trypsinization and plating at dilutions . Treated cells were grown for 9–11 days before staining with 0 . 25% crystal violet . Colonies were scored by eye at 4× magnification . Only colonies with more than 50 cells were counted and all experiments were repeated at least twice . For MCF10A cells , various concentrations of cells were plated in 6 well dishes and were allowed to attach overnight . Cells were treated with varying concentrations of MMS for two hours . Following treatment , cells were rinsed with PBS and replaced with fresh media . Cells were allowed to grow for 12–14 days before staining with crystal violet as detailed above . MCF10A pools were plated at a density of 25 , 000 or 50 , 000 cells per 10 cm dish . Cells were counted every day for 5 consecutive days . Data were plotted as change in cell number per day . pET28a plasmids with human Pol β with 242Pro ( WT ) and 242Arg ( 242Arg ) cDNA were transformed into Rosetta ( DE3 ) cells . Protein expression was induced by addition of 1 . 0 mM isopropyl β-D-thiogalactopyranoside ( IPTG ) and grown at 37°C for 2 hours before harvesting via centrifugation . Cell pellets were dried and stored at −80°C . Protein induction was verified by using 10% SDS-PAGE stained with Coommassie Blue . Protein was purified using fast protein liquid chromatography . Crude proteins were run through a HiTrap Chelating HP column ( GE Healthcare ) charged with Ni2+ using a linear imidazole gradient ( from 5 to 500 mM ) in 40 mM Tris-HCl pH 8 . 0 , 500 mM NaCl . The His-fusion proteins eluted at approximately 277 mM imidazole over five or six fractions of two ml each . The fractions were combined , concentrated to less than one milliliter and diluted into nine milliliters low salt buffer ( 50 mM Tris-HCl pH 8 . 0 , 1 mM EDTA , 10% glycerol , 0 . 1 M NaCl ) . The diluted protein was then run on a SP HP column ( GE Healthcare ) using a linear NaCl gradient from 100 mM to 2000 mM . Purified protein fractions eluted at approximately 1000–1200 mM NaCl in two fractions , which were combined and concentrated to less than one milliliter . Glycerol was added to a final concentration of 10–15% and aliquots were flash frozen in liquid N2 and stored at −80°C . All proteins were purified to >90% homogeneity based on Coomassie Blue staining of 10% SDS-PAGE gels . Oligonucleotides were synthesized by the W . M . Keck facility and purified by polyacrylamide gel electrophoresis as described previously [34] . Briefly , primer oligos were radiolabelled with γ-32P ATP using T4 polynucleotide kinase ( New England Biolabs ) and downstream oligos were kinased using non-radioactive ATP . Kinased oligonucleotides were purified using Microspin columns ( Biorad ) . Primer , template , and downstream oligos were annealed by denaturing at 95°C for 5 minutes , slow cooling to 50°C for 30 minutes , holding at 50°C for 20 minutes , and then resting on ice . Complete substrate annealing was confirmed using 12% native polyacrylamide gel electrophoresis and visualized using autoradiography . Radiolabeled 1 bp gapped DNA ( 300 nM 45AG [35] ) and Pol β ( 100 nM ) were combined with the correct dNTP and 10 mM MgCl2 in a KinTek Chemical Quench-Flow apparatus at 37°C . The reactions were quenched by the addition of 0 . 5 M EDTA . The reaction products were separated on a 20% denaturing polyacrylamide gels , visualized , and quantified using a Storm 860 Phosphorimager with ImageQuant software . Data were fitted to the burst equation:where A is the amplitude , kobs is the observed rate constant of the exponential phase , and kss is the rate constant of the linear phase [36] . The DNA binding constant was determined by gel electrophoretic mobility shift assay as described previously [34] . The dissociation constant for DNA ( KD ) was determined by fitting the fraction bound protein ( Y ) versus protein concentration with the equation:wherer Y is the amount of bound protein , m is a scaling factor , and b is the apparent minimum Y value [37] . High molecular weight DNA was harvested from C127λb cells using standard phenol chloroform extraction follwed by dialysis against TE buffer as previously described [16] . Sonication and freeze-thaw packaging extracts were prepared as in [28] , [29] using NM759 and BHB2688 E . coli strains . To rescue the λ vector from the DNA , approximately 5 µg of high molecular weight genomic DNA was added to 60 µl of sonication extracts mixed with 40 µl of freeze-thaw extracts . Reactions were incubated at 32°C for 90 minutes . An additional 100 µl of packaging extracts was added and the reactions were incubated for an additional 90 minutes at 32°C . For cII gene mutation detection , the in vitro packaged phage were diluted , adsorbed in G1250 bacteria , and plated in 0 . 4% top agar on TB plates . Plates were incubated at 37°C overnight to determine the infection titre , because due to a termperature-sensitive mutation in the cII gene in the phage vector , all phage are lytic and form plaques at this termperature . Plates were also incubated at 24°C . The WT phage do not form plaques at this temperature whereas phage with mutations in the cII gene will . cII mutants were obtained from at least three independent packages . Mutant plaques were replated in fresh cultures . Phage DNA was harvested and used as a template for PCR . The cII genes were amplified by PCR and sequenced at the W . M . Keck Sequencing facility . Two-tailed t-tests and two-way analysis of variance ( ANOVA ) were used as appropriate to determine whether the mean of each cell line was different from the empty vector cells . Bonferroni's post hoc test was used to determine significant differences between the means of each group . All statistics were performed using GraphPad Prism version 5 ( GraphPad Software , San Diego , CA ) . Data are represented as mean ± SEM . | Cancer is the second leading cause of death in the United States . The maintenance of genomic integrity is dependent on faithful DNA replication and repair . The base excision repair ( BER ) pathway is responsible for repairing at least 20 , 000 lesions per cell per day . DNA polymerase beta ( Pol β ) is the main polymerase in the BER pathway and is mutated in up to 40% of human tumors . However , little is known regarding any germline mutations found in the human population . Here , we provide evidence that the germline variant of Pol β , P242R , has a slower catalytic rate than the wild-type Pol β . This reduced rate induces an increase in chromosomal aberrations , a type of genomic instability that can lead to cancer . We also show that expressing P242R in human cells induces cellular transformation , anchorage-independent growth , and an increased rate of proliferation: all hallmarks of tumorigenesis . Together , our data suggest that the germline Pol β variant can drive carcinogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"cancer",
"genetics",
"genetic",
"mutation",
"genetic",
"polymorphism",
"nucleic",
"acids",
"genetics",
"dna",
"biology",
"dna",
"repair",
"population",
"genetics",
"genetics",
"and",
"genomics"
] | 2012 | A Germline Polymorphism of DNA Polymerase Beta Induces Genomic Instability and Cellular Transformation |
Using small molecule probes to understand gene function is an attractive approach that allows functional characterization of genes that are dispensable in standard laboratory conditions and provides insight into the mode of action of these compounds . Using chemogenomic assays we previously identified yeast Crg1 , an uncharacterized SAM-dependent methyltransferase , as a novel interactor of the protein phosphatase inhibitor cantharidin . In this study we used a combinatorial approach that exploits contemporary high-throughput techniques available in Saccharomyces cerevisiae combined with rigorous biological follow-up to characterize the interaction of Crg1 with cantharidin . Biochemical analysis of this enzyme followed by a systematic analysis of the interactome and lipidome of CRG1 mutants revealed that Crg1 , a stress-responsive SAM-dependent methyltransferase , methylates cantharidin in vitro . Chemogenomic assays uncovered that lipid-related processes are essential for cantharidin resistance in cells sensitized by deletion of the CRG1 gene . Lipidome-wide analysis of mutants further showed that cantharidin induces alterations in glycerophospholipid and sphingolipid abundance in a Crg1-dependent manner . We propose that Crg1 is a small molecule methyltransferase important for maintaining lipid homeostasis in response to drug perturbation . This approach demonstrates the value of combining chemical genomics with other systems-based methods for characterizing proteins and elucidating previously unknown mechanisms of action of small molecule inhibitors .
Methyltransferases are a large class of enzymes comprising 0 . 6–1 . 6% of protein coding genes in most sequenced organisms [1] . S-adenosyl methionine ( SAM ) -dependent methyltransferases regulate a dynamic network of cellular signaling events and are required to maintain intracellular homeostasis in the face of external perturbations by catalyzing the methylation of a wide variety of substrates ( proteins , nucleic acids , lipids and small molecules ) [2]–[4] . The characterization and understanding of the roles of most methyltransferases remains challenging , however , due to their dispensability in standard growth conditions . Numerous studies from our lab and others have demonstrated that chemogenomic profiling of the Saccharomyces cerevisiae yeast deletion collection [5] is a powerful approach for the identification and subsequent characterization of genes required for growth in the presence of bioactive compounds [6]–[15] . Moreover , while most yeast genes ( ∼80% ) are dispensable for growth in standard laboratory conditions , the presence of chemical and/or environmental perturbations of the cell , 97% of the yeast genome exhibits a fitness defect that would not otherwise have been revealed [15] . Well-established chemogenomic assays in yeast , such as drug-induced Haploinsufficiency Profiling ( HIP ) , Homozygous Profiling ( HOP ) and Multicopy Suppression Profiling ( MSP ) are designed to identify small molecule-gene interactions . For example , HIP assay is used to detect compounds that target essential genes , and HOP and MSP are suitable for identification genetic modifiers of drug resistance [8]–[10] , [13] . The combination of these chemogenomic assays allowed us to identify a novel gene , YHR209W , that we subsequently named CRG1 ( Cantharidin Resistance Gene 1 ) , due to its requirement for growth in the presence of the small molecule cantharidin [14] . Specifically , both CRG1 heterozygous and homozygous deletion strains exhibited sensitivity to the drug , and the overexpression of CRG1 conferred resistance to the drug . Nonetheless , Crg1 is uncharacterized , except for annotation derived from large-scale analyses [15]–[17] . Based on its primary sequence , Crg1 is predicted to encode a Class I S-adenosyl-methionine ( SAM ) -dependent methyltransferase [18] . Crg1 shares close sequence homology with trans-aconitate methyltransferase Tmt1 ( BLAST-P expect value 2×10−31 and 3×10−34 for the full proteins and the methyltransferase domains , respectively ) . Tmt1 is known to modify and detoxify small molecules by methylation [19]–[21] . We have previously shown indirectly that Crg1 does not likely possess Tmt1 methyltransferase activity towards trans-aconitate , 3-isopropylmalate , and isopropylmaleate , indicating that these closely related proteins have divergent substrates [19] , [21] . Bioinformatics analysis from our group has shown , however , that Crg1 clusters with a family of eight methyltransferases based on their methyl-accepting substrate specificity , including Tmt1 , the lipid methyltransferases ( Coq3 , Coq5 , and Erg6 ) , and a tRNA methyltransferase , Trm9 [22] . All of these proteins methylate carboxylic acids present in small molecules to form methyl esters , suggesting that Crg1 might have a similar biochemical activity and catalyze the formation of a methyl ester . Cantharidin , a natural product produced by Chinese blister beetles of the Meloidae family of Coleoptera , is used in Traditional Chinese Medicine for the treatment of a variety of cancers [23] . Cantharidin has potent anticancer activity characterized by cell cycle arrest in G2/M phase , apoptosis , and DNA damage , presumably as a result of the generation of reactive oxygen species [23]–[29] , yet its use is limited due to renal and mucous membrane toxicity . Although the activity of cantharidin is usually attributed to its high affinity towards Type 1 and 2A serine/threonine protein phosphatases [30] , [31] , several studies suggest that cantharidin has additional cellular targets . Specifically , cantharidin has been reported to stimulate xanthine oxidase activity and to inhibit N-acyltransferase and cAMP phosphodiesterase in liver cells , suggesting a complex mode of action [32]–[34] . Using the HIP and HOP genome-wide assays , we discovered that a surprisingly large number of methyltransferase deletion mutants are sensitive to cantharidin , suggesting that , as a class , these enzymes may interact directly or indirectly with cantharidin and participate in the response to cantharidin stress [15] . Notably , among these methyltransferases only the overexpression of CRG1 is able to confer resistance to cantharidin . To further explore the function of Crg1 and the mechanism of cantharidin cytotoxicity , we employed chemical genomics tools combined with conventional biological techniques . We demonstrated that Crg1 methylates cantharidin in vitro , and identified cantharidin-specific CRG1 genetic interactors . To extend our chemogenomic results we analyzed the lipid profile of mutants grown in the presence of cantharidin , and demonstrated that cantharidin resistance involves Crg1-dependent maintenance of lipid homeostasis .
To confirm our published cantharidin-specific response of CRG1 [14] , we measured the growth of three strains 1 ) wild-type diploid strain BY4743 , 2 ) a crg1 Δ/Δ homozygous deletion strain , and 3 ) a crg1Δ/Δ homozygous deletion strain overexpressing CRG1 ( 2 µ plasmid ) as a function of cantharidin concentration . We observed that the gene dosage of the putative SAM-dependent methyltransferase CRG1 correlated with the sensitivity/resistance of these strains to cantharidin ( Figure 1A ) . In agreement with this gene-dose dependent effect , crg1Δ/CRG1 heterozygous mutants grew worse than the wild-type strain but better than a crg1Δ/Δ homozygous mutant in the presence of cantharidin ( 500 µM ) ( Figure S1A ) . We found that cantharidin is more potent against cells grown in synthetically defined ( SD ) medium than in YPD medium ( 5 µM and 250 µM , IC20 for wild-type in SD and YPD , respectively; Figure S1B ) . The observed differential drug sensitivity in defined media and rich YPD media is a common phenomenon in our drug screens ( unpublished data ) . We also tested structural analogues of cantharidin , including cantharidic acid and norcantharidin , and found that these compounds produced a similar gene-dose dependent response in crg1 mutants ( Figure S1C ) . Because our data suggested that CRG1 responds to cantharidin in a gene dose-dependent manner , we next tested whether the transcription of CRG1 is induced in the presence of the drug . qRT-PCR analysis showed that the relative abundance of CRG1 transcripts increased drastically in the wild-type strain after 60 min of the drug treatment ( 250 µM ) compared to the DMSO control ( P-value <0 . 02; Figure 1B , left panel ) . Importantly , a gene-dose dependent effect in the response to cantharidin was also observed for CRG1 transcript levels in crg1Δ/CRG1 heterozygous and CRG1-overexpressing mutants ( Figure S2A ) . In agreement with the qRT-PCR data , we also observed induction of Crg1 at the protein level . GFP-tagged Crg1 protein increased from undetectable levels prior to treatment and accumulated to high levels ( restricted to the cytoplasm ) following 1 hour of cantharidin treatment ( Figure S1B ) . Given the known high affinity of cantharidin towards Type 2A protein phosphatases ( PP2A ) and to a lesser degree towards Type 1 ( PP1 ) [30] , [31] , we tested if CRG1 induction was mediated by chemical inhibition of protein phosphatase function . We phenocopied cantharidin treatment using a panel of protein phosphatase homozygous deletion strains . Consistent with the results of chemical inhibition of protein phosphatases with cantharidin , we found that the homozygous deletion strains sit4Δ/Δ ( PP2A ) , ptc1Δ/Δ ( PP2C ) and the heterozygous deletion strain glc7Δ/GLC7 ( PP1 ) also resulted in transcriptional upregulation of CRG1 in the absence of cantharidin ( Figure 1B , right panel ) . It is important to note that perturbation of these protein phosphatases accounted for only ∼20% of the transcript induction observed by cantharidin . Furthermore , the treatment with calyculin A , a structurally distinct PP2 and PP1 inhibitor [35] , known to interact with the yeast PP1 GLC7 [14] , resulted in an increase of CRG1 transcript level to a similar degree as in glc7Δ/GLC7 mutant ( ∼2 . 5 fold; Figure S2C ) . This observation opens up the possibility that cantharidin acts independently of this PPase . This hypothesis is also supported by our observation that overexpression of GLC7 confers resistance to calyculin A , but not to cantharidin [14] . These results also suggest that these protein phosphatases are likely to be negative regulators of the cellular pathway regulating CRG1 induction . We also observed that cantharidin-induced transcription of CRG1 follows a temporal pattern characteristic of diverse environmental stress responses [36] , following a peak at 60 min of treatment the transcript levels began to decrease at 120 min ( ∼40 fold , P-value<0 . 01; Figure 1B ) . Indeed , a comprehensive genome-wide analysis of diverse environmental stresses from publicly available expression data [36] revealed that the transcription profile of CRG1 in diverse stress conditions correlates highly ( r = 0 . 8 ) with a well-characterized stress-responsive gene , the heat shock protein SSE2 ( Figure S2D ) , suggesting that CRG1 is also transcriptionally activated by other stress conditions in addition to cantharidin . Because CRG1 is annotated ( based on its amino acid sequence ) as a putative SAM-dependent methyltransferase [18] , we next asked whether its methyltransferase domain is required for cantharidin tolerance by mutating amino acids ( D44A , D67A , E105A-D108A ) within the conserved motifs ( Figure 1C ) . These amino acids have previously been shown to be critical for activity of other methyltransferases [37] . Overexpression of these crg1 site-specific mutants in a crg1Δ/Δ strain failed to confer cantharidin resistance while , in contrast , mutation of a non-conserved residue ( G96A ) in the methyltransferase domain showed resistance equivalent to wild-type CRG1 ( Figure 1C ) . The observed decrease in resistance to cantharidin was not due to reduced expression of the mutated Crg1 proteins ( Figure S2E ) , suggesting that the methyltransferase domain of Crg1 is both functional and important for cellular survival in the presence of the drug . To identify other potential cellular factors important for Crg1-mediated cantharidin resistance , we profiled the complete yeast transcriptome using whole-genome tiling microarrays . Transcriptional changes in wild type , crg1Δ/Δ deletion and CRG1-overexpressing crg1Δ/Δ strains were analyzed after 1 hour of exposure to the drug . To ensure that the transcriptome datasets for the different strains are comparable , the IC20 for wild type ( 250 µM ) was applied to all strains . Even at this high dose , the hypersensitive crg1Δ/Δ strain is viable , after 1 hour of exposure ( Figure S3A ) . When applied for an extended period , this dose is , in fact , inhibitory for growth of crg1Δ/Δ strains ( Figure 1A and Figure S1A ) . Furthermore , the treatment of crg1Δ/Δ strain with a lower dose ( 30 µM , the IC20 for this mutant ) resulted in quantitative difference in the transcriptome profile rather than in any qualitative differences , suggesting that the transcriptional changes are consistent across a range of concentrations . In particular , this observation was relevant to downregulated genes ( Figure S3B and S3C ) . It is also worth noting that the expression of most genes was not affected by crg1 mutation . To uncover cantharidin-specific genes in our transcriptome analysis , we eliminated Environmental Stress Response ( ESR ) genes known to be activated by a large number of stresses , such as genes required for vacuole biogenesis , response to stress , ribosome biogenesis , and RNA processing [36] . We also eliminated those genes that did not demonstrate at least two-fold difference in the presence of cantharidin or if their differential expression failed to show statistical significance . To detect genes and biological processes that are differentially expressed among the strains and treatments , the enrichment of genes for Gene Ontology ( GO ) term Biological process in the transcriptomes of wild type , crg1Δ/Δ and CRG1-overexpressing crg1Δ/Δ strains in the presence and absence of cantharidin were compared ( Table S1 ) . We also clustered genes according to their expression pattern . The clustering and GO term comparative analysis revealed that significantly downregulated genes ( log2 ( drug/DMSO ) <−1 , P-value <0 . 05 ) in crg1Δ/Δ mutant and the wild type were enriched for the genes of amino acid process ( multiple-testing corrected P-value <1 . 0×10−7 and P-value <7 . 0×10−7 , respectively ) , while the transcriptional profile of cantharidin-resistant CRG1-overexpressing crg1Δ/Δ strain did not demonstrate a similar enrichment ( Figure 1D and Figure S3D ) . Of particular interest , most of the genes that comprise methionine biosynthetic process ( MET6 , MET17 , MMP1 , STR3 , ADE3 , SAM1 , SAM2 , SAH1 , MET22 , MET31 ) were differentially expressed between cantharidin-resistant CRG1-overexpressing crg1Δ/Δ , wild type and cantharidin-sensitive crg1Δ/Δ strain in the presence of cantharidin ( P-value <0 . 05; Table S2; Figure 1E ) . One noteworthy example is STR3 , a cystathionine beta-lyase , the gene that demonstrated the most differential expression in the strains . STR3 was significantly induced by cantharidin in CRG1-overexpressing crg1Δ/Δ strain ( log2 ( drug/DMSO ) = 4 . 7 , P-value <0 . 022 ) and downregulated in the wild type ( log2 = −0 . 6 ) and crg1Δ/Δ ( log2 = −0 . 55 ) . The observed differential expression of STR3 was further confirmed by qRT-PCR ( Figure S3F ) . Str3 is of interest because it functions in methionine biosynthesis by converting cystathionine into homocysteine , a precursor for methionine , which is a substrate for the generation of SAM . SAM is required as a methyl donor for methylation reactions ( Figure 1E ) . To further explore the role of the Crg1 SAM-dependent methyltransferase , we treated wild-type cells with a combination of cantharidin and S-adenosyl homocysteine ( SAH ) , a non-specific methyltransferase inhibitor . We found that wild-type strains were more sensitive to the cantharidin/SAH combination compared to either single agent ( Figure S3F ) . These observations confirm the requirement of SAM-dependent methyltransferase activity in response to cantharidin , and suggest that Crg1 is a functional methyltransferase that catalyzes a SAM-dependent methylation reaction important for cantharidin resistance . Given that CRG1 provides cantharidin resistance in a gene dose-dependent manner and because of its close sequence homology to TMT1 , a small molecule methyltransferase that catalyzes the formation of methyl esters ( Figure 2A ) , we hypothesized that Crg1 might methylate cantharidin because this drug bears some structural similarity to the substrates of Tmt1 ( Figure 2B ) . To test this possibility , we performed in vitro biochemical assays with purified Crg1 , cantharidin , and S-adenosyl-[methyl-14C]methionine ( Figure 3A ) . These in vitro reactions were separated via reverse phase liquid chromatography and the radioactivity of the collected fractions was quantified with a scintillation counter . We detected a unique peak of radioactivity eluting in the 18–20 min fraction ( Figure 3B ) . The appearance of this peak was both cantharidin and Crg1-dependent , suggesting that it could correspond to methylated cantharidin . To confirm that the novel activity was catalyzed by Crg1 rather than by a co-purifying protein , we repeated the methylation reactions with mutant forms of Crg1 containing amino acid substitutions at critical residues within the methyltransferase domain . As described earlier , the D44A and E105A-D108A mutations abolished resistance to cantharidin ( Figure 1C ) , so we assessed whether these mutated proteins were able to methylate the drug molecule . We prepared in vitro reactions containing varying concentrations of cantharidin and quantified the amount of acid-labile volatile radioactivity because methyl esters are known to readily hydrolyze in both strongly acidic and basic conditions to yield methanol [19] , [38] . Unlike the reactions performed with wild-type Crg1 , addition of cantharidin to the reactions with mutant forms of the enzyme showed no increase in acid-labile radioactivity ( Figure 3C ) , strongly suggesting that a functional methyltransferase domain in Crg1 is required for cantharidin methylation . To definitively determine whether cantharidin is a substrate of Crg1 , we prepared and analyzed unlabeled reactions containing purified Crg1 , cantharidin , and SAM by liquid chromatography-mass spectrometry . We first looked at the extracted ion chromatogram expected for unreacted cantharidin ( C10H13O4+; m/z = 197 . 0814±100 ppm ) and found a large peak in cantharidin-containing reactions with an elution time of 18 . 6–18 . 8 min ( Figure S4B ) . The combined spectra of this peak in the complete reaction mixture contained several species: m/z = 197 . 0942 and 215 . 1065 , corresponding to the m/z for cantharidin and hydrated cantharidin ( C10H15O5+; m/z = 215 . 0919 ) , respectively ( Figure 3D ) . Because these species co-elute at 18 . 6–18 . 8 min in the complete reaction and the control reactions containing cantharidin ( Figure S4B and S4C ) and because sterically hindered anhydrides do not favor hydrolysis [39] , a likely explanation is that cantharidin forms a co-eluting water adduct during ionization . Next , we analyzed the full reaction to determine whether Crg1 methylates cantharidin . Specifically , we analyzed the extracted ion chromatogram expected for methyl cantharidin ( C11H15O4+; m/z = 211 . 0970±100 ppm ) . We identified a peak eluting after cantharidin at 19 . 2 min corresponding to the mass of methyl cantharidin in the complete reaction mixture ( Figure S4D ) . Importantly , this species was absent in each of our control reactions lacking cantharidin , SAM , or Crg1 ( Figure S4D ) . This is strong evidence that cantharidin is indeed methylated by Crg1 . Finally we analyzed the combined spectra of the 19 . 2-min peak ( Figure 3E ) . In addition to the m/z = 211 . 1105 species , we observed m/z = 229 . 1215 and m/z = 197 . 0930 species , corresponding to hydrated methyl cantharidin ( C11H17O5+; m/z = 229 . 1076 ) and cantharidin , respectively . Importantly , all of these species co-elute ( Figure S4B , S4D and S4E ) . Based on its close sequence homology to Tmt1 , we suspect that Crg1 catalyzes the formation of a cantharidin methyl ester ( Figure 2 ) . In solution , this putative methyl ester is likely in equilibrium with its ring-closed methyl anhydride form . If the equilibrium favors the ester form , some fraction of the cantharidin methyl ester could undergo ring-closing elimination reactions during ionization to yield methyl cantharidin and cantharidin products . Likewise , if the equilibrium favors the methyl anhydride , it may possibly undergo in-source fragmentation to give cantharidin , and like cantharidin , it may simply form a water adduct during ionization . Although our mass spectrometry data strongly support our hypothesis that Crg1 methylates cantharidin in vitro , additional analysis is needed to determine the structure of the methylated drug molecule . It is possible that the observed methyl cantharidin and hydrated methyl cantharidin species are ionization products of another methylated cantharidin derivative that is the actual product of the Crg1-catalyzed reaction . To identify cellular processes required for cantharidin resistance and to define the spectrum of genes that compensate for the absence of CRG1 in the presence of cantharidin , we used a chemogenomic approach to analyze genetic interactions between CRG1 and each of the ∼4800 non-essential yeast genes both in the presence and absence of the drug . A genetic interaction between two genes occurs when the phenotype of the double deletion mutant shows significant deviation in fitness compared with the expected ( multiplicative ) effect of combining two single mutants ( e . g . sickness or synthetic lethality ) [17] , [40] . When synthetic lethality is observed , it suggests that the genes may have overlapping functions . By analogy , identification of drug-gene interactions will similarly uncover genes that act in parallel with a gene of interest and these interactions can illuminate a compound's effect on a cell . To perform this experiment , we generated double deletion mutants ( with a crg1Δ strain as the query ) using Synthetic Genetic Array ( SGA ) technology [40] , pooled all viable double deletion mutants and analyzed their growth in a competitive fitness assay in the presence and absence of cantharidin ( Figure 4A ) [41] . Six highly reproducible and independent crg1ΔxxxΔ pools ( r = 0 . 72 , Figure S5A ) were further averaged yielding 70 double deletion mutants ( Table S3; Datasets S1 and S2 ) that showed significant growth defects ( log2 ( drug/DMSO ) <−1 , P-value <0 . 05 ) in the presence of an IC20 dose of cantharidin ( 30 µM ) in YPD when grown in a pool . Noteworthy , these genes were not sensitive as single deletion mutants ( P-value <0 . 025; Figure 4B and 4C , Table S3 ) , and , thus , the effect was specific to the double mutant combination . To obtain a general overview of “aggravating” ( negative ) interactors of CRG1 in the presence of cantharidin , we categorized this set of genes according to their GO term Biological process ( Figure 4B ) . This dataset comprised diverse biological processes , including vesicle-mediated transport ( P-value <0 . 008 ) , chromosome organization ( P-value <0 . 001 ) , response to chemical stimulus ( P-value <0 . 019 ) , lipid metabolic process ( P-value <2 . 0×10−5 ) , response to stress ( P-value <0 . 018 ) , and protein modification process ( P-value <0 . 003 ) . We also identified the serine/threonine kinase DBF2 as a strong suppressor of CRG1-dependent cantharidin toxicity ( log2 ( drug/DMSO ) = 1 . 12 , P-value <4 . 0×10−5; Figure 4B ) . This interaction was confirmed by evaluating the fitness of individual strains in liquid and on solid SD medium in the presence of 25 µM and 10 µM cantharidin , respectively ( lethal doses for crg1Δ strain in these media conditions; Figure 4C and Figure S5B ) . Furthermore , the alleviating interaction between DBF2 and CRG1 was not observed at 37°C , indicating cantharidin-specific nature of this interaction ( Figure S5B ) . In addition to the well-characterized roles of Dbf2 in the mitotic exit network [42] , this newly uncovered interaction suggests that this kinase may have an opposing function to the protein phosphatases ( PP2A and PP1 ) , the primary targets of cantharidin [30] , [31] . As independent evidence for an interaction with cantharidin , we found that DBF2 transcript levels were significantly decreased in a crg1Δ/Δ mutant in the presence of cantharidin ( 250 µM ) ( ∼2-fold , P-value <0 . 013 ) compared to DMSO control , and that change was not detected in other strains ( Figure S5C ) . This observation confirms the role of Crg1 in phosphorylation/dephosphorylation homeostasis during cantharidin stress . To identify genes specifically required for growth in the presence of cantharidin , we removed the genes that behave as multidrug resistance genes ( MDR ) from our chemogenomic dataset described above . MDR genes are defined here as those that are required for growth in the presence of multiple stress conditions ( at least 20% of tested conditions for homozygous deletion strains ) [15] . This filtering removed apparent enrichment of genes involved in vesicle-mediated transport genes ( P-value = 0 . 216 ) , response to stress ( P-value = 0 . 024 ) , chromosome organization ( P-value = 0 . 075 ) and protein modification process ( P-value = 0 . 021 ) . Following this , we found that cantharidin-specific CRG1 interactors are significantly enriched for genes required for lipid metabolic process ( multiple testing corrected P-value <0 . 0003 ) ( Figure S6A ) . In particular , lipid methyltransferases ( CHO2 , ERG6 ) , glycosylphosphatidylinositol ( GPI ) lipid biosynthesis genes ( ARV1 , GUP1 , PER1 ) and lipid-related genes ( SAC1 , MOT3 , DEP1 , RVS167 , YTA7 ) are essential in the double deletion strains in the presence but not absence of cantharidin treatment ( Figure 5A ) . Furthermore , we demonstrated that an increase in cantharidin concentration ( 10 µM ) did not result in cantharidin sensitivity for these genes compared to wild type ( Figure 5A ) , confirming the dependence of the detected interactions on the presence of CRG1 . To explore further the role of CRG1 in lipid metabolism and related processes , we compared the lipid content ( or “lipidome” ) of wild type , crg1Δ/Δ homozygous deletion and CRG1-overexpressing crg1Δ/Δ strains in the presence and absence of cantharidin ( 250 µM ) using electrospray ionization tandem mass spectrometry ( ESI-MS/MS ) analysis [43] . We observed significant changes in the abundance of most glycerophospholipids and sphingolipids in both the wild type and crg1Δ/Δ strains after growth in cantharidin-containing medium ( P-value <0 . 05 , Kruskal-Wallis test ) . The strains with a CRG1-overexpressing construct did not exhibit significant cantharidin-induced lipid alterations ( P-value >0 . 08 , Kruskal-Wallis test; Figure 5B , Table S4 , Dataset S3 ) . Specifically , in both the wild type and crg1Δ/Δ strains , cantharidin measurably increased the levels of short chain phosphatidylcholine ( PC ) , phosphatidylethanolamine ( PE ) , and phosphatidylinositol ( PI ) species , while the levels of long-chain PCs and PIs were reduced ( Figure 5B ) . In the crg1Δ/Δ strain we also noted a substantial decrease in the levels of mixed size phosphatidylserine ( PS ) species after cantharidin stress , while the wild type and crg1Δ/Δ strain had increased levels of saturated short chain ( C16 and C18 ) PI species compared to mono-unsaturated short chain PIs in cantharidin ( Table S4 ) . Such abundance changes with respect to acyl chain length and saturation were not observed in the CRG1-overexpressing mutant , suggesting that extra copies of CRG1 complemented the cantharidin-induced defects . It has been previously reported that phospholipid and sphingolipid biosynthetic pathways are interconnected ( Figure 5C ) [44]–[46] . One way in which this interconnection is seen is when , a single gene deletion or chemical perturbation of cells results in the so-called “ripple effect” [45] characterized by lipidome-wide perturbations . We see evidence of this effect: the amounts of the most abundant sphingolipid inositolphosphoceramide ( IPC ) and mannosyl-inositolphosphoceramide ( MIPC ) were also affected by cantharidin in a crg1Δ/Δ mutant ( Figure 5D ) . To investigate if other lipid intermediates are affected by the drug in a similar manner in crg1 mutants , we analyzed both sterol content and the formation of lipid droplets , which serve as storage pools of triacylglycerols and steryl esters [47] . We found no obvious changes in these lipid species in the presence of drug ( Figure S6B and S6C ) . Taken together , these results demonstrated that cantharidin's effect is specific towards phospholipids and sphingolipids in crg1 mutants . To test if cantharidin-induced alterations in yeast lipidomes are evolutionally conserved , we examined the lipidome of the human fungal pathogen Candida albicans in response to cantharidin . A C . albicans homozygous crg1 deletion ( orf19 . 633Δ/Δ ) displayed similar growth defects to those observed in S . cerevisiae when challenged with cantharidin ( Figure S7A ) . Lipidomic analysis demonstrated that cantharidin treatment ( 2 mM , IC20 for C . albicans wild type ) resulted in the significant changes in most phospholipid species in both wild type and orf19 . 633Δ/Δ homozygous mutant ( P-value <0 . 05 ) . Furthermore , although to a more modest degree than seen in S . cerevisiae , we found that C . albicans CRG1 may account for some difference between wild type and a mutant strain ( P-value <0 . 05; Figure S7B; Dataset S4 ) . In addition , we have shown previously that the overexpression of C . albicans ORF orf19 . 633 restored cantharidin resistance in S . cerevisiae crg1Δ/Δ mutant [14] , further suggesting that the lipid homeostasis functions of this C . albicans putative SAM-dependent methyltransferase are conserved . One of the phospholipids manifesting substantial changes in our lipidome analysis was phosphatidylinositol ( PI ) ( Figure S7C ) . PI is an essential phospholipid with multiple roles in the biosynthesis and metabolism of phosphoinositides ( PIP ) , inositol polyphosphates ( IPs ) , complex sphingolipids and glycerophosphoinositols ( GPIs ) ( Figure 5D ) [48] . It has been previously reported that phosphorylated derivatives of PI species ( mainly PI ( 4 , 5 ) P ) are well-conserved second messengers involved in the regulation of the actin cytoskeleton in Pkc1-dependent manner ( Figure 6A ) [48] , [49] . Therefore , to examine one of the physiological consequences of altered levels of PI , we tested if cantharidin affects the actin cytoskeleton . Microscopy of FITC-phalloidin stained cells revealed that crg1Δ/Δ strain treated with 250 µM cantharidin for 1 hour lacked actin patches and displayed highly disorganized actin cables compared to wild type . Overexpression of CRG1 in crg1Δ/Δ strain restored the number of actin patches close to that seen in the wild-type strain without cantharidin ( Figure 6B ) . These results demonstrate that Crg1 is critical for both actin patch and actin cytoskeleton integrity during cantharidin stress . Although , the observed role of Crg1 in cytoskeleton organization might be indirect , in our genome-wide screen ( without cantharidin ) we found that positive genetic interactions ( alleviating ) of CRG1 were significantly enriched for the genes involved in the actin cytoskeleton , bud emergence , and cell polarity ( P-value <1 . 0×10−5; Figure S8A; Dataset S5 ) . In particular , the deletion of RVS167 , a well-characterized actin patch and lipid-interacting protein , manifested fitness defects that are suppressed by the deletion of CRG1 ( Figure S8B ) [50] , [51] . These findings further support the role of Crg1 in actin-related biological process . Finally , to determine how Crg1 is regulated at the transcriptional level in response to cantharidin , we explored which pathways , if any , are required for cantharidin resistance . Based on our observation that the homozygous deletion strains slt2Δ/Δ and bck1Δ/Δ ( both CWI kinases ) are hypersensitive to cantharidin [14] , [15] , combined with the fact that the promoter region of CRG1 contains a binding site for Rlm1 ( a transcriptional regulator of CWI pathway ) [52] , we asked if CRG1 expression is activated by cantharidin via the CWI pathway . We found that deletion of these genes blunted the increase of CRG1 transcript in response to cantharidin ( 250 µM ) compared to the wild type ( Figure 6C ) , indicating that CWI pathway components are required for CRG1 expression in the presence of cantharidin . The CRG1 promoter also contains a binding site for Yap1 , a transcription factor required for cadmium tolerance and the oxidative stress response . In contrast to Rlm1 and Slt2 , the relative amount of CRG1 transcript in the yap1Δ/Δ mutant was unchanged in the presence of cantharidin ( Figure S9A ) . While these data suggest that Crg1 may be regulated via the CWI pathway and is transcriptionally responsive to numerous cell wall stressing agents ( Figure S9B ) , we did not detect any drastic fitness defects when crg1 mutants were grown in the presence of cell wall perturbing agents ( Figure S9B ) . However , overexpression of CRG1 in the crg1Δ/Δ mutant did confer resistance to lithium chloride and fenpropimorph , both of which are known perturbants of the cell membrane and other lipid processes ( Figure S9C ) [53]–[58] . Together these results supports a model in which Crg1 is involved in lipid-related processes that indirectly impinges on cell wall integrity .
In this study we demonstrated that yeast genetic and chemical genome-wide approaches , when combined with rigorous biological follow-up , can effectively characterize a novel gene that , despite being subject to numerous large-scale phenotypic studies , had little functional annotation . Our previous work demonstrated that Crg1 , a putative SAM-dependent methyltransferase , was a novel mediator of resistance to the protein phosphatase inhibitor , cantharidin [14] . Here we show that the mechanism of Crg1-cantharidin interaction is through direct methylation of the compound , and that , furthermore , Crg1 plays an essential role in the cellular response to cantharidin-induced lipid alterations . Our initial observation that cantharidin cytotoxicity is suppressed by overexpression of CRG1 suggested a specific , although not necessarily direct , cantharidin-Crg1 interaction in vivo [14] . Here , we demonstrate that Crg1 is able to interact with cantharidin in vitro , resulting in the formation of a methylated cantharidin species . Modification by methylation is known to remove negative charges on diverse molecules which can alter hydrophobicity and modulate cellular pathways and processes . Given the clear phenotype of Crg1-deficient cells and the results from our in vitro biochemical characterization of Crg1 , we hypothesize that methylation of cantharidin alters its physical properties such that it is no longer harmful to cells . In a manner similar to other methyltransferases that are known to detoxify small molecules [17] , [59]–[62] , chemical modification of cantharidin provides some insight regarding how its methylation may modify its activity . For example , endothall , an unmethylated , ring-opened form of cantharidin , has been assayed for protein phosphatase inhibition [26] and the methyl , ethyl , and propyl esters of endothall are still potent inhibitors of PP1 and PP2A . Several lactol derivatives of norcantharidin ( the anhydride form of endothall ) formed by reducing one of the carbonyl groups to a hydroxyl group have been synthesized and characterized . Modification of the free hydroxyl to form methyl , ethyl , and propyl ethers sharply reduced the ability of the drug derivatives to inhibit protein phosphatases . While the unmodified lactol form inhibited PP2A with an IC50 of 5 µM , the IC50 for the methyl ether lactol form was >1000 µM . Collectively , these observations suggest that methylation of closed-ring forms , not open-ring forms , reduces cellular toxicity . Cantharidin is more sterically hindered than norcantharidin , and as such , we would expect that its equilibrium would favor the closed-ring anhydride form more than that of norcantharidin . Accordingly , we are intrigued by the possibility that the methyl cantharidin product of the reaction catalyzed by Crg1 resembles the closed-ring lactol ether compounds that are less potent inhibitors of both growth and protein phosphatase activity . Further study to elucidate the structure of this product will enhance our understanding of how the methylation of cantharidin by Crg1 facilitates its detoxification . Looking ahead , it will be interesting to explore if similar mechanisms of cellular detoxification in mammalian cells are mediated by methyltransferases , such as METTL7A or METTL7B , which both share sequence homology to CRG1 . Interestingly , METTL7B , also known as ALDI , was reported to be highly expressed in kidney and liver , and associated with hepatic lipid droplets [63] , providing a provocative link between Crg1-like methyltransferases cantharidin toxicity and lipid process . In addition to characterizing the direct interaction of Crg1 with cantharidin , we investigated cellular pathways of Crg1-mediated cantharidin resistance using cells sensitized with a crg1 deletion allele . This analysis revealed that genes involved in lipid-related processes are required for survival under cantharidin-induced stress in the absence of CRG1 ( CHO2 , OPI3 , ERG6 , SAC1 , ARV1 , GUP1 , PER1 , MOT3 , DEP1 ) . Because CRG1 is both not essential and also shows very few genetic interactions under standard laboratory conditions , the identification of these genes required condition-specific assays . Our chemogenomic data are supported by lipidome-wide analysis , which demonstrated that cantharidin-induced alterations in glycerophospholipids and sphingolipids occur in a CRG1 gene dose-dependent manner . Specifically , we observed the accumulation of short chain phospholipids in the crg1Δ/Δ mutant , suggesting that the drug affects fatty acid elongation in a Crg1-dependent fashion . Consistent with this result , we also observed that overexpression of CRG1 confers resistance to lipid stressing agents such as lithium ions and the ergosterol inhibitor fenpropimorph ( Figure S9C ) . Resistance to fenpropimorph is acquired by mutations in the fatty acid elongase FEN1 ( ELO2 ) , which is known to be involved in sphingolipid biosynthesis [64] . Thus , it will be informative to test if CRG1 and FEN1 have overlapping functions in lipid biosynthesis . Another possible explanation for our observation that cantharidin-induced lipidome alterations can be suppressed by increasing the gene dose of CRG1 can be found in the transcriptional changes that occur in these strains . Genes involved in methionine biosynthesis are differentially expressed in CRG1-overexpressing strains in the presence of the drug compared to wild type and the crg1Δ/Δ mutant . This is of particular interest because changes in methionine metabolism can regulate methylation reactions by altering levels of SAM , a methyl donor [65] , [66] . For example , Tehlivets et al . showed that defects in methionine cycling enzymes result in an imbalance of phospholipid and triacylglycerol synthesis [67] , [68] . The mechanisms underlying these relationships are not yet clear , but it is possible that cells sense that the level of SAM is depleted via Crg1 activity , which results in transcriptional changes in methionine biosynthesis genes , in particular , the cystathionine beta-lyase Str3 . These findings suggest that overexpression of Crg1 may buffer cantharidin-treated lipidome alterations in part through changes in the methionine cycle . To define the ‘core’ buffering network to CRG1 in the presence of cantharidin , we compared the transcriptome and cantharidin SGA profiles . Although we did not find any obvious overlap in GO term biological processes between these datasets , in our cantharidin SGA one of the most sensitive mutants was MET22 ( Figure 4C ) , a gene with a role in sulfur assimilation and methionine biosynthesis . This gene was also differentially expressed in CRG1-overexpressing mutant vs . wild-type strain ( P-value <0 . 02; Table S2 ) . Our chemical genomics results were corroborated by traditional SGA analysis . This analysis demonstrated that CRG1 has an alleviating ( suppressing ) genetic interaction with RVS167 . It is well established that a similar phenotype is observed when RVS167 is deleted in combination with genes involved sphingolipid biosynthesis ( e . g . SUR1 , SUR2 , FEN1 , ELO3 and IPT1 ) , implicating sphingolipid biosynthesis in the regulation of the actin cytoskeleton [49] , [50] , [69]–[71] . Similarly to S . cerevisiae and C . albicans , studies in the ciliate Tetrahymena showed that cantharidin treatment also influences PI metabolism and the actin cytoskeleton [72] , demonstrating the conservation of cantharidin-lipid-actin interactions . Understanding the transcriptional regulation of CRG1 during cantharidin stress adds many layers to the picture of the complex physiological roles of this methyltransferase . CRG1 transcription is activated by cantharidin via the conserved MAPK family components of the CWI signaling pathway [52] , [73] . Hoon et al . previously demonstrated that deletion of slt2 and bck1 results in cantharidin sensitivity , suggesting that this pathway is critical for cantharidin resistance [14] . In mammalian cells , several studies have reported that the MAP kinases ERK and JNK are also activated by cantharidin [27] , [29] , likely as a consequence of the inhibition of protein phosphatases . Moreover , other studies reported that an intact CWI cascade is essential for maintaining lipid homeostasis [74] . It remains to be determined what specific steps are involved in the activation of CRG1 by cantharidin . One possible scenario is that the CWI pathway is activated by the accumulation of aberrant lipid species in a manner analogous to previous reports that suggest that long chain bases induce the Pkc1-MAPK CWI pathway in yeast [75] , [76] . Based on our observations , we propose the following mechanism for Crg1-cantharidin interaction ( Figure 6D ) . Cantharidin treatment inhibits PP2A and PP1 , resulting in the perturbation of both lipid homeostasis and actin cytoskeleton organization . This perturbation activates the CWI pathway , which in turn induces of CRG1 transcription . The resulting Crg1 protein directly methylates cantharidin , alleviating its cytotoxicity and restoring lipid homeostasis , actin cytoskeletal architecture , as well as other cantharidin-associated effects . In summary , our study demonstrates the value of combining classic biology approaches and chemical genomics with other “omic”-based methods for de-orphaning proteins and elucidating previously unknown mechanisms of therapeutics action .
Yeast strains and plasmids used in this study are described in Table S5 and Table S6 . Unless otherwise stated , wild-type ( wt ) strain is BY4743; crg1Δ/Δ was derived from BY4743 . Yeast cells were grown in YPD ( 2% yeast extract , 1% peptone , 2% glucose ) or in synthetically defined medium , SD ( 0 . 67% yeast nitrogen base , 2% glucose , and amino acids ) . Cantharidin , norcantharidin , cantharidic acid , and fenpropimorph were purchased from Sigma Aldrich ( Toronto , Canada ) . Lithium chloride was purchased from Teknova ( Hollister , CA , USA ) . Cantharidin , cantharidic acid , norcantharidin , and fenpropimorph were dissolved in DMSO and stored at −20°C . The IC20 of cantharidin in YPD for wild-type is 250 µM , in SD it is 5 µM , both determined in liquid culture as described [77] . CRG1 was amplified from wild-type strain using primers ( Table S7 ) with homology to the vector p426-GAL1-TAP at the 5′ end . The amplified CRG1 and HindIII linearized vector were directly co-transformed into a crg1Δ/Δ mutant and transformant colonies were selected in synthetic defined media lacking uracil ( SD-URA ) . CRG1 was cloned downstream of a GAL1 inducible promoter and in frame with the TAP coding sequence . Transformants were screened by PCR and for cantharidin resistance . CRG1 missense mutants were prepared using the QuickChange Lightning Site-directed mutagenesis kit ( Stratagene - Agilent Technologies Company , La Jolla , CA , USA ) . Clones were sequenced to verify the mutations . To express Crg1 , transformants were grown to mid-exponential phase in SD-URA and raffinose ( 2% ) , then induced by the addition of galactose to a final concentration of 2% . 30 µM of cantharidin was used to test sensitivity of mutants . Cells were harvested after 3 hours of induction , and Crg1 expression was verified by Western blots of 12% SDS-PAGE gels using anti-TAP antibodies ( OpenBiosystems – Thermo Fisher Scientific , Huntsville , AL , USA ) . Cells grown to mid-exponential phase in YPD medium were incubated with or without cantharidin ( 250 µM ) for various amounts of time , harvested by centrifugation , frozen in liquid N2 and stored at −80°C . RNA was extracted with hot acidic phenol [78] and treated with the Turbo DNA-free kit ( Ambion – Applied Biosystems , Austin , TX , USA ) . RNA purity was tested using a spectrophotometer and integrity was evaluated by denaturing gel electrophoresis . First-strand cDNA was synthesized from 1 µg of DNase-treated RNA with 0 . 5 µg of oligo ( dT12–18 ) primers ( Invitrogen , Burlington , ON , Canada ) using 200 units of Superscript II Reverse transcriptase ( Invitrogen , Burlington , ON , Canada ) . Real-time PCR analysis was conducted with Power SYBR Green PCR master mix ( Applied Biosystems , Foster City , CA , USA ) and gene-specific primers ( Table S7 ) at a final concentration of 250 nM . qRT-PCR was carried out on a 7900HT Fast system ( Applied Biosystems ) using Sequence Detection System software version 2 . 3 . Fold change in CRG1 transcript level normalized to ACT1 was calculated using the 2−ΔΔCt method . At least three independent replicates of each reaction were performed . Student's t-test was applied for statistical analysis ( paired for drug vs . DMSO treatments , and unpaired for mutants vs . wild type ) . Cells grown to mid-exponential phase in YPD medium were incubated with or without cantharidin ( 250 µM ) for 1 hour , harvested by centrifugation . Isolation of RNA and hybridization to the tiling arrays was performed as described [79] , except that actinomycin D was added in a final concentration of 6 µg/mL during cDNA synthesis to prevent antisense artifacts [80] . Two independent replicates were used for the analysis . Hybridization to Affymetrix Tiling Arrays using the GeneChip Fluidics Station 450 ( Affymetrix ) was followed by the extraction of intensity values for the probes using the GeneChip Operating Software ( Affymetrix ) . Acquisition and quantification of array images were performed using the Affymetrix tiling analysis software ( http://www . affymetrix . com/support/developer/downloads/TilingArrayTools/index . affx ) . The resulting . BAR files containing probe position and intensities were further analyzed by aligning the probes that match the position of the S . cerevisiae Genome Database list of defined ORFs ( http://downloads . yeastgenome . org/chromosomal_feature/saccharomyces_cerevisiae . gff ) . The log2 of signal intensity of each ORF was defined as the average across the probes associated with the ORF . Quantile normalized datasets were clustered with a correlation similarity metric and the average linkage method using Cluster 3 . 0 software ( http://bonsai . hgc . jp/~mdehoon/software/cluster/software . htm ) . The cluster was visualized using TreeView software ( http://jtreeview . sourceforge . net/ ) . The significance for differential expression was set as log2 ( drug/DMSO ) >1 and <−1 , P-value <0 . 05 as determined by Student's t test . Significantly up- and downregulated transcripts were further tested for Gene Ontology ( GO ) biological process enrichment using FunSpec ( http://funspec . med . utoronto . ca/ ) with P-value cutoff of 0 . 01 and multiple testing correction ( Bonferroni ) ( Table S1; Dataset S6 ) . The probability was calculated using a test employing hypergeometric distribution ( see below ) . To detect cantharidin-specific genes the genes involved in ESR [36] were eliminated from the gene-set . Purification of 6xHis-Crg1 was performed as previously described [81] . Wild-type yeast strain Y258 carrying a vector pBG1805-GAL1-CRG1 with a triple affinity tag at C-terminal ( His6-HAepitope-3Cprotease site-ZZprotein A ) was grown in 660 mL of synthetically defined medium ( SD-URA and 2% raffinose ) to mid-exponential phase at 30°C . To induce expression of CRG1 340 mL of 3x YP ( yeast extract and peptone ) and 6% galactose was added to a final concentration of 2% . Cells were harvested by centrifugation at 3 , 000 rpm for 5 min . All steps following harvest were performed at 4°C . Cells were washed with PBS buffer , resuspended in 7 mL of resuspension buffer ( 20 mM HEPES pH 7 . 5 , 1 M NaCl , 5% glycerol ) , and lysed using an acid washed Zirconia beads in the presence of protease inhibitors ( 1 mM Pefablock , 2 . 5 µg/mL pepstatin A , 2 . 5 µg/mL leupeptin , 1 mM PMSF ) . The cell lysate was centrifuged at 20 , 000 rpm for 45 min , and diluted two fold with binding buffer ( 20 mM HEPES pH 7 . 5 , 40 mM imidazole , 5% glycerol ) . 300 µL of Ni Sepharose 6 Fast Flow beads ( 50% slurry in 20% ethanol ) was added to the sample and rotated for 1 . 5 hours , followed by three washes with 40 mL of wash buffer ( 20 mM HEPES pH 7 . 5 , 40 mM imidazole , 5% glycerol , 0 . 5 M NaCl ) . To elute Crg1 the Ni beads were resuspended in 1 mL of elution buffer ( 20 mM HEPES pH 7 . 5 , 250 mM imidazole pH 7 . 7 , 5% glycerol , 0 . 5 M NaCl ) , and Crg1p was released by rotating the mixture for 15 min at 4°C . The protein was further concentrated with Amicon Ultra tubes ( 10 K ) ( Millipore , Etobicoke , ON , Canada ) to 100 µL . Purification of TAP-tagged wild-type and mutant forms of Crg1 was performed in BY4743 carrying p426-GAL1-CRG1-TAP as described in Rigaut et al . [82] . Cell growth , induction of Crg1 with galactose ( 2% ) , and preparation of cell lysates were performed as described for the Crg1-6xHis fusion . 300 µL IgG-agarose was added to the extract and incubated for 2 hours followed by a triple wash with 25 mL of Low Salt and High Salt Wash Buffer ( 50 mM HEPES pH 7 . 5 , 10% glycerol , 150 mM/750 mM NaCl , 0 . 1% Tween20 ) . The final wash was performed with 15 mL TEV Cleavage Buffer ( 10 mM , Tris pH 8 . 0 , 150 mM NaCl , 0 . 05% Tween20 , 10% glycerol , 0 . 5 mM EDTA , 1 mM DTT ) . The extract was incubated overnight with 100 U TEV protease ( Invitrogen , #12575-015 ) . 1 . 2 mL CaM Binding Buffer ( 10 mM Tris pH 8 . 0 , 150 mM NaCl , 0 . 05% Tween20 , 10% glycerol , 1 mM MgOAc , 1 mM imidazole pH 8 . 0 , 2 mM CaCl2 , 1 mM DTT ) and 2 . 4 µL 1 M CaCl2 were added to the protein eluates . The eluates were then incubated with 400 µL ( 50% slurry ) Calmodulin Sepharose in 5 mL CaM Binding Buffer for 2 hours , followed by a wash with 25 mL CaM Binding Buffer and elution with 5×200 µL Elution Buffer ( 10 mM Tris pH 8 . 0 , 150 mM NaCl , 0 . 05% Tween20 , 10% glycerol , 1 mM MgOAc , 1 mM imidazole pH 8 . 0 , 2 mM EGTA , 1 mM DTT ) . Protein eluates were stored at −20°C in the presence of 50% glycerol . Recovery of Crg1 was determined using Bradford reagent ( BioRad Laboratories , Mississauga , ON , Canada ) and its integrity and purity was assessed with 12% silver-stained SDS-PAGE gel . In vitro enzymatic reactions were prepared with 0 . 09 µg 6×His-tagged Crg1 , 0 . 2 mM cantharidin dissolved in DMSO ( Sigma Aldrich , St . Louis , MO , USA ) , and 20 µM S-adenosyl-[methyl-14C]methionine ( 55 . 8 mCi/mmol ) ( GE Healthcare , Piscataway , NJ , USA ) in 0 . 1 M sodium phosphate , pH 7 . 4 with a final volume of 50 µL and 2% DMSO . The complete reactions and relevant controls were incubated at 30°C for 2 hours . Following incubation , the enzymatic reactions were separated by reverse-phase high-performance liquid chromatography ( Series II 1090 Liquid Chromatograph , Hewlett Packard , Palo Alto , CA , USA ) . The chromatography gradient was adapted from [83] , except that the flow rate was 1 mL/min . Mobile phase A contained 0 . 1% trifluoroacetic acid in water and mobile phase B was 0 . 1% trifluoroacetic acid in acetonitrile . A BetaBasic-18 column ( 250 mm×4 . 6 mm; 5-µm particle size ) ( Thermo , Waltham , MA ) was used . 40 µL was injected , 2-minute fractions were collected , and 350 µl of each fraction was mixed with 5 mL of Safety-Solve ( Research Products International , Mt . Prospect , IL , USA ) before quantification of radioactivity with a LS6500 liquid scintillation counter ( Beckman Coulter , La Brea , CA , USA ) . Each fraction was counted three times for 3 minutes . Other in vitro reactions were prepared in an identical manner with 0 . 09 µg of either mutant or wild-type TAP-tagged Crg1 and varying concentrations of cantharidin ( USB , Cleveland , OH , USA ) . After incubation at 30°C for 2 hours , 40 µL 2 N HCl was added to each 50 µL reaction . Immediately , 80 µL of this mixture was transferred to a 1 . 9-cm×9-cm folded piece of filter paper in the neck of a scintillation vial containing 5 mL of Safety-Solve and the vials were capped . After 4 hour incubation at room temperature , the pieces of filter paper were removed from the neck of each vial and the acid-labile volatile radioactivity was quantified with a liquid scintillation counter as described above [84] . In vitro enzymatic reactions using unlabeled SAM ( Sigma Aldrich ) were prepared in a similar manner with 0 . 09 µg of 6xHis-tagged Crg1 , 200 µM cantharidin ( Sigma Aldrich ) , and a SAM concentration of 1 . 6 mM . These reactions were quenched with addition of 200 µL acetonitrile , and 12 . 5 µL of 15% ammonium bicarbonate was added to reduce product degradation . After concentration with a vacuum centrifuge and resuspension in 50 µL of water , these reaction mixtures were analyzed by liquid chromatography-tandem mass spectrometry ( 1100 Series Liquid Chromatograph , Agilent , Santa Clara , CA; QSTAR Elite Mass Spectrometer , Applied Biosystems , Foster City , CA ) in positive ionization mode with a Turbo Spray source . 8 µL of sample was injected onto a reverse-phase column ( Luna C18 ( 2 ) , 150 mm×1 mm , 5-µm particle size , Phenomenex , Aschaffenburg , Germany ) with a flow rate of 50 µL/min and the following gradient: t = 0–1 min , 2% mobile phase B; t = 10 min , 35% B; t = 14–16 min , 90% B; t = 16 . 5–35 min , 2% B . Mobile phase A was 0 . 1% formic acid in 98% water and 2% acetonitrile , while mobile phase B was 0 . 1% formic acid in 98% acetonitrile and 2% water . Spectra were collected with the instrument information-dependent acquisition mode ( full scan: m/z = 70–600 , 1 . 000071 s; 3 product experiments: m/z = 65–500 , <2 s per experiment ) . The following mass spectrometer parameters were used: declustering potential , 85 V; focusing potential , 300 V; declustering potential II , 15 V; ionspray voltage , 5500 V; ion source gas , 45 units; ion source gas II , 5 units; curtain gas , 45 units; collision gas , 7 units . A pool of double deletion mutants ( crg1ΔxxxΔ ) was prepared by generating viable double deletion mutants using Synthetic Genetic Array ( SGA ) technology with crg1Δ as a query strain [40] ( Text S1 ) . ∼4800 viable double deletion mutant colonies were collected , normalized to 50 OD's/mL and stored at −80°C in media containing 7% DMSO . Two independent pools were generated for the analysis . Each was tested in triplicate . The pooling of the strains was possible due to the presence of strain-specific sequence tags flanking each gene deletion region [5] . The double deletion pool was treated with 30 µM of cantharidin , a dose which inhibits growth of the crg1 double deletion pool by ∼20% . Fitness analysis using a tag-specific algorithm that takes into account the intensities of each tag in cantharidin-treated cells compared to non-treated cells was performed as described [41] . Hybridization to Affymetrix Gene Chips using GeneChip Fluidics Station 450 ( Affymetrix ) was followed by the extraction of intensity values for the probes using the GeneChip Operating Software ( Affymetrix ) . The data was quantile normalized , outliers ( one standard deviation off ) were omitted , and fitness defect scores as the log2 ratio between the mean signal intensities of the control ( DMSO ) and the drug were calculated for each deletion strain in the pool as previously described [41] . As a control , the relative fitness of the double-gene deletion mutants exhibiting high sensitivity to cantharidin ( log2 ( drug/DMSO ) <−1; P-value <0 . 05 ) was compared to the relative fitness of the corresponding single gene deletion mutants . For a given double deletion mutants , the resulting interaction was evaluated using comparison of the observed double mutant growth rate to the expected assuming that there is no interaction exists . Log2 ratio <−1 represents those strains with a measurable growth defect ( or lethality ) and log2 ratio >1 demonstrates resistance to the drug . Cantharidin-specific CRG1 negative genetic interactors were further verified as individual clones in liquid growth assays and/or by spot dilutions . To evaluate the chemogenomic dataset for statistical significant enrichment for general biological processes ( Gene Ontology Slim mapper ) , we used a standard hypergeometric test , that asses the probability that the intersection of given list with any given functional category occurs by chance . Obtained P-values were corrected for multiple testing correction ( Bonferroni ) by multiplying P-values with the number of genes in the test . The probability was calculated as follows: the P-value of observing x genes , belonging to the same functional category , is:where M is the total number of genes involved in a functional category , n is the total number of genes in the cluster , and N is the total number of yeast ORFs . Cells were grown to mid-exponential phase and treated with cantharidin ( 250 µM ) for 2 hours . Lipids were extracted from 25 OD600-equivalent of cells and analyzed as described [43] . All lipid standards were obtained from Avanti Polar Lipids ( Alabaster , AL , USA ) , with the exception of dioctanoyl glycerophosphoethanolamine , which was obtained from Echelon Biosciences ( Salt Lake City , UT , USA ) . Quantification of individual molecular species was carried out using multiple reaction monitoring ( MRM ) with an Applied Biosystems 4000 Q-Trap mass spectrometer ( Applied Biosystems , Foster City , CA , USA ) . 25 µl of samples were subjected to analysis as described previously [43] , [85] . Lipid levels in each sample were normalized to internal standards . For each lipid species , the mean normalized signal from the wild type and mutant strains grown in the presence or absence of cantharidin were calculated . Three independent experiments were used for analysis . Lipid levels were calculated relative to relevant internal standards . The quantities of lipids are expressed as ion intensities relative to the levels without cantharidin , and then converted to a log2 ( drug/DMSO ) scale . The difference in levels of individual lipid species in DMSO vs . cantharidin was determined with Kruskal-Wallis test ( Table S4 ) . Similarly , the difference between wild type and mutants was assessed statistically using the Kruskal-Wallis test , with a P-value cutoff of 0 . 05 . Cells were grown to mid-exponential phase in YPD media , and treated with and without cantharidin ( 250 µM ) for 1 hour . Cells were then fixed by addition methanol-free formaldehyde ( Polysciences , Warrington , PA ) to 4% for 1 hour , centrifuged at 3 , 000 rpm 5 min and washed with PBS buffer three times . Cells were permeabilized with 0 . 2% Triton X-100 in PBS at 25°C for 15 min , washed with PBS three times and a normalized number of cells were stained with Alexa Fluor 488 phalloidin ( Invitrogen , Burlington , ON ) in the dark at 25°C for 1 hour . Cells were observed with 100× objective , and fluorescence images were acquired using AxioVision software on an Axiovert 200 M fluorescence microscope ( Carl Zeiss ) using a 1 . 5 s exposure for all images . The average number of actin patches per cell was determined by dividing the total number of actin patches per total number of cells ( N = 3 , n≥300 cells ) . | Chemical genetics uses small molecules to perturb biological systems to study gene function . By analogy with genetic lesions , chemical probes act as fast-acting , reversible , and “tunable” conditional alleles . Furthermore , small molecules can target multiple protein targets and target pathways simultaneously to uncover phenotypes that may be masked by genes encoding partially redundant proteins . Finally , potent chemical probes can be useful starting points for the development of human therapeutics . Here , we used cantharidin , a natural toxin , to uncover otherwise “hidden” phenotypes for a methyltransferase that has resisted characterization . This enzyme , Crg1 , has no phenotype in standard conditions but is indispensible for survival in the presence of cantharidin . Using this chemical genetic relationship , we characterized novel functions of Crg1 , and by combining diverse genomic assays with small molecule perturbation we characterized the mechanism of cantharidin cytotoxicity . These observations are relevant beyond yeast Crg1 because cantharidin and its analogues have potent anticancer activity , yet its therapeutic use has been limited to topical applications because of its cytotoxicity . Considering that methyltransferases are an extremely abundant and diverse class of cellular proteins , chemical probes such as cantharidin are critical for understanding their cellular roles and defining potential points of therapeutic intervention . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cellular",
"stress",
"responses",
"genetic",
"networks",
"functional",
"genomics",
"enzymes",
"chemical",
"biology",
"toxicology",
"toxic",
"agents",
"gene",
"function",
"genome",
"analysis",
"tools",
"enzyme",
"metabolism",
"genetics",
"and",
"genomics",
"lipids",
"... | 2011 | A Systems Biology Approach Reveals the Role of a Novel Methyltransferase in Response to Chemical Stress and Lipid Homeostasis |
Many arboviruses transmitted by mosquitoes have been implicated as causative agents of both human and animal illnesses in East Africa . Although epidemics of arboviral emerging infectious diseases have risen in frequency in recent years , the extent to which mosquitoes maintain pathogens in circulation during inter-epidemic periods is still poorly understood . This study aimed to investigate whether arboviruses may be maintained by vertical transmission via immature life stages of different mosquito vector species . We collected immature mosquitoes ( egg , larva , pupa ) on the shores and islands of Lake Baringo and Lake Victoria in western Kenya and reared them to adults . Mosquito pools ( ≤25 specimens/pool ) of each species were screened for mosquito-borne viruses by high-resolution melting analysis and sequencing of multiplex PCR products of genus-specific primers ( alphaviruses , flaviviruses , phleboviruses and Bunyamwera-group orthobunyaviruses ) . We further confirmed positive samples by culturing in baby hamster kidney and Aedes mosquito cell lines and re-sequencing . Culex univittatus ( 2/31pools ) and Anopheles gambiae ( 1/77 pools ) from the Lake Victoria region were positive for Bunyamwera virus , a pathogenic virus that is of public health concern . In addition , Aedes aegypti ( 3/50 ) , Aedes luteocephalus ( 3/13 ) , Aedes spp . ( 2/15 ) , and Culex pipiens ( 1/140 ) pools were positive for Aedes flaviviruses at Lake Victoria , whereas at Lake Baringo , three pools of An . gambiae mosquitoes were positive for Anopheles flavivirus . These insect-specific flaviviruses ( ISFVs ) , which are presumably non-pathogenic to vertebrates , were found in known medically important arbovirus and malaria vectors . Our results suggest that not only ISFVs , but also a pathogenic arbovirus , are naturally maintained within mosquito populations by vertical transmission , even in the absence of vertebrate hosts . Therefore , virus and vector surveillance , even during inter-epidemics , and the study of vector-arbovirus-ISFV interactions , may aid in identifying arbovirus transmission risks , with the potential to inform control strategies that lead to disease prevention .
The East African Great Lakes region is a recognized hotspot for a broad diversity of arthropod-borne viruses ( arboviruses ) [1] that affect humans and animals [2] and are transmitted by several mosquito genera ( mostly Culex Linnaeus , Aedes Meigen , Anopheles Meigen , Mansonia Blanchard , and Aedeomyia Theobald species ) [3–5] . Some mosquito species are capable of naturally maintaining viruses in circulation through vertical transmission [6–9]–up to 38 generations for San Angelo ( SA ) virus in Aedes albopictus , though with progressive decline in filial infection rate ( FIR ) in laboratory population bottlenecks [10] . The Lake Victoria and Lake Baringo regions of Kenya have historically been associated with arboviral diseases [11] and have unique lake and island biogeographies [12] in which arboviruses exist [5] . Outbreaks in the 1960s around the Lake Victoria basin involved Semliki Forest , chikungunya , and o’nyong-nyong viruses that are vectored by Culex , Aedes , and Anopheles mosquito species , respectively [13] . More recent studies have found seropositivity for arboviruses in humans [14–16] . During the recent 2006–2007 Rift Valley fever ( RVF ) outbreak in Baringo County , 10 mosquito species were implicated as potential vectors , among which Aedes pembaensis Theobald , Culex univittatus Theobald , and Culex bitaeniorhynchus Giles were reported as potential vectors for the first time [11] . Although widespread arboviral activity in human populations has been documented in the Lake Victoria and Lake Baringo basins , the role of vertical transmission among mosquito vectors in the maintenance of arboviruses within ecologies remains poorly understood [17] . To ascertain the competence of mosquitoes to horizontally transmit arboviruses between hosts , many methods have been used to collect and test different mosquito body parts ( abdomen , saliva , and legs ) for arboviruses [18] . However , vertical transmission of arboviruses from adult female mosquitoes to their offspring can also maintain viruses in circulation for generations within mosquito populations [6–10] . To investigate how vertical transmission in different mosquito species in Homa Bay and Baringo counties of Kenya may be maintaining endemic arboviruses in circulation , we set out to identify arboviral infections in laboratory-reared adults of field-caught larvae and pupae .
In 2012 , immature mosquitoes were sampled from islands and mainland shores of Lake Baringo ( in Baringo County along the Great Rift Valley ) and Lake Victoria ( in Homa Bay County ) of Kenya ( Fig 1 ) during the rainy season . In Baringo County , samples were collected in July and October 2012 from Kokwa Island , Nosuguro , Salabani , Kampi ya Samaki , Sirata , and Ruko . In Homa Bay County , samples were collected in April , May and November 2012 from Ringiti , Chamaunga , Kibuogi , Rusinga , Takawiri , Mfangano and Ngodhe Islands , and Ungoye , Luanda Nyamasare , Mbita and Ngodhe mainland sites on the Kenyan part of Lake Victoria . Sampling was conducted on unprotected public land concurrently with an adult mosquito genetic diversity survey conducted in the same study areas [19] . We collected eggs , larvae , and pupae with 350-ml standard dippers ( Bioquip Products , USA ) from their breeding sites and transported them to the Martin Lüscher Emerging Infectious Disease ( ML-EID ) Laboratory at the Duduville campus of the International Centre of Insect Physiology and Ecology ( icipe ) in Nairobi , Kenya . In the laboratory , we reared them to adults in their field-collected breeding water at 28°C temperature , 80% relative humidity , and 12-hour day and night cycles [20 , 21] . Before sampling , we obtained ethical clearance for the study from the Kenya Medical Research Institute ( KEMRI ) ethics review committee ( Approval Ref: Non-SSC Protocol #310 ) and no protected species were sampled . All reared adult mosquitoes were identified and sorted using morphological keys [22–25] in petri-dishes on frozen ice packs to keep them cold and to avoid degradation of any viruses in the samples . The ice packs were wrapped with paper towels to absorb moisture and prevent frosting of the petri-dishes . We stored pools of ≤25 reared adult mosquitoes in well-labelled 1 . 5 ml microcentrifuge tubes according to species , larval collection sites , sex , and dates in tubes in a -80°C freezer . Ten pieces of 2 . 0-mm yttria-stabilized zirconia beads ( Glen Mills , Clifton , NJ ) and 400 μl of cold homogenization media ( 2% L-glutamine , 15% fetal bovine serum ) ( Sigma-Aldrich , St . Louis , USA ) were added to each tube , which were placed on ice to keep them cold . The mosquito pools were then homogenized for 10 seconds in Mini-BeadBeater-16 ( BioSpec , Bartlesville , OK , USA ) followed by centrifugation for 10 seconds in a bench top centrifuge ( Eppendorf , USA ) at 1 , 500 relative centrifugal force ( rcf ) and 4°C . Aliquots of 210 μl of each homogenate were used for nucleic acid extraction and the remaining aliquots were stored in -80°C freezer as stock . Nucleic acid ( NA ) was extracted from the 210-μl mosquito homogenate aliquots using the MagNA 96 Pure DNA and Viral NA Small Volume Kit ( Roche Applied Science , Penzberg , Germany ) in a MagNA Pure 96 automatic extractor ( Roche Applied Science ) and eluted into a final volume of 50 μl according to the manufacturer’s instructions . A reverse transcription-multiplex polymerase chain reaction with high-resolution melting ( RT-PCR-HRM ) analysis based arbovirus screening protocol recently developed by Villinger et al . [26] was used to rapidly screen many samples and detect the presence of four arbovirus genera , namely , Alphavirus ( family Togaviridae ) , Flavivirus ( family Flaviviridae ) , Bunyamwera-group Orthobunyavirus ( family Peribunyaviridae ) , and Phlebovirus ( family Phenuiviridae ) . Briefly , the High Capacity cDNA Reverse Transcription ( RT ) kit ( Life Technologies , USA ) was used to synthesize complimentary DNA ( cDNA ) of the nucleic acid extracts . cDNA synthesis from 5 μl of extracted nucleic acids was performed in 10-μl reaction volumes with final concentrations of 1x RT Buffer , 4 mM dNTP mix , 2 . 5 U/μl MultiScribe Reverse Transcriptase , 1 U/μl RNase Inhibitor , 600 μM non-ribosomal random hexanucleotide primers [27] . Reverse transcriptions were performed in a Veriti 96-Well Thermal Cycler ( Applied Biosystems , Singapore ) at 25°C for 10 minutes , 37°C for 2 hours , 85°C for 5 minutes and held at 4°C . We used established multiplex RT-PCR thermocycling conditions [26] in a HRM capable Rotor-Gene Q real-time PCR thermocycler ( Qiagen , Redwood city , CA , USA ) to screen for virus sequences in cDNA templates . Ten microliter reactions consisting of 1 μl cDNA template , 5 μl 2x MyTaq HS Mix ( Bioline , UK ) , 1 μl of 50 μM SYTO-9 saturating intercalating dye ( Life Technologies ) , and multiplex PCR primers at concentrations given in Table 1 . The QIAgility robot ( Qiagen ) for liquid handling was used to set up the reaction mixture . Touchdown PCR cycling conditions as detailed by Villinger et al . [26] included an initial denaturation at 95°C for 5 minutes , followed by 50 cycles of denaturation at 94°C for 20 seconds , annealing at 63 . 5–47 . 5°C for 20 seconds , and extension at 72°C for 5–30 seconds , followed by a final extension at 72°C for 3 minutes . Immediately after PCR , the product was held at 40°C for 1 minute before HRM analyses of PCR product double stranded DNA stability by measuring SYTO-9 fluorescence at 0 . 1°C temperature intervals increasing every 2 seconds from 75°C to 90°C . PCR grade water was used as negative control , and Bunyamwera ( Orthobunyavirus ) , dengue and West Nile ( Flavivirus ) , sindbis and Middelburg ( Alphavirus ) , and Rift Valley fever ( Phlebovirus ) viruses were used as positive controls . Positive samples were re-run in singleplex reactions ( using primers from only one genus; Table 1 ) . Amplicons from singleplex runs were purified with ExoSAP-IT for PCR Product Kit ( Affymetrix Inc . , USA ) and Sanger-sequenced at Macrogen ( Korea ) . Samples that were positive for the Flavivirus genus by HRM analysis were further sequenced from nested PCR products using the 2NS5F ( 5’-GCNATNTGGTWYATGTGG-3’ ) and 2NS5Re ( 5’-TRTCTTCNGTNGTCATCC-3’ ) primers that amplify longer nucleotide fragments ( ~930 nt ) of Flavivirus NS5 genes [28] . Resulting nucleotide sequences were edited using Geneious R7 . 1 . 9 software ( created by Biomatters ) [29] . To validate that the sequenced targets were truly viral and not viral genome segment inserts in the mosquito genome , a fraction of the original mosquito homogenates that were PCR-positive for potential arboviruses were subjected to cell culture in vertebrate BHK-21 ( Kidney of Syrian hamster , Lot: 59300875 from ATCC ) and Ae . albopictus clone C6/36 ( Whole larva of Asian tiger mosquito , Lot: 60400699 from ATTC ) cell lines . Stock mosquito homogenates of 19 samples with sequences that aligned with known viruses on GenBank [32] and RNA virus databases were subjected to cell culture . The homogenates were thawed on ice and clarified by centrifugation at 15 , 000 rcf and 4°C in a bench top centrifuge ( Eppendorf 5417R ) for 5 minutes . One hundred microlitres of the clarified supernatant were aseptically inoculated in each of sub-confluent BHK-21 and C6/36 cell lines in a 24-well culture plate . The BHK-21 cells were initially aseptically grown in growth media ( GM; pH 7 . 5 ) made of 2% Minimum Essential Media ( MEM; +Eagle’s salt , +25 Mm HEPES ) with 10% FBS , 2% L-glutamine and 1% antimycotic ( Sigma-Aldrich ) . The C6/36 GM contained same proportions of respective constituents as the BHK-21 GM , but with the addition of 1% non-essential amino acids ( GIBCO , UK ) . The inoculated cell lines were incubated for 14 days and observed daily for any change in the morphology of the cell line caused by viral infection , also known as the cytopathic effect ( CPE ) . Virus presence was ascertained as CPE . During the initial 14-day incubation period , any contaminated cell culture was purified using a 0 . 22 μm syringe filter [33] and re-tested . Further , RNA was extracted from cell culture wells that showed CPE and tested in single-genus arbovirus RT-PCR-HRM reactions and re-confirmed by sequencing , as described above . Using Basic Local Alignment Search Tool ( BLAST ) [34] , initial searches were performed for comparison of all obtained virus sequences with those in GenBank . This was followed by sequence alignments using the default settings of the MAFFT v7 . 017 [35] plugin in Geneious software , to identify virus segments . Maximum likelihood phylogenetic relationships of the study’s insect-specific flaviviruses ( ISFVs ) NS5 sequences with those of related ISFVs were analyzed using PhyML version 3 . 0 [36] , employing the Akaike information criterion [37] for automatic selection of the general time reversible ( GTR ) sequence evolution model . Tree topologies were estimated using nearest neighbour interchange ( NNI ) improvements over 1000 bootstrap replicates . Rooting the phylogeny to the yellow fever vaccine strain sequence ( GenBank accession NC_002031 ) as an outgroup , the phylogenetic tree was depicted using FIGTREE version 1 . 4 . 2 [38] .
A total of 4 , 453 adult mosquitoes comprised of nine Aedes , six Anopheles , 16 Culex and one Mimomyia species were reared from immatures ( Table 2 ) . Among 612 pools of ≤25 mosquito samples per pool , 92 pools were from Baringo County and 520 pools were from Homa Bay County . Among mosquito pools from 32 species sampled in Homa Bay County , Bunyamwera virus ( Orthobunyavirus ) was the only vertically transmitted arbovirus ( pathogenic to vertebrates ) detected . It was identified by HRM analysis ( Fig 2A ) , culture , and DNA sequencing ( 143 nt; 100% identity to GenBank accession KM507344 , S1 Fig ) from female Anopheles gambiae from Luanda Nyamasare ( 1/77 pools ) and Cx . univittatus from Rusinga ( 2/31 pools ) ( Table 2 ) that were reared from larvae sampled in November 2012 . However , no vertically transmitted pathogenic arbovirus was detected in Baringo County samples . Further , we detected ( Fig 2B ) and sequenced ISFV NS5 sequences from 12 mosquito pools ( Table 2 , GenBank accessions: MG372051-MG372060 , MK015647-MK015648 ) among May 2012 collections . Among Baringo samples , we sequenced three ISFVs from female An . gambiae mosquitoes ( 3/15 pools; one pool from Ruko and two pools from Kampi ya Samaki ) collected in October 2012 that were closely related to Anopheles gambiae flaviviruses ( An ( g ) FV ) that were previously detected in mosquitoes sampled from Kenya’s North-Eastern Province [26] , as well as Western and Coastal Provinces ( Fig 3 ) . Among Homa Bay County samples , we found Aedes flavivirus ( AeFV ) NS5 sequences in Ae . luteocephalus ( 3/13 pools; two pools from Ungoye and one pool from Mbita ) and Aedes sp . ( 1/15 pools; from Takawiri Island ) , as well as in Cx . pipiens ( 1/140 pools; Rusinga Island ) . We also found cell fusing agent virus ( CFAV ) , the first ISFV originally identified in Ae . aegypti using an Ae . albopictus cell line ( C6/36 ) [39] , among Homa Bay County Ae . aegypti ( 3/50 pools; Mfangano Island ) and Aedes spp . ( 1/15 pools; from Ungoye ) samples .
We identified natural infections of Bunyamwera virus and ISFVs in diverse anopheline and culicine mosquito species reared to adults from field-collected larvae , demonstrating that these viruses persist transstadially through development to adult stages from naturally infected immature life stages . Since vertical transmission was first identified of vesicular stomatitis virus by phlebotomine sandflies [40] followed by La Crosse virus in Aedes triseriatus [41 , 42] , this mode of maintaining arboviruses within ecosystems has been observed in numerous arboviruses of medical importance circulating in East Africa , including West Nile virus by Culex and Aedes mosquitoes [43–47] , Ndumu virus [48] by Cx . pipiens , and Zika [49 , 50] , dengue [51–53] , chikungunya [54] , and RVF [8] viruses by Aedes mosquitoes [55] . However , how widespread or important this mode of transmission is in natural ecologies remains poorly understood . While we attribute the naturally occurring virus infections that were transstadially transmitted from immature life stages in this study to vertical transmission from their parents , we cannot completely rule out the possibility that the immature mosquitoes were infected with these viruses from viral contamination in their aquatic environment during early development . However , this mode of transmission if far less likely as past studies indicate that such infection of immature mosquitoes requires unrealistically high viral doses in their aquatic environment [56] . We documented the vertical transmission of the Orthobunyavirus , Bunyamwera virus , from naturally occurring infections in two mosquito species–An . gambiae and Cx . univittatus–the former of which has previously been found to competently transmit Bunyamwera virus during blood-feeding on suckling mice [57] . This is of public health importance and needs to be monitored closely , as Bunyamwera is an important cause of acute febrile illness in humans ( Bunyamwera fever ) [58] that is able to reassort with closely related arboviruses to form new viruses , such as Ngari virus , which can cause haemorrhagic fever in humans [59] . With the well-established role of vertical transmission in Ae . triseriatus mosquitoes of the closely related Orthobunyavirus , La Crosse virus [60] , the potential of Bunyamwera virus to remain in circulation by vertical transmission within mosquito populations in East Africa , highlights the importance of control strategies focused on vectors and the replication of arboviruses within the vector . Recent laboratory vector competence studies have found that Bunyamwera virus can be competently transmitted by An . gambiae and Ae . aegypti mosquitoes [57] , and can naturally infect Aedeomyia africana , Anopheles coustani , and Mansonia africana mosquitoes [5] . However , Culex quinquefasciatus was found to be refractory to Bunyamwera virus infection experimentally [57] . Our findings demonstrate that Bunyamwera infection persists from larval stages to adults in Cx . univittatus mosquitoes as well as in Bunyamwera competent An . gambiae . This expands the mosquito species , and indeed genera , that may play key roles in maintaining Bunyamwera virus in circulation . Though the vectorial competence of Cx . univittatus to transmit Bunyamwera virus has not been established , the species is thought to prefer birds as a source of bloodmeals [61] and has recently been found to also feed on dogs , donkeys , sheep , and toads [5] , as well as humans [62] . Therefore , Cx . univittatus may have a greater potential for transmitting arboviruses between birds and other vertebrates to humans , in contrast to the more anthropophilic An . gambiae . The vertically transmitted ISFVs , AeFV and CFAV , were only detected in samples from the Lake Victoria region , not only in Aedes mosquitoes , but also in Cx . pipiens ( AeFV ) , though we cannot fully rule out accidental Aedes mosquito contamination in the Cx . pipiens sample . While vertical transmission of ISFVs has been reported experimentally [63–66] , which may be as high as 90% [67] , this study corroborates its occurrence in natural ecologies [64 , 65 , 68 , 69] . Although ISFVs do not infect mammals and generally have been found to cluster within distinct phylogenetic clades associated with distinct mosquito genera [70–72] , Aedes flavivirus , which is phylogenetically distinct from related Culex flaviviruses , has previously also been found in Cx . pipiens mosquitoes sampled in Italy [73] . Our findings therefore support not only the vertical transmission of ISFVs in mosquitoes , but also the potential of occasional horizontal transmission between mosquito species and genera . Therefore , ISFVs in mosquito populations represent a promising model for the study of the evolution of host specificity of flavivirus infectivity [72] . Some ISFVs ( Palm Creek flavivirus and Culex flavivirus ) have been found to inhibit replication of West Nile and Murray Valley encephalitis viruses in the Ae . albopictus C6/36 cell line and in Cx . pipiens mosquitoes [65 , 66] . In contrast , CFAV , also identified in this study , has recently been found to increase susceptibility of dengue virus in an Ae . aegypti cell line ( Aa20 ) [74] and to be inhibited by the Wolbachia endosymbiont ( wMelPop ) used for dengue control in Ae . aegypti mosquitoes [75 , 76] . Because there is considerable variability in how ISFVs effect arbovirus superinfections , how vertical transmission of ISFVs affects the competence of mosquito populations to transmit arboviruses , either horizontally to vertebrate hosts or vertically to the next generation , remains largely unknown . We also detected An ( g ) FVs only in mosquito populations from the Lake Baringo region , despite the more than seven times greater sample size of An . gambiae tested from the malaria endemic Lake Victoria region . While it is curious that this ISFV was only detected in malaria mosquitoes from regions with relatively low malaria transmission rates [77] , they have been previously identified in An . gambiae and Anopheles squamosus mosquitoes from malaria endemic North-Eastern Province [26] , and Coastal and Western Provinces of Kenya ( Fig 3 ) . Other closely related Anopheles flaviviruses ( AnFVs ) ( Fig 3 ) have since been reported in anopheline mosquitoes from Australia [71] , Liberia and Senegal [78] , and Turkey [79] ( Fig 3 ) . Furthermore , transcriptionally active Flavivirus-derived endogenous viral elements have been identified in Anopheles minimus and Anopheles sinensis genomes via in silico and in vivo analyses [80] , which suggests a historical presence of ISFVs in anopheline mosquitoes . Though ISFVs may have important implications in the transmission of medically important arboviruses [70] , the study of AnFVs has been limited by their inability to replicate in standard Aedes cell line cultures , or even in cell lines of heterologous Anopheles species [26 , 71] . Appropriate Anopheles cell line cultures for the in vitro replication of the AnFVs will have to be established to further study their role in co-infection with other arboviruses , and possibly malaria parasites [26] . We recorded more diverse vector mosquito species and viruses in samples from Homa Bay County ( Table 2 ) , which concurs with reports from previous studies around the Lake Victoria basin [5 , 16 , 19 , 81] . Although adult Aedes mosquitoes have been sampled in both study areas [5 , 82] , we only sampled Aedes spp . larvae from Lake Victoria . In a previous study , we found that many of the suitable larval habitats for Ae . aegypti sampled in the Lake Victoria region correlated with increased ammonium and phosphate levels , which are key components of commonly used fertilizers [83] . Our larval sampling strategy may have been more favourable for sampling Aedes mosquitoes in the Lake Victoria region where agricultural activity is more intensive in comparison to the Lake Baringo region . Though there was an RVF virus outbreak in Baringo County in 2006/2007 and surveillance studies around the area reported possible mosquito vectors [11 , 84] , none of our mosquito samples from Baringo County tested positive for any pathogenic virus . Our identification of both a pathogenic arbovirus and three ISFVs in larval mosquitoes from both lake basins suggests complex ecologies involved in their circulation and maintenance . Although Omondi et al . [5] did not detect any virus from blood-fed mosquitoes around the Lake Victoria region where we found vertical transmission of Bunyamwera virus , AeFVs , and CFAV , the study found Bunyamwera virus in blood-fed mosquitoes from the Lake Baringo region , where we found no Bunyamwera infected larvae . Though these discrepant findings may be a result of inadequate sample size required to reliably identify specific arboviruses circulating in a region , the conditions for the maintenance of arboviruses by vertical transmission may depend on environmental factors of the mosquito vector’s reproductive environment . Nonetheless , our findings indicate that in the Lake Victoria region environmental context , An . gambiae , and possibly Cx . univittatus , can act as a reservoir that can both vertically and horizontally transmit Bunyamwera virus , ISFVs , and possibly other arboviruses . This is important towards understanding how arboviruses are maintained and geographically spread in different ecological contexts and can be used to forecast risks and improve prevention and other vector management strategies to mitigate future outbreaks . Continued arbovirus surveillance in diverse mosquito and other arthropod vector species in the region will help to more accurately identify the most important vectors of arboviruses possibly associated with febrile illnesses , while a better understanding of the role of ISFVs in the vertical transmission of arboviruses may open new control strategies . Insect-specific flavivirus NS5 gene sequences from twelve mosquito pools were deposited into the GenBank nucleotide database ( accessions MG372051- MG372060 , MK015647- MK015648 ) . | The East African region is endemic to diverse mosquito-transmitted arboviruses , though little is known about the role of vertical transmission in maintaining these viruses within mosquito vector populations during inter-epidemic periods . We sampled mosquito larvae from the Lake Baringo and Lake Victoria regions of Kenya and reared them to adults in the laboratory before screening them for mosquito-associated viruses by multiplex RT-PCR-HRM , cell culture , and sequencing . From the Lake Victoria region , we detected the arbovirus , Bunyamwera , which can cause febrile illness in humans , in Culex univittatus and vector competent Anopheles gambiae mosquitoes . We also identified diverse insect-specific flaviviruses in Aedes aegypti , Aedes luteocephalus , Aedes spp . and Culex pipiens mosquitoes . From the Lake Baringo region , we detected Anopheles flavivirus in An . gambiae mosquitoes . These findings demonstrate that naturally occurring vertical transmission potentially maintains viruses in circulation within the sampled vector species populations . Therefore , mosquitoes may potentially transmit a pathogenic arbovirus during their first bite after emergence . Because various insect-specific flaviviruses have recently been found to either inhibit or enhance replication of specific arboviruses in mosquitoes , their vertical transmission , as observed in this study , has implications as to their potential impact on both horizontal and vertical transmission of medically important arboviruses . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"pathogens",
"population",
"genetics",
"microbiology",
"animals",
"alphaviruses",
"gene",
"pool",
"viruses... | 2018 | Vertical transmission of naturally occurring Bunyamwera and insect-specific flavivirus infections in mosquitoes from islands and mainland shores of Lakes Victoria and Baringo in Kenya |
Dengue is a frequent cause of acute febrile illness with an expanding global distribution . Since the 1960s , dengue in Sri Lanka has been documented primarily along the heavily urbanized western coast with periodic shifting of serotypes . Outbreaks from 2005–2008 were attributed to a new clade of DENV-3 and more recently to a newly introduced genotype of DENV-1 . In 2007 , we conducted etiologic surveillance of acute febrile illness in the Southern Province and confirmed dengue in only 6 . 3% of febrile patients , with no cases of DENV-1 identified . To re-evaluate the importance of dengue as an etiology of acute febrile illness in this region , we renewed fever surveillance in the Southern Province to newly identify and characterize dengue . A cross-sectional surveillance study was conducted at the largest tertiary care hospital in the Southern Province from 2012–2013 . A total of 976 patients hospitalized with acute undifferentiated fever were enrolled , with 64 . 3% male and 31 . 4% children . Convalescent blood samples were collected from 877 ( 89 . 6% ) . Dengue virus isolation , dengue RT-PCR , and paired IgG ELISA were performed . Acute dengue was confirmed as the etiology for 388 ( 39 . 8% ) of 976 hospitalizations , with most cases ( 291 , 75 . 0% ) confirmed virologically and by multiple methods . Among 351 cases of virologically confirmed dengue , 320 ( 91 . 2% ) were due to DENV-1 . Acute dengue was associated with self-reported rural residence , travel , and months having greatest rainfall . Sequencing of selected dengue viruses revealed that sequences were most closely related to those described from China and Southeast Asia , not nearby India . We describe the first epidemic of DENV-1 in the Southern Province of Sri Lanka in a population known to be susceptible to this serotype because of prior study . Dengue accounted for 40% of acute febrile illnesses in the current study . The emergence of DENV-1 as the foremost serotype in this densely populated but agrarian population highlights the changing epidemiology of dengue and the need for continued surveillance and prevention .
Dengue virus ( DENV ) , a frequent cause of acute febrile illness with an expanding global distribution , is a flavivirus with 4 antigenically distinct serotypes ( DENV-1–4 ) that is transmitted by Aedes aegypti and Aedes albopictus mosquitoes [1] . Infection with a single serotype leads to long-term protective immunity against the homologous serotype but not against other serotypes [2] . In the past decades , urbanization , globalization , and lack of control of Aedes aegypti have contributed to the global spread of the disease [3] . It is now estimated that 50% of the world’s population live in areas where dengue virus has the potential to be transmitted [4] . Dengue was first serologically confirmed in Sri Lanka in 1962 and the initial outbreak was recorded in 1965 [5] . Although all four serotypes were known to be co-circulating in the country , dengue hemorrhagic fever/ dengue shock syndrome ( DHF/DSS ) was rare in Sri Lanka before 1989 [6] . From 1989 through 2008 , epidemics with periodic shifting of serotypes and increasing magnitude and severity occurred every few years , with the later epidemics being associated with a new clade of DENV-3 . However , since 2009 , there has been a dramatic increase in dengue transmission with an average of 33 , 000 cases reported per year . The predominant virus associated with these recent outbreaks has been a newly introduced genotype of DENV-1 [7] . In 2007 , our research team conducted etiologic surveillance of acute febrile illness at the largest tertiary care hospital in the Southern Province . Dengue was confirmed in only 6 . 3% of febrile patients seeking acute care at the hospital and no cases of DENV-1 were identified [8] . To re-evaluate the importance of dengue as an etiology of acute febrile illness in the Southern Province , we renewed fever surveillance at this large tertiary care hospital in 2012 . We report here the epidemiologic features of an epidemic of DENV-1 that occurred during our latter surveillance period .
We conducted a cross-sectional surveillance study for febrile illness from June 2012 through May 2013 in the adult and pediatric wards of Teaching Hospital Karapitiya , the largest ( 1 , 500 bed ) tertiary care hospital in the Southern Province of Sri Lanka . This hospital provides both primary and tertiary care to a catchment area of approximately 1 million people in the Galle District . The hospital also serves as the primary teaching hospital for the Southern Province ( total population 2 . 5 million ) and receives referrals from other hospitals in the province [9] . Consecutive patients ≥1 year of age with documented fever at presentation or within 48 hours of hospital admission were eligible for enrollment . We excluded patients who presented with focal bacterial infections , such as pneumonia or soft tissue infection . MBBS-qualified study physicians recorded epidemiologic and clinical data and trained phlebotomists collected an acute blood sample at the time of enrollment . Patients returned 2–4 weeks after enrollment for a convalescent blood sample or were visited at home if their address was known . Serum samples were stored promptly at -80°C and shipped on dry ice to the Duke-NUS Graduate Medical School in Singapore for laboratory testing . We required definitive serologic evidence ( IgG seroconversion ) , definitive virologic evidence ( positive PCR and either isolation or positive PCR with a second target ) , or both serologic and virologic evidence ( positive convalescent IgG and either positive PCR or isolation ) to confirm acute dengue . Acute primary ( first episode ) and acute secondary ( recurrent ) dengue were distinguished by the absence or presence of IgG in acute-phase serum samples , respectively , as defined previously [8] . Those with inconclusive acute dengue ( virologic evidence of dengue or flavivirus infection but not confirmed ) were excluded . Patients with positive IgG in both acute and convalescent sera but without evidence of acute infection were classified as past dengue . Seroprevalence was defined as the presence of IgG in acute-phase serum samples , as we have done previously [8] . The proportions of patients who met criteria for confirmed acute dengue were calculated . Associations between patients’ sociodemographic characteristics and dengue infection were evaluated . We compared epidemiological characteristics of patients with confirmed acute dengue versus no acute dengue . Categorical variables were compared using the Chi square test or Fisher exact test and continuous variables were compared using the t-test or Kruskall-Wallis test . We performed multivariable logistic regression to identify characteristics associated with acute dengue versus no dengue , adjusting for sex and age as a continuous variable . Rainfall data for the Galle region during the study months were obtained from the Sri Lanka Department of Meteorology , Colombo [13] . STATA , version 11 ( STATACorp , College Station , Texas ) was used for all statistical analyses . Written informed consent was obtained from patients or their guardians ( for patients <18 years of age ) and written assent was obtained from patients aged 12–17 years . The institutional review boards of Ruhuna University ( Sri Lanka ) , Johns Hopkins University , and Duke University Medical Center approved the study .
We enrolled 976 patients and obtained convalescent samples from 877 ( 89 . 6% ) . A total of 628 ( 64 . 3% ) were male and 306 ( 31 . 4% ) were children < 18 years . The median duration of hospitalization was 5 ( IQR 4–6 ) days . Six ( 0 . 6% ) patients were transferred to an intensive care unit and 4 ( 0 . 4% ) died in hospital . The median day of illness the acute sample was drawn was 4 days ( IQR 3–6 ) , and the median time between acute-phase and convalescent follow up was 23 days ( IQR 16–39 ) . The likelihood of follow-up was greater in children than in adults ( 94 . 8% versus 87 . 6% , p<0 . 001 ) , but did not differ by sex ( p = 0 . 61 ) , level of education ( p = 0 . 15 ) , or reported urban vs . rural residence ( p = 0 . 86 ) . The reported duration of fever ( p = 0 . 41 ) and illness ( p = 0 . 73 ) was similar in patients who did and did not return for follow-up . Overall , acute dengue was confirmed as the etiology for 388 ( 39 . 8% ) of the 976 hospitalizations for acute febrile illness during the study period . Acute dengue was either confirmed or excluded in 937 patients , since 39 had inconclusive evidence of acute dengue , Of the 388 with acute dengue , 100 were confirmed by IgG seroconversion . Most were virologically confirmed and by multiple methods , since 291 were confirmed by dengue PCR with a positive convalescent IgG , 161 by viral isolation with a positive convalescent IgG , 166 were both virus isolation and dengue-specific PCR positive , and 136 were dengue PCR and flavivirus PCR-positive . Few ( 15 ) virologically confirmed cases of acute dengue were viral isolation-positive but dengue PCR-negative; of these . 5 were flavivirus PCR-positive and the remainder had supportive serology ( convalescent IgG positive ) . The mean reported duration of illness was shorter ( 4 . 0 d versus 5 . 2 d , p = 0 . 006 ) among those confirmed by isolation and PCR versus than among those confirmed by seroconversion alone . Among 351 patients with laboratory-confirmed acute dengue who were dengue PCR or virus isolation-positive , 320 ( 91 . 2% ) were DENV-1 , 25 ( 7 . 1% ) were DENV-4 , and 6 ( 1 . 7% ) were DENV-2 . A dendrogram of representative isolates of DENV-1 is shown in Fig 1 . Among the 388 patients with acute dengue , 103 ( 26 . 5% ) had primary dengue , 245 ( 63 . 1% ) had secondary dengue , and 40 ( 10 . 3% ) could not be classified because insufficient acute sample was available to test for IgG . Among 937 patients in whom acute dengue was either confirmed or excluded ( 39 with inconclusive acute dengue removed from analysis dataset ) , 606 ( 64 . 7% ) were male and 645 ( 68 . 8% ) were adults . The median age of enrolled patients was 27 . 2 years ( IQR 14 . 3–42 . 2 ) . Dengue accounted for over half of febrile illnesses in adolescents and young adults ( those 15 to 39 years of age ) ; the proportion of acute febrile illness attributable to acute dengue in each age group is depicted in Fig 2 . Sociodemographic characteristics of those with acute dengue versus no acute dengue are compared in Table 1 . Reported residence in a rural area ( 38 . 7% versus 24 . 4% ) , a history of travel in the previous 30 days ( 27 . 9% versus 17 . 2% ) , and residing further from the hospital ( 25km versus 18km ) were more common in those with acute dengue than in those without acute dengue in bivariable analyses ( p<0 . 001 for all ) as well as in analyses adjusted for age and sex ( Table 1 ) . A total of 255 patients without confirmed acute dengue had laboratory results consistent with past dengue . The proportion of patients who were seropositive at enrollment increased with age from 30 . 7% in those <5 years old to 85 . 4% in those 45–49 years old ( Fig 3 ) . Presence of IgG at enrollment was similar in male and female patients ( 56 . 9% versus 54 . 1% , respectively ) and in rural and urban dwellers ( 54 . 9% versus 56 . 4% , respectively ) . Acute dengue occurred during each month of the study , with a clear increase associated with the seasonal monsoons . Acute dengue accounted for the largest proportion of febrile illness cases in October 2012 ( 82 . 3% ) , which was also the month with the greatest rainfall during our study period ( 519 . 5mm ) . Monthly variation in proportion of acute dengue and monthly rainfall for the Galle district are depicted in Fig 4 . Dengue virus isolates from this Galle cohort were closely related to each other as well as to viruses isolated from other areas in Sri Lanka . The other viruses from Sri Lanka were all isolated from patients in the Western Province , with the exception of one for whom the location could not be traced ( KP398852 ) . One closely related genotype 1 sequence was from Germany ( KJ468234 ) , but was isolated from a peripheral blood stem cell recipient with acute myeloblastic leukemia whose donor had recently contracted acute dengue while traveling in Sri Lanka [14] . The clade containing the Sri Lankan and German sequences also included three sequences from China and one from Taiwan . This clade is most closely related to sequences from China and Southeast Asia , especially Thailand and Vietnam ( Fig 1 ) , as has also been noted by others who , during the same period as this study , sequenced isolates from the Western Province and found them also most closely related to those from China [15] . Interestingly , samples from India were more distantly related despite the geographical proximity of Sri Lanka to India and primacy as trade partners ( https://atlas . media . mit . edu/en/profile/country/lka/ ) .
By repeating acute febrile illness surveillance at the largest tertiary care center in the Southern Province of Sri Lanka , we were uniquely positioned to describe the changing epidemiology of dengue at this hospital , which serves a large proportion of the region’s population . Comparing our study in 2012 with our previous study in 2007 , we detected an increase in acute dengue as a cause of AFI ( from 6 . 3% to 40% ) , a shift in serotype to DENV-1 in a susceptible population ( no DENV-1 identified in 2007 ) , and a new association of dengue with reported rural rather than urban residence . In short , in agrarian but still densely populated southern Sri Lanka , we identified dengue 1 as the cause of epidemic dengue associated with reported rural rather than urban residence . Expanded virologic testing and the high proportion of patients in whom virologic confirmation was achieved ( dengue virus isolated or PCR-positive or both ) provided conclusive evidence for DENV-1 as the cause of the dengue epidemic . This is the first report to describe epidemic DENV-1 in the Southern Province of Sri Lanka . The high proportion with acute dengue in 2012 versus 2007 ( 39 . 8% vs . 6 . 3% ) and seroprevalence in 2012 similar to that observed in the Western Province in 2012 rather than that observed in 2007 are consistent with an island-wide dengue epidemic that began in 2009 [16] . In 2012 , 44 , 461 cases of dengue were reported to the Sri Lankan Ministry of Health from throughout the country . Although this represents the largest number of cases of dengue reported to date in a single year in Sri Lanka , the majority of cases ( 70% ) were only clinically diagnosed . In contrast , we used rigorous diagnostic methods to conservatively but conclusively document the magnitude of the DENV-1 epidemic at this hospital , which provides both primary and tertiary care for a large proportion of patients residing in the region [17] . We assert that the population studied is representative of the general population in the area , since only 6% of hospital admissions during 2012–2013 were due to transfers from other hospitals . Additionally , care at government run Teaching Hospital Karapitiya is free for all . Finally , although we enrolled only those hospitalized in this study , 92 . 6% of those with illness subsequently confirmed as acute dengue were hospitalized in our prior study [8] . This is in accord with Sri Lankan guidelines for management of dengue , wherein those with platelet counts of less than or equal to 100 , 000 are hospitalized . In our study , dengue accounted for at least 10% of acute febrile illness during each month of the study period , with a high of 82% in October 2012 and a low of 10% in March 2013 . Notably , October 2012 was the month with the greatest rainfall during our study period ( 519 . 5mm ) whereas March 2013 had relatively less rainfall ( 114 . 1mm ) [13] . The association between dengue and rainfall has been previously described , since rainfall produces conditions that are favorable for reproduction and survival of the vector mosquitoes [18 , 19] . Further surveillance studies must be continued in this region to delineate patterns of disease associated with seasonal and climatic changes such that appropriate prevention measures can be planned and carried out . Traditionally , dengue has been considered an urban disease , and disease surveillance and prevention efforts have focused on the control of Aedes aegypti larval habitats in urban-associated containers used for storing water , used tires , and non-biodegradable plastic and metal packaging for consumer goods [20] . Although dengue has been reported from throughout Sri Lanka , most cases have been recognized in the heavily urbanized Western Province [7 , 21–24] . In this study , reported rural residence was surprisingly more common in those with acute dengue than in those with no evidence of acute dengue . There have been increasing reports of dengue in rural areas worldwide , but this pattern of transmission is not yet fully understood [25–28] . Travel to urban centers that have a heavy burden of disease , secondary vectors such as Aedes albopictus , and stored water containers because of uncertain water supply have all been implicated as reasons for the increasing prevalence of dengue in rural areas [25] . In our study , a history of recent travel prior to illness was associated with acute dengue . Although we collected information on occupation and travel within 30 days and found these variables to be associated with higher risk of acute dengue infection , we could not determine the exact location where dengue had been acquired . However , in Sri Lanka up to 28% of the labor force in the urban Western Province consists of workers from outside the region who commute for work [29] . Galle is located only 116 km from the urbanized capital of Colombo and is agrarian but still densely populated . As new routes of transportation , such as the Southern Expressway built in 2013 , are introduced between the two cities , it is expected that inter-city travel will increase . Further study of patterns of human travel as well as on the characterization of the vector may be important in control efforts . The vast majority of dengue in our study was caused by DENV-1 serotype , which represents a marked shift from our findings in 2007 when only DENV-2 , DENV-3 , and DENV-4 were isolated and DENV-3 was the most prevalent [8] . This shift in serotype is similar to what has been observed in the Western Province [6 , 7] . Although DENV-1 has been co-circulating in Sri Lanka for decades , large annual epidemics of DENV-1 have only occurred since 2009 with the appearance of a new DENV genotype-1 strain that may have been introduced from China or Thailand [15] . Our phylogenetic analysis revealed that the DENV-1 genotype in southern Sri Lanka is similar to the one that is circulating in the Western Province with both most closely related to those from China , which supports the theory that migration of humans between the two cities may have contributed to the spread of the disease [15] . Our prior surveillance in 2007 suggested that our study population lacked DENV-1 antibodies and was susceptible to the new DENV-1 strain , since dengue antibodies are generally type-specific [2] . In 2007 , we found that the dengue seroprevalence reached a plateau of 70% by 40 years of age [8] . In contrast , in this study , nearly 70% were seropositive by age 20 and the seroprevalence reached 90% in those 60–65 years of age . Strengths of our study include reproducible and objective enrollment criteria ( unselected patients with documented fever of defined magnitude ) and use of gold standard diagnostic criteria ( paired IgG serology as well as isolation and PCR ) , which was made possible by a high proportion ( 89 . 6% ) in whom convalescent clinical and serological follow-up was achieved . In 2012 as in 2007 we tested paired sera by IgG ELISA as per Chungue et al . [10] , distinguished primary vs . secondary dengue based on the absence or presence of IgG in acute-phase sera , and performed isolation and PCR to serotype dengue . Our estimate of secondary dengue in 2012 is relatively conservative , since we used IgG ELISA as a qualitative test and some proportion of those IgG-positive in both acute and convalescent sera but with negative virologic testing may have had secondary acute dengue . Identification of epidemic dengue 1 at our site in southern Sri Lanka is consistent with reports of epidemic dengue elsewhere in the country by the Sri Lankan Ministry of Health [30] . In conclusion , we describe the first documented epidemic of DENV-1 in the Southern Province of Sri Lanka in a population susceptible to this serotype . Our report expands understanding of the changing epidemiology of acute dengue in the region with acute dengue infections in southern Sri Lanka associated with reported rural ( not urban ) residence , rainfall , travel , and residence in rural areas ( median~ 25km ) outside of Galle . Our epidemiologic and phylogenetic analyses highlight the importance of recognized internal travel but also unrecognized international travel , as evidenced by the relatedness of isolates from Sri Lankans to that a German stem cell transplant recipient who had not traveled but his donor had [14] . Our continued surveillance allowed us to detect and describe a marked change in the epidemiology of dengue with the emergence of DENV-1 in this region of southern Sri Lanka . Further characterization of isolated strains and larger population-based epidemiologic studies will be critical to elucidate the geographic expansion of DENV-1 , its pathogenicity relative to other serotypes and independent of whether associated with secondary vs primary dengue , and the most effective methods to prevent and control the disease . | Dengue is a globally emerging cause of fever that is caused by four serotypes of virus ( DENV1-4 ) . Since the 1960s , dengue in Sri Lanka has been documented primarily along the heavily urbanized western coast . Outbreaks from 2005–2008 were attributed to a new type of DENV-3 and more recently to a newly introduced type of DENV-1 . Having identified serotypes other than DENV-1 as an important cause of fever in southern Sri Lanka in 2007 , our prospective surveillance in 2012 enabled detection and characterization of an epidemic of dengue that caused 388 ( 39 . 8% ) hospitalizations for fever . DENV-1 accounted for 91 . 2% of 351 virologically confirmed cases; isolates were most closely related to sequences described from China and Southeast Asia , not nearby India . Acute dengue was associated with self-reported rural residence , travel , and rainfall . Emergence of DENV-1 as the foremost serotype in the densely populated but agrarian southern Sri Lanka highlights the changing epidemiology of dengue and need for continued surveillance and prevention . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"sequencing",
"techniques",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"geographical",
"locations",
"microbiology",
"tropical",
"diseases",
"viruses",
"rna",
"viru... | 2016 | Emergence of Epidemic Dengue-1 Virus in the Southern Province of Sri Lanka |
Different data types can offer complementary perspectives on the same biological phenomenon . In cancer studies , for example , data on copy number alterations indicate losses and amplifications of genomic regions in tumours , while transcriptomic data point to the impact of genomic and environmental events on the internal wiring of the cell . Fusing different data provides a more comprehensive model of the cancer cell than that offered by any single type . However , biological signals in different patients exhibit diverse degrees of concordance due to cancer heterogeneity and inherent noise in the measurements . This is a particularly important issue in cancer subtype discovery , where personalised strategies to guide therapy are of vital importance . We present a nonparametric Bayesian model for discovering prognostic cancer subtypes by integrating gene expression and copy number variation data . Our model is constructed from a hierarchy of Dirichlet Processes and addresses three key challenges in data fusion: ( i ) To separate concordant from discordant signals , ( ii ) to select informative features , ( iii ) to estimate the number of disease subtypes . Concordance of signals is assessed individually for each patient , giving us an additional level of insight into the underlying disease structure . We exemplify the power of our model in prostate cancer and breast cancer and show that it outperforms competing methods . In the prostate cancer data , we identify an entirely new subtype with extremely poor survival outcome and show how other analyses fail to detect it . In the breast cancer data , we find subtypes with superior prognostic value by using the concordant results . These discoveries were crucially dependent on our model's ability to distinguish concordant and discordant signals within each patient sample , and would otherwise have been missed . We therefore demonstrate the importance of taking a patient-specific approach , using highly-flexible nonparametric Bayesian methods .
Molecular data show great promise to stratify patients into distinct subgroups that are indicative of disease development , response to medication and overall survival prospects [1] . Such subgroups are highly useful in informing treatment decisions [2] , [3] . Most current computational diagnostic approaches are based on gene expression data and cluster patients by co-expression of genes . For example , multivariate gene expression signatures have been shown to discriminate between disease subtypes , such as recurrent and non-recurrent cancer types or tumour progression stages [3]–[6] . In addition to expression data there are also many other data types that can be informative about a patient's disease status . For example , somatic copy number alterations provide good biomarkers for cancer subtype classification [7] . For this reason , the focus of research has recently shifted towards integrative clustering of complementary data types , e . g . [8] . The goal of integrative analysis is to identify clusters of samples that share not only expression profiles , but also other molecular characteristics such as copy number alterations . The subtypes of tumours identified in this way are more likely to share the same regulatory programs and underlying genomic alterations . Data integration for subtype discovery poses several challenges that we address in this paper . Challenge 1: Separating concordant from contradictory signals . While different molecular data are expected to share complementary information on common cellular processes , they can also contain contradictory signals because of the complexity of living cells and noise in the data . For example , genomic gains and losses may or may not be accompanied by concordant expression changes of the genes in the altered regions . The level of concordance may differ dramatically from patient to patient due to cancer heterogeneity . However , most existing integrative methods force different data types to be fused in all samples without reference to whether the data are concordant or contradictory in each patient . Challenge 2: Selecting informative features . Identifying which measurements are informative about the underlying subtypes is particularly important when using genomic data because the number of measurements can be very large , e . g . in the tens of thousands or more in the case of microarrays . Because a priori we expect only a fraction of measurements to contain useful clustering information , extracting these features accurately will improve the quality and stability of clustering outcome . Additionally , identifying the relevant biological features can inform us about the underlying processes driving the disease . Challenge 3: Estimating the number of subtypes . In many clustering algorithms this number is a parameter that needs to be set by the user [8] . Afterwards , the quality of the clusterings need to be compared , e . g . using stability indices [9] . However , jointly estimating the clusters together with their optimal number in a unified framework can improve results , because the most likely number of clusters can be inferred directly from the data . These three challenges are not independent of each other: Whether or not the data show concordant signals for a subgroup of patients has a direct effect on which features should be selected as informative , which in turn has a direct influence on the estimate of the number of clusters . Thus , all three challenges need to be treated in an unified model . Our approach is Patient-specific Data Fusion ( PSDF ) by Bayesian nonparametric modeling . In this paper , we propose a statistical model based on a two-level hierarchy of Dirichlet Process ( infinite mixture ) models ( DPMs ) [10] , [11] that integrates copy number and expression data to jointly classify patients into cancer sub-groups . This model is an extension of the model presented in [12] , modified to include a method of feature selection and adjusted to address a different problem with a number of advantages: Thus , the model not only identifies copy number alterations driving gene expression changes but simultaneously finds differences in regulation that distinguish one cancer subtype from the other . In doing so it explores the basic scientific question to which extend copy number data can be fused with expression data in integrative cancer studies . everal integrative clustering approaches have been proposed in the literature [8] , [13] , [14] . A recent method is iCluster [8] . iCluster is based on a k-means approach that is extended to include more than one data type and performs feature selection in each data type independently . iCluster is fast and easily applied to more than two data types . However , compared to iCluster we have a more flexible mixture model underlying our own approach that in particular does not need the number of clusters ( the ‘k’ in ‘k-means’ ) to be specified beforehand . In contrast to our model , iCluster assumes that both data are informative for all patients without checking for patient-specific consistency . In two case studies with cancer data sets [7] , [15] , we will show what impact these differences have and that our model compares favourably with iCluster in clinically important analysis results .
Bayesian nonparametric modeling provides a principled way to learn unknown structure in the data . Dirichlet Process ( infinite mixture ) models ( DPMs ) [10] , [11] are Bayesian nonparametric models that have been widely used for clustering [18]–[25] . DPMs give us a sound interpretation of common cluster membership , that the data for those samples are drawn from the same underlying distribution . They also allow us to infer the most likely number of clusters given the data as part of the unified model . PSDF groups patient samples on the basis of both gene expression and copy number alteration data . It also simultaneously distinguishes , on a sample-by-sample basis , between samples that can share concordant signal across the data types ( fused ) and those for which there is contradiction ( unfused ) . We note that throughout this paper we will use the following terminology , relating to the concordance ( or otherwise ) of the two data sets for a given patient . The breast cancer data from [15] contains both copy number and expression data for 106 tumour samples , with 26 , 755 copy number probes and 37 , 411 expression probes . Even for a clustering method with feature selection capability , it is convenient to remove the mostly obviously uninformative “noise” features . To preselect features with functional implications in a principled , controlled manner , we take the following steps . First , copy number data are filtered based on whether there is a concomitant change between a locus's copy number and its own expression . This is to exclude passenger events without explicit downstream effects . Each expression probe is matched to its nearest copy number probe allowing for multiple matches , i . e . a copy number probe can be matched to multiple expression probe . This resultes in 37 , 411 matched pairs of copy number and expression data annotated by expression probes . We then calculate the adjusted -values of the correlations of each pairs of copy number and expression probes , and a copy number probe is selected if the corresponding -value is smaller than 0 . 1 . Still there are highly similar copy number profiles among the selected copy number probes . To remove redundancy , copy number data of the selected probes are then merged based on their similarity using CGHregions [26] , which results in 379 regions . Finally , both of the copy number signatures from the merged regions and all expression profiles passing the above -value threshold are ranked by the Wald test in predicting breast-cancer-specific survivals . The best 200 of each type of data are used for clustering . For the prostate cancer data set , there are 150 tumour samples with both copy number and expression data [7] . The expression data were profiled with Affymatrix Human Exon 1 . 0 ST array which contains 229 , 581 probes after quality filtering . For the copy number data , there are 43 , 416 probes on Agilent 244K array comparative genomic hybridization array . To extract features , we use a slightly different approach since the scale of this data set is much larger than that of the breast cancer data . Substantially larger number of probes compared to the breast cancer study means that the probe-centric method is not suitable , hence we take a gene-centric method by aggregating copy number and expression data to 12 , 718 genes based on array annotation . For copy number data , the aggregation is done by taking the median for probes within a gene . For the expression , the probe most highly correlated with the copy number profile of a gene is chosen to represent this gene . Even if so , only modest correlations are observed between the two data types . Finally , 286 genes with highly correlated copy number and expression ( adjusted ) from the two data sets are used as clustering input .
This paper explores the potential of patient-specific data fusion to enhance prediction power in cancer subtype discovery . Cancer subtype discovery combining both genomics and transcriptomics leads to a more comprehensive understanding of the heterogenous cellular contexts . By using a flexible , nonparametric model such as the model presented in this paper , we can learn both the concordant and contradictory structures underlying those multiple data types . This structure leads to an improved understanding of the functional components and pathway regulations for each cancer subtype , something that is essential for the future development of targeted therapeutics . Our contributions are therefore as follows . With both breast cancer and prostate cancer data , PSDF is able to discover poor outcome subtypes with early-stage , highly frequent recurrences/deaths . These subtypes are not identified by other methods which either force to fuse data on all samples , or cluster patients based on single data type . We show that there exist both concordant and contradictory signals in these data , which , when forced to cluster together , can result in inferior subtype identification . Moreover , data fusion is necessary in predicting both events and timing of cancer survivals/recurrrences . Hence , taking this approach is vital in the discovery of new disease subtype consisting of early-stage events . A promising aspect of studying cancer subtypes is the identification of key pathways altered unique to this subtype . Our network analyses show functionally interacting genes in the subtype-specific network modules whose deregulations may contribute to the poor outcome of a cancer subtype . The pathway enrichment analysis facilitates functional interpretation of the new clusters/subtypes in a coherent manner with the network modules . Under-lying driver events for poor outcome may be revealed during this process , such as the over-expression of the Cell Cycle pathway in breast cancer , and the under-expression of Endocytosis and Chemokine signaling pathway in prostate cancer . Further exploration of these results may lead to the discovery of new genes participating in the cancer-related pathways , as well as the identification of treatment target and the development of pathway inhibitors . Our analysis results also highlight the difference between different cancer types . Previously , relatively low concordance between prostate cancer copy number and expression has been reported [17] , in contrast to the high-level correlations generally observed in breast cancer . In addition , unlike breast cancer where RNA expression are predictive of recurrence , copy number changes in prostate cancer have been found to outperform expression in prediction [7] . Different degrees of concordance in the data lead to significantly different clustering results – while fused clusters in highly concordant breast cancer data are prognostic , an unfused subtype in prostate cancer turns out to be extremely aggressive . The results from the breast and prostate cancer data sets are in fact strong statements that different cancer types should be treated differently by statistical methods . Hence , a versatile tool such as PSDF is particularly suitable for this field .
The naive Bayes data model used in [12] models data for a given feature as being drawn from a multinomial distribution with unknown class probabilities . Choosing a conjugate ( Dirichlet ) prior , these unknown class probabilities can be marginalised out to give a marginal likelihood for each feature in each cluster . ( 1 ) Where and , is the index over features and is the index over discrete data values . The are the Dirchlet prior hyperparameters , which in this case are set to match the known proportions of each data value in the data set ( which is prior knowledge here , as we define the data discretisation ) . These proportions are scaled to sum to 1 . 5 , which is the sum of the Jeffreys' value ( 0 . 5 ) over the three possible data values , hence representing only a weakly-informative constraint . To perform feature selection , we will consider two different likelihoods for a given feature , corresponding to the feature being off/on , as denoted by an indicator variable . For , we simply use the multinomial-Dirichlet marginal likelihood , as before . For , we fix the class probabilities to the expected prior values , given the spread of discrete input values for the given feature . ( 2 ) Where again is the index over features and is the index over discrete data values . The are simply taken as the proportion of each data value in a given feature across the whole data set , with a minimum count of one assigned to each data value . ( 3 ) Where and are required to have minimum of one count per class . This has the effect of defining an ‘indifference’ likelihood , where it makes no difference to the overall posterior ( for the given feature ) to which cluster any given sample is assigned . It is straightforward to write down the conditional distribution for a single indicator variable , so we Gibbs sample each in turn when producing a new MCMC sample . The switching on/off of a given feature can be regarded as a kind of model selection . Considering the limit of many samples ( and hence negligible uncertainty in the value of the class probabilities for ) , the ‘indifference’ likelihood is simply the expected case if the samples are randomly assigned to clusters . For finite numbers of samples , the ‘indifference’ likelihood is inherently simpler ( in the sense that the class probabilities are known ) , so the feature selection becomes a competition between this simplicity and the greater ability of the case to explain non-random cluster assignments . To give improved mixing , we run 50 MCMC chains for each analysis . The chains are samples long , with the first removed as a burn-in . The remainder are sparse-sampled by a factor of 10 for computational convenience and then used to produce the outputs . All chains are examined using the R package CODA . In particular , the time-series and histograms for each parameter/chain pair are examined by eye for any obvious anomolies that would indicate incomplete mixing . The multiple MCMC chains are used to compute uncertainties in statistics of interest ( for example , the probability that a given feature is selected ) . This gives us a direct measure of chain mixing quality . Each chain runs to completion in less than 48 hours on nodes of the University of Warwick's high performance computer cluster . In order to validate our model , we performed a simulation study . We constructed a pair of synthetic data sets . For each synthetic data set , we started with the 106 signal items and 200 signal features in the copy number variation data from [15] ( which is also analysed in Section . These items will therefore ( by construction ) be fused as they share identical clustering structure across the two synthetic data sets . We note that this is a reasonable test of the method because in the real analyses both copy number and gene expression data sets are discretised into three levels . These synthetic data represent a good way of constructing items that share concordant signals across the two data sets . To each synthetic data set , we then added 50 noise items . These items are drawn by replacement from the signal items and are drawn separately for each synthetic data set . For example , a given noise item may be a copy of signal item 15 in the first synthetic data set , and signal item 59 in the second synthetic data set . These noise items are therefore drawn from the existing clustering structure of each synthetic data set , but in general they will not be fused ( excepting the case where by coincidence they are both drawn from the same underlying cluster ) . This then gives us 156 items in total . Finally , we added to each synthetic data set 200 noise features . The data for these features are drawn with replacement from the original data . Therefore , while they reflect the distribution of data values in the signal features , they are entirely random and without clustering structure . As such , we expect them o be rejected by feature selection . Table 1 shows the results of an analysis of these synthetic data . The method successfully rejects all 400 noise features across the two data sets . 8 signal features are also rejected at this level , but we note that some level of feature rejection is expected of signal features , as some of them will be uninformative . The method successfully finds 105 of the 106 fused items . It also identifies 17 of the noise items as being fused . We note that we expect some level of coincidental fusion for the noise items , where they happen to have been drawn from the same cluster . For example , if we assume there are 5 ( equally-sized ) underlying clusters in the copy number data , we expect coincidentally fused noise items . We note that here , 25 MCMC chains of length samples are sufficient to achieve reasonable convergence . We conclude that our method performs well in identifying both fused/unfused items and selecting appropriate features in each data set . | The goal of personalised medicine is to develop accurate diagnostic tests that identify patients who can benefit from targeted therapies . To achieve this goal it is necessary to stratify cancer patients into homogeneous subtypes according to which molecular aberrations their tumours exhibit . Prominent approaches for subtype definition combine information from different molecular levels , for example data on DNA copy number changes with data on mRNA expression changes . This is called data fusion . We contribute to this field by proposing a unified model that fuses different data types , finds informative features and estimates the number of subtypes in the data . The main strength of our model comes from the fact that we assess for each patient whether the different data agree on a subtype or not . Competing methods combine the data without checking for concordance of signals . On a breast cancer and a prostate cancer data set we show that concordance of signals has strong influence on subtype definition and that our model allows to define prognostic subtypes that would have been missed otherwise . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"algorithms",
"medicine",
"oncology",
"computer",
"science",
"mathematics",
"basic",
"cancer",
"research",
"statistics",
"biology",
"computational",
"biology",
"biostatistics"
] | 2011 | Patient-Specific Data Fusion Defines Prognostic Cancer Subtypes |
Metabolic fingerprinting analysis can offer insights into underlying reactions in a biological system; hence it is crucial to the understanding of disease pathogenesis and could provide useful tools for discovering biomarkers . We sought to examine the urine and plasma metabolome in individuals affected by urogenital schistosomiasis and its associated-bladder pathologies . Blood and midstream urine were obtained from volunteers who matched our inclusion criteria among residents from Eggua , southwestern Nigeria . Samples were screened by urinalysis , microscopy , PCR and ultrasonography , and categorised as advanced ( urogenital schistosomiasis associated-bladder pathologies ) , infection-only ( urogenital schistosomiasis alone ) and controls ( no infection and no pathology ) . Metabolites were extracted and data acquired with ultra high-performance liquid chromatography coupled with Thermo Q-Exactive orbitrap HRMS . Data was analysed with MetaboAnalyst , Workflow4Metabolomics , HMDB , LipidMaps and other bioinformatics tools , with univariate and multivariate statistics for metabolite selection . There were low levels of host sex steroids , and high levels of several benzenoids , catechols and lipids ( including ganglioside , phosphatidylcholine and phosphatidylethanolamine ) , in infection-only and advanced cases ( FDR<0 . 05 , VIP>2 , delta>2 . 0 ) . Metabolites involved in biochemical pathways related to chorismate production were abundant in controls , while those related to choline and sphingolipid metabolism were upregulated in advanced cases ( FDR<0 . 05 ) . Some of these human host and Schistosoma haematobium molecules , including catechol estrogens , were good markers to distinguish infection-only and advanced cases . Altered glycerophospholipid and sphingolipid metabolism could be key factors promoting the development of bladder pathologies and tumours during urogenital schistosomiasis .
Among the most prominent neglected tropical diseases ( NTDs ) is schistosomiasis , a helminthic disease caused by Schistosoma spp . Its urinary form , caused by S . haematobium and known as urogenital or urinary schistosomiasis , is widespread in Africa and the Middle East . In chronic cases , infected persons may experience abdominal pain , enlarged liver , paralysis , granuloma formation , blood in the urine and the risk of early onset and aggressive bladder cancer [1] . A population of more than 200 million in different countries is at risk of schistosomiasis [1] and more than 100 million are said to be affected by urogenital schistosomiasis [2] . Several studies in different parts of Nigeria have reported moderate to high prevalence of urogenital schistosomiasis [3 , 4] . There is as yet no effective vaccine , and there are some reports of drug resistance due to over-reliance on praziquantel , the major drug in use [1 , 5] , therefore there is a continuous search for molecular targets for the development of vaccines and chemotherapeutics . Urogenital schistosomiasis has been associated with different forms of bladder pathologies in Nigeria and with bladder tumours in parts of Africa [6 , 7] . The molecular intricacies involved in the development of bladder tumours during urogenital schistosomiasis are not clearly defined , and the tumours are usually preceded by abnormal morphologies or pathologies especially in the bladder . Recent studies on the mechanisms of tumour development in schistosomiasis have highlighted the role of estrogen-related molecules from the parasite , as such molecules were found in many urine samples from infected persons in Angola [8 , 9 , 10] . Also , it was recently shown that an inflammation-regulatory microbiome could play important roles in the maintenance or development of bladder pathologies during infection [11] . Analysing the metabolome , the milieu of compounds in a body fluid , gives a comprehensive idea about the internal body reactions and its products [12] . This can be influential in biomarker discovery , diagnosis of disease and health conditions , and offer insights into disease mechanisms [13] . Indeed , it has led to efforts such as the Human Serum Metabolome and the Human Urine Metabolome projects [14] . Given the pathology of urogenital schistosomiasis , a deep understanding of the metabolome is important . The aim of the present study is to examine the metabolome features in urine and plasma samples from persons infected with urogenital schistosomiasis and related pathologies in a rural population in Nigeria; and offer insights into host-parasite interaction and induction of bladder pathologies during urogenital schistosomiasis .
The study protocol was approved by the University College Hospital/University of Ibadan Review Committee , as well as the Ogun State Ministry of Health . Ethical considerations were reported earlier [11] . Briefly , all adult participants were recruited into the study after giving written informed consent . Participants were informed of the purpose of the study , the health risks associated with the sample collection methods , and the potential benefits to public healthif biomarkers were to be discovered . Interviews and questionnaires were administered in the local language . Participants were recruited from Eggua community in Ogun State , southwestern Nigeria . Sampling was carried out between December 2014 and June 2015 . Sampling procedures and detection of S . haematobium infection and bladder pathologies were carried out as reported earlier [11] . Samples were immediately anonymised and aliquoted for microscopy . An aliquot of samples was immediately kept in ice chest and transported in dry ice to the laboratory , where they were kept at -80°C prior to analysis . Participants were interviewed to obtain information on demographics and lifestyle . They provided blood samples from which plasma was isolated and midstream clean-catch urine samples in the morning hours upon instructions . Urine microscopy and PCR were used to confirm infection status . Bladder scans were carried out with TitanUltraSystem ( Sonosite , USA ) by a radiologist . No confirmed bladder cancer cases were detected but various forms of pathologies ( abnormal morphologies ) were observed; hence , samples were grouped based on presence of infection with pathologies ( advanced ) , infection without pathologies ( infection-only ) , pathologies without infection ( pathology-only ) and controls ( no infection or pathology ) . Both urine and plasma samples were prepared using chilled methanol:water mixture ( 4:1 ) , following previous methods [15] . An Accela ultra high-performance liquid chromatography ( UHPLC ) system ( ThermoFisher , USA ) , coupled online via heated electrospray ionization source ( HESI ) to a mass spectrometer was employed for non-targeted metabolomics profiling . Separation was achieved with 150mm x 2 . 1mm , 1 . 9μ HypersilGold column with a 5μl injection volume . The temperature of column oven was set at 40°C and the sample manager was maintained at 4°C . Gradient elution was performed using 0 . 1% formic acid in water ( A ) and acetonitrile ( B ) as mobile phase after modification and optimization of previous methods [9] , and the elution was run witha mobile phase gradient of 0-6min , 100% A; 6-8min , linear gradient from 100% to 80% A; 8-12min , linear gradient from 80% to 40% A; 12–14min , linear gradient from 40% to 70% A; 14-16min , linear gradient from 70% to 80% A; 16-20min , linear gradient from 80% A to 100% B . The column was washed between each sample for stability and to eliminate any carry-overs . The flow rate was 0 . 35 ml/min . MS acquisition was performed on the Q-Exactive orbitrapmass spectrometer ( ThermoFisher , MA , USA ) operated in positive and negative electrospray ionization ( ESI ) modes . The sample sequence was set to random and samples were run in triplicates . Each sample type and ESI mode were run in a batch . In the ESI+ mode , the MS spray voltage was 4 . 2 KV while it was 3 . 6 KV in the ESI− mode . The capillary temperature was set at 300°C and probe heating temperature at 320°C with the sheath gas at 45 arbitrary units . For ESI+ and ESI− mode , the aux gas was set at 5 and 12 arbitrary units respectively . The tube lens was set to 50V and the mass scan range was set from 70 to 1000 m/z . The resolution of the orbitrap was set at 70 , 000 . Both ESI modes were used to analyse urine and plasma samples ( Table 1 ) , but forty-two of the urine samples were analysed in the negative run due to inadvertent loss of samples . Profile mode raw data were converted to centroid mode mzXML files with MSConvert and subjected to XCMS and CAMERA for pre-processing [16 , 17] . Peaks were identified with the xcmsSet algorithm using the centwave method , a mass tolerance of 3 ppm , and peakwidth range between 10–50 seconds . Peaks were matched ( bandwidth = 5 , mzwid = 0 . 015 ) , retention time aligned using obiwarp method , peaks were regrouped and filled . They were annotated using xsAnnotate , groupFWHM ( perfwhm = 0 . 6 ) , findIsotopes ( mzabs = 0 . 01 ) , groupCorr ( cor_eic_th = 0 . 75 ) , and findAdducts parameters . The Workflow4Metabolomics galaxy server ( https://galaxy . workflow4metabolomics . org/ ) [18] and MetaboAnalyst ( http://www . metaboanalyst . ca ) [19] were also used for processing and identification . Data was filtered ( relative standard deviation ) and normalized . Identification of metabolites were at least Level 2 identification . Accurate mass , spectral library match ( low level MS/MS spectra ) , retention time , annotations ( adduct , isotope combinations ) , and biological context ( organism and type of body fluid of previous identification ) were all used in comparison to available data with similar analytical procedures in The Human Metabolome Database [14] , LipidMaps and Metlin database [20] , allowing for molecular weight tolerance of 0 . 5Da . The naming of the metabolites are putative and metabolite family name is provided when more than one match occurs . Metabolite features that varied significantly between sample groups were evaluated with Mann Whitney ( two groups ) or Kruskal-Wallis ( more than two groups ) tests with False Discovery Rate ( FDR ) correction , and inter-group separation based on the features were modeled with Partial Least Square Discriminant Analysis ( PLSDA ) and Principal Component Analysis . Analysis of feature importance for selection was carried out with PLSDA , Significance Analysis of Microarrays ( and Metabolites ) SAM and RandomForest trees [21 , 22 , 23] . Selected features were evaluated for use as biomarkers using multivariate ROC curve analysis [24] with logistic regression models .
There were differences in LC-MS features of urine and plasma samples from healthy controls , urogenital schistosomiasis and urogenital schistosomiasis induced-bladder pathologies ( Fig 1A and 1B ) , especially from 6 . 5 to 12 minutes in urine samples ( Fig 1A ) . The differential features , after univariate analysis ( p<0 . 01 ) , were subsequently filtered using FDR correction ( FDR<0 . 01 ) ( S1 File ) . In pairwise comparison of the study groups , the features which differentiate induced-bladder pathologies ( advanced ) from infection-only were far fewer in both modes and sample types than those which differentiate advanced from pathology-only ( S2 File ) . Pathway analysis of the peaks detected is presented in S3 File . The number of the filtered significant features ( FDR<0 . 01 , fold change >2 ) was different depending on whether ionization mode was positive or negative . For plasma , among the four sample groups , 26% ( 2334 ) of the detected features were significantly different in the ESI negative mode , and 15% ( 983 ) in the positive mode ( FDR<0 . 05 ) . When plasma samples were simply grouped into two , infected and non-infected , 18% ( 1562 ) of the detected features were significantly different in the negative mode , and 4% ( 237 ) of the features detected in positive mode were significantly different ( FDR<0 . 05 ) . In urine samples , 648 ( 10% ) differentiated the four groups in negative mode ( FDR<0 . 05 ) . When urine samples were grouped into two , infected and non-infected , 834 ( 13% ) differentiated infected and non-infected in negative mode; and 891 ( 7% ) differentiated the same groups in positive mode . For multivariate analyses , using spectral data , PCA and PLSDA were performed and 95% confidence ellipses drawn . Both methods showed a trend of inter group axes separation , but as may be expected , PLSDA gave better axes separation than PCA ( Fig 2 ) . For PLSDA , Q2 validation was highest in urine samples in negative mode ( > 0 . 99 ) and least for plasma in positive mode ( 0 . 8 ) ( Fig 3 ) . The scores plots of the PLSDA model also show that the inter-group discrimination ( and therefore , the choice of representative metabolites for the various groups ) could be maximized with several distinct metabolome features ( Fig 2 ) . To select the most important metabolites in the high dimensional metabolomics data , PLSDA models , Significance analysis of Microarrays and Metabolites ( SAM ) and RandomForest ( RF ) trees were utilized . For SAM , a minimal false positive value ( ≤ 1 ) and relatively high delta ( delta>2 ) were set in order to obtain accurate measurement of differentially expressed features . Also , the metabolite features were subjected to Random Forest’s algorithm , using 500 trees and 7 predictors , to determine the metabolites with the most important contribution . The highest RF misclassification occurred with plasma samples , with Out of the Bag ( OOB ) error of 0 . 0631 . For PLSDA models , Variable Importance in Projection ( VIP ) scores were set at >2 . As a result of variation in the top features selected by the three algorithms , metabolites which were eventually selected for biomarker evaluation were chosen systematically . Biomarkers are expected to be robust in order to fit varying conditions . Hence , the goal of the systematic choice was to take advantage of overlapping top-ranked metabolites from three algorithms from a pool of metabolites detected in both positive and negative modes . First , metabolites from both ESI modes were combined , then those that were ( 1 ) significantly different among all the four study groups ( FDR<0 . 05 ) , ( 2 ) significantly different between infection-only and advanced cases ( FDR<0 . 05 ) , and ( 3 ) ranked high in two of three multivariate feature selection methods ( SAM delta>3 , PLSDA VIP score >1 . 5 or RF rank <200 ) were chosen ( Tables 2 and 3 ) . Several of the highly differential metabolites have no known matches in the Human Metabolome Database and are probable molecules produced by S . haematobium during urogenital schistosomiasis infection ( Table 3 ) . The suitability of the selected features as biomarkers of urogenital schistosomiasis associated-bladder pathology and urogenital schistosomiasis were then analysed with Receiver Operating Characteristics ( ROC ) curve analysis . The performance characteristics of these probable biomarkers using logistic regression algorithm are presented in Table 4 . Of interest in the current study were the S . haematobium catechol estrogen and related molecules identified previously in Santos et al . ( 2014 ) and Gouveia et al . ( 2015 ) , in which urine was examined in ESI negative mode only . In the current study , 11metabolites appeared to match those highlighted in the two aforementioned studies ( Table 5 ) .
In this study , LC-MS was used to examine urine and plasma metabolites in samples from people with different states of urogenital schistosomiasis , and those without infection ( controls ) . Important metabolites were putatively identified by peak annotations , from searching metabolome databases and mass spectral matching . In general , in both urine and plasma a large number of chromatographic peaks were detected and hence , a large number of metabolites . The large number of differential metabolites among study groups ( S1 and S2 Files ) which were statistically significant indicates that there were many peaks that could be explored for biomarker use . Nevertheless , a higher number of unique metabolites were found in urine compared to plasma , probably because the parasite inhabits the bladder environment . It is clear from the data that abnormal lipid regulation in the human host is important in both urogenital schistosomiasis infection and urogenital schistosomiasis associated-bladder pathologies as several of the most dysregulated metabolites from the metabolome were lipids ( Table 2 ) . Many proteins , including those involved in developing pathological conditions , interact with lipids of the cell membranes and these protein–lipid interactions are susceptible to modifications [25] . Thus , alterations in lipid levels observed in this study would definitely lead to changes in protein activity . A similar situation of altered lipid metabolism has been associated with the development of cardiovascular pathologies , such as hypertension , atherosclerosis , coronary heart disease and thrombosis , as well as tumours [26 , 27] . There is also evidence that the changes in phospholipids may occur before morphological changes in tumours [28] , and there are attempts to target them with drugs [25] . Given that no bladder carcinomas were recorded among participants , we suggest that specific lipid metabolites highlighted in this study may serve as early warning metabolites prior to the development of bladder cancer , during chronic urogenital schistosomiasis . They may be useful as early diagnostic markers or therapeutic targets . Among the most downregulated metabolites in urogenital schistosomiasis and urogenital schistosomiasis- associated bladder pathology cases were human steroid hormone precursors ( Table 2 ) and this observation is supported by the number of metabolic pathway hits ( S3 File ) . These steroid precursors are required to produce estrogens , estradiol and testosterone . Thus , the finding in this study is that host steroids are much reduced as a result of urogenital schistosomiasis . Given an earlier study [8] which found increased infertility occurrence along with urogenital schistosomiasis , one of the observations of the present study i . e . low level of host sex hormone precursor in infection , provides more insight into a possible mechanism by which urogenital schistosomiasis causes infertility . In the study , Santos et al . [8] found that 17 of 29 Angolan women infected with urogenital schistosomiasis had self-reported infertility , compared to 8 of 24 women infected with urogenital schistosomiasis who showed no signs of infertility . It had been suspected that hormonal imbalances may be a factor in this form of infertility [10] . To the best of our knowledge , this is the first study to provide evidence of significant reduction in the levels of human sex hormone precursors in urogenital schistosomiasis infection ( Table 3 ) and its associated bladder pathologies . Furthermore , several of the important metabolites which distinguished infection were those ostensibly produced by the parasite S . haematobium , and which could not be identified on currently curated databases such as HMDB , Metlin or LipidMaps . Eleven of these were matched to the catechol estrogens associated with infection in [8] and Gouveia et al . [9] ( Table 5 ) , although conditions of LC-MS analysis were different . However , the presence of trace amounts of such metabolites in some non-infected persons is an indication that better diagnostic tools are still needed for urogenital schistosomiasis and that its burden in Nigeria may be underestimated . Based on the current study ( Table 5 ) , four of these 11 estrogen-related parasite molecules detected , m/z 269 , 228 , 204 and 369 , were significant enough to be considered as biomarkers . In the report of Gouveia al . [9] , urine samples from participants with urogenital schistosomiasis-associated hyperplasia , metaplasia squamous or urothelial carcinoma were analysed , but the current study involved participants with healthy controls , ordinary infection or infection with the bladder pathologies which precede carcinoma , rather than confirmed carcinoma cases . Therefore , it is likely that the full repertoire of the schistosome catechol estrogens is abundant only in the matured/developed carcinoma cases . In addition , the differences in parts of the analytical conditions used in the study may result in differences in results . Another parasite molecule , m/z 335 , a putative catechol , proved to be a good biomarker of urogenital schistosomiasis associated bladder pathology ( Table 3 ) . Many of these S . haematobium molecules in the current study will require further validation research which would involve comparison with standards and targeted tandem MSn in order to determine their structure or whether they are modified forms; although in this study , the peak annotations and chemical identity have been revealed . Because high levels of potential parasite estrogen-related molecules and low levels of host estrogen-related molecules were observed in the present data , we suggest that estrogen metabolism could be a key influential reaction in host-parasite relationship during urogenital schistosomiasis and could be important as the infection progresses into tumour . Considering the evidence from the current study and that from previous studies , we suggest that S . haematobium adult worm infection reduces human sex hormone availability either by utilising host steroid hormones or blocking its utilisation or production; and therefore , the worm produces related molecules . This proposed hypothesis is strengthened by earlier research [29] which showed that flatworms such as Taenia spp or hookworms may depend extensively on host sex steroids for growth . Such dependence could be expected to lead to lower levels of the host steroids . Among the most upregulated metabolites in urogenital schistosomiasis and urogenital schistosomiasis associated bladder pathology cases were molecules belonging to catechols , cyclic aromatic hydrocarbons , benzenoids or quinones . These included many parasite and some host metabolites ( Table 4 ) . Abnormal levels of such metabolites are known to be involved in carcinogenesis . There was an abundance of two related glycerophospholipids in advanced cases and to a lesser extent , in infection-only cases . Also , glycerophospholipid metabolism had high number of metabolic pathway hits ( S3 File ) . High levels of phosphatidylcholine ( PC ) and phosphatidylethanolamine ( PE ) were found strongly associated in urogenital schistosomiasis associated-bladder pathology ( Advanced ) cases ( Table 4 ) . PE may be converted to PC by phosphatidylethanolamine-N-methyltransferase [30] . PC is formed from phosphocholines and catabolised back to phosphocholines . Increased levels of phosphocholines are associated with proliferation , and there are reciprocal interactions between oncogenic signalling and phosphocholine metabolism [31] . Increase in PCs , one of the major forms in the alteration of choline metabolism , involving specific phospholipases , transporters , kinases , was recently shown in induction of cancer [32] , including colorectal cancers and non-small-cell lung cancer [30] . Due to oxidative stress in a tumour microenvironment , PE becomes highly expressed on endothelial cells as they are redistributed from the inner to the outer membrane leaflet [33] . It is suggested that increased PC and PE in bladder endothelial cells is one of the mechanisms for cancer induction in chronic urogenital schistosomiasis . Gangliosides are glycosphingolipids containing sialic acid found mainly in the plasma membrane and having functions in cell recognition and signalling . From our data , N-Glycoloylganglioside GM2 ( GGM ) , a ganglioside , was abundant in advanced cases ( Table 2 ) . GM2 gangliosides are over- expressed and abundant in different forms of carcinoma , including melanoma , neuroblastoma and breast carcinoma; they promote T cell dysfunction and have been utilised in vaccine trials [34] . There are also reports that these gangliosides are involved in pathological processes , because they can be receptors for viruses , toxins , and autoantibodies , and that they can suppress availability of innate and adaptive immune molecules [35] . In this study , GGM were found in bladder pathologies with no tumours observed; it was shown in breast cancer research by Azordegan et al . [28] that changes in phospholipids may occur before morphological changes in tumours . This study is the first to report a ganglioside strongly associated with urogenital schistosomiasis associated-bladder pathology . In a similar vein , increased levels of benzenamines , putatively identified as Adrenochrome and 3-Succinoylpyridine , associated strongly with urogenital schistosomiasis infection alone . The abundance of these molecules is biologically relevant . Adrenochrome ( also Adrenochrome-O-quinone; AQ ) , is a toxic quinone metabolite of catecholamines , specifically epinephrine . It is formed as a result of oxidation activities , is neurotoxic and has psychotomimetic properties [36] . Like other such quinones , AQ is capable of forming reactive oxygen species with pathological consequences . Glutathione transferases ( GSTs ) may prevent pathologies by catalysing the formation of glutathione conjugates of o-quinones [36] . AQ abundance may be affected by other factors . GSTs attempt to scavenge free radical forming agents to prevent pathologies or cancer and some free radical species inhibit GST to prevent this . Polymorphisms in GST determine the efficiency of these processes and also determine the susceptibility to cancer . Hence , the efficiency may actually reduce the formation or availability of these agents , but further research will be needed to confirm this . However , abundance of AQ in infection-only cases would indicate that there is an increased amount of free reactive oxygen species and would therefore hasten bladder pathologies . It would also indicate inadequate activity , decreased activity or inhibition of GSTs . Such a decrease in GST activity due to S . haematobium infection was reported earlier [37] . 3-Succinoylpyridine ( 3SP ) a nicotine metabolite and by-product of N-nitrosamine formed by the action of cytochrome P450 . N-nitrosamines ( and especially N-nitrosodimethylamine , NDMA ) , nitrite and nitrate were detected in significant amounts in the urine of schistosomiasis patients in the studies [38] and the authors suggested they have roles in carcinogenesis . In the current study 3SP was found in abundance in infection-only cases; it is known to be a nicotine metabolite expected in the urine of tobacco smokers [39] . Participants in the current study did not indicate having a smoking history . 3SP is produced by hydroxylation of methylnitrosaminopyridylbutanone ( NNK ) , a nitrosamine which was not among those reported by Mostafa et al . [38] to be in the urine of urogenital schistosomiasis patients . Thus , it has been suggested that quinones could be formed and N-nitrosamines are present in urogenital schistosomiasis infection [38] . This study is the first report of the abundance of Adrenochrome-o-quinone and NNK by-product 3SP in urogenital schistosomiasis . Furthermore , a modified peptide , indolylacryloylglycine ( IAG ) , was abundant in urogenital schistosomiasis associated-bladder pathology . IAG is suspected to be a by-product of tryptophan metabolism and glycine conjugation , a process which may or may not involve gut bacteria action [40] . IAG levels in urine were reported to be increased in muscular pathologies , autism , skin tuberculosis , and probably reduced in tumour [40 , 41] However , its use as a marker may be of limited value because its concentration varies seasonally , possibly due to higher solar radiation , and depending on age [41] . A putative naphthalene based compound , 1-Nitro-5 , 6-dihydroxy-dihydronaphthalene ( DDN ) , was abundant in , and a putative biomarker of , infection-only cases . Napthalene is classified by the International Agency for Cancer Research as a 2B carcinogen because at proper doses , naphthalene metabolites show genotoxic and/or mutagenic activity [42] . Napthalene is normally expected as an environmental pollutant in air and sometimes water effluents , and chronic inhalation of naphthalene can induce respiratory tract tumours [43] . In a similar mechanism as other weak carcinogens such as estrogens and benzene , naphthalene is metabolically activated by cytochrome P450 ( CYP ) and metabolites formed react with DNA to form depurinating adducts [43] . In the KEGG database , several metabolites of naphthalene including dihydroxy-dihydronapthalenes are annotated in the metabolism of xenobiotics by CYP ( map00980 ) , but not DDN . Since the participants in this study live in a rural setting and relatively far from air pollutants such as naphthalene ( although air pollution from a nearby cement factory cannot be ruled out ) , we suggest that just as estrogen-like molecules were found to be produced by S . haematobium , benzenoids and related molecules may also be produced by the parasite and metabolised by human CYP , leading to their activation . A limitation of the current study is lack of complementary data such as a validation set of completely different samples and higher order MS2 or MS3 data . This would have further enhanced definitive identification status of all metabolites . In summary , unique putative metabolites with potential value as biomarkers were identified in this study; when these metabolites are completely validated , the molecules and associated proteins could be further characterised and studied for future use as therapeutic or diagnostic targets , or in vaccine development . | Obtaining specific molecules having a strong association with a disease condition i . e . biomarkers , is usually a major step to developing new modes of therapy or diagnosis for the disease . In this study , samples from individuals with S . haematobium infection , some of whom had developed bladder pathologies , were compared to controls . From the blood and urine samples analysed by mass spectrometry , we highlight important human and Schistosoma haematobium metabolites , small molecules smaller in size than most proteins , which associate strongly with individuals having schistosomiasis induced-pathology or schistosomiasis alone . This report of the potential biomarkers will also add to the current understanding of the molecular events leading to schistosome associated-bladder cancer . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"schistosoma",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"helminths",
"tropical",
"diseases",
"bladder",
"biomarkers",
"parasitic",
"diseases",
"animals",
"urine",
"metabolites",
"neglected",
"tropical",
"diseases",
"urogenital",
"schisto... | 2018 | Metabolite profiling for biomarkers in Schistosoma haematobium infection and associated bladder pathologies |
Cones with peak sensitivity to light at long ( L ) , medium ( M ) and short ( S ) wavelengths are unequal in number on the human retina: S cones are rare ( <10% ) while increasing in fraction from center to periphery , and the L/M cone proportions are highly variable between individuals . What optical properties of the eye , and statistical properties of natural scenes , might drive this organization ? We found that the spatial-chromatic structure of natural scenes was largely symmetric between the L , M and S sensitivity bands . Given this symmetry , short wavelength attenuation by ocular media gave L/M cones a modest signal-to-noise advantage , which was amplified , especially in the denser central retina , by long-wavelength accommodation of the lens . Meanwhile , total information represented by the cone mosaic remained relatively insensitive to L/M proportions . Thus , the observed cone array design along with a long-wavelength accommodated lens provides a selective advantage: it is maximally informative .
Human perception of color comes from comparing the signals from cones with different peak sensitivities at long ( L ) , medium ( M ) , and short ( S ) wavelengths . While three cone types are required to support trichromatic color vision , the three types distribute unequally . The central human fovea contains only ∼1 . 5% S cones , with the fraction increasing to ∼7% at greater retinal eccentricities [1] . In most dichromatic mammals , S cones are similarly rare ( e . g . , [2]–[4] , but see [5] for exceptions ) . Meanwhile , the mean ratio of L cones to M cones varies widely between primate species ( majority L in humans , majority M in baboons [6]–[11] ) . Amongst human individuals , the L∶M ratio varies between 1∶4 and 15∶1 without loss of normal color vision [12] and similar variation is seen in New World monkeys [6] . We asked: why are S cones rare , why can the L/M ratio be so variable , and why does the S cone fraction increase with retinal eccentricity ? A possible explanation for the rarity of S cones has been proposed: the lens is accommodated to focus long-wavelength ( red ) light; thus short wavelengths are blurred and the Nyquist limit on sampling predicts fewer S cones [13] . While plausible , this explanation seems incomplete for three reasons: ( i ) In human , the blur radius for blue light is ∼1 . 5 times the blur radius for red light ( [14] , [15] , [16] , see Results ) , giving , via a Nyquist sampling argument , an ( L+M ) /S ratio of ∼ ( 1 . 5 ) 2∼2; this implies ∼33% blue cones which is 5–10 times too high , ( ii ) The accommodation wavelength of the eye ( the wavelength at which light is most sharply focused ) is under behavioral control [17] , and aberration could be minimized for blue light , reversing the sampling argument , if this improved vision , ( iii ) The sampling argument ignores noise and correlations and thus could be entirely wrong if natural scenes filtered through the ocular media had low power or greater spatial correlations at long wavelengths . To model noise and correlations , we must consider additional key factors – optical properties of the ocular media , and correlated chromatic structure in natural scenes . Thus , we asked if these two factors , in combination with chromatic aberration in the lens , might suffice to explain long wavelength accommodation and the structure of the cone array . Our analysis treated as fixed three characteristics of the eye: ( i ) number of cone types; ( ii ) cone spectral sensitivities; and ( iii ) transmittance of the ocular media . Previous work has suggested that these characteristics are evolutionary adaptations to the structure of the environment . Specifically , researchers have considered the relation between the number of cone types in the retina , the spectral sensitivities of these cones , how the cone signals are processed by ganglion cells , and the statistical structure of naturally occurring spectra [18]–[20] . Researchers have suggested that the peak sensitivities of primate cone photopigments are optimally placed for encoding visual information in natural environments [21]–[24] or to facilitate crucial behavior under the constraints of chromatic aberration , diffraction , and input noise , particularly in dim light [25] . Finally , it is believed that the ocular media , especially the macular pigment , transmit less short wavelength light in order to protect the retina against damage from UV light [26]–[28] . To model the correlations and intensity distributions in the natural world , we based our analyses on a database of high-resolution chromatic images accumulated in a riverine savanna habitat in the Okavango Delta , Botswana ( Fig . 1a ) . These images showed similar power when integrated through the spectral sensitivities of the three human cone classes . But including the selective absorption of short wavelengths by the ocular media breaks this symmetry , leading to fewer photoisomerations per second , and hence lower signal-to-noise ratio , in S cones . Measuring spatial correlations within , and between , model cone classes responding to natural images , we found surprising similarity – long-distance correlations , approximately scale invariant , prevailed between pairs of cones of any type . Given these characteristics of cone responses to natural images , we asked , what combination of lens accommodation wavelength and cone mosaic would best support vision . Since visually guided behavior is limited by the amount of information available from the retinal cone array [29]–[31] , we formulated a precise question by looking for the mosaic and lens accommodation wavelength that jointly maximized information about natural images . First , we computed the signal-to-noise ratio , or equivalently , the information rate , in single cone responses . Second , we summarized the effects of correlations in natural images by measuring how information in a cone array scaled with array size after discounting for redundancies between cones . From these data we found that in the absence of chromatic aberration , information about natural scenes was maximized by an array with ∼40% S cones and about equal numbers of L and M cones . We then included the point-spread function of the human eye accommodated to different wavelengths [32] . The resulting blurring of light leads to redundancies in the responses of neighboring cones . Including these redundancies , information was maximized when the lens was accommodated to long wavelengths while the mosaic contained just a few percent of S cones . Meanwhile , after including optical blur , the amount of information was largely independent of the L/M ratio , allowing substantial differences in this ratio between near-optimal mosaics . In addition , plausible variations in our parameters gave a slight advantage to having a majority of either L or M cones ( as seen on average in , e . g . , human vs . baboon ) . We also modeled the topographic variation of the retina , and found that as the cone density decreased towards the periphery , the advantage of L cones decreased too . Thus , information was maximized by an S cone fraction that increased towards the periphery of the mosaic . All of these features – low S cone fraction , large variation in L/M ratios , and peripheral increase of S cones – are seen in the primate retina . The match between our analysis and the observed structure of the cone mosaic adds to a growing body of evidence , which started with pioneering work [29] , [30] , [31] , investigating how the constraints and properties of the biological hardware interact with the statistical properties of the natural environment to shape organization of the brain ( e . g . , [21] , [33]–[45] ) .
Our analyses are based on a new high-resolution ( camera resolution: 2014×3040; 46 pixel separation = one degree of visual angle ) , database of color images from which we selected 176 daylight scenes of a riverine habitat in dry-season Botswana . Although our image resolution was less than that of the primate fovea , we used the scale invariance of natural images to treat the pixel spacing as being equivalent to the foveal cone spacing for a more distant observer [46] , [47] . We required that the camera response be in the linear range and that fewer than 0 . 5% of the pixels be saturated . Most images in our database that were acquired under daylight conditions had these properties . While our images are qualitatively different from those taken in other environments ( e . g . , urban scenes , the van Hateren database [48] , the McGill Calibrated Colour Image Database [49] ) , we have tested ( but do not show here ) that these image databases share the main statistical properties ( distributions of light intensity and spatial correlations ) that drive our analysis . From the red , green , and blue camera response at all pixels in each image we estimated the equivalent L , M , and S cone photoreceptor response . First we calculated the best choice amongst linear maps between camera and cone spectral sensitivities [50] and checked the accuracy using patches from the Macbeth Color Checker imaged with both our camera and a spectral-radiometer . Photoisomerization rates ( R* s−1 ) were estimated using the procedures described in [51] for guinea pig , but substituting appropriate parameters for human foveal cones ( human peak photopigment sensitivities and ocular media transmittance ) . See Materials and Methods for details of camera calibration and image processing . To determine the characteristics of a cone array that maximizes information about natural scenes , we considered in turn: ( i ) the information represented by a single , independent cone signal; ( ii ) an array of cones of the same type; and ( iii ) a mixed array . The single cone analysis incorporated the attenuation of short wavelength light by the ocular media , while the cone array analyses incorporated spatial correlations in natural images . We then found the LS and LM arrays that maximized transmitted information . The optimization analysis was carried out with and without including accommodation of the lens to understand how chromatic aberration interacts with other factors . Having estimated information in single cones and in arrays of one type of cone , we asked how a mosaic with two cone types should be organized to maximize information . To separate out the effects of accommodation wavelength of the lens we first studied the optimal mosaic without chromatic aberration . The information in a mixed mosaic of two cone types is the sum of the information in each array minus the redundant mutual information between them . For example , the information transmitted by a mixed array of X and Y cones ( where X and Y can be L , M or S ) is given by ( 6 ) Here SX , Y are the sets of X and Y array responses while S = {SX , SY} represents the set of joint responses; IX and IY represent the mutual information between the responses of the X and Y arrays and the image input; and IXY is the mutual information between responses of the X and Y subarrays:Here p ( sx , sy ) is the joint response distribution of X and Y cone arrays , and p ( sx ) and p ( sy ) are marginal response distributions of each cone type . In deriving ( 6 ) we assumed that noise in X and Y cones is uncorrelated , so that the conditional response probability factorizes: p ( sX , sY|E ) = p ( sX|E ) p ( sY|E ) . Neglecting chromatic aberration , the information scaling for each cone type ( see ( 4 ) and Fig . 2 ) then allows us to write ( 7 ) Here , I1X is the information transmitted by a single X cone; I1Y is the information transmitted by a single Y cone; and δX and δY are scaling exponents . The number of cones is N = NX+NY , while the average distance in pixels between neighboring cones is dX = √ ( N/NX ) for X cones , and dY = √ ( N/NY ) for Y cones . The exponents δX , Y are functions of dX , Y ( see Fig . 2 ) . We measured the mutual information IXY in a 6-pixel array by varying the proportion of each kind of cone and their geometric arrangement . The cone signals were discretized to reflect the number of signaling levels in each cone class , commensurate with their different estimated information transmission rates . These discrete signals were then used to directly compute mutual information using ( 5 ) . In this way , the mutual information between the L and M and between L and S cones in mixed arrays was estimated directly from the image data . To do this we used ( 5 ) to compute the total information ( I ) in mixed arrays , as well as the information in the subarrays of each type . From ( 6 ) , this gave IXY ( SX , SY ) = IX ( SX , E ) +IY ( SY , E ) −I ( S , E ) . We averaged over all geometric arrangements of arrays with the same cone proportion , to smooth out effects of pixelation . As expected , the mutual information between cone types vanishes for arrays with only one cone type and peaks in between , giving a domed shape ( i . e . , mutual information between two cone classes was highest when the array had an approximately even mix of the two classes of cones; Fig . 3 ) . The above analysis estimates how the mutual information between X and Y cones changes with the proportion of X and Y cones in a small array . Now we need to know how the mutual information in a large array changes with the proportion of cones of each type . This is hard to measure directly but will have the same qualitative domed form as for small arrays . Thus to extrapolate to large arrays we made the simplifying assumption that the ratiois a function only of the relative fraction of X cones , x = NX/N ( or Y cones , ( 1-x ) = NY/N ) . Using this form of the mutual information between X and Y cones , we can rewrite ( 7 ) asFor arrays in which the scaling exponents of the two subarrays are similar ( δX≈δY = δ ) , our simplification is equivalent to assuming that total information in the array also scales as Nδ . The similar correlations between L , M , S cones and slow variation of δ with spacing in Fig . 2 , thus imply that our simplification should be valid for mixed arrays with roughly similar numbers of cones of each type , and for sparse arrays , since δX≈δY = δ in these cases . We checked that our final results were self-consistent within this domain of validity . We also checked that our final results depended largely on the qualitative shape of the X-Y mutual information curve ( which we infer from small arrays ) , rather than the precise values . We confirmed this by repeating our analyses with various assumed domed shapes for the mutual information . To find the optimal mixed cone mosaics we first measured the ratio ρ directly from the result in 6-pixel LM and LS arrays ( Fig . 3 ) and used our estimates of single cone SNR to writeWe then obtained the optimal mosaic by maximizing I ( S , E ) with respect to the proportion of L cones . We obtained an optimal LM mosaic with 52% L cones and 48% M cones . Using the same technique , but substituting S cones for M cones , we found an optimal LS mosaic with 61% L cones and 39% S cones ( Fig . 4 ) . In both cases , the small excess of L cones was driven by two factors: ( a ) the similarity in power and correlations between L , M and S sensitivity bands in natural scenes , and ( b ) the selective attenuation of short wavelength light by the ocular media , which breaks the symmetry between L , M , and S . Because L and M bands are so similar , the optimum contained about equally many of each type of cone . Meanwhile , the higher SNR in L responses resulted in 20% more L cones in the optimal mosaic . This mosaic , which is well adapted to the symmetric statistics of natural images and to the attenuation of the blue light in the ocular media , differed in two respects from the observed characteristics of the human eye: ( a ) the S cone fraction is an order of magnitude too high , and ( b ) there is no indication that the L/M ratio can be any more variable that the L/S ratio without detriment . But this analysis omitted one further key factor – chromatic aberration in the lens . The lens of the eye blurs light of different wavelengths to different degrees . Such chromatic aberration can affect the cone proportions in the optimal mosaic because the amount of blurring differs for each cone channel [61] . The presence of blur modifies the problem of calculating information in an array: within a blurred region the information conveyed by pixels is highly redundant . For each color channel the extent of the optical blur was estimated from measurements of optical aberrations in human observers ( [32]; data and code provided by H . Hofer; see Materials and Methods ) . While the eye can accommodate to various wavelengths , in white light and under normal viewing conditions the eye tends to focus for longer wavelengths , near the peak sensitivities of L and M cones ( e . g . [15] , [16] ) . For this reason , we used the mean chromatic point spread function ( PSF ) averaged across 13 subjects when accommodation focuses light best on the L and M cones . PSFs are highly kurtotic , and so values far away from the central peak are important for characterizing the PSF width . As an estimate of the region over which the blur is large , we chose one half of the radius that enclosed 90% of the PSF as a measure of this width . This choice corresponds roughly to twice the full-width at half-height of the PSF . For each color channel , this estimate of the spatial extent of the significant chromatic aberrations ( i . e . , blur ) gavewith the conversion to pixels obtained assuming a foveal cone-to cone spacing of ∼2 . 2 cones/arcmin [62] . To estimate the information in a blurred chromatic mosaic , we made the approximation that chromatic aberrations render L , M , and S cones separated by distances less than dL , dM , and dS respectively , completely redundant . We also made the approximation that , for each cone class , the blur has no effect beyond this distance . First we consider cones of one class separated by distances less than dX . In our approximation such cones are transmitting completely correlated signals but have independent noise . Averaging n redundant signals that are each corrupted by an independent noise source ( each with the same average magnitude ) will increase the signal-to-noise ratio by a factor of n . Thus , in our analysis , averaging across a block of n redundant cone signals increases the signal-to-noise ratio relative to a single cone so that the block of cones representsbits of information . The information represented by a block of ( dX ) 2 pixels is thereforeThis block information can be thought of as the information represented by a single ‘effective pixel’ that includes all the redundant pixels in a small region . In mixed arrays of cones of different types , a given block may only contain a fraction x of cones of a particular type . Then , gives the information represented by pixels of that type within the block . To compute the total information in an array of cones , we group pixels into blocks of area and treat the blocks as mutually correlated in a scale invariant way . Specifically , each of the blocks defined by the blur space constant is treated as a single effective pixel transmitting Iblock bits . Then , taking these blocks as being spatially correlated as in Fig . 1d ( because of the scale invariance of natural scenes ) , the same treatment as for arrays of single cones can be applied to arrays of blocks of cones . Thus , following ( 4 ) for single cones , the total information transmitted by cones of one type in the blurred array is given by a power-law in the number of blocks:where N is the total number of pixels in the array , and so N/dX2 is the number of blocks in the array and δX is the scaling exponent from Fig . 2 ( bottom ) for a pixel separation equal to the spacing of the blocks . For L , M blocks spaced at 4 . 8 pixels , this gave δL , M = 0 . 89 , while for S blocks spaced at 7 . 3 pixels , δS = 0 . 91 . The total information in a mixed array is then approximated as the sum of the information in the sub-arrays of each type minus the mutual information between the sub-arrays . This mutual information was taken to be the same fraction of the total information as measured before blurring . This approximation reproduced the general domed shape of the mutual information as a function of the fraction of cones of each type . Thus the information in an array with two kinds of cones and blurred optics was estimated as ( 8 ) where x = NX/N and 1-x = NY/N = ( N-NX ) /N are the fractions of each kind of cone . Including blur in this way increases the redundancy in each channel , predominantly among nearby pixels . We asked what cone fractions maximized information when the optics are accommodated to focus light best on L and M cones . Using our estimated values for the blur ( dL , M , S ) , the scaling exponents ( δL , M , S ) , the SNRs , and the redundancy ( ρLM , ρLS ) we plotted the total information conveyed by LM and LS arrays ( Fig . 5 ) . The increased redundancy in the L and M channels produces a broad range of equally effective LM mosaics ( Fig . 5 , top ) . That the L/M cone ratio has little effect on the information transmitted by a cone mosaic is consistent with the large variability in L/M ratios in primates with normal color vision [6] , [12 , ] and with the observation that human performance on some psychophysical tasks is invariant with respect to cone ratio [63] , [64] . Meanwhile , the blur reduces the information transmitted by the S channel more than by the L and M channels since the blur in the S channel extends further . Thus its inclusion reduces the number of S cones in the optimal mosaic as compared to Fig . 4 . That most information is transmitted by an array with few S cones ( ∼6 . 5% - Fig . 5 , bottom ) is consistent with the rarity of S cones in most mammalian cone mosaics ( e . g . [53] ) . The advantage of L cone domination is small but significant – using our parameters , a mosaic with 90% L cones conveys 10% more information per cone than a mosaic with 90% S cones . This result confirms the basic intuition of Yellott et al . [13] , that chromatic aberration plays a key role in the organization of the cone mosaic , but includes additionally the effects of spatial correlations and noise . One limitation of our analysis was that we analyzed a mixed mosaic of just L and S cones; if we were to consider all three cone types simultaneously , the fraction of S cones in the optimal mosaic would likely decrease . This is because an optimally organized LM mosaic transmits slightly more information per cone than a mosaic with only L cones . Thus an optimal trichromatic mosaic would have still fewer S cones than the optimal LS mosaic . Likewise , because the fraction of S cones in the optimal array is small , it is unlikely to significantly affect the L/M ratio in the trichromatic array . Our results for the optimal mosaic are due to the interaction of four factors: ( i ) correlations within a cone class ( summarized by the scaling exponents δL , M , S ) ; ( ii ) correlations between cone classes ( summarized by the redundancy factor ρLS , LM ) ; ( iii ) optical blur ( summarized by the blur widths dL , M , S ) ; ( iv ) power in different chromatic bands and attenuation by the ocular media ( summarized by single cone SNRs ) . All of these factors were modeled and estimated , rather than directly measured , and might vary between individuals and species . Thus , to test the relative importance of each of these factors in determining the optimum we systematically varied each one while keeping the others fixed and determined the consequences for the optimal array . The analysis above was carried out with a lens that focused long wavelengths best , as appropriate for the normal accommodative state of the eye [15] , [16] . However , the accommodation wavelength at which light is most focused by the lens is under behavioral control . Since the single cone SNRs , provided they were large , were a sub-leading determinant of the optimal cone fractions , we wondered whether there is any advantage to long-wavelength accommodation . Thus we explored how different accommodation wavelengths affect the distribution of cones in the optimal LS mosaic . Using the polychromatic PSFs computed for various accommodation wavelengths ( see Materials and Methods ) , we repeated the optimization procedure ( described above ) for mixed LS arrays . Since long and short wavelengths cannot be focused simultaneously , we expected to find optimal arrays that are dominated by either L or S cones , depending which channel is best focused . Our measure of the spatial extent of the blur in each color channel was again half of the 90% width of the PSF . The width of the L cone PSF is plotted against the width of the S cone PSF in Fig . 7a . For each accommodation wavelength , we then estimated the information per cone as a function of the L cone fraction following Eq . 7 ( Fig . 7b ) . When the lens accommodated to the L cone peak sensitivity , the mosaic maximizing information per cone had mostly L cones , while a lens accommodated to S cone peak sensitivity led to an optimal mosaic with mostly S cones . The information per cone in the optimal mosaic for each accommodation wavelength ( parameterized as S cone PSF width ) is plotted in ( Fig . 8 ) . Information transmission rates were highest when L cone light was focused sharply and S cone light was blurred . Although the per cone advantage of focusing long wavelength light is small ( ∼3% ) , multiplying by the number of cones in a retina gives a significant increase in the total amount of transmitted information . Interestingly , the worst choice is to accommodate between the L and S peak sensitivities . Focusing short wavelength light could be advantageous if , due to some other constraint , it was impossible to focus long wavelength light sufficiently . In this case , the optimal retina has mostly S cones , which may be related to the existence of a few species with S cone dominated retinas [5] . One might wonder whether the apparently small excess of information ( ∼3% ) in the optimal accommodation wavelength in Fig . 8 actually confers a significant selective advantage . It is worth noting that small selective advantages have a multiplicative effect over generations , and , just like compound interest , can pay large dividends over evolutionary time . Our analysis thus far has considered the overall proportions of cones of different classes . An additional observation is that the fraction of S cones in the human retina increases somewhat with eccentricity [1] , [65] , [66] . We wondered whether this observation might also be accounted for by our theory . Two relevant factors that are known to decrease with eccentricity are the overall cone density of the mosaic [1] and the optical density of short-wavelength filtering macular pigment [50] . The effect of reducing macular pigment density will be to reduce the SNR advantage of the L and M cones over the S cones , and to the extent this has an effect this would tend to increase the relative proportion of S cones . Our analysis of robustness presented above , however , indicates that this effect will be small but in the right direction , and preliminary calculations ( not presented here ) indicated that alone it would be insufficient to account for the increase in ∼1 . 5% to ∼7% S cone percentage from the central fovea to the periphery . We thus focused on the effect of the decrease in overall cone density . As the distance between cones becomes large relative to the blur , the number of cones in each blurred and redundant block decreases , reducing the significance of the blur in the optimization . Since chromatic aberration has a greater effect on the S-channel , increased sparseness of the array tends to increase the fraction of S cones in the optimal mosaic . To see this , we kept the extent of the blur fixed , and used the scale invariance of natural images to treat the image pixels as having the separation of cones at larger eccentricities that are viewing the same scene from a greater distance . Repeating the analysis for the optimal mosaic , we found that the predicted S cone fraction increases with decreasing cone density ( Fig . 9 ) . Overall the predicted optimal cone fractions are somewhat higher than seen in Curcio et al . [1] , but these measurements were accumulated from only two retinas and variations should be expected between individuals . Moreover , our estimates of the exact optical parameters to use for the periphery are not currently precise enough to support inferences about the significance of predicted differences of a few percent . The key point we wish to emphasize at this juncture is thus that our theory is qualitatively consistent with an increase of S cone proportion with eccentricity . The results presented here depend on many estimated and modeled quantities and thus it was important to test how plausible variations in these quantities might affect the optimal mosaic . First , we checked that our results were insensitive to the details of the model of the eye's optics , and confirmed that essentially the same results were obtained when we used the Marimont and Wandell [61] model of the eye's chromatic aberrations . We also approximated the cone signal as a Gaussian channel so that an explicit functional form for information could be manipulated . As a check , we directly estimated the information in the cone signal from the histogram of isomerizations rates binned according to the Poisson noise at each rate . This procedure gave a similar result to the Gaussian approximation – our results follow from the similarity of SNR in the L and M cone channels , and the smaller SNR in the S cone channel . The relative sizes of SNR in each channel are a consequence of the cone spectral sensitivities ( which peak at similar wavelength for L and M cones , but at significantly shorter wavelength for S cones ) and the transmittance properties of the ocular media , which selectively attenuate short wavelengths . To extrapolate information to large arrays we used the power law that gave a good fit for small arrays . We checked that the results were insensitive to the overall size of the array by varying the number of model cones ( N in ( 4 ) ) between 1000 and 1 , 000 , 000 . In these extrapolations we treated noise in photoreceptors as being dominated by photon noise and therefore independent . It should be kept in mind that noise correlations can significantly affect the total information transmitted by a population of cells [67] . However , substantial noise correlations are not expected in cone isomerization rates , since in daylight these fluctuations are primarily controlled by the stochastic arrival of photons . In treating chromatic aberration we modeled optical blur as making all cones within the scale of the blur completely redundant . In fact , the redundancy decreases with separation even within blurred regions . However , since the significant factor is the relative range of the L , M and S channel blurs , we do not expect more detailed modeling to affect our conclusions . Like the lens of the eye , our camera lens also exhibits chromatic aberration , blurring short wavelengths more than longer ones . The effect is small , and only apparent at the highest spatial frequencies . We estimated the effect of camera blur on our results in two ways . First , we deblurred the S cone channel ( adding power at high spatial frequencies ) to compensate for the reduction in power at high spatial frequency due to the camera lens . Our analysis was robust to this correction . Second , we blurred the L and M cone channels ( reducing power at high spatial frequencies ) to match the blur in the S cone channel . Again , the results and conclusions were unchanged .
Many mammals exhibit a significant excess of L cones over S cones ( Fig . 10 ) . Meanwhile , humans exhibit , on average , only a small excess of L cones over M cones [1] but the relative proportion of L and M cones varies significantly between individuals [12] . There is also topographic variation in cone proportions within a mosaic – e . g . , in human retina the proportion of S cones in the central retina exceeds the proportion of S cones in the peripheral retina [1] . All these basic facts about the design of the photoreceptor mosaic seem to be explained by a single hypothesis – given the filtering properties of the eye and the optics of chromatic aberration , the overall mosaic arrangement combines with lens accommodation to maximize information transmitted from natural scenes . Any optimization argument of this kind must hold fixed some characteristics of the system while varying others . We held fixed the number of cone types , their spectral sensitivities and the absorption of the ocular media and tested how varying cone proportions and lens accommodation wavelength changed the amount of information conveyed by the array . We treated these factors as variable because we were seeking underlying rationale for the observed cone proportions and because accommodation wavelength is under behavioral control . We could have instead varied the absorption of the ocular media or the number of cone types while holding the other factors fixed . This type of analysis will be interesting for deriving the predictions of the theory for other species . As in the present work , we expect that in most vertebrate eyes a larger fraction of the light in natural scenes to which S cones are most sensitive never reaches the photoreceptor layer , giving an small advantage to long wavelength cones . This is because ocular media ( cornea , aqueous humor , lens and vitreous ) filter out more short wavelength light than long wavelength light [54] , [68] . Humans , lower primates , and diurnal sciurids ( squirrels ) , have an additional macular pigment that filters out even more short wavelength light to protect the retina from UV radiation [26]–[28] , [69]–[71] . Of course , if natural scenes had much more power at short wavelengths , S cones would still have an advantage despite attenuation in the optical media . Thus , the disadvantage for S cones is the combined result of similar power at short and long wavelengths , and selective attenuation . The present analysis offers a unified explanation for why S cones are rare , why they increase toward the periphery , and why a large variation in L/M ratio can be tolerated . A useful way to consider these results is to imagine how one might “build” a retina , cone-by cone , with the goal of transmitting as much information as possible . First consider the case where only L and S cones are available . Since the signals S cones receive are smaller , L cones are individually more valuable . Consequently , a builder would begin by using only L cones . However , as the array of L cones becomes large , each additional L cone adds progressively less value because its signals become increasingly redundant with its neighbors . Eventually the value of an additional L cone decreases sufficiently so that adding an S cone becomes advantageous , despite its smaller relative information capacity . The end result is an optimal array with mostly L cones , and a few S cones . When optical blur is included in the analysis , the redundancy in S cone signals is increased relative to the redundancy in L cone signals , making S cones even less valuable . Thus L cones dominate the array . For L and M cones the situation is different . Since these two cone types carry similar amounts of information , adding an M cone instead of an L cone becomes advantageous much sooner , and the optimal array is more evenly mixed . Furthermore , optical blur affects L and M cone signals similarly because their spectral sensitivities are similar . The blur renders L and M cone signals - already quite redundant - even more redundant . The result is that , within the spatial extent of the blur , L and M cones have roughly equivalent value . Consequently , the information transmitted by the array changes little over a wide range of L/M proportions , although plausible variations in the parameters can give a small advantage to L or M cones . The latter might explain why , despite large variations between individuals , the human eye has , on average , more L cones , while the baboon eye has more M cones [9] . Our findings fit with a growing body of evidence that the retina allocates limited resources to maximize the information transmitted from natural scenes , subject to biophysical constraints . For example , this principle seems to explain the excess of OFF ganglion cells in the retina [72] , the overlap of ganglion cell receptive fields [58] , cone density distribution [73] , the distribution of information traffic in the optic nerve [74] , and the distribution of its axon calibers [75] ( see the review [76] ) . Snyder , Stavenga , & Laughlin [33] pioneered this approach to understanding the design of photoreceptor arrays as maximizing information under various constraints . Their effort considered trade-offs between spatial acuity and contrast sensitivity given white noise stimuli at different intensities , but did not consider the chromatic organization of natural scenes . In a related approach to analyzing the photoreceptor array , Bayesian decision theory was used to investigate tradeoffs between monochromatic and dichromatic vision [77] . The present findings also extend a large body of work on the evolution of wavelength sensitivity . There are many examples where the peak sensitivity of a photopigment matches the most prevalent wavelength in the environment ( e . g . , cones of fish in Lake Baikal , [78] ) . There are also examples where the behavioral niches of organisms seem to influence their photoreceptor sensitivities ( e . g . , UV receptors in insects and birds for seeing flower patterns , and the UV receptor of falcons which detects vole urine trails that fluoresce in the ultraviolet [79] . A number of authors have , for various species , considered the optimal choice of cone opsin spectral sensitivity [21]–[24] , [80] , [81] . For primates , it has been suggested that trichromacy evolved to assist detection of ripe fruit on a green background [21] . It has been further argued that the spectral sensitivities of the three cone types in human might maximize information transmission from natural scenes under the constraints of chromatic aberration , diffraction , and input noise particularly in dim light [25] . These arguments suggest that the molecular properties of the cone opsin are shaped to maximize the information they transmit about behaviorally relevant stimuli , and here we find the same for the structure of the photoreceptor array .
Our main results can be derived using an alternative estimate of the single cone information that does not make use of the Gaussian channel approximation , and instead directly applies Shannon's formula to the individual cone isomerization distributions . The cone signal distributions are discretized into bins with boundaries placed 2 noise standard deviations from each bin's center . This standard deviation was determined by assuming Poisson photon noise , for signals with mean intensity equal to the intensity at the center of each bin . Fig . 11 is a plot of the main result from our paper , using this method for calculating the single cone entropies . The results are qualitatively and quantitatively very similar . Note that the precise values of the cone SNRs were shown in Results to have little influence on the on the optimal cone proportions . Our natural image database consists of images taken in a variety of environments and lighting conditions . Most are daylight images from dry-season Botswana , but we also have images of Botswana at other times of year , Philadelphia , and locations in Southern India . Within this set we have collected low light intensity ( dusk / dawn ) images , close-ups , and images of the horizon with split sky/ground . The diversity of the images allows us to investigate how different color environments , behavioral needs , and activity periods ( nocturnal vs . diurnal vs . crepuscular ) affect the demands on spatial-chromatic information processing . Images were acquired with a Nikon D70 digital camera writing to “RAW” format . This format gives approximately 9 . 5 bits per pixel for each color channel ( see http://www . majid . info/mylos/weblog/2004/05/02-1 . html ) . Images were collected on a 2014×3040 photocell array with interleaved red , green , and blue sensors , then interpolated ( using nearest-neighbor interpolation within each sensor class ) to estimate the full red , green , and blue camera response at each pixel location . Following this interpolation , each image was downsampled by a factor of 2 to minimize aliasing artifacts from the interleaved red , green , and blue sensor sampling of the camera . At the down-sampled resolution ( 1007×1520 ) , the camera resolution was 46 pixels per degree of visual angle . In our analyses we used scale invariance of natural scenes to regard the pixels as having the separation of foveal cones viewing the same scenes from a greater distance . Additional detail follows . Point spread functions ( PSFs ) were estimated from wavefront aberration measurements obtained for 13 subjects , reported by Chen , et al . [14] . For each subject , the wavefront aberration measurements characterize individual observer optical aberrations and allow calculation of the PSF for various choices stimulus wavelength , pupil diameter , and accommodative state . All of our calculations were performed for a 3mm pupil . The code to convert wavefront aberration measurements to polychromatic PSF as a function of accommodation was kindly provided to us by Heidi Hofer , along with the tabulated wavefront aberration measurements required for the calculations . We were interested in how the PSF seen by the L cones traded off the PSF seen by the S cones as accommodation varied . For each observer , we computed the PSF for each stimulus wavelength ( 372 nm to 700 nm at 4 nm steps ) for a range of accommodative states ( nominal wavelengths of accommodation 372 nm to 700 nm at 4 nm steps ) . For each observer and accommodative state , we then obtained the PSF seen by each cone class by weighting the individual wavelength accommodative states by the spectral sensitivity of that cone class . This provided individual observer data on how the PSFs seen by the two cone classes varied with accommodation . To combine data for each observer , we found for each observer the accommodative state that minimized various weighted sums of the widths of the PSFs seen by the L and S cones . PSF width for this optimization was obtained from the circular average of each computed PSF . For each choice of L and S cone weights in the minimization , the optimized PSFs for each cone class were circularly averaged within each observer and then averaged across observers . The result of this analysis is shown in Fig . 6 , where the computed points trace out the desired tradeoff , with widths obtained from the final average across observer PSFs for each choice of S and L cone weights . The endpoints of the curve shown provide the minimum width PSF obtainable for each of two cone classes , when no weight was attached to the width of the PSF for the other class . It was important to separately optimize the accommodative state for each observer for each weight choice , because for any particular weight choice different observers' PSFs were optimized by different nominal wavelengths of accommodation . The smooth fit to the PSF tradeoff shown in Fig . 6 was used to resample the tradeoff frontier for the calculations shown in Fig . 7 . We also used the general procedure described above to find the PSFs seen by the L , M and S cones when the accommodative state was set to minimize the average width of the L and M cone PSFs . | Human color perception arises by comparing the signals from cones with peak sensitivities , at long ( L ) , medium ( M ) and short ( S ) wavelengths . In dichromats , a characteristic distribution of S and M cones supports blue-yellow color vision: a few S and mostly M . When L cones are added , allowing red-green color vision , the S proportion remains low , increasing slowly with increasing retinal eccentricity , but the L/M proportion can vary 5-fold without affecting red-green color perception . We offer a unified explanation of these striking facts . First , we find that the spatial-chromatic statistics of natural scenes are largely symmetric between the L , M and S sensitivity bands . Thus , attenuation of blue light in the optical media , and chromatic aberration after long-wavelength accommodation of the lens , can give L/M cones an advantage . Quantitatively , information transmission by the cone array is maximized when the S proportion is low but increasing slowly with retinal eccentricity , accompanied by a lens accommodated to red light . After including blur by the lens , the optimum depends weakly on the red/green ratio , allowing large variations without loss of function . This explains the basic layout of the cone mosaic: for the resources invested , the organization maximizes information . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"&",
"Methods"
] | [
"neuroscience/sensory",
"systems"
] | 2010 | Design of a Trichromatic Cone Array |
The bacterial pathogen Pseudomonas syringae modulates plant hormone signaling to promote infection and disease development . P . syringae uses several strategies to manipulate auxin physiology in Arabidopsis thaliana to promote pathogenesis , including its synthesis of indole-3-acetic acid ( IAA ) , the predominant form of auxin in plants , and production of virulence factors that alter auxin responses in the host; however , the role of pathogen-derived auxin in P . syringae pathogenesis is not well understood . Here we demonstrate that P . syringae strain DC3000 produces IAA via a previously uncharacterized pathway and identify a novel indole-3-acetaldehyde dehydrogenase , AldA , that functions in IAA biosynthesis by catalyzing the NAD-dependent formation of IAA from indole-3-acetaldehyde ( IAAld ) . Biochemical analysis and solving of the 1 . 9 Å resolution x-ray crystal structure reveal key features of AldA for IAA synthesis , including the molecular basis of substrate specificity . Disruption of aldA and a close homolog , aldB , lead to reduced IAA production in culture and reduced virulence on A . thaliana . We use these mutants to explore the mechanism by which pathogen-derived auxin contributes to virulence and show that IAA produced by DC3000 suppresses salicylic acid-mediated defenses in A . thaliana . Thus , auxin is a DC3000 virulence factor that promotes pathogenicity by suppressing host defenses .
Plant pathogens have evolved a variety of strategies to ensure a successful interaction with their hosts . These include the delivery of virulence proteins directly into host cells through the type III secretion system and production of plant hormones or hormone mimics . Both strategies are important for suppressing host defenses and/or modulating host physiology to promote colonization and disease development [1–3] . For example , the bacterial pathogen Pseudomonas syringae , the causal agent of bacterial speck disease [4 , 5] produces the phytotoxin coronatine , a molecular mimic of the plant hormone jasmonic acid-isoleucine [6 , 7] . Production and secretion of coronatine modulates host jasmonic acid signaling and is important for P . syringae pathogenesis [8–10] . Many plant-associated microbes also have the ability to synthesize indole-3-acetic acid ( IAA ) , a common form of the phytohormone auxin , and in several cases production of IAA has been implicated in pathogen virulence [11 , 12] . IAA synthesis in microbes has been well characterized , with five biosynthetic pathways for IAA utilizing the amino acid tryptophan ( Trp ) as the precursor ( Fig 1 ) identified to date [11] . These include the indole-3-acetamide ( IAM ) , the indole-3-acetonitrile ( IAN ) , the indole-3- pyruvate ( IPyA ) , the tryptamine ( TAM ) , and the tryptophan side-chain oxidase ( TSO ) pathways [13] . Detailed analyses of the IAM and IPyA pathways helped elucidate the role of bacterial IAA production in plant-microbe interactions . Two enzymes responsible for converting Trp to IAA via the IAM pathway are tryptophan 2-monoxygenase ( TMO ) and IAM hydrolase ( IAH ) , encoded by the iaaM and iaaH genes respectively [14] . Cloning of the iaaM and iaaH genes , as well as ipdC genes encoding IPyA decarboxylase [15–17] , from a wide range of bacteria and the characterization of their encoded proteins provided insight on the various roles for IAA synthesis during pathogenesis [11 , 18 , 19] . Auxin is involved in a broad range of growth and developmental processes in plants , including cell division and expansion and responses to a variety of environmental stimuli [20–22] . Auxin is also important in several plant-microbe interactions . For example , IAA produced by plant growth promoting rhizobacteria such as Azospirillum brasilense stimulates root growth [23] . IAA also promotes plant cell proliferation during gall formation caused by Rhizobium radiobacter ( formerly Agrobacterium tumefaciens ) [24] , Pantoea agglomerans [18] and P . savastanoi [19 , 25] . More recently auxin has been shown to promote virulence of P . syringae pv . tomato strain DC3000 . Exogenous application of auxin enhances disease susceptibility on Arabidopsis thaliana [26–28] and transgenic A . thaliana lines that over-express the YUCCA1 auxin biosynthesis gene and accumulate elevated levels of IAA exhibit enhanced susceptibility to DC3000 [29] . Additionally , impairment of auxin signaling in the plant can reduce susceptibility to P . syringae pv . tomato and maculicola [27 , 30] . Nonetheless , the role of pathogen-derived auxin in promoting P . syringae virulence remains to be elucidated . We sought to take advantage of the well-established DC3000-A . thaliana interaction to investigate the role of pathogen-derived IAA during pathogenesis . Here , we demonstrate that DC3000 produces IAA and identify an indole-3-acetaldehyde dehydrogenase , AldA , that catalyzes the NAD-dependent formation of IAA from indole-3-acetaldehyde ( IAAld ) . The x-ray crystal structure of AldA provides insight on the biochemical function of this enzyme . We show that disruptions of aldA and a close homolog ( aldB ) lead to reduced IAA production in DC3000 and reduced virulence in A . thaliana . Furthermore , we explore the mechanism by which pathogen-derived auxin contributes to DC3000 virulence and show that auxin produced by DC3000 suppresses salicylic acid ( SA ) -mediated defenses in A . thaliana .
Many P . syringae strains produce IAA in culture , and synthesize elevated levels of IAA when supplemented with Trp [31]; however , it has not been determined whether P . syringae pv . tomato strain DC3000 can synthesize IAA . To examine this , we grew DC3000 in Hoitkin-Sinden minimal media containing citrate ( HSC ) with shaking for 48 hours at 28°C . We chose this media as it is reported to more accurately reflect growth conditions in the intercellular space ( e . g . the apoplast ) of leaves colonized by P . syringae [32] . IAA concentrations in culture supernatants harvested at 24 and 48 hours were determined by LC-MS/MS . As observed for many other P . syringae strains , the level of IAA produced by DC3000 was significantly higher ( ranging from 100- to 200-fold greater , depending on the experiment ) when provided with Trp than in unsupplemented media ( Table 1 ) . The observation that DC3000 produces IAA in culture led us to investigate which pathway ( s ) DC3000 uses to synthesize IAA ( Fig 1 ) . The DC3000 genome annotation includes a TMO enzyme ( PSPTO0518; iaaM; [33] ) , but the predicted protein exhibits limited amino acid identity to enzymes with demonstrated IAA biosynthetic activity and is more closely related to a second group of TMO homologs that may function in pathways other than IAA synthesis [13] . Thus , it is unclear whether DC3000 uses the IAM pathway to synthesize IAA . To identify the IAA biosynthetic pathway ( s ) used by DC3000 , we performed IAA precursor feeding experiments using Trp , IAM , IAN , IPyA , TAM , or IAAld and analyzed DC3000 for IAA production by LC-MS/MS . Cultures supplemented with IAM , IAN , and TAM produced small but detectable amounts of IAA compared to cultures grown in HSC alone; however , these levels were relatively low compared to cultures fed with Trp ( Table 1 ) . In contrast , at least 100- to 500-fold higher levels of IAA , depending on the incubation time , were produced when DC3000 was grown in media supplemented with IAAld . This indicates that IAAld is an important intermediate for DC3000 IAA synthesis in culture . The feeding experiments with IPyA were inconclusive , as IPyA is unstable in solution [34] and high amounts of IAA accumulated spontaneously in HSC media containing IPyA , but lacking DC3000 ( Table 1 ) . Given the absence of an obvious ipdc gene in the DC3000 genome , it is unlikely that DC3000 uses IPyA to synthesize IAAld . Thus , we hypothesize that DC3000 synthesizes IAA via a pathway involving conversion of Trp to IAAld through a TSO activity [35 , 36] ( Fig 1 ) . We cannot rule out the ability of DC3000 to produce IAA through alternative pathways using IAM , IAN and/or TAM , as it is possible that limitations in the ability of DC3000 to take up these intermediates is responsible for the low level of IAA accumulation in feeding experiments with these intermediates . However , based on the results of our feeding studies these pathways do not appear to contribute significantly to IAA synthesis in culture . Our studies indicate that DC3000 synthesizes IAA via one or more pathways that involve IAAld as an intermediate ( Fig 1 ) . Thus , we predicted that disrupting the final step , which converts IAAld to IAA , would decrease IAA biosynthesis in DC3000 . To investigate this , we sought to identify the gene ( s ) encoding the enzyme ( s ) responsible for the conversion of IAAld to IAA . Previously , an Azospirilum brasilense mutant ( aldA ) with decreased IAA production was identified and the mutation mapped to a gene encoding a protein with ~80% amino acid identity to an annotated aldehyde dehydrogenase from Xanthobacter autotrophicus GJ10 [37] . Aldehyde dehydrogenases ( ALDs ) generally catalyze the conversion of aldehydes to carboxylic acids [38 , 39] . We predicted that a similar enzyme might metabolize IAAld to IAA in DC3000 , and thus utilized the amino acid sequences of the ALDs from A . brasilense and X . autotrophicus to identify putative ALDs in DC3000 . Using BLAST , we identified PSPTO_0728 , a putative ALD with ~70% amino acid identity to the ALD from X . autotrophicus . We then used the PSPTO_0728 sequence to search the DC3000 genome and identified 5 additional putative ALD homologs , PSPTO_0092 , PSPTO_2673 , PSPTO_3064 , PSPTO_3323 , and PSPTO_3644 , with ~30–40% amino acid identity to PSPTO_0728 . None of these proteins had previously been demonstrated to have dehydrogenase activity , nor were they described as involved in either auxin biosynthesis or DC3000 virulence . We examined whether these proteins could convert IAAld to IAA by expressing each gene individually in E . coli , growing the strains in LB media supplemented with 0 . 25 mM IAAld , and assaying the resulting strains for IAA production by LC-MS/MS . Background levels of IAA were produced by E . coli carrying the empty expression vector ( Fig 2 ) , consistent with previous reports [17 , 31] . Upon induction of expression of the ALDs from DC3000 ( S1 Fig ) , we observed increased IAA levels for three of the six proteins . The strains expressing either PSPTO_2673 or PSPTO_3644 showed ~10- and 5-fold increases in IAA levels , respectively ( Fig 2A ) . Cells expressing PSPTO_0092 showed the greatest accumulation of IAA with an ~200-fold increase in IAA over the empty vector control ( Fig 2B ) . Thus , PSPTO_0092 , PSPTO_2673 , and PSPTO_3644 can convert IAAld to IAA and likely function in DC3000 auxin biosynthesis . We refer to PSPTO_0092 , PSPTO_2673 , and PSPTO_3644 as AldA , AldB , and AldC , respectively , throughout this study . Based on sequence comparisons , AldA-C are members of the aldehyde dehydrogenase enzyme superfamily [38 , 39] ( S2 Fig ) . To examine the biochemical activity of the three putative ALDs from DC3000 , these proteins were expressed in E . coli as a N-terminal hexahistidine-tagged proteins and purified by nickel-affinity and size-exclusion chromatographies . Each of the putative ALDs was isolated with a monomer Mr~56 kDa , as determined by SDS-PAGE ( S3A Fig ) , which corresponds to the estimated molecular weights of AldA ( Mr = 52 . 7 kDa ) , AldB ( Mr = 53 . 1 kDa ) and AldC ( Mr = 51 . 8 kDa ) plus the addition of a His-tag . Size-exclusion chromatography of AldA and AldB indicates that each protein functions as a tetramer and that AldC is dimeric ( S3B Fig ) . In vitro assays of purified AldA , AldB and AldC using IAAld with either NAD+ or NADP+ as substrates confirm the major activity of AldA as that of an IAAld dehydrogenase , as each protein converted NAD ( P ) + to NAD ( P ) H only in the presence of the IAAld ( S3B Fig ) . Each Ald used NAD+ with a 10- to 40-fold preference versus NADP+ , but AldA had a specific activity ( 3 . 52 μmol min-1 mg protein-1 ) using IAAld as a substrate that was 100- and 800-fold higher than AldB and AldC , respectively . AldA-C displayed no changes in specific activities in the presence of calcium , magnesium , manganese , cobalt , nickel , and copper , which suggests that these proteins function as non-metallo NAD+-dependent ALDs . None of the three Alds showed detectable activity with IAA ( at 1 mM ) and NADH ( at 200 μM ) , indicating a clear preference for the formation of IAA compared to the reverse reaction . Kinetic analysis showed that AldA had a catalytic efficiency ( kcat/Km ) with IAAld as a substrate that was 130- and 710-fold higher than AldB and AldC , respectively ( S1 Table ) . AldA also showed more than a 300-fold higher kcat/Km with NAD+ compared to NADP+ . A similar cofactor preference was observed for AldB and AldC . The low activities of AldB and AldC did not allow for accurate determination of kinetic parameters for NADP+ . These biochemical comparisons suggest that AldA functions as an IAAld dehydrogenase and that AldB and AldC likely prefer other aldehyde substrates in vivo . To explore the molecular basis of IAAld dehydrogenase activity of AldA , its three-dimensional structure was determined by x-ray crystallography . The x-ray crystal structures of AldA in the apoenzyme , NAD+ complex , and NAD+•IAA complex forms were determined ( S2 Table ) . In each structure , two AldA monomers were in the asymmetric unit and packed to form a dimer , which then form a tetramer by crystallographic symmetry ( Fig 3A ) . The interface between two monomers buries ~2 , 450 Å2 of surface area with a ~3 , 800 Å2 interface between each of the dimer units . The overall fold of AldA shares structural similarity with ALDH2-3 ( 4PXL; 1 . 2 Å r . m . s . d . for ~480 Cα-atoms; 46% identity ) and ALDH2-6 ( 4PZ2; 1 . 3 Å r . m . s . d . for ~484 Cα-atoms; 46% identity ) from Zea mays , along with multiple human ALD structures ( 1 . 4–1 . 5 Å r . m . s . d . for ~400 Cα-atoms; 43–46% identity ) [40 , 41] . The AldA monomer adopts a canonical aldehyde dehydrogenase fold ( Fig 3B ) , which contains an NAD+-binding domain with a Rossmann-fold motif of a central β-sheet ( β10-β9-β8-β11-β12 ) surrounded by α-helices , a mixed α/β domain with the catalytic cysteine residue ( Cys302 ) , and an oligomerization domain with a protruding β-sheet ( β6-β7-β23 ) . The AldA•NAD+ and AldA•NAD+•IAA crystal structures define the position of the active site between the catalytic and cofactor binding domains ( Fig 3B and 3C ) . Although the ligand binding sites occupy two separate pockets on opposite sides of the monomer ( Fig 3C ) , both sites are linked by a ~25 Å tunnel that places the reactive groups of the co-substrates in proximity to Cys302 ( Fig 3D ) . Comparison of the AldA crystal structures suggests that ligand binding results in structural changes that order the active site ( S3D Fig ) . The α11-β14 loop ( residues 297–305 ) , which contains Cys302 , is disordered in the apoenzyme structure and has average temperature factors ~1 . 8-fold higher than surrounding residues . Likewise , a ~50 amino acid region of the catalytic domain ( residues 348–397; α13-β15-β16-α15-β17-β19 ) is disordered in the apoenzyme structure and displays elevated B-factors in ligand bound structures . Unambiguous electron density in the AldA•NAD+ and AldA•NAD+•IAA crystal structures identifies the respective ligand binding sites ( Fig 4A ) . In the NAD+ binding site , the cofactor is bound in a hydrophobic tunnel ( Fig 4B ) . The adenine ring of NAD+ lies in an apolar region that provides multiple van der Waals contacts . The adenine ring also forms two hydrogen bonds between the hydroxyl group of Tyr255 and a water . The adenine-ribose rings provide extensive polar interactions with AldA . The 2’-hydroxyl hydrogen bonds with Lys191 and Glu194 . Interactions with Ser193 , Ser245 , and Trp167 position the phosphate backbone in the binding site . The nicotinamide-ribose forms a bidentate interaction with Glu401 and the nicotinamide ring is bound by a water-mediated interaction with Thr243 and through a hydrogen bond from Glu267 . Sequence comparisons show a conserved NAD+ binding site in AldA , AldB and AldC ( S2 Fig ) . These interactions place the nicotinamide ring in proximity to the invariant catalytic cysteine ( Cys302 in AldA ) [38] . Crystallization of a ‘dead-end’ complex ( i . e . , AldA•NAD+•IAA ) provides insight on the IAAld binding site ( Fig 4A and 4C ) . Electron density was observed near the reactive Cys302 and modeled as IAA for refinement . In contrast to NAD ( H ) binding , the IAAld/IAA site is formed predominantly by apolar residues . The carboxylic acid of IAA forms hydrogen bonds with the sulfhydryl group of Cys302 , the amide side-chain of Asn168 , and the backbone nitrogen of Cys302 . Multiple aromatic and apolar residues , including Phe169 , Met173 , Trp176 , Val301 , and Phe467 , surround the indole moiety . Computational docking of IAAld into the active site yielded a solution that matched the crystallographically observed position of IAA ( Fig 3D ) . The docked IAAld overlays with IAA and positions the reactive aldehyde group of the substrate near Cys302 for subsequent catalysis . To understand the different activity with IAAld displayed by the three ALDs , homology models of AldB and AldC based on the AldA structure were generated . Although the NAD ( H ) binding sites of AldA-C are highly conserved , the residues in the aldehyde binding site of each enzyme displays greater variability ( S2 Fig ) . Compared to AldA , sequence differences in AldB and AldC alter the hydrophobicity , electrostatics , and surface shape of the site ( Fig 4D–4I ) . For example , the calculated hydrophobicity values of the IAAld/IAA binding site are 7 . 51 in AldA , -2 . 99 in AldB , and 2 . 78 in AldC ( Fig 4D–4F ) . Likewise , the surface electrostatics of AldB and AldC are more basic than AldA ( Fig 4G–4I ) . In addition , the shape of the site in each enzyme differs . The largely apolar IAAld/IAA binding site of AldA best fits the substrate molecule . Amino acid changes in the AldB may widen the substrate binding pocket . The wider and more basic nature of this site likely reduces catalytic efficiency of AldB with IAAld . Whereas , substitutions in the AldC substrate binding site likely constrict access to the catalytic cysteine and result in the even lower activity of this enzyme with IAAld . Thus , structural differences in the substrate binding sites of these ALD result in the preference of AldA for IAAld . To study the role of these ALDs in DC3000 IAA biosynthesis , we generated plasmid disruption mutants in aldA ( PSPTO_0092 ) , aldB ( PSPTO_2673 ) and aldC ( PSPTO_3644 ) ( S4 Fig ) . The mutant strains were indistinguishable from wild-type DC3000 in their growth rates in both rich ( NYG ) and defined media ( HSC ) ( S4E and S4F Fig ) . We monitored the ability of each mutant strain to produce IAA in culture when provided with IAAld . Only two mutants displayed reduced levels of IAA when compared to DC3000 ( Fig 5 ) . The aldA mutant displayed a ~70–80% reduction in IAA levels compared with DC3000 , whereas the aldB mutants exhibited a ~10–15% reduction in IAA levels . Although the levels of IAA that accumulated in the aldB mutant culture were always lower than wildtype , the values were not significantly different from wild-type in all experiments ( Fig 5A ) . These results indicate that AldA plays a major role in IAA synthesis in DC3000 , and that AldB may make only a small contribution to IAA synthesis . AldC does not appear to be involved in IAA synthesis . Previous studies indicate that auxin of plant origin promotes susceptibility to DC3000 and P . syringae pv . maculicola ES4326 [26–30]; however , it is unknown whether auxin produced by these strains contributes to their virulence . To examine this , we assayed the aldA and aldB mutants for altered virulence on A . thaliana plants . DC3000 grew to high levels when infiltrated into A . thaliana plants ( Fig 6 ) , while the aldA and aldB mutants exhibited a ~5-fold reduction in growth in multiple independent experiments . Although the reduced virulence phenotype was observed in most experiments ( in 11 of 18 experiments for aldA and in 10 of 17 for aldB ) , as is often the case for mutants with subtle virulence phenotypes , the level of in planta bacterial growth of the mutants was not always significantly different from that observed for wildtype DC3000 . Surface inoculation experiments were also performed to assay development of disease symptoms , but we did not reproducibly observe reduced symptom severity in plants infected with either ald mutant . Both the reduced IAA synthesis and reduced virulence phenotypes of the aldA mutant were complemented by introduction of the wild-type aldA genomic clone ( S5 Fig ) , indicating that aldA contributes to the virulence of DC3000 in Arabidopsis . We tested whether the ald genes have an additive effect on IAA synthesis and virulence by generating an aldA aldB double mutant in DC3000 . We monitored the ability of the double mutant to produce IAA in culture when fed with IAAld , and observed that IAA production was not significantly lower in the aldA aldB double mutant than in either single mutant ( Fig 5C ) . Thus , the contribution of AldB to overall IAA synthesis appears to be minor compared to that catalyzed by AldA . The aldA aldB double mutant also did not exhibit a further reduction in bacterial growth on A . thaliana plants compared to the single mutants ( Figs 6B and 7B ) . IAA may contribute to pathogenesis by suppressing host defenses mediated by the defense hormone SA [27 , 42] . We hypothesized that if pathogen-derived IAA promotes pathogen growth in planta by suppressing SA-mediated defenses , then the reduced growth of the DC3000 ald mutants in planta would be associated with elevated SA-mediated defenses due to an impairment in the ability to suppress SA-mediated defenses . To investigate this , we monitored the expression of PR1 , a commonly used marker for SA-mediated defenses in A . thaliana [29] , in plants infected with wild-type DC3000 and the aldA and aldB mutants 24 hours after inoculation . PR1 expression was induced by 24 hrs in plants infected DC3000 compared to mock treatment ( Fig 7A ) . Expression of PR1 was significantly higher in plants infected with the aldA mutant . There was also a significant increase in PR1 expression in plants infected with the aldB mutant; however , this increase was not as large as observed for the aldA mutant . These results suggest that DC3000-derived IAA is required for normal virulence via a mechanisms involving suppression of SA-mediated defenses . Given these findings , we predicted that the growth of the ald mutants would be restored to wild-type levels on A . thaliana mutants with impaired SA-mediated defenses . To test this , we inoculated the sid2-2 mutant , which carries a mutation in the ICS1 SA biosynthesis gene [43] , with DC3000 and the ald mutants and monitored bacterial growth . Wild-type DC3000 grew to higher levels in sid2-2 mutants plants than in wild-type Col-0 ( Fig 7B ) , consistent with previous reports that the sid2-2 mutant exhibits increased disease susceptibility to P . syringae [29 , 43] . Consistent with our earlier results , the aldA , aldB and aldA aldB double mutants exhibited reduced growth on wild-type plants compared to DC3000 . The aldA mutant grew to levels comparable to wild-type DC3000 on sid2-2 plants , but growth of the aldB and the aldA aldB double mutants only reached levels similar to that of wild-type DC3000 on Col-0 plants ( Fig 7B ) . The observation that the reduced growth of the aldA mutant is restored to normal levels in plants impaired for SA-mediated defenses suggests that AldA promotes pathogen virulence by suppressing SA-mediated defenses . The finding that the reduced growth of the aldB and the aldA aldB double mutants is only partially rescued in sid2-2 plants indicates that the reduced virulence of these mutants is not due solely to a defect in suppressing SA-mediated defenses .
We identified a family of ALDs that catalyze the oxidation of IAAld to IAA . Of this family , AldA is the enzyme primarily responsible for IAA synthesis from IAAld in culture ( Fig 5 ) . A second enzyme , AldB may also contribute to IAA synthesis , but seems less important than AldA , based both on its lower activity in vitro ( S3B Fig ) and on the observation that IAA production by the aldB mutant is only moderately reduced ( Fig 5 ) . The two enzymes do not function redundantly in culture , as IAA synthesis is not significantly further reduced in the aldA aldB double mutant . The observation that the double mutant still accumulates some IAA in cultures fed with IAAld suggests there may be one or more additional genes encoding IAAld dehydrogenase activity . Biochemically , aldehyde dehydrogenases ( ALDs ) are a large enzyme superfamily that convert aldehydes to carboxylic acids on a broad array of molecules [38 , 39 , 44] . In diverse organisms , multiple ALDs function in various metabolic pathways and provide house-keeping functions , such as the detoxification of reactive aldehydes produced by lipid peroxidation . As with other enzyme superfamilies , the aldehyde dehydrogenases are an excellent example of how evolution of different substrate specificity while retaining common reaction chemistry leads to functional diversity and tailoring of biological function [45] . This appears to be the case for the ALDs in DC3000 , as AldA has a specialized role in IAA biosynthesis and pathogenesis that is distinct from AldB and AldC . Structurally , AldA shares the same overall three-dimensional fold as other ALDs ( Fig 3 ) and functions as an NAD ( H ) -dependent enzyme ( S3B Fig; S1 Table ) . Although AldA shares ~40% amino acid identity with both AldB and AldC ( S2 Fig ) , kinetic analysis of AldA demonstrates a distinct preference for IAAld as a substrate compared to the other two enzymes . The x-ray crystal structure of AldA in complex with NAD+ and IAA reveals the molecular basis for the activity of this protein ( Fig 4 ) . In the reaction sequence catalyzed by AldA , substrate binding leads to conformational changes that order the active site for catalysis ( S3D Fig ) . The chemical mechanism would proceed as described for other aldehyde dehydrogenases [46] . For conversion of IAAld to IAA , the active site cysteine ( Cys302 ) acts as a nucleophile to attack the substrate aldehyde moiety . This leads to formation of a covalent intermediate . Subsequence transfer of a hydride from the substrate to NAD+ and nucleophilc attack by an activated water molecule on the resulting carbonyl of the intermediate releases the carboxylic acid product with the thiol acting as a leaving group . Comparison of the structure and sequence of AldA with AldB and AldC shows how changes alter the size , shape , hydrophobicity , and electrostatics of the binding pocket ( Fig 4D–4I ) . Thus , the evolution of the AldA substrate binding site leads to a preference for IAAld . Additional studies are needed to identify the preferred substrates of AldB and AldC . Overall , the biochemical and structural data presented here indicate that in P . syringae strain DC3000 AldA functions as an IAAld dehydrogenase in IAA biosynthesis . Recently , the dhaS gene from Bacillus amyloliquefaciens , encoding a putative IAAld dehydrogenase , was implicated in IAA synthesis [47] . However , studies to characterize the biochemical activity of this enzyme have not been reported . AldA and DhaS are the first ALDs described in either plants or microbes and suggests that the evolution of different metabolic routes to IAA synthesis can be exploited by microbial plant pathogens . We propose that AldA-dependent IAA synthesis in DC3000 involves the direct conversion of Trp to IAAld through TSO activity ( Fig 1 ) , as the DC3000 genome does not encode an obvious IPDC , nor do our feeding studies implicate TAM as an intermediate ( Table 1 ) . The TSO pathway , which has been reported in several P . fluorescens strains [11] , is not well characterized . A Tn5 mutant lacking TSO activity was identified in P . fluorescens strain CHA0 [35]; however , a gene encoding this activity has not been described . Future investigation of TSO activity in DC3000 will provide additional insight into IAA synthesis in P . syringae and other bacteria . We also investigated the hypothesis that DC3000 utilizes the IAM pathway , as this pathway is used by other IAA producing bacteria , including several Pseudomonas strains [12 , 31] . Neither our feeding studies nor recent bioinformatic and genetic analyses provide support for the existence of an IAM pathway in DC3000 . Patten et al . [13] noted that PSPTO_0518 , which is annotated as encoding a TMO ( Fig 1; iaaM , [33]; http://www . pseudomonas . com ) , shares only ~30% amino acid identity with enzymes with demonstrated TMO activity . PSPTO_0518 is more closely related to a second family of monooxygenases that may function in other pathways [13] . Further , our observation that mutation of PSPTO_0518 does not alter accumulation of IAA in cultures fed with Trp provides additional evidence for the absence of the IAM pathway in DC3000 [48] . Likewise , our feeding studies do not implicate the IAN pathway as a major contributor to IAA synthesis in DC3000 ( Table 1 ) . Many Pseudomonads , including P . syringae , P . fluorescens , P . putida , and P . aeruginosa , have genes predicted to encode proteins with ~90–95% sequence identity to AldA , including a nearly invariant conserved IAAld binding site . A survey of The Pseudomonas Genome Data Base ( www . pseudomonas . com ) revealed that AldA homologs are much more common in these genomes than TMO , which is only found in a few P . syringae or P . savastanoi strains [13] . Thus , we speculate that the AldA-dependent IAA biosynthesis pathway is the predominant IAA synthesis pathway in Pseudomonads . The role of IAA production in the biology of these microbes is yet to be elucidated; however , in the case of plant-associated bacteria , modification of the biology of their plant hosts seems likely . Alternatively , or additionally , IAA may be involved in signaling with other microbes in the soil or leaf epiphytic community [12 , 19] . Our observation that the aldA and aldB mutant strains exhibit reduced growth on A . thaliana plants ( Fig 6 ) suggests that AldA and AldB play important roles during pathogenesis . The observation that the ald mutants did not exhibit altered growth in culture ( S4 Fig ) and the fact they grow to high levels in sid2 plants ( Fig 7 ) indicates that the reduced growth of these strains in wild type plants does not reflect a general growth defect . Thus , both Ald activities contribute to DC3000 virulence on A . thaliana . Although kinetic comparisons indicate that AldA is more specific than AldB for IAAld , differences in protein expression in the microbe ( i . e . , high levels of AldB ) could allow for the less efficient enzyme to contribute to IAAld conversion to IAA . However , given that our biochemical studies suggest that AldB is not likely to use IAAld as a substrate ( S3B Fig ) , the role of AldB during pathogenesis is not clear . It is possible that oxidation of some other aldehyde by AldB contributes to virulence . We have not demonstrated that AldA catalyzes IAA production in planta , as it is technically difficult to distinguish pathogen-derived from plant-derived auxin in plant tissue . However , it is reasonable to expect that this is the case , as both Trp and IAAld are present in significant amounts in A . thaliana tissue [49 , 50] . Our observation that plants infected with the aldA mutant express elevated levels of PR1 mRNA ( Fig 7A ) suggests that pathogen-derived IAA promotes virulence by suppressing SA-mediated defenses . Consistent with this , we also showed that growth of the aldA mutant is restored to wild-type levels on SA-deficient plants ( Fig 7B ) . These findings agree with results from earlier studies demonstrating that exogenous application of auxin down-regulated SA-mediated defenses [27 , 51] . The observation that growth of the aldB and aldA aldB double mutants was only partially restored to wild-type levels in sid2-2 plants suggests that aldB plays some role in suppression of SA-mediated defenses , but is also involved in promoting virulence via an SA-independent process . We are currently investigating the role of the AldB enzyme in virulence . The finding that pathogen-derived IAA promotes DC3000 virulence by suppressing SA-mediated defenses contrasts with results from our previous studies with transgenic plants that overexpress the YUCCA1 ( YUC1 ) IAA biosynthesis gene and accumulate elevated levels of IAA [52] . We observed that YUC1 overexpressing plants exhibited increased susceptibility to DC3000 , but that neither SA accumulation nor SA-responsive gene expression was suppressed in these plants [29] . Further , plants carrying both the YUC1 overexpression construct and the sid2 mutation exhibited additive effects of enhanced susceptibility due to both elevated IAA and impaired SA-mediated defenses . These results suggest that in these plants , IAA promotes pathogen growth through a mechanism that functions independently of suppression of SA-mediated defenses [29] . The apparent discrepancy between these studies can be resolved by proposing that: 1 ) auxin promotes DC3000 virulence via multiple different mechanisms , and 2 ) pathogen-derived auxin and plant-derived auxin play different roles during pathogenesis ( Fig 8 ) . Our data suggest that the stimulatory effect of AldA-dependent DC3000-synthesized IAA on virulence acts via suppressing SA-mediated defense signaling , while auxin produced by the plant ( e . g . YUC1-dependent ) promotes pathogen growth via a mechanism that acts independently or down-stream of SA-mediated defenses . Another possible role for IAA during pathogenesis is through a direct effect on the pathogen , for example by regulating virulence gene expression . Previous studies have shown that IAA also acts as a microbial signaling molecule , and a variety of plant-associated bacteria respond to IAA [11 , 19 , 53] . Future studies examining the impact of the source , the targets , and possibly also the form of auxin during pathogenesis will provide important insight into the roles of auxin in promoting disease development by DC3000 . It will also be of interest to investigate whether auxin plays multiple roles in other plant-microbe interactions . DC3000 synthesizes IAA via the activity of the aldehyde dehydrogenase AldA . The DC3000 aldA mutant exhibits reduced virulence on A . thaliana and plants infected with aldA express elevated SA-mediated defenses , suggesting that pathogen-derived IAA promotes virulence by suppressing SA-mediated defenses . Previous studies have shown that exogenous application of auxin promotes disease [26 , 30] and inhibits SA-mediated defenses [27] , but that in transgenic plants overexpressing the YUCCA1 ( YUC1 ) IAA biosynthesis gene and that accumulate elevated IAA , increased susceptibility to DC3000 occurs via a mechanism that does not involve suppression of SA-mediated defenses [29] . Together , these observations suggest that pathogen-produced auxin and plant-produced auxin promote disease via different mechanisms . SA , salicylic acid; ICS1/SID2 , ISOCHORISMATE SYNTHASE 1 , PR1 , PATHOGENESIS RELATED 1
The bacterial strains and plasmids used in this study are summarized in S3 Table . P . syringae strain DC3000 wild-type and mutant strains were grown on Nutrient Yeast Glycerol Medium ( NYG ) [54] or Hoitkin Sinden ( HS ) Medium ( with appropriate carbon sources added ) at 28°C . HS was prepared as described in [55] . Escherichia coli strains were maintained on Luria Broth ( LB ) medium at 37°C . Antibiotics used for selection of P . syringae strains include: rifampicin ( Rif , 100 μg mL-1 ) , kanamycin ( Kan , 25 μg mL-1 ) , and tetracycline ( Tet , 16 μg mL-1 ) . Antibiotics used for selection of E . coli strains were ampicillin ( Amp , 100 μg mL-1 ) , Kan ( 25 μg mL-1 ) and chloramphenicol ( Cm , 20 μg mL-1 ) . A modified version of the pJP5603 suicide vector [56] , pJP5603-Tet , in which the KanR cassette was replaced with the TetR gene , was constructed for generation of double insertion/disruption mutants . The pJP5603-Tet vector was made by digesting pJP5603 with XbaI and BglII to release the ~1 . 3kb KanR cassette , and an ~2 . 9kb XbaI and BglII fragment containing the TetR gene from pME6031 was inserted in its place . P . syringae strains were grown in NYG medium with Rif in overnight cultures . Cells were collected by centrifugation from each overnight culture , washed twice with 10 mM MgCl2 , re-suspended at a density of ~1 x 107 cells mL-1 in HS minimal media containing 10 mM citrate and incubated with shaking for 48 hrs at 28°C . The culture medium was supplemented with 0 . 25 mM L-Tryptophan ( Trp , Sigma Aldrich , Cat No . T-0254 ) , Indole-3-acetamide ( IAM , Sigma Aldrich , Cat No . 286281 ) , 3-Indoleacetonitrile ( IAN , Sigma Aldrich , Cat No . 129453 ) , Tryptamine hydrochloride ( TAM , Sigma Aldrich , Cat No 246557 ) or Indole-3-acetaldehyde–sodium bisulfite addition compound ( IAAld , Sigma Aldrich , Cat No . I1000 ) , as indicated . One mL samples were taken at 24 and 48 hrs after incubation , centrifuged to pellet the cells and the resulting supernatants frozen in liquid nitrogen and stored at -80°C . Growth of cultures was monitored by reading the OD600 at regular intervals with a spectrophotometer . The samples were prepared and analyzed for IAA production by LC-MS/MS as described in Supplemental Information ( S1 Text ) . BLASTP searches were performed using the National Center for Biotechnology Information ( NCBI ) server to search non-redundant databases for P . syringae DC3000-specific sequences . P . syringae strain DC3000 sequence information was obtained from Kyoto Encyclopedia of Genes and Genomes ( KEGG; www . genome . jp/kegg ) and the Pseudomonas-Plant Interaction website ( PPI; www . pseudomonas-syringae . org ) . Accession numbers for genes used in this study are: aldehyde dehydrogenase ( AldA ) from A . brasilense: AY850388; chloroacetaldehyde dehydrogenase ( AldA ) from X . autotrophicus: AF029733; DC3000 PSPTO_0092 ( AldA ) : NP_789951 . 1; DC3000 PSPTO_2673 ( AldB ) : NP_792480 . 1; DC3000 PSPTO_3644 ( AldC ) : NP_793419 . 1 . To make the pET21a-0092 ( AldA ) expression plasmid , the full-length coding sequence ( CDS ) from PSPTO_0092 was amplified from DC3000 genomic DNA with primers 0092NdeI F and 0092XhoIR ( S4 Table ) . The resulting ~ 1 . 5 kb PCR fragment was cloned into the pBlunt II-TOPO vector ( Invitrogen ) , transformed into E . coli DH5α and plated on LB media containing Kan . The resulting pTOPO-0092 plasmid was sequenced to confirm that no PCR-derived mutations were introduced into the clone , and then was digested with NdeI and XhoI and the approximately ~1 . 5 kb insert corresponding to the PSPTO_0092 CDS was ligated into the pET21a vector cut with the same enzymes to generate pET21a-0092 . The pET21a-0092 plasmid was transformed into E . coli BL21 ( DE3 ) . The same strategy was used to generate pET21a-0728 , pET21a-2673 ( AldB ) , pET21a-3064 , pET21a-3323 and pET21a-3364 ( AldC ) ( see S3 and S4 Tables for primers and strains ) . For E . coli expression assays to monitor IAA production , the E . coli strains carrying the pET21a-DC3000 putative aldehyde dehydrogenase ( Ald ) constructs were grown in triplicate cultures overnight in LB media containing Amp with shaking at 37°C . Overnight cultures were diluted 1/100 and incubated with shaking until an OD600nm 0 . 4–0 . 6 was reached . Cultures were induced with IPTG ( 1 mM final concentration ) , supplemented with 0 . 25 mM IAAld and incubated with shaking for an additional 24 hrs . One mL samples were taken 1 . 5 hrs after IPTG induction to verify induction of the putative Ald proteins . This was done by centrifuging the samples , boiling the resulting cell pellets in SDS-PAGE buffer and loading equal amounts of cell lysate on an acrylamide gel for visualization of protein . Additional 1mL samples were taken at 24 hrs after IPTG induction , centrifuged to pellet cells and the resulting supernatants were frozen in liquid nitrogen and stored at -80°C . The samples were analyzed for IAA production by LC-MS/MS [57] . The pET28a-AldA , pET28a-AldB , and pET28a-AldC constructs used to express protein for biochemical experiments were generated using NdeI and XhoI enzyme sites and transformed into E . coli BL21 ( DE3 ) cells ( Agilent Technologies ) . Cells were grown at 37°C in Terrific broth containing 50 μg mL-1 Kan until OD600nm = 0 . 8 and induced with 1 mM IPTG at 18°C . Cells were harvested by centrifugation ( 4 , 500 x g; 15 min ) and re-suspended in lysis buffer ( 50 mM Tris , pH 8 . 0 , 500 mM NaCl , 25 mM imidazole , 10% glycerol , and 1% Tween-20 ) . After sonication and centrifugation ( 11 , 000 x g; 30 min ) , the supernatant was loaded onto a Ni2+-NTA column ( Qiagen ) previously equilibrated with lysis buffer . Wash buffer ( lysis buffer without Tween-20 ) was used to remove unbound proteins , and then bound Ald protein was eluted using wash buffer containing 250 mM imidazole . The His-tagged Ald protein was loaded onto a Superdex-200 26/60 size-exclusion column ( GE healthcare ) equilibrated in 25 mM Hepes ( pH 7 . 5 ) and 100 mM NaCl . Fractions with Ald protein were pooled , concentrated to 10 mg mL-1 , and stored at -80°C . Protein concentrations were determined using molar extinction coefficients at A280nm for each Ald , as calculated using ProtParam . Enzymatic activity of each Ald was measured by monitoring NADH formation ( ε340 = 6220 M−1 cm−1 ) at A340nm on an Infinite M200 Pro plate reader ( Tecan ) . Standard assay conditions for Ald were 100 mM Tris•HCl ( pH 8 . 0 ) , 100 mM KCl in 200 μL at 25°C . For specific activity determinations , the following substrate concentrations were used: 1 mM IAAld and either 1 mM NAD+ or 1 mM NADP+ . For determination of steady-state kinetic parameters , reactions were performed in standard assay conditions with either fixed NAD+ ( 1 . 0 mM ) and varied IAAld ( 0 . 05–2 . 5 mM ) or with fixed IAAld ( 1 . 0 mM ) and varied NAD+ ( 0 . 05–2 . 5 μM ) . All data were fit to the Michaelis-Menten equation , v = ( kcat[S] ) / ( Km + [S] ) , using SigmaPlot . Crystallization of AldA was performed at room temperature using the vapor diffusion method in hanging drops of a 1:1 mixture of protein ( 10 mg mL-1 ) and crystallization buffer . Crystals of the AldA apoenzyme were obtained in 10% ( w/v ) PEG-8000 , 100 mM HEPES , pH 7 . 5 , and 8% ( v/v ) ethylene glycol . Crystals of the AldA•NAD+ and AldA•NAD+•IAA complexes were obtained in 8% ( w/v ) PEG-8000 and 100 mM Tris•HCl ( pH 8 . 5 ) supplemented with either 5 mM NAD+ or 5 mM NAD+ and 5 mM IAA , respectively . Crystals were stabilized in cryoprotectant ( crystallization solution with either 30% glycerol or 30% ethylene glycol ) before flash freezing in liquid nitrogen for data collection at 100 K . Diffraction images were collected at beamline 19ID of the Advanced Photon Source at the Argonne National Lab . Diffraction data were indexed , integrated and scaled using HKL3000 [58] . The structure of AldA in complex with NAD+ was were solved by molecular replacement using PHASER [59] with betaine aldehyde dehydrogenase from Staphylococcus aureus , which shares 40% amino acid identity with AldA , as a search model ( PDB: 4MPB; [60] . For iterative rounds of manual model building and refinement , COOT [61] and PHENIX [62] were used , respectively . The resulting model of AldA was used to solve the structures of the apoenzyme and NAD+•IAA complex by molecular replacement with PHASER . Model building and refinement was as described above . Data collection and refinement statistics are summarized in S2 Table . Atomic coordinates and structure factors were deposited in the RCSB Protein Data Bank ( www . rcsb . org ) as follows: AldA ( 5IUU ) ; AldA•NAD+ ( 5IUV ) ; and AldA•NAD+•IAA ( 5IUW ) . Molecular homology models of AldB and AldC were generated using the homology-modeling server of SWISS-MODEL with the 1 . 93 Å resolution crystal structure of AldA• NAD+•IAA ( chain B ) as a template . Molecular docking experiments were performed by Autodock vina ( Version 1 . 1 . 2 ) [63] with standard protocols . Docking of IAAld ( substrate ) into the AldA active site used a 30 × 30 × 30 Å grid box with the level of exhaustiveness = 20 . The position of NAD+ was fixed based on its position in the AldA• NAD+•IAA structure . Docking of IAAld yielded a calculated affinity of -5 . 9 to -4 . 8 kcal mol-1 . To generate the aldA::pJP5603 insertion disruption strain , an ~0 . 5 kb SacI-XbaI genomic fragment internal to the aldA ( PSPTO_0092 ) CDS was amplified from P . syringae DC3000 genomic DNA with the primers 0092SacIF and 0092XbaIR ( see S4 Table for primer sequences ) . The resulting PCR fragment was cloned into the pBlunt II-TOPO vector ( Invitrogen ) , transformed into E . coli DH5α and plated on LB media containing Kan . Several pTOPO-0092int clones were sequenced to verify that there were no PCR-derived mutations . The genomic fragment was then cloned into the pJP5603 KanR suicide vector [56] by digesting the pTOPO-0092int clone with SacI and XbaI and ligating the resulting genomic fragment into pJP5063 digested with SacI and XbaI to generate pJP5603-0092int . The pJP5603-0092int plasmid was transformed into E . coli DH5α λpir and introduced into P . syringae DC3000 via bacterial conjugation using the helper strain MM294A ( pRK2013 ) ( S3 Table ) [64] . DC3000 trans-conjugates were selected for Rifr and Kanr resistance on NYG media containing Rif and Kan at 28°C . The same strategy was used to generate aldB::pJP5603 and aldC::pJP5603 single mutants , as well as aldA::pJP5603 aldB::pJP5603-Tet , double mutant strains . To generate double mutants , a TetR version of the pJP5603-aldB insertion disruption suicide plasmid was used ( see S3 and S4 Tables for primers and strains ) . Plasmid disruption of aldA by pJP5603 was confirmed by PCR using primers M13F , 0092seqF , and 0092seqR . Disruption of the wild-type genomic copy was verified by amplification of an ~1 . 1 kb fragment with M13F and 0092seqF primers in the aldA:pJP56023 strain and the absence of a band of this size in wild-type DC3000 and aldB::pJP5603 strains ( S4C and S4D Fig ) . The same strategy was used to confirm all of the additional single and double ald mutants , using M13F and 2673SeqF to amplify the aldB::pJP5603 and aldB::pJP5603TetR disruptions , and M13F and 3644seqF to amplify the aldC::pJP5603 disruption ( see S3 and S4 Tables for strains and primers ) . When generating the aldA aldB double mutant , special attention was given to identifying strains in which the pJP5603Tet-2673int disruption plasmid had integrated into the chromosome at the aldB locus , rather than via homologous recombination with the pJP5603 vector sequences in pJP5603-0092int integrated at aldA . To generate the aldA complementing clone , pAldA , the aldA coding sequence and 5’ regulatory region were amplified from genomic DNA using primers 0092XhoIF and 0092EcoRIR . The resulting ~2 kb PCR product was cloned into the pBlunt II-TOPO vector ( Invitrogen ) to generate pTOPO-0092comp . This plasmid was then digested with XhoI and EcoRI and the 2 kb insert ligated into the broad host range plasmid pME6031 vector with Xho1 and EcoRI compatible ends to generate pME6031-0092 ( pAldA ) ( S3 Table ) . The pAldA plasmid was introduced into the aldA::pJP5603 mutant strain via bacterial conjugation using the helper strain MM294A ( pRK2013 ) . DC3000 trans-conjugates were selected for Rifr , Kanr and Tetr resistance on NYG media containing Rif , Kan and Tet at 28°C . All A . thaliana transgenic lines and mutants used in this study were in the Col-0 background . The 35S:YUC1 overexpression line [52] was obtained from Yunde Zhao . The sid2-2 mutant [43] was obtained from Mary Wildermuth . Plants were grown on soil in a growth chamber with a short-day photoperiod ( 8 h light/16 h dark ) at 21°C and 75% relative humidity , with a light intensity of ~ 130 μEinsteins sec-1 m-1 . A . thaliana plants were infected at approximately four weeks of age . For surface inoculations plants were dipped into a solution containing P . syringae at approximately 3x108 cells mL-1 ( OD600nm = 0 . 3 ) , 10 mM MgCl2 and 0 . 02% ( v/v ) Silwet L-77 [65] . For syringe infiltrations , a solution containing 104–105 cells mL-1 ( OD600nm = 10−5–10−4 ) in 10 mM MgCl2 was injected into leaves using a 1-mL needleless syringe . To quantify bacterial growth in the plant , whole leaves were sampled at various time points after inoculation , weighed to determine leaf mass , ground in 10 mM MgCl2 and then plated in serial dilutions on NYG media with rifampicin . Between four and eight leaves were sampled per treatment , depending on the experiment . To generate the composite growth curves shown in Figs 6A and 7B , data from independent experiments in which wild-type DC3000 grew to similar levels ( e . g . ~ 1 x 105 cfu/mg leaf tissue ) were combined . Quantification of disease symptoms following dip inoculation was carried out four days post inoculation . Leaves were categorized based on the presence and amount of chlorosis or yellowing of the leaf . For ~ 10 plants per each treatment , each leaf was individually assessed for percent of the leaf exhibiting chlorosis , ranging from leaves with no yellowing to leaves displaying >75% chlorosis . The Student’s t-test was used for all statistical analysis . | Pathogens have evolved multiple strategies for suppressing host defenses and modulating host physiology to promote colonization and disease development . For example , the plant pathogen Pseudomonas syringae uses several strategies to the manipulate hormone signaling of its hosts , including production of virulence factors that alter hormone responses in and synthesis of plant hormones or hormone mimics . Synthesis of indole-3-acetic acid ( IAA ) , a common form of the plant hormone auxin , by many plant pathogens has been implicated in virulence . However , the role of pathogen-derived IAA during pathogenesis by leaf spotting pathogens such as P . syringae strain DC3000 is not well understood . Here , we demonstrate that P . syringae strain DC3000 uses a previously uncharacterized biochemical pathway to synthesize IAA , catalyzed by a novel aldehyde dehydrogenase , AldA , and carry out biochemical and structural studies of the AldA protein to investigate AldA activity and substrate specificity . We also generate an aldA mutant disrupted in IAA synthesis to show that IAA is a DC3000 virulence factor that promotes pathogenesis by suppressing host defense responses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"crystal",
"structure",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"pathogens",
"brassica",
"microbiology",
"condensed",
"matter",
"physics",
"organic",
"compounds",
"hormones",
"plant",
"science",
"mo... | 2018 | Indole-3-acetaldehyde dehydrogenase-dependent auxin synthesis contributes to virulence of Pseudomonas syringae strain DC3000 |
The SWI/SNF chromatin remodeling complexes regulate the transcription of many genes by remodeling nucleosomes at promoter regions . In Drosophila , SWI/SNF plays an important role in ecdysone-dependent transcription regulation . Studies in human cells suggest that Brahma ( Brm ) , the ATPase subunit of SWI/SNF , regulates alternative pre-mRNA splicing by modulating transcription elongation rates . We describe , here , experiments that study the association of Brm with transcribed genes in Chironomus tentans and Drosophila melanogaster , the purpose of which was to further elucidate the mechanisms by which Brm regulates pre-mRNA processing . We show that Brm becomes incorporated into nascent Balbiani ring pre-mRNPs co-transcriptionally and that the human Brm and Brg1 proteins are associated with RNPs . We have analyzed the expression profiles of D . melanogaster S2 cells in which the levels of individual SWI/SNF subunits have been reduced by RNA interference , and we show that depletion of SWI/SNF core subunits changes the relative abundance of alternative transcripts from a subset of genes . This observation , and the fact that a fraction of Brm is not associated with chromatin but with nascent pre-mRNPs , suggest that SWI/SNF affects pre-mRNA processing by acting at the RNA level . Ontology enrichment tests indicate that the genes that are regulated post-transcriptionally by SWI/SNF are mostly enzymes and transcription factors that regulate postembryonic developmental processes . In summary , the data suggest that SWI/SNF becomes incorporated into nascent pre-mRNPs and acts post-transcriptionally to regulate not only the amount of mRNA synthesized from a given promoter but also the type of alternative transcript produced .
Messenger RNAs ( mRNAs ) are synthesized in eukaryotic cells as precursor RNA molecules ( pre-mRNAs ) , which are then assembled into ribonucleoprotein complexes ( pre-mRNPs ) during transcription . The newly synthesized pre-mRNAs are modified by capping , splicing and 3′-end maturation reactions that involve cleavage and polyadenylation of the 3′-end of the transcript . The existence of alternative splicing and alternative polyadenylation sites in many pre-mRNAs is a major source of protein variability , and the regulation of pre-mRNA processing is a major mode of genetic regulation [1] , [2] . Many observations during the last decade have indicated that some of the mechanisms that regulate the processing of the pre-mRNA are related to the transcription process itself , and that chromatin dynamics , transcription and pre-mRNA processing are functionally connected [reviewed in 3] . Transcription can influence the usage of alternative splice sites in the pre-mRNA [4] through several mechanisms . Promoter-specific coregulators can recruit splicing factors to transcribed genes or they can themselves play dual roles in the regulation of transcription initiation and alternative splicing [review by 5] . For example , the hnRNP-like protein CoAA , which coactivates the transcription of multiple genes regulated by steroid hormones [6] , is recruited to the promoters of its target genes by the coactivator TRBP/NCoA6 and regulates splice site selection [7] . The CAPER proteins , members of the U2AF65 family , and the polyC-RNA-binding protein 1 , PCBP1 , are further examples of transcriptional coactivators that influence alternative splicing in a promoter-dependent manner [8] , [9] . The RNA polymerase itself can also participate in the recruitment of pre-mRNA processing factors . A reported case is that of SRp20 , a member of the SR protein family of splicing regulators . SRp20 is recruited to transcribed genes through interaction with the C-terminal domain ( CTD ) of the large subunit of RNA polymerase II [10] . It has been proposed that these proteins are recruited to the promoter , travel along the gene with the transcription machinery , and are eventually delivered to the nascent pre-mRNA for splicing regulation [5] . Low transcription elongation rates favor the usage of proximal splice sites by increasing the time during which the proximal sites are exposed to the splicing machinery before more distal sequences are synthesized . This provides a further mechanism by which transcription influences pre-mRNA processing [reviewed in 11] . A recent study has shown that the human Brahma ( hBrm ) protein , a chromatin remodeling factor , regulates the alternative splicing of several genes in human cells [12] . Overexpression and depletion experiments have shown that hBrm , together with the mRNA-binding protein Sam68 , favors the accumulation of RNA pol II at specific gene positions and decreases the elongation rate of the RNA pol II . These effects favor the inclusion of variable exons with weak splice sites [12] . The hBrm protein and its paralog hBrg1 are the ATPase subunits of the SWI/SNF chromatin remodeling complexes in human cells [reviewed in 13] . The SWI/SNF complexes regulate the transcription of many genes in mammalian cells by remodeling nucleosomes at promoter regions [reviewed in 14] . The Brm protein in Drosophila melanogaster ( dBrm ) is associated with transcribed loci in the polytene chromosomes and plays an important role in ecdysone-dependent transcription regulation [15] , [16] . Two types of dBrm-containing complexes , BAP and PBAP , are present in Drosophila . They share seven core subunits , including dBrm , but differ in the presence of additional signature subunits [17 and references therein] . We have analyzed in situ the association of Brm with the actively transcribed Balbiani ring ( BR ) genes of the dipteran Chironomus tentans ( C . tentans ) in order to obtain further insight into the mechanisms by which Brm influences pre-mRNA splicing . The BRs are giant puffs that form by active transcription of the BR genes in the polytene chromosomes of the salivary gland cells [18] . The BR pre-mRNAs synthesized in the BR1 and BR2 puffs have all the features of typical pre-mRNAs . They are capped at the 5′-end [19] , spliced , cleaved and polyadenylated at the 3′-end [20] , released from the chromosome , and finally exported to the cytoplasm [reviewed in 21] . It is possible to visualize using transmission electron microscopy ( TEM ) how the BR pre-mRNPs are synthesized along the BR genes [22] , and it is possible to study the association of defined proteins with nascent BR pre-mRNP particles in situ using specific antibodies . The Brm protein of C . tentans ( ctBrm ) is associated with the BR puffs and is widely distributed along the active BR genes , as shown by immuno-electron microscopy ( immuno-EM ) and chromatin immunoprecipitation ( ChIP ) , which suggests that ctBrm has further roles in addition to that of regulating transcription initiation [23] . We have analyzed the location of ctBrm in the actively transcribed BR genes in more detail , and we have shown that a fraction of Brm is associated with the nascent transcripts in both insect and mammalian cells . We have also determined whether Brm plays a role in pre-mRNA processing in insects and we have analyzed the expression profiles of D . melanogaster cells in which the levels of dBrm have been reduced by RNA interference ( RNAi ) . We show that depletion of dBrm affects not only the splicing but also the usage of alternative polyadenylation sites .
We used three different antibodies against ctBrm to study the association of this protein with the BR genes of C . tentans . The first one , Ab1 , was raised against the rat Brg1 protein [24] . The second antibody , Ab2 , was raised against the C-terminal part of the Ct-BRM protein ( Figure S1 ) . The third one , Ab3 , was raised against dBrm and has been characterized by Zraly et al . [25] . The specificity of the antibodies was tested by Western blot against nuclear protein extracts prepared from C . tentans cultured cells . The three antibodies recognized one major band of approximate molecular mass 200 kDa , the expected molecular mass of ctBrm ( Figure S2A ) . The same band was present in protein extracts prepared from larval salivary glands ( Figure S2B ) . Moreover , the ctBrm protein immunoprecipitated by Ab1 was recognized by Ab2 and by Ab3 ( Figure S2C ) . We concluded that the three antibodies recognized the same protein , ctBrm . Immunofluorescent staining of isolated polytene chromosomes gave a banded pattern from all three antibodies , and the antibodies stained many chromosomal bands . The actively transcribed BR puffs were among the most intensely stained loci ( Figures 1A–C and S3A–D ) . In some cases , the chromosomes were co-stained with a mAb against Hrp45 , an hnRNP protein used as a marker to visualize the BRs in chromosome IV ( Figures 1A and S3C–D ) . Preparations of isolated polytene chromosomes were digested with RNase A before immunostaining to determine whether the association of Brm with the chromosomes was mediated by RNA ( Figure 1D ) . The chromosomes were co-stained with Y12 , a mAb against core snRNP proteins , to monitor the effect of the RNase treatment . Control chromosomes were incubated in parallel in the absence of RNase A . The snRNP staining ( red in Figure 1D ) was abolished by the RNase treatment , as expected . The intensity of the Brm staining ( green in Figure 1D ) was reduced . However , a part of the Brm staining was resistant to the RNase treatment . These results suggest that there are two modes of interaction of Brm with the chromosomes . One mode is independent of the presence of RNA and may be explained by a direct association of Brm with the chromatin . The other mode requires RNA . We have previously mapped the association of ctBrm with the BR1 gene by immuno-EM using the anti-Brg1 antibody , Ab1 , and we have shown that ctBrm is widely distributed along the entire BR1 gene [23] . We have here confirmed this observation by extending the immuno-EM analysis to BR1 and BR2 , where we have used the C . tentans-specific antibody Ab2 . The results obtained are summarized in Figure 2 . The BR1 and BR2 genes are approximately 40 kb long and they are transcribed simultaneously by several RNA polymerases . The different regions of the gene show specific morphological features due to the progressive growth and assembly of the nascent BR pre-mRNPs ( Figure 2B ) . Full-length genes are not available in the sections used for TEM , but partial gene segments are observed ( Figure 2C–D ) . The gene segments can be classified into proximal , middle and distal segments , based on the morphology of the nascent pre-mRNPs . The pre-mRNPs in the proximal region appear as growing fibers with increasing length , whereas the pre-mRNPs in the middle and distal regions appear as stalked granules of increasing diameter . We isolated chromosome IV from salivary glands and stained the isolated chromosomes with either Ab1 or Ab2 . The antibody-binding sites were revealed using a gold-conjugated secondary antibody . Control chromosomes were processed in parallel in order to assess the specificity of the immunolabeling . The chromosomes were embedded in plastic after the immunolabeling and sectioned for TEM analysis . Photographs were taken at random positions , and each gold particle was classified according to its association with a proximal , middle or distal gene segment ( Figures 2C and 2D ) . ctBrm was present on the proximal , middle and distal regions of the BR genes . Similar results were obtained with Ab1 and Ab2 ( Figure 2E ) . These results confirm that ctBrm is widely distributed along the BR genes . We wanted to determine whether the ctBrm protein located at the BR genes was associated with the chromatin or with the nascent BR pre-mRNPs . Detailed analysis of immuno-EM data provides enough resolution to distinguish between labeling of the pre-mRNPs and labeling of the chromatin , as shown by Wetterberg et al . [26] . We selected 60 distal BR segments in which the relative positions of the chromatin axis and the nascent pre-mRNPs could be identified , and we determined for each segment whether the gold markers were close to the chromatin ( within 50 nm of the axis ) or distant from the chromatin ( more than 50 nm from the axis ) . The dimensions of the antibodies mean that this latter group contains only gold markers associated with BR pre-mRNP . In contrast , the markers close to the chromatin may label ctBrm molecules bound to the stalk of the pre-mRNP , bound to the chromatin , or bound to the transcription machinery . The gold markers were distant from the chromatin in 20 out of 60 analyzed cases ( 33% ) . A fraction of ctBrm was associated with the nascent pre-mRNPs and this association was confirmed with all three anti-Brm antibodies , as shown in Figure 3 . This result agrees with the results of the RNase A digestion experiments shown in Figure 1D . We conclude that a fraction of ctBrm is associated with the nascent BR pre-mRNPs and is not in contact with the chromatin . We used a cell fractionation assay to confirm the association of ctBrm with RNPs . We isolated nuclei from C . tentans tissue culture cells and prepared two types of protein extracts ( Figure 4A ) . One of the extracts contained soluble nuclear proteins ( soluble ) and the other extract contained proteins bound to the chromosomes via RNA ( chromosomal RNP ) . These proteins could be released by RNase A digestion . The proteins in each fraction were resolved by SDS/PAGE and analyzed by Western blotting . Coomassie Blue staining showed that each fraction contained a different set of proteins ( Figure 4B ) . Antibodies against Hrp36 , an abundant member of the hnRNP A family , histone H3 and the TATA-binding protein ( TBP ) were used as controls to assess the quality of the fractions . Hrp36 was present in both the soluble and the chromosomal RNP preparations , as expected , whereas histone H3 and TBP were found in the soluble fraction and in the pellet ( which contained nuclear components that were not extracted by RNase , such as chromatin and the nuclear envelope ) . ctBrm was present in the soluble nuclear fraction and in the chromosomal RNP fraction ( Figure 4C ) . The low abundance of ctBrm in the pellet ( lane 3 , Figure 4C ) was unexpected considering that a fraction of ctBrm remained associated with the chromosomes after RNase digestion ( Figure 1D ) . The signal in the pellet was considerably increased by using a sample buffer supplemented with 8 M urea , as shown in the bottom panel of Figure 4C . This observation is consistent with the immunofluorescence experiments and indicates that a fraction of ctBrm is highly insoluble . We repeated the fractionation experiments using Drosophila S2 cells ( Figure 4D ) and human HeLa cells ( Figure 4E ) . The mRNA-binding proteins Hrp59 and SAP155 were used as RNP controls in Drosophila and human extracts , respectively . In S2 cells , dBrm was clearly present in the soluble and chromosomal RNP fractions . In HeLa cells , hBrm and hBrg1 were also present in both soluble and chromosomal fractions . We analyzed the presence of other core SWI/SNF subunits in the fractions from HeLa cells . All the analyzed subunits were present in the soluble fraction , as expected , and all of them were also present in the chromosomal RNP fraction ( Figure 4E ) . In summary , we conclude that a fraction of the Brm protein is associated with chromosomal RNPs and that the association of Brm with the RNPs is conserved from insects to mammals . Our results also suggest that hBrm and hBrg1 are not bound to nascent RNPs as individual proteins in human cells , but as components of SWI/SNF complexes . The hBrm and hBrg1 proteins interact with snRNPs [12] , [27] . We showed that they interact with snRNPs in HeLa cells using immunoprecipitation experiments with the Y12 antibody against core snRNP proteins ( Figure 4F ) . We showed also that ctBrm interacts with snRNP complexes in C . tentans ( Figure 4G ) . We prepared a soluble RNP extract , immunoprecipitated snRNPs with Y12 , and probed the immunoprecipitated proteins with anti-Brm antibodies . ctBrm was co-immunoprecipiated with snRNPs in the soluble fraction ( Figure 4G , lane 4 ) . To determine whether the interaction of ctBrm with snRNPs is a direct protein-protein interaction or whether it requires RNA , the immunoprecipitation was carried out as above and the bound material was treated with RNase A before elution . As shown in Figure 4G , lane 5 , the signal intensity was significantly reduced in the RNase-treated sample . We assessed the specificity of the interaction by re-probing the blot with an antibody against TBP , a protein that is not expected to interact with snRNPs . We conclude that ctBrm is associated with snRNPs and that the association is RNA-dependent . We next wanted to determine whether Brm affects pre-mRNA processing in Drosophila . Moshkin and coworkers have determined the expression profiles of Drosophila S2 cells after depletion of individual SWI/SNF subunits by RNAi and microarray hybridization using the Affymetrix Drosophila Genome 2 arrays . Three independent RNAi experiments followed by RNA extraction and microarray hybridization were carried out for dBrm , and six independent experiments were carried out for mock-treated cells [28] . The data from these experiments is available at Array Express , E-TABM-169 ( http://www . ebi . ac . uk/microarray-as/aew/ ) . We investigated the effects of dBrm depletion on the relative abundances of alternatively spliced transcripts by mining the E-TABM-169 data and selecting those genes that were represented by more than one probe set in the Drosophila Genome 2 arrays ( 974 genes ) . In many cases , the multiple probe sets targeted different parts of the same transcript , pseudogenes or alternative transcripts derived from alternative promoters of the same gene . We found evidence that SWI/SNF regulates the activity of many gene promoters , as expected ( not shown ) . We could also identify genes for which the multiple probe sets targeted transcripts that had originated by alternative splicing or alternative polyadenylation of a single pre-mRNA . We selected those genes that displayed changed expression levels specific for at least one transcript with p<0 . 02 . Fifteen of these genes showed transcript-specific expression changes in the dBrm-depleted cells ( Table 1 ) . We then used the annotations available at FlyBase ( http://flybase . bio . indiana . edu/ ) to analyze the qualitative differences between the transcripts affected . The transcripts affected show differences in their patterns of alternative , including the use of alternative 3′ slice sites , exon skipping and intron retention ( Figure S4 ) . However , the most striking observation was that the processing of the affected transcripts also involved the alternative use of polyadenlation signals , which suggests that dBrm influences not only the splicing but also the formation of the 3′-end of the transcripts . We validated the microarray results by silencing the expression of dBrm in S2 cells and analysing the expression of four selected genes in which the absence of dBrm affected pre-mRNA processing in different ways , according to the microarray experiments . Mock-treated cells and control cells treated with dsRNA for GFP were analyzed in parallel to assess the specificity of the depletion effects . The levels of dBrm RNA and protein were significantly reduced after 4 days of treatment with dsRNA , as shown by RT-PCR and Western blot , respectively ( Figure 5A–B ) . We designed PCR primers for each of the selected genes in order to amplify specific transcripts and we analyzed the effects of dBrm depletion by RT-PCR . The results of the RT-PCR analyses agreed with the microarray data ( Table 1 , Figures 5C and S5 ) . In summary , depletion of dBrm affects the relative abundances of alternatively spliced and/or alternatively polyadenylated transcripts . Are the effects of SWI/SNF depletion on pre-mRNA processing direct or indirect ? Pre-mRNA splicing often occurs co-transcriptionally [20] , [29] . We thus asked whether dBrm was associated with the gene regions involved in the alternative processing events that were affected by dBrm depletion , and we carried out chromatin immunoprecipitation ( ChIP ) experiments to analyze the association of dBrm with the three genes shown in Figure 5 . ChIP can detect proteins that are bound to the DNA as well as proteins associated with the nascent pre-mRNA [see for example 30] . For each gene analyzed , we used primer-pairs to detect the proximal promoter , the internal region affected by the alternative processing and the 3′-end of the gene ( Figure S6 ) . dBrm associated with all regions of the three genes ( lanes 2 , 6 and 10 in Figure 6 ) . We used an antibody against the C-terminus of the largest subunit of RNA pol-II as a positive control for the ChIP reactions ( lanes 3 , 7 and 11 in Figure 6 ) . An unrelated anti-rabbit antibody was used as a negative control ( lanes 4 , 8 and 12 in Figure 6 ) . Additional controls were carried out by analyzing the association of dBrm with the actin gene , a housekeeping gene whose expression is not regulated by SWI/SNF . The RNA pol-II ( lane 15 ) associated with the actin gene while the dBrm did not ( lane 14 in Figure 6 ) . An intergenic region located far from any annotated genes was devoid of both dBrm and RNA pol-II . We next asked whether dBrm alone or the entire BAP/PBAP complex is responsible for the effects detected at the level of pre-mRNA processing . We mined the data from the E-TABM-169 microarray experiment and asked whether depletion of other SWI/SNF subunits had any effects on the processing of the pre-mRNAs derived from the CG8092 , CG8421 and CG9380 genes . Depletion of either Mor or Snr1 , two SWI/SNF core subunits , induced changes very similar to those induced by dBrm , whereas depletion of the signature subunits Osa , Bap170 or PB gave milder and in many cases non-significant effects ( Figure 7 ) . The effect of Snr1 depletion on the abundances of the CG8421 and CG9380 transcripts was validated by RNAi and RT-qPCR ( Figure S7 ) . These results suggest that dBrm does not regulate pre-mRNA processing alone: it is part of the core SWI/SNF complex . The 15 genes identified above were tested for enrichment of gene ontology ( GO ) terms for biological processes and molecular functions [31] . The expected number of genes associated with a given term by random was compared with the observed number of SWI/SNF-regulated genes that were associated with that particular term using the Fisher's exact test . Several GO terms for biological processes were very significantly enriched , including positive regulation of developmental process ( GO:0051094 , p = 0 . 00003 ) , programmed cell death ( GO:0008219 , p = 0 . 0005 ) and instar larval or pupal morphogenesis ( GO:0048707 , p = 0 . 0008 ) . The test for enrichment of molecular functions also revealed significant associations . Seven of the genes are predicted to code for proteins with catalytic activity . Three of them have phosphatase activity ( GO:0016791 , p = 0 . 0009 ) and three have RNA polymerase II transcription factor activity ( GO:0003702 , p = 0 . 002 ) . And nine out of the fifteen gene products were found to have metal ion binding activity ( GO:0046872 , p = 0 . 00002 ) . In agreement with this finding , a search for conserved protein domains revealed that five of the genes regulated by SWI/SNF post-transcriptionally , including the known transcription factors broad/CG11491 , lola/CG12052 , mod ( mdg4 ) /CG32491 and hr39/CG8676 , code for proteins that contain zinc finger domains . In summary , the tests for ontology enrichments indicate that the genes that are regulated post-transcriptionally by SWI/SNF are primarily enzymes and transcription factors that function in the regulation of postembryonic developmental processes . It is worth mentioning that one of them , broad/CG11491 , is a key regulator of metamorphosis [32] .
The hBrm and hBrg1 proteins are the catalytic subunits of the SWI/SNF chromatin remodeling complexes and much of what is known about their function comes from studies of transcriptional regulation [33] , [34] . SWI/SNF participates in regulatory networks that can result in either the activation or the repression of a gene , depending on the genomic context and the activities of additional co-regulators [35] . One of the functions of SWI/SNF is to remodel the structure of nucleosomes at promoter regions in an ATP-dependent manner [36] , [37] . In some genes , SWI/SNF is associated also with downstream regions of the genes and influences transcription elongation [38] . Recent studies have shown that hBrm and hBrg1 regulate the alternative splicing of several pre-mRNAs in human cells [12] , [39] . The current proposed model ( Figure 8A ) suggests that hBrm acts together with mRNA-binding proteins such as Sam68 or p54nrb to decrease the elongation rate of RNA pol II and to induce the accumulation of RNA pol II at specific positions in the gene . This facilitates the assembly of the splicing machinery at weak splice sites , which favors the inclusion of proximal exons [12] , [39] . Inactivation of the ATPase activity of hBrm does not affect its ability to regulate alternative splicing [12] , which indicates that the mechanism by which hBrm regulates pre-mRNA processing is independent of its nucleosome remodeling activity . We have now studied Brm in two insect model systems , D . melanogaster and C . tentans , and our results show that Brm becomes incorporated into nascent pre-mRNPs during transcription . This conclusion is based on several observations . Firstly , immunofluorescence experiments combined with RNase A digestion show that the association of ctBrm with the polytene chromosomes of C . tentans is partially mediated by RNA . Secondly , immuno-EM reveals that a fraction of ctBrm is associated with the BR pre-mRNPs , not with the chromatin . Thirdly , biochemical fractionation experiments show that Brm is present in the chromosomal RNP fraction in C . tentans , D . melanogaster and H . sapiens . The fact that Brm interacts directly with the nascent pre-mRNP suggests that Brm regulates gene expression post-transcriptionally . An interesting question is whether the post-transcriptional role of Brm is mediated by the Brm protein alone or in complex with other SWI/SNF subunits . Fractionation of nuclear extracts from HeLa cells showed that several SWI/SNF core subunits , Brm , Brg1 , Baf155 , Baf170 and INI1/SNF5 , are associated with nuclear RNPs . This observation suggests that , at least in human cells , the hBrm and hBrg1 proteins that are present in the RNP-associated fraction are part of a SWI/SNF complex . This seems to be the case also in insect cells , as judged by the effects of depletion of individual SWI/SNF subunits on pre-mRNA processing ( see below ) . We have analyzed the expression profiles of S2 cells in which either Brm or other subunits of SWI/SNF have been silenced by RNAi , and we have identified 15 genes that show changes in the relative abundance of alternatively processed transcripts . The number of genes that are affected by dBrm depletion is likely to be underestimated because our study is based on the use of gene expression arrays that do not fully cover all the transcriptome . Our results clearly show that Brm influences the levels of alternatively processed mRNAs in Drosophila cells . Interestingly , the analysis of the expression profiles of S2 cells depleted of other SWI/SNF core subunits revealed effects similar to those induced by Brm depletion . This finding suggests that the role of Brm in pre-mRNA processing is mediated by a core SWI/SNF complex . What is the functional significance of the alternative processing events regulated by SWI/SNF ? For some of the identified genes , the alternative transcripts code for different protein isoforms . For instance , the CG8092 gene encodes two different proteins . The longer isoform , CG8092-PA , contains an AT-hook motif and a zinc finger domain whereas the shorter isoform , CG8092-PB , lacks the zinc finger . Considering that CG8092 is an essential gene in D . melanogaster ( http://flybase . org/reports/FBal0211894 . html ) and that the CG8092 proteins resemble transcription factors , changes in the relative abundance of the CG8092 isoforms are likely to be biologically significant . In other cases , the differences among the alternative transcripts regulated by SWI/SNF lie outside the ORF , in the 3′ UTRs of the mRNAs . The proteins encoded are thus identical , but the stability of the transcripts may differ . A search in miRBase ( http://microrna . sanger . ac . uk ) , a database for microRNA ( miRNA ) data , revealed that several of the genes identified in our study are predicted targets for miRNA regulation and that in most cases the miRNAs are specific for each alternative transcript ( data not shown ) . One example of this is broad/CG11491 , a gene with seven alternative mRNAs with five different 3′ UTRs . Interestingly , all five 3′ UTRs contain predicted miRNA targets [40] . These observations lead us to speculate that in some cases the regulation of alternative processing mediated by SWI/SNF acts in concert with the miRNA pathway to fine-tune the abundances of key proteins with catalytic and/or regulatory activities . Depletion of SWI/SNF does not change the levels of all the mRNAs that have originated from a pre-mRNA in the 15 genes identified , but only a few . Indeed , in most cases only one mRNA is affected , which indicates that the step that is affected is not only the synthesis of the pre-mRNA but its processing into alternative mRNAs . SWI/SNF plays an important role in the transcription of many genes , and it may be that the alterations of pre-mRNA processing that we have observed are a consequence of alterations in the synthesis of specific pre-mRNA processing factors . However , we have shown by ChIP that dBrm is physically associated with both the proximal promoter and downstream sequences of the genes affected . We have also shown by immuno-EM of the BR genes of C . tentans that ctBrm is associated with nascent pre-mRNPs . These observations strongly suggest that Brm acts directly at the mRNA level . Our results indicate that dBrm affects both alternative splicing and alternative polyadenylation sites . Many of the alternative processing events regulated by SWI/SNF involve the use of mutually exclusive splicing and polyadenylation sites . The use of the proximal site is favored by dBrm in some of the genes , as would be expected if dBrm acts by reducing the elongation rate of the Pol-II at certain positions , as has been proposed for the regulation of the CD44 pre-mRNA in human cells [12] . However , depletion of dBrm has the opposite effect in other cases ( such as CG18251 , CG9380 , CG3665 ) , which is difficult to reconcile with a model of kinetic regulation based on modulation of the Pol-II elongation rate . This observation , and the finding that a significant fraction of Brm is associated with nascent pre-mRNPs , lead us to propose that dBrm regulates pre-mRNA processing in a more direct manner . Several mechanisms can be envisioned by which Brm , being part of the pre-mRNP complex , could influence the usage of alternative splicing or polyadenylation sites ( Figure 8B ) . In one possible scenario , SWI/SNF could work as an RNP remodeling factor to modulate interactions between specific processing factors and their target RNA sequences . However , this possibility is unlikely because experiments in human cells suggest that the role of hBrm in pre-mRNA processing does not require a functional ATPase domain [12] . Alternatively , SWI/SNF could influence the structure of the pre-mRNP . For instance , SWI/SNF could recruit pre-mRNA processing factors to the nascent transcript , or prevent the interactions of processing factors with their target RNA sequences . The interaction of Brm with each pre-mRNP is likely to depend on the sequence of the transcript and/or the specific combination of proteins that forms the pre-mRNP . As a consequence , the action of Brm and the specific outcome of the processing reactions will depend in each case on the specific features of the pre-mRNP . The same type of context-dependent mechanism has been proposed to explain the complex function of SWI/SNF in transcription regulation [35 and references therein] . SWI/SNF acts as a co-activator in the transcription of certain genes , but represses the transcription of certain other genes [for example 41] , and these two opposite effects are mediated by interactions with different types of co-regulators . In a similar manner , SWI/SNF can either repress or activate the choice of splice sites and/or polyadenylation sites in a gene-specific manner . The role of SWI/SNF in pre-mRNA processing affects a specific subset of pre-mRNAs [12 and our present results] . Ontology analysis revealed that many of the transcripts regulated post-transcriptionally by SWI/SNF in D . melanogaster code for proteins that are implicated in postembryonic developmental processes . One of them , broad , is a transcription factor that plays a central role in the cross-talk between ecdysone and juvenile hormone , the two hormones that coordinate insect growth and development [reviewed in 42] Dubrovsky 2005 . This is particularly interesting , since hBrm and hBrg1 have also been identified as key regulators of growth control and differentiation in mammals [43] , [44] . hBrm and hBrg1 are differentially expressed during development , and their expression is altered in cancer cells , which leads to deregulation of genetic programs [reviewed in 45] . In summary , SWI/SNF appears to act both transcriptionally and post-transcriptionally to fine-tune the expression of genes with key regulatory functions in development . In this way , SWI/SNF can regulate gene expression at two levels by determining not only the amount of mRNA synthesized from a given promoter but also the type of alternative transcript produced . Acting at the pre-mRNA processing level , SWI/SNF can rapidly modulate the abundance and activity of the resulting protein products by acting on genes that are already active .
Chironomus tentans were cultured as described by Meyer et al . [46] . The salivary glands used for study were isolated from 4th instar larvae . C . tentans tissue culture cells were grown in ZO medium at 24°C as described by Wyss et al . [47] . D . melanogaster S2 cells were cultured at 28°C in Schneider's medium ( Invitrogen ) . Degenerate primers for nested PCR were designed based on conserved residues in the C-terminal part of D . melanogaster Brm ( CG5942 ) and Anopheles gambiae Brm ( AGAP010462 ) . The sequences of the primers are provided in the Supplementary Materials and Methods . A PCR product of about 750 bp was amplified from a total cDNA preparation made from C . tentans tissue culture cells . The sequence of the PCR product encoded a partial protein corresponding to amino acids 1252–1455 in dBrm . The PCR product was cloned into pET21b ( Novagen ) , expressed in BL21 E . coli cells ( Novagen ) and used to immunize rabbits . The antibody against the C-terminal part of ctBrm was raised in rabbit following standard procedures ( AgriSera , Sweden ) . The anti-rat Brg1 antibody was raised and characterized by Östlund Farrants et al . [24] . The anti-dBrm antibody was raised and characterized by Zraly et al . [25] . The antibodies against hBrm , Baf47/INI1 , Baf180 , Baf155 , Baf170 and Sap155 were characterized by Sif et al . [48] , de la Serna et al . [49] , Nicolas et al . [50] , Shanahan et al . [51] and Will et al . [52] , respectively . The anti-Baf170 was from Santa-Cruz ( sc-10757 ) . The mAbs 10:3G1 and 2E4 against Hrp36 and Hrp45 , respectively , have previously been characterized [53] , [54] . The Y12 antibody against the Sm epitope of snRNP core proteins was characterized by Lerner et al . [55] . The rabbit anti-Hrp59 was the Y38 antibody raised by Falk et al . [56] . The anti-Pol-II antibody was purchased from Abcam ( Ab5408 ) . The anti-TBP was from Santa-Cruz Biotechnology ( sc-204 ) . FITC-conjugated , Texas Red-conjugated and gold-conjugated secondary antibodies were from Jackson ImmunoResearch Laboratories . The secondary antibodies conjugated to alkaline phosphatase and horseradish peroxidase were from DakoCytomation . Salivary glands were pre-fixed with 2% formaldehyde in TKM buffer ( 10 mM triethanolamine-HCl , 100 mM KCl and 1 mM MgCl2 ) , permeabilized and disrupted by pipetting in 0 . 25% NP40 in TKM . Individual chromosomes were isolated and fixed with 4% parafomaldehyde in TKM . The chromosomes were then blocked in 2% bovine serum albumin ( BSA ) in TKM and incubated with antibodies following standard procedures . The secondary antibodies were conjugated to FITC or Texas Red . The immunostained chromosomes were mounted in Vectashield ( Vector Laborarories ) . Preparations were analyzed and images were taken with a laser scanning microscope ( model LSM 510; Carl Zeiss MicroImaging , Inc . ) equiped with PlanApochromat objectives 40×/1 . 0 oil and 63×/1 . 4 oil , using immersion oil Immersol 518F ( Carl Zeiss MicroImaging , Inc . ) . The optical sections were approximately 1 µm thick . Photoshop software ( Adobe ) was used for the preparation of composite images and for adjustment of intensity and contrast . Salivary glands were prefixed and permeabilized , and the polytene chromosomes were isolated by pipetting in the same way as those intended to be used in immunofluorescence experiments . The isolated chromosomes were fixed with freshly prepared 4% paraformaldehyde in TKM . The chromosome preparations were blocked in 2% BSA in TKM for 30 min , incubated with primary antibody , washed and incubated with an anti-rabbit IgG conjugated to 6-nm gold particles . The control chromosomes were incubated with either secondary antibody only or with a pre-immune serum . The stained chromosomes were fixed with 2% glutaraldehyde and embedded in Agar 100 . Thin sections ( 70 nm ) of plastic-embedded chromosomes were stained with uranyl acetate and examined in a FEI 120 kV TECNAI electron microscope . Images were recorded using a Gatan US 1000P CCD camera . For quantitative purposes , the BR genes were photographed at random areas and the numbers of gold markers in the proximal , middle , and distal segments of the BR genes were counted . The specificity of the immuno-EM results was supported by negative control preparations that were processed in parallel and incubated with either the pre-immune serum or with only secondary antibody . The labeling density in the negative controls was calculated and was found to be between 18 . 5 and 7 . 5% , respectively . C . tentans tissue culture cells , Drosophila S2 cells or human HeLa cells were homogenized in PBS containing 0 . 2% NP-40 . The homogenate was centrifuged at 1500 g for 10 min at 4°C . The pellet containing the nuclei was resuspended in PBS , sonicated and centrifuged at 16 , 300 g for 10 min at 4°C . The resulting supernatant was the soluble nuclear extract . The pellet was resupended in PBS , digested with RNase A ( 100 µg/ml ) and centrifuged at 16 , 300 g for 10 min at 4°C . The supernatant was the chromosomal RNP fraction and contained proteins that were retained in the pellet through RNA-dependent interactions . For the experiment shown in Figure 4G , the RNase A digestion was allowed to run for 15 min at either 4°C or 25°C . Immunoprecipitation experiments were carried out following standard procedures . Soluble nuclear extracts and chromosomal RNP extracts were prepared as described above , supplemented with 0 . 1% NP40 and used as input . The bound proteins were eluted , precipitated with acetone and analyzed by SDS-PAGE and Western blotting . Protein extracts were separated by SDS-PAGE and transferred to polyvinylidenefluoride membranes ( Millipore ) following standard procedures . The NBT/BCIP system was used for detecting secondary antibodies conjugated to alkaline phosphatase . The ECL system ( GE Healthcare ) was used for the chemiluminiscent detection of horseradish peroxidase . Chromatin was prepared from S2 cells after cross-linking with 2% formaldehyde . The chromatin was sheared by sonication to a DNA size of 250–1000 bp and pre-cleared . Chromatin fragments were precipitated with antibodies against either rBrg1 ( Ab1 ) or Pol-II ( Abcam ) using protein A/G-Sepharose beads ( 50% of each ) . The precipitated DNA fragments were purified and amplified by PCR using primers for the CG8092 , CG8421 and CG9380 genes . Actin 5C ( CG4027 ) was used as a control . The PCR conditions were optimized to avoid saturation . See The microarray data was extracted from Array Express , E-TABM 169 ( http://www . ebi . ac . uk/microarray-as/aew/ ) . Drosophila Genome 2 . 0 Arrays ( Affymetrix ) were hybridized with total RNA purified from Drosophila S2 cells treated with dsRNA corresponding to dBrm or to other subunits of SWI/SNF [28] . dsRNAs against dBrm and GFP were prepared by in vitro transcription from PCR products with T7 promoters on both ends of the amplimers , using the Megascript RNAi kit ( Ambion ) . The sequences of the PCR primers are provided in the Supplementary Materials and Methods . The RNAi treatment was performed as described by Clemens et al . [57] . In brief , 20 µg of dsRNA was applied to S2 cells and the cells were harvested after 48 h . Total RNA from S2 cells was extracted , reverse transcribed and used as a template for PCR reactions using primers specific for selected transcripts . Quantitative real-time PCR was carried out in an ABI7000 system using SYBR Green ( Applied Biosystems ) . The RNAi experiments were repeated three times to confirm the reproducibility of the observations . A detailed description of the Materials and Methods , including primer sequences , are provided as Supporting Information ( Text S1 ) . Seven Supporting Figures are also provided ( Figures S1 , S2 , S3 , S4 , S5 , S6 , and S7 ) . | Genetic programs in multicellular organisms often involve different levels of regulation . The expression of many genes is regulated by factors that remodel the structure of the chromatin at the promoter . SWI/SNF is one such factor , and it is highly conserved in eukaryotes . Studies in human cells suggest that Brahma , the catalytic subunit of SWI/SNF , regulates the processing of precursor mRNAs ( pre-mRNAs ) . We have studied Brahma in two insect model systems to further elucidate the mechanisms by which SWI/SNF regulates gene expression . We show that depletion of SWI/SNF subunits changes the relative abundances of alternative transcripts from a subset of pre-mRNAs that code for proteins that regulate the postembryonic development of the flies . We also show that a fraction of Brahma is not associated with chromatin but with nascent pre-mRNPs—both in insects and mammals—which suggests that SWI/SNF acts at the RNA level to regulate pre-mRNA processing . These findings illustrate the dual role of a chromatin remodelling factor; SWI/SNF acts both at the transcriptional level and post-transcriptionally to regulate not only the amount of mRNA synthesized from a given promoter but also the type of alternative transcript produced . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genetics",
"and",
"genomics/epigenetics",
"cell",
"biology/gene",
"expression",
"genetics",
"and",
"genomics/gene",
"expression"
] | 2009 | SWI/SNF Associates with Nascent Pre-mRNPs and Regulates Alternative Pre-mRNA Processing |
Apart from a single report , the last publication of cutaneous leishmaniasis ( CL ) in Mali dates back more than 20 years . The absence of information on the current status of CL in Mali led us to conduct a cohort study in Kemena and Sougoula , two villages in Central Mali from which cases of CL have been recently diagnosed by Mali's reference dermatology center in Bamako . In May 2006 , we determined the baseline prevalence of Leishmania infection in the two villages using the leishmanin skin test ( LST ) . LST-negative individuals were then re-tested over two consecutive years to estimate the annual incidence of Leishmania infection . The prevalence of Leishmania infection was significantly higher in Kemena than in Sougoula ( 45 . 4% vs . 19 . 9%; OR: 3 . 36 , CI: 2 . 66–4 . 18 ) . The annual incidence of Leishmania infection was also significantly higher in Kemena ( 18 . 5% and 17% for 2007 and 2008 , respectively ) than in Sougoula ( 5 . 7% for both years ) . These data demonstrate that the risk of Leishmania infection was stable in both villages and confirm the initial observation of a significantly higher risk of infection in Kemena ( OR: 3 . 78; CI: 2 . 45–6 . 18 in 2007; and OR: 3 . 36; CI: 1 . 95–5 . 8 in 2008; P<0 . 005 ) . The absence of spatial clustering of LST-positive individuals in both villages indicated that transmission may be occurring anywhere within the villages . Although Kemena and Sougoula are only 5 km apart and share epidemiologic characteristics such as stable transmission and random distribution of LST-positive individuals , they differ markedly in the prevalence and annual incidence of Leishmania infection . Here we establish ongoing transmission of Leishmania in Kemena and Sougoula , Central Mali , and are currently investigating the underlying factors that may be responsible for the discrepant infection rates we observed between them . ClinicalTrials . gov NCT00344084
Leishmaniasis is endemic in 88 countries with an estimated global burden of approximately two million disability-adjusted life years [1] . No vaccine is currently available for leishmaniasis , and chemotherapy—if accessible—is costly and has considerable side effects [2] , [3] . Leishmaniasis is a vector-borne disease transmitted to the host via the bite of a Leishmania-infected phlebotomine sand fly . Depending on the Leishmania species , humans can develop visceral or cutaneous forms of the disease . In the Old World , cutaneous leishmaniasis ( CL ) is a self-healing disease characterized by skin lesions that can develop into unsightly scars , typically on the face and extremities . CL was first reported from Mali , West Africa , in 1958 [4] . Since then there have been sporadic reports of CL in Mali [5]–[13] , with the majority of cases originating from 5 of 42 districts ( Kayes , Bafoulabe , Segou , Gao , and Nara ) [6] . Leishmania major was first identified as the causative agent of CL in Mali following its isolation from two skin lesions , one from a tourist and the other from a permanent resident of Mali [10] , [11] . Both individuals were likely infected in the Mopti region of Central Mali . The parasite isolates were characterized by isoenzyme analysis as L . major MON-26 and MON-25 [10] , [11] . Recently , two additional zymodemes of L . major , MON-74 and MON-117 , were reported from Mali[13] . Keita et al . [12] reported on CL cases referred to the Centre National d'Appui à la Lutte contre la Maladie ( CNAM ) , the only dermatological reference center in Mali , from 1997 to 2001 , showing that CL remains widely distributed in the country . Thirty of the 251 CL cases ( 12% ) reported in this study originated from the Segou district of Mali [12] . The present study demonstrates the active transmission of Leishmania infection in two neighboring villages ( Kemena and Sougoula ) in the Segou district of Central Mali .
The study protocol ( NCT00344084 , available at http://www . clinicaltrial . gov/ ) was approved by the Institutional Review Boards of the National Institute of Allergy and Infectious Diseases ( USA ) and the University of Bamako ( Mali ) . The study was externally monitored for protocol agreement , data integrity , and protection of human subjects . The study was carried out in two villages , Kemena ( 13°7′30 . 50″N , 6°54′46 . 30″W ) and Sougoula ( 13°5′24 . 97″N , 6°53′11 . 86″W ) , located 180 km northeast of Bamako , the capital of Mali ( Figure 1 ) . The villages are 5 km apart and share the same topography and climate . The villages are situated on a flat plain , and the natural vegetation is characterized by savannah grasses and shrubs . The climate is subtropical to arid and is hot and dry from February to June ( 27–34°C ) , mild , rainy , and humid from June to November ( 27–29°F ) , and cool and dry from November to February ( 25–28°C ) . The population consists mainly of farmers , and the basic food crops are millet , sorghum , peanuts , and peas . Cattle , goats , and sheep are maintained inside the villages . Both villages have similar multi-ethnic societies , represented by a Bambara majority and Peulh and Sarakole minorities . Prior to the study , a census was conducted of all permanent residents in the two villages . The position of every house within each village was mapped using GPS and the houses were numbered sequentially . Local guides invited families from adjacent houses to the local school house located at the perimeter of each village to insure full coverage of the population . Families were welcomed at the school and provided with a protocol identification card comprising their name , location and a color picture . Thereafter , they received an explanation of the aims of the study and those agreeing to participate were asked to sign a consent form . Parents or guardians were asked to sign on behalf of participating children . All children younger than one year old were excluded from the study . The LST ( leishmanin , LOT 124; Institute Pasteur , Tehran ) was performed at the beginning of the study and at approximately 1 and 2 years later . Briefly , 0 . 1 ml of leishmanin was injected intradermally in the left forearm . Readings were taken 48 to 72 hours after the injection using a ball point pen to determine the size of the induration . Measurements with a diameter greater than 5 mm were considered positive [14] . Only individuals who were negative during the preceding LST survey were re-tested the following year . Results were analyzed using the Statistical Package for the Social Sciences ( SPSS , Chicago , IL , USA ) . Fisher's exact test was used to assess the association between infection and demographic variables . Age means between villages were compared using an independent samples t-test . One-way ANOVA with Bonferroni's multiple comparison test was applied to evaluate the difference in the mean size of the LST reaction between the two study sites . Values below P<0 . 05 were considered statistically significant . To assess geographic clustering of cases , coordinates of households were determined using a GPS receiver ( Trimble , Sunnyvale , CA , USA ) . The coordinates were superimposed onto satellite images acquired from DigitalGlobe Incorporated ( images were taken May 2006 by the QuickBird satellite ) . Spatial-scan statistical software ( SaTScan™ v8 . 0 software ) was used to map the location of the LST-positive individuals . The performed scans were purely spatial and accordingly were tested under the Poisson model [15] . The spatial software calculates a P value and log likelihood ratio to determine the statistical significance of any detected clusters .
We enrolled a total of 1530 individuals , 663 from Kemena and 867 from Sougoula . The mean age of all study participants was 20 years ( SD±19 ) and included 1- to 92-year-old individuals ( Table 1 ) . Seventy-five percent of the population was <30 years old and , nearly 53% were female . No differences in the age mean or sex distribution were found between the two villages ( Table 1 ) . In May 2006 , we determined the prevalence of Leishmania infection by conducting a baseline LST survey of Kemena and Sougoula residents . The proportions of LST-positive ( LST+ ) individuals differed by more than twofold between the villages: 45 . 4% in Kemena and 19 . 9% in Sougoula ( Table 2 ) . There was an age-associated increase in the frequency of LST+ individuals in both villages , although it was more pronounced in Kemena ( Figure 2 ) . Analysis of these data suggested that residents of Kemena have a more than threefold increased risk of Leishmania infection compared with residents of Sougoula ( OR: 3 . 36; CI: 2 . 66–4 . 18; P<0 . 0005 ) . Furthermore , the population of Kemena presented a significantly higher prevalence of infection ( P<0 . 05 ) among all age groups except children less than 3 years old ( Figure 2 ) . Importantly , there was no statistical significance in the mean size of the LST reaction in inhabitants of Kemena compared to Sougoula ( Figure 3 ) . To estimate the annual incidence of Leishmania infection in the two study villages , we re-tested the LST-negative ( LST− ) individuals 1 and 2 years later . The annual incidence of Leishmania infection in both villages combined was 9 . 9% ( 85/853 ) in 2007 and 8 . 9% ( 59/659 ) in 2008 ( Table 2 ) . In both years , we observed a higher incidence of Leishmania infection in Kemena than in Sougoula ( 18 . 5% vs . 5 . 7% in 2007 and 17 . 0% vs . 5 . 7% in 2008; P<0 . 001 ) ( Table 2 ) . Analysis of these data indicated that the risk of Leishmania infection was significantly higher in Kemena than in Sougoula during 2007 ( OR: 3 . 78; CI: 2 . 45–6 . 18; P<0 . 005 ) and 2008 ( OR: 3 . 36; CI: 1 . 95–5 . 80; P<0 . 005 ) . The consistent incidence rates provide evidence that the rate of Leishmania transmission was stable in both villages over the 2-year study period . As was observed for prevalence , the mean size of the LST reaction was not significantly different when comparing the populations of Kemena and Sougoula for both incidence surveys ( Figure 3 ) . Kemena and Sougoula households each occupy an area of roughly 0 . 4 km2 and radiate from a central mosque in a circular pattern . Figure 4 shows a map of the number of LST+ individuals by household for Kemena ( Figure 4A ) and Sougoula ( Figure 4B ) for the combined incidence data of 2007 and 2008 . Yellow circles represent population by household , and red circles represent LST+ individuals . Spatial analysis determined that there is no geographical clustering of LST+ individuals within either village .
Apart from a single publication [12] , there has been no further report of CL in Mali the past 20 years . In this study , we used the LST to determine infection by Leishmania parasites [16]–[19] . A positive LST reaction relies on an in vivo cellular immune response specific to parasite molecules and is a measure of exposure to Leishmania parasites irrespective of disease presentation . Based on two annual incidence rates , our data suggest that transmission of Leishmania is active in the villages of Kemena and Sougoula . The high prevalence of LST+ individuals in these two villages was similar to prevalence data reported previously from other parts of the country [6] . This may be a reflection of the stable endemicity of Leishmania transmission in Mali over time . In our study villages , we have shown an increase in the number of LST+ individuals with age , suggesting that this is not a recent focus . The presence of LST− individuals in the older age group ( >41 years ) may indicate a moderate force of infection . Our data show that most transmission is occurring after the age of 3 years with males and females equally affected , suggesting that transmission is not related to occupational or other gender-associated tasks . The statistically significant difference in prevalence of infection between the two villages was unexpected , because they are located only 5 km apart and share similar geographical and human behavioral features . Sougoula was founded more than 300 years ago during the Bambara Empire [20] , [21] and Kemena was established a couple of decades later . Both villages retain urbanization patterns from ancient times , and there have been no substantial differences between the two villages that could account for the increased risk of Leishmania infection observed in Kemena compared with Sougoula . The higher risk of infection in Kemena was maintained in two subsequent surveys , with very similar incidence rates over time . These differences may be due to a higher concentration of rodent reservoirs and , in turn , a higher infection rate of vector sand flies in Kemena . Indeed , the uniform distribution of LST+ individuals throughout the villages ( Figure 4 ) may reflect peridomestic transmission , with reservoir and vector populations present either in or around houses . We have previously established the presence of P . duboscqi , an incriminated vector of CL in sub-Saharan Africa [23]–[26] , in Kemena [27] . Additionally , examination of households in the villages revealed the presence of active rodent burrows . This may be of consequence , as rodents are the established reservoirs of CL due to L . major [22] . Incrimination of rodents in the village as reservoirs and finding infected sand flies inside the villages will support the hypothesis that transmission is peridomestic . Moreover , comparison of the density and infection rates of reservoir and vector populations may help elucidate the cause of the observed discrepancy in the incidence of Leishmania infection in the two villages . It is important to note that the mean size of the LST reaction was similar for the two villages in all the conducted surveys ( Figure 3 ) . This suggests that the observed difference in LST positivity was not the result of the non-responsiveness of either population to leishmanin . In conclusion , the results of this study highlight the fact that there is active transmission of Leishmania in Kemena and Sougoula , located in Central Mali . Our findings show a significant discrepancy in CL prevalence and incidence . Studies are under way to elucidate the cause of these differences . | Leishmaniasis is a vector-borne disease transmitted to humans by the bite of an infected sand fly . Leishmaniasis is present in more than 88 countries and affects more than 12 million people . Depending on the species of Leishmania , the host can develop cutaneous leishmaniasis ( CL ) , which is characterized by skin ulcers in uncovered parts of the body or a more severe form , visceral leishmaniasis , which affects the liver and spleen and is fatal if not treated . This study aims to establish the past and present infection with Leishmania parasites in two villages where recent cases have been diagnosed by the dermatology center ( CNAM ) in Bamako . This was achieved using a Leishmania-specific skin test that was administered annually to permanent residents of Kemena and Sougoula villages from 2006 to 2008 . The results show that transmission of Leishmania is active and stable in these two villages . Moreover , despite sharing similar cultural and environmental features , the individuals from Kemena presented three times the risk of Leishmania infection compared with those from Sougoula . Our findings raise awareness of the continued presence of CL in Mali . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/protozoal",
"infections",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases"
] | 2009 | Discrepant Prevalence and Incidence of Leishmania Infection between Two Neighboring Villages in Central Mali Based on Leishmanin Skin Test Surveys |
As part of a study to investigate drivers of dengue virus ( DENV ) transmission dynamics , this qualitative study explored whether DENV-infected residents of Iquitos , Peru , considered it acceptable ( 1 ) to participate in direct mosquito feeding experiments ( lab-reared Aedes aegypti mosquitoes fed directly on human volunteers ) and ( 2 ) to provide blood meals indirectly ( Ae . aegypti fed on blood drawn from participants by venipuncture ) . Twelve focus group discussions ( FGDs; 94 participants: 82 females and 12 males ) were conducted in January 2014 to explore six themes: ( 1 ) concerns and preferences regarding direct mosquito feeds and blood draws , ( 2 ) comprehension of and misconceptions about study procedures , ( 3 ) motivating factors for participation , ( 4 ) acceptability of children’s participation , ( 5 ) willingness to provide multiple samples over several days , and ( 6 ) preference for direct feedings in homes versus the study laboratory . Results of FGDs , including one with 5 of 53 past direct mosquito feed participants , indicated that mosquito feeding procedures are acceptable to Iquitos residents when they are provided with information and a few key messages are properly reinforced . FGD participants’ concerns focused primarily on safety issues rather than discomfort associated with mosquito bites . A video explaining the study dramatically increased comprehension of the study procedures . The majority of participants expressed a preference for mosquito feeding over venipuncture . Adults supported child participation if the children themselves assented . For most participants , home feedings were preferred over those in a laboratory . A major impetus for participation was the idea that results would contribute to an improved understanding of DENV transmission in Iquitos . Findings from our study will support future large-scale studies that employ direct mosquito feeding , a low-risk , non-invasive procedure that is experimentally superior to artificial mosquito feeding methods .
Vector competence studies on the intrinsic ability of mosquitoes to transmit human pathogens have increased in importance and frequency with the emergence of epidemic Aedes aegypti-borne viruses such as dengue ( DENV ) , chikungunya ( CHIKV ) , Zika ( ZIKV ) , and yellow fever ( YFV ) viruses [1] . Historical methods to determine vector competence had important experimental constraints that limited researchers’ ability to extrapolate to natural transmission and to understand the significance of data from prospective epidemiological studies for humans . Limitations of laboratory-based vector competence studies include using artificial infectious blood meals that are not the same as feeding directly on a human host and laboratory-reared mosquitoes from colonies that are many generations removed from field populations and do not represent vector competence of wild mosquitoes [2] . Artificial infectious blood meals , often composed of cultured virus mixed with animal blood and presented to mosquitoes across a skin-simulating membrane , use virus passaged in cell culture , which can select for viruses that are not maintained in nature [3] , altering the ability of a virus to infect and/or replicate in a mosquito [4] . The common use of defibrinated blood in artificial blood meals can similarly change the distribution of virus in a mosquito midgut , resulting in a systematic misrepresentation of experimentally exposed mosquito infection rates [5] . The most realistic way to overcome these limitations is to allow uninfected mosquitoes to take a blood meal directly from a naturally infected human volunteer or , alternatively , to imbibe blood from an artificial feeding apparatus that was drawn from an infected person . These kinds of experiments present logistical and ethical challenges . We know of no published recommendations to guide Institutional Review Boards ( IRBs ) in evaluating the use of human participants in these procedures [6] . Researchers in several studies explored interactions between malaria parasites ( Plasmodium spp . ) and anopheline mosquitoes by feeding laboratory-reared mosquitoes on people with active malaria infections , alone or in comparison with indirect feeding ( infectious blood drawn from a participant and delivered by artificial feeder ) [7–9] . The fact that treatment is available for malaria makes these procedures more acceptable to IRBs than when treatment is not available , as is the case for arboviral diseases . Since publication of the Belmont Report in 1978 [10] , which marked the initiation of IRB review of human use protocols , we are aware of only three research groups that have carried out experiments of direct mosquito feeding on human subjects naturally infected with DENV [11–13] . All of these were conducted in Southeast Asia . In Vietnam , 407 direct mosquito feedings were carried out on 208 hospitalized dengue patients , ranging from 19–30 years of age , with no adverse events reported [12] . In Singapore , direct feeding was completed on 26 hospitalized adults [13] . In Cambodia , 164 direct mosquito feedings were carried on household contacts of known dengue cases . These participants had not received laboratory results about their infection status at the time of feeding; 89% of the participants were less than 16 years of age [11] . In late 2010 , our research group initiated the process to obtain IRB approval for a pilot study designed to compare direct with indirect mosquito feeding methods using blood from naturally infected study subjects recruited from our ongoing community- and clinic-based febrile surveillance protocols . Our initial objective was to obtain preliminary data for a large-scale research program that would resolve the long-standing enigma about the contribution of people with inapparent and mild symptomatic infections to DENV transmission dynamics . We aimed to determine whether indirect methods could be used in lieu of direct mosquito feeding experiments . Approval to carry out our study was granted in May 2011 with multiple IRB stipulations , including close monitoring [34] . We enrolled our first participants during September 2012 , and after a year of participant interactions realized that direct mosquito feedings were potentially acceptable on a large scale . As we continued to enroll participants in the companion vector competence study , we requested IRB permission to ( 1 ) expand inclusion criteria to younger participants , ( 2 ) move mosquito feedings from the laboratory to participants’ homes , and ( 3 ) conduct focus group discussions ( FGDs ) on the acceptability of direct mosquito feeding experiments . Here , we present results from 12 FGDs conducted during January 2014 . Our goals were to ( 1 ) identify community concerns and misconceptions associated with the direct mosquito feeding procedures , ( 2 ) identify key messages to ensure comprehension of our study , ( 3 ) assess willingness to participate in study protocols requiring direct mosquito feeds and multiple blood draws over the course of a single dengue infection , ( 4 ) determine the acceptability of allowing young children to participate in mosquito feeding experiments , and ( 5 ) determine the acceptability of conducting direct mosquito feedings in the participants home environment rather than the study laboratory . Our long-term objective was to use community-derived opinions to inform subsequent IRB applications and to finalize protocols for planned , follow-up larger-scale vector competence studies .
The study protocol was approved by the U . S . Naval Medical Research Unit No . 6 ( Protocol #NAMRU6 . 2011 . 0002 ) Institutional Review Board , which includes Peruvian representation and complies with US Federal and Peruvian regulations governing the protection of human subjects . IRB authorization agreements were established between U . S . Naval Medical Research Unit No . 6 , University of California , Davis , and Institute Pasteur . The protocol was reviewed and approved by the Loreto Regional Health Department ( LRHD ) , which oversees health research in Iquitos . Consent without written documentation ( verbal ) was approved by the NAMRU-6 IRB so that no names were recorded or stored by the investigator . This study took place in Iquitos , Peru , located in the northeastern Amazon basin , where the human population is approaching 400 , 000 inhabitants [14] and divided into four districts: Maynas , Punchana , Belen , and San Juan Bautista . Detailed descriptions of the city , its Aedes aegypti population , and local DENV transmission have been published elsewhere [15–23] . This region is geographically isolated; it can only be reached by boat or airplane . The main industries in this region are small commercial entreprises , and of extractive ( logging , mining ) or agricultural nature . Iquitos has experienced rapid urbanization in the past three decades [14] , from the neighborhoods around the city center/commercial zones in the districts of Maynas and Punchana to areas on the river ( Belen and parts of Punchana ) and to the South ( San Juan Bautista ) . The recruitment neighborhoods described in the next section were located in the more developed and central neighborhoods of Maynas and Punchana , which are relatively homogenous , with a patchwork of households ranging from wood structures with dirt floors to brick and concrete and ceramic floors . The Peruvian Statistics and Information Institute states that about 30% of the urban jungle population of Peru has at least one unmet basic need , 18 . 2% live in poverty , and 3% of the population in extreme poverty [24] . Evidence of extreme wealth is not observed in Iquitos , and luxury items such as air conditioners are rarely observed . Other indicators for the Maynas and Punchana districts include that 93% of structures are individual houses ( row houses with shared walls ) , 82–38% have corregated metal roofs , 90–97% have electricity , and literacy rates are 88–92% [25] . Purposive sampling , a non-probabilistic sampling method commonly used in qualitative research [26] , was used to recruit focus group participants from neighborhoods in the Amazonian city of Iquitos , Peru . We carried out 12 FGDs with a total of 94 residents over a single week during January 2014 . To facilitate recruitment and transportation of individuals to the NAMRU-6 conference room , 2–3 residents per block within 5–6 contiguous blocks were recruited in person , door-to-door , 2–3 hours before initiation of each FGD ( for a total of 7 FGDs and 49 participants ) from three neighborhoods where prospective cohort studies have been ongoing since 2007 [19 , 22 , 27] , and two neighborhoods ( for a total of 4 FGDs and 40 participants ) where no dengue studies have been conducted by our group since 2007 . All five neighborhoods were located in the districts of Maynas and Punchana in the center of the city , representing more developed and economically stable neighborhoods than observed in neighborhoods located in the far north and south and river edges of the city . The neighborhoods without ongoing research studies were either next to or within a few kilometers of the neighborhoods with ongoing research . For context , neighborhoods with ongoing prospective studies were visited by our staff approximately three times per week to identify people with febrile illness . Our experience in Iquitos is that women usually make the health-related decisions for the family , including children , hence we intended to have a larger representation of women when requesting household participation in FGDs . One FGD was conducted with previous mosquito feeding participants ( n = 3 ) and their mothers if the participants were minors ( n = 2 ) . At this time , the Asian-American genotype of DENV serotype 2 was circulating and Zika virus had not been detected in the city [28] . A Peruvian social scientist with over a decade of work experience conducting FGDs in Iquitos ( VAPS ) facilitated the FGDs in Spanish . A Spanish-speaking expert in the mosquito feeding procedures ( ACM ) was present in all of the FGDs to answer technical questions . Two research team members took detailed notes of the discussion and two more assisted in recording ideas and thoughts on large sheets of paper ( that all could see ) and in role-play exercises . Before each FGD , an IRB-approved consent form was read and participants and parents of minor participants provided verbal consent to participate in the study and be audiotaped . We developed and applied an FGD guide to ensure we covered the same topics in each FGD . Each FGD began with a brief introduction and an explanation that we wanted to learn more about dengue and how DENV is transmitted from infected people to mosquitoes , followed by a brief description of the direct and indirect mosquito feeding methods . Role-playing exercises were used to simulate an invitation to participate in the project . Initially , we also intended to evaluate a mosquito feeding consent video that we had developed for the pilot project ( see description below , S1 Video , Fig 1 ) , but , after our first two FGDs in which the video was shown at the end for discussion and feedback , we decided to show this video immediately after the brief introduction because it became apparent that it was an effective tool for introducing the project and procedures to the group . This also mirrored how the study was presented to potential participants in the field . To assess people’s understanding of the direct mosquito feeding procedures , after the brief discussion and viewing of the consent video , we asked participants to describe the purpose of this study , the procedures , and their initial reactions . To ensure that these hypothetical questions felt real to the FGD participants , one member of our research team ( ACM ) had mosquitoes feeding on her during the FGD to show them what the mosquitoes in a container looked like , and what her legs looked like post-feeding . We noted the types of questions and discussion among the participants . We wanted to assess how well people understood the purpose and procedures , identify concerns participants had about the process , assess whether they would participate or allow their children to do so , and identify any additional concerns regarding their children’s participation . Groups were asked to state their preferences for providing a venous blood sample to feed to mosquitoes indirectly or feeding the mosquitoes directly on their arms or legs , and discuss the pros and cons of each method . Individuals were asked about how many consecutive days they would be willing to provide both venous blood samples and directly feed mosquitoes . Regarding the mosquito feeding consent video , our primary objective was to clearly show the mosquito feeding procedures to potential participants . We wanted people to see a full cup of mosquitoes biting an arm and becoming engorged with blood . The video includes the following: ( 1 ) the title and purpose of the project , ( 2 ) a brief description of the DENV transmission cycle and how one can become infected , ( 3 ) an explanation of what is expected from a participant in the project ( mosquito feeding and provision of blood samples ) , ( 4 ) the study’s risks and benefits , ( 5 ) a visual of mosquito feeding procedures ( i . e . , mosquitoes feeding on an arm and mosquitoes feeding on blood through an artificial feeder ) , ( 6 ) clear statements that the mosquitoes used in the experiments were laboratory reared and laboratory tested to be free of DENV , and ( 7 ) an explanation that after the mosquitoes are fed they are held in a secure laboratory environment . Between September 2012 and January 2015 , we enrolled 58 DENV-positive subjects from a total of 197 people ( 50% female ) who agreed to participate in direct mosquito feeds if they were identified as having an active DENV infection [34] , representing about 70% of the febrile people invited to participate in the study . Of these , 53 subjects ( 35 males , 18 females ) participated in direct mosquito feeds . The remaining five ( 4 females , 1 male ) agreed to only participate in indirect feeds ( drawing their blood and artificially feeding it to mosquitoes ) . During and directly after the procedure , we asked participants how they felt about doing the feeds , whether they wanted to continue , and if they would do it again . After the FGDs , the entire team present during FGDs compiled all notes to produce a detailed report of each session and discussions that took place . Audiotapes were not transcribed , but were used to fill in gaps in notes and to obtain exact quotes . Data was segregated by themes in the focus group guide: ( 1 ) comprehension of project , based on misconceptions expressed by participants , ( 2 ) questions and concerns about direct and indirect mosquito feeds , ( 3 ) willingness to allow children to participate , ( 4 ) number of times individuals would be willing to participate in daily indirect and direct mosquito feeds , and ( 5 ) preference for participation in their home versus in the laboratory . Codes were developed based on these themes , as well as subthemes , by the two lead investigators ( VAPS , ACM ) . Data was further stratified by participant—those from surveillance and non-surveillance areas . All notes were coded by one member of our research team who was not present during the FGDs , using Dedoose qualitative analysis software , version D . 7 . 5 . 16 . Additional sub-themes that emerged during coding were discussed and added if appropriate , and the person coding returned to previous transcripts to ensure all sub-themes were included . Any questions in the coding process were resolved by discussions between the two lead investigators present at FGDs , along with the person coding . Results are presented based on these main themes .
We determined that the most effective way to communicate the purpose and procedures of the mosquito feeding experiments was to early in the FGD present the video ( Fig 1 , S1 Video ) used in the informed consent process . We decided that the video should be preceded by a brief explanation of its content and why it was important . Based on the types of questions answered after the video , it was clear that many participants required further explanation , interaction , and reinforcement of key points by the research team . To test comprehension , we asked participants “why are we doing the study ? ” Initial responses varied widely , but the most common responses indicated that participants captured two key concepts: ( 1 ) we wanted to know more about dengue , and ( 2 ) we were trying to understand what would happen to the mosquitoes after they fed on an infected person . Participants used phrases like “you want to see what kind of reaction the mosquitoes will have” or “[you want to see] if we can give the mosquitoes dengue . ” Although it was clear that participants in many cases tried to repeat messages from the consent video , their focus on transmission from humans to mosquitoes demonstrated an understanding of why feeding mosquitoes blood , either directly or indirectly , was an essential part of the project . Comprehension of this key message was incomplete initially , with some participants expressing interesting misconceptions ( see below and Table 3 ) . In most groups , however , individuals who grasped the concept that we were interested in what happened to mosquitoes , rather than what happened to human participants , provided explanations to others in the group who did not understand this concept . This method of peer-to-peer explanation helped researchers assess how well some participants had understood , as well as allowed us to document how the concept was expressed among the participants themselves so we could use their language in information materials developed in the future . When asked what would motivate them to participate in a study like ours , participants felt it was important to better understand dengue and this would in turn help their community . This seemed to be a satisfactory reason for most to participate . One participant seemed taken aback when probed about her willingness , she replied “why wouldn’t I participate ! ” As described previously , once the FGD participants had their initial concerns and questions addressed and understood the scientific objectives of the study , enthusiasm to participate increased . Clinical attention that would be received during the study was seen as valuable . FGD participants liked the idea that they would be monitored by a doctor and our research team until they were dengue free . There was interest in knowing how their dengue was progressing and knowing if they were recovering . For example , some expressed interest in knowing if they had anemia , and those who had heard about platelet counts wanted to know if that kind of information would be provided . We clarified that clinical monitoring was not dependent on participating in mosquito feeds or intensive blood draws , because all participants in our studies get medical attention . Most FGD members continued to express willingness to participate . A related motivation was obtaining a dengue diagnosis . Before we clarified that for our future experiments we would need to directly feed mosquitoes and take tubes of blood , more participants thought feeding mosquitoes directly would more desirable than providing blood samples . It was seen as less painful and would not require as much blood . Some participants also expressed curiosity , stating , “I want to know how it feels to feed the mosquitoes” , or “I would want to see how many mosquitoes got infected if I were sick” . We implemented a variety of exercises to discern people’s attitudes about providing blood samples and feeding mosquitoes , and their perceived advantages and disadvantages for each . These questions were posed after the FG facilitator ( VAPS ) felt that the participants understood the procedures would only be conducted on DENV-infected people and that the mosquitoes used in the experiments were clean . In the initial FGDs , we asked people which of the two options , direct mosquito feeding or blood collection , they would prefer , followed by an explanation that we would need both and a question about their willingness to provide both . In later FGDs , we started by saying we needed people to feed mosquitoes and give blood samples . We asked whether they would be willing to do both , and the pros and cons of each . About half of participants said they would be willing to do both , whereas a little over a quarter said they would only be willing to feed mosquitoes directly and a little under a quarter said they would only be willing to provide venous blood samples for indirect feeding . Most parents stated that they were comfortable allowing their children to participate in direct feeding if the child could decide . There was some discussion among participants about the age at which children would understand the procedures , but parents felt that , as long as the child’s choice was respected , they would allow their participation . At the same time , many parents were skeptical that their child would choose to participate or be able to sit still throughout the process , with a few FGD participants mentioning how the sight of a nurse caused their children to cry or run away . In contrast , a few parents said that their child would want to participate because they were “brave” or that it would be interesting to them . A few people said they would want to experience a feed before allowing their child to participate: “I want to know what it feels like before giving authorization for my child . ” Some parents wanted to ensure that less blood would be taken from children than adults , “If they take my blood for three days , then the child should only do two days… . ” As with adults , some children were recognized as sensitive to mosquito bites; parents of these children would not allow them to participate . We used a variety of strategies to probe participants on their willingness to participate in direct mosquito feeding and provide blood samples multiple times . These strategies included asking people to raise their hands and playing games with individuals jumping to the “yes” or “no” side of a line on the floor based on their responses . Responses ranged from none to as many times as necessary ( explained by the facilitator as every day they had a fever ) . Those who answered as often as necessary qualified their responses by stating they would participate “as long as I still have dengue , ” “until my platelets are okay , ” or “until my hemoglobin is okay . ” There was a strong connection between willingness to participate and the clinical follow up that would accompany study participation . About 65% of participants who indicated a willingness to participate said they would be willing to repeat the procedures at least three times . Others were willing to have samples taken every other day or at the beginning and end of their illness . We asked participants their opinions about conducting direct mosquito feeds in their own homes versus a laboratory setting to determine whether people had safety concerns about performing the experiments in their communities . Overwhelmingly , people preferred their homes . Everyone expressed a desire to participate in the fastest and most convenient procedure . One FGD participant stated a preference for home feeds “so that my family sees this . ” The few individuals who expressed a preference for participating in the procedures in the laboratory gave the following reasons: “because it is safer and I would be afraid if mosquitoes escaped , my family could get sick , ” “because people in my family would gossip , ” “I would have more confidence , it is more credible… , ” and “so I can get more information . ” In one FGD , the issue of prying neighbors and their possible negative reactions to mosquito feedings was raised . Although we did not include questions about incentives for providing blood samples or participating in direct mosquito feeds in our original FG guide , the issue of incentives emerged during one FG discussion in a group where since 2007 , participants received small thank you gifts ( i . e . , powdered milk , vitamins ) after providing blood samples . We asked participants what they thought about incentives and whether they should be given for participation in direct mosquito feeds . One participant stated it would not motivate her , but that “some people would do the mosquito feeding for money or necessity . ” We then asked the group directly whether we should provide incentives , and the group answered , definitively and in unison , “No ! ” Although none of these participants suggested we should stop providing milk and vitamins for our ongoing studies , there were a variety of strong responses that indicated incentives were not the reason they would participate and that they often had to set their gossiping neighbors straight about this .
The impact of dengue in Iquitos is large and far-reaching , making people willing collaborators in studies to reduce the local burden of the disease . This , combined with the normalcy of mosquito bites in Iquitos , made direct mosquito feeding experiments , which might seem peculiar to some , acceptable to the majority of our FGD participants , and for parents , also an acceptable experiment to be carried out among children who could assent to the process . Taken in context with the response to direct feeding experiments [34] , in Iquitos , when conducted by experienced researchers using a DENV-free generation of mosquitoes raised in a secure insectary , this is a low-risk , non-invasive procedure that is experimentally superior to artificial mosquito feeding methods . The use of formative research , to identify and develop effective communication strategies for “unusual” procedures provides and effective path to opening up new research procedures to communities where they might be appropriate . | Approximately half of the world’s population is at risk of contracting dengue virus ( DENV ) . Ethical and logistical concerns with feeding lab-raised mosquitoes directly on naturally infected human subjects , and the lack of a relevant animal model for DENV experimental infection , are important obstacles to better understanding DENV transmission from humans to mosquitoes . Results from artificial infectious blood meals can bias estimates of mosquito infection and transmission rates . Based on 12 focus group discussions , we determined that the practice of feeding uninfected lab-raised mosquitoes on naturally infected human subjects is highly acceptable to people living in Iquitos , Peru , especially after common concerns are addressed . The majority of participants were willing to have mosquitoes feed on them directly and to give venous blood samples to feed to mosquitoes indirectly . Most participants stated a preference for direct feeding . This formative research , including recognition of and addressing common misconceptions , will help guide future development of protocols using biologically relevant direct mosquito feeding methods . | [
"Abstract",
"Introduction",
"Methods",
"and",
"materials",
"Results",
"Discussion"
] | [
"invertebrates",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"pruritus",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"togaviruses",
"pathogens",
"neighborhoods",
"geographical",
"locations",
"microbiology",
"social",
"sciences",
"neuros... | 2019 | Acceptability of Aedes aegypti blood feeding on dengue virus-infected human volunteers for vector competence studies in Iquitos, Peru |
Lectins and adhesins are involved in bacterial adhesion to host tissues and mucus during early steps of infection . We report the characterization of BC2L-C , a soluble lectin from the opportunistic pathogen Burkholderia cenocepacia , which has two distinct domains with unique specificities and biological activities . The N-terminal domain is a novel TNF-α-like fucose-binding lectin , while the C-terminal part is similar to a superfamily of calcium-dependent bacterial lectins . The C-terminal domain displays specificity for mannose and l-glycero-d-manno-heptose . BC2L-C is therefore a superlectin that binds independently to mannose/heptose glycoconjugates and fucosylated human histo-blood group epitopes . The apo form of the C-terminal domain crystallized as a dimer , and calcium and mannose could be docked in the binding site . The whole lectin is hexameric and the overall structure , determined by electron microscopy and small angle X-ray scattering , reveals a flexible arrangement of three mannose/heptose-specific dimers flanked by two fucose-specific TNF-α-like trimers . We propose that BC2L-C binds to the bacterial surface in a mannose/heptose-dependent manner via the C-terminal domain . The TNF-α-like domain triggers IL-8 production in cultured airway epithelial cells in a carbohydrate-independent manner , and is therefore proposed to play a role in the dysregulated proinflammatory response observed in B . cenocepacia lung infections . The unique architecture of this newly recognized superlectin correlates with multiple functions including bacterial cell cross-linking , adhesion to human epithelia , and stimulation of inflammation .
The Burkholderia cepacia complex ( Bcc ) is a group of Gram-negative bacteria comprising at least 17 species [1] . Bcc species are common in the environment and can be isolated from various sources including water , soil and vegetation . Bcc bacteria are involved in symbiosis and other interactions with plants that are beneficial for agriculture , but they are also recognised as important opportunistic human pathogens . In particular , B . cenocepacia causes infections in patients suffering from chronic granulomatous diseases [2] and cystic fibrosis [3] with significant morbidity and mortality . This is in part due to the extreme resistance of B . cenocepacia strains to almost all clinically useful antibiotics and their transmissibility between patients [4] . B . cenocepacia isolates survive either extracellularly in the airways or intracellularly within epithelial and phagocytic cells [5] . Among the virulence factors of B . cenocepacia [6] , soluble lectins bind to carbohydrates present on epithelial cells and mucus [7] . A family of four soluble lectins has been identified in B . cenocepacia , all of them containing at least one domain with strong sequence similarity with LecB ( PA-IIL ) from Pseudomonas aeruginosa . LecB is a tetrameric fucose-binding lectin with unusually high affinity for carbohydrate mediated by two bridging calcium ions in its binding site [8] . Its structure has been elucidated and the involvement in biofilm formation and epithelial cell adhesion has been demonstrated [9] , [10] . The four soluble lectins of B . cenocepacia are designated BC2L-A , -B , -C and –D . BC2L-A , consisting of one LecB like domain , associates as a dimer and binds mannose and oligomannose-type N-glycans [7] , [11] . The three other lectins have additional N-terminal domains . The N-terminal domain of BC2L-C has been recently characterized as a novel fucose binding domain with a TNF-α-like fold [12] . Thus , BC2L-C encompasses two lectin domains probably assembling as oligomers , but nothing is known about its architecture and biological role . Here , we characterize the structure and specificity of the C-terminal domain of BC2L-C . We also compare the specificity of the whole lectin with that of each domain . The overall hexameric architecture of the lectin was determined in solution by electron microscopy and SAXS . Finally , we determined that this lectin binds to the bacterial cell surface and elicits a proinflammatory response in cultured respiratory epithelial cell cultures . Together , we conclude that BC2L-C is a novel superlectin with multiple specificities and biological functions .
BC2L-C has an N-terminal region of 155 amino acids ( the TNF-α-like lectin ) , a 28-aa linker region , and a 115-aa C-terminal region ( Fig . 1A ) [12] . The C-terminal domain has sequence similarity with two-calcium bacterial lectins such as LecB/PA-IIL from P . aeruginosa ( 43% identity ) and related ones ( Fig . 1B ) . The Ala-Ala-Asn sequence in the “specificity loop” [13] suggests that this domain is specific for mannose . BC2L-C interacts strongly with surfaces modified by mannose and fucose residues but not by galactose , as shown by Surface Plasmon Resonance ( SPR ) experiments , while the C-terminal domain , BC2L-C-ct binds only to mannose-coated chips ( Fig . 2 ) . Since the isolated BC2L-C-nt domain binds strongly to fucose but not to mannose and galactose [12] , BC2L-C is therefore a novel type of lectin consisting of independent fucose and mannose-binding domains . The fine specificity of BC2L-C was determined using the Glycan Array facility of Consortium for Functional Glycomics with 377 carbohydrates available . BC2L-C bound to oligosaccharides containing terminal mannose or fucose residues ( Fig . 3A ) . In contrast , the glycans bound by BC2L-C-ct included oligomannose-type N-glycans and their terminal fragments ( Fig . 3B ) . The monosaccharide α-d-mannose ( Man ) was the shortest fragment recognized , albeit not efficiently ( #8 in Fig . 3B ) . The recognised disaccharides were Manα1–2Man , Manα1–3Man and Manα1–6Man indicating that the specificity for the linkage is not strict . Hybrid structures with galactose or sialic acid on one antenna and α-mannose on the other one were also bound . The BC2L-C-nt domain is a fucosylated oligosaccharide binding lectin [12] . Therefore , bound fucosylated epitopes encompassed all fucosylated human histo-blood group epitopes such as blood group O ( H ) and Lewis oligosaccharides , with some preference for the Fucα1-2Gal epitope ( Fig . 3C ) . The specificity charts for the separated domains do not overlap but their superimposition clearly explains that the specificity of the whole protein is determined by the contributions of the specificities of each domain . The interaction of BC2L-C-ct with different carbohydrates was characterised by isothermal titration microcalorimetry . All thermograms display exothermic peaks with saturation of binding sites at the end of titration ( Fig . S1 ) . Affinity values and thermodynamics parameters are reported in Table 1 . BC2L-C-ct bound to Man and α-methyl-mannoside ( αMeMan ) with a strong affinity in the micromolar range but with a stoichiometry close to 0 . 5 , indicating that only one binding site per dimer is accessible . Using the whole protein , the same stoichiometry and affinity was measured for mannoside , but not for Lewis Y , a fucosylated oligosaccharide that binds to the other domain with a stochiometry of one ( n = 1 ) . The branched trimannoside ( Manα1-3 ( Manα1-6 ) Man ) exhibited an even lower stoichiometry ( n = 0 . 22 ) demonstrating that the two terminal mannose residues bind to two BC2L-C-ct dimers , as observed previously for BC2L-A/trimannose interaction [11] . Binding was also tested towards l-glycero-d-manno-heptopyranose ( Hept ) since this residue is similar to mannose differing only in an additional hydroxymethyl group at C-6 and several Hept residues are in the B . cenocepacia lipopolysaccharide ( LPS ) [14] , [15] . The methylated monosaccharide ( αMeHept ) bound with an affinity of 150 µM and the α1-3 linked disaccharide bound with an affinity of 88 µM , indicating that heptose-containing LPS could be a candidate substrate for BC2L-C-ct binding . The crystal structure of BC2L-C-ct was solved at 1 . 9 Å resolution ( Table 2 ) , demonstrating a nine-stranded antiparallel β-sandwich fold that is similar to the two-calcium lectins characterised in P . aeruginosa [16] , Chromobacterium violaceum [17] , and Ralstonia solanacearum [18] ( Fig . 4A ) . The dimeric association displays close similarity to the lectin BC2L-A , described previously [7] , [11] . In contrast to other crystal structures in this family , no electron density for calcium ions and monosaccharide ( d-mannose ) was found in the final model , presumably due to the presence of citric acid in the crystallization buffer . Without the stabilizing effect of the two calcium ions in the binding site , the acidic amino acids that mediate their binding appear to point in all directions ( Fig . 4B ) . A sulphate ion was observed close to one of the two binding sites , establishing hydrogen bonds with Gln241 , His177 and two bridging water molecules . Modelling the complex with αMeMan was possible since BC2L-C-ct has strong sequence similarity to the other lectins of the family ( Fig . 1B ) , in particular with the R . solanacearum RS-IIL for which a crystal structure with calcium and mannose is available [18] . The modelled binding site was built by reorienting the amino acids side chains and slightly modifying the conformations of the loops ( Fig . 4C ) . The main difference with the mannose-binding site of RS-IIL is the presence of His177 ( Asn in all other lectins ) . In the absence of calcium , this histidine interacts with one sulphate ion but also modifies the conformation of the C-terminus of the other chain . Since the C-terminal carboxyl group has an essential role in binding mannose and calcium in all the other similar lectins , the putative destabilizing role of His177 could account for the observed non-even stoichiometry of the dimer . Modeling the interaction between BC2L-C-ct and αMeHept was achieved by extending the hydroxymethyl group at C5 of mannose in a glycolyl one . The binding site can accommodate this bulky group with no steric hindrance and the mode of binding of αMeHept displays the same hydrogen bond network that is observed for mannose ( Fig . 4D ) . The oligomeric state of BC2L-C and the C- and N-terminal domains were analysed by size exclusion chromatography combined with multi-angle laser light scattering ( SEC-MALLS ) and refractometry ( RI ) ( Fig . S2 ) . BC2L-C-ct is dimeric in solution with a molecular mass of 22±1 kDa whereas BC2L-C-nt is trimeric with a molecular mass of 41±1 kDa ( the expected monomeric masses are 12 . 4 kDa and 19 . 3 kDa , respectively ) . The SEC/MALLS profile of the whole BC2L-C analysis indicates a hexamer in solution with a molecular mass corresponding to 145±4 kDa ( the expected mass of the monomeric form is 28 . 2 kDa ) . These results are consistent with the crystallographic data of each domain; the dimeric C-terminal domain established here and the previously determined trimeric N-terminal domain [12] . The overall shape of the hexamer in solution was determined by small angle X-ray scattering ( SAXS ) and validated by negative stain electron microscopy ( EM ) . The Guinier analysis ( Fig . S3 and Table S1 ) suggests an Rg of ∼5 nm with the absence of severe aggregation effects that allowed for the ab initio shape reconstruction to be performed using the idealized SAXS curve . The refined ab initio envelope is elongated ( max length ∼160 Å ) with a pseudo 3-fold axis in the long direction and three bulges protruding from the middle of this long axis . ( Fig . 5A ) . Negative stain electron microscopy analysis validated the SAXS results . Indeed , the three-dimensional reconstruction of BC2L-C at 20 Å resolution ( Fig . 5B and 5D ) shows the same global shape as the SAXS envelope ( Fig . 5C ) . These complementary results confirmed the size and overall shape of the molecule allowing manual fitting of the domains using the combined EM and SAXS reconstructions as the template . Positioning two trimers of BC2L-C-nt on the large axial bulges and three dimers of BC2L-C-ct on the equatorial ring-like envelop fitted well within the envelope ( Fig . 6 ) . However , attempts to mathematically optimize this model were partially successful probably because of the absence of the linker moieties in the model . By adding random chains for the missing linkers ( 6 chains each 28 residues in length ) an acceptable value of chi = 3 . 5 could be attained ( Fig . S4 ) . Distances between domain extremities were checked in the final model , the maximum one being 50 Å , a value that allow for the 28 amino acid linkers to fit . The positions of the linkers are not presented as definitive as the entire complex appears to be flexible , especially in the central part of the molecule . Our results suggest that the low-resolution shape with the two trimers separated along the long axis and three dimers in the middle is the conformation adopted in solution under physiological conditions . A general 3-fold axis is visible , passing by through the trimers and by the center of the donut , but the symmetry is broken by the twisted orientations of dimers . Since the linkers could not be located using available methods , two possible architectures ( mode I and mode II ) can be proposed for the BC2L-C hexamer ( Fig . 6B ) . However , our shape reconstruction indicates that mode I is more probable to occur in solution as the mode II would generate more extended structures of high conformational variability . The expression of BC2L-C in B . cenocepacia J2315 was previously demonstrated by classical proteomics [12] . Western blots were performed with the purified recombinant lectins BC2L-A , -B and -C using rabbit antisera prepared against formalin-fixed intact B . cenocepacia K56-2 cells ( clonally related to J2315 ) . The purified recombinant lectins BC2L-B and -C are strongly detected by anti-K56-2 antibodies , while BC2L-A is barely detectable ( Fig . 7 ) . This result agrees with the observed expression levels of BC2L-A , -B , and -C lectins in B . cenocepacia K56-2 . Since intact bacterial cells were used in the immunisation procedure , we conclude that the lectins are present on the cell surface . BC2L-A , -B and -C lectins were tagged with a FLAG epitope at their N-terminus and expressed as recombinant lectins in B . cenocepacia K56-2 . Culture supernatant analysis by Western blot using anti-FLAG and anti-RNA polymerase alpha subunit antibodies ( used as a cell lysis control ) revealed that the three lectins are secreted or released into the extracellular medium without detectable cell lysis ( Fig . 8A ) . To determine whether BC2L-A , -B and C lectins associate to the bacterial cell surface , bacterial cells expressing the FLAG-tagged lectins were incubated with buffer or with 50 mM d-mannose for 5 min . Western blot using anti-FLAG and anti-RNAPα revealed that BC2L-B and -C lectins are released into the supernatant upon incubation with mannose but not with buffer only . That BC2L-B and -C lectins are specifically released upon mannose treatment without any detectable cell lysis suggests that BC2L-B and -C lectins are located on the surface of B . cenocepacia ( Fig . 8B ) . Since the crystal structure of BC2L-C-nt demonstrated a TNF-α-like fold [12] , the immunostimulatory activities of BC2L-C and its domains were tested on epithelial cells . A markedly increase in IL-8 production was observed in cells exposed to BC2L-C and this activity was attributed to the N-terminal domain ( Fig . 9 ) . Attempts to inhibit the IL-8 production by carbohydrate ligands were unsuccessful ( data not shown ) , indicating that the carbohydrate-binding and pro-inflammatory eliciting activities reside in different parts of the molecule . Attempts to inhibit IL-8 production using siRNA directed against the TNF-α receptor ( TNFR1 ) also resulted in negative results ( Fig . S5 ) . The immunostimulatory activity of BC2L-C is therefore mediated by its N-terminal domain in a carbohydrate-independent manner , and does not appear to be mediated by selective binding to TNFR1 .
To our knowledge , BC2L-C is the first protein identified harbouring two different lectin domains with distinct specificity , for which we refer to it to as a “superlectin” . Multispecificity was only found in lectins with duplicated domains resulting from divergent evolution , such as some human galectins and plant lectins [19] , [20] . The association of two functionally and structurally distinct domains in BC2L-C is therefore the paradigm for a new class of lectins . The cellular localisation of the soluble lectins produced by opportunistic bacteria remains an open question . While the lectins have a role in host recognition , they are present in large quantity in the cytoplasm and do not contain any canonical secretion signals . Previous work demonstrated the location of LecB on the outer membrane of P . aeruginosa [21] and recent work suggested that transient glycosylation of the lectin is required for transportation [22] . Our data indicate that B . cenocepacia lectins are also located at the bacterial surface . Control data monitoring the RNA polymerase alpha subunit ( cytoplasmic protein used as cell lysis control ) demonstrate that the lectin does not exit by simple cell lysis . Therefore , the lack of typical secretion sequences in these lectins suggests they are secreted by one or more specialized secretion systems that are yet to be identified . We also demonstrate that surface localisation depends on the mannose-binding site in the C-terminal domain , since treatment of bacterial cells with d-mannose results in the release of the lectins . Since this binding site has strong affinity for l-d-heptose , an abundant component of the B . cenocepacia LPS [14] , it is possible that LPS may provide an attachment site on the bacterial surface . However , attempts to demonstrate lectin binding to LPS were unsuccessful ( data not shown ) , suggesting that the lectin may bind to a different bacterial surface molecule . The unique hexameric architecture of BC2L-C is well suited for cross-linking between bacteria and epithelial cells ( Fig . 10 ) . The advantage of such flexible structure is that all carbohydrate binding sites can be exposed at the surface and free to interact . Also , a flexible linker could adapt its conformation under shear force and provide tight binding as observed in some pili adhesins [23] . The three mannose/heptose binding sites , responsible for bacterial surface binding , are located in the external part of the middle ring , while the fucose binding sites , that binds to H-type 1 and other fucosylated epitopes on glycolipids , are present at each extremity . These TNF-α-like N-terminal domains have a strong pro-inflammatory effect , as determined by IL-8 release by epithelial cells . Lung infection by B . cenocepacia in CF patients is characterized by strong inflammation [24] . In addition to the classical activation of Toll-like receptors by LPS and flagella [25] , it has been recently demonstrated that B . cenocepacia activates the TNFR1 signalling in cystic fibrosis airway epithelial cells [26] . In conclusion , our study opens many questions about the biological function of super lectins in opportunistic bacteria . Future work will unravel the binding epitope on the bacterial cell surface and provide more details on the physiological role of the super lectin in the infection processes .
Monosaccharides ( Sigma ) , trimannoside ( Dextra ) and Lewis Y ( Elicityl ) were obtained from commercial sources . Methyl l-glycero-α-d-manno-heptopyranoside and allyl l-glycero-α-d-manno-heptopyranosyl- ( 1→3 ) -l-glycero-α-d-manno-heptopyranoside were synthesized according to published procedures [27] , [28] that are briefly described in Text S1 . The gene encoding full-length BC2L-C was synthesized by GenScript Corp with optimization for expression in E . coli and contained flanking NdeI and HindIII sites . This synthetic gene was cloned into pRSET vector ( Invitrogen ) , resulting in pRSET_bc2l-c , which was used as a template to clone pRSET_bc2l-c-ct encoding BC2L-C-ct ( Table S2 ) . E . coli BL21 ( DE3 ) cells containing plasmid pRSET_bc2l-c and/or pRSET_bc2l-c-ct were cultured in LB broth low salt medium ( Duchefa Biochemie ) containing 100 µM ampicillin at 37°C until the OD600 reached ∼0 . 5 . After the addition of 0 . 5 mM IPTG ( Duchefa Biochemie ) , cells were cultured for an additional 3 hours at 30°C , harvested by centrifugation and resuspended in 20 mM Tris buffer containing 100 mM NaCl and 100 µM CaCl2 , pH 7 . 5 . Cells were disintegrated by ultrasonic vibration and the soluble fraction was collected by centrifugation at 21000 g at 4°C for 30 min . Harvested cells were stored in plastic falcons at −20°C . Recombinant BC2L-C and/or BC2L-C-ct were purified by affinity chromatography on the mannose-agarose ( Sigma-Aldrich ) column using an FPLC system ( ÄCTA , GE Healthcare ) . After washing , specific elution was carried out with 20 mM Tris pH 7 . 5 , 100 mM NaCl and 10 mM EDTA . The protein was dialysed in 10 mM Tris pH 7 . 5 , 20 mM NaCl and 1 mM CaCl2 . Protein purity was assessed by SDS-PAGE ( 12% gel ) and stained with Coomassie Brilliant Blue R-250 ( Sigma Aldrich ) . Preparation of BC2L-C-nt was performed as described previously [12] . SPR experiments were performed on a BIAcore 3000 instrument ( GE Healthcare ) at 25°C using a running buffer HEPES - Buffered Saline ( HBST ) ( 10 mM HEPES and 150 mM NaCl , pH 7 . 5 containing 0 . 005% ( v/v ) Tween 20 ) and a flow rate of 5 µL per minute . Two different chips presenting monosaccharides have been used . Biot-PAA ( biotinylated polyacrylamide ) probes bearing sugar moieties ( Lectinity Corp , Russia ) were trapped on a CM5 ( BIAcore Life Science ) sensor chip that was coated with streptavidin using the standard procedure ( Amine coupling , BIAcore Sensor Surface Handbook ) . Each Biot-PAA-monosaccharide ( 50 µL at concentration 200 µg/mL ) was injected to the selected channel . Direct binding curves of the lectin to immobilised sugars were measured over the concentration range 0 . 35–0 . 45 mg/mL . Samples were injected ( 20 µL , KINJECT ) onto the CM5 chip at a flow rate 5 µL/min . The chip was regenerated using 100 mM EDTA and 50 mM NaOH . Binding of the proteins to the immobilised sugars was determined by resonance units ( RU ) over time and data were evaluated using the BIAevaluation Software ( version 4 . 1 ) . Purified BC2L-C lectin samples were labeled with Alexa Fluor 488-TFP ( Invitrogen , CA ) according to manufacturer's instructions and re-purified on a D-Salt polyacrylamide desalting column ( Pierce , Rockford IL ) . Alexa-labeled proteins were used for glycan-array screening with standard procedure of the Core H of the Consortium for Functional Glycomics ( Emory University , Atlanta , GA , http://www . functionalglycomics . org ) . The screening of the printed glycan microarray chip ( version 3 . 2 , with 377 glycans from a library of natural and synthetic glycans ) was performed with a concentration of BC2L-C of 200 µg/mL dissolved in 20 mM HEPES , 140 mM NaCl , 5 mM CaCl2 , pH 7 . 5 for all samples . Experiments were performed at 25±0 . 1°C using VP-ITC and ITC200 calorimeters ( Microcal , GE Healthcare ) . Saccharides and proteins were dissolved in the same buffer ( 20 mM Tris pH 7 . 5 , 20 mM NaCl and 0 . 03 mM CaCl2 ) . Protein concentration for measurements varied from 125 to 400 µM . Aliquots of 2 or 10 µL of sugar solution at various concentrations from 1 . 56 to 50 . 0 mM , were added automatically to the protein solution present in the calorimeter cell . Stirring was adjusted to at 300 and 1000 rpm . Titration of BC2L-C and BC2L-C-ct was performed with αMeMan , d-mannose and α-methyl-l-fucoside , trimannose and diheptose . Control experiments performed by injections of buffer in the protein solution yielded insignificant signals . Integrated heat effects were analysed by non-linear regression using a single-site binding model ( Microcal Origin 7 ) . The experimental data fitted to a theoretical titration curve brought up the association constant Ka and the enthalpy of binding ΔH . The other thermodynamic parameters such as free energy ΔG and entropy ΔS were calculated from the equation: ΔG = ΔH - TΔS = -RTlnKa , where T is the absolute temperature and R is molar gas constant ( 8 . 314 J . mol−1 . K−1 ) . All experiments were performed with c values between 10< c <100 [29] . At least two or three independent titrations were performed for each tested ligand . Lyophilized BC2L-C-ct was solubilised ( 14 . 5 mg/mL ) in 5 mM Tris buffer ( pH 7 . 5 ) containing 4 . 0 mM αMeMan and 0 . 1 mM CaCl2 . Initial crystallization conditions were determined using commercial crystallization screens ( Hampton Research and Molecular Dimension Limited ) using a Mosquito robot ( TTP LabTech Ltd ) . Protein crystals in the form of thin baguettes with hexagonal profile appeared after several weeks at 17°C in the presence of 100 mM Sodium Citrate pH 5 . 5 and 2 . 5 M Ammonium Sulphate . These initial conditions were optimized and scaled up to 4 µL hanging drops resulting in bigger crystals with the same shape . Crystals were cryocooled at 100 K in liquid nitrogen after soaking them for as short a time as possible in 30% ( v/v ) glycerol mixed with precipitant solution . Diffraction data for BC2L-C-ct were collected on the beamline ID14–1 at ESRF ( Grenoble ) using an ADSC Q210 CCD detector ( Quantum Corp . ) . Diffraction images were integrated using MOSFLM [30] , scaled and converted into structure factors using the CCP4 program suite [31] . Protein crystallised in the hexagonal space group P65 ( a = b = 100 . 814 Å , c = 47 . 313 Å , γ = 120 . 0° ) with two monomers in the asymmetric unit . The 1 . 9 Å structure was solved by molecular replacement using the MOLREP program [32] , [33] . A monomer of CVIIL lectin ( PDB: 2BV4 ) from Ch . violaceum [17] was used as a search model . Crystallographic refinement was carried out with the program REFMAC5 [34] alternated to manual rebuilding using WinCoot [35] . The solvent model was built automatically with the program ARP/wARP [36] and revised manually with WinCoot . Stereochemical verification was performed with the PROCHECK program [37] . Details about data collection and refinement statistics are available in Table 2 . The final model for the apo-form of BC2L-C-ct was deposited in the PDB database with accession code 2XR4 . A model of the binding site in complex with αMeMan and calcium ions was produced combining the -apo structure from the present structure combined with that from the complex between R . solanacearum RS-IIL and αMeMan ( pdb code 1UQX ) [18] . Briefly , a monomer of RS-IIL complexed with αMeMan was fitted on one monomer of BC2L-C-ct and the side chains of amino acids in the binding site of BC2L-C-ct were adjusted to match those of RS-IIL . Coordinates for monosaccharide and calcium ions were merged with those of BC2L-C-ct . Hydrogen atoms and partial charges were added using Sybyl software ( Tripos Inc , St Louis ) using Amber parameters for the protein and PIM parameters for carbohydrates [38] . Energy minimisation was performed with geometry optimisation of all hydrogen atoms , monosaccharide and side chains in the binding site . Graphical representations are performed with Sybyl and Pymol ( Pymol . org ) . BC2L-C whole protein and its separate domains were analysed on the Superdex 200 ( GE Healthcare ) column equilibrated with 20 mM Tris , 250 mM NaCl , 1 mM CaCl2 , pH 7 . 5 using the FPLC system ( ÄCTA , GE Healthcare ) . A 200 µL sample was loaded at a flow rate of 0 . 4 mL/min . Molecular weights were determined using gel filtration standard ( Bio-Rad ) . Fractions corresponding to molecular mass of hexamer , dimer and trimer , respectively , were concentrated by centrifugation ( Vivaspin , Sartorius Stedim Biotech ) up to the concentration of 4 . 2 mg/mL for BC2LC , 7 . 5 mg/mL for BC2L-C-nt and 10 . 7 mg/mL for BC2L-C-ct , respectively , and used for SEC-MALLS analysis . 100 µL of each sample was loaded at a flow rate of 0 . 4 mL/min . On-line MALLS detection was performed with a DAWNEOS detector ( Wyatt Technology Corp . ) using a laser emitting at 690 nm and by refractive index measurement using an RI2000 detector ( Schambeck SFD ) . Weight-averaged molar masses ( Mw ) were calculated using the ASTRA software ( Wyatt Technology Corp . ) . The BC2L-C protein was purified by gel filtration ( as previously described ) immediately prior to the SAXS experiment . The central fraction of the peak containing the hexameric BC2L-C ( 1 . 27 mg/mL ) was collected and used to prepare two additional dilutions ( 0 . 66 and 0 . 31 mg/mL ) with sample concentrations verified using a spectrophotometer ( NanoDrop Technologies ) . The rest of the fractions containing the hexameric BC2L-C were collected and concentrated to 4 . 2 mg/mL . SAXS data from the resulting samples ( from 0 . 33 to 4 . 2 mg/mL ) were collected at the ESRF BioSAXS station ( ID14EH3 , http://www . esrf . fr/UsersAndScience/Experiments/MX/About_our_beamlines/ID14-3 ) at fixed energy wavelength ( 13 . 32 keV , λ = 0 . 931 Å ) . Samples were exposed using 30 µl of protein solution loaded into a 2 mm quartz capillary mounted in vacuum using an automated robotic system ( developed as part of a trilateral collaboration between ESRF and EMBL Hamburg and Grenoble Outstation ) which enables the sample to pass through the beam during exposure to minimise the effect of radiation damage . 2D scattering images were collected on a Pilatus 1M detector ( Dectris ) 1 . 83 m from the sample . Standard data collection was used for all data ( 10 frames each 10 second in duration ) . Individual time frames are processed automatically and independently by the data collection software ( BsxCUBE ) developed at the ESRF , yielding individual radially averaged curves of normalised intensity versus scattering angle ( s = 4πSinθ/λ in nm ) . Time frames are combined excluding any data points affected by aggregation induced by radiation damage to give the average scattering curve for each measurement . The scattering from the buffer alone was measured before and after each sample measurement and the average of the scattering before and after each sample was used for background subtraction , the different concentrations were then compared and merged to obtain the idealized scattering curve using the program PRIMUS ( 13 ) form the ATSAS package developed by EMBL Hamburg . Ab-initio models were produced with DAMMIF ( 14 ) and averaged with DAMAVER [39] . Rigid body modeling was undertaken using MASHA [40] with 6 additional random chains of 28 residues created by ranch13 ( also part of the ATSAS package from EMBL-Hamburg ) to represent the linkers . The plot of the fits was produced with the beta version of SASPLOT from the upcoming cross-platform release of the ATSAS package developed at EMBL-Hamburg . For preparation of negatively stained BC2L-C , the purified sample was diluted to 0 . 05 mg/mL , applied to the clear side of carbon on a carbon-mica interface and stained with 2% ( w/v ) sodium silicotungstate at pH 7 . Images were recorded under low-dose conditions with a JEOL 1200 EX II microscope at 100 kV and at nominal 40000× magnification . Selected negatives were digitized on a Zeiss scanner ( Photoscan TD ) at a step size of 14 micrometer giving a pixel size of 3 . 5 Å at the specimen level . A generous semi-automatic particle selection with the EMAN boxer routine [41] lead to an extraction of a total of 18426 subframes of 56×56 pixels containing individual BC2L-C complex particle frames which were CTF-corrected with CTFFIND3 [42] and bsoft [43] , and low-path-filtered at 15 Å with Imagic-5 . Subsequent data processing was performed with the Imagic-5 software package [44] . The data set was translationally but not rotationally aligned relative to the rotationally averaged total sum of the individual images . This translationally centered data set was subjected to multivariate statistical analysis and classification . Characteristic class averages were then used as a set of references for multireference alignment of each subframe with Spider [45] , [46] and the new translational parameters were used to update the boxer coordinates and extract better centered particles . This procedure was repeated several times until the classes became stable and the individual frames well centered . At this point , the class averages were compared to projections of the current SAXS model of BC2L-C filtered to 25 Å resolution . Given a notable similarity between them , the SAXS model was filtered to 80 Å resolution in order to resemble a nearly featureless blob of density but conserve the particle dimensions . This blob was used as an initial model for iterative projection matching with Spider [45] , [46] . The resolution of the final 3D reconstruction of the negatively stained BC2L-C was estimated via Fourier shell correlation to be around 20 Å according to the conservative 0 . 5 criterium . The epitope FLAG-containing sequence was excised from plasmid pBADNTF [47] and subcloned into pDA12 [48] using EcoRI and HindIII restriction enzymes , giving rise to pEL-1 . The lectin encoding genes ( BCAM0184 ( bc2l-b ) , BCAM0185 ( bc2l-c ) and BCAM0186 ( bc2l-a ) ) were PCR amplified by use of B . cenocepacia J2315 genomic DNA as template and sense and antisense primers with BamHI and HindIII restriction sites , respectively that were designed for each gene . Primer pairs were as follows: ( 5′TTTAGGATCCTGCTGATTCTCAAACGTCATCCA-3′ ) and ( 5′-TTTTAAGCTTAACGTG CGTCAGGTCAGC-3′ ) for bc2l-a; ( 5′-TTTTGGATCCTTCACAACCCTTTACCCACGA-3′ ) and ( 5′-TTTTAAGCTTGTGATGTAACGGCGAAGACC-3′ ) for bc2l-b; ( 5′-TTTTGGATCCTC CCCTCCTTTCGGCTTCGAT-3′ ) and ( 5′-TTTTAAGCTTGTACAGCAGTGGGACTGCAA-3′ ) for bc2l-c . Amplicons were digested with BamHI and HindIII and ligated into similarly digested pEL-1 giving rise to pBC2L-AFLAG , pBC2L-BFLAG and pBC2L-CFLAG plasmids , which encode BC2L-A , BC2L-B and BC2L-C , respectively N-terminally fused to the FLAG epitope . Plasmids were mobilized into B . cenocepacia K56-2 by triparental mating using E . coli DH5α carrying the helper plasmid pRK2013 [49] as previously described . Exconjugants were selected onto tetracycline 100 µg/ml and gentamicin 50 µg/mL containing plates . Culture supernatant proteins were precipitated with trichloroacetic acid as described previously [48] The protein concentration was determined by Bradford assay ( Bio-Rad ) and 4 µg of protein were loaded on a 18% SDS-PAGE gel . After electrophoresis , gels were transferred to nitrocellulose membranes for immunoblot analysis . The membranes were incubated with the 4RA2 monoclonal antibody ( Neoclone ) cross-reacting with the B . cenocepacia RNA polymerase subunit alpha ( cytosolic protein , cell lysis control ) and the FLAG M2 monoclonal antibody ( Sigma ) . The Alexa Fluor 680 goat anti-mouse IgG ( Molecular Probes ) was used as a secondary antibody . Detection was performed using the Odyssey Infrared Imager ( LI-COR Biosciences ) . Overnight cultures were diluted to an OD600 nm of 0 . 03 in 50 mL LB and grown at 37°C for 8 h . Cells were then centrifuged at 5000 g for 10 min . The pellet was washed twice with 25 mL of phosphate buffered saline ( PBS ) and finally resuspended in 1 . 5 mL PBS . Five hundred µL aliquots were placed into two eppendorf tubes to which 500 µL of PBS or 500 µL of 100 mM d-mannose made in PBS ( 50 mM final concentration ) was added . Samples were gently mixed by inversion and incubated for 5 min at room temperature . Samples were centrifuged at 6000 g for 5 min , supernatants were collected ( 800 µL ) and filter-sterilized using 0 . 2 µM filters . Proteins were precipitated overnight at 4°C with trichloroacetic acid ( 10% final concentration ) . Samples were centrifuged at 16 000 g for 30 min at 4°C . Each pellet was then washed with 1 mL of ice-cold acetone , air-dried and resuspended in 15 µL of sodium phosphate buffer 0 . 1 M pH 7 . 2 . The totality of the samples were loaded on a 18% SDS-PAGE gel . BC2L-C and its separate domains were used for the stimulation of the human bronchial cell line BEAS-2B obtained from the American Type Cell Collection ( Manassas , VA ) . Cells were maintained in serial passage in F-12K culture medium supplemented with 10% FCS , 1% penicillin and streptomycin , 1% glutamine and 10 mM HEPES in 75 cm2 culture flasks and seeded at 5×104 on 24-well plates 3 days before stimulation . In all experiments , BEAS-2B cells were stimulated during 15 hours with the different agonists in a 300 µL medium . IL-8 concentrations in cell culture supernatants were determined using a Duo-Set ELISA kit . Duo-Set ELISA kit and the recombinant human TNF-α were obtained from R&D Systems ( Minneapolis , MN ) . | The glycoconjugates that cover the surface of eukaryotic cells are a target for pathogens that use protein receptors for binding to the carbohydrate moieties exposed . Opportunistic bacteria such as Pseudomonas aeruginosa and Burkholderia species of the B . cepacia complex display a wide range of adhesins and soluble lectins that are specific for human oligosaccharides . We characterized the complex architecture of one Burkholderia cenocepacia soluble lectin , and analysed the specificity of two different lectin subdomains . We propose that one of the subdomains attaches to sugars present on the bacteria surface , enabling bacterial aggregation in microcolonies . The other subdomain attaches to sugars in human airways . In addition , this domain can elicit an inflammatory response in airways cells . Burkholderia cenocepacia causes lethal infections in cystic fibrosis patients and soluble lectins may be novel therapeutics targets . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"biochemistry",
"infectious",
"diseases",
"biology",
"microbiology"
] | 2011 | Burkholderia cenocepacia BC2L-C Is a Super Lectin with Dual Specificity and Proinflammatory Activity |
The success of Staphylococcus aureus , as both a human and animal pathogen , stems from its ability to rapidly adapt to a wide spectrum of environmental conditions . Two-component systems ( TCSs ) play a crucial role in this process . Here , we describe a novel staphylococcal virulence factor , SpdC , an Abi-domain protein , involved in signal sensing and/or transduction . We have uncovered a functional link between the WalKR essential TCS and the SpdC Abi membrane protein . Expression of spdC is positively regulated by the WalKR system and , in turn , SpdC negatively controls WalKR regulon genes , effectively constituting a negative feedback loop . The WalKR system is mainly involved in controlling cell wall metabolism through regulation of autolysin production . We have shown that SpdC inhibits the WalKR-dependent synthesis of four peptidoglycan hydrolases , SceD , SsaA , LytM and AtlA , as well as impacting S . aureus resistance towards lysostaphin and cell wall antibiotics such as oxacillin and tunicamycin . We have also shown that SpdC is required for S . aureus biofilm formation and virulence in a murine septicemia model . Using protein-protein interactions in E . coli as well as subcellular localization in S . aureus , we showed that SpdC and the WalK kinase are both localized at the division septum and that the two proteins interact . In addition to WalK , our results indicate that SpdC also interacts with nine other S . aureus histidine kinases , suggesting that this membrane protein may act as a global regulator of TCS activity . Indeed , using RNA-Seq analysis , we showed that SpdC controls the expression of approximately one hundred genes in S . aureus , many of which belong to TCS regulons .
Two-component systems ( TCSs ) are composed of a histidine kinase , usually membrane-bound and acting as an environmental sensor , which phosphorylates a coupled response regulator , often controlling gene transcription . Although these systems have been extensively studied and play an essential role in bacterial adaptation to the environment , the signal ( s ) to which they respond and additional factors positively or negatively controlling their activities remain mostly unknown . Staphylococcus aureus , a major human pathogen , causes diseases ranging from superficial cutaneous abscesses to life-threatening infections affecting all major organs [1] . S . aureus is also a commensal bacterium , colonizing approximately half the human population asymptomatically , essentially within the anterior nares [2] . In addition to its considerable arsenal of virulence factors , S . aureus must rapidly adapt to environmental conditions encountered during host colonization . Among the 16 TCSs encoded by the S . aureus genome [3] , the SaeSR , AgrCA , and WalKR systems are particularly important for controlling virulence and innate immune evasion factors [4–8] . The WalKR system is the only S . aureus TCS shown to be essential for cell viability , suggesting that its activity may respond not only to environmental conditions but could also be controlled by intrinsic bacterial factors [9–11] . Indeed , it is becoming increasingly apparent that so-called two-component systems frequently involve additional proteins regulating the phosphorylation levels of the response regulator [12 , 13] . These include accessory phosphatases such as CheZ , Spo0E or RapA , respectively dephosphorylating the E . coli CheY [14] , B . subtilis Spo0A [15] and Spo0F [16] response regulators , or antikinases such as KipI and Sda that inhibit B . subtilis KinA [17 , 18] and FixT which inhibits the Sinorhizobium meliloti FixJ kinase [19] . A sub-class of histidine kinases , known as intra-membrane sensing kinases [20] , require the permease component of an associated ABC transporter for signal sensing , such as the S . aureus BraS [21] and GraS [22] kinases . Many TCS histidine kinases act as so-called bifunctional sensors , acting on their cognate response regulators both as kinases and phosphoprotein phosphatases [23] . Accordingly , several accessory proteins act by binding to the histidine kinase to inhibit its kinase activity or stimulate its phosphatase activity towards the response regulator . These include the PII protein acting on the NtrB kinase to control nitrogen assimilation by dephosphorylating NtrC [24] , the Streptococcus agalactiae Abx1 Abi-domain membrane protein which interacts with the CovS histidine kinase to inhibit activity of the CovR response regulator [25] , and the SaePQ protein complex , which stimulates the phosphatase activity of the SaeS histidine kinase in S . aureus [26] . In Bacillus subtilis , the WalK histidine kinase is thought to coordinate cell wall plasticity with cell division , with two membrane-bound accessory proteins , WalH and WalI , inhibiting WalK kinase activity [11 , 27–29] . However , the WalH and WalI proteins of S . aureus share no significant sequence similarities with those of B . subtilis and their role is not as clear-cut . Indeed , although they are also membrane proteins and interact with the WalK kinase , WalH and WalI do not seem to play a major role in negatively controlling WalKR activity , suggesting that their functions have evolved [30] . In an effort to identify additional factors controlling the WalKR system , we showed that the S . aureus SpdC Abi-domain protein negatively affects WalK activity and expression of WalKR-regulated genes . We showed that SpdC , previously identified as playing a role in the display of surface proteins [31] , forms a complex with the WalK histidine kinase and that the two membrane proteins preferentially localize at the division septum , suggesting that this interaction regulates WalK histidine kinase activity . The ΔspdC mutant displays a pleiotropic phenotype , including altered resistance towards compounds targeting the cell wall , as well as strongly diminished biofilm formation and virulence . Using RNA-Seq analysis , we showed that SpdC controls the expression of approximately one hundred genes in S . aureus . Indeed , SpdC activity appears to extend well beyond the WalKR system , since we have shown it also interacts with several other S . aureus histidine kinases suggesting it could be involved in controlling multiple regulatory/adaptive pathways .
We previously performed an extensive transcriptome analysis in order to define the scope of the S . aureus WalKR regulon [6] . Our results showed that expression of the spdC gene ( SAOUHSC_02611 ) was increased 3 . 5-fold in a S . aureus strain producing a constitutively active form of the WalR response regulator ( D55E ) [6] . SpdC , a membrane-anchored protein with 8 predicted transmembrane segments and an Abi domain ( CAAX protease self-immunity ) , ( Fig 1A ) , was previously reported as playing a role in the display of surface proteins such as protein A [31] . In order to confirm that spdC is a member of the WalKR regulon , we used quantitative real-time PCR ( qRT-PCR ) to measure its expression in a S . aureus strain where the walRKHI operon is placed under the control of the IPTG-inducible Pspac promoter [9] . Cells were grown overnight in TSB with 0 . 05 mM IPTG , and cultures were inoculated at OD600nm = 0 . 05 , with or without different IPTG concentrations ( 0 . 05 and 1 mM ) to induce expression from the Pspac promoter . RNA samples were prepared from exponentially growing cells harvested at OD600nm = 0 . 5 , before cessation of growth of the culture lacking IPTG , and walR and spdC mRNA levels were measured by qRT-PCR . As shown in Fig 1B , walR transcription is increased approximately 4-fold when cells are grown with 0 . 05 mM IPTG and 8-fold at 1 mM IPTG . Under the same conditions , spdC expression followed that of walR , and was increased 2-fold and 5-fold with 0 . 05 or 1 mM IPTG , respectively ( Fig 1B ) , confirming positive regulation by the WalKR system . As shown above , the WalKR TCS controls spdC expression . In Streptococcus agalactiae , another bacterial Abi-domain protein , Abx1 , has been shown to inhibit activity of the CovSR two-component system [25] . In order to determine whether SpdC has a regulatory role in S . aureus , we generated a ΔspdC mutant in strain HG001 and performed a comparative RNA-Seq analysis . The ΔspdC mutant did not display any gross morphological changes or growth defects . Indeed , although it had a slight lag during the first hour post inoculation , the growth rate and final OD600nm were not significantly different from those of the parental strain ( S1 Fig ) . Doubling times ( http://www . doubling-time . com/compute . php ) calculated during the exponential growth phase ( S1 Fig , 90 min to 210 min ) gave identical values of 32 min for both strains . The HG001 strain and ΔspdC mutant were grown in TSB until early exponential phase ( OD600nm = 1 ) and total RNA was extracted for RNA-Seq analysis ( See Materials and Methods ) . We verified that spdC is well expressed under these conditions using a lacZ reporter fusion with the spdC promoter region ( S2 Fig ) . Three biological replicates were analyzed by RNA-Seq for each strain and the data presented as the mean fold-change . Using a value cut-off greater than 2 with a P value less than 0 . 05 , we found that the expression of 42 genes was lowered in the ΔspdC mutant strain and that of 65 increased ( Table 1 ) . In order to perform a general analysis of the transcriptomic data we generated an ontological grouping of SpdC-regulated genes ( Fig 2A ) . Among the genes positively controlled by SpdC , 10 are known virulence factors , suggesting that SpdC may influence S . aureus pathogenicity , and 11 are involved in capsular biosynthesis . The S . aureus capsule is known to impede phagocytosis and promote host colonization , however although the HG001 strain used in this study carries the serotype 5 capsule gene cluster , a missense mutation in the cap5E gene prevents capsular biosynthesis [32–34] . Expression of several S . aureus prophage genes was also increased in the ΔspdC mutant strain: 11 for phage Φ13 and 17 for Φ12 , ( Table 1 ) . As shown above , spdC expression is controlled by the WalKR system . The RNA-Seq data analysis of SpdC-regulated genes reveals that 25 of these belong to the WalKR regulon ( indicated by an asterisk in Table 1 ) . In particular , the expression of 4 WalKR-dependent cell wall hydrolase genes ( sceD , ssaA , lytM and atlA ) is increased in the ΔspdC mutant , suggesting that SpdC negatively controls WalKR activity ( Table 1 ) . It is interesting to note that among the genes positively controlled by SpdC , many are also WalKR-activated genes ( Table 1 ) . All of these are classified as virulence genes , however they are not preceded by the WalR consensus binding site , and we have previously shown that several of these are not directly regulated by the WalKR system but through the SaeRS two-component system instead [6] . We used qRT-PCR to verify SpdC-dependent regulation for 3 positively ( spa , hlgC , sdrD ) and 3 negatively ( atlA , sceD , lytM ) controlled genes , in the ΔspdC strain compared to the HG001 parental strain , grown under the same conditions as for the RNA-Seq analysis ( Fig 2B ) . We observed a perfect correlation with the RNA-Seq data: expression of the spa gene encoding protein A was very strongly lowered in the ΔspdC mutant strain ( about one hundred-fold ) , while sdrD and hlgC expression levels were 5- to 10-fold less . The atlA , sceD and lytM cell wall hydrolase genes were more highly expressed in the absence of SpdC ( 2- , 12- and 5-fold , respectively ) in agreement with the RNA-Seq analysis ( Fig 2B ) . In order to confirm SpdC-dependent regulation at the protein level , we chose two genes that were positively or negatively controlled by SpdC , spa and lytM , respectively , and performed Western blot analyses . Whole cell extracts were prepared from cultures of strains HG001 , the ΔspdC mutant and the complemented mutant strain ( ΔspdC/pMK4Pprot-spdC ) and subjected to SDS-PAGE and immunoblotting . As shown in the top panel of Fig 2C , LytM levels are higher in the ΔspdC strain compared to the parental strain ( lane 2 ) , and in the complemented strain the LytM level is reduced to a level lower than in the parental strain ( lane 3 ) , likely reflecting the higher production of SpdC in the complemented strain . Indeed , under these conditions , spdC mRNA levels were increased more than 100-fold as measured by qRT-PCR as compared to the parental HG001 strain . For studying levels of protein A , known to be covalently anchored to the cell wall ( LPxTG sortase motif ) , identical quantities of cell wall fractions of the HG001 parental strain , ΔspdC deletion mutant , and complemented strain ( ΔspdC/pMK4Pprot-spdC ) were subjected to SDS/PAGE and compared by Western blot . As expected , protein A levels were significantly lower in the ΔspdC mutant than in the parental and complemented strains ( Fig 2C , lower panel ) , in agreement with the RNA-Seq and qRT-PCR results . These data indicate that SpdC is involved in controlling gene expression through potential interactions with regulatory systems , and the WalKR two-component system in particular . As shown above , we have uncovered a regulatory link indicating that SpdC negatively controls activity of the WalKR two-component system , strongly suggesting that the proteins interact . In order to test possible interactions between SpdC and the WalKR proteins , we used the bacterial adenylate cyclase two-hybrid system ( BACTH ) [35] . We fused the full-length membrane-bound WalK histidine kinase or the WalR cytoplasmic response regulator to the C-terminal domain of the T25 subunit of the Bordetella pertussis adenylate cyclase and full-length SpdC to the C-terminal domain of the T18 subunit , using plasmids pKT25 and pUT18c respectively . To probe putative interactions , E . coli strain DHT1 was co-transformed with combinations of the pKT25 and pUT18c derivatives carrying the translational fusions . Upon protein-protein interactions , the close proximity between the T18 and T25 subunits restores adenylate cyclase activity , leading to cAMP synthesis and activation of the lactose operon . Interactions were tested both by spotting the resulting strains on LB plates containing X-Gal and by measuring β-galactosidase activity . To determine pair-wise interactions , we chose a cut-off value of 100 Miller Units as indicating a positive interaction between the protein fusions . As shown in Fig 3A , strong β-galactosidase activity was only observed for the plasmid combination co-producing the membrane anchored proteins SpdC and WalK while no interactions between SpdC and the cytoplasmic regulator WalR could be detected . SpdC is annotated as an Abi domain protein ( CAAX protease self-immunity ) in genome databases . The S . aureus HG001 genome encodes 4 Abi domain proteins , three of which , SpdA ( SAOUHSC_01900 ) , SpdB ( SAOUHSC_02587 ) , and SpdC , have been reported as being involved in surface protein display , whereas the fourth ( SAOUHSC_02256 ) has no known function [31] . In order to test whether WalK also interacts with the other three Abi domain proteins , we constructed translational fusions for the remaining Abi proteins with the T18 domain of adenylate cyclase . As shown in Fig 3A , the combinations of SpdA , SpdB or SAOUHSC_02256 with WalK did not generate significant levels of β-galactosidase activity , demonstrating that SpdC is the only S . aureus Abi domain protein specifically interacting with WalK . We have previously shown that the S . aureus WalK histidine kinase is mainly localized at the division septum [30] . Since SpdC and WalK interact , we studied the subcellular localization of SpdC in S . aureus by constructing a translational fusion with the GFP fluorescent protein using the pOLSA vector ( See Materials and Methods ) , under the control of the cadmium-inducible Pcad promoter . The resulting plasmid , pOLSA-spdC was then introduced into the HG001 strain . Expression of the gene fusion was induced by addition of CdCl2 ( 0 . 25 μM ) , cells were harvested during exponential growth ( OD600nm = 1 . 5 ) and observed by fluorescence microscopy . As shown in Fig 3B , SpdC is preferentially localized at the division septum , with a mean septum/membrane fluorescence ratio of around 2 . 6 . Taken together , our results indicate that SpdC and WalK interact and are localized at the division septum . As shown above , the expression of several cell wall hydrolase genes is significantly increased in the ΔspdC mutant strain , suggesting that sensitivity to compounds targeting the cell wall might also be affected . We followed bacterial lysis during incubation in the presence of a non-anionic detergent , Triton X-100 , thought to trigger cell lysis by favoring endogenous autolysin activity [36] . However , Triton X-100 induced lysis for the HG001 and ΔspdC mutant strains was not significantly different ( S3 Fig ) . We also tested sensitivity to lysostaphin , a glycyl-glycine endopeptidase that hydrolyzes the peptidoglycan pentaglycine interpeptide crossbridge , leading to cell lysis [37] . The HG001 , ΔspdC mutant and complemented strains were grown in TSB until OD600nm ≈ 1 and cells were then harvested and resuspended in PBS in the presence of lysostaphin . As shown in Fig 4A , the ΔspdC mutant was less sensitive to lysostaphin-induced lysis than the parental HG001 and complemented ΔspdC strains , suggesting that the absence of SpdC leads to cell wall modifications . We then tested sensitivity to antibiotics targeting the cell wall . As shown in Fig 4B the ΔspdC mutant displayed increased sensitivity to the β-lactam antibiotic oxacillin , whereas the parental and complemented strains were able to grow at the concentration tested ( 0 . 1 μg/ml ) . No difference in sensitivity between the strains was seen using fosfomycin , an antibiotic inhibiting MurA , which catalyzes the very first step of peptidoglycan biosynthesis ( Fig 4B ) . These results suggested that the ΔspdC mutant strain may either be affected in the later steps of peptidoglycan biosynthesis or may exhibit a cell wall structure modification leading to a difference in accessibility of antibiotics acting extracellularly . Wall teichoic acids ( WTAs ) are anionic sugar rich cell surface polymers that can alter accessibility to the cell wall . We therefore tested resistance to tunicamycin , an antibiotic targeting biosynthesis of WTAs . As shown in Fig 4B the ΔspdC mutant strain was highly sensitive to tunicamycin , in contrast to the parental and complemented strains . Taken together , these results suggest that the S . aureus cell envelope structure is altered in the absence of SpdC . Since our results indicate that the ΔspdC mutation may modify the S . aureus cell surface , we tested whether the absence of SpdC may have an effect on biofilm formation . Strains were grown statically in TSB , supplemented with glucose and NaCl , for 24 h at 37°C in PVC microplates . As shown in Fig 5 , biofilm formation was strongly decreased in the absence of SpdC ( approximately 7-fold ) . Complementation of the ΔspdC mutant with the pMK4Pprot-spdC plasmid restored biofilm formation to levels comparable to those of the parental HG001 strain ( Fig 5 ) . These results are consistent with a modification of the S . aureus cell surface in the absence of SpdC , which could influence resistance against antimicrobial compounds targeting cell surface structures as well as the capacity to form biofilms . In order to determine which biofilm component is affected , biofilm detachment experiments were carried out ( S4 Fig ) by treatment with proteinase K , DNaseI and sodium metaperiodate ( a carbohydrate-modifying agent ) . Under our conditions , biofilm production was lowered three-fold after treatment with DNaseI , and more than 10-fold when treated with proteinase K , but not significantly modified after treatment with sodium metaperiodate . Thus , biofilms formed under our conditions by the HG001 parental strain are essentially protein-based , and , to a lesser extent , due to extracellular DNA . We observed reduced biofilm formation for the ΔspdC mutant even after DNAseI treatment , but not after proteinase K treatment ( S4 Fig ) , suggesting that SpdC affects the production of proteins important for biofilm formation . Cell surface modifications are known to impact virulence . Likewise , the capacity to form robust biofilms favors bacterial colonization of the host . Additionally , our RNA-Seq analysis revealed that the expression of at least 10 genes directly involved in bacterial virulence upon infection was lowered in the ΔspdC mutant , strongly suggesting that SpdC plays a role in virulence . We used a murine sepsis model to compare virulence of the HG001 and ΔspdC strains . SWISS mice were infected intravenously with 5 . 107 cfu and mortality was monitored over 9 days post infection . As shown in Fig 6 virulence of the ΔspdC mutant was strongly diminished . Indeed , following infection with the HG001 parental strain , significant mortality occurred in the first 3 days post-infection ( greater than 60% ) , whereas only a single mouse out of 14 died within five days after infection with the ΔspdC mutant . After the sixth day , a moderate mortality was observed for the group infected with the ΔspdC mutant , with only 36% mortality at the end of the assay ( compared to 72% mortality for mice infected with the parental HG001 strain ) . This significant difference indicates that SpdC is a novel virulence factor in S . aureus . As shown above , SpdC localizes at the division septum and interacts with the WalK histidine kinase , negatively controlling WalKR activity . However , many of the SpdC-regulated genes identified by RNA-Seq do not belong to the WalKR regulon , but are known to be controlled by other two-component systems ( Table 1 ) suggesting SpdC may interact with other TCS regulatory pathways . The S . aureus HG001 genome encodes 16 two-component systems [33] and we constructed translational fusions for each of the histidine kinase genes with the carboxy-terminal region of the adenylate cyclase T25 domain . Each of the resulting plasmids was co-transformed in combination with the pUT18c-spdC plasmid into E . coli strain DHT1 . As shown in Fig 7 , in addition to WalK , we detected interactions between SpdC and nine additional histidine kinases: YesM , GraS , SaeS , DesK , ArlS , SrrB , PhoR , VraS , and BraS . These interactions appear to be specific , since no interactions were detected between SpdC and the remaining six histidine kinases , as shown in Fig 7 ( LytS , AirS , AgrC , KdpD , HssS , and NreB ) . Among the genes positively regulated by SpdC ( Table 1 ) , the expression of at least fifteen is also activated by the SaeSR TCS , including the spl operon , sak , the hlgBC operon and chp [7 , 38–40] . This suggests that SpdC activates the SaeSR TCS , in contrast to its role in negatively controlling activity of the WalKR system . Histidine kinases have different combinations of signaling domains such as HAMP or PAS domains [41] in addition to the conserved H , N , G1 , F and G2 boxes of the phosphoacceptor/dimerization ( HisKA ) and catalytic ( HATPase_C ) domains [42–44] . We have shown that SpdC interacts with 10 of the 16 S . aureus histidine kinases , which do not share any strong amino acid sequence similarities other than the conserved histidine kinase HisKA and HATPase_C domains . Since SpdC interacts with some , but not all of the S . aureus histidine kinases , this protein-protein contact must involve some other domain . We focused our analysis on the WalK and SaeS histidine kinases . WalK has two transmembrane domains ( amino acids 14–34 and 183–203 ) , with an extracellular loop of 148 amino acid residues , a HAMP domain involved in signal transduction ( 204–256 ) , a PAS domain ( 261–331 ) a PAC domain ( 314–378 ) and a histidine kinase domain ( 382–600; Fig 8A ) . In contrast , SaeS is a member of the intra-membrane sensing kinases [45] , with two transmembrane domains ( amino acids 9–29 and 40–60 ) separated by only ten amino acids , as well as a HAMP domain ( 61–114 ) and a histidine kinase domain ( 129–348; Fig 8A ) . We tested the interactions of SpdC with the N-terminal domains of WalK and SaeS containing the transmembrane regions ( WalK1-203 and SaeS1-64 , respectively ) . The truncated proteins were fused to the T25 domain of adenylate cyclase , and the resulting plasmids were co-transformed into E . coli strain DHT1 together with the pUT18c-spdC plasmid . As shown in Fig 8B , the first 203 amino acids of WalK are sufficient to allow stable interactions with SpdC . For SaeS , the N-terminal domain containing only the two transmembrane segments ( SaeS1-64 ) did not lead to interaction with SpdC , but a longer fragment of the protein ( SaeS1-120 ) gave rise to a stable interaction with SpdC and high β-galactosidase activity ( Fig 8B ) . These results suggest that WalK and SaeS interact with SpdC through their transmembrane domains . The negative interaction results obtained with the truncated SaeS protein containing only the transmembrane domains suggest that since this kinase lacks an extracellular loop , the HAMP domain may be required for proper membrane insertion of the fusion protein . In order to identify which domain of SpdC interacts with the WalK and SaeS kinases , we compared interactions with full-length SpdC ( pUT18c-spdC ) and a carboxy-terminal truncated SpdC consisting only of the eight transmembrane domains ( SpdC1-252; pUT18c-spdC1-252 ) . We noted self-interaction of SpdC following co-transformation of pKT25-spdC with either pUT18c-spdC or pUT18c-spdC1-252 ( Fig 8B ) , indicating that the SpdC transmembrane domains are involved in these self-interactions . Similar results were obtained when testing interactions with WalK and SaeS , i . e . the transmembrane domains of SpdC are sufficient to allow interactions with the histidine kinases ( Fig 8B ) . Taken together , these results indicate that SpdC is a membrane-bound protein that interacts with itself and several histidine kinases through transmembrane domain contacts .
Abi-domain proteins constitute a large family whose functions are mostly unknown . SpdC was initially designated LyrA for Lysostaphin resistance A and identified by screening a bursa aurealis transposon mutant library for increased lysostaphin resistance [46] . An independent study aiming at characterizing proteins involved in the display of surface proteins led to the identification of three proteins , SpdA , SpdB and SpdC ( Surface protein display A , B and C ) , playing a role in protein A levels at the staphylococcal cell surface [31] . These proteins share a similar structural organization with 6 to 8 transmembrane domains and an Abi-domain embedded within the hydrophobic region . An additional protein with a similar organization is encoded by the S . aureus genome , SAOUHSC_02256 , but appears to have no role in controlling protein A levels [31] . We previously characterized the essential WalKR two-component system in S . aureus and highlighted its major role in controlling cell wall homeostasis [9 , 47] . Transcriptome analysis revealed that spdC expression is positively controlled by WalKR [6] . We show here that the SpdC membrane protein and the WalK histidine kinase interact and that SpdC negatively controls WalKR activity and expression of WalKR-regulated genes . Interaction of the WalK histidine kinase with SpdC is specific since no interaction was seen with the other Abi-proteins ( SpdA , SpdB and SAOUHSC_02256 ) . Accordingly , we also showed that SpdC and WalK are both localized preferentially at the division septum . Interestingly , an RNA-Seq analysis of a ΔsdpC mutant revealed that the expression of 107 genes varied compared to the parental strain . Among these , 24 ( more than 20% ) are controlled by the WalKR system . Since SpdC appears to negatively control WalKR activity by interacting with WalK , this septal localization is consistent with the previously suggested cell wall metabolism-related activation signal of the WalK histidine kinase in Bacillus subtilis [11 , 48] . Indeed , in cocci , cell wall synthesis has been shown to exclusively occur at the division septum in an FtsZ-dependent manner [49] , suggesting that a peptidoglycan metabolism related signal at the septum may relieve negative control of WalK activity by SpdC . Histidine kinases often act as phosphoprotein phosphatases towards their associated response regulator . WalK was previously classified as a kinase/phosphatase «bifunctional sensor» [23] and the PAS domain of Streptococcus pneumoniae WalK plays a role in its phosphatase activity [50 , 51] . In S . aureus , we have previously shown that WalK acts as a WalR phosphoprotein phosphatase upon entry into stationary phase in order to shut off WalR activity [6] . Interactions between SpdC and WalK can either interfere with signal perception by the sensor histidine kinase , inhibit its kinase activity or increase its phosphatase activity towards WalR , thus negatively controlling WalKR-dependent gene expression . SpdC is unlikely to directly regulate gene expression since it is a membrane protein lacking any typical DNA-binding domain . To understand how the other SpdC-dependent genes were controlled ( 83 of the genes identified by RNA-Seq are not regulated by the WalKR system ) , we tested interactions of SpdC with the other S . aureus histidine kinases and found that it interacts with 10 of the 16 encoded in the genome . No obvious structural motifs or domain sequences were common to those that interacted with SpdC compared to those that did not . Two histidine kinases , WalK and SaeS , were chosen for further analysis of their interactions with SpdC . Our results indicate that the two transmembrane domains of WalK are sufficient to allow interaction with SpdC , whereas a greater amino-terminal fragment of SaeS was required , encompassing the HAMP domain . This result suggests that although the transmembrane domains of SaeS are likely involved in interactions with SpdC , a longer fragment may be necessary to ensure proper membrane insertion of the truncated protein . We also showed that a truncated form of SpdC , containing only the eight transmembrane domains , was sufficient for self-interaction and interaction with WalK and SaeS , indicating that transmembrane domains are involved in the interactions between SpdC and the histidine kinases . Interestingly , the only two S . aureus cytoplasmic histidine kinases which lack transmembrane domains , AirS and NreB , did not interact with SpdC under our conditions , in agreement with our results indicating transmembrane segments are involved in the interactions . The only other example of an Abi domain protein interacting with a histidine kinase is Abx1 of Streptococcus agalactiae , which interacts with the CovS kinase [25] . The two transmembrane domains of CovS were shown to be necessary and sufficient for these interactions [25] . At least fifteen genes belonging to the SaeSR regulon , including the spl operon , sak , the hlgBC operon and chp [7 , 38–40] , are also positively controlled by SpdC ( Table 1 ) , indicating that SpdC likely activates the SaeSR TCS , in contrast to its role in negatively controlling activity of the WalKR system . Since WalR controls spdC expression , this is consistent with our previous results showing that constitutive activation of WalR generates a signal leading to activation of the SaeSR TCS and a corresponding increase in SaeSR regulon expression [6] . The localization of SpdC at the division septum and its role in gene regulation through interactions with sensor kinases of two-component systems led us to speculate that SpdC may interfere with bacterial division sensing and impact cell wall metabolism . Accordingly , the ΔspdC mutant displays increased resistance against lysis when treated with lysostaphin , in agreement with the original phenotype characterized by transposon insertion [46] . Additionally , the absence of SpdC was reported to lead to increased cross wall abundance and thickness [31] . We tested sensitivity of the ΔspdC mutant to antibiotics targeting the cell wall . The ΔspdC mutant is highly sensitive to oxacillin and tunicamycin , but not to fosfomycin , which inhibits the first step of cell wall biosynthesis . In agreement with the sensitivity of the ΔspdC mutant to tunicamycin , which inhibits wall teichoic acid synthesis , the spdC gene was also identified as a candidate using a screen to identify synthetically lethal mutations with teichoic acid biosynthesis defects [52] . We have shown that expression of cell wall hydrolase genes is increased in the ΔspdC strain ( sceD , ssaA , lytM , atlA ) . This may lead to increased cell wall degradation , which could explain the lowered resistance to oxacillin . This could also lead to the mutant’s increased sensitivity to tunicamycin . Indeed , teichoic acids are key elements for the proper localization of AtlA to the division septa , where cell wall biosynthesis takes place , since cell wall plasticity is essential for cell division [53] . In the presence of tunicamycin , the absence of wall teichoic acids results in a delocalized distribution of AtlA across the cell surface . Since atlA is more highly expressed in the ΔspdC stain , this could explain why this strain is more sensitive to tunicamycin . Taken together with the strong links to the WalKR TCS , these results indicate that SpdC is involved in bacterial cell envelope homeostasis . The importance of the bacterial cell envelope in host-pathogen interactions cannot be over-emphasized: it is the first layer of contact between the bacterium and its host , containing an array of cell wall-linked or associated toxins and virulence factors , the first and major bacterial line of defense against threats from the host or environment , and is also both the target of choice for antibiotic treatment and the source of many antibiotic-resistance pathways . We have shown that SpdC is required for biofilm formation , an important step in S . aureus pathogenesis . We also showed that SpdC is a novel S . aureus virulence gene , required for the infectious process in a mouse septicemia model . The loss of virulence observed with the ΔspdC mutant may involve both its diminished capacity to form biofilms as well as the lowered expression of multiple virulence genes as shown by RNA-Seq analysis . The regulatory mechanism mediated by SpdC remains to be determined . Abi-domain-containing proteins have been extensively studied in eukaryotes . They are involved in CAAX-protein maturation by cleaving the C-terminal AAX tripeptide after addition of an isoprenyl group on a cysteine , the last step consisting in methylation of the new C- terminus . These three modification steps are termed prenylation [54] . This post-translational maturation has a crucial role in maintaining cellular homeostasis by controlling the localization and activity of a large range of proteins . In particular , by adding a lipid group at the carboxy-terminal end of proteins , it favors their interactions with membranes , which have a high concentration of signaling molecules [55] . Prenylation has been recently described in prokaryotes and a geranyltransferase , IspA , has been identified in S . aureus [56] . Putative methyltransferase and CAAX-protease encoding genes ( including spdC ) are also present in the S . aureus genome . Deletion of the ispA gene has pleiotropic effects such as a growth defect , increased sensitivity to oxidative stress and an altered cell envelope [56] . Of note , the absence of IspA or SpdC both lead to increased cell wall antibiotic sensitivity . The RNA-Seq transcriptome analysis of the ΔispA strain [56] shows similarities with that of the ΔspdC mutant . Indeed , one of the most regulated genes in both cases is spa , encoding the immunoglobulin G binding protein A . We also noted that 21 Φ13 genes are up-regulated in the ΔispA mutant whereas we characterized 11 Φ13 genes up-regulated in the ΔspdC mutant , with several in common . These data suggest that SpdC and IspA could be involved in the same cellular pathway . Interestingly , the transcriptome analysis of the ΔispA mutant revealed modified expression of a large number of genes involved in regulatory circuits and particularly increased expression of the walR , walH , walI and walJ genes of the wal locus [56] . While no direct link between prenylation and Abi domain proteins has been shown in prokaryotes , there are several links with the WalKR system , either by physical interaction and negative control of activity , for SpdC , or at the transcriptional level for the IspA geranyltransferase . This study identifies a membrane-bound protein with an Abi domain , SpdC , at the core of an interaction network that coordinates bacterial division with cell envelope metabolism and host interactions . Further studies are required in order to decipher the molecular mechanism and consequences of these interactions .
Escherichia coli K12 strain DH5α ( Invitrogen , Thermo Fisher Scientific , Waltham , MA ) was used for cloning experiments . Staphylococcus aureus strain HG001 [57] was used for genetic and functional studies . Plasmids were first passaged through the restriction deficient S . aureus strain RN4220 before introduction into the HG001 strain . E . coli strains were grown in LB medium with ampicillin ( 100 μg/ml ) added when required . S . aureus strains and plasmids used in this study are listed in Table 2 . S . aureus strains were grown in Trypticase Soy Broth ( TSB; Difco; Becton , Dickinson and Co . , Franklin Lakes , NJ ) supplemented with chloramphenicol ( 10 μg/ml ) or erythromycin ( 1 μg/ml ) as required . E . coli and S . aureus strains were transformed by electroporation using standard protocols [58] and transformants were selected on LB or Trypticase Soy Agar ( TSA; Difco ) plates , respectively , with the appropriate antibiotics . Expression from the Pcad promoter was induced by adding cadmium chloride ( CdCl2 ) at a final concentration of 0 . 25 μM . Expression from the Pspac promoter was induced by addition of isopropyl β-D-1-thiogalactopyranoside ( IPTG ) . Oligonucleotides used in this study were synthesized by Eurofins Genomics ( Ebersberg , Germany ) and their sequences are listed in Table 3 . S . aureus chromosomal DNA was isolated using the MasterPure Gram-positive DNA purification Kit ( Epicentre Biotechnologies , Madison , WI ) . Plasmid DNA was isolated using a QIAprep Spin Miniprep kit ( Qiagen , Hilden , Germany ) and PCR fragments were purified using the Qiaquick PCR purification kit ( Qiagen ) . T4 DNA ligase and restriction enzymes , PCR reagents and Q5 high-fidelity DNA polymerase ( New England Biolabs , Ipswich , MA ) were used according to the manufacturer's recommendations . Nucleotide sequencing of plasmid constructs was carried out by Beckman Coulter Genomics ( Danvers , MA ) . For construction of the ΔspdC mutant strain , two 800 bp DNA fragments were generated by PCR using oligonucleotide pairs OP375/OP376 and OP377/OP378 , respectively ( see Table 3 ) , corresponding to the DNA regions located immediately upstream and downstream from the spdC gene . These DNA fragments were cloned in tandem in two consecutive steps , between the BamHI and NcoI restriction sites of the pMAD vector . The resulting plasmid was introduced by electroporation into S . aureus and transformants were selected at 30°C on TSA plates containing erythromycin ( 1 μg/ml ) and 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside ( X-Gal , 100 μg/ml ) . Integration and excision of the plasmid were then performed as previously described [59] , yielding mutant strain ST1317 ( ΔspdC ) . A complementation plasmid pMK4Pprot-spdC was constructed by cloning the entire spdC coding sequence ( amplicon OP404/OP405 ) in plasmid pMK4Pprot , under the control of the constitutive Pprot promoter [60] . The plasmid was introduced into the ST1317 ΔspdC mutant , generating the ST1375 complemented strain . Expression of the walKRHI in strain HG001 was placed under the control of the IPTG-iducible Pspac promoter by Φ80α phage transduction [61] using strain ST1000 ( RN4220 PspacwalRKHI; [9] ) as a donor and strain HG001 as the recipient , yielding strain ST1017 ( HG001 PspacwalRKHI ) . Plasmid pSA14 [62] is a derivative of shuttle vector pMK4 [63] , carrying a promoterless E . coli lacZ gene and was used to construct transcriptional lacZ reporter fusions . The spdC promoter region was amplified by PCR using oligonucleotides OSA512/OSA513 ( see Table 3 ) and cloned between the PstI/BamHI restriction sites of the pSA14 vector , yielding plasmid pSD3-41 ( Table 2 ) . For β-galactosidase assays in S . aureus , strain ST1386 carrying the spdC’-lacZ fusion was grown in TSB at 37°C and cells were harvested by centrifuging 2 ml culture samples ( 2 min; 5 , 400 x g ) . Assays were performed as previously described [21] and β-galactosidase specific activities expressed as Miller units mg−1 protein [64] . Protein concentrations were determined using the Bio-Rad protein assay ( BioRad , Hercules , CA ) [65] . Strains were grown in TSB , supplemented with IPTG when specified , at 37°C with aeration until OD600nm = 1 . Cells were pelleted by centrifugation ( 2 min , 20 , 800 x g ) and immediately frozen at -20°C . RNA extraction was then performed as previously described [66] , followed by DNaseI treatment with the TURBO DNA-free reagent ( Ambion , Austin , TX ) in order to eliminate residual genomic DNA . cDNA synthesis was carried out as previously described [47] . Oligonucleotides were designed with the BEACON Designer 7 . 91 software ( Premier Biosoft International , Palo Alto , CA ) in order to synthesize 100–200 bp amplicons ( see Table 3 ) . Quantitative real-time PCRs ( qRT-PCRs ) , critical threshold cycles ( CT ) and n-fold changes in transcript levels were performed and determined as previously described using the SsoFast EvaGreen Supermix ( Bio-Rad , Hercules , CA ) and normalized with respect to 16S rRNA whose levels did not vary under our experimental conditions [47] . All assays were performed using quadruplicate technical replicates , and repeated with three independent biological samples . Three independent biological replicates were used for RNA-Seq analysis of the parental HG001 and ΔspdC strains . Strains were grown in TSB until OD600nm = 1 . Total RNA was isolated as described above , and 7 μg were treated using the MicrobExpress kit ( Ambion , Austin , TX ) in order to remove rRNA . The rRNA depleted fraction was used for construction of strand specific single end cDNA libraries using the Truseq Stranded Total RNA sample prep kit according to the manufacturer’s instructions ( Illumina , San Diego , CA ) . Libraries were sequenced using an Illumina Hiseq2000 sequencer ( multiplexing 6 samples in one lane ) according to the manufacturer’s instructions ( Illumina , San Diego , CA ) . Sequences were demultiplexed using the Illumina alignment and sequence analysis pipeline ( GERALD , included in CASAVA version 1 . 7 ) giving FASTQ formatted reads . Reads were cleaned by removing adapter and low quality sequences using an in-house program ( https://github . com/baj12/clean_ngs ) . Only sequences with a minimum length of 25 nucleotides were considered for further analysis . Bowtie ( version 0 . 12 . 7 , -m50—chunkmbs 400 -a—best -q -e50 ) [67] was used for alignment with the reference Staphylococcus aureus subsp . aureus genome ( gi|88193823 ) . Only uniquely aligning reads where considered for counting . HTseq-count ( version 0 . 5 . 4p5 , parameters: -m intersection-nonempty , -s yes , -t CDS -I locus_tag ) was used for counting genes [68] . Statistical analysis was performed with R version 3 . 0 . 2 [69] and DESeq2 version 1 . 2 . 10 [70] . Data were first normalized with DESeq2 and the default parameters . Dispersion estimation and statistical testing were performed using the Generalized Linear Model with default parameters . Independent filtering was performed with default parameters to exclude transcripts with very low count values . Raw P-values were then adjusted according to the Benjamini and Hochberg procedure [71] and transcripts were considered differentially expressed when their adjusted P-value was lower than 0 . 05 . For testing protein interactions using the Bacterial Adenylate Cyclase Two-Hybrid System ( BACTH ) , genes encoding the proteins of interest were cloned into plasmids pKT25 and pUT18c leading to translational fusions with the T25 or T18 domains of the Bordetella pertussis adenylate cyclase [35] . DNA fragments corresponding to the coding sequences were amplified by PCR using chromosomal DNA from strain HG001 and specific oligonucleotide pairs ( see Table 3 ) . Fragments were digested with BamHI and EcoRI or KpnI ( indicated in Table 3 ) for cloning into plasmids pKT25 or pUT18c . The resulting plasmids were co-transformed into E . coli strain DHT1 [72] to detect protein-protein interactions and transformants were selected on kanamycin ( 50 μg/ml ) for pKT25 derivatives and ampicillin ( 100 μg/ml ) for pUT18c derivatives . The resulting strains carrying combinations of pKT25 and pUT18c derivatives were tested for cyclic AMP-dependent activation of lacZ expression . For tests on plates , strains were grown in LB liquid medium supplemented with ampicillin ( 100 μg/ml ) and kanamycin ( 50 μg/ml ) . Overnight cultures were then spotted on LB-agar plates containing IPTG ( 0 . 5 mM ) , ampicillin ( 100 μg/ml ) , kanamycin ( 50 μg/ml ) , and X-Gal ( 100 μg/ml ) . Plates were incubated for 24 H at 30°C and examined for appearance of the characteristic blue color indicative of β-galactosidase activity through X-Gal hydrolysis . Quantitative β-galactosidase assays were performed on exponentially growing E . coli liquid cultures . Cells were grown in LB with IPTG ( 0 . 5 mM ) , ampicillin ( 100 μg/ml ) and kanamycin ( 50 μg/ml ) at 30°C under aeration until OD600nm = 1 and assays performed on SDS/chloroform permeabilized cells as previously described [64] . Enzymatic activities are represented relative to negative and positive controls , respectively a strain carrying the empty pKT25 and pUT18c vectors ( activity = 0 , arbitrary unit ) , and a strain with the pKT25-zip and pUT18c-zip plasmids [35] ( activity = 1000 , arbitrary unit ) . The pOLSA plasmid was used to produce a fluorescent SpdC-GFP fusion protein [30] . The translational fusion was constructed by PCR amplification using HG001 chromosomal DNA and oligonucleotide pair OSA417/OSA404 ( Table 3 ) . The amplicon was cloned into pOLSA between the SalI and XmaI restriction sites , allowing transcription from the Pcad promoter and production of the SpdC-GFP fusion protein . Subcellular protein localization of SpdC was performed in S . aureus HG001 transformed with pOLSA-spdC . Fluorescence microscopy was carried out on cells grown in liquid cultures in TSB supplemented with CdCl2 ( 0 . 25 μM ) to induce gene fusion expression . When cells reached OD600nm ≈ 1 . 5 ( exponential growth phase ) , they were harvested and concentrated 20 times in PBS . Cell suspensions were mixed with Vectashield mounting media ( Vector Laboratories , Burlingame , CA ) and used for microscopic observations with a Nikon Eclipse E600 . Images were acquired with a Nikon DXM1200F Digital Camera . ImageJ software was used for quantifying fluorescence ( http://imagej . nih . gov/ij/index . html; [73] ) . Fluorescence ratios were calculated by measuring fluorescence at the division septa divided by the fluorescence at the lateral wall after subtracting background fluorescence . Quantification was performed for 33 cells and two independent biological replicates and plotted using GraphPad Prism ( GraphPad Software , San Diego , CA; http://www . graphpad . com ) . The HG001 , ΔspdC and ΔspdC/pMK4Pprot-spdC strains were grown overnight at 37°C with aeration in TSB medium , with chloramphenicol ( 10 μg/ml ) when required . Bacterial suspensions diluted from 10−2 to 10−7 were spotted ( 3 μl ) onto TSA plates with the indicated antibiotic concentrations and incubated for 15 hours at 37°C . S . aureus whole cell lysates were prepared as previously described [47] . Briefly , 5 ml of a cell culture grown to stationary phase were harvested by centrifugation ( 10 min; 3 , 000 x g ) , resuspended in 2X Laemmli SDS sample buffer ( 0 . 2 ml ) and heated at 99°C for 10 min . Supernatants containing SDS-soluble proteins were collected following centrifugation ( 10 min; 20 , 800 x g ) , and used for further analysis . Cell wall extracts were prepared from 50 ml of the same cultures . Cells were pelleted and resuspended in 4 ml of digestion buffer ( 50 mM Tris-HCl pH 8 , 145 mM NaCl , 30% sucrose , 160 ng/ml DNaseI , 250 μg/ml lysostaphin ) and incubated for 60 min at 37°C . Supernatants corresponding to cell wall extracts were then harvested by centrifugation ( 10 min; 3 , 000 x g ) . Protein extracts were separated by SDS-PAGE on a 12% polyacrylamide gel , followed by Coomassie Brilliant Blue staining to verify that the quality and quantity of loaded extracts was equivalent for the different strains . For immunoblotting experiments , protein extracts were transferred to a nitrocellulose membrane after SDS-PAGE using a semi-dry blotter ( Bio-Rad , Hercules , CA ) and the following buffer: 25 mM Tris , 192 mM glycine , 20% ethanol . The LytM protein was detected using a purified polyclonal rabbit antibody [74] and horseradish peroxidase-coupled anti-rabbit secondary antibodies ( Zymed , ThermoFisher , Waltham , MA ) and the Pico chemiluminescence Western blot kit ( Pierce , ThermoFisher , Waltham , MA ) . Detection of Spa was carried out directly using the secondary antibodies . Purified Staphylococcus aureus Protein A was obtained from Sigma-Aldrich ( St . Louis , MO ) . Strains were grown in TSB , with 10 μg/ml chloramphenicol for the complemented strain , at 37°C under aeration . When the OD600nm reached 1 , bacteria were harvested by centrifugation , ( 10 min; 3000 x g ) , washed in PBS , and resuspended in the same volume of PBS ( control ) or PBS containing 200 ng/ml lysostaphin followed by incubation at 37°C . Lysis was monitored by measuring the decline in OD600nm over time and indicated as a percentage of the initial OD ( measured OD600nm/initial OD600nm ) . Biofilm assays were performed by growing cells in PVC microtiter plates ( 200 μl per well ) in TSB with 0 . 75% glucose and 3 . 5% NaCl . After 24 h static growth at 37°C , adherent biomass was rinsed twice with PBS , air dried , stained with 0 . 1% crystal violet for 15 min , resuspended in ethanol-acetone ( 80:20 ) and quantified by measuring OD595nm , normalized to the OD600nm of each culture ( growth rates for the different strains were the same ) . Seven-week-old female RjOrl:SWISS mice ( Centre d’Elevage Roger Janvier , Le Genest-St . -Isle , France ) were inoculated intravenously with the S . aureus HG001 parental strain and the otherwise isogenic ΔspdC mutant . Groups of seven mice were infected with 5 . 107 cfu per mouse in 0 . 2 ml . Survival was monitored daily over 9 days post-infection and three independent experiments were carried out . Virulence of the complemented strain could not be carried out since we have shown that in vivo , in the absence of selection pressure , the complementation plasmid was lost over the assay period . Indeed , after nine days post-infection with the ST1375 complemented strain , animals were sacrificed and the kidneys removed and homogenized for determination of bacterial CFU ( total and chloramphenicol resistant ) per kidney , revealing that 97% of the bacteria had lost the pMK4Pprot-spdC complementation plasmid . Animal experiments were conducted at the Institut Pasteur in compliance with French legislation ( Decree N° 2001–464 05/29/01 ) and European Union guidelines on handling of laboratory animals: ( http://ec . europa . eu/environment/chemicals/lab_animals/index_en . htm ) . Animals were sacrificed by increasing carbon dioxide concentrations . Protocols were approved by the Institut Pasteur ethics committee ( Authorization N° 2013–0032 ) . The complete RNA-Seq dataset was deposited in the EMBL European Nucleotide Archive ( accession number PRJEB11849 ) and is accessible at the following URL: http://www . ebi . ac . uk/ena/data/view/PRJEB11849 | Staphylococcus aureus is a major human pathogen , and has become a significant worldwide health concern due to the rapid emergence of antibiotic resistant strains . Like most bacteria , S . aureus adapts to its environment by adjusting its genetic expression through sensing and regulatory systems . We show here that the SpdC membrane protein is a novel virulence factor of S . aureus , controlling biofilm formation and pathogenesis . We show that SpdC interacts with the WalK histidine kinase to inhibit activity of the WalKR two-component system . SpdC also interacts with nine other histidine kinases of S . aureus , suggesting it acts as a pleiotropic global regulator . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"cell",
"walls",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"gene",
"regulation",
"pathogens",
"microbiology",
"organic",
"compounds",
"staphylococcus",
"aureus",
"plasmid",
"construction",
"mutation",
... | 2018 | SpdC, a novel virulence factor, controls histidine kinase activity in Staphylococcus aureus |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.